Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006

FREE INQUIRY IN CREATIVE SOCIOLOGY

Volume34 Number I, May 2006 SN 0736-9182

Cover design: Hobart Jackson, University of Kansas School ofArchitecture

AUTHOR TABLE OF CONTENTS PAGE

Ira Sommers & The Health and Social Consequences of Methamphetamine 003 Deborah Baskin Use Among Young Adults

Marcela Raffaelli & Reducing Women's Risk of Heterosexual Transmission 015 Jill R. Brown of HIV in the U.S.

Michael Duke, Wei Patterns of Intimate Partner Violence Among Drug 029 Teng, Scott Clair, Using Women Hassan Saleheen, Pamela Choice, & Merrill Singer

Merrill Singer A Dose of Drugs, A Touch of Violence, A Case of AIDS, 039 Part 2: Further Conceptualizing the SAVA Syndemic

Byron L. Zamboanga & Applying Aspects of Problem Behavior Theory to Latino 055 Gustavo Carlo Youth: Theoretical, Methodological, and Sociocultural Considerations

Sanna J. Thompson & Runaway Youth Admitted to Juvenile Detention: Factors 069 Liliane Cambraia Associated with Cigarette, Alcohol, and Marijuana Use Windsor

Richard C. Cervantes, Key Risk and Protective Factors Among Multi-Ethnic, 077 Ann Del Vechio, Jose Elementary Aged Children: Findings from New Mexico's Esquibel, & Tony Rey Behavioral Health Services Prevention Bureau

Nickalos A. Rocha, Trauma Registries as a Potential Source for Family 087 Alberto G Mata, Jr., Violence and Other Cases of Intimate Partner Violence Alan H. Tyroch, Susan for Border Communities: Indicator Data Trends From Mclean, & Lois Blough 2000-2002

Subscriber and Author Manuscript Form 103 2 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

MISSION: To give an efficient outlet for informative, innovative articles legible to lay readers.

MANUSCRIPTS: Send three paper copies and one copy on a 3 1/2" computer disk, IBM compatible, saved in a rich text file format, use ASR guidelines. Manuscript fees: current subscribers $25 (US $); authors who are not current subscribers $40 (US $) which includes a current subscription.

WORLD REFEREED AND WORLD DISTRIBUTED. Abstracted in SOCIOLOGICAL ABSTRACTS. Over 150 major academic library subscribers, including 15 libraries in Europe, Asia, Africa and the Americas.

SUBSCRIPTION RATES: Individuals: United States $30. Foreign Surface Mail $35 (US$). Institutions: United States $40, Foreign Surface Mail $45 (US$). All subscriptions must be prepaid and are by calendar year only.

ADDRESS: Editor, FICS PHONE: 405-744-6126 Oklahoma State University Department of Sociology 006CLB Stillwater, OK 74078-4062

Q uESTIONs/CoMMENTS: [email protected]

GENERAL EDITOR: John R. Cross, Oklahoma State University SPECIAL EDITION EDITOR: Albert G. Mata, Jr. , University of Oklahoma ARABIC EDITOR: Arabic mss; lbtihaj Arafat, City College, New York EUROPEAN EDITOR: Charles Kaplan, Johann Wolfgang Goethe Universitaet, Germany ASSOCIATE EDITORS: Bruce L. Berg, California State University-Long Beach; Bonnie Berry, Social Problems Research Group; Don Drennon-Gala, Drennon-Gala & Associates; Beth Hartung Freimuth, California State University, Fresno; Cecil Greek, Florida State University; Thomas D. Hall, DePauw University; Charles Harper, Creighton University; Jeffrey R. McDade, Graceland University; Ralph O' Sullivan; Carl Pope, University of Wisconsin; William A. Reese, II , Augusta State University; Beverly Stiles, Midwestern State University; Donald Yates, Albany State University; and Marjorie Zatz, Arizona State University. TYPESETTER: Deborah Sweet, Oklahoma State University COVER DESIGN: Hobart Jackson, University of Kansas School of Architecture

BOARD OF GOVERNORS: Patricia Bell, Oklahoma State University; Alvin Turner, East Central University; Susan Chase, University of Tulsa; Marvin Cooke, Tulsa Community Coll ege; Kay Decker, Northwestern OSU; David Ford, University of Central Oklahoma; Lawrence Grear, Langston University; Philip Holley, Southwestern OSU; Jody Horn, Oklahoma City University; Tina Winn, University of Science & Arts of Oklahoma; Chuck Ziehr, Northeastern State University; Ed Mauzey, Southeastern OSU; Gordon Crews, Cameron University; Craig St. John, University of Oklahoma; Abe Marrero, Rogers State University.

OKLAHOMA SOCIOLOGY ASSOCIATION 2006 OFFICERS: Donna Barlow, Southwestern OSU President; Robert Bausch, Cameron University, President Elect; Monica Varner, Rogers State University, Secretary-Treasurer.

PUBLISHED: May and November by the Oklahoma Sociology Association and the Consortium of University Sociology Departments and Programs in the State of Oklahoma.

© 2006 Oklahoma State University Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 3 THE HEALTH AND SOCIAL CONSEQUENCES OF METHAMPHETAMINE USE AMONG YOUNG ADULTS

Ira Sommers, California State University, Los Angeles Deborah Baskin, California State University, Los Angeles

ABSTRACT

The current research analyzed the relationship between methamphetamine use and health and social o utcomes. Interviews were conducted with a sample of I 06 respondents. Virtually all of the respondents experienced negative consequences of methamphetamine use. The most serious, but least prevalent, methamphetamine-related health problems were seizures and convulsions. The most prevalent health effect was weight loss. A substantial number of respondents experienced severe psychological symptoms: depression, hallucinations, and paranoia. Of the I 06 respondents, 34.9 percent had committed violence while under the influence of methamphetamine. The data suggest that methamphetamine-based violence was more likely to occur within private domestic contexts, both family and acquaintance relationships. It is apparent from the findings that methamphetamine use heightens the risk for negative health, psychological, and social outcomes. Having said this, it is crucial to acknowledge that there was no evidence of a single, uniform career path that all chronic methamphetamine users follow. Furthermore, a significant number of sample members experienced limited or no serious social, psychological, or physical dysfunction as a result of their methamphetamine use. The use of a variety of drugs by adoles­ percent, respectively. Also during 2004, 6.2 cents and young adults continues to be an percent of high school seniors reported us­ important public health problem. Drug use ing methamphetamine within their lifetime. may have important implications for the fu­ During 2002, 11 .9 percent of college students ture health and well-being of many adoles­ and 14.8 percent of young adults (ages 19- cents and young adults as they negotiate the 28) reported using methamphetamine at transition to adulthood. Adolescents and least once during their lifetimes. young adults who use drugs may have espe­ Despite these reports indicating a greater cially high risks of developing mental or physi­ availability and consumption of methamphet­ cal problems that interfere with educational amine, little is known about the association and occupational pursuits, and which under­ of its use and health over time, particularly mine long-term life chances. during the formative stages of adolescence Although the use of certain types of drugs and young adulthood. The present research has decreased recently (National Institute on examined the inter-relationships among Drug Abuse [NIDA) & University of Michigan methamphetamine use, physical symptoms, 2005), there is evidence that methampheta­ and psychological and social well-being in a mine use is becoming more prevalent. Ac­ community sample of young adults living in cording to the National Survey on Drug Use Los Angeles. and Health (Substance Abuse and Mental Health Services Administration [SAMSHA) POTENTIAL CONSEQUENCES OF 2005), 4.9 percent (over 12 million people) METHAMPHETAMINE USE of the U.S. population reported trying meth­ Methamphetamine is a powerfully addic­ amphetamine at least once in their lifetime. tive stimulant that dramatically affects the The highest rate of methamphetamine use central nervous system. The drug is made was among the 26 to 34 age group, with 6.7 easily in clandestine laboratories with rela­ percent reporting lifetime methamphetamine tively inexpensive over-the-counter ingredi­ use during 2002. The second highest group ents. These factors combine to make meth­ was young adults (18-25), with 5.7 percent amphetamine a drug with high potential for reporting lifetime methamphetamine use widespread abuse. The effects of metham­ during 2002. According to the 2004 Monitor­ phetamine use can include addiction, psy­ ing the Future Study (NIDA & Univeristy of chotic behavior, and brain damage. Metham­ Michigan 2005), 6. 7 percent of high school phetamine is highly addictive and users try­ seniors reported using methamphetamine ing to abstain from use may suffer withdrawal within their lifetime. Lifetime use among 8th symptoms that include depression, anxiety, and 1Oth graders was 3.5 percent and 6.1 fatigue, paranoia, aggression, and intense 4 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology cravings for the drug (Katsumata, Sato, & experience with cocaine. Recently Perdue Kashiwade 1993). Chronic methamphet­ and colleagues (2003) studied the associa­ amine use can cause violent behavior, anxi­ tions between being high on methamphet­ ety, confusion, and insomnia. Users can also amine and engaging in risky sexual prac­ exhibit psychotic behavior including auditory tices. Their findings indicate that individuals hallucinations, mood di sturbances, delu­ high on methamphetamine were significantly sions, and paranoia, possibly resulting in more likely to have unprotected sex and to homicidal or suicidal thoughts (Albertson, have multiple sex partners than their coun­ Walby, & Derlet 1995). terparts not high on methamphetamine. According to the Drug Abuse Warning Network (DAWN) 2002 mortality data, areas RESEARCH METHODS with the highest number of methamphet­ The current research analyzed the health amine mentions in drug-related deaths were and social consequences of methamphet­ those in the Midwest and Western areas. amine use among a sample of young adults. Methamphetamine emergency department Interviews were conducted with a targeted (ED) mentions have fluctuated since 1995, sample of 106 respondents. This section when there were 15,933 mentions. Metham­ outlines sampling and data collection pro­ phetamine ED mentions declined to 10,44 7 cedures, as well as measures of variables during 1999. This number has since in­ used in the analysis. creased to 17,696 in 2002. Methamphetamine users in treatment Th e Sample have reported physical symptoms associ­ Location and Recruitment Methods ated with the use of methamphetamine in­ The research was based primarily on in cluding weight loss, tachycardia (abnormal depth, life-history interviews with 106 individ­ rapidity of heart action), tachypnea (abnor­ uals who used methamphetami ne for a mini­ mal rapidity of respiration) , hyperthermia (un­ mum of three months and who re sid ed in usually high fever) , insomnia, and muscular Los Angeles County. The respondents were tremors. The behavioral and psychiatric recruited from two social settings: 1) meth­ symptoms reported most often include vio­ amphetamine users participating in ADAPT, lent behavior, repetitive activity, memory loss, a drug treatment program for metha mphet­ paranoia, delusions of reference, auditory amine users and 2) methamphetamine us­ hallucinations, and confusion or fright. Em­ ers at liberty in the community and having pirical studies, however, concerning the little or no contact with treatment or criminal health and social consequences of metham­ justice institutions. phetamine use are sparse. The data collection process began with One significant finding common to the few the recruitment of a sample of methamphet­ ethnographic studies on methamphetamine amine users from a drug treatment program. use is its relationship to violent behavior. Arrangements for respondent recruitment Morgan's (1997) study of methamphetamine were made with the ADAPT program, a drug use in San Francisco, Honolulu and San Di­ treatment program for methamphetamine ego indicates a significant relationship be­ users in Los Angeles County. Meetings were tween methamphetamine use and violence held between the senior research staff and for both males and females. For example, the treatment program Director and program 53 percent and 44 percent of males and fe­ participants. The research study was ex­ males, respectively, in the Honolulu sample plained in detail and contact letters were left reported engaging in violent acts due to meth­ with the program participants. Potential re­ amphetamine use. Furthermore, a majority spondents were instructed to call for appoint­ of respondents across all sites reported ex­ ments, at which time they were screened for periencing major psychological problems. eligibility (i.e., used methamphetamine for Overall, 58 percent of the males and 52 per­ at least 3 months, age 18-26) and arrange­ cent of the females reported paranoia due to ments were made for the interview. Once the their methamphetamine use. initial pool of respondents were identified, Similarly, an ethnographic study in Arizona they were asked to nominate or refer "some­ (Castro 1997) su ggests that methamphet­ one like them" who also has been involved amine users burn out even faster and often in methamphetamine use. Thus, the initial develop higher levels of paranoia than they sample was comprised of treatment pro- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 5

Table 1: Sample Characteristics (N=1 06) Sex(%) Male 59.40 Female 40.60 Age (mean) 21.58 (median) 22.00 Race (%) white 30.20 black 7.50 hispanic 62.30 Education (mean years completed) 11 .88 School dropout (%) 17.00 Marital status at interview (%) married/living together 26.40 never married 69.80 other 3.80 Children have children (%) 34.00 number (mean) 2.10 Employment history (%) never worked 17.00 sales/cashier/foodworker 24.50 clerical 9.40 non-skilled 21 .70 skilled 10.40 semi professional/professional 17.00 Problems while in school Prevalence (%) Age at initiation (mean) fighting 72.60 11 .64 weapons possession 27.40 13.34 alcohol use 45.30 13.69 drug use 78.40 13. 86 Intact Family (%) 82. 10 Family problems (%) someone arrested 48.10 substance abuse 53.70 family mental health 11 .30 famiy violence while using drugs/alcohol 26.40 gram participants and "chain referrals" from sessing, on average, 25 months of work ex­ these treatment respondents. perience (see Table 1). The youngest re­ A broader community sample was recruit­ spondent was 18 years old and the oldest ed through advertising in local university 25; the median age was 22 years. (California State University, Los Angeles, Most of the respondents worked in a legiti­ University of Southern California) newspa­ mate job (83%). Approximately three in five pers. This tactic helped expand our sample respondents (66%) worked in unskilled and to unknown members of the population who semi-skilled occupations (e.g., clerical, sales have no contact with formal treatment or crimi­ and factory jobs). However, approximately nal justice institutions. Chain referral or 20% of the sample worked in semi-profes­ "snowball" sampling techniques also was sional and professional jobs (e.g., counse­ used with this sample. lor, teacher, accountant). The sample contains 55 respondents Table 2 shows self-reported lifetime prev­ (51 .9%) in drug treatment and 51 (48.1 %) alence of drug use, drug selling, and non­ active community methamphetamine users. violent and violent crimes. Respondents re­ T he majority of respondents were male ported that they were engaged in a wide (59.4%), Hispanic (62.3%), high school grad­ range of criminal and deviant activities. Nearly uates (83.0%), in their twenties (86.2%), pos- all said they were experienced drug users. 6 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Table 2: Crime, Drug Use and Drug Selling History (N=106) Prevalence (%) Age at Initiation Non-Violent Crimes Auto theft 42 .5 14.36 Shoplifting 68.9 12 .62 Forgery 8.5 19.67 Prostitution 0.9 19.00 Burglary 13.2 15.79 Violent Crimes Assault 36.7 15.71 Robbery 16.0 15.59 Weapons possession 54.3 15.34 Attempted murder 16.0 16.18 Murder 6.6 15.86 Drug Used Alcohol 100.0 13.59 Marijuana 96.2 13.95 Inhalants 28.3 14.87 Hallucinogens 55.2 15.74 PCP 29 .3 15.77 Methamphetamine 100.0 16.80 Depressants 17.9 16.05 Cocaine 76.2 16.92 Crack 50 .9 16.95 Heroin 2.8 20.00 Drug Sold Methamphetamine 60 .9 Cocaine 16.9 Crack 14.6 Marijuana 32.0

This is not surprising since the criterion for Interview Protocol inclusion in this study was methamphet­ The primary goal of this research was to amine use. Seventy-six percent used co­ capture thick descriptions of the relationship caine, 51 percent used crack, 5 percent used between methamphetamine use, health, hallucinogens, and 96 percent used mari­ and high-risk behaviors. Depth interviewing juana. Of the 106 people interviewed, 67.9 was used to record information about spe­ percent (N=72) had committed at least one cific events and allowed respondents to re­ violent crime. Sixteen percent reported in­ flect on those events. Structured , but open­ volvement in robbery, 16 percent reported in­ ended interview guides were used. The volvement in attempted murder, 6 percent in open-ended technique created a context in murder, 37 percent had committed assault, which respondents were able to speak freely and 54 percent had carried weapons. How­ and in their own words. Furthermore, it facili­ ever, only twenty-three percent (N=24) of the tated the pursuit of issues that were raised sample were ever arrested for a violent by the respondents during th e interview but crime. Eighty-three percent (N=88) of the re­ are not recognized beforehand by the re­ spondents were involved in nonviolent crime. searchers. Table 2 also shows lifetime participation The interviews included items on person­ rates in drug selling by drug type. Sixty-one al demographics, family background, de­ percent had sold methamphetamine. Thirty­ tailed life history information about prior in­ two percent of the respondents had sold volvement in drug use, questions about life­ marijuana and 15 percent and 17 percent style, health and psychological problems, sold crack and cocaine, respectively. The and items on violence toward others. Partici­ mean age of initiation into dealing was be­ pants who reported health, psychological, fore 17 years of age. and social problems were asked to provide a description of the problem event/act (the Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 7

Table 3: Characteristics of office in a university. In order to convey the Methamphetamine Use neutrality and anonymity of the study, we N Percent avoided offices of either criminal justice agen­ Frequency of Use cies or clinical settings. The participants Weekends 20 18.9 were given a travel allowance ($5), regard­ 3-6 days/week 13 12.2 less of the length or duration of their trip. A Daily 73 68.9 stipend of $25 for the interview was paid at Weekly Cost the conclusion of the interview, although it Range $0-800 was not contingent on completion of the inter­ Mean $136 view. Median $60 Prmiary Method of Use Getting Into Methamphetamine Snort 82 77.4 Although no one process of initiation was Smoke 20 18.9 uniformly experienced by all our sample Inject 4 3.7 members, some themes were common to Binge most accounts. First, the vast majority of the Never 3 2.8 respondents were seasoned drug users, first 2-5 days 78 73.5 alcohol and marijuana and later cocaine. 6-10 days 25 23.7 "Getting high" was part of their life experi­ ences, so it was not a giant step for them to range 2-21 days indulge their curiosity about a new high. Sec­ mean 4.18 days ond, a high percentage of individuals sought median 3.00 days out methamphetamine on their own the first time. Although initiation was often self-moti­ most recent one for multiple episodes) and vated, it nearly always was part of some so­ its consequences. A narrative account of how cial situation with friends and/or acquaintan­ these drugs and drug states were related to ces. It was a rare case in which one's first the event also were obtained. use of methamphetamine was occasioned Sample members were asked if they had by a stranger. Typically, methamphetamine experienced any of 13 drug-related problems was first tried at the respondent's or friend's while using methamphetamine. The 13 prob­ house; a safe, private, comfortable location. lems covered a wide range of intrapsychic, An overwhelming number of respondents personal and interpersonal difficulties. Fac­ increased their use of methamphetamine tor analysis with varimax rotation and a Kai­ within days of their initial experience. This ser criterion was used to create indices of pattern of rapid escalation not only can be drug problems. For example, intrapsychic attributed to the physical and psychological problems related to methamphetamine use effects of the drug but also with the general included depression, paranoia, hallucina­ availability of methamphetamine. tions, anxiety/irritability, and sleeplessness. A key factor often cited as contributing to A second factor involved difficulties in social escalating use was the seductive nature of functioning and in fulfilling role obligations, the drug itself. The word often used to de­ including trouble at school, trouble at work, scribe methamphetamine by the respond­ family problems, and financial problems. ents was "seductive." Most stated that meth­ In addition, respondents were asked to amphetamine effects offered not only in­ describe the relationship, if any, between the creased energy, but a sense of well-being problems and methamphetamine use, in­ and a feeling of mastery and power that was cluding amounts of specific substances in­ so reinforcing it often led them to use more gested prior to the time of the incident by the frequently than they expected. Casual week­ respondent, the state of intoxication or other end use often led to greater use during the drug states (e.g., 'crashing') manifested by week. Even those who initially limited their the respondent prior to the reported behav­ use to specific situations- parties, sexual ior. activities, work- gradually found themselves using methamphetamine in a variety of activ­ lnteNiew Procedures ities. People who went on periodic binges Interviews were conducted in a neutral lo­ sometimes found their binges stretching cation such as a library, park, or a private over longer periods at higher dosages. All 8 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology this helps to explain why many users esca­ long car trips lated their use over time. It is important to -To socialize with friends note, however, that such escalation was not inevitable; approximately 20 percent of the Many of the sample members felt ener­ respondents maintained stable use patterns gized by methamphetamine and reported a over many years without increasing doses heightened sense of accomplishment while (Table 3). using the drug. Joe, a 25 year old mechanic: The binge is the continuation of the meth­ amphetamine high. The user maintains the Methamphetamine kept me awake for a long high by using more methamphetamine. Af­ time . It allowed me to do things when I didn 't ter each use of the drug, a smaller euphoric want to . Like other drugs would relax me , rush than the initial rush is experienced un­ give me a weird feeling , make me lazy or til , finally, there is no rush and no high. Dur­ sleep a lot. Meth gave me a feeling like my ing the binge, the user becomes hyperactive body was constantly charged no matter both mentally and physically. Matt, a 25 year what I did for a long period of time . old construction worker described a typical binge episode: Val , a 30 year old employment agency super­ visor: I would go on meth spree for about a week and couldn 't control my usage . It was like I Methamphetamine made me fell like I was had to constantly have to be snorting or finally capable of doing a lot of things all in smoking the meth . In my mind it tells me to do one day. Especially since I have to deal with some more to function a lot better and faster. a lot of people and paperwork. I believe that meth is one of the most psy­ chologically addictive drugs around . When­ Terri , a 31 year old child care supervisor: ever I get tired or wish I had more energy, I always think how nice it would be to have It keeps me going . Lets me feel like I'm al­ some speed . In that respect, I am addicted , ways energized . It allows me to finish all because it is definitely a part of my thought my chores after coming home from work pattern . Meth is very seductive. It makes and able to play with my kids at all times . you feel energized and powerful. Once you take it a few times, you will continue to think For Jill, a 20 year old sales clerk, meth­ about it after you stop . amphetamine use was simply a method to loose weight. Within three months, she was Many users emphasized its energizing ef­ using $1 00 a day. fects, some its euphoria, and others its sex­ ual effects. Unlike alcohol, marijuana, and Meth made me loose my appetite. I felt I could hallucinogens, methamphetamine did not quit as soon as I got down to the weight diminish a person's basic competence in that I was satisfied with . But then I couldn 't daily life-functions unless and until it was stop. I had to have it daily. used in excess. The sample members spoke of methamphetamine as a general pick-me­ Still for others, methamphetamine use up in a variety of circumstances: cleaning was a way to achieve a satisfying high. one's house, studying, keeping up with the Martha, a 23 year old office worker: kids, enhancing a marijuana high, etc. The repsondents reported at least seven uses At first, I liked the way methamphetamine for methamphetamine. tasted with marijuana . Then I began smok­ ing meth by itself, with my boyfriend and I -To party in varying public and private set- liked the way it got me energized daily. I tings liked the way it energized during sex . Also , -To enhance sex I was able to do plenty of errands. -To work -To diet Bob, a 29 year old architect: -To get high -To sustain themselves for laborious tasks, I smoke coca ine but I stopped liking it. So I such as studying , writing , child care , went to meth in order to get a good high . I Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 9

Table 4: Methamphetamine-Related of this level of use, it is not surprising that the Problems (N=106) respondents reported a wide range of side Health Problems Percent effects from methamphetamine use. Their Seizures/convulsions 3.8 experiences are summarized in Table 4. Dehydration 8.5 Virtually all the repsondents experienced Sleep 93.4 negative consequences of methampheta­ Weight 55.7 mine use. The most serious, but least preva­ Depression 36.8 lent, methamphetamine-related health prob­ Paranoia 62.3 lem was seizures and convulsions. Four re­ Hallucinations 37.7 spondents reported that they had suffered Irritability 79.3 from some form of convulsion or seizure as Social Problems a consequence of methamphetamine use. Family 49.1 Nine respondents reported fairly serious School 15.1 episodes of dehydration. The most preva­ Work 7.6 lent health effect was weight loss. Fifty-six Financial 23.6 percent of the respondents reported weight Psychological Problem Index loss. (Sleep, depression , paranoia, hallucinations, With long-term use the psychological ef­ irritability) fects of methamphetamine can be severe. # of Problems Psychological symptoms specific to meth­ 1 11.8 amphetamine can include suspicion, anxi­ 2 19.4 ety and hallucinations. Much more acute 3 23.7 symptoms can be changes in lifestyle and 4 24.7 eventually in personality. Insomnia was the 5 20.4 most frequent mental health problem re­ Social Problem Index ported by the sample members. This finding (Family, school, work, financial) is not surprising since methamphetamine # of Problems is a central nervous system stimulant that is 0 19.4 valued precisely for its energizing effects. Ir­ 1 31.2 ritability was reported by 70 percent of the 2 35.5 sample. Irritability was described as feeling 3 9.7 "moody," having a "short fuse," and being ar­ 4 4.3 gumentative. A substantial number of respondents ex­ liked it because a small amount would get perienced severe psychological symptoms: me high. What caught me about meth is the depression, hallucinations, and paranoia. feeling of invulnerability. I got from meth The most frequently mentioned form of para­ what most cocaine users get from coke. noia was fear of others; feeling that people The feeling of being on top of the world. wished harm to or threatened the respond­ ent. This type of psychotic symptom has par­ Consequences of Methamphetamine Use ticular relevance to violent behavior. Previ­ Perhaps the most important theme in ous research suggests that when a person their description of the methamphetamine fears personal harm or feels threatened by high was the "intensity" of the euphoria. The others, interpersonal violence becomes positive characteristics of methamphet­ more likely (Link & Stueve 1998). In addition , amine use - the euphoria, energy, empow­ violence is more likely when internal con­ erment-, however, were often overshad­ trols that might otherwise block the expres­ owed by the negative effects of long-term use. sion of violence break down. Sample members were asked to report on Approximately 38 percent of the respond­ side effects during their heaviest period of ents reported experiencing some form of hal­ methamphetamine use. Respondents var­ lucination. Hallucinations usually took the ied in their length of use (the average length form of hearing voices familiar to the re­ of use was 3.8 years). Ninety-seven percent spondent that make insulting remarks or of the sample reported that they engaged in command the respondent to do certain binge behavior. Approximately four days was things. Depressed users often had halluci­ the average reported binge duration. In light nations with themes of guilt and personal 10 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Table 5: Mean Scores of Ma les and Females on Frequency of Methamphetamine Use and Drug Problem Indices Psychological Problem Social Problem Index Index Frequency of Use Male Female Male Female Weekends 2.00 2.50 0.93 0.00 3-6 days I week 2.96 3.80 1.00 1.00 Daily 3.29 3.90 1.91 1. 52 inadequacy, such a hearing voices berating of methamphetamine. Males comprised two­ them for their shortcomings. thirds of the 37 respondents (N=24). Of the Despite the high level of addiction among total sample, 38 percent of males and 30 sample members, the social effects of meth­ percent of females committed methamphet­ amphetamine use were surprising small. amine-related violence , respectively. Seven­ Nineteen percent of the sample reported no teen of the 37 respondents who committed social effects and approximately 31 percent methamphetamine-related violence (45.9%) reported experiencing only one social prob­ reported that they had never committed a vio­ lem related to methamphetamine use. Meth­ lent crime prior to the methamphetamine­ amphetamine use seemed to have the least based events. However, 12 (70.5%) of these impact on school, work and finances. Meth­ respondents had committed aggressive acts amphetamine-related problems with while under the influence of other drugs. Over­ spouses, lovers, or friends were more ap­ all , the 37 respondents reported 54 sepa­ parent. One in two respondents reported that rate violent events while using methamphet­ methamphetamine use had negative effects amine. Of these 54 events, 33 (61.1 %) acts on their interpersonal relationships. of violence involved domestic relationships, Overall, the sample members that re­ 9 (16.7%) of the violent events were drug re­ ported the greatest number of psychological lated, 7 (1 3%) were gang related, and 5 and social problems are the respondents (9.3%) involved random acts of violence (e.g. , that reported the greatest methamphetamine road rage, stranger assault). use (see Table 5) . Regardless of sex, the It has been suggested that in contrast to mean scores for psychological and social crack, methamphetamine produces a longer problems increase as the level of metham­ lasting high . As a result, methamphetamine phetamine use increases. users are able to remain away from the mar­ While the psychological effects of meth­ ket environment longer as they are not con­ amphetamine use are quite similar to those stantly "chasing the pipe". Consequently, of crack cocaine, the social consequences methamphetamine users are more likely to seem to be quite different. Regardless of how return to work, school, or home settings while crack use was initiated, the vast majority of high. Thus, in contrast to their crack using crack users ended up in the same role-as counterparts, they are less likely to be en­ street addicts. The crack user became in­ trenched in street networks yet more likely to creasingly immersed in their addiction at the engage in violent behavior at home, in the exclusion of almost all else. Economic prob­ workplace, or within other more mainstream lems and the culture of addiction justified social settings. Study data suggest that meth­ the use of virtually any means to get money amphetamine-based violence may indeed in order to support crack habits. For many, be more likely to occur within private domes­ the problem of maintaining an addiction took tic contexts, both family and acquaintance precedence over all other interests and over relationships. Thirty-eight (70.4%) of the 54 participation in other social worlds. Crack violent events occurred in private homes, users often became enmeshed in deviance seven (14.3%) at parties, one (1.9%) at work, and further alienated, both socially and psy­ and eight (14,8%) in public settings (e.g., chologically, from conventional life . parks, street, roadways).

Prevalence of Methamphetamine-related The Social Context of Methamphetamine­ Violence Related Violence Of the 106 respondents, 37 (34.9%) had Methamphetamine affects were evident in committed violence while under the influence decision making, cognition, intensified emo- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 11 tional states, exaggerated affect, and dimin­ friends. My mom said "you're not going out ished capacity for self-regulation. For ex­ until you run some errands for me. You ample, respondents indicated that language haven't done anything, so you will do this when intoxicated was more provocative, and for me." I just snapped and called my mom a language often "amped up" otherwise minor fuckin' asshole and pushed her into my disputes into violent encounters. dresser. I never did anything like this be­ Phillip, a 30 year old restaurant worker, fo re. I hadn't slept for a couple of days, used methamphetamine for two years. doing meth continuously.

I was tired and exhausted from working, A fairly common effect of methamphet­ so I went out with a few co-workers for amine was paranoia. Paranoia contributed some drinks. I did a few lines of meth and to hostile attributions that created an air of had a few beers. I got home late. I got un­ danger and threat, leading to defensive or dressed and my wife asked me where I pre-emptive violence. Veronica, a 20 year old was and don't you know how to call. I told receptionist, high on three lines of metham­ her to shut up and I will do as I please. We phetamine and three beers was relaxing began arguing back and forth and she called outside her house: me a drunk and druggie. I lost control. I slapped her and kicked her in the stomach. I was relaxing at home when a girl passed I threw her down, then I left the room. by. I thought she gave me a dirty look. She continued looking at me as she passed by. I Methamphetamine use often increased can 't stand it when individuals give me dirty the stakes in everyday interactions, trans­ looks when I don't know them. I shouted forming them from non-challenging verbal what are you looking at you dumb bitch and interactions into the types of "character con­ socked her in her face. I pushed her around tests" whose resolution often involved vio­ and she went runn ing away. lence. Methamphetamine exaggerated the sense of outrage over perceived transgres­ Similarly, Bernard, a 28 year old lab techni­ sions of personal codes (respect, space, cian, imagined that people were evil. verbal challenges), resulting in violence to exert social control or retribution. I was on vacation in Rosarito, Mexico. I be­ In the following account, the social iden­ gan doing meth on Friday night and now it tity of Larry, a 21 year old real estate assis­ was Sunday afternoon. I also drank a few tant, was challenged by his girlfriend. beers. I was paranoid the whole day. I thought others were up to something. I was Me and my girlfriend were coming home watching everybody, th inking and looking from the doctor's from getting her pregnancy to see if anyone was doing something bad . test. We were talking about how she was I had evil thoughts. I was thinking evil and pregnant and how she needed to stay fo­ thought others were doing something cused in school and with her health. Then wrong . So I got bottles and started break­ she said and "your ass better be able to be ing them over people's heads. a responsible father and keep your ass out of trouble." I simply slapped her in her face. Several sample members reported that I couldn't deal with the insult. their decision making within violent events was comprised. Perhaps the most common Some people simply made bad decisions language respondents used to describe their while high, leading to fights that might have behavior was "loss of control." The respond­ been avoided in other circumstances. Martha, ents spoke in terms of "being out of control, " a 23 year old student: "blowing up," or having an "outburst of rage." Alicia, a 24 year old clerical assistant, used I was inside my room, getting ready to take methamphetamine for four years. She de­ a shower. I was making plans for the even­ scribed a minor dispute with her boyfriend ing, when my si ster said that "I couldn't that erupted into violence. Both had snorted leave. Mom said you can't go anywhere." methamphetamine for two hours. My mom walked in and said "where do you think you 're going?" I said out with some Me and boyfriend were having some finan- 12 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

cial problems and we were discussing the I was relaxing at my house with a few money I spent. We were talking about the friends . We were drinking some beer and money that I spent on clothes on a weekly doing a few lines of meth . I was chopping basis. My boyfriend said it was unneces­ up some lines for me and my friends. We sary. I yelled at him and he got upset and were just getting ready to do the lines when pushed me. Then I punched him and he hit my friend noticed that mine was the biggest me and threw me to the ground . We contin­ of all. He got upset because they got little ued to fight for a few minutes. lines and I got a fatter one . He said that it wasn't fair. We argued and then started While cognitive impairment was evident punching and fighting . Things got out of for many, others noted that their decisions to control. use violence reflected the normative process of gang conflicts. Gangs provide a social Joey, a 12th grade student, described a simi­ context in which the potential for violence re­ lar conflict over methamphetamine use. sults from any number of concerns includ­ ing: territorial battles, initiation and detach­ It was during lunch at school. I was relax­ ment rituals, attaining status and social iden­ ing with my girlfriend and friends at the table. tity, material gain, expressions of grievances, We were talking and I wanted to do some retribution, reinforcement of collective iden­ more meth . I was asking my friend to kick tity, etc. me down with a little bit of his stuff and I For Javier, a 26 year old mechanic, drug would get him back when I got my sack. My use and violence was part of the normative friend refused to give me any. I got upset gang process: and began to lose my patience. I told him he should give me some. He said , "no and what Me and my homeboys were kicking back at are you going to do if I don 't. " I said , "I will a park doing some methamphetamine, smok­ kick your ass ." He said , "well then come ing weed , and drinking beer. We were all on ." So I punched him first and we began shit-faced and bullshiting . And then we de­ fighting . cided to go on a drive-by that night. We were talking about a homeboy that had just Finally, in only two cases, violence was recently passed away and how we should used as means to obtain money for meth­ go and get those guys back. We went to amphetamine. Mario, a 25 year old sales our houses, got our weapons and cars, clerk, described such an incident: then met up at the park. We had about three to four full car loads ... We shot at about 10 I was on my way to a friends house, walk­ guys and hit a couple. ing down the street, when I noticed some guy that looked like he had money. I was In this case, aggression was perceived as a thinking about how I could but more meth . form of retaliation for a previous wrongdo­ So I decided to steal this guy's wallet. I put ing. By retaliating, the gang members "saved a gun to his to face and told him to give me face" were able to nullify the image of being everything or else I would shoot him . weak and ineffectual. The relationship between drug use and Allen, a 26 year old furniture mover, had simi­ violence has been observed in literally hun­ lar motivations: dreds of empirical studies. People who use and sell drugs are more likely to engage in I was coming down from meth and had no violence than non-drug involved individuals. money. Had no sleep, no meth , I needed Accordingly, although individuals generally more. So I went to a local drug spot with have low base rates of violence (Sommers two friends. I pretended I was going to buy & Baskin 1992), their entry into drug use or some meth. After discussing the price, the selling increases the risks of violence. dealer gave me the drugs and then I pulled In some cases, interpersonal violence out my 9mm . I hit him with it in the face and occurred within the system of drug use. jumped in my friends car. Frank, a 25 year old cook, talked about one incident: The above accounts indicate that meth­ amphetamine use provided several mecha- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 13

nisms for motivating violence. Cognitive ef­ are mediated by users' norms, values, prac­ fects included: inhibition of cues that nor­ tices, and circumstances. No matter how se­ mally control behavior, increased arousabil­ ductive methamphetamine is, it is always ity, interference with communication and inter­ used in social contexts that shape how it is personal interactions, and intensification of used and what its effects are taken to mean emotions. The findings suggest that a meth­ by users. amphetamine-related violent event results The variation in intoxicated behaviors with­ from the interaction of the individual, the sub­ in social contexts suggests that the context stance, and the situation. itself exerts a powerful influence on the vio­ In the present study, methamphetamine lence outcomes of methamphetamine situ­ was more often present in violent events that ations. This study has shown that the impor­ occurred in peoples' homes and between tance of social context for methamphet­ known individuals. Similar to previous re­ amine-related violence lies in the mediating search on assaultive behavior, the picture processes that shape behaviors as well as that emerges from these analyses is not one in the specific interactions leading to violence of blind irrational behavior. Rather, the ration­ between offenders and victims. Violent be­ al character of these events is evidenced in havior resulted from a complex interaction a person's image maintenance in the face among a variety of social, personality, envi­ of challenge. It is clear from the accounts ronmental, and clinical factors whose rela­ that interactions between victim and offender tive importance varied across situations and played a fundamental role in violent inci­ time. dents. To a large extent, these sample mem­ Furthermore, research on intoxication of­ bers were not roaming willy-nilly through the ten has overlooked the distinction between streets engaging in "unprovoked" violence. acute and chronic intoxication and their differ­ ential effects on affective or personality states. CONCLUSIONS The most significant pharmacologic deter­ Study findings suggest that methamphet­ minants of the methamphetamine-violence amine use has serious negative conse­ link are the dose and the chronicity of expo­ quences for health and psychological func­ sure to the drug. At acute low doses, meth­ tioning. It is also apparent from our findings amphetamine produced cognitive and mood that methamphetamine use heightens the alterations but tended not to increase offen­ risk for violence. Everyone we interviewed sive-aggressive behavior. With increasing agreed that methamphetamine has clear dose and long-term use, methamphetamine abuse and violence potential. Almost all of users tended to display psychological and our respondents knew people who had gone physical deterioration, as well as changes "too far" with methamphetamine even if they in their social behavior. Correspondingly, themselves had not. Having said this, it is chronic use tended to reduce impulse con­ crucial to reiterate that we could find no evi­ trol and produce exaggerated defensive pos­ dence of a single, uniform career path that tures that deviated from a respondent's ex­ all chronic methamphetamine users follow. pected behavioral repertoire. It is important Progression from controlled use to addic­ to note that sample members also reported tion is not inexorable. Furthermore, a signifi­ that high acute methamphetamine doses cant number of sample members experi­ and binging often induced paranoia that was enced limited or no serious social, psycho­ directly linked to aggressive and violent be­ logical, or physical dysfunction as a result of havior. their methamphetamine use. Most germane to this study, we found that violence is not an REFERENCES inevitable outcome of even chronic metham­ Albertson T, W Walby & R Derlet. 1995. Stimulant­ phetamine use. induced pulmonary toxicity. Chest 108 11 40-9. Our findings suggest clearly that pharma­ Castro F. 1997. Methamphetamine trends in five cology is not destiny. As Fagan (1993) and western states and Hawaii. National Trends Zinberg (1984) have shown, the interaction in Drug Abuse. Office of National Drug Control Policy Pulse Check Washington, DC: ONDCP. between the pharmacological properties of Dru g Abuse Warn ing Network (DAWN). 2002. a substance and the physiological character­ Emergency department visits involving meth­ istics of a user accounts for only part of a amphetamine. SAMHSA. drug's effects. Drug effects and outcomes Fagan J. 1993. Set and setting revisited: influ- 14 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

ences of alcohol and illicit drugs on the social Perdue T, H Hagan , H Thiede & L Valleroy. 2003. context of violent events. In Alcohol and Inter­ Depression and HIV risk behavior among Se­ personal Violence: Fostering Multidisciplinary attle-area injection drug users and young men Perspectives. S. Martin , ed . Rockville, MD: US who have sex with men. AIDS Education Pre­ Department of Health and Human Services. vention 15 1 81-92. Katsumata S, K Sato, & H Kashiwade. 1993. Sud­ Sommers I & D Baskin. 1992. Sex, race , age and den death due presumably to internal use of violent offending. Violence & Victims 7 191- methamphetamine. Forensic Sci 62 209-15. 202 . Link B & A Stueve. 1998. Psychotic symptoms and Substance Abuse and Mental Health Services Ad­ the violenUillegal behavior of mental patients ministration. 2005. Results From the 2004 Na­ compared to community controls. In Violence tional Survey on Drug Use and Health: Na­ and Mental Disorder. J Monahan & H Steadman, tional Findings. Rockville , MD: Office of Ap­ eds. Chicago: U Chicago Press. plied Studies. Morgan S. 1997. Methamphetamine trends in five Zinberg N. 1984. Drug, Set and Setting: The Basis western states and Hawaii. National Trends for Controlled Intoxicant Use. New Haven: Yale in Drug Abuse. Office of National Drug Control U Press. Policy Pulse Check Washington , DC : ONDCP. National Institute on Drug Abuse & University of Michigan. 2005. Monitoring the Future National Survey Results on Drug Use, 1975-2002, Vol­ ume II: College Students & Young Adults Ages 19-40. NIDA. Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 15 REDUCING WOMEN'S RISK OF HETEROSEXUAL TRANSMISSION OF HIV IN THE U.S.

Marcela Raffaelli and Jill R. Brown, University of Nebraska- Lincoln

Assuming no major scientific breakthrough women, while many millions more are at in tools that protect against the sexual trans­ risk for infection at this time. Women who mission of HIV in the near future, to prevent engage in heterosexual intercourse are the HIV infection among women in the United group most rapidly becoming infected with States, it will be necessary not only to pro­ human immunodeficiency virus (HIV) in the mote changes that support the active role United States. (O'Leary & Jemmott 1995 ix) of women in prevention but also to promote change in the attitudes and behaviors of Since the beginning of the HIV/AIDS epi­ men. (Amaro 1995 445) demic, there have been 886,575 AIDS diag­ noses and half a million deaths due to AIDS Well into the third decade of the AIDS pan­ in the U.S. (CDC 2002). Despite treatment demic, it has become clear that stemming advances that contribute to increased sur­ heterosexual transmission of HIV, the virus vival among those infected with HIV and re­ that causes AIDS, will require innovative ways duce the risk of maternal-child transmission of thinking about the myriad factors that influ­ of the virus, HIV/AIDS represents a contin­ ence sexual behavior. We focus on the grow­ ued threat to the health of the nation. This is ing HIV/AIDS epidemic among heterosexual reflected in the fact that in 2001 , AIDS was women in the U.S. Building on prior work in the 6th leading cause of death among both this area, we begin at the point where many male and female 15-to-44-year-olds in the considerations of this topic end: the need to United States (CDC 2004). shift the focus of prevention efforts from From its initial emergence, AIDS has been women to their male partners. This shift does unequally distributed among different seg­ not reflect a belief that women are passive ments of the U.S. population. According to participants in sexual encounters, but rather statistics compiled by the Centers for Dis­ the recognition that many women who be­ ease Control, the majority of the 384,906 come infected with HIV do not engage in any people living with AIDS at the end of 2002 "risk behaviors" other than unprotected sex­ were adolescents or adults, with 43 percent ual intercourse with a primary partner. Thus, of cases occurring in the 35-44 age group interventions must focus not only on and 30 percent in the 45-54 age group (CDC women's sexual behavior but also on reduc­ 2002; unless otherwise noted all statistics ing the male partner's risk of HIV exposure in this and the next paragraph are from this and fostering sexual behavior change at the report). In terms of gender, over three quar­ level of the couple. We consider the implica­ ters (78.3%) of adolescents and adults liv­ tions of moving from prevention strategies ing with AIDS are male; 21 .7 percent are fe­ that make women the "gate-keepers" of sex­ male. The majority of AIDS cases among ual risk reduction efforts to approaches that men are attributed to male-male sexual con­ take into account the gendered and interac­ tact (58%), injecting drug use (23%), or both tive nature of sexuality. We first describe cur­ (8%); just 10 percent of AIDS cases among rent and future dimensions of the HIV/AIDS men are attributed to heterosexual exposure, epidemic in the U.S. Following this, we re­ and 2 percent to other or unknown risks. view prevention efforts aimed at sexual risk Among women, the majority of AIDS cases reduction among heterosexual women. We are attributed to heterosexual exposure then discuss programs that focus on hetero­ (61 %); 36 percent are attributed to injecting sexual men. Finally, we consider future chal­ drug use and 3 percent to other or unknown lenges for the prevention of HIV/AIDS among risks. Members of ethnic minority groups are heterosexual women.1 over-represented in AIDS statistics: in 2002, estimated rates of AIDS were highest among PAST, CURRENT, AND FUTURE DIMENSIONS African American men (108.4 per 100 ,000) OF THE HIVIAIDS EPIDEMIC IN THE U.S. and women (48.6), followed by Hispanic The acquired immunodeficiency syndrome men (39.7) and women (11 .3), American In­ (AIDS) is devastating the lives of millions of dian/Alaska Native men (16.9) and women 16 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

(5 .8) , White men (12.3) and women (2 .1) and case fo r HIV/AIDS. Although highly active anti­ Asian/Pacific Islander men (8 .6) and women retroviral therapy (HAART) represents a sig­ (1.5) . nificant treatment advance, an effective vac­ Because of the long incubation period , cine against HIV has not yet been developed reporting delays, and changes in the case (UNFPA 2001 ) and a cure for AIDS remains definition for AIDS introduced in 1993, these elusive. As a result, prevention remains the statistics do not accurately represent the fu­ focus of HIV/AIDS intervention efforts (Auer­ ture of the epidemic, particularly as it relates bach & Coates 2000; DiClemente 2000). to gender. AIDS diagnoses have increased Physical barriers rema in the primary among women each year between 1998 and method for preventing the spread of HIV dur­ 2002 , from 23.8 percent to 26 percent of all ing heterosexual intercourse. To date, pro­ cases, with a corresponding decrease moting the use of male condoms has been among men. Foreshadowing the probable at the core of most HIV prevention efforts future of the epidemic, surveillance of new (UNAIDS 2001 ). Because use of the male HIV infections among 13-24 year olds indi­ condom requires the partner's cooperation cates that almost half (47%) are occurring -which may be difficult to enlist- there is a among young women (CDC 2003a). The long-recognized need for female-controlled changing gender dynamics of the HIV/AIDS prevention methods (Gollub 1995; Gupta & epidemic are linked to a shift in exposure Weiss 1995; Stein 1990). The female con­ categories over time. Early on , the epidemic dom was introduced in the early 1990s and was concentrated among gay and bisexual represents an alternative to the male con­ men and their partners; however, as HIV dom. However, although research studies spread into the general population, this pat­ suggest that women are willing to try them tern changed . The incidence of exposure (e.g ., Hoffman, Exner, Leu , Ehrhardt, & Stein through heterosexual contact increased be­ 2003), female condoms are not widely used tween 1998 and 2002: from 12.3 percent to due in part to their high cost, difficulty of in­ 15.9 percent of AIDS cases among men and sertion experienced by some women , and from 61 .3 percent to 68.2 percent of AIDS visibility to the male partner (Logan , Cole, & cases among women. Given these trends, Leukefeld 2002; O'Leary & Wingood 2000). the HIV/AIDS epidemic in the U.S. is likely to Moreover, condom use is not a viable long­ increasingly resemble the pattern observed term strategy for a woman who wants to be­ in much ofthe rest of the world , where women come pregnant. Thus, development of an un­ account for approximately half of the H IV/AIDS obtrusive, inexpensive, female-controlled cases and heterosexual contact is the pri­ barrier - most likely in the form of a vaginal mary mode of transmission. This changing microbicide that would prevent transmission reality will pose a formidable challenge to of HIV while permitting conception to occur public health efforts and require that HIV/AIDS - remains a high priority. According to the prevention efforts be refined and redirected . National Institute of Allergy and Infectious Diseases (2004), several topical microbi­ PREVENTING HETEROSEXUAL TRANSMIS­ cides are currently being tested in clinical SION OF HIV/AIDS AMONG WOMEN: trials to evaluate acceptability, safety, and ef­ APPROACHES AND CHALLENGES fectiveness. Evidence is clear that it is more difficult for In the absence of new prevention tech­ women to ach ieve condom use in an ongo­ nologies, ind ividuals can reduce their risk of ing stable relationship. Unfortunately, many exposure to HIV by engaging in safer sexual (probably most) women who have become practices. Many intervention approaches infected globally have been infected by a aimed at changing sexual behavior among primary partner. (O'Leary & Wingood 2000 different populations have been developed. 195-196) Interventions draw on a variety of theoretical bases, including individual-level (Fisher & Within the last 100 years, medical science Fisher 2000) and community-level (Rogers has allowed humans to eradicate most in­ 2000) models. Of greatest relevance to this fectious diseases that caused the majority paper, a large number of programs have of deaths throughout history, either through been implemented with women at risk of the development of vaccines or the discov­ heterosexual transmission of HIV (for re­ ery of effective cures. This has not been the views, see lckovics & Yoshikawa 1998; Lo- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 17 gan et al 2002; O'Leary & Wingood 2000). ships. Most U.S.-based intervention programs The accumulated evidence indicates that aimed at heterosexual women have focused even the most successful programs are of on individual-level factors and been delivered limited success in promoting sexual behav­ in small-group settings (lckovics & Yoshika­ ior changes among general populations of wa 1998). In their review of HIV/STI interven­ heterosexual women. Ten years ago, Amaro tions for sexually active heterosexual women (1995) proposed that one major reason for (excluding commercial sex workers), O'Leary this lack of success is that programs tend to and Wingood (2000) noted that the majority be based on theoretical models of sexual recruited women in urban health-care set­ behavior that are of limited applicability to tings, with a primary focus on the promotion women because they ignore the fact that of condom use. sexual behavior takes place within social and Several recent reviews provide informa­ cultural contexts; assume that sexual activity tion regarding the success of behavior is under an individual's control (rather than change interventions for heterosexual being the result of impulse or coercion); and women. A meta-analysis of 30 randomized ignore the influence of sociocultural factors intervention trials implemented with high-risk (e.g., gender roles, values, norms) on sex­ heterosexual adults (Logan et al 2002) re­ uality. Reflecting on the modest results of vealed small but significant overall effects, their meta-analysis, Logan and colleagues indicating that interventions led to increases (2002) echoed these themes, stressing the in condom use and decreases in the num­ need for prevention programs that acknowl­ ber of sex partners. There were no gender edge the relational context in which sexual differences in effect sizes for condom use (r activity typically occurs, target men or couples, = .059); the mean weighted effect size for and take a comprehensive approach to HIV number of partners was significant for prevention (e.g., combining HIV interventions women (r = .073, p < .005) but not men (r = with substance abuse treatment or primary .002, ns), suggesting that women were more care). O'Leary and Wingood emphasized the likely than men to reduce their number of sex need to partners after participating in interventions. The findings are encouraging, but the au­ begin to address the economic, cultural, and thors noted that there was little evidence that sociopolitical issues that underlie the diffi­ initial behavior changes are sustained over culty many women face in their efforts to time. Moreover, although the small number protect themselves, (2000 196) of studies in each category made drawing firm conclusions impossible, interventions including power disparities and violence. involving women who were most likely to per­ Feminist scholars (e.g., Gupta 2000; Heise ceive themselves at risk of HIV infection ap­ 1995) have pointed out how traditional gen­ peared to be most effective: no significant der relations place women at a disadvan­ effects for either condom use or number of tage in sexual relationships. In response to partners were found for programs whose these concerns, recent interventions have participants were recruited in primary health attempted to go beyond the individual level care settings as compared to STD clinics, and consider broader influences on sexual­ drug treatment facilities, or low-income hous­ ity. ing (Logan et al 2002). Reviewing programs Some programs have focused on build­ for women around the world, lckovics and ing social support and community norms that Yoshikawa (1998) reported that 32 of 51 inter­ foster women's sexual risk reduction efforts. ventions led to "slight to moderate" changes For example, Sikkema and Kelly (2000) de­ in at least one measure of sexual behavior scribed a multisite community level interven­ (usually condom use), with the most suc­ tion implemented in 18 low-income housing cessful programs targeting commercial sex developments in five U.S. cities. After base­ workers rather than general populations of line assessments were completed, hous­ heterosexual women. Based on their review, ing developments were randomized to one O'Leary and Wingood (2000) concluded that of two conditions. Women living in housing interventions tended to be most effective at developments assigned to the experimental promoting sexual behavior change among group ("intervention developments") received women who were not in long-term relation- a one-year community-level intervention that 18 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology involved opinion leaders in risk reduction no significant differences were observed be­ workshops and the formation of Women's tween women who received the intervention Health Councils to undertake prevention out­ with their partners and those who received reach activities and organize community­ the intervention alone. The authors specu­ based events. Condoms and AIDS educa­ late that this was due to similar intervention tional materials were made available to all content (both the couples and woman-alone women in the intervention housing develop­ conditions focused on relationship issues, ments. Women living in housing develop­ and participants completed homework as­ ments assigned to the control group ("com­ signments with their partners) and self-se­ parison developments") were mailed a cou­ lection (only 56% of eligible couples partici­ pon for 10 free condoms and HIV/AIDS bro­ pated , and half of the cases of non-participa­ chures. Baseline assessments and 12- tion were due to male partner refusal, so month follow-ups conducted with 690 prima­ women with highly resistant or abusive part­ rily African American women revealed that ners were probably under-represented in the women in the intervention developments sig­ study). The study findings are encouraging nificantly decreased their HIV risk behavior, because they suggest that it is possible to whereas those in the comparison develop­ promote condom use among couples in ments showed little change. For example, long-term relationships by focusing on how the percentage of intercourse acts protected issues of trust and intimacy may act as bar­ by condoms during the past two months in­ riers to HIV/STI protection. At the same time, creased from 30 percent at baseline to 47 the challenges of involving couples in HIV/ percent at follow-up among women in the AIDS interventions are apparent. intervention developments; those in compari­ Other interventions explicitly address gen­ son developments showed little change der-related aspects of sexual relationships. (34% to 36%) . The effects were magnified One promising model is Connell's (1987) among women who had been exposed to theory of gender and power, which elucidates more of the intervention activities. Results sexual inequality and gender and power im­ from this study suggest that community-level balances in relationships between men and interventions can be successful in bringing women. Connell identified three major so­ about sexual behavior change among cial structures that characterize male-female women. relationships: sexual division of labor, divi­ Responding to the call that interventions sion of power, and the structure of cathexis target couples rather than individuals, a re­ (i.e. , affective aspects of the relationship). cent study examined the efficacy of a rela­ Wingood and DiClemente (2000) identified tionship-based program for heterosexual HIV/AIDS interventions for women that re­ couples (EI-Bassel, Witte , Gilbert, Wu, flected components of the theory of gender Chang , Hill, & Steinglass 2003). Women in and power. Although few interventions were long-term relationships with partners they explicitly based on the theory of gender and knew or suspected engaged in at least one power, there was support for individual com­ HIV/STI risk behavior (e.g ., sex with another ponents of the model. Recent interventions partner, injecting drug use) were recruited based on the theory of gender and power and asked to recruit their partner into the have shown success in promoting sexual study. Both partners completed baseline risk reduction among incarcerated women measures and then couples were randomly (St. Lawrence, Eldridge, Shelby, Little, assigned to one of three study arms: couples Brasfield, & O'Bannon 1997) and women liv­ condition (six session intervention delivered ing with HIV (Wingood, DiClemente, Mikhail, to the couple) ; woman-alone condition (the Lang , McCree, Davies, Hardin, Caliendo, woman attended the six session interven­ Hook, & Vernumd 2004), suggesting the po­ tion without her male partner); education con­ tential of this approach for HIV prevention with trol (the woman received a one-session in­ general populations of women (Harvey formation session). Both partners completed 2000). follow-up assessments three months post­ In recent years, the construct of empower­ intervention. Participants in the couples and ment has been incorporated into HIV inter­ woman-alone conditions showed signifi­ ventions for women. An example is a multi­ cantly safer sexual behavior than those in faceted program for Latina immigrant women the control condition at follow-up; however, in San Francisco named Mujeres Unidas y Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 19

Activas [active and united women] (MUA) HIV/AIDS but were convinced by their hus­ (Gomez, Hernandez, & Faigeles 1999). MUA bands that the couple was not at risk, and is an education, organizing and advocacy one-third reported that their husbands had project aimed at empowering immigrant and agreed "in principle" to use condoms. Par­ refugee Latina women and the broader ticipants noted a need to organize workshops Latino immigrant community to bring about for men, emphasizing the challenges women economic, political and social equality. HIV encountered in attempting to negotiate safer prevention was not the sole focus of the pro­ sex with their spouses. gram, but was integrated into MUA activities, As these examples illustrate, efforts to ex­ which included group meetings, support ses­ pand beyond the individual level show prom­ sions, HIV workshops, leadership training, ise for reducing women's risk of HIV infec­ community participation activities, and a vol­ tion . However, because many women can­ unteer HIV peer educator program. An evalu­ not control their male partners' behavior, ation of 74 women who participated in there is an urgent need to focus prevention baseline, 3-month, and 6-month interviews efforts on heterosexual men and address revealed that consistent condom use was how male sexuality and gender-related fac­ low at baseline (41% of participants had ever tors contribute to the transmission of HIV. used a condom, and 21% used one at last intercourse) and remained low at the 6- REFOCUSING HIV PREVENTION EFFORTS month follow-up (48% lifetime, 26% last inter­ ON MEN course). Participants did show significant in­ For all their sexual activity, for all the in­ creases in sexual communication comfort stances of sexual distress and anguish they and sexual comfort over time, and decreased inflict upon young women, young men pur­ endorsement of traditional sexual gender sue and are left to pursue sex and their norms. Among women with male partners in understanding of it in almost total silence the past year, fear of coercion decreased; and in the absence of support. It is not sur­ among those in steady relationships, the prising, therefore, that they get it wrong so male partner's decision-making power was often. (Dowsett, Aggleton, Abega, Jenkins, reported to have decreased. The findings in­ Marshall, Runganga, Schiffer, Tan & Tarr dicate that community-initiated programs 1998 305) may be a valuable way of reaching women, but underscore the need for identifying meth­ It has long been recognized that health ods of bringing about sexual behavior change outcomes are directly affected by gender-re­ within more general programs. lated factors, yet it is only recently that male Another participatory approach is action gender roles and notions of masculinity have research, which is rare in the U.S. but has been explicitly considered in HIV prevention. been used effectively in other countries. One This shift was prompted by the realization example is the program developed by the that a woman who is at risk from her partner's CONNAISSIDA group to address the HIV/ behavior rather than her own, or who lacks AIDS epidemic in Kinshasa, Zaire. Action re­ power in her personal relationships, will have search "begins with the principle that people limited ability to reduce her risk of HIV infec­ already know a great deal about their own tion (Ehrhardt 1992; Logan et al 2002; Seal situations" (Schoepf 1993 1403) and builds & Ehrhardt 2004). To address this reality, inter­ on this knowledge through social interaction ventions must begin to address gender-re­ to develop a critical consciousness about, in lated issues among men as well as women this case, sexual behavior and HIV and its (Gupta 2000). causes. The group created culturally appro­ Few interventions that focus specifically priate risk-reduction workshops targeting dif­ on men have been developed, but discus­ ferent groups of women (e.g. , sex workers, sions of how best to do so are becoming church group members). Evaluations were more common in the literature. A meta-analy­ conducted with 60 women who participated sis of 30 randomized HIV prevention inter­ in the church group workshops. One-third ventions for adult heterosexuals identified reported that their husbands refused to use just three programs exclusively for men (Lo­ condoms and/or reacted with anger to the gan et al 2002). Recently, Seal and Ehrhardt request that they do so, one-third said they (2004) recommended that efforts to promote were able to talk to their husbands about behavior change among heterosexual men 20 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology incorporate risk reduction messages that Mpowerment program, which builds support­ take into account existing sexual behavior ive risk-reduction communities of young gay patterns (e.g ., promote condom use with non­ and bisexual men (Hayes, Rebchook, & primary partners rather than attempting to Kegeles 2003), has been successful in fos­ eliminate sexual activity with multiple part­ tering safer sexual behavior. The program ners) and build upon societal gender role was initially implemented in Euguene, OR; norms (e.g., expand the notion of men as the delayed intervention comparison site was initiators of sexual activity by teaching men Santa Barbara, CA (Kegeles & Hays 1996). to initiate safer sex efforts). Other scholars The program was designed by a "Core stress the need to identify approaches that Group" of young men with input from a Com­ help heterosexual men recognize how their munity Advisory Board. Activities were aimed sexuality contributes to their own and their at diffusing HIV/AIDS prevention information female partners' risk of HIV/AIDS (e .g., and condom promotion messages through Campbell1997; Gupta 2000). Attempts to do peer outreach, establishment of social set­ this can articulate how male sexuality puts tings where young men could gather, and both men and women at risk of HIV trans­ organization of events that would provide out­ mission by raising several key con­ reach opportunities, including activities as siderations. First, it should be emphasized diverse as house parties, picnics and art that masculinity is culturally, socially and his­ shows. Prevention messages were trans­ torically constituted (Kilmartin 2000) and can mitted through small groups, informal out­ be expressed in many different forms reach , and a media campaign. Approximately (Courtenay 2000). Next, it must be acknowl­ 100 gay men aged 18-29 from each com­ edged that men's behavior is constrained by munity were assessed at pre- and post-inter­ traditional expectations about gender and vention via mail surveys. The intervention site these traditional models of masculinity place had a decrease in the proportion of men en­ both men and women at heightened risk for gaging in unprotected anal intercourse over­ HIV infection (Gupta 2000). In general, mas­ all (41 % to 30%), with non-primary partners culinity has been linked to poor health be­ (20% to 11 %) and with a boyfriend (59% to haviors. Relevant to sexual behavior, men's 45%). No comparable changes occurred in health behaviors often demonstrate a domi­ the comparison group, suggesting the power nant (hegemonic) notion of masculinity: a of a collective approach that builds on com­ denial of weakness, virility, appearance of munity strengths. being strong, emotional and physical con­ There are also powerful lessons to be trol. It is often in the pursuit of power and learned by looking at how countries around privilege that men are led to harm them­ the world are addressing the heterosexual selves (Courtenay 2000; Kandrack, Grant, & HIV/AIDS epidemic. For example, social mar­ Segall 1991) and subsequently women. Tra­ keting campaigns have been successfully ditional masculinities sometimes encourage implemented in several countries in Africa. men to force unwilling partners into sex, re­ In Za ire, a television soap opera was devel­ ject condom use and view drinking as a con­ oped as part of a larger campaign to pro­ firmation of their manhood (Campbell 1997; mote gender equity and safer sex. After fa­ Heise 1995). Finally, it must be recognized mous actors were seen negotiating condom that masculinities are tied to hierarchy and use in bedroom scenes, almost three-quar­ power relations: each culture or group shows ters of viewers said they were motivated to more dominant and subordinate forms of change their behavior (Communication Ini­ masculinity (e .g., the masculinity of a sports tiative 2004); furthermore, reports of mutual figure may be valued more highly than that of fidelity increased from 28.5 percent to 46 a schoolteacher). percent and condom sales increased by 443 Although HIV prevention programs for percent within a one-year period (Ferreros, heterosexual men have not been explicitly Mivumbi, Kakera, & Price 1990, in Wingood based on these constructs, community & DiClemente 2000). National level pro­ based interventions and empowerment ap­ grams in other countries, including Switzer­ proaches aimed at changing men's sexual land's "STOP AIDS" campaign and Thai­ behavior have been undertaken in the gay land's "100 % Condom Program" support the and bisexual communities in the U.S. (Sik­ notion that it is possible to bring about be­ kema & Kelly 2000) . For example, the havior change in heterosexual men (Auer- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 21

bach & Coates 2000). linked to heterosexual transmission of HIV. The international community has also re­ Comparable interventions are lacking in the sponded to the message that men are espe­ U.S. so it is impossible to tell how success­ cially crucial in the prevention of HIV. One ful this type of program might be in the U.S. example is a project being implemented by context. Nonetheless, the development of Institute PROMUNDO, a non-governmental male-focused interventions represents an organization based in Rio de Janeiro, Brazil important direction for future research. One (Horizons 2004). This ongoing project is concern is whether men will participate in aimed at reducing HIV risk behaviors and such interventions. There are some indica­ sexual violence by changing young men's tions that men are less likely than women to attitudes about gender roles and sexual rela­ take part in interventions. For example, half tionships. In a quasi-experimental study, 708 of non-participation in EI-Bassel et al.'s young men from three different low-income (2003) couples study was due to male part­ areas of Rio de Janeiro were assessed be­ ner refusal. In the NIMH Multisite HIV Preven­ fore their communities were assigned to one tion Trial (1998), which recruited individuals of three intervention conditions: group ses­ to participate in small group interventions, sions led by adult facilitators; community­ only 86 percent of potential male participants wide social marketing campaign plus group who were screened at STI clinics and deter­ education; and delayed intervention (control mined to be eligible for the study actually community). Participants are being as­ completed baseline interviews, compared to sessed at 6- and 12-months on key indica­ 91 .5 percent of eligible female participants tors (e.g., gender role beliefs, violence, HIV­ screened in the same clinics and 95 percent related behavior) and qualitative interviews of women recruited in health serving organi­ are being conducted with a sub-sample of zations. Innovative strategies may be needed participants and their regular sexual part­ to overcome men's reluctance to take part in ners. At baseline, endorsement of less eq­ health promotion programs. uitable gender norms was significantly asso­ One potential avenue for reaching gen­ ciated with higher incidence of sexually trans­ eral populations of men in HIV prevention mitted infections (STI), lack of contraceptive efforts involves bringing prevention pro­ use, and physical and sexual violence against grams to men's natural settings. One obvi­ a current or recent partner. Preliminary find­ ous setting is the workplace, which has been ings from the two intervention sites indicate the context for information-based HIV pro­ that a significantly smaller proportion of re­ grams in the U.S. (Wilson, Jorgensen, & Cole spondents supported traditional gender 1996). Taking this idea further, an innovative norms after being exposed to the interven­ program for migrant gold miners in South tion. Based on 6- and 12-month follow-up Africa focused on behavioral change and at­ data, significant reductions in STI symptoms tempted to take into account the broader so­ were seen in both intervention communities, cial contexts affecting HIV transmission and significant increases in condom use (Campbell & Williams 1999). Preliminary with a primary partner were seen in the com­ work by Campbell and colleagues examined bined intervention community. Moreover, mine workers' notions of masculinity and changes in gender norms were associated found that HIV preventions programs that tar­ with changes in HIV/STI risk outcomes. The get individual behavior change are minimally findings demonstrate that interventions fo­ effective (Campbell 1997). Instead, the pro­ cused on gender dynamics can be success­ ject viewed HIV transmission as a commu­ ful in reducing men's sexual risk behaviors. nity problem. In addition to the mine work­ The larger magnitude of effects in the com­ ers, the program targeted the communities bined intervention as compared to the group surrounding the mines, where workers con­ education intervention suggests the impor­ ducted their social and sexual lives. The pro­ tance of community support for individual ject used both traditional healers and bio­ behavior change. medical practitioners; was managed not only These examples support the notion that by the mine management but by unions, focusing on gender-related considerations grassroots organizations, and national and exploring the implications of traditional health workers; and utilized community­ masculinity for men's sexual behavior may based and peer outreach. Although the inter­ lead to changes in attitudes and behaviors vention showed only limited success (Will- 22 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology iams, Taljaard , Campbell, Gouws, Ndhlove, glected populations of women are needed. van Dam, Carael, & Auvert 2003), the strat­ These include older women , women who do egy of embedding HIV prevention programs not self-identify as heterosexual yet engage in the workplace and the larger community in sexual intercourse with male partners, and offers a model for future programs. teenage girls. In this final section, we briefly Other projects have identified additional highlight considerations for prevention in venues for engaging men in HIV/AIDS pre­ these groups. vention programs. One program in Kenya Sexuality is a core aspect of the human uses soccer as a natural meeting place for experience throughout the lifespan (Ehrhardt discussing HIV/AIDS and safer sex (UNICEF & Wasserheit 1991; Levy 1994), yet the litera­ 2004). The Kibera Community Self-Help Pro­ ture is largely silent on the issue of HIV/AIDS gramme, a local non-governmental organi­ among older women. In an early analysis of zation supported by UNICEF, helps young women infected with HIV through heterosex­ volunteers conduct informal meetings about ual contact, women aged 50 and over at the HIV after soccer matches in communities time of their AIDS diagnosis were compared that have been heavily affected by HIV/AIDS. to women under 50 years old at the time of The volunteers create songs and impromptu diagnosis (Schable, Chu , & Diaz 1996). The theater in the hopes of changing the behav­ two groups differed in ethnicity: women aged ior and attitudes of their friends and kins­ 50 or older at the time of diagnosis were men. Unstructured grassroots programs more likely to be White or Hispanic/Latina, such as this do not readily lend themselves whereas those under 50 at initial diagnosis to randomized controlled evaluations and the were more likely to be African American. Older impact of many similar programs around the women were more likely to have been ex­ world has not been assessed; despite this, posed to HIV through sex with an HIV-infected they offer creative ideas for how to reach men. man whose risk of exposure was unknown In sum , the increased attention paid in (59% vs. 42% of women aged under 50 at recent years to developing interventions that the time of diagnosis), and less likely to re­ explicitly address gender-related issues and port sex with an injection drug user (31% vs . target heterosexual men has yielded encour­ 48%) . Hillman and Broderick (2002) identi­ aging results and provides a basis for future fied HIV risk factors for post-menopausal interventions that will reduce women's risk women , including biological risks (e.g., thin­ of exposure to HIV. At the same time as the ning vaginal walls due to lower levels of es­ field of HIV/AIDS prevention advances to stem trogen after menopause are more likely to the heterosexual transmission of HIV, tear during intercourse, facilitating transmis­ though, the epidemic is shifting Cjnd new sion of HIV) , condom-related factors (e .g., populations of women at heightened risk are lack of contraceptive need and limited so­ emerging. cialization for sexual communication with potential partners) , and partner-related is­ FUTURE CHALLENGES FOR HIV/AIDS sues (e .g., lack of knowledge about partners' PREVENTION AMONG WOMEN IN THE U.S. potential risk behaviors). Because of in­ I don't even know if I made the conscious creased survival rates as AIDS treatments effort to decide yes, this is the time and I'm improve, and the possibility of higher levels actually gonna do it [have intercourse]. It of sexual activity as elderly men take advan­ just kind of happened which as I look back tage of newly developed and aggressively I feel that's unfortunate. I wish I had been marketed anti-impotence drugs, HIV/AIDS able to think about it more and had been among older women is likely to increase in more assertive in saying , "whoa, I don't the future. Thus, an important direction for know if I'm ready at this particular time ." future prevention efforts will involve develop­ (Latina woman describing her first sexual ing interventions for older women and their intercourse; Raffaelli 2001 ) partners. Women who have sex with women repre­ It is impossible to predict with any degree sent another group that will need increased of certainty how the HIV/AIDS epidemic will attention as the epidemic progresses. This unfold in the next 25 years in the U.S. How­ may seem counter-intuitive, as there are few ever, there are indications that prevention documented cases of female-to-female programs addressing several currently ne- transmission of HIV. However, case reports Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 23 of female-to-female transmission, as well older partners are particularly vulnerable to as the well documented female-to-male engaging in unprotected sex (Darroch, transmission, show that vaginal secretions Landry & Oslak 1999). Intervention programs and menstrual blood are potential paths of for adolescents must also consider how sex­ transmission through mucous membranes ual behavior changes as youth gain sexual (Hughes & Evans 2003; Morrow 1995). In an experience and begin forming partnerships. attempt to better understand the association For example, condom use is common when between female-female sexual behavior and teens start having sex, but as adolescent girls sexually transmitted infections, Bauer and begin having sex regularly they tend to shift Welles (2001) studied 286 women recruited to other methods of birth control (Jemmott & at a gay/lesbian/bisexual pride rally. One key Jemmott 2000). Condom use is also more finding was that 13 percent of women with typical at the start of relationships; as couple only female partners reported a history of STI, becomes more committed and trust is es­ a rate clearly counter to a "no risk" group. tablished, condom use decreases and use Frequency of female-female sexual expo­ of other methods becomes more common sure was independently associated with in­ (e.g., Fortenberry, Wanzhu, Harezlak, Katz, & creased odds of STI when controlling for fe­ Orr 2002). It is not clear to what extent HIV male-male sexual behavior. Moreover, in 98 risk reduction interventi ons are effective in percent of cases of AIDS among women who addressing these aspects of adolescent have sex with women, risk factors other than sexuality. female-female sex were present (e.g., injec­ A number of interventions have demon­ tion drug use, sex with high-risk men) (CDC strated success in fostering safer sexual be­ 2003b). Because sexual identity and sexual havior among adolescents, although stud­ behavior may differ, there is a need to inter­ ies are often limited by short follow-up peri­ vene with women who identify as lesbian but ods and low rates of sexual activity at base­ engage in activities that put them at hetero­ line. One review found that over half (57%) of sexual risk of HIV infection. 23 HIV interventions evaluated in random­ Adolescents represent perhaps the most ized controlled trials achieved significant promising population for intervention efforts, sexual risk reductions (Pedlow & Carey 2003; because they are in the process of establish­ Kim , Stanton, Li , Dickersin, & Galbraith 1997). ing lifetime sexual behavior patterns. This A meta-analysis of interventions conducted population has not been neglected; as dis­ in community samples of adolescents re­ cussed below, many interventions targeting vealed a significant overall effect for condom young people have been developed. Yet there use (Jemmott & Jemmott 2000). Finally, in a are indications that current HIV-prevention review of school-based interventions aimed programs are not having a substantial im­ at decreasing teen pregnancy and HIV, Kirby pact on young people in the U.S. As noted (2000) reported that some comprehensive earlier, young women are increasingly con­ sex education programs were successful in tracting HIV through heterosexual transmis­ promoting sexual risk reduction. sion, and account for almost half of new HIV Despite these encouraging overall re­ infections among 13-24 year olds. A discus­ sults, it is unclear whether teenage girls and sion of factors contributing to young women's boys typically show the same level of risk risk of HIV/AIDS is beyond the scope of this reduction. The effectiveness of interventions paper; risk factors include biological vulner­ for female adolescents is not reported sepa­ ability due to the immaturity of the reproduc­ rately in most meta-analyses and reviews, tive tract, which facilitates transmission of but there is some evidence that intervention HIV, as well as psychological and social fac­ effects tend to differ by gender. In some cases tors linked to sexual behavior (see Rotheram­ intervention effects are weaker for girls than Borus, O'Keefe, Kracker, & Foo 2000). Pre­ boys (e.g., Stanton, Xiaoming, Ricardo, vention efforts for adolescent girls are com­ Galbraith, Feigelman & Kaljee 1996); in other plicated by several considerations. First, cases, interventions show differential effects sexual intercourse in this age group may be for male and female teens. For example, St. the result of male pressure or coercion, par­ Lawrence and colleagues (1995) compared ticularly among young teens (SIECUS 1997). the efficacy of a single-session HIV/AIDS Differential power between partners is also education session with an eight-session a factor; young women who have sex with behavioral skills training intervention in a 24 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology sample of 246 African American adolescents recruited in a public clinic serving low-in­ CONCLUSIONS come populations. In this randomized con­ It is becoming increasingly clear that pre­ trol trial, participants in the behavioral skills vention efforts must move beyond a focus on intervention showed reduced levels of risk individual women to curb heterosexual trans­ behaviors compared to youth in the control mission of HIV. Interventions must consider group through the one-year follow-up period. the realities of women's lives, take into ac­ However, boys and girls in the intervention count the contexts in which sexual encoun­ group showed different patterns of behavior ters occur, and involve men as well as at baseline, and there were gender differ­ women. Moreover, many women's only risk ences in observed behavior post-interven­ of HIV infection results from having unpro­ tion such that boys decreased their sexual tected sex with a primary partner; thus, broad­ risk-taking whereas girls maintained their based prevention efforts will be needed to initial (lower) levels of sexual risk-taking . In supplement intensive interventions aimed at contrast, girls and boys in the educational "high risk" populations. control condition tended to increase their sex­ The move from targeted to general inter­ ual risk-taking over time (if already sexually ventions is likely to be challenging in the U.S. active at baseline) or initiate sexual inter­ Because the HIV/AIDS epidemic has been course (if they were inexperienced at base­ concentrated among population sub-groups, line). Thus, the intervention appeared to help interventions have tended to be focused girls avoid increased sexual risk over time. rather than general. There has been resis­ Although it is impossible to tease apart which tance to using the mass media for HIV pre­ of the eight sessions are responsible for the vention in the U.S. (Ehrhardt 1992), despite specific effects, the intervention included evidence of the effectiveness of marketing sessions on behavioral skills to resist sexual approaches in changing attitudes and be­ pressure and verbal coercion, and the devel­ havior (Auerbach & Coates 2000). School­ opment of specific plans for avoiding un­ based interventions - which have the po­ wanted sex. tential to reach virtually all individuals in the The need to address gender-related is­ country in a cost-effective manner- are sub­ sues is reflected in results of a recent inter­ ject to political pressures that reduce their vention with sexually experienced African effectiveness (DiClemente 1993). Currently, American young women (aged 14-18) who local and federal policies in the U.S. con­ were randomly assigned to participate ei­ strain the content of school-based programs; ther in an intervention that emphasized eth­ federal funding is available to states willing nic and gender pride, HIV-related knowledge to provide "abstinence-only-until-marriage" and skills, and healthy relationships or in a sexuality education (but not comprehensive nutrition and exercise control intervention sexuality education), and 35 percent of (DiClemente, Wingood , Harrington, Lang , school districts require abstinence-only pro­ Davies, Hook, Oh, Crosby, Hertzberg, Gor­ grams (Pardini 2002/2003). This is unfortu­ don, Hardin , Parker & Robillard 2004). Rela­ nate because there is mounting evidence tive to the control participants, intervention that such programs are largely ineffective, participants showed decreases in multiple whereas comprehensive sex education is measures of sexual risk behaviors (e.g., un­ effective in fostering risk reduction without protected sex, inconsistent condom use) that increasing sexual activity among teens (see were sustained over a 12-month follow-up Jemmott & Jemmott 2000; Kirby 2000). period. In countries that hold more open sexual Based on these findings, it appears that attitudes, comprehensive programs aimed additional consideration of gender-related at the general population have shown con­ issues could provide valuable insight regard­ siderable success. For example, in Sweden ing how best to promote sexual risk reduc­ (which has a low overall rate of HIV/AIDS), tion among both female and male adoles­ sexual health is a national priority and man­ cents. By fostering safer sexual behavior in datory sex education in all public schools teenagers, prevention programs should ulti­ pre-dates the onset of the HIV/AIDS epidemic mately contribute to lower HIV/AIDS rates as (Herlitz & Steel 2000). The curriculum is age­ young people grow up and move into adult­ graded and developmentally appropriate, hood . beginning in primary school with basic re- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 25 productive information and progressing to J Public Health 91 1282-1286. more detailed information about contracep­ Campbell C. 1997. Migrancy, masculine identities tion and sexually transmitted infections in and AIDS: the psychosocial context of HIV secondary school. A network of regional transmission on the South African gold mines. health clinics provides access to reproduc­ Social Science Medicine 45 273-281. Campbell C & B Williams. 1999. Beyond the bio­ tive information and care. The effectiveness medical and behavioural: towards an integrated of Sweden's HIV/AIDS program has been approach to HIV prevention in the Southern evaluated through periodic cross-sectional African mining industry. Social Science Medi­ surveys of random population samples. cine 48 1625-1639. Among other observed changes, condom CDC. 2002. HIVIAIDS surveillance report, 14. At­ use among young people increased from lanta, GA: Centers for Disease Control and Pre­ 1987 to 1997. It is noteworthy that Sweden's vention. Retrieved May 21 , 2004, from http:// openness toward sexuality is associated not www. cdc.g ov/h iv/stats/hasrli n k. htm. with sexual permissiveness, but "rather the -...,..-,c--· 2003a. HIVIAIDS Among US Women: Minority and Young Women at Continuing Risk. ability to make informed and responsible Retrieved May 14, 2004, from http://www.cdc. decisions regarding sexuality" (Herlitz & Steel govlhiv/pubs/facts/women.htm. 2000 889). -.,...,---· 2003b. HIVIAIDS and US Women Who As heterosexual transmission of HIV be­ Have Sex With Women (WSW) . Retrieved May comes more common in the U.S., it will be 14, 2004, from http://www.cdc.gov/hiv/pubs/ necessary to initiate a national dialogue facts/wsw.htm. about how to protect everyone at potential -..,....,.....,.--· 2004. 10 Leading Causes of Death, risk from exposure to HIV/AIDS. New risk re­ United States. Retrieved May 25, 2004, from duction approaches are needed that take into http://webappa.cdc.gov/sasweb/ncipc/lead caus1 O.html. account the complex interplay of factors con­ Communication Initiative. 2004. Impact data- Mass tributing to HIV risk among heterosexual Media Campaign HIVIAIDS- Zaire. Retrieved women. In particular, there is an urgent need on June 15, 2004, from http://www.comminit. to refocus prevention efforts on men and com/idmay15/sid-2258.html/ boys, and identify theoretical models that pro­ Connell RW. 1987. Gender and Power. Stanford, vide avenues for behavior change interven­ CA: Stanford U Press. tions among heterosexual males. In this Courtenay W. 2000. Constructions of masculinity paper, we have suggested that program de­ and their influence on men's well being : a velopers and policy makers in the U.S. can theory of gender and health. Social Science Medicine 50 1385-1401 . learn from the innovative strategies that have Darroch JE, OJ Landry, & S Oslak. 1999. Age dif­ been implemented around the globe. By iden­ ferences between sexual partners in the tifying and implementing participatory and United States. Family Planning Perspectives empowerment-based approaches, it will ulti­ 31 160-167. mately be possible to reduce the impact of DiClemente RJ . 1993. Preventing HIV/AIDS among heterosexual transmission of HIV/AIDS in the adolescents: schools as agents of behavior U.S. change (Editorial). JAMA 270 760-762. --:--:-::.· 2000. Looking forward: future directions ENDNOTE for HIV prevention research. Pp. 311-324 in Handbook of HIV Prevention. JL Peterson & 1 This paper was completed in 2004, but publica­ tion was delayed due to factors beyond the RJ DiClemente, eds. NY: Kluwer. author's control. Therefore, work published DiClemente RJ, GM Wingood, KF Harrington, DL after 2004 was not considered. Lang, SL Davies, EW Hook, MK Oh, RA Crosby, VS Hertzberg, AB Gordon, JW Hardin, SParker, REFERENCES & A Robillard. 2004. Efficacy of an HIV-preven­ tion intervention for African American adoles­ Amaro H. 1995. Love, sex, and power: consider­ cent females: a randomized controlled trial. ing women's realities in HIV prevention. Amer JAMA. Psychologist 50 437-447. Dowsett GW, P Aggleton, S Abega, C Jenkins, TM Auerbach JD & T J Coates. 2000. HIV prevention Marshall, A Runganga, J Schifter, ML Tan , & CM research: accomplishments and challenges for Tarr. 1998. Changing gender relations among the third decade of AIDS. A mer J Public Health young people: the global challenge for HIV/AIDS 90 1029-1 032. prevention. Critical Public Health 8 291 -309. Bauer GR & SL Welles. 2001 . Beyond assumption Ehrhardt AA. 1992. Trends in sexual behavior and of negligible risk: sexually transmitted diseases the HIV pandemic (Editorial). Amer J Public and women who have sex with women. Amer Health 82 1459-1461 . 26 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Ehrhardt AA & JN Wasserheit. 1991 . Age, gender, associated with HIV and sexual risk behaviour and sexual risk behaviors for sexually trans­ from 1987 to 1997. A/OS 14 881-890. mitted diseases in the United States. Pp. 97- Hillman JM & KJ Broderick. 2002. HIV among eld­ 121 in Research Issues in Human Behavior erly women: ignored and overlooked by health and Sexually Transmitted Diseases in the care p(oviders and public policy makers. Pp. A/OS Era. JN Wasserheit, SOAral, KK Holmes, 193-215 in Charting a New Course for Femi­ & PJ Hitchcock, eds. Washington, DC : Ameri­ nist Psychology. LH Collins, MR Dunlap, & JC can Society for Microbiology. Chrisler, eds. Westport, CT: Praeger. EI-Bassel N, SS Witte, L Gilbert, E Wu , M Chang , J Hoffman S, TM Exner, C Leu , AA Ehrhardt, & Z Hill, & P Steinglass. 2003. The efficacy of a Stein. 2003. Female-condom use in a gender­ relationship-based HIVISTD prevention program specific family planning clinic trial. Amer J Public for heterosexual couples. Amer J Public Health 93 1897-1903. Health 93 963-969. Horizons. 2004. Promoting healthy relationships Ferreros C, N Mivumbi, K Kakera, & J Price. 1990. and HIVISTD prevention for young men: posi­ Social marketing of condoms for AIDS preven­ tive findings from an intervention study in Brasil. tion in developing countries: The Zaire experi­ Retrieved June 12, 2004, from http:// ence. Paper presented at the Sixth Interna­ www. popcou neil. orglh orizonslhorizons. htm I. tional Conference on AIDS, San Francisco, CA. Hughes C &A Evans. 2003. Health needs of women Fisher JD & WA Fisher. 2000. Theoretical ap­ who have sex with women . Brit Medical J proaches to individual-level change in HIV risk 327 939-941 . behavior. Pp . 3-55 in Handbook of HIV Preven­ lckovics JR & H Yoshikawa. 1998. Preventive in­ tion. JL Peterson & RJ DiClemente, eds. NY: terventions to reduce heterosexual HIV risk Kluwer. for women : current perspectives, future di­ Fortenberry JD, T Wanzhu, J Harezlak, BP Katz, & rections. A/OS 12 Supp S197-S208. DP Orr. 2002. Condom use as a function of time Jemmott JB & LS Jemmott. 2000. HIV behavioral in new and established adolescent sexual re­ interventions for adolescents in community lationships. Amer J Public Health 92 211-213. settings. Pp. 103-127 in Handbook of HIV Pre­ Gollub EL. 1995. Women-centered prevention tech­ vention. JL Peterson & RJ DiClemente, eds. NY: niques and technologies. Pp . 43-82 in Women Kluwer. at Risk: Issues in the Primary Prevention of Kandrack MA, KR Grant, & A Segall. 1991 . Gender A/OS. A O'Leary & LS Jemmott, eds. NY: Ple­ differences in health related behavior: some num. unanswered questions. Soc Sci Medicine 32 Gomez CA, MH Hernandez, & BF Faigeles. 1999. 579-590. Sex in the new world: an empowerment model Kegeles SM & RB Hays. 1996. The Mpowerment for HIV prevention in Latina immigrant women . project: a community-level HIV prevention in­ Health Education Behavior 26 200-213. tervention for young gay men . Amer J Public Gupta GR. 2000. Gender, sexuality, and HIVIAIDS: Health 86 1129-1136. The what, the why, and the how. Plenary ad­ Kilmartin C. 2000. The Masculine Self. Boston : dress delivered at the International AIDS Con­ McGraw Hill. ference, Durban, South Africa. Retrieved June Kim N, B Stanton, X Li , K Dickersin, & J Galbraith. 5, 2004, from www.icrw.org . 1997. Effectiveness of the 40 adolescent AIDS­ Gupta GR & E Weiss. 1995. Women's lives and risk reduction interventions: a quantitative re­ sex: implications for AIDS prevention. Pp. 259- view. J Adolescent Health 20 204-215. 270 in Conceiving Sexuality: Approaches to Kirby D. 2000. School-based interventions to pre­ Sex Research in a Postmodern World. RG vent unprotected sex and HIV among adoles­ Parker & JH Gagnon , eds. NY: Routledge. cents. Pp. 83-101 in Handbook of HIV Preven­ Harvey SM. 2000. New kinds of data , new options tion. JL Peterson & RJ DiClemente, eds. NY: for HIV prevention among women : a public Kluwer. health challenge. Health Edu Behavior 27 566- Levy JA. 1994. Sex and sexuality in later life 569. stages. Pp . 287-309 in Sexuality Across The Hayes RB, GM Rebchook, & SM Kegeles. 2003. Life Course. AS Rossi ed . Chicago, IL: U Chi­ The Mpowerment project: community-building cago Press. with young gay and bisexual men to prevent Logan TK, J Cole, & C Leukefeld . 2002. Women, HIV. Amer J Community Psychology 31 301- sex, and HIV: social and contextual factors, 312. meta-analysis of published interventions, and Heise LL. 1995. Violence, sexuality, and women's implications for practice and research . Psy­ lives. Pp. 109-134 in Conceiving Sexuality: Ap­ chological Bull 128 851-885. proaches to Sex Research in a Postmodern Morrow KM . 1995. Lesbian women and HIVIAIDS: World . RG Parker & JH Gagnon, eds. NY: an appeal for inclusion. Pp . 237-256 in Women Routledge. at Risk: Issues in the Primary Prevention of Herlitz CA & JL Steel. 2000. A decade of HIVIAIDS A/OS. A O'Leary & LS Jemmott, eds. NY: Ple­ prevention in Sweden: changes in attitudes num. Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 27

National Institute of Allergy and Infectious Dis­ St. Lawrence JS, TL Brasfield, A Shirley, KW eases. 2004. HIV infection in women. Re­ Jefferson, E Alleyne, & RE O'Bannon. 1995. trieved May 14, 2004, from http://www.niaid. Cognitive-behavioral interventions to reduce nih.gov/factsheets/womenhiv. htm. African American adolescents' risk for HIV NIMH Multisite HIV Prevention Trial Group. 1998. infection. J Consulting Clinical Psychology 63 The NIMH Multisite HIV Prevention Trial: Reduc­ 221-237. ing HIV sexual risk behavior. Science 280 1889- St. Lawrence JS, GD Eldridge, MC Shelby, CE Little, 1894. TL Brasfield, & RE O'Bannon. 1997. HIV risk O'Leary A & LS Jemmott. 1995. Preface. Pp. ix-x in reduction for incarcerated women: a compari­ Women at Risk: Issues in the Primary Pre­ son of brief interventions based on two theo­ vention of AIDS. A O'Leary & LS Jemmott, eds. retical models. J Consulting Clinical Psych 65 NY: Plenum. 504-509. O'Leary A & GM Wingood. 2000. Interventions for Stanton BF, L Xiaoming, I Ricardo, J Galbraith, S sexually active heterosexual women. Pp. 179- Feigelman, & L Kaljee. 1996. A randomized, 200 in Handbook of HIV prevention. JL Peterson controlled effectiveness trial of an AIDS pre­ & RJ DiClemente, eds. NY: Kluwer. vention program for low-income African-Ameri­ Pardini P. 2002/2003. Abstinence-only education can youths. Archives Pediatric & Adolescent continues to flourish. Rethinking Schools 17 Medicine 150 363-372. 1-8. Retrieved June 21 , 2004, from http://www. Stein Z. 1990. HIV prevention: the need for meth­ rethinkingschools .org/archive/17 _02/ ods women can use. Amer J Public Health 80 Abst172.shtml. 460-462. Pedlow CT & MP Carey.· 2003. HIV sexual risk­ UNAIDS. 2001. The Global Strategy Framework reduction interventions for youth: a review and on HIVIAIDS. Retrieved June, 4, 2004 from methodological critique of randomized controlled www. u nfpa. org/h iv. 2002 update/4e. htm. trials. Behavior Modification 27 135-190. UNFPA. 2001 . HIV prevention now, Programme Raffaelli M. 2001 . Dimensions of acculturation re­ Briefs, No. 1. Retrieved June 5, 2004 from lated to Latino sexual behavior. Unpublished www. unfpa. org/hiv/prenention/hivprev1 b. htm. data. UNICEF. 2004. Football and HIV/AIDS Prevention Rogers EM. 2000. Diffusion theory: a theoretical in Kenya . Retrieved June 5, 2004 from http:// approach to promote community-level change. www. u n ice f . o rg /fo otba II . world /kenya­ Pp. 57-65 in Handbook of HIV Prevention. JL kennedy.html. Peterson & RJ DiClemente, eds. NY: Kluwer. Williams BG, D Taljaard, CM Campbell, E Gouws, L Rotheram-Borus MJ, Z O'Keefe, R Kracker, & H Ndhlove, J van Dam, M Carael, & B Auvert. Foo. 2000. Prevention of HIV among 2003. Changing patterns of knowledge, re­ adolescents. Prevention Sci 1 15-30. ported behaviour and sexually transmitted in­ Schable B, SY Chu, & T Diaz. 1996. Characteristics fections in a South African gold mining commu­ of women 50 years of age or older with nity. A/OS 17 2099-2107. heterosexually acquired AIDS. Amer J Public Wilson MG, C Jorgensen, & G Cole. 1996. The Health 86 1616-1618. health effects of worksite HIV/AIDS interven­ Schoepf BG. 1993. AIDS action-research with tions: a review of the research literature. Amer women in Kinshaha, Zaire. Soc Sci Medicine J Health Promotion 11 150-157. 37 1401-1413. Wingood GM & RJ DiClemente. 2000. Application Seal OW & AA Ehrhardt. 2004. HIV-prevention of the theory of gender and power to examine sexual health promotion for heterosexual men HIV related exposures, risk factors, and ef­ in the US: pitfalls and recommendations. fective interventions for women. Health Edu & Archives Sexual Behavior 33 211-222. Behavior 27 539-565. SIECUS. 1997. Male involvement in teen pregnancy. Wingood GM, RJ DiClemente, I Mikhail, DL Lang, SHOP Talk (School Health Opportunities and DH McCree, SL Davies, JW Hardin, AM Caliendo, Progress) Bulletin 2. EW Hook, & S Vernumd. 2004. A randomized Sikkema KJ & JA Kelly. 2000. Outcomes of a controlled trial to reduce HIV transmission risk randomized community-level HIV prevention behaviors and STDs among women living with intervention for women living in 18 low-income HIV: The Willow program. JAIDS. housing developments. Amer J Public Health 90 57-64. 28 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

tim ~lication is available in microform

Please send me additional information. University Microfilms International 300 North Zeeb Road 18 Bedford Row Dept. P.R. Dept. P.R. Ann Arbor, Ml 48106 London, WC1A 4EJ U.S.A. England

Na~------lnstttutlon Street ______City ______State ______Zip ----- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 29 PATTERNS OF INTIMATE PARTNER VIOLENCE AMONG DRUG USING WOMEN

Michael Duke, Prevention Research Center; Wei Teng, Yale University Hospital; Scott Clair, Iowa State University; Hassan Saleheen, Hispanic Health Council; Pamela Choice, HTA Consulting; and Merrill Singer, Hispanic Health Council

ABSTRACT

Following from growing concern with the role of violence in intimate relationships, this paper examines the relationship between partner violence dynamics and illicit drug use among women in Hartford, CT. Based on an interview sample of 497 street-recruited, not-in-treatment, drug-involved women, the paper compares drug use and health risk among women in four types of relationships: I) those without self­ reported violence; 2) those in which there is violence targeted at a female partner; 3) those in which the violence is perpetrated by the female partner; and 4) those in which there is bi-directional or mutual violence. Findings suggest that drug treatment programs that serve women should therapeutically address the issue of intimate partner violence.

Recent scholarship in the domestic vio­ the role of batterer. Likewise, as Tjaden and lence literature has explored the complex re­ colleagues (1999) have indicated, rates of lationship between substance abuse and inti­ partner violence in female same-sex rela­ mate partner violence (IPV) (Amaro, Fried, tionships are roughly comparable to those Cabral & Zuckerman 1990; Bennett 1995; of heterosexual couples. Finally, acts of vio­ Caetano, Cunradi, Clark, & Schafer 2000; lence within an intimate relationship can be Cunradi, Caetano, Clark, & Schafer 1999; reciprocal, rather than unidirectional. In or­ Cunradi, Caetano, & Schafer 2002; EI­ der to understand the relationship between Bassel, Gilbert, Witte, Wu, Gaeta, Schilling & substance abuse and intimate partner vio­ Wada 2003; Goldberg 1995; Leadley, Clark, lence in all its complexity, it is thus important & Caetano 2002; Lown & Vega 2001 ; Sharps, to explore the full panorama of relationship Campbell, Campbell, Gary, & Webster 2001). dynamics vis-a-vis IPV, including: relation­ Although the initial focus of this work was on ship history; social support; directionality of the substance abusing behaviors of batter­ partner violence; prior exposure to sexual vio­ ers, there is a growing interest, as well, in lence; and drug procurement and sharing substance misuse by victims of such vio­ behaviors of romantic partners. lence, particularly women (Cunradi et al2002; This paper examines the relationship be­ EI-Bassel et al 2003; Gilbert, EI-Bassel, Ra­ tween partner violence dynamics and illicit jah, Fontdevila, Foleno, & Frye 2000). Spe­ drug use among substance-involved women cifically, this latter work has examined both in Hartford, CT. Utilizing a sample of 497 the ways in which the grim necessity of addic­ street-recruited, not-in-treatment heroin and/ tion leaves certain women vulnerable to IPV or cocaine (including crack cocaine)-involved (in that these women may be dependent on women, the paper compares substance their partner for money, shelter, protection, abuse and related behaviors among women or access to drugs), and the uses of mood in four alternative (current or most recent') altering substances by victims to self-medi­ relationship types: 1) those in which there is cate the deleterious emotional effects of vio­ no reported physical violence; 2) those in lence victimization (Duke 2002; Wu, EI­ which there is unidirectional violence di­ Bassel, Witte, Gilbert, & Chang 2003; rected against the woman; 3) those in which Romero-Daza, Weeks, & Singer 2003; Singer there is unidirectional violence by the woman 2006). However, intimate relationships are against her partner; and 4) those in which not uniform in terms of the direction of physi­ there is reciprocal or mutual violence. cal violence, particularly in regards to the re­ lationship between partner violence victim­ Methods ization and gender. In other words, although The study described here2 was imple­ women in heterosexual relationships are mented in the city boundaries of Hartford, more likely to be victims of partner violence CT. Hartford currently has a population esti­ than are men, some nonetheless take on mated at approximately 130,000 people, with 30 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Table 1: Binomial Physical Violence Scores and Intimate Partner Violence (IPV) Group Assignment IPV Group Ego-as-Victim Ego-as-Perpetrator Partner Violence Victim (PW) 1 0 Partner Violence Perpetrator (PVP) 0 Mutual Partner Violence (MPV) 1 Non-Abusive Relationsh ip (NAR) 0 0 an ethnic composition that is 45-50 percent in the informed consent process and then African American, 30-35 percent Hispanic were interviewed with the project instrument (primarily Puerto Rican) , and 20 percent battery. After the hour long interview, if deem­ White and other. Hartford is estimated to be ed appropriate by the interview coordinator the fourth poorest moderate-sized city in the and project coordinator, project staff made a U.S. , with high rates of unemployment, com­ voluntary referral for intervention services or, munity violence, drug abuse, and AIDS cases if needed, emergency services. (Himmelgreen & Singer 1998). Assignment of relationship type was deter­ Participants were recruited in areas of the mined by responses to the physical violence city known from past studies to have com­ prevalence items of the partner violence sub­ paratively high numbers of drug users, drug­ scale (Form N) of the Conflict Tactics (CT) related activities, and drug use/acquisition Scale developed by Straus (Straus 1979; sites. Outreach workers-who matched the Straus & Gelles 1990). The CT scale is a target population by gender, language, and widely used instrument for measuring intra­ ethnicity-walked through these areas and family conflict and violence. The subscale walked up to and engaged in conversation elicits the frequency, recency, and duration of with women encountered on the street. Usu­ specific minor (e .g., verbal abuse, pu shing, ally, conversations began with the offer of con­ shoving) and severe (e.g. , beatings, threats doms and led quickly to a brief description of or actual use of weapons) acts of violence the project. Potential participants were asked directed toward them by their domestic part­ a brief set of questions to determine their ner. For each item the participant was asked eligibility for the study. Women were deemed to identify prevalence, frequency, and sever­ eligible if they met the following criteria: ity of violence committed against her by her partner (ego-as-victim) . Cronbach's alpha for 1) between 18-58 years old ; the Violence subscale is .80. In the version 2) reported having used heroin or cocaine modified for this study, for each item partici­ during the previous 30 days; pants were also asked whether they had 3) reported not being in drug treatment (in ­ committed those acts of violence against cluding detoxification and self help pro­ their partner (ego-as perpetrator). grams) during the last 30 days. In order to describe patterns of intimate partner violence among drug using women, Candidates were excluded from partici­ data from the current and recent relationships pating: 1) if project staff concluded-based were combined into a single category. The on their observations-that a woman was first step of the analysis was descriptive. Par­ unable to comprehend the informed consent ticipants were described regarding selected process (because she offered inappropri­ socio demographic characteristics (e.g. age, ate responses to consent questions); or 2) if ethnicity), substance abuse history and types the candidate participant made verbal threats of their IPV relationship. In the second step, or actually engaged in violent behavior (nei­ ch i square tests were undertaken to address ther of which occurred). differences in terms of lifetime history of vio­ Woman interested in participating who lence and also address the group differ­ met the inclusion criteria and were not elimi­ ences in term of the drug using behaviors of nated by the exclusion criteria were given an respondents' of currenUmost recent partner. appointment to be interviewed at the offices A two tailed alpha of p<0.05 was considered of the Hispanic Health Council. At the time of statistically signifi cant in bivariate analysis. appointment, women were again screened All data was analyzed using SPSS software using the inclusion and exclusion criteria , version 10 .0. and , if accepted into the study, participated Based upon their responses to the physi- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 31

Table 2: Relationship Group Distribution for Last Three Relationships Relationship Current/Recent 2nd Most Recent 3rd Most Recent Type Relationship Relationship Relationship (N=499) (N=440) (N=250)* PW 10.7 18.9 14.8 PVP 08.5 03.9 04.8 M~ ~ . 0 ~ . 5 ~A NAR 41 .9 41.8 40.0 *Percentages may not sum to 100% due to rounding. cal violence items3, two binomial scores ences between racial/ethnic groups in terms were then calculated (O=answered negatively of IPV relationship patterns were not statisti­ to all of the physical violence items; 1=an­ cally significant. The average age of the swered affirmatively to one or more physical women in our sample was 37.8 years (stand­ violence items) for both the ego-as-victim and ard deviation = 8.0) . There was no signifi­ ego-as-perpetrator scales. A score of 1 in cant relationship between age, race/ethnicity, the former scale and 0 in the latter assigned or education and IPV group membership. Of that participant to the Partner Violence Victim the 497 respondents, 279 (56.1 %) were in (PW) category, while a reverse score (ego romantic relationships at the time of their in­ suffered no physical violence victimization, terview5. Ten percent of respondents report­ but engaged in physical violence against her ed that their current or most recent relation­ partner) placed her in the Partner Violence ship was with a ·woman, while 0.2 percent Perpetrator (PVP) group. Respondents with reported that their current/most recent part­ a score of 1 for both ego-as-victim and ego­ ner was transgendered. However, there was as-perpetrator scales were assigned to the no significant group difference between het­ Mutual Partner Violence (MPV) group, while erosexual and same-sex partners, nor het­ those scoring 0 for both scales constituted erosexuals and transgendered partners in the Non-Abusive Relationship (NAR) group. terms of their assignment in one of the four This categorization scheme was utilized for partner violence groups. participants' current or most recent relation­ Turning to the four IPV groups, 41 .9 per­ ships, as well as their second and third most cent (n=208) reported that there were no in­ recent relationships. The assignment of re­ cidents of physical violence between them­ spondents to each of the four groups is sum­ selves and their current or most recent part­ marized in Table 1. ner, and were thus assigned to the Non-Abu­ There are certain limitations to this ana­ sive Relationship group (NAR). In contrast, lytical strategy, particularly in terms of recip­ the PVV or Partner Violence Victim group rocal violence. For example, the survey data (those who reported being physically abused do not measure the sequence of violence by their partner but not vice versa) comprised (i.e. , whether the participant or her partner 10.7 percent of the sample (n= 53). A some­ initiated acts of violence) nor whether one what smaller percentage (8.5%, n=42) re­ member of the romantic dyad engaged in ported physically abusing their current or partner violence as a form of self-defense most recent partner, although not vice versa against the other. Nonetheless, a focus on (Partner Violence Perpetrator group or PVP). each of these four types of relationships Finally, a full 39 percent (n= 194) reported mu­ yields clear differences in terms of history of tual physical violence in their current/most childhood sexual violence, drug use pat­ recent relationship (Mutual Partner Violence terns, economic strategies, relationship dy­ group or MPV) (see Table 2). namics and social support. Because the physical violence scale en­ compasses a wide range of injury, it is im­ Research Sample portant to distinguish between relationships In terms of the racial/ethnic distribution of solely consisting of moderate violence and the sample, 38.6 percent were African Ameri­ those which include reported incidents of can, 39.4 percent were Hispanic/Latino, and what we term "severe intimate partner vio­ 17.4 percent were Euro-American, which lence" (SIPV), which we define as those roughly corresponds to the demographic where beatings, stabbings and/or shootings composition of Hartford as a whole•. Differ- reportedly occurred. Among the PVV group, 32 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Table 3: Distribution of IPV Group Cate­ unclear from the survey data whether re­ gories in the Second Most Recent Relation­ duced likelihood of non-reciprocal violence ship by Current/Most Recent Relationship victimization and increased likelihood of uni­ Group Membership directional violence perpetration in respond­ IPV Group 2nd Most Recent Relationship ents' current/most recent relationships is a (N=438) PW PVP MPV NAR product of response bias (i. e., respondents PW (n=48) 29.2 2.1 31.3 45. 7 may be less likely to characterize their cur­ PVP (n=38) 13.2 13.2 42 .1 31.6 rent partner as abusive or themselves as MPV (n=179) 16.8 4.5 36.9 41. 9 victims) or whether prior abuse for some NAR (n=1 73) 19.7 1.2 33. 5 45.7 woman may result in exercising control over Bold indicates remaining in same relationship their current or most recent partner. type. Percentages may not sum to 100% due to However, the likelihood of remaining in rounding . the same category from one relationship to the next is not as apparent as the above fig­ nearly half (45.5%, n=55) were victims of SIPV, ures may suggest, since there is a notable while 30 .2 percent (n=43) of those in the PVP degree of fluctuation between relationship group initiated severe forms of violence category membership between respond­ against their partners. Among the Mutual ents' current/most recent relationship and Partner Violence group, 35 .0 percent of re­ their second most recent relationship (see spondents were victims of SIPV, while 40.3 Table 3) . For current/most recent PVVs, for percent engaged in SIPV against their part­ example, the highest percentage (37.5%) of ner, meaning that in over five percent of MPV group membership in their second most re­ relationships the severity of self-perceived cent relationship was in the Non-Abusive respondent violence perpetration was group, a pattern also found in the MPV group greater than the severity of their victimization. (41.9%). It is worth noting that, although the largest percentage of the Non-Abusive Re­ History of Violence lationship (NAR) group remained in the We encountered significant differences same relationship category between their between participants in the four relationship current/most recent and second most recent groups in terms of lifetime history of violence. relationships, the majority of their previous For example, there are significant group dif­ relationships contained incidents of IPV ferences (x 2=.001) in terms of being victims (45.7%). of sexual abuse prior to age 18 (n=121). Re­ spondents in the Partner Violence Perpetra­ Survival Strategies and Health Status tor (PVP) group were much more likely to There was a significant group difference report victimization (45.2%) than those whose in terms of receiving money from "hustling," current or most recent relationship was non­ a proxy for both legal (e.g., panhandling, abusive (NAR) (17.7%). Those in the victim bottle collecting) and illegal (e .g., theft, drug (PVV) group and mutual partner violence sales, commercial sex work) money-mak­ (MVP) group were in between, at 28.3 per­ ing strategies within the informal economy cent and 27.4 percent, respectively. (x 2=.002). The highest percentage of women In terms of prior relationship history, re­ who utilized this economic strategy were in spondent distribution in the four relationship the MPV group (65 .2%), followed by the PVP categories is generally consistent between group (61 .9%) . In contrast, this strategy was cu rrent/most recent relationships and par­ utilized by less than half (46.3%) of the Non­ ticipants' second and third most recent rela­ Abusive Relationship (NAR) group and 55.8 tionships (see Table 2). However, there is a percent of the PW group. In response to the somewhat lower likelihood of PVV group question , "Have you ever given sex for drugs membership in current/most recent relation­ or a place to stay?", there was a moderate ships (10.7%), than in the second and th ird group difference (x 2=.044), with those in the most recent relationships (18.9% and 14.8%, PVV group being more likely to have done so respectively). In contrast, respondents are (53.8%) than those in the NAR group, who much more likely to be in the PVP group in were the least likely (38 .8%). Positive re­ their current/most recent relationship, than sponses for the MPV and PVP groups were they were during their second (3.9%) and 51 .5 percent and 42.9 percent, respectively. third (4.8%) most recent re lationships . It is The distribution of responses was some- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 33

Table 4: Conflicts Between Romantic Partners Over Splitting Drugs (In Percents) Taking the other's drugs Taking more than "fair share" without asking of split drugs Partner blamed Ego blamed Partner accused Ego accused IPV Group ego (N=327)** partner (N=329)** ego (N=328)** partner (N=327)** PW 38.2 25.7 57 .1 51 .4 PVP 22.2 29.6 37.0 55.6 MPV 41 .8 43.2 55.2 54.9 NAR 13.3 09.1 23.1 17.4 **p<.01

what different in response to the question, percent, respectively. "Have you ever given sex for money?" While Contrary to expectations, those in relation­ 67.9 percent of the PW group had engaged ships with a drug using partner in which there in sex work for money, the lowest percent­ is no partner violence (NAR) were much less age of positive responses were from those likely to assist their partners in securing in the PVP group (47.6%). Of the MPV group drugs than those in other types of relation­ 63.9 percent and 50.7 percent of the NAR ships. Also somewhat surprising is that re­ group had carried out sex work (x 2=.011 ). spondents who have been the victims of non­ There was no significant group difference reciprocated violence (PVV group) were not in terms of self-report for most of the thirteen the most likely to engage in such activities. illnesses associated with drug use, includ­ For example, in response to the question ing, Hepatitis B, Hepatitis C, and HIV. How­ "Did you ever sell drugs in order to get drugs ever, there were moderately significant group for this partner?" (N=328), only 11 .6 percent differences (x 2=.045) in terms of whether of those in non-violent relationships an­ respondents had ever been diagnosed with swered affirmatively, as compared to 28.6 a sexually transmitted disease (STD) (MPV= percent of those in the PVV group (MPV= 24%; PVP=19%; PW=17%; NAR=13%) or a 32.4%, PVP=29.6%; x 2=.001). Likewise, only mental illness (MPV=50%; PVP=45.2%; 11 .8 percent of the NAR group reported that NAR=37.6%; PW=32.1%). As these figures they ever sold sex for money or drugs in or­ indicate, those in relationships in which re­ der to get drugs for their partner, while 22.9 ciprocal violence occurs were more likely to percent of PW group did so (PVP=25.9%; have been diagnosed with an STD or a men­ MPV=25.5%; x 2=037). In both instances, a tal illness than their counterparts in the other significant percentage of participants in rela­ groups. Although the survey did not include tionships in which mutual violence had oc­ questions regarding types of mental illness, curred (MPV group) participated in these ac­ the latter is of particular interest in that there tivities on behalf of her partner. Even more is a sharp divide between those respondents surprising, a significant percentage of those who engage in IPV (either mutually or uni­ who had engaged in non-reciprocated vio­ directionally) and those who do not. lence against their partner (PVP group) sold sex or drugs in order to secure drugs for their Partners' Substance Abuse and IPV partners. However, when the partner pres­ The data yielded a moderate group differ­ sures the respondent, the distribution is no­ ence in terms of the illicit drug using behav­ tably different. Drug using partners in the iors (not including marijuana) of respond­ nonreciprocal violence victims group, for ex­ ents' currenUmost recent partner. Of 4 73 re­ ample, were significantly more likely than spondents who answered the survey ques­ those in the other groups (20%) to have in­ tion regarding whether their partner had ever sisted that the participant boost or steal in used illicit drugs during the relationship, 329 order to get drugs for him/her (N=328; MPV= (69.6%) responded affirmatively (x 2=.015). 15.9%; NAR=4.1 %; PVP=3.7%; (x 2=.003). Those in the Mutual Partner Violence group Among women whose romantic partners were much more likely to have a drug using used heroin and/or cocaine (including crack partner (77.2%) than were those Non-Abu­ cocaine), there was no significant group dif­ sive Relationship group (62.1 %). Partner vio­ ference in terms of whether these partners lence victims (PW) and perpetrators (PVP) split drugs with them. Despite this, tensions were in between, at 68.6 percent and 71 .1 surrounding the splitting and sharing of 34 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Table 5: Mean Scores of Social Support and Action Towards Leaving the Relationship Social Support Action Towards Leaving (N=497)* (N=297)** Standard Standard IPV Group Mean Deviation Mean Deviation PW 2.53 0.39 2.61 0.88 PVP 2.69 0.43 2.1 4 0.77 MPV 2.68 0.45 2.36 0.76 NAR 2.75 0.43 1. 95 0.68 *p<0.05; **p<0.01

drugs seem to provide a nexus of physical Thus, for women in abusive relationships, conflict. In response to the question, "Has the mediating factor that is most likely to pro­ this partner ever blamed you for taking his/ vide emotional and material support can be her drugs without asking him/her?", for ex­ difficult to maintain (EI-Bassel et al 2003). ample, those in the MPV and PVV groups For a drug involved woman, accessing net­ were much more likely to have responded works of social support entails particular affirmatively (41.8% and 38 .2%) than those challenges, since she may be dependent in the remaining two groups (see Table 4) . on her partner for money, alcohol, or drugs, Likewise, those in the Mutual Partner Vio­ and her substance misuse may have alien­ lence group were more likely to blame their ated her from friends and family. Often her partner for taking their drugs without asking circle of friends consists of other substance (43. 2%) than th ose in the other groups. For abusers, who may not able to be provide the each of the four sharing conflict items, the level of material and emotional support that percenta ges of affirmative responses in the she needs. NAR group were significantly lower than The Social Support Behavior Scale (SS- those in the groups where IPV has occurred. 8) (Vaux, Riedal , & Stewart 1987) was used This strongly suggests a positive associa­ to measure the extent to which participants tion between conflicts over drug sharing and had access to supportive networks. Using a partner violence history within that relation­ four-point Likert scale participants were ship, whether directed towards ego, ego's asked to use past experience to indicate the partner, or both. likelihood that a relative or friend would per­ form specific supportive activities. The scale Social Support and Action Toward Leaving taps emotional support, socializing levels, the Relationship practical assistance, financial assistance, For people in abusive relationships, ac­ and provision of advice, and has an internal cess to social support networks can play a consistency of .90. In order to assess ac­ critical role in moderating the negative ef­ tions taken by participants to end their rela­ fects to well-being that result from IPV expo­ tionships, we used the Action Toward Leav­ sure, as well as providing the emotional and ing (ATL) scale, a 14-item measure of termi­ material resources to leave that relationship. nation strategies developed by Wilmot and However, women in abusive relationships colleagues (1985). Participants were asked may feel a high degree of anxiety, embarrass­ to indicate, via a four-point Likert scale, the ment, or other forms of reluctance, in asking frequency of use of three factor-analyzed cat­ members of their social circle for help egories of tactics to terminate the relation­ (Choice & Lamke 1999). Furthermore, abus­ ship: verbal directness, verbal ind irectness , ing partners frequently exert considerable ef­ and nonverbal withdrawal (Cronbach's al­ fort to keep their partners socially isolated pha=.94). ATL thus measures communica­ from family and friends, in order to increase tive acts engaged in by the respondent to their dependence (Avni 1991 ; Mitchell & emotionally and socially disengage from the Hodson 1983; Hilberman & Munson 1977- relationship. The scale was used only for 1978). Indeed, as Tan and colleagues have participants who were in a romantic relation­ noted (1995), increases in the rate of physi­ ship at the time of their interview. cal violence by abusive partners are associ­ As shown in Table 5, availability of social ated with increased withdrawal from social support is strongly associated with relation­ support networks on the part of the victim. ship type, with those in the Partner Violence Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 35

Victim group scoring lower than those in the NAR group participants shared relatively other groups, while those in Non-Abusive low levels of participation in the informal Relationships scored highest. Action toward economy (apart from their role as consum­ leaving the relationship was moderately as­ ers of illicit drugs) as compared with women sociated with relationship type, with those in in the remaining relationship groups. For ex­ the NAR group scoring lowest and those in ample, NAR group participants were signifi­ the PVV group highest. This distribution indi­ cantly less likely to engage in "hustling" or to cates that, like those in physically abusive have ever sold sex than their counterparts in relationships in the general population, sub­ the other three groups. Interestingly, this ten­ stance-involved women who are victims of dency to keep the informal economy at arm's IPV face patterns of systematic estrangement length also extends to selling sex or drugs in from their social network, quite apart from order to secure drugs for their partner. Thus, the loosening of social bonds resulting from contrary to the expectation that drug-involved, their addiction. The fact that the PW group non-violent romantic partners would provide scored significantly higher on the ATL scale mutual support in helping the other secure is particularly intriguing, because it indicates drugs, participants in this relationship group that women in abusive relationships are not had little involvement in their partners' drug merely passive victims. Rather, they practice procurement. Likewise, the NAR group was what might be termed "everyday forms of re­ the least likely to have conflicts over the split­ sistance" (Scott 1985) within the relation­ ting and sharing of drugs, indicating that the ship, using avoidance, emotional withdrawal, tensions of procuring, splitting, and sharing and other tactics to, if not leave the relation­ are important features of violent conflict. Drug ship, then at least create a degree of subjec­ splitting occasions can be tense because tive autonomy within it. they often occur when users are experienc­ ing drug craving and have the cure for their Discussion problem at hand. Couples that either avoid The analysis presented above outlines a sharing or share drugs without conflict tend number of group differences between the to avoid partner violence. four relationship types. However, while the Partner Violence Victims (PVV) Group­ causes of some of these differences are fairly Comprising nearly 11 percent of the research intuitive, others are less so. The latter is due sample, members in this group were more in no small measure to the fact that respond­ likely to report severe intimate partner vio­ ents have belonged to different relationship lence (which involves being beaten, stabbed , groups throughout the life course, and thus or shot) than those in the Mutual Partner Vio­ it is likely that their attitudes and behaviors in lence group: 45.5 percent vs. 35.0 percent. the current/most recent relationship are in­ The association between being a non-recip­ fluenced by past experiences. In this sec­ rocated victim of partner violence and vio­ tion, we will discuss the findings for each of lence severity undoubtedly stems from the the groups in turn. lack of physical sanctions faced by the perpe­ Non-Abusive Relationship (NAR) Group­ trator (or, conversely, awareness of a part­ Despite the fact that there is significant move­ ner's proclivity for extreme violence may intim­ ment across relationship groups from one idate a woman from even attempting defen­ relationship type to the next, the NAR group sive violence). However, relationships in was by far the largest group for each of the which there is non-reciprocated violence are three sequential relationships examined for also quite different-interpersonally, psycho­ each participant (current/most recent; sec­ logically, and in terms of power relations­ ond most recent; third most recent). This dis­ than those in which mutual violence occurs. tribution indicates that, contrary to popular Johnson and Ferraro (2000) refer to system­ stereotypes, not all drug involved women are atic, unidirectional domestic violence against condemned to a life of violence and abuse at women by their partners (measured in terms the hand of their romantic partners, but are of frequency, severity, recency, and duration) capable of participating in stable romantic as intimate terrorism. Intimate terrorism is re lationships. Even when a woman's part­ grounded in one partner's motivations for ner is also a drug user, the relationship is power and control over the other partner. The not necessarily unstable or prone to violence psychosocial dynamics of intimate terrorism (Simmons & Singer 2006). are therefore distinct from relationships in 36 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology which each partner engages in violent be­ to have accused their substance using part­ havior against the other, since the former ner of taking more than their fair share of consists of asymmetrical physical-and con­ drugs that they have split suggests that ac­ sequently behavioral and psychological­ cusation involving drug sharing is a poten­ control over a partner. This sense of control tial trigger of violence. is reflected in the fact that members of this Mutual Partner Violence (MPV) Group­ group are much more likely to have a drug The largest of the groups in which partner using partner insist that she boost or steal violence occurred , a key feature of MPV re­ in order to secure drugs than those in the spondents is that they were much more likely other groups. Thus, over the long run inti­ to have drug using partners than were those mate terrorism tends to produce victim de­ in the other three groups. This strongly sug­ pression and learned helplessness (Walker gests the significant role that the pain and 1984). tension of drug withdrawal plays in trigger­ Not altogether surprising was the fact that ing episodes of mutual partner violence. the PVV group had the lowest measure of MPV participants with drug using partners social support, reflecting the sometimes engaged in significant acts of mutual sup­ considerable effort on the part of batterers to port. For example, respondents in this group keep their partners socially isolated from fam­ were the most likely to sell drugs in order to ily and friends, in order to increase their de­ obtain drugs for their partner. In addition, a pendence (Avni 1991 ; Mitchell & Hodson nearly equal percentage with those in the 1983; Hilberman & Munson 1977-1 978; MPV group sold sex for money or drugs in Dobash, Dobash, & Cavanagh 1985; Tan et order to obtain drugs for their partner. From al 1995). As stated above, it is particularly this standpoint, MPV participants place them­ noteworthy that PVV respondents were much selves at considerable risk (of street vio­ more likely to have taken action toward leav­ lence, of possible arrest, of disease expo­ ing their relationship than those in the other sure) in order to secure drugs for their part­ groups, indicating that these women con­ ners, regardless of whether they are also tinue to exercise a significant degree of engaging in these activities in order to ob­ agency within the confines of an abusive re­ tain drugs for themselves. lationship, parti cularly in terms of seeking However, the strain of addiction places ways to extricate themselves from that rela­ unique strains on the relationships of drug tionship (Choice & Lamke 1999). using partners which can , in turn , lead to mu­ Partner Violence Perpetrator (PVP) tual acts of violence. As noted above, the axis Group-The smallest of the four relationship of tension-and, potentially, of violence-in groups (8.5%) , PVP respondents were the these relationships seems to revolve around least likely to have been in a nonviolent rela­ the splitting and sharing of drugs. Of the four tionship in their prior relationship. Th is sug­ relati onship categories, for example, the MPV gests that their response to prior partner vio­ group was more likely to have blamed their lence-whether as victim, perpetrator, or in partners, or to have been blamed by their a mutually violent relationship-is to take on partners, for taking the other's drugs without the role of batterer in the current or most re­ asking. In addition, they were also the most cent relationship. Interestingly, despite the likely to have been accused by their partner fact that members of this group were the least of having taken more than their "fair share" of likely to have ever engaged in sex work for drugs that they had split. Thus, the grinding money, they were the most likely to have sold pursuit of cash to purchase drugs, coupled sex for money or drugs specifically in order with the tension of providing mutual support to secure drugs for their partners. This ap­ with in the relationship while also obtaining parent altruism in terms of providing drugs a sufficient quantity of drugs for each part­ for their partner seems to contradict their role ner, ca n lead to significant levels of stress, of batterer in the relationship, although it is and at least potentially to physical confronta­ likely that engaging in these activities would tion . result in resentment, wh ich may on some occasions result in violence, particularly Conclusion against partners who, apparently, are not Female drug users are highly diverse in likely to reciprocate with violence of their own. terms of their romantic relationship patterns, Furthermore, the fact that they are most likely particularly in terms of physical intimate part- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 37 ner violence. Contrary to popular stereotypes (Weeks, Grier, Romero-Daza, Puglisi, & of drug users as socially isolated or as Singer 1998). It is thus critically important people who would take advantage of anyone that partner violence prevention and advo­ to get drugs, our findings show that drug cacy accompany AIDS prevention and drug users are capable of fully participating in ro­ abuse intervention programs, in order to re­ mantic relationships, even when both mem­ duce violence in the lives of their clients. bers of the romantic dyad are substance in­ volved. Furthermore, intimate violence is not ENDNOTES an inherent feature in these relationships, 'current and most recent relationships were co l­ as evidenced by the fact that the Non-Abu­ lapsed into a single category since there was sive Relationship group was the largest of no significant group difference between these the four in terms of participants' current or two groups in terms of their demographic char­ acteristics, drug use patterns, and distributions most recent relationship. Further research in the four IPV relationship categories. is therefore necessary in understanding 2 This study was funded by the National Institute these relationships. In particular, research on Drug Abuse, Merrill Singer, Principallnvesti- should focus on indigenous forms of conflict 3 gator. resolution and mutual support among sub­ Although we recognize verbal abuse as a form stance involved romantic partners, in order of violence, we exclude the verbal abuse items to understand better the ways in which po­ of the Partner Violence scale from the current tentially violent situations are avoided or dif­ analysis since, in the absence of context, acts fused within those relationships. are ambiguous in terms of whether they con­ However, the fact that drug involved stitute abuse or result from extenuating cir­ cumstances (misunderstandings, etc.) Further­ women are much more likely to be victims of more, the entire sample of respondents who intimate partner violence throughout the life reported physical violence victimization and/or course than US women as a whole also in­ perpetration also reported positively to the ve r- dexes the importance of designing violence 4 ba l abuse items. prevention and protection programs that take The remaining 4.6 percent of respondents were addiction status into account. Few programs 5 re-calculated as Other. exist, for example, that offer stress and an­ Respondents were allowed use their own crite­ ger management or self defense for women ria of what constitutes a romantic relationship, in addiction. Even more glaring is the lack of provided that the relationship in question had lasted at least two weeks. access to safe, anonymous, and well-pro­ tected shelter (e.g. "battered women's shel­ REFERENCES ters") for women in addiction, the latter of Amaro H, LE Fried, H Cabral, & B Zuckerman. 1990. which is a fundamental resource for those Violence during pregnancy and substance use. wishing to leave an abusive relationship. Be­ Amer J Public Health 80 5 575-579. cause of multiple liability, childcare, security, Avni N. 1991 . Battered wives: the home as a total and logistical concerns, substance abusing in stitution. Violence Victims Summer 6 2 137- victims of partner violence are barred from 49. admission to these facilities. It is therefore Bennett LW. 1995. Substance abuse and the do­ imperative that research-based harm reduc­ mestic assault of women. Social Work 40 760- 772. tion strategies be initiated to address part­ Caetano R, CB Cunradi, CL Clark, & J Schafer. ner violence in all its manifestations among 2000. Intimate partner violence and drinking this vulnerable population. patterns among wh ite , black, and Hispanic Our findings also have relevance for un­ couples in the US. J Substance Abuse 11 2 derstanding drug use, commercial sex, and 123-138. AIDS risk as reflecting far more than individ­ Choice P & L Lamke. 1999. Stay/leave decision­ ual choice or morality. In the case of relation­ making processes in abusive dating relation­ ships involving violence victimization, but in ship. Personal Relationships 6 351-367. relationships where violence among part­ Cunradi CB, R Caetano, CL Clark, & J Schafer. 1999. Alcohol-related problems and intimate ners is mutual as well, interpersonal violence partner violence among white, black, and His­ may be an important force driving individual panic couples in the US. Alcoholism, Clinical behavior. Women who are victimized by part­ and Experimental Res 23 9 1492-1501 . ner violence, and then face social opprobrium Cunradi CB, R Caetano, & J Schafer. 2002 . Alco­ for self-medicating drug use or for engaging hol-related problems, drug use, and male inti­ in risky behavior, are doubly victimized mate partner violence severity among US 38 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

couples. Alcoholism. Clinical and Experimen­ Scott J. 1985. Weapons of the Weak. Everyday tal Res 26 4 493-500. Forms of Peasant Resistance. Yale U Press. Dobash R, R Dobash, & K Cavanagh. 1985. The Sharps PW, J Campbell, D Campbell, F Gary, & D contact between battered women and the social Webster. 2001 . The role of alcohol use in inti­ and medical agencies. In Private Violence and mate partner femicide. Amer J Addictions 10 2 Public Policy: The Needs of Battered Women 12-135. and The Response of the Public Services. J Simmons J & M Singer. 2006. I love you ... and Pahl, ed. London: Routledge. heroin: care and collusion among drug-using Duke M. 2002. Establishing emergency shelter ser­ couples. Substance Abuse Treatment Preven­ vices for substance abusing victims of domes­ tion Policy 1 7. tic violence: structural, political, and cultural Singer M. 2006. The Face of Social Suffering: Life barriers. Society for Applied Anthro Ne wslet­ History of a Street Drug Addict. Prospect ter 13 3 August. Heights, IL: Waveland Press. EI-Bassel N, L Gilbert, S Witte, E Wu , T Gaeta, R Straus M. 1979. Measuring intrafamily conflict and Schilling , & T Wad a. 2003. Intimate partner vio­ violence: the conflict tactics (CT) scales. J Mar­ lence and substance abuse among minority riage Family 41 75-88. women receiving care from an inner-city emer­ Straus MA & RJ Gelles. 1990. Physical Violence in gency department. Womens Health Issues American Families: Risk Factors and Adap­ Jan-Feb 13 1 16-22. tations to Violence in 8, 145 Families. New Gilbert L, N EI-Bassel , V Rajah , J Fontdevila, A Brunswick, NJ : Transaction. Foleno, & V Frye. 2000. The converging epi­ Tan C, J Basta , C Sullivan, & W Davidson. 1995. demics of drug use, HIV and partner violence: The role of social support in the lives of women a conundrum for methadone maintenance treat­ exiting domestic violence shelters. J Interper­ ment. Mt. Sinai J Medicine 67 452-464. sonal Violence 10 4 437-451 . Goldberg M. 1995. Substance-abusing women: Tjaden P, N Thoennes, & CJAIIison. 1999. Compar­ false stereotypes and real needs. Social Work ing violence over the life span in samples of 40 6 789-798. same-sex and opposite-sex cohabitants. Vio­ Hilberman E & K Munson. 1977-8. Sixty battered lence Victims Winter 14 4 413-25. women . Victimology: Internal J 2 460-470. Vaux A, S Riedal, & D Stewart. 1987. Modes of Himmelgreen D & M Singer. 1998. HIV, AIDS and social support: the social support behavior (SS­ other risks : findings from a multisite study. Amer B) scale. Amer J Community Psycho/ 15 209- J Drug Alcohol Abuse 24 2 187-197. 237. Johnson MP & KJ Ferraro. 2000. Research on do­ Walker L. 1984. The Battered Woman Syndrome. mestic violence in the 1990s: making distinc­ NY: Springer. tions. J Marriage Family 62 948-963. Weeks M, M Grier, N Romero-Daza, M Puglisi, & M Lead ley K, CL Clark, & R Caetano. 2002. Couple's Singer. 1998. Streets, drugs, and the economy drinking patterns, intimate partner violence, and of sex in the age of AIDS . Women Health 27 1/ alcohol-related partnership problems. J Sub­ 2 205-228. stance Abuse 11 3 253-263. Wilmot W, D Carbaugh, & A Baxter. 1985. Commu­ Lown AE & WA Vega. 2001. Alcohol abuse or de­ nication strategies used to terminate romantic pendence among Mexican American women relationships. Western J Speech Communi­ who report violence. Alcoholism, Clinical and cation 49 204-216. Experimental Res 25 10 1479-1486. Wu E, N EI-Bassel, SS Witte, L Gilbert, & M Chang . Mitchell R & C Hodson. 1983. Coping with domes­ 2003. Intimate partner violence and HIV risk tic violence : social support and psychological among urban minority women in primary health health among battered women . Amer J Com­ care settings. AIDS and Behavior Sep 7 3 munity Psycho/ 11 629-65. 291 -301 . Romero-Daza N, M Weeks, & M Singer. 2003. No­ body gives a damn if I live or die: violence, drugs, and street-level prostitution in inner-city Hartford , Connecticut. Medical Anthro 22 233- 259. Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 39 A DOSE OF DRUGS, A TOUCH OF VIOLENCE, A CASE OF AIDS, PART 2: FURTHER CONCEPTUALIZING THE SAVA SYNDEMIC

Merrill Singer, Hispanic Health Council

ABSTRACT

This paper builds on prior discussion of the concept of syndemics in the social science and public health literatures to further define and extend the utility of this construct in analyzing the relationships among substance abuse, violence and HIV/AIDS. The term syndemic refers to a set of closely entwined and mutual enhancing health problems that "working together" in a context of noxious social and phys ical conditions can significantly affect the overall disease burden and health status of a popul ation. The paper focuses on the emergence of wh at is termed the SAVA (s ubstance abuse, violence, AIDS) syndemic among several populations that are at high risk because they are subj ect to social di scrimination, stigmatization, and subordination, namely abused children and battered women, men who have sex with men, illicit drug users, and commercial sex workers.

In the course of a multi-year, 1988-pres­ Greek word demos ("the people," but in a ent, program on HIV risk prevention research health context refers to a disease that is dif­ among inner city drug users in Hartford, CT­ fusing in a population), while the prefix is a community-based program that has in­ taken from the Greek term for "working to­ cluded a number of federally funded studies gether." of drug use patterns among several different In short, a syndemic is a set of closely populations of drug users-it became ap­ entwined and mutual enhancing health prob­ parent that studying HIV in isolation of other lems that "work together" in a context of nox­ diseases and conditions also prevalent in ious social and physical conditions, can sig­ these populations was a distortion. Indeed, nificantly affect the overall disease burden it became clear that even the term epidemic and health status of a population. For ex­ does not sufficiently describe the contempo­ ample in January, 2004, the World Health rary US inner city health AIDS crisis. A crisis Organization announced a decision to sup­ that may be characterized by the spread of port expanded collaboration between tuber­ AIDS that is closely associated with a set of culosis and HIV/AIDS programs to curb the other endemic and epidemic conditions (e.g., growing spread of TB/HIV co-infections. The TB, STDs, hepatitis, cirrhosis, infant mortal­ new WHO policy guidelines define the pub­ ity, drug abuse, suicide, homicide, etc.). lic health activities that were needed to ad­ These conditions are intertwined and dress what is now referred to as "the dual strongly influenced and sustained by a epidemic of TB and HIV." According to Lee broader set of societal factors. Factors that Jong-Wook, Director-General of the WHO, include high rates of unemployment, poverty, homelessness, residential overcrowding, TB/HIV is a deadly combination and needs substandard nutrition, infrastructural deterio­ to be tackled with an approach treating the ration and loss of quality housing stock, whole person. (World Health Organization forced geographic mobility, family disruption, 2004 1) attenuation of social support networks, health care inequality, racism, and domestic But this syndemic is just one of several in as well as street violence (Bourgois 1995; which HIV is a primary disease component. Stall & Purcell 2000; Wallace 1990; Recently, in our efforts to futher delineate Waterston 1993). the concept of syndemic, we have drawn at­ As a result, our Hispanic Health Council tention to the fact that disease interaction oc­ research team proposed the term "syndemic" curs at both the population and individual (Singer 1994, 1995) which refers to the inter­ levels (Singer & Clair 2003). At the popula­ related complex of health and social crises tion level, the term syndemic refers to two or facing the urban poor. Like the terms epi­ more epidemics interacting synergistically demic and pandemic (i .e., spreading health and contributing as a result to an excess dis­ problems of local or extra-local distribution), ease load in a population (Frumkin 2002; the suffix of syndemic is derived from the Homer & Milstein 2002; MacQueen 2002). 40 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

As Milstein observes, these other diseases. The added damage to the immune system produced by co-infec­ Syndemics occur when health-related prob­ tion with HPV and herpes facilitates the rapid lems cluster by person, place or time. The development of HIV infection and subsequent problems-along with the reasons for their devastating consequences. Other studies by clustering-define a syndemic and differen­ this research team on the effects of co-infec­ tiate one from another (though they may tion found that HIV+ women infected with have nested or overlapping re lationships). human herpesvirus type 8 (HHV-8) exhibited (2001 2) accelerated deterioration of their immuno­ logic and hematologic conditions when com­ As this definition suggests, at the population pared to HIV+ women coinfected with other level HIV maybe entwined with several differ­ sexually transmitted diseases (Pugliese, ent other diseases, each forming a some­ Torre, Saini, Pagliano, Gallo, Pistono, & Paggi what different distributional cluster (e.g., hepa­ 2002b). In other words, the important issue titis in one part of a geographically dispersed at the individual level is not just co-infection or socially segmented population and one or co-presence of two or more diseases, but or more sexually transmitted diseases in an­ the enhanced infection and physical conse­ other part of a population). In time, of course, quences due to disease interactions. these independent syndemics may merge Social context, including both the physi­ into what might be called a "superdemic." An cal conditions in which people live their lives, occurence that syndemic theory predicts is as well as the hierarchical structure of social most likely in populations that suffer multi­ relations and their consequences in every­ ple structural disadvantages and the result day life, is a critical component of the syn­ is in the interconnected breakdown of social demic concept. In contrast to traditional clini­ structures, social relationships, and immune cal approaches to conceptualizing disease, defenses. the types of social conditions that increase At the individual level, the term syndemic the likelihood that various diseases will be refers to the health consequences of the bio­ concentrated in a population and that certain logical interactions that occur when two or populations will be particularly vulnerable to more diseases or health conditions are co­ co-terminus diseases, are of central concern present in multiple individuals within a popu­ in syndemic research and public health/medi­ lation (Aicabes, Schoenbaum & Klein 1993; cal response. For example, researchers at Ensoli & Sirianni 2002; Farzio, Bueler, Cham­ Johns Hopkins University School of Public berland, Whyte, Sivanajan Froelicher, Hop­ Health followed a cohort of over 450 prima­ kins, Reed, Mokotoff, Cohn, Troxler, Phelps, rily Black (95.8%) male (76.3%) injection drug & Berkelman 1992). For example, in a sub­ users in Baltimore during the years between study of over 5,000 men that were enrolled in 1988 to 1999. At intake, participants were all the Multicenter AIDS Cohort Study between HIV+, but were asymptomatic. Two years af­ 1994 and 2000, Thio and co-workers (2002) ter enrollment in this study, 32 of the partici­ divided the sample into four groups: individu­ pants had progressed to an AIDS diagnosis, als with HIV only, those with hepatitis B, those for a cumulative incidence rate of 7.1 per­ with both infections, and those who were dis­ cent. Questionnaires completed by partici­ ease-free. They found that liver disease-re­ pants showed that 43.8 percent of those who lated death was highest in the dual infected had developed full-blown AIDS reported a subgroup and was especially high among high level of psychological distress at intake those with low CD4 cell counts, a sign of compared to 22 percent of individuals who advanced HIV infection. Men infected with did not convert to AIDS. Multiple regression hepatitis Band H IV were 17 times more likely analyses found that reporting distress in to die of liver disease than those infected one's life was associated with a significantly with just hepatitis B. Similarly, Pugliese and elevated risk (adjusted hazard ratio = 2.39) colleagues (2002a), have shown that HIV+ for the development of AIDS (Golub, Astem­ women who are also infected with both the borski, Hoover, Anthony, Vlahov, & Strathdee human papillomavirus (HPV), a cause of cer­ 2003). vical cancer, and herpes simplex virus (type In our own studies, we have found that 2) have a higher level immunodepression inner city injection drug users tend to experi­ than those who are not also co-infected with ence many risks and stressors (e .g., home- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 41 lessness, uncertain access to needed er 1996), the purpose of this paper is to con­ drugs, police harassment, threat of infection) tribute to the further conceptualization of but being at immediate and continual risk of syndemics generally and the SAVA syndemic violence (in various forms) is particularly dis­ specifically by examining expressions and tressing. Consequently, exposure to violence dimensions of SAVA in several populations (directly, as a victim, or even indirectly, as a that are subjected to social discrimination, witness to the victimization of others, espe­ stigmatization, and subordination, namely cially significant others) was identified early abused children and battered women, men in our work on syndemics as a condition of who have sex with men, illicit drug users, particular importance. Indeed, in our studies and commercial sex workers. of drug abuse, violence, and AIDS in Hartford over the last 17 years, it has become clear SAVA AND VICTIMS OF INTERPERSONAL that these three grave threats to health and VIOLENCE well-being are not really disconnected phe­ Domestic violence has become a grave nomena. As a result, we proposed the term concern in the societies of North America and SAVA (substance abuse, violence, and AIDS) elsewhere. Research that speaks to the role to label these as a syndemic comprised of of domestic violence in the SAVA syndemic three closely linked and interdependent includes both studies of childhood sexual health conditions that co-exist in the bodies and physical abuse, on the one hand, and and the social worlds of many low-income intimate partner violence, on the other. The individuals in urban environments (Singer headlines of newspapers emphasize grim 1996). Of course, the SAVA syndemic is not statistics: studies show that child abuse oc­ confined to Hartford. Rather, this widespread curs in 30-60 percent of family violence cases national and even global syndemic has taken that involve families with children (Carter, Wei­ a devastating toll on the lives of the urban thorn, & Behrman 1999). Approximately four poor in many countries. Barring significant million teenagers in the U.S. have been vic­ health and social interventions, SAVA threat­ tims of a serious physical assault, while nine ens to continue to wreck pain and havoc into million have been witnesses to severe vio­ the future. lence during their lifetimes (Kilpatrick & Some dimensions of the enmeshed rela­ Saunders 1997). Each year, 3-10 million chil­ tionship among substance abuse, violence dren in the U.S. witness domestic violence. and AIDS have been studied. These include In the literature on domestic violence, a the social conditions that led to direct and common explanatory theme is that ]violence indirect sharing of drug injection equipment begets violencei (Widom 1989). Usually, this and the consequent spread of AIDS; the role relationship is thought of in interpersonal of crack-cocaine in particularly demeaning terms. Thus, researchers have noted that low sex for drugs/money transactions that lead self-esteem, in conjunction with limited so­ to AIDS transmission; the contribution of turf­ cial support, is closely linked to violence vic­ wars and broken contractual agreements timization and the development of a life pat­ among drug sellers to drug-related violence; tern of revictimization (Sobo 1995). Prior life and the role of an AIDS diagnosis in enhanc­ history is a critical feature used in explana­ ing levels of drug use in some individuals for tions of domestic violence. Those who com­ some period of time. Other suspected con­ mit child abuse often have histories of hav­ nections are unclear, such as the frequency ing been abused as ch ildren themselves. of violence against women who proposed Consequently, interventions often have condom use by condom-resistant men; the ibreaking the cycle of violencei as their ob­ precise role of violence victimization in the jective. initiation and continuation of drug use as a While it is likely that the psychological in­ form of self-medication; the impact of vari­ juries of abuse find expression in respon­ ous expressions of structural violence on sive acts of violence, a narrow focus on inter­ AIDS risk behavior; differences in level of personal violence ignores another important drug withdrawal agitation and intense drug source of violence that many people, partic­ craving in interpersonal violence; and the con­ ularly the poor and working classes, people tribution of childhood sexual abuse on adult of color, and women and sexual minorities drug use and AIDS risk. Building on our ear­ endure, namely structural violence perpetu­ lier assessment of the SAVA syndemic (Sing- ated by the major institutions in society 42 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology against denigrated and subordinated popu­ lems). A number of studies of adult female lations. As Farmer explains, structural vio­ victims of child sexual abuse have found lence refers to higher rates of both alcohol and other drug abuse than women who were not sexually a host of offenses against human dignity abused or women in the general population [including): extreme and relative poverty, so­ (Briere & Runtz 1987; Brown & Anderson cial inequalities ranging from racism to gen­ 1991 ; Goodwin, Cheeves, & Connell 1990; der inequality, and the more spectacular Pribor & Dinwiddie 1992). Further, in a gen­ forms of violence that are uncontested hu­ eral population study, the Los Angeles Epi­ man rights abuses ... (2003 1) demiologic Catchment Area (ECA) survey (1983-1984), found that 6.8 percent of It is within this contexts of structural violence women participants reported that they were that the intersection of domestic violence, victims of "forced sexual contact" before the substance abuse, and AIDS risk is dispro­ age of 16. In this study there was a signifi­ portionately common . Structural violence, in cant association identified between report­ short, begats much interpersonal domestic ing such a history and the later development violence and the accompanying faciliators of alcohol and drug dependence (Burnam , and consequences of violence including Stein, Golding, Siegel, Sorenson, Forsythe, drug use and AIDS risk behavior. & Telles 1988; Scott 1992; Stein et al 1988). In recent years, there has been a prolifer­ Similarly, a national telephone survey (Kil­ ation of research on childhood sexual abuse, patrick, O'Neill , Beak, Resnick, Stugis, Best, with considerable attention paid to the life­ & Saunders 1990) found that childhood time consequences for victims. Research sexual assault was significantly associated findings show a relationship between child with current substance abuse among sexual abuse victimization and emotional women . disturbances like anxiety and depression, Intimate partner violence (IPV) is another sexual problems ranging from risky practices critical component of the SAVA syndemic. Es­ to sexual dysfunction, and substance-related timated rates of partner violence among illnesses like food disorders, alcohol abuse, women who use drugs are two or three times and drug abuse (Beitchman, Zucker, Hood, greater than in general population samples Dacosta, Akamn , & Cassavia 1992; Briere & of women (Bennett & Larson 1994; Brewer, Runtz 1987; Brown & Anderson 1991 ; Bush­ Fleming, Haggerty, & Catalano 1998). The nell, Wells, & Oakley-Browne 1992). Com­ relationship between partner violence, sub­ munity studies have shown that 7 to 33 per­ stance abuse and HIV risk is complex. On cent of adults report childhood sexual abuse the one hand, it can involve substance abuse (Russell 1983; Stein , Golding, Siegel, Bur­ by either perpetrators, or victims, or both, and , nam, & Sorenson 1988; Wyatt 1985). Women on the other hand, can involve clear cut vio­ who report that they were subjected to child­ lence victimization or reciprocal violence hood sexual abuse have been found con­ among partners (Amaro, Fried , Cabral, Zuck­ sistenly to be more likely than women in the erman 1990; EI-Bassel, Gilbert, Schilling, & general population to seek treatment for al­ Wada 2000; Gilbert, EI-Bassel, Rajah, Foleno, cohol and drug-related problems (Kovach Fontdevila, Frye , & Richman 2000). Where 1983; Miller, Downs, Gondoli, & Keil 1987; violence victimization is a factor, a dynamic Rohsenow, Corbett, & Devine 1988; Sterne, process can be activated, involving: a) a part­ Schaefer, & Evans 1983). For example, Miller ner, battering that is triggered by the and co-workers (1993) found that women in perpetrator's use of drugs; b) illicit drug use alcohol treatment reported significantly high­ by the victim to self-medicate the damaging er rates of childhood sexual abuse than ei­ emotional effects of violence victimization; ther women in the general population or and c) engaging in risky sexual and drug­ women without an alcohol-related problem related behaviors. While women in hetero­ receiving treatment for a mental health is­ sexual relationships are usually the victims sue. These researchers found that the asso­ of partner violence, this is not always the ciation between childhood sexual abuse and case. alcohol-related problems remained even af­ To examine the relationship between sub­ ter they controlled for sociodemographic and stance abuse, violence and HIV risk more genetic factors (e .g., parental alcohol prob- closely, EI-Bassel and co-workers (2000) in- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 43 terviewed 31 women in drug treatment who this behavior compared to 28.6 percent of reported physical or sexual violence com­ those women who were victims of intimate mitted by an intimate partner. Of those women partner violence. Similarly, only 11 .8 percent who recalled recent experiences of intimate of women who reported that they ever sold partner violence, almost all (83.8%) reported sex for money or drugs in order to get drugs drug use during the incident. In 40 percent of for their partner were in non-violent re lation­ these cases, both partners were using drugs, ships, while 22.9 percent of those who were while in 35 percent it was only the perpetra­ victims of partner violence did so (p<.01 ). tor who was using drugs. About a fifth of the In short, we found that drug-involved women (19.3%) indicated that they used women who were in abusive relationships drugs immediately after the violence had were significantly more likely to engage in ended as a way of dealing with their emotion­ risky behaviors raising drug money for their al upset and physical pain. Additionally, about partner than other women. This finding sug­ a fifth of the women reported that they had gests that one of the ways the SAVA syndemic been forced to have unprotected sex during unfolds in this population is that some drug­ the most recent incident of violence or just involved men use particularly severe forms after it ended. These researchers note of violence with their female partners. And these women, in turn, are more likely to put women in our sample attributed their expe­ themselves at risk for HIV or for street vio­ riences of abuse to their partneris drug use lence than are other women who use drugs. and to a lesser extent to their own drug use Severe intimate partner violence begets HIV OWomen in this study are at very high risk risk and risk for additional violence. of contracting HIV and HCV, fo r multiple rea­ sons. Only a minority of our sample have SAVA AMONG MSM ever used condoms with their partners al­ The highest absolute number of both new though a majority reported that they or their HIV infections and AIDS cases occur among partners have had outside re lationships. men who have sex with men (MSM). For the (Gilbert et al 2000 406) most part, studies of HIV risk among MSM have focused on sexual risk with compara­ In our own research on drug using women tively little attention given to the dual risk cat­ at the Hispanic Health CounciP , we found it egory of MSM drug users, men who are important to differentiate the women into one placed in harmis way both through drug use of four relationship groups. About two-fifths and sexual behavior. However, the CDC of the women (41 .9%) reported that there (2002) reports that increasing proportions of were no incidents of physical violence be­ HIV infections are occurring among men who tween themselves and their current or most report dual risks from both drug injection and recent sex partner. Ten percent of the women risky sex with men, especially for men of color. reported that they were the victims of partner In their comparison of sexual risk behaviors violence. Another 8.5 percent of the women among MSM who also inject drugs with MSM reported that they physically abused their cur­ who do not inject drugs, OiConnell and col­ rent or most recent partner, but they were not leagues (2004) found that the former are themselves victims of partner violence. Fi­ younger and more likely to be HIV-sero­ nally, almost 40 percent reported mutual positive than the latter. physical violence in their current or most re­ In Stueve and co-workersi study (2002) of cent relationship (Duke, Teng, Clair, Saleheen, 3,075 MSM aged 15-25 years, study partici­ Choice, & Singer 2006). pants were asked about th eir last sexual con­ Women who were subject to violence vic­ tact with primary and secondary partners, in­ timization by partners were more likely to re­ cluding whether they were ihigh on drugs or port suffering more severe forms of violence, alcoholi at the time. Almost one fifth (18.6%) including being beaten, stabbed, or shot than who reported having a pri mary partner that those women who were involved in mutual they used drugs during their last sexual en­ violence with their partners: 45.5 percent vs. counter, and 25 percent said they had anal 35.0 percent. In answering a question about sex without a condom. Among men without whether they had sold drugs in order to raise primary partners, 29.3 percent reported drug money for their partner, only 11 .6 percent of use during their last sexual episode, and only those in non-violent relationships reported 12.3 percent reported unprotected anal in- 44 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology tercouse. Using drugs was associated with to depression in adulthood, entrance into unprotected receptive anal intercourse with abusive adult relationships, the use of multi­ nonprimary partners (odds ratio = 1.66, p = ple drugs, and experience with high levels of .02). Some drugs like crystal methamphet­ HIV risk and infection. This suggests that the amine have played a particularly significant factors interact, are mutually reinforcing, and role in persistent high risk sex among MSM are best addressed in tandum rather than (Reback & Grella 1999; Shoptaw, Reback, & as separate threats to health. Freese 2002). Use of this drug, for example, Similarly, Relf and co-workers (2004) mea­ is common among men diagnosed with HIV sured the prevalence of battering victimiza­ and other sexually transmitted infections tion6which they defined as the experience (Bernstein, Tulloch , Montes, Golan, Dyer, of psychological/symbolic, physical, and sex­ Lawrence, Dodagoda, Rottblatt, Kerndt, ual battering6in the same sample of MSM Funn , DeAugustine, & Weismuller 2001). analyzed by Stall. They found that rates of Among AIDS cases in the U.S., currently battering were quite high compared to het­ eight percent fall into the dual risk category erosexual men and that HIV serostatus was of being an IOU and MSM . Unfortunately, HIV/ associated with being the victim of physical AIDS surveillance data do not provide infor­ and psychological/symbolic violence, but not mation on noninjecting drug use and HIV in­ sexual violence. Further they found that bat­ fection among MSM . However, our research tering victimization is the key mediating vari­ at the Hispanic Health Council with MSM in able between being subjected to childhood Hartford, Connecticut suggest that trading sexual abuse, having a gay identity, having sex for drugs and/or money or engaging in various adverse early life experiences, and high risk sexual practices as a result of drug subsequent HIV risk behaviors. In short, use is disproportionately common in this these studies among MSM suggest the im­ population (Singer & Marxuach-Rodriquez portance of a set of syndemic factors begin­ 1996; Clair & Singer 2004). ning with childhood exposure to abuse, later Several factors have been found to in­ exposure to intimate partner violence, par­ crease H IV risk among men who have sex ticular psychological reactions, drug use, and with men in the United States, including mul­ high-risk sexual behavior. tiple drug use, partner violence, childhood sexual abuse and depression (Carballo­ SAVA AND STREET DRUG USERS Dieguez & Dolezal 1995; Dilorio, Hartwell, & Tony, a participant in Hispanic Health Hansen 2002; Jinich, Paul, Stall, Acree, Council drug research (Singer 2006), ex­ Kegeles, Hoff, & Coates 1998; Relf, Huang, plained his most recent bout with violence, a Campbell, & Catania 2004; Stall, Mills, near-fatal revenge stabbing initiated by a drug Williamson, & Hart 2003). These factors are dealer that Tony had ibeati [stolen drugs thought to interact producing an increase from] as follows: both in drug-related risk and high-risk sexual behaviors (Barthalow, Doll , Joy, Douglas, When I was walking down the street, wait­ Bolan , Harrison, Moss, & McKirnan 1994; ing for her [h is girlfriend] to come back from Cohen & Densen-Gerber 1982). Stall and co­ her trick [commercial sex], I was going up workers found these associations in a towards Washington Street. ... There is like household telephone survey of 2,881 MSM this little alleyway. I take that alleyway be­ in New York City, Chicago, Los Angeles and cause itis a short cut, everybody knows San Francisco. Moreover, the percentage of that. Thatis where they got me. They started MSM in the study reporting high-risk sex be­ to attack me and one dude sliced me like havior increased steadily from 7.1 percent that [indicating a jagged 12 inch slash across among those with none of the four health his chest on the left side]. problems to 33.3 percent for those suffering from all four. For men who lacked any of the A 38-year old man of mixed Italian back­ co-factors, 13 percent were HIV+ compared ground, Tony had been using drugs heavily to 25 percent who reported all four co-fac­ for twenty-five years, had both been victim tors. Consequently, these workers affirmed and a perpetrator of drug-related violence the existence of a SAVA syndemic among since childhood. The violence began with MSM that has its roots in childhood sexual harsh beatings administered by his father; abuse. Childhood sexual abuse contributes daily beatings intended to correct his alleged Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 45 transgressions. These continued during heroin use was less common, with a me­ adolescence as he defended his ground in dian of 10 times per 30 days. Frequency of the bellicose world of street-corner drug alcohol consumption had a U-shaped distri­ dealing, a practice taught to him by his fa­ bution, with participants at the bottom quartile ther. During his young adulthood, violence, reporting drinking on 1 or 2 days during the in the form of brutal assaults of wayward last month and those in the top quartile re­ members, was a regular part of his role as porting drinking on 25 or more days. an "enforcer" in a drug-selling street gang. Turning to the issue of violence, we iden­ Indeed, violence, in one form or another, was tified a wide range of violence exposure and an enduring component Tony's life until he involvement types among study participants. contracted AIDS through his daily drug injec­ Seventy-four percent of our participants re­ tion .2 ported witnessing fighting in the streets of This study explains differences in vio­ their neighborhood during the last 4 months lence, drug/alcohol use, and HIV risk among (Romero-Daza, Weeks & Singer 1998). Vio­ study participants and in participant personal lence in the streets was said to be especial­ networks (Singer 1999b). Data collection ly common by participants, with "once or twice was targeted at three levels: a) at the individ­ a week" being the median frequency. The ual level, by implementing a prospective other most common type of recently wit­ study design that allowed systematic quanti­ nessed violence was domestic violence, tative and qualitative data collection every four which, notably, was reported by 54 percent of months with a street outreach-recruited study participants. Gang violence (45%), rob­ sample of drug users from targeted neigh­ bery and muggings (42%), and beatings or borhoods; b) at the social relationship level, stabbing (31 %) were the next most common by identifing and interviewing a set of index types of violence participants reported they individuals and members of their personal had witnessed. As these findings indicate, drugs and sex network (Singer et al 1999b); street drug users are experientially exposed and c) at the social context level, by assess­ to a considerable amount of violence on the ing key context characteristics (e.g., unem­ streets and in their homes in which they are ployment, crime) and contextual threats ex­ neither victim nor perpetrator. It became evi­ perienced by sample participants in four tar­ dent in our study that witnessing violence is get neighborhoods. The final sample in­ an important element in assessing the inter­ cluded 224 participants with data collected relationship of violence, drug use and HIV at intake and 4, 8, and 12 months follow-up. risk. Heroin was found to be the most com­ Considering all forms of direct "involve­ monly used drug (64% of participants) dur­ ment" in violence (including emotional ing the 30 days period prior to the initial inter­ abuse)-as either victim or perpetrator-, 39 view of study participants, followed closely percent of the sample reported being a vic­ by alcohol. In descending order of impor­ tim over the past 4 months, while 30 percent tance, other commonly used drugs in this reported being a perpetrator of some form of sample were speedball, crack, powder co­ violence. Specific rates of violence victimiza­ caine, and tranquilizers. As compared with tion affirm that exposure to violence is not a our prior studies of street drug users in Hart­ distant phenomenon in the lives of street drug ford over the last 10 years, in which we have users. A third of participants reported being consistently found comparatively high rates the victims of emotional abuse during the of heroin injection among Puerto Rican drug prior 4 months. Additionally, 14 percent re­ users, we found a significant number of non­ ported being the targets of physical violence injection heroin users - 29 percent of heroin­ and 7 percent indicated they suffered seri­ users in the sample - this suggests either a ous physical violence during this period. shift in the pattern of consumption or the ten­ Indeed, violence is a component of every­ dency of network methods to tap a different day life among drug users as indicated by strata of drug users than street outreach. For the following participant who explained his those who inject heroin, the median rate of strategy for defense against the constant injection was 70 times during the previous threat of violent attack: 30 days. Thirty percent of injecting partici­ pants injected 120 or more times during the I was up against the corner, and I was sit­ last month (Singer 1999a). Non-injection ting on that little bench, the little couch. A 46 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

guy came up and said , 'Give me everything , them to escape exposure to violence (38% your watch, everything.' He had a knife .. . vs . 27%) (Dushay, Singer, Weeks, Rohena, He had me trapped in the corner. And the & Gruber 2001 ). way he had me , you know. It was like, 'give In an ethnographic component of the it up , and this and that.' And see, if I had study, we conducted qualitative interviews seen it coming, I would grab .. . you see, I with a subsample of 30 participants, most of always carry a bottle .... I'll crack that over whom were men . These individuals provided someone's head. They'll think twice about graphic descriptions of the experience of vio­ robbing me with a knife or not .... I think he lence in their lives (Romero-Daza et al 1998). was using "ready" [cocaine]. He probably For example, several participants described wanted to get a hit, because I had dope on witnessing murders. One participant, who me and he came in with a girl. And he was suffered the emotional consequences of wit­ like, 'Give me the dope too !.' nessing violence, stated:

Significantly, nine of the drug users in our This is haunting me still about when I seen sample responded that they had been the they killed this man and everything .... They target of attempted murder. Participants also beat this man up and he was dead. I think all revealed their own role as perpetrators of the blows and everything, and they took violence against others during the last four his head and hit him on the floor and that months. Ten percent admitted committing killed him .. .. And now one of my brothers is acts of violence, while two respondents indi­ in jail cause he shot another man. Cause if cated that they had attempted murder during he wouldn't have shot that man , he would the 4 month period . have killed my brother for a bike the other Of the incidents of violence victimization man wanted. reported by participants, 71 percent of physi­ cal violence involved the use of drugs or al­ Another stated, cohol. In the reported incidents of serious physical violence, the rate of substance use I used to sell drugs [and] this guy killed this rose to 75 percent. In cases where the study woman in the alleyway behind my door and participant was the perpetrator, the reported I seen it happen and , you know, I was use of psychoactive substances was 75 per­ scared. cent when they committed emotional abuse, 80 percent for acts of physical violence Prison is an environment in which many against another person , and 100 percent in of our participants witnessed a considerable serious physical violence and attempted amount of horrific violence, as one study par­ deadly acts of violence. Notably, 44 percent ticipant indicated: of our participants indicated that involvement in violence (as either victim or perpetrator) Well, in jail sometimes the people would get contributed to increases in their rate of drug together, you know, a lot of people, and consumption. Also of interest, 14 percent re­ they used to beat up a lot of people, they ported a decrease in the frequency of drug would hit them , they would rob them and use as a result of involvement in violence, they would do all kinds of barbaric things. including two participants who gave up us­ Sometimes they would rape them also . I ing drugs as a result (Duke, Teng , Simmons, saw a lot of things, a lot of fights, I also & Singer 2003). saw, they would stab them also, I saw Among women participants (about 1/4 of knives also. the sample) , 16 percent reported being vic­ tims of violence during the last 4 months. In Noting the reflective emotional effects of cases of violence against women , the per­ being a perpetrator of violence, another par­ petrator was more likely to be a family mem­ ticipant stated: ber or someone known to the woman than was the case in victimization among men So then I started shooting up and that's (1 00% vs . 75%). Notably, women were also when I started going crazy ... you know like more likely to increase drug use following getting sick, real sick, starting to do bad violence victimization. Additionally, women things, stealing robbing .. .like taking money were more likely to report that it was hard for away from people ... and I used to have a Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 47

gun. My cousin had a gun and he used to made me feel I needed to escape. I wanted give it to me so I can go rob people, like drug to be like a free bird and when I finally found dealers, take their drugs. A lot of crazy stuff myself free thatis when everything hap­ like that I come to think about now and I be pened. I started with marijuana and then I like, damn, man, I could have been dead. moved to crack and then to heroin. Every time I had a chance to get out I would buy Gang involvement, which is extensive some rock [crack] and use it in secret. I felt among younger drug users in Hartford, was trapped. I was with him for 20 years. I related to a considerable amount of violence couldnit talk to anyone. If he saw me talking perpetration by our participants. This is ex­ to anyone, he would say I had something emplified by one individual who participated with that person; man or woman . He in a number of gang ibeatdownsi and stab­ wouldnit hit me in front of people but as bings. In one instance, he was called upon soon as we got home he would hit me. Heid by gang leaders to beat another inmate. He throw dishes at me. He would hit me with recalled: his fists.

There was one guy one day, he molested Community-based studies, including eth­ somebody - a little kid ... and we put a blan­ nographic examination of actual temporal, ket over his head, and we started beating sequential, or other associations between down. We took some socks and put some drug use, HIV risk and violence, are needed locks inside the socks and started hitting to further clarify the actual nature of the rela­ the guy... tionships that exists between these inter­ twined epidemics that have had a notable This individual also reported involvement effect on morbidity and mortality in low in­ in a drive-by shooting as a result of his gang come and minority communities. However, it affiliation, stating: is evident that the SAVA syndemic is a signifi­ cant aspect of the life experience of street They just send me to go and shoot some­ drug using populations, including contexts body. I never did it, the other two guys did it. away from the street like jails and prisons. We went to the block and started shooting Particularly severe forms of violence are everybody in there. found in this population punctuating the everyday violence associated with conflicts Partner violence was also reported dur­ over drug deals, disagreements during the ing qualitative interviews. One woman partici­ sharing of drugs, police harassment, lack of pant reported, reliable shelter, and drug user on drug user use of force to extract items of any value, I thought that if I would leave him, my kids, however minimal. you know, theyire going to suffer because they didnit have a father and stuff like that, SAVA AND COMMERCIAL SEX so I stayed but, after three, four years, I left As Silliman and Bhattacharjee empha­ ... He broke my leg. He pushed me down the size, stairs and broke my leg. You know how you get black and blue and stuff like that? women in prostitution are particularly at risk He used to hurt me like that. My body was of gender-based violence6including physi­ all sore. So, I took the train and went to cal, psychological and economic violence6 New York, to my cousin. from pimps, buyers, police and boyfriends. (2002 210) Nilda, a 26-year old woman who had been using heroin for 7 years at the time she was This fact has often been hidden behind a interviewed, reported: public health focus on prostitution as a vec­ tor of disease. Notes Janice Raymond and My husband was twice my age. He used to colleagues, beat me up all the time. He was a very jeal­ ous man. He wouldnit even let me look out The minimal documentation of the harm and the window. He will tear my clothes off and trauma of prostitution and trafficking may in he will keep me locked in the house. That large part be due to the fact that prostitution 48 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

has not been recognized as a form of vio­ being witnesses to domestic violence , los­ lence against women and the ambivalence, ing their primary home , being runaways, on the part of many researchers , NGOs having a difficult home life, and economic and governments, to view prostitution as a destitution ... Much of their current substance violation of women 's human rights. (Ray­ abuse results from the accretion of abuse : mond , D'Cunha , Dzuhayatin , Hynes, Rodri­ sexual , physical, mental and economic prior guez, & Santos 2002 296) to and within prostitution. Further, some re­ port that prostitution worsens their drug None the less, violence is a common ex­ habits, forcing them to escape longer and perience among commercial sex workers. deeper from the consciousness of their en­ For example, Parrirot (1994) who interviewed trapment. The worsening of drug problems 68 women in Minneapolis/St. Paul who had ultimately traps them within prostitution , been involved in commercial sex for at least sapping them of the stamina and will to get six months in various setting, including the out. (2002 197) "street," massage parlors, and escort ser­ vices, found that 62 percent had been raped , Women in their study (Raymond et al 2002 half had been physically assaulted; and one 196-197) reported: third were assaulted by customers at least several times each year. About one fourth of • ''They just broke me down , shattered my the women suffered broken bones and two will and hopes. I was humiliated." were beaten into a coma (Parriott 1994). • "They didn't push me to take drugs, they Similarly, a survey of 55 commericial sex just made me an injection about 2 weeks workers in Portland, Oregon, found that the after arriving. " majority (78%) reported being raped by • "The bosses ... they used to say: 'Re ­ pimps and male customers on average 49 member, there was a girl working for times a year. Additionally, 84 percent were us . You should know, she is not here the victims of aggravated assault, often suffi­ anymore because she did something she cient to require emergency room treatment; was not supposed to ."' 53 percent were sexually abused or tortured; and 27 percent were mutilated (Hunter 1993). Commercial sex workers at special risk Based on a study of commercial sex in for substance abuse, violence, and AIDS are four countries, Raymond and her co-work­ those who get caught up in the international ers report the existence of a complex relation­ cross-border commercial sex trade (Singer, ship between substance abuse and violence Salaheen, & He 2004). Generally speaking, in this population: there is a strong link between migration and the geographical spread of HIV/AIDS and Some [commercial sex workers] encour­ other infections. Studies have shown that ex­ aged buyers to use crack so buyers "would tended or repeated overnight travel away from forget about sex altogether." Most of the one's home community is associated height­ women were habitual drug users ( ... 77%), ened risk for HIV infection. However, in the and ... used alcohol and drugs to deaden case of commercial sex trafficking-i.e., the their feelings . [As one participant ex­ movement, usually of women and girls, plained],"lt would end up that I would just across national boundries for use as com­ drink to get drunk to cover up what I was mercial sex workers-the link is particularly feeling-which was dirty and ashamed ." strong, and its causes identifiable. Although many U.S. women said that they Whatever their country of origin or ultimate used drugs and alcohol prior to entering destination, women ensnared in the cross­ prostitution ... , it is simplistic to assume that border sex trade tend to come from impover­ they entered prostitution to support a drug ished families and the poorest regions of habit. The cycle of substance abuse in their home countries, have limited formal which they are caught has its roots in the education, and to have their roots in rural life history of abuse , neglect, and severe areas and in subordinated ethnic minority stress which all of the respondents ... de­ groups within their countries of origin. These scribed when asked about prior sexual factors, demarcating the weak social re­ abuse before entering the sex industry. sources women bring with them into the [Many had) experiences of rape , incest, arena of commercial sex, are magnified Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 49 many times as a result of cross border com­ ford using street outreach (to contact initial mercial sex trafficking, which involves: isola­ participants) and the personal networks of tion from any means of traditional social sup­ intial contacts to recruit subsequent partici­ port, often having illegal status in a foreign pants and found that 15 percent of the individ­ country, often having limited linguistic or cul­ uals in our sample had ever been involved in tural skills in the new context, being trapped commercial sex. Age of initiation into com­ in some form of debt to the traffickers, having mercial sex ranged from 14 to 35 years of limited knowledge of HIV prevention, and age, with a mean age of 24 years. Level of possessing little ability to negotiate preven­ involvement in commericial varied, but about tive behaviors with clients or access to medi­ 40 percent reported they had traded sex for cal care through their handlers or elsewhere. drugs or money many times. Almost all of In assessing the ability individuals have these individuals (91 %) reported that drug to protect themselves from HIV infection­ and alcohol use was part of their involve­ measured in terms their HIV/AIDS knowledge ment in commercial sex. Violence in various and learned prevention skills, social posi­ forms (observed, attacks on friends, street tion and ability to command the labor of oth­ fights, and personal victimization) was very ers on their behalf, level of emotional and common in the whole sample, with over 10 material social support, possession or con­ percent reported seeing street robberies al­ trol over material recourses including pre­ most every day while growing up while 38 vention materials, freedom of movement; percent stated that they observed fighting in protection from violence, and overall health the street almost every day as children. About status-women in the commercial sex trade 30 percent of participants reported that they are clearly at high risk. Their limited social had witnessed a murder. In this sample of options and resources makes them highly Puerto Rican drug users, 16 percent had vulnerable to HIV infection and to disease been diagnosed with HIV/AIDS, 19 percent progression. Consequently, when the AIDS had been diagnosed with another sexually epidemic in Thailand was at its peak, over transmitted infection, and 14 percent had 80 percent of HIV/AIDS cases in the country been diagnosed for Hepatitis B. As these data were attributed to commercial sex workers suggest, commericial sex, violence, and vari­ and clients (Viravaidya 1993). In India, infec­ ous infectious disease are commonly inter­ tion rates among commercial sex workers twined with drug use, a finding replicated in in some locales such as Mumbai (Bombay) all of our studies. exceeds 50 percent (DevNews Media Cen­ ter 2002). CONCLUSION Similarly high rates of STD infection have AIDS has been conceptualized in several been found in migrant sex workers in Italy different ways since it first gained medical (Matteelli 2003). Notably in their comparative and ultimately public recognition early in the study of commercial sex trafficking in Indo­ 1980s. Initially, because its symptoms were nesia, the Philippines, Thailand, Venezuela, so unexpected (e.g. , a cancer associated and the United States, Janice Raymond and with old age showing up in young men, a workers (2002) found that the highest rate of lung infection associated with recent surgery, physical violence was against women traf­ especially organ transplant, and weakened ficked to the United States. It is not quite clear immune capacity induced to avoid organ re­ why this is the case, but it may be a conse­ jection being diagnosed in people who had quence of the opportunities for these women not undergone surgery nor taken immune to escape from their pimps in a wealthy coun­ blocking drugs}, that physicians and epide­ try with many service and feminist organiza­ miologists were uncertain how to conceptu­ tions. Under such conditions, pimps may use alize the new diease. high doses of violence to dissuade women Before long because it appeared that all from seeking ouside contacts. of the initial sufferers were gay men (in fact, In our study of the relationships between they were not), AIDS-under the rubric of GRID violence, drug/alcohol use, and HIV risk (Gay Retroviral Immune Deficiency)- be­ among active drug users in the Puerto Rican came the reigning conceptualization. How­ community (Singer 1996 ; Singer et al1 999a), ever, the obvious spread of the diease be­ our research team at the Hispanic Health yond gay men soon led to a new conceptual­ Council recruited 224 participants in Hart- ization based on the existence of so-called 50 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

"risk groups," most notably, initially, the 4-H As a result, it would not be inappropriate club (homosexuals, hemophiliacs, heroin in­ to talk about the existence of multiple SAVA jectors, and Haitians, although, by this point epidemics, each driven by its own configura­ the disease had intact spread to many other tion of social conditions and relationships. groups as well) . This recognition points to the importance of Alternate conceptions also arose, includ­ public health responses that: 1) go beyond ing those with conservative religious bent focusing just on HIV but instead respond to (AIDS as God's punishment for sin) and the roles of drugs and violence in undermin­ those with a political foundation (AIDS as ing the effectiveness of narrowly pitched pre­ conspiracy against gays, or ethnic minori­ vention initatives; and 2) being sensitive to ties, the primary sufferers) . Eventually, as the specific population of immediate con­ aspects of the new epidemic began to be cern and the particular expression(s) of SAVA clear, the term AIDS was introduced and the in this population. What is called for, then , is retrovirus that caused the new disease was a two directional approach involving both a identified. At this point, AIDS came to be con­ broadening of focus to approach AIDS pre­ ceptualized like other infectious diseases, a vention/intervention in terms of a syndemic distinct entity with an identified pathogenic model of responding to multiple, interacting cause that could potentially be contracted by dieases as a single entity, and a narrowing anyone engaged in a set of "risk behaviors" of focus to match prevention/intervention to (ignoring the fact that many are infected by specific populations in social context. In other conventional behaviors like sexual inter­ words, prevention efforts must be guided by course with their spouse or being born to an a keen awareness of and response to the infected mother). social, cultural , and health conditions of tar­ From the perspective of biomedicine, get populations. So too, AIDS care. each known disease is a discrete, objective, and clinically identifiable phenomenon. Nor­ ENDNOTES mal practice in biomedicine is guided by the 1 We used street outreach to recruit a sample of conceptualization of diseases as disjunctive 500 not-in-treatment heroin and/or cocaine us­ entities that exist (in theory) separate from ing women over the age of 18 (average age = other diseases and from the social groups 37 .8 years) in the greater Hartford area. The and social contexts in which they are found sample reflects the ethnic composition of Hart­ at any point in time. Introduction of the term ford , with 38 .6% of the women being African American, 39.4% being Hispanic, and 17.4% syndemic, and SAVA as one example of a being non-Hispanic white. syndemic, was specifically intended to fur­ 2Tony was a participant a NIDA-funded study by ther refine our conceptualization, beyond con­ our research team at the Hispanic Health Coun­ ventional thinking about bounded, indepen­ cil of relationships between substance abuse, dent disease entities and to a realization of violence, and HIV risk among not-in-treatment interlocking, mutually advancing threats to Puerto Rican street drug users in Hartford (NIDA health in conducive social contexts. #R01 DA 10438) (Singer 1996; Singer, Simmons, With a syndemic understanding, AIDS is Duke, & Broomhall 1999a; Singer, Duke, Soto, conceived not in isolation as a specific dis­ & Weeks 1999b). ease with particular properties but rather in terms of its relationship to other diseases REFERENCES Alcabes P, E Schoenbaum & R Klein . 1993. Corre­ and social conditions. In the case of SAVA, it lates of the rates of decline of CD4+ lympho­ is the relationship among HIV/AIDS, violence, cytes among injection drug users infected with and drug use that is of primary analytic con­ human immunodeficiency virus. Amer J Epide­ cern. AIDS, drug use, and violence are con­ miology 137 989-1000. ceived not as distinct "things in the world" Amaro H, L Fried, H Cabral, & B Zuckerman. 1990. but as phenomena in tandem, the essence Violence during pregnancy and substance use. of each being significantly shaped by the Amer J Public Health 80 575-579. presence, nature and influence of the oth­ Barthalow B, L Doll, D Joy, J Douglas, G Bolan, J ers. As argued in this paper, the actual ex­ Harrison, P Moss, & D McKirnan. 1994. Emo­ pression of the SAVA syndemic is shaped as tional behavior and HIV risks associated with sexual abuse among adult homosexual and bi­ well by the social context, including both the sexual men. Child Abuse & Neglect 18 747- population being affected and the social con­ 761 . ditions faced by the population of concern . Beitchman J, K Zucker, J Hood, G Dacosta, D Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 51

Akamn, & E Cassavia. 1992. Review of the Structural and interpersonal violence among long-term effects of child sexual abuse. Child Puerto Rican drug users. Practicing Anthro­ Abuse & Neglect 16 101-118. pology 25 3 28-31 . Bennett L & M Larson. 1994. Barriers to coopera­ Duke M, WTeng, S Clair, H Saleheen, P Choice, &M tion between domestic violence and substance Singer. 2006. Patterns of intimate partner vio­ abuse programs. Family in Society 75 277- lence among drug using women. Free lnq Great 286. Social 34 1 pp-pp. Bernstein K, R Tulloch, J Montes, G Golan, I Dyer, Dushay RA, M Singer, M Weeks, L Rohena, & R M Lawrence, D Dodagoda, H Rotblatt, P Kerndt, Gruber. 2001 . Lowering HIV risk among ethnic R Gunn, N DeAugustine, & P Weismuller. 2001 . minority drug users: comparing culturally tar­ Outbreak of syphylis among men who have geted intervention to a standard intervention. sex with men-southern California. Morbidity Amer J Drug Alcohol Abuse 27 3 504-524. & Mortality Weekly Rev 50 117-120. EI-Bassel N, L Gilbert, R Schilling, & T Wad a. 2000. Bourgois P. 1995. In Search of Research: Selling Drug abuse and partner violence among women Crack in El Barrio. Cambridge: Cambridge U in methadone treatment. J Family Violence 15 Press. 3 209-225. Breire J & M Runtz. 1987. Post-sexual abuse Ensoli F & MC Sirianni. 2002. HIV/HCV co-infec­ trauma: data and implications for clinical prac­ tion: clinical and therapeutic challenges. AIDS tice. J Interpersonal Relations 2 367-379. 16 10 1419-1420. Brewer D, C Fleming, K Haggerty, & R Calalano. Farmer P. 2003. Pathologies of power: health, hu­ 1998. Drug use predictors of partner violence man rights and the new war on the poor. North in opiate dependent women. Violence Victim­ Amer Dialogue 6 1 1-4. ization 13 2 107-115. Farzio K, J Bueler, M Chamberland, B Whyte, E Brown G & B Anderson. 1991. Psychiatric morbid­ Sivanajan Froelicher, S Hopkins, C Reed, E ity in adult inpatients with childhood histories Mokotoff, D Cohn, S Troxler, A Phelps, & R of sexual and physical abuse. Amer J Psy­ Berkelman. 1992. Spectrum of diseases in per­ chiatry 148 55-61 . sons with human immunodeficiency virus in­ Burnam M, J Stein, I Golding, J Siegel, S Sorenson, fection in the United States. J Amer Medical A Forsythe, & C Telles. 1988. Sexual assault Assoc 267 1798-1805. and mental disorders in a community popula­ Frumkin H. 2002. Urban sprawl and public health. tion. J Consulting Clinical Psycho/ 56 843- Public Health Rep 117 23-35. 850. Gilbert L, N EI-Bassel, V Rajah, A Foleno, J Fontde­ Bushnell J, J Wells, & M Oakley-Browne. 1992. vila, V Frye, & B Richman. 2000. The converg­ Long-term effects of intrafamilial sexual abuse ing epidemics of mood altering-drug use, HIV, in childhood. Acta Psychiatry Scandenavia 85 HCV, and partner violence. Mount Sinai J Medi­ 136-142. cine 67 5&6 452-464. Carballo-Dieguez A & C Dolezal. 1995. Association Golub E, J Astemborski, D Hoover, JC Anthony, D between history of childhood sexual abuse and Vlahov, & S Strathdee. 2003. Psychological dis­ adult HIV-risk sexual behavior in Puerto-Rican tress and progression to AIDS in a cohort of men who have sex with men. Child Abuse injection drug users. J Acquired Immune Defi­ Neglect 19 595-605. ciency Syndromes 32 4 429-434. Carter L, L Weithorn L, & R Behrman. 1999. Do­ Goodwin J, K Cheeves, & V Connell. 1990. Bor­ mestic violence and children: analysis and rec­ derline and other severe symptoms in adult ommendations. Future Children 9 3 4-20. survivors of incestuous abuse. Psychiatric An­ Centers for Disease Control and Prevention. 2002. nals 20 22-32. HIV/AIDS Surveillance Report. Atlanta, GA. Homer J & B Milstein. 2002. Communities with Mul­ Clair S & M Singer. 2004. HIV status, risk, and tiple Afflictions: A System Dynamics Approach prevention needs among Latino and non-Latina to the Study and Prevention of Syndemics. Pre­ MSM in Connecticut. ln. AIDS, Culture, and Gay sented at the International System Dynamics Men. D Feldman, ed. Greenwood Press. Conference, Palermo, Italy, 2002. Cohen F & J Densen-Gerber. 1982. A study of the Hunter SK. 1993. Prostitution is cruelty and abuse relationship between child abuse and drug to women and children. Portland: Feminist addiction in patients: preliminary results. Child Broadcast Quarterly. Abuse Neglect 6 383-387. Jinich S, J Paul, R Stall, M Acree, S Kegeles, C DevNews Media Center. 2002. SAR Puts HIV/A/DS Hoff, & T Coats. 1998. Childhood sexual abuse at Top of Agenda. Washington, DC : The World and HIV risk-taking behavior among gay and Bank Group. bisexual men. AIDS Behavior 2 41 -51 . Dilorio C, T Hartwell, & N Hansen. 2002. Childhood Kilpatrick D & B Saunders. 1997. The Prevalence sexual abuse and risk behaviors among men and Consequences of Child Victimization: at high risk for HIV infection. Amer J Medicine Summary of a Research Study. Washington 337 501-502. DC: U.S. Dept. of Justice, Office of Justice Pro­ Duke M, W Teng , J Simmons, & M Singer. 2003. grams, National Institute of Justice. 52 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Kilpatrick D, H O'Neill, S Beak, H Resnick, E Stugis, Reback C & C Grella. 1999. HIV risk behaviors of C Best, & B Saunders. 1990. Parental sub­ gay and bisexual male methamphetamine us­ stance abuse, personal victimization , and ers contacted through street outreach. J Drug women's risks for substance abuse: results Issues 29 155-166. from a national probability sample. Paper pre­ Relf M, B Huang, J Campbell, & J Catania. 2004. sented at the Association for Advancement of Gay identity, interpersonal violence, and HIV Behavior Therapy, San Francisco, Calif, No­ risk behaviors: an empirical test of theoretical vember. relationships among a probability-based sample Kovach J. 1983. The relationship between treat­ of urban men who have sex with men . J Assoc ment failures of alcoholic women and incestu­ Nurses AIDS Care 15 2 14-26. ous histories with possible implications for post Rohsenow D, R Corbett, & D Devine. 1988. Mo­ traumatic stress disorder symptomatology. lested as children : a hidden contribution to sub­ Doctoral dissertation, Wayne State University stance abuse? J Substance Abuse Treatment Graduate School. 5 13-18. MacQueen K. 2002. Anthropology and public health. Romero-Daza N, M Weeks, & M Singer. 1998. Much In Encyclopedia of Public Health. L Breslow, more than HIV! the reality of life on the streets L Green, W Keck, J Last, & M McGinnis, eds. for drug-using sex workers in inner city Hart­ NY: Macmillan. ford. lnternat Qtrly Community Health Edu 18 Matteelli A. 2003. Chlamydia trachomatis genital 1 107-119. infection in migrant female sex workers in Italy. Russell D. 1983. The incidence and prevalence of International J STOIA/OS 14 9 591-5. intrafamilial and extrafamilial sexual abuse of Miller B, W Downs, & M Testa. 1993. Interrelation­ female children. Child Abuse Neglect 7 133- ships between victimization experiences and 146. women's alcohol use. J Studies Alcohol Supp Scott K. 1992. Childhood sexual abuse: impact on No. 11109-117. a community's mental health status. Child Miller B, W Downs, D Gondoli, & A Keil. 1987. The Abuse Neglect 16 285-295. role of childhood sexual abuse in the develop­ Shoptaw S, C Reback, & T Freese. 2002. Patient ment of alcoholism in women. Violence Vic­ chacteristics, HIV serostatus, and risk behav­ timization 2 157-171 . iors among gay and bisexual males seeking Milstein B. 2001 . Introduction to the Syndemics treatment for methamphetamine abuse and Prevention Network. Atlanta: Centers for Dis­ dependence in Los Angeles. J Addictive Dis­ ease Control and Prevention. eases 21 91-105. O'Connell J, T Lampinen, A Weber, K Chan, M Miller, Silliman J & A Bhattacharjee. 2002. Policing the M Schechter, & R Hogg. 2004. Sexual risk pro­ National Body: Sex, Race, and Criminalization. file of young men in Vancouver, British Colum­ Cambridge, MA: South End Press bia, who have sex with men and inject drugs. Singer M. 1994. AIDS and the health crisis of the AIDS Behavior 8 1 17-23. US urban poor: the perspective of critical medi­ Parriott R. 1994. Health Experiences of Twin Cit­ cal anthropology. Soc Sci Medicine 39 7 931- ies Women Used in Prostitution: Survey Find­ 48. ings and Recommendations. St. Paul, MN : ------· 1995. Providing Substance Abuse Breaking Free. Treatment to Puerto Rican Clients Living in Pribor E & S Dinwiddie. 1992. Psychiatric corre­ the U.S. 1 In Providing Substance Abuse Treat­ lates of incest in childhood . Amer J Psychiatry ment in the Era ofAIDS . Washington, DC: CSAT 149 52-56. --:--· 1996. A dose of drugs, a touch of vio­ Pugliese A, LAndronico, L Gennero, G Pagliano, G lence, a case of AIDS: conceptualizing the Gallo, & D Torre. 2002a. Cervico-vaginal dys­ SAVA syndemic. Free lnq Creat Social 24 2 plasia-papillomavirus-induced and HIV-1 infec­ 99-110. tion : role of correlated markers for prognostic ---:-:c,.---,-· 2006. The Face of Social Suffering: Life evaluation. Cell Biochemical Function 3 233- History of a Street Drug Addict. Prospect 236. Heights, IL: Waveland Press. Pugliese A, D Torre, A Saini, G Paglia no , G Gallo, P Singer M & S Clair. 2003. Syndemics and public Pistono, & C Paggi. 2002b. Cytokine detection health: reconceptualizing disease in bio-social in HIV-1/HHV-8 co-infected subjects. Cell Bio­ context. Medical Anthro Qtrly 17 4 423-441 . chemical Function 20 3 191-4. Singer M, M Duke, M Soto, & M Weeks. 1999b. Raymond J, J D'Cunha, SR Dzuhayatin, HP Hynes, Violence in the lives and social networks of ZR Rodriguez, & A Santos. 2002. Comparative street drug users. Bull Alcohol Drug Study Study of Women Trafficked in the Migration Group 34 3 8-11. Process: Patterns, Profiles and Health Con­ Singer M & L Marxuach-Rodriquez. 1996. Apply­ sequences of Sexual Exploitation in Five ing anthropology to the prevention of AIDS: the Countries (Indonesia, the Philippines, Thai­ Latino gay men's health project. Human Orga­ land, Venezuela and the United States). Coali­ nization 55 2 141-148. tion Against Trafficking in Women. Singer M, H Salaheen, & Z He. 2004. The Special Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 53

HIV and Related Health Vulnerabilities of Sex Golding, ed. Lancaster, England: MTP Press Workers Caught in the International Sex Trade Limited. Industry. Presented at the Society for Applied Stueve A, L O'Donnell, R Duran, A San Doval, J Anthropology. Dallas, Texas. Geier, & Community Intervention Trial for Youth Singer M, J Simmons, M Duke, & L Broomhall. 1999a. Study Team. 2002. Being high and taking sexual The challenges of street research on drug use, risks: findings from a multisite survey of urban violence, and AIDS risk. In Qualitative Methods young men who have sex with men. AIDS Edu in Drug Research. Prevention 14 6 482-95. Sobo E. 1995. Finance, romance, and social sup­ Thio C, E Seaberg, R Skolasky, J Phair, B Visscher, port and condom use among impoverished in­ A Munoz, & Multicenter AIDS Cohort Study. 2002. ner city women. Human Organization 63 115- HIV-1 , hepatitis B virus, and risk of liver-re­ 128. lated mortality in the Multicenter Cohort Study Stall R, T Mills, J Williamson, & T Hart. 2003. Asso­ (MACS). Lancet 14 1921-1926. ciation of co-ocurring psychosocial health prob­ Viravaidya M. 1993. The economic impact of AIDS lems and increased vulnerability to HIV/AIDS on Thailand. In Economic Implications of AIDS among urban men who have sex with men. in Asia. DE Bloom & JV Lyons, eds. Amer J Public Health 93 6 88-99. Wallace R. 1990. Urban desertification, public Stall R & D Purcell. 2000. Intertwined epidemics: a health and public order: planned shrinkage, vio­ review of research on substance use among lent death, substance abuse and AIDS in the men who mave sex with men and its connec­ Bronx. Soc Sci Medicine 31 801-813. tion to the AIDS epidemic. AIDS Behavior4181- Waterston A. 1993. Street Addicts in the Political 192. Economy. Philadelphia: Temple U Press. Stein J, J Golding, J Siegel, M Burnam, & S Widom C. 1989. Does violence beget violence? a Sorenson. 1988. Long-term psychological se­ critical examination of the literature. Psycho­ quelae of child sexual abuse: the Los Angeles logical Bull 106 1 3-28. epidemiologic catchment area study. Pp. 135- World Health Organization. 2004. WHO Pushing 154 in Lasting Effects of Child Sexual Abuse. to Rapidly Scale-up Measures to Fight TB and G Wyatt & G Powell, eds. Newbury Park, CA: HIV. Press release, January 21 . Geneva. Sage. Wyatt G. 1985. The sexual abuse of Afro-Ameri­ Sterne M, S Schaefer, & S Evans. 1983. Women's can and white-American women in childhood. sexuality and alcoholism. Pp. 421-425 in Alco­ Child Abuse Neglect 9 507-519. holism: Analysis of a World-Wide Problem. P 54 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

INTERNATIONAL REVIEW OF MODERN SOCIOLOGY & INTERNATIONAL INTERNATIONALJOURNALOF SOCIOLOGY OF THE FAMILY REVIEW OF MODERN MANUSCRIPTS should be submitted in duplicate and in acceptab~ form. REFERENCES and FOOT­ SOCIOLOGY NOTES should be according to the FORMAT used in AMERICAN SOCIOLOGICAL REVIEW. Styles instruction for guidance in preparing manuscripts will be provided upon request to the EDITOR INTERNATIONAL PROCESSING FEE of$25.00 must accompany each paper submitted either to INTERNATIONAL JOURNAL OF REVIEW OF MODERN SOCIOLOGY or INTERNATIONAL JOURNAL OF SOCIOLOGY OF THE FAMILY. The manuscripts not accompanied by the fee will not be processed until the fee is received SOCIOLOGY by the EDITOR OF THE FAMILY

EDITORIAL OFFICE: Address correspondence on manuscripts, news and announcements. Mail to: Editor: Dr. ManS. Das; Sociology Department; Northern Illinois University; DeKalb, Illinois 60115-2854, USA. BUSINESS OFFICE: Address correspondence on subscriptk>ns, change of address, renewals , advertising , reprints of individual artic~s and permission to quote. Mail to: International Journals; PRINTS INDIA, 11 DARYAGANJ ; NEW DELHI1100021NDIA. SUBSCRIPTION RATES: Annual RS 200, $40; single issueRS 100,$20. Biannual. MISSION: Both journals are devoted to encourage cross-cultural, cross-national and inter-disciplinary research and exchange of information concerning significant developments in comparative sociology and sociology of marriage and the family. The journals publish theoretical, methocological and empirical articles, book reviews , letters to the Editor, comments, rejoinders , annotated bibliographies, news, and announcements. Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 55 APPLYING ASPECTS OF PROBLEM BEHAVIOR THEORY TO LATINO YOUTH: THEORETICAL, METHODOLOGICAL, AND SOCIOCULTURAL CONSIDERATIONS

Byron L. Zamboanga, Smith College, and Gustavo Carlo, University of Nebraska-Lincoln

ABSTRACT

Problem Behavior Theory (PBT; Jessor, Donovan, & Costa 1991 ) is a common and influential perspective designed to further our understanding of problem behaviors among youth. However, few scholars have directly examined the validity of PBT to use with Latino youth. The present chapter critically examines the basic tenets of PBT and its relevance to understanding problem and conventional behavioral outcomes in Latino youth. A brief overview of PBT is presented, foll owed by an in-depth, critical discussion of its application to research on Lati no youth. Several conceptual and methodological recommendations for future research are di scussed including definitional issues and relevant social, cultural, and demographic influences (e.g., peer and famil y influences, socioeconomic status, ethnic identity, immigration and acculturation, and gender) on problem and conventional behaviors. In addition, the links between problem and conventional behaviors are critically investigated. Consideration of these various factors will further enhance the ecological and cultural validity of Latino youth development theories and research.

INTRODUCTION retical models on problem and conventional Problem behaviors (e.g., substance use, behaviors were formulated and tested pri­ physical fighting, binge drinking, risky sex be­ marily with White, middle-income popula­ haviors) among Latino1 adolescents has tions. And second, although a number of cul­ been an ongoing concern for many practition­ turally-relevant variables have been shown ers and health professionals. While recent to be associated with ri sk-taking and social statistics indicate slight declines in the preva­ competence behaviors among ethnically di­ lence rates of problem behaviors among verse youth, many mainstream theories do Latinos during the last several years, the per­ not account for culturally-relevant variables. centage of Latino adolescents who engage In order to enrich our theoretical understand­ in problem behaviors such as physical fight­ ing of problem and conventional behaviors ing, cocaine use, substance use (alcohol and among ethnically diverse youth, it is impor­ marijuana) on school property, and teen preg­ tant that researchers closely examine exist­ nancy remain high relative to White adoles­ ing models and arrive at alternative formula­ cents (CDC 2002). Furthermore, Latinos rep­ tions that take into account culturally-relevant resent one of the largest and fastest grow­ methodological and conceptual factors. ing ethnic minority groups in the U.S. (U .S. The present paper examines Problem Be­ Census Bureau 2001a). Hence, issues re­ havior Theory (PBT: Jesser & Jesser 1977; lated to social development and cultural ad­ Jesser, Donovan, & Costa 1991 ) and its rele­ justment will become increasingly relevant vance to Latino youth. The focus of the in research and service delivery with this pop­ present paper is PBT because it is one of ulation. Finally, Latino youth make up an in­ the most widely studied conceptual and influ­ creasing percentage of the U.S. population ential models in adolescent and young adult (U .S. Department of Health and Human Ser­ development. This paper begins with a brief vices 2001 ). Taken together, these statistics overview of relevant problem behavior re­ highlight the importance of furthering re­ search with Latino adolescents. Methodo­ searchers' and health professionals' under­ logical considerations in the conception of standing of problem behaviors in Latino PBT, as well as culturally-relevant factors youth. known to be linked to Latino adolescents' There have been a number of theories risk-taking behaviors are presented. Finally, that have been developed and tested that implications for theory are discussed. attempt to explain adolescent social behav­ iors (e.g. , Hawkins, Catalano, & Miller 1992). OVERVIEW OF PROBLEM BEHAVIOR However, there are several limitations in the THEORY existing literature on adolescent behaviors Problem Behavior Theory is a psychoso­ that are worth highlighting. First, many thee- cial model that attempts to explain behav- 56 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology ioral outcomes such as substance use, de­ tion of independence from parents and soci­ viancy, and precocious sexual behavior etal influence. In contrast, conventional be­ among adolescents (Jessor & Jessor 1977) havior structures consist of behaviors ori­ and young adults (Jessor et al 1991 ). The ented towards society's traditional standards model includes two antecedent-background of appropriate conduct such as church at­ factors, and three independent, but related tendance and high academic performance. systems of psychosocial components. Ante­ An important premise of PBT is that prone­ cedent-background variables consist of de­ ness to specific problem behaviors entails mographic factors (e .g., parental education involvement in other problem behaviors and and occupational levels, family structure), less participation in conventional behaviors and socialization influences that encompass (Jessor 1987). This premise is a central te­ parental ideologies with respect to tradition­ net of PBT and has important implications ality, religiosity, tolerance for deviance, home for research and intervention. Because of the climate, and peer and media influences. The significance of this claim, the present paper three psychosocial systems consist of per­ will focus on examining the validity of this sonality, perceived environment, and behav­ tenet, especially when applied to understand­ ior systems, each including variables that ing Latino youth . contribute to the likelihood that problem be­ haviors will occur. CONCEPTUAL ISSUES IN PROBLEM The personality system consists of three BEHAVIORS components. The Motivational-Instigation Definitions and Classifications Structure encompasses an individual's set According to Jessor, problem behavior is of values and expectations regarding aca­ defined as demic achievement, independence, and level of peer affection. The Personal Belief behavior that departs from the norms-both Structure consists of a person's social criti­ social and legal-of the larger society; it is cisms (i.e., the acceptance or rejection of behavior that is socially disapproved by the society's norms, values, and practices), level institutions of authority and tends to elicit of alienation from others, self-esteem, and some form of social control response internal/external locus of control. The third whether mild reproof, social rejection , or component of the personality system is the even incarceration. (1987 332) Personal Control Structure, which entails an individual's attitude and tolerance towards These behaviors include, but are not limited deviance, level of religiosity, and positive func­ to, substance use (e.g., tobacco, alcohol, and tions (e.g., drinking reduces stress and anxi­ illicit drugs), general deviant behaviors (e.g ., ety) and perceived effects of risk behaviors. vandalism, stealing), and precocious sexual The perceived environmental system con­ intercourse. sists of distal and proximal components that reflect social influences. The distal structure PBT and Latinos is comprised of contextual social factors re­ Relatively few studies have directly exam­ garding an individual's level of parental-, fa­ ined the generalizability of PBT in Latino milial-, or peer-orientation . In contrast, the youth . A few studies have examined the rela­ variables in the proximal structure encom­ tions among problem behaviors and the fac­ pass approval or disapproval from parents, tor structure posited by PBT in Latino youth. family, or peers regarding problem be­ For example, Dinh and colleagues' (2002) havior(s) . study with Latino youth (primarily Mexican The third component of Problem Behav­ American youth) revealed that substance use ior Theory, the behavior structure system, attitudes, association with delinquent peers, consists of problem and conventional be­ externalizing problem behaviors, and gang havioral structures that work in opposition to involvement loaded on a single-factor which one another. Examples of the problem be­ was descriptive of "problem behavior prone­ havior structure include illicit drug use, to­ ness." Furthermore, findings indicated that bacco use, alcohol abuse, deviancy, and pre­ problem behavior proneness was stable over cocious sexual behavior. Jessor and his col­ a one-year time period. leagues postulate that these problem be­ Although prior research has shown evi­ haviors stem from an individual's affirma- dence for problem behavior proneness Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 57 among Latino youth, it has been suggested ation of the definition of risky-sexual behav­ that the structure of problem behaviors may iors in this population is imperative. Taken differ across ethnic groups (Barrera, Biglan, together, it is important to consider the defi­ Ary, & Li 2001). Newcomb noted that, nition of problem and normative behaviors in the appropriate ethnic and developmental On the basis of the differential association contexts. observed between drug use and other Another variable in the behavior structure types of delinquency or problem behav­ system of PBT worth noting is general devi­ iors ... it seems possible that this syndrome ant behaviors. According to Jessor and col­ may have different patterns for various eth­ leagues (1977, 1991 ), general deviant behav­ nic groups. (1995 126) iors are behaviors that violate social and le­ gal norms but do not involve substance use Indeed, prior research with Latino youth and misuse. Jessor and Jessor (1977) showed marijuana use loaded higher on a measured general deviance by using a second factor with arrest history than on the multi-item scale that assessed diverse prob­ first factor with alcohol and tobacco use (Ebin, lem behaviors such as trespassing, vandal­ Sneed, Morisky, Rotheram-Borus, Magnus­ ism, lying, stealing, threatening a teacher, son, & Malotte 2001 ). Finally, a study with and skipping school without a valid excuse. Latino college students showed variations Jessor and colleagues (1991) used similar in the number of factor structures in problem items (except skipping school and threaten­ and conventional behaviors across Latino ing a teacher) and added initiating fights to subgroups, suggesting that the behavior sys­ measure general deviant behaviors in young tem may operate differently across these adults. Such conception of general deviant groups (Zamboanga, Carlo, & Raffaelli 2004). behavior is consistent with a "syndrome" view On the basis of those and other empirical of problem behaviors; however given the pau­ findings, researchers have raised questions city of research on the structure of problem about the generalizability of PBT to other eth­ behaviors in Latino adolescents and young nic groups (e.g., Barrera et al 2001 ; Mitchell adults, it remains unclear whether this pur­ & Beals 1997; Newcomb 1995; Williams, ported structure of general deviancy has func­ Ayers, Abbott, Hawkins, & Catalano 1996). tional and structural equivalence with this One area of particular concern is that the population. operational definition of problem behavior may be too narrow to apply to Latino groups. PROBLEM BEHAVIORS IN LATINOS: For example, in their follow-up work with SOCIAL CONSIDERATIONS young adults, Jessor and colleagues omit­ Jessor and colleagues (1991 ) acknowl­ ted sexual intercourse from the behavior edged the limited attention given to "distal" structure system. They argued that factors (i .e., social environment) in their early formulation of PBT. They noted that such with development from adolescence to young adulthood, the very same behavior­ decision was partly influenced by the logic sexual intercourse-shifts from problem to of causality and partly by the relative homo­ normative behavior. (1991 24) geneity of social background of our in­ school, relatively middle-class population of However, it should be noted that risky sexual yo uth . (1991 19) behaviors are problematic, particularly among Latino populations (Raffaelli, Zam­ Although Jessor and colleagues (1991 ) in­ boanga, & Carlo 2005). For example, Latinas cluded a number of demographic variables are more likely to engage in unprotected in­ (e.g., education, occupation, religious affil i­ tercourse than women from other ethnic ation, and family structure) in their PBT model, groups (cf. Raffaelli et al2005). Furthermore, there are several cultural and sociodemo­ sexually active Latina college students are graphic factors that are also relevant in our less likely to use condoms than their female understanding of problem behaviors in Latino counterparts from other ethnic groups (CDC youth. The following section provides ex­ 1997). Although there might be many rea­ amples on the relevance of social, cultural, sons that account for frequent unsafe sexual and demographic factors to problem behav­ behaviors among Latinas, careful consider- iors such as substance use in Latino youth. 58 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Peer Influences ant peers. Consistent with this suggestion , Researchers have argued that one of the research with Hispanic, White, and Ameri­ strongest influences on youth problem be­ can Indian adolescent boys and girls indi­ haviors is involvement with delinquent peers cated that inadequate parental monitoring (Barrera et al 2001 ). Consistent with prior was directly and indirectly (through its effect studies in non-Latino populations, research on adolescents' association with delinquent with White, Hispanic, and American Indian peers) associated with problem behaviors adolescents revealed that association with (Barerra et al 2001) . delinquent peers was associated with in­ volvement in general problem behaviors Family Relationships (substance use, poor academic perfor­ According to scholars, family plays a cen­ mance, and antisocial behavior) (Barrera et tral role in shaping Latinos' experiences al 2001). Research with Latino adolescents (Carlo, Carranza, & Zamboanga 2002; Fu­ revealed strong relations between peer al­ ligni, Tseng, & Lam 1999). This value is re­ cohol and peer marijuana use and adoles­ flected in familism-the strong identification cent drinking and marijuana use, respectively with, and attachment and loyalty to , one's (Frauenglass, Routh , Pantin , & Mason 1997). family, which has also been well-documented A large-scale study with African American and among Latinos (e.g., Sabogal, Marin , Otero­ Puerto Rican adolescents showed tolerance Sabogal, Marin, & Perez-Stable 1987; Suarez­ of deviance and peer modeling of substance Orozco & Suarez-Orozco 1995). The quality use and deviance to be predictive of sub­ of family and peer relationships is an impor­ stance use and delinquency (Brook, White­ tant consideration in youth problem behav­ man, Balka, & Cohen 1997). This study also iors in Latino populations. While parent in­ examined differences in the magnitude of fluences, particularly parent support and con­ the relations between risk factors with sub­ trol are relevant aspects of the perceived en­ stance use and delinquency; results showed vironment system of PBT, they are consid­ that association with marijuana using peers ered "distal" structures in the model. Further­ was more strongly related to substance use more, the personality system of PBT entails than delinquent behaviors, while involvement values and expectations placed on indepen­ with deviant peers had a stronger associa­ dence. Traditional Latino values of familism tion with delinquency than substance use. In and cultural emphasis placed on family inter­ short, research findings with Latino adoles­ dependence and connectedness thus war­ cents are consistent with PBT's hypothesized rants further consideration with respect to influence of personality (e.g ., attitudinal tol­ health-risk behaviors among Latino youth . erance of deviance) and the perceived envi­ Negative family relationships such as in­ ronment (peer approval and models of prob­ terparental conflict can threaten Latino ado­ lem behavior) on problem behaviors in this lescents' emotional well-being and therefore population. increase their risk for problem behaviors Although negative peer influences have (see also research on family violence by Cae­ been linked to youth problem behaviors in tano, Field , & Nelson 2003 and literature re­ Latinos, family factors can influence such re­ view by Salzinger, Feldman , Stockhammer, lations. For example, one study showed that & Hood 2002). For example, one study re­ family support moderated the relations be­ vealed that Latino adolescents who were tween peer substance use (marijuana and exposed to parental arguments about them tobacco) and adolescent substance use. In also reported higher levels of substance use other words, as the number of substance (alcohol , tobacco, and marijuana use) and using peers increased, higher levels of fam­ elevated sexual experience (Tschann, Flores, ily support were related with lower levels of Marin, Pasch, Baisch, & Wibbelsman 2002). youth substance use (Frauenglass et al Furthermore, Tschann et al (2002) found that 1997). Researchers have also highlighted adolescents who were more involved in their the impact of family conflicts on family rela­ parents' conflicts (e.g. , siding with a parent) tionships (e.g ., open parent-youth communi­ also reported higher levels of emotional dis­ cations) which in turn can lead to inadequate tress and in turn, experienced higher levels parental monitoring (see Barerra et al 2001 ). of substance use and had more sexual ex­ Inadequate parental monitoring increases perience. Hence consistent with prior re­ the likelihood of youth involvement with devi- search with non-Latino adolescents, inter- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 59 parental conflict is associated with negative would broaden the utility of PBT for all ethni­ psychological and behavioral outcomes in cally diverse populations. Latino adolescents (Tschann et al 2002). Positive family relationships can help pro­ Immigration and Acculturation tect Latino adolescents' from becoming in­ Another set of factors that must be con­ volved in problem behaviors. Studies have sidered when theorizing about youth risk be­ shown that higher family support, strong fam­ haviors in Latinos are those associated with ily connectedness, and higher parental moni­ the dynamics of culture adaptation and toring is associated with lower alcohol and change. Acculturation2 is the process of psy­ substance use and less gang involvement chological and behavioral adaptation that among Latinos (e.g., Frauenglass et al1997; occurs when two cultures come into contact, Kerr, Beck, Shattuck, Kattar, & Uriburu 2003). as happens when immigrants arrive in a new A study with Latino adolescents revealed a country or one group is colonized by another significant association between positive (Marin & Marin 1991 ). Researchers have ar­ family attitudes (i.e. , familism) and lower gued that odds for lifetime marijuana use (but only for those who possessed high or moderate [a]cculturation is one of the most important knowledge of the drug) (Ramirez, Grano, factors that explain risk behavior and health Quist, Burgoon, Alvaro, & Grandpre 2004). In status of Latinos. (Suarez & Ramirez 1999 essence, family relationships can greatly 120) impact Latino adolescents' development and thus warrant important consideration when Research suggests that immigration sta­ theorizing about problem behaviors with this tus and acculturation (commonly assessed population. through language use or generation status) into U.S. society play a role in youth problem PROBLEM BEHAVIORS IN LATINOS: behaviors in Latinos (e.g., Ebin et al 2001 ; SOCIODEMOGRAPHIC AND CULTURAL see also De La Rosa 2002, and Epstein, CONSIDERATIONS Botvin, & Diaz 2001 for reviews). For example, Socioeconomic Status compared to adolescents with two or more Latinos are overrepresented in the lower years of U.S. residency, foreign-born Cuban socioeconomic sector in the U.S. According and other Hispanic adolescents who had to the U.S. Census Bureau (2001 b) , 22.8 lived in the U.S. for two years or less had the percent of Hispanics were living in poverty in lowest overall lifetime prevalence rates of 1999, compared to only 7.7 percent of non­ substance use (Khoury, Warheit, Zimmer­ Hispanic Whites. The number of Latino fe­ man, Vega, & Gil1996). In another study, Ep­ male-headed households is high and when stein et al's (2001) large-scale, longitudinal employed, Latino women are likely to work investigation with Latino adolescents re­ in low-status, low-paying jobs (Padilla & vealed that Latino adolescents who spoke Salgado de Snyder 1995). Such challenges English (only or mostly) with their parents may contribute to parental absence, reduced reported higher levels of marijuana use than maternal involvement, and increased family adolescents who Spanish (only or mostly) distress. Consistent with this suggestion, with their parents. Moreover, one-year follow­ research with a nationally representative up results showed that Latino adolescents sample of White, Black, and Hispanic ado­ who spoke English with their parents en­ lescents showed that living in a single-par­ gaged in higher levels of polydrug use than ent home and being Latino were associated those who spoke Spanish with their parents with higher levels of involvement in violence, (Epstein et al 2001). Finally students who independent of income (e.g., Blum, Beuhring, spoke both English and Spanish with their Shew, Bearinger, Sieving, & Resnik 2000; parents reported higher lifetime polydrug use Smith & Krohn 1995). Although these stud­ compared to those who spoke Spanish with ies highlight the importance of the family in their parents (Epstein et al 2001 ). Although preventing Latino youth risk-taking, it can be the majority of studies show positive rela­ argued that the socioeconomic consider­ tions between acculturation and substance ations outlined above are not unique to use, it should be noted that not all studies Latino adolescents, and that a fuller consid­ support the contention that acculturation is eration of the impact of economic factors associated positively with Latino adolescent 60 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology problem behaviors (e.g., Ramirez et al 2004; negative peer hassles mediated the relation Zapata & Katims 1994; see De La Rosa between acculturation (as measured by lan­ 2002, for a review) . Thus, it is clear that the guage use and generational status) and de­ association between acculturation and sub­ linquency in Mexican American adolescents. stance use is complex. In another study, Dinh et al (2002) showed The majority of past acculturation and that parental involvement mediated the rela­ problem behavior research has focused pri­ tion between acculturation and problem be­ marily on the direct relation between these havior proneness a year later. two variables (or some other outcome vari­ able) (Dinh et al 2002); hence pathways of Gender mediation or mechanisms that explain the Differential standards and values regard­ link between acculturation and problem be­ ing alcohol use are known to vary by gender haviors remain unclear (McQueen, Getz, & among Hispanics. In general, women and Bray 2003). Given the central role of family in children are typically socialized to abstain Latino culture and the influence offamily vari­ from drinking (Gilbert & Collins 1997). Flores­ ables (e .g., parental monitoring and involve­ Ortiz's (1994) study with Latina adolescents ment, family relationships) on problem be­ in California noted inherent gender double haviors in Latino youth , the mediating influ­ standards regarding drinking among Mexi­ ence of family factors warrants much needed can American families. The general reported attention (Dinh et al 2002). Indeed, scholars consensus among the Latinas in this study contend that children acculturate faster than was that Latino cultures condoned drinking their parents (McQueen et al 2003; Padilla & among men but not women. Consistent with Salgado de Snyder 1995), and highlight this suggestion, it is argued that gender can Latino parents' concerns regarding their chil­ moderate the relation between acculturation dren's acculturation into mainstream Ameri­ and substance use in Latino youth and young can society: adults. Latino youth and young adults accul­ turate into a U.S. culture that is less prohibi­ To say that parents do not get concerned tive (compared to traditional Latino cultures) about the changing family values would be about the use of alcohol by women. As such, to ignore a real tension that haunts immi­ Latinas may modify their drinking behavior grant parents in particular. During the pre­ by adopting more liberal attitudes and behav­ school years, parents are able to exert a iors toward drinking. Conversely, Latino youth strong influence on their children . As chil­ and young adults who acculturate to the U.S. dren get older, parents fear that they will are therefore less likely to undergo signifi­ become too Americanized and forget their cant changes in their drinking because they language and culture . (Delgado-Gaitan 1993 are acculturating into a U.S. society where, 425) much like their Latin country of origin, there are no strict cultural sanctions against drink­ Parental concerns may give rise to parent­ ing for males. child conflicts, especially if pressures to as­ Research with Mexican American adoles­ similate outside the home are present. Par­ cents highlighted other gender differences ent-child acculturation gaps are believed to with respect to the link between accultura­ give rise to problems in family communica­ tion and problem behaviors (McQueen et al tion and parent-child conflicts (Negy & Woods 2003). Their findings revealed that family 1992; Szapocnik, Santisteban, Rio, Perez­ conflict mediated the association between Vidal , & Kurtines 1989). Such challenges can acculturation (as measured by language) disrupt family connectedness which in turn and marijuana use and deviant behavior for places these youth at high risk for problem males, but not females. Furthermore, gen­ behaviors. eration status was unrelated to problem be­ Additional empirical evidence has high­ haviors, family conflict, and separation for lighted the mediating role of family and peer males; however for females, acculturation relationships with respect to the association (as measured by generation status) was in­ between acculturation and problem behav­ directly associated with substance use and iors. The Samaniego and Gonzales (1999) deviant behaviors through its effect on family study revealed that family conflict, low mater­ conflict and separation. nal monitoring, inconsistent discipline, and While prior research has indicated gen- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 61 der differences in adolescent substance Associations Between Problem and use, researchers have argued that gender Conventional Behaviors in Latinos alone has limited utility in predicting most According to the behavioral structure sys­ substance use outcomes in Latino youth tem of PBT, there is a direct relation between (Kulis, Marsiglia, & Hurdle 2003). Research problem and conventional behavior such that with Mexican American adolescents in the youth can be expected to be more likely to Southwest revealed that gender identity (ag­ engage in one but not both types of social gressive masculinity, assertive masculinity, behaviors. This assumption has implications affective femininity, and assertive femininity) for theory, methodology, and intervention pro­ was a stronger predictor of substance use grams and policy making. First, the assump­ than gender alone. In particular, aggressive tion implies understanding the development masculinity was associated with increased of conventional or problem behaviors will risk for substance use, regardless of accul­ lead to an understanding of the development turation level. Findings also revealed that af­ of both types of behaviors. Second, mea­ fective femininity and submissive femininity sures that tap into either set of social behav­ appear to have a protective effect against iors will suffice in our understanding of youth substance use for Mexican American ado­ development. And third, the tenet suggests lescent boys and girls, particularly among that programs or policy decisions designed those who are highly acculturated (Kulis et to address either the promotion of conven­ al 2003). Hence in order to make PBT more tional behaviors or the reduction of problem applicable to Latino adolescents, it is impor­ behaviors will affect both set of behaviors. tant that gender-identity (not just gender However, there are concerns that this as­ alone per se) and acculturation factors be sumption may oversimplify the challenges considered as integral aspects of PBT's con­ of understanding problem and conventional ceptual model. behaviors. The complexity and challenges of under­ CONCEPTUAL ISSUES IN CONVENTIONAL standing problem behaviors among Latinos BEHAVIORS can be exemplified by observing behaviors Definitions and Classifications among gang members. Although gang mem­ Jesser and colleagues (1991) define bers often exhibit antisocial behaviors, it is conventional behaviors as behaviors consis­ also clear that gang members frequently en­ tent with societal and legal norms as en­ gage in sharing, comforting, protective and dorsed by social institutions of authority. The supportive behaviors, and even risk their own most common operational definition is lives for the good of the group or for others in church attendance, although political and the group (i.e., altruistic behaviors). Those health behaviors are included in their con­ exhibited behaviors suggest that gang mem­ ceptual model. These scholars acknowl­ bers are capable of prosocial (i .e., behav­ edge that their primary interest is in "prone­ iors that benefit others) and socially accept­ ness to behavior system." That is, ultimately, able behaviors; however, prosocial behav­ the focus is on the individual's involvement iors are often reserved for members of their in problem behaviors relative to his or her ingroup and antisocial behaviors are often involvement in conventional behaviors. manifested towards outgroup members (in­ Therefore, according to PBT scholars, youth cluding majority society). who are engaged in problem behaviors are The research evidence that supports the less likely to engage in conventional behav­ incongruity between antisocial and prosocial iors. However, this notion does not neces­ behaviors is well-documented. Youth who sarily hold up conceptually nor when one engage in prosocial behaviors do not auto­ considers existing empirical literature. More­ matically engage less in antisocial behav­ over, the emphasis on church attendance as iors and vice versa. Scholars have suggested one of the primary markers of conventional­ that children sometimes engage in both pro­ ity in the behavior structure system presents social and antisocial behaviors in order to conceptual and methodological challenges have greater impact on their peer group activ­ for researchers interested in understanding ities and to gain approval from their peers positive youth development among ethnic (Carlo 2006). Furthermore, empirical re­ minorities. search on the association between prosocial and antisocial behaviors (such as aggres- 62 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology sion) often yields modest correlations (Carlo, may help to reinforce negative stereotypes, Hausmann, Christiansen, & Randall 2003; racist attitudes, and stigmatization that are Crick & Grotpeter 1995; Wyatt & Carlo 2002). already prevalent in sections of our society. There are some youth who engage in high A broader and more comprehensive ap­ levels of both antisocial and prosocial be­ proach to studying problem behaviors would haviors and there are other youth who en­ create opportunities to deepen our under­ gage in low levels of both sets of behaviors. standing of positive behavioral outcomes as The research suggests that aggression is well as promote our understanding of vari­ not just the flip side of prosocial behaviors ables that could buffer negative symptoma­ or vice versa. Thus, existing developmental tology. Furthermore, studying a broader ar­ research on prosocial and aggressive be­ ray of behaviors would provide an ecological haviors suggests that the relations between valid and more balanced understanding of conventional and problem behaviors will ei­ Latino youth development that acknowledges ther be nonsignificant or modest at best. the strengths and complexity of these indi­ There is additional evidence on the mod­ viduals. In addition, movement towards more est and sometimes nonsignificant relations complex models of Latino youth development between conventional and problem behav­ would help us account for the wide individual iors. For example, a perusal of the relations differences in social behaviors among La­ between conventional and problem behav­ tinos. Thus, there is great importance to un­ iors (a number of measures of substance derstanding positive social development, and alcohol use and deviant behaviors) including conventional behaviors among showed that the correlations ranged from Latinos. -.13 to -.38 (mean correlation = -.24) in a One major limitation of PBT is the some­ sample of high school students and from what narrow operational definition of conven­ -.06 to -.24 (mean correlation = -.15) in a tional behaviors. Turiel (1983) and his col­ sample of college students (Jessor et al leagues proposed that behaviors can be di­ 1991 ; see also Costa, Jessor, Fortenberry, & vided into several categories depending Donovan 1996). Similarly, in a sample of La­ upon the obligatory nature and the surround­ tino college students, Zamboanga et al ing social norms. Conventional behaviors (2004) found modest relations between prob­ were defined as actions guided by prevail­ lem and conventional behaviors. Ebin et al ing informal social norms and customs. (2001) yielded evidence that adaptive health Moral behaviors are defined by formal soci­ behaviors were modestly (mostly nonsignifi­ etal laws or rules that have strong socially cantly) associated with problem behaviors obligatory characteristics. Actions in the per­ among Latino adolescents. These findings sonal domain reflect individual preferences demonstrate a modest overlap between con­ and biases with no grounding in formal so­ ventional and problem behaviors. cietal rules or laws. Finally, prudential ac­ tions are those behaviors that subscribe to Towards a Broader Conception of considerations of the child's safety or well­ Conventionality being. An additional set of behaviors is pro­ Scholars have long noted the overempha­ social behaviors (i .e., behaviors intended to sis on negative and risk behaviors by re­ benefit others). Those latter behaviors can fit searchers who study ethnic minority popula­ under the rubric of either conventional or tions (Allen & Mitchell 1998; McLoyd 1990). moral domains (Carlo 2006) . The strength There have been a number of important con­ of this typology is that conventional behav­ sequences that stem from this kind of re­ iors are not simply considered as a unidi­ search emphasis. First, researchers have mensional construct. Instead, behaviors are noted the lack of theories that foster our un­ classified accordingly to reflect the social derstanding of normative development contextual circumstances, the underlying in­ among those youth. Second, some of the tentions, and the consequences. existing research has been characterized as In an attempt to broaden their focus to reinforcing or creating deficit models-mod­ additional conventional behaviors, Jessor, els that depict ethnic minorities as deficient Turbin , and Costa (1998) examined the asso­ relative to non-ethnic minority youth . Third, ciations between several conventionality-re­ an overemphasis on negative and risky be­ lated variables (e.g. , school and parent orien­ haviors among ethnic minority populations tation , positive relations with adults, friends Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 63 as models for conventional behaviors, pro­ Furthermore, peers provide direct and indi­ social activities, and church attendance) and rect social feedback (social rewards and health-related risk factors (e.g., stress, peer punishers) on prosocial and conventional pressure susceptibility, parents smoking be­ behaviors. Moreover, because youth con­ haviors) and health enhancing behaviors stantly engage in social comparisons, peers (e .g., seat belt use, good dental hygiene, can influence youth by providing standards sleep, exercise). The researchers demon­ and norms for social behaviors (Carlo et al strated that conventionality-related behaviors 1999). The influence of peers is likely to be were positively related to health-enhancing exacerbated or mitigated by the degree of behaviors. perceived similarity or admiration for the peer Although the aforementioned study is one or peer group. of the first to directly examine positive traits One specific dimension along which the and behaviors from a PBT perspective, there strength of the influence of peers might vary are several issues worth noting. First, church among Latino youth is ethnic identity. Latino attendance was included as a predictor youth might be more susceptible to peer in­ rather than a criterion variable as part of the fluence to the degree that the youth identifies behavioral system as proposed by PBT. Sec­ with their culture of origin and the peer group ond, the conceptualization of the study de­ reflects the strength of that ethnic identity. sign was somewhat ambiguous because the Peer groups that exhibit behaviors or cultural conventionality-related variables were con­ pride that reflect closely the youth's ethnic ceptualized as protective factors but schol­ identity may be more apt to their influence ars have noted that protective factors are vari­ and vice versa. Furthermore, the influence of ables that protect against negative symp­ peers may also depend on the congruency tomatology under adverse conditions (e.g., between the youth's ethnic identity and his Masten & Reed 2002). The adverse condi­ or her parents' ethnic identity. The greater tions of the sample in the study were not the disparity, the greater the distance be­ established-thus, the operationalization of tween parents and their youth, and in turn, protective factors is subject to question. And this might lead to greater impact by the peer third, although the study of protective and group. Although some research has been buffer factors is important in its own right, conducted on peer influence in Latino gang understanding the development of conven­ affiliation, research on the influence of peers tional behaviors requires that conventional on Latino youth normative development is behaviors are the focus outcome of research. lacking. Further research is needed to ex­ Thus, there is a need to carefully distinguish amine these processes among normative between the different system levels of the groups of Latinos. structure of PBT and to consider the broad array of behaviors that fall under the rubric of Family Relationships conventional behaviors. The importance of In contrast to the lack of research on the these issues becomes more evident when influence of peers on positive Latino youth we attempt to understand the development development, there is a body of research on of conventional behaviors among Latino the influence of family on positive Latino youth. youth development. Conceptually, parents and family members are expected to impact CONVENTIONAL BEHAVIORS IN LATINOS: Latino children's development, particularly SOCIAL CONSIDERATIONS early in life. By adolescence, youth renegoti­ Peer Influences ate their relationships with their parents and Although there is a substantial body of family members and become increasingly evidence suggesting that peers influence the influenced by peers (Youniss 1980). development of prosocial behaviors (see However, as mentioned previously, schol­ Carlo, Fabes, Laible, & Kupanoff 1999), re­ ars have noted that close family relationships search on the influence of peers on prosocial are the hallmark of many Latino families and and conventional behaviors among Latinos that most Latino families foster familial inter­ is nonexistent to our knowledge. However, dependence (e.g., Knight, Bernal, & Carlo as peers become more influential with age, 1995; Raffaelli, Carlo, Carranza, & Gonzalez­ one might expect that peers serve as mod­ Kruger Forthcoming). Furthermore, studies els for prosocial and conventional behaviors. suggest that parents may still be influential 64 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

even in adolescence (Carlo et al 1999) and turative stress might have detrimental im­ this might be particularly true among some pact on prosocial and conventional behav­ Latinos. For example, there is some research iors in Latino youth . One might discover that shows that Latinas (relative to Latinos) strong associations between SES and pro­ remain closely monitored by their parents social and conventional behaviors as a re­ and maintain close relationships with their sult of the strong association between accul­ parents (see Carlo et al 1999). Recently, de turative stress and prosocial and conven­ Guzman and Carlo (2004) showed that fam­ tional behaviors and the association be­ ily adaptability was associated positively with tween SES and acculturation. prosocial behaviors in a sample of Latino Consistent with expectations, there is evi­ adolescents. The finding suggests that, dence that more acculturated Latinos are among Latinos, families who are flexible in less cooperative and prosocial and less ac­ responding to the youth 's specific circum­ culturated individuals are more competitive stances may be more adept at fostering pro­ (de Guzman & Carlo 2004; Knight & Kagan social behaviors. Given the potential chal­ 1977). However, to our knowledge , there is lenges posed by intergenerational and inter­ no research that examines associations cultural value conflicts, family adaptability among acculturative stress, SES , and pro­ may become a more pressing characteristic social and conventional behaviors. Clearly, to foster positive behavioral outcomes. more research is needed to examine those possibilities. CONVENTIONAL BEHAVIORS IN LATINOS: SOCIODEMOGRAPHIC AND CULTURAL Gender CONSIDERATIONS According to gender socialization theo­ Socioeconomic Status rists (Gilligan 1982; Maccoby & Jacklin 1974), Researchers have shown links between girls are socialized differently than boys and family economic status and maladjustment this has important implications for the devel­ in children and adolescents (e .g. , Elder & opment of conventional and pro social behav­ Conger 2000). According to those scholars, iors. For example, girls are encouraged to economic strain on the family fosters paren­ express sadness more than boys, which is tal depression, which in turn, impedes effec­ associated with prosocial responding . Fur­ tive parenting and leads to negative sympto­ thermore, in many societies, girls are as­ matology in children . However, to our knowl­ signed to caring and nurturing responsibili­ edge, no research has been conducted to ties and expected to fulfill those duties more examine whether there is a similar mecha­ than boys (see Carlo et al 1999). Indeed, nism that impacts prosocial or conventional prosocial and conventional behaviors (e .g. , behaviors. Research examining the poten­ comforting, caring) are perceived as more tial impact of economic strain on prosocial consistent with girls' gender role than boys' and conventional behaviors among Latinos gender role (Eisenberg & Fabes 1998). Par­ is therefore needed. ticularly among Latino families, gender typed There is another mechanism that would expectations are strong . Although there is imply a strong association between SES and little or no research that focuses on Latinos, prosocial and conventional behaviors, espe­ scholars have noted that Latinas are prob­ cially among Latinos. As new Latinos enter ably more strongly encouraged to fulfill fam­ the U.S., many immigrants acculturate to the ily and household responsibilities (includ­ majority society. However, acculturation of­ ing caring and nurturing siblings) than boys ten induces acculturative stress (i.e ., taxing (Knight et al1995). Other scholars have noted demands that result from adapting to the new parental expectations for Latino boys to ex­ majority society). Thus, in addition to pos­ press strong masculine-typed traits and be­ sible economic strain from low starting haviors. Taken together, those practices and household income, Latino families might ex­ expectations foster greater likelihood of pro­ perience stress resulting from discrimina­ social and conventional behaviors in Latino tion, prejudice, or harassment experiences. girls rather than Latino boys. Therefore, for new immigrants, there might There is considerable empirical evidence be strong correlations between SES , accul­ that girls exhibit higher levels of prosocial turation , and acculturative stress. Any nega­ behaviors than boys, especially during ado­ tive consequences that result from accul- lescence (e .g., Carlo 2006). During late ado- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 65 lescence and young adulthood, however, REFERENCES scholars have found that boys do express Allen L & C Mitchell. 1998. Racial and ethnic differ­ higher levels of instrumental and risky pro­ ences in patterns of problematic and adaptive social behaviors than girls; whereas, girls development: an epidemiological review. Pp. express higher levels of nurturing and car­ 29-54 in Studying Minority Adolescents: Con­ ceptual, Methodological, and Theoretical Is­ ing prosocial behaviors than boys (Eagly & sues. V McLoyd & L Steinberg, eds. Mahwah, Crowley 1986). Unfortunately, to our knowl­ NJ: Lawrence Erlbaum Associates. edge, studies that directly examine gender Barrera M, A Biglan, D Ary, & F Li. 2001 . Replica­ differences in prosocial and conventional tion of problem behavior model with American behaviors among Latino youth are lacking. Indian, Hispanic, and Caucasian youth. J Early Adolescence 21 133-157. SUMMARY AND CONCLUSIONS Berry JW. 2002. Conceptual approaches to un­ Taken together, research on problem and derstanding acculturation. Pp. 17-37 in Accul­ conventional behaviors among Latino turation: Advances in Theory, Measurement, and Applied Research. KM Chun, PB Organista, samples suggest a more differentiated be­ & G Marin, eds. Washington, DC: American Psy­ havior system than that which is proposed chological Association. by PBT scholars. The conceptual and meth­ Blum RW, T Beuhring, ML Shew, LH Bearinger, RE odological limitations of PBT highlighted in Sieving, & MD Resnik. 2000. The effects of this paper have important implications for race/ethnicity, income, and family structure on the development of valid measurement tools. adolescent risk behaviors. Amer J Public The existing research suggests that further Health 90 462-468. assessment of the psychometric properties Brook JS, M Whiteman, EB Balka, & P Cohen. 1997. of measures to use with Latinos is neces­ Drug use and delinquency: Shared and un­ shared risk factors in African American and sary. These issues, if not properly addressed, Puerto Rican adolescents. J Genetic Psychol­ could create stigmatization and rejection to­ ogy 158 25-39. wards Latinos if inadequate measures Caetano R, CA Field, & SNelson. 2003. Associaiton present Latinos as problematic or "prone" to between childhood physical abuse, exposure problem behaviors. Finally, the study of prob­ to parental violence, and alcohol problems in lem and conventional behaviors in Latinos adulthood. J Interpersonal Violence 18 240- requires careful consideration of peer, fam­ 257. ily, sociodemographic, gender, and culture­ Carlo G. 2006. Care-based and altruistically-based relevant (e.g. , acculturation, ethnic identity) morality. Pp. 551-579 in Handbook of Moral Development. M Killen & JG Smetana, eds. variables. Although such research increases Mahwah, NJ: Lawrence Erlbaum Associates. the complexity of existing theory and meth­ Carlo G, MA Carranza, & BL Zamboanga. 2002. odology, the resulting empirical research will Culture, ecology, and Latinos on the Great more adequately reflect the multidimen­ Plains: an introduction. Great Plains Res J sional nature of Latino youth development. Natural & Soc Sci 12 3-12. Carlo G, RA Fabes, OJ Laible, & K Kupanoff. 1999. ENDNOTES Early adolescence and prosocial/moral behav­ 1 The term "Latino/a" are used in this paper to ior II: the role of social and contextual influ­ refer to persons of Mexican, Puerto Rican, ences. J Early Adolescence 19 133-147. Cuban, or other Central and South American, Carlo G, A Hausmann, S Christiansen, & BA Randal l. and/or other Spanish origin with the understand­ 2003. Sociocognitive and behavioral correlates ing that there are cultural differences between of a measure of prosocial tendencies for ado­ Latino subgroups. We also used the terms lescents. J Early Adolescence 23 107-134. Latino and Hispanic interchangeably in the text Center for Disease Control. 1997. Youth risk be­ of the manuscript. havior surveillance: national college health risk 2 Scholars have debated the complexities sur­ behavior survey-United States, 1995. MMWR rounding acculturation as well as appropriate 46 (SS-6). ways to measure it (e.g., language use, gen­ -....,...,....,.,..-· 2002. Youth risk behavior surveillance­ eration status, ethnic identity, ethnic loyalty and United States, 2001. MMWR 51 (SS-4). awareness) in Latino populations. The theo­ Costa FM, R Jessor, JD Fortenberry, & JE Donovan. retical significance surrounding these measure­ 1996. Psychological conventionality, health ori­ ment debates cannot be overemphasized, but entation, and contraceptive use in adoles­ because they are beyond the scope of this cence. J Adolescent Health 18 404-416. paper, readers are directed to other sources Crick NR & JK Grotpeter. 1995. Relational aggres­ (e.g. , Berry 2002; Negy & Woods 1992) for in­ sion, gender, and social-psychological adjust­ depth information about acculturation. ment. Child Development 66 71 0-722. 66 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology de Guzman MRT & G Carlo. 2004. Family, peer, and bridge, MA: Harvard U Press. acculturative correlates of prosocial develop­ Hawkins JD, RF Catalano, & JY Miller. 1992. Risk ment among Latino youth in Nebraska. Great and protective factors for alcohol and other Plains Res J Natural & Soc Sci 14 185-202. drug problems in adolescence and early adult­ De La Rosa M. 2002. Acculturation and Latino ado­ hood: implications for substance abuse pre­ lescents' substance use: a research agenda vention. Psychological Bull 112 64-105. fo r the future. Substance Use Misuse 37 429- Jessor R. 1987. Problem-behavior theory, psycho­ 456. social development, and adolescent problem Delgado-Gaitan C. 1993. Parenting in two genera­ drinking. Brit J Addiction 82 331-342. tions of Mexican American families. lnternat J Jessor R, JE Donovan, & FM Costa. 1991 . Beyond Behavioral Develop 16 409-427. Adolescence: Problem Behavior and Young Dinh KT, MW Roosa, JY Tein , & VA Lopez. 2002. Adult Development. Cambridge, England : Cam­ The relationship between acculturation and bridge U Press. problem behavior proneness in a Hispanic Jessor R & S Jessor. 1977. Problem Behavior and youth sample: a longitudinal mediation model. J Psychosocial Development: A Longitudinal Abnormal Child Psycho/ 30 295-309. Study of Youth . NY: Academic Press. Eagly AH & M Crowley. 1986. Gender and helping Jessor R, MS Turbin, & FM Costa. 1998. Protective behavior: a meta-analytic review of the social factors in adolescent health behavior. J Per­ psychological literature. Psychological Bull1 00 sonality Soc Psycho/ 75 788-800. 283-308. Kerr MH, K Beck, TO Shattuck, C Kattar, & D Uriburu. Ebin VJ , CD Sneed, DE Morisky, MJ Rotheram­ 2003. Family involvement, problem and prosocial Borus, AM Magnusson, & CK Malotte. 2001 . behavior outcomes of Latino youth. Amer J Acculturation and interrelationships between Health Behavior 27 55-65. problem and health-promoting behavior among Khoury EL , GJ Warheit, RS Zimmerman, WA Vega, Latino adolescents. J Adolescent Health 28 & AG Gil. 1996. Gender and ethnic differences 62-72. in the prevalence of alcohol , cigarette, and il­ Eisenberg N & RA Fabes. 1998. Prosocial devel­ licit drug use over time in a cohort of young opment. Pp. 701 -778 in Handbook of Child Hispanic adolescents in South Florida. Women Psychology, Vol. 3: Social, Emotional, and and Health 24 21-40. Personality Development. 5th ed. W Damon Knight G, M Bernal, & G Carlo. 1995. Socialization (series ed .) & N Eisenberg (vol. ed .). NY: John and the development of cooperative, competi­ Wiley. tive, and individualistic behaviors among Mexi­ Elder GH , Jr, & RD Conger. 2000. Children of the can American children. Pp . 85-102 in Meeting Land: Adversity and Success in Rural America. the Challenge of Linguistic and Cultural Diver­ Chicago, IL: U Chicago Press. sity in Early Childhood Education, Vol. 6: Year­ Epstein JA, GJ Botvin, & T Diaz. 2001 . Linguistic book in Early Childhood Education. EE Garcia acculturation associated with higher marijuana & BM McLaughlin, eds. NY: Teachers College and polydrug use among Hispanic adolescents. Press. Substance Use Misuse 36 477-499. Knight GP & S Kagan. 1977. Acculturation of pro­ Flores-Ortiz YG. 1994. The role of cultural and social and competitive behaviors among sec­ gender values in alcohol use patterns among ond- and third-generation Mexican-American Chicana/Latina high school and university stu­ children. J Cross-Cultural Psycho/ 8 273-284. dents: implications for AIDS prevention. lnternat Kulis S, FF Marsiglia, & D Hurdle. 2003. Gender J Addictions 29 1149-1171 . identity, ethnicity, acculturation, and drug use: Frauenglass S, OK Routh, HM Pantin, & CA Mason. exploring differences among adolescents in 1997. Family support decreases influence of the Southwest. J Community Psycho/ 31 167- deviant peers on Hispanic adolescents' sub­ 188. stance use. J Clinical Child Psycho/ 26 15- Maccoby EE & CN Jacklin. 1974. The Psychology 23. of Sex Differences. Stanford, CA: Stanford U Fuligni AJ, V Tseng , & M Lam. 1999. Attitudes to­ Press. ward family obligations among American ado­ Marin G & BV Marin. 1991. Research with His­ lescents with Asian, Latino American, and Eu­ panic Populations. Newbury Park: Sage. ropean backgrounds. Child Development 70 Masten A& MJ Reed . 2002. Resilience in develop­ 1030-1044. ment. Pp. 74-88 in Handbook of Positive Psy­ Gilbert JM & RL Collins. 1997. Ethnic variation in chology. CR Snyder & S Lopez, eds. London : women's and men's drinking. Pp . 357-378 in Oxford U Press. Gender and Alcohol: Individual and Social Per­ McLoyd VC. 1990. Minority children: introduction spectives. RW Wilsnack & SC Wilsnack, eds. to the special issue. Child Development 51 263- New Brunswick, NJ : Rutgers Center of Alco­ 266. hol Studies. McQueen A, JG Getz, & JH Bray. 2003. Accultura­ Gilligan C. 1982. In a Different Voice: Psychologi­ tion , substance abuse, and deviant behavior: cal Theory and Women's Development. Cam- examining separation and family conflict as Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 67

mediators. Child Development 7 4 1737-1750. Suarez-Orozco C & M Suarez-Orozco. 1995. Mitchell CM & J Beals. 1997. The structure of prob­ Transformations: Immigration, Family Life, lem and positive behavior among American In­ and Achievement Motivation Among Latino dian adolescents: gender and community dif­ Youths. Stanford, CA: Stanford U Press. ferences. Amer J Community Psycho/25 257- Szapocnik J, D Santisteban, A Rio, A Perez-Vidal, 288. & WM Kurtines. 1989. Family effectiveness Negy C & OJ Woods. 1992. The importance of training: an intervention to prevent drug abuse acculturation in understanding research with and problem behaviors in Hispanic adolescents. Hispanic-Americans. Hispanic J Behavioral Sci Hispanic J Behavioral Sci 11 4-27. 14 224-247. Tschann JM, E Flores, BV Marin, LA Pasch, EM Newcomb MD. 1995. Drug use etiology among eth­ Baisch , & CJ Wibbelsman. 2002. lnterparental nic minority adolescents: risk and protective conflict and risk behaviors among Mexican factors. Pp . 105-129 in Drug Abuse Preven­ American adolescents: a cognitive-emotional tion with Multiethnic Youth. GJ Boivin, S model. J Abnormal Child Psycho/ 30 373-385. Schinke, & MA Orlandi, eds. Thousand Oaks, Turiel E. 1983. The Development of Social Knowl­ CA: Sage. edge: Morality and Convention. Cambridge, En­ Padilla AM & VN Salgado de Snyder. 1995. Hispan­ gland: Cambridge U Press. ics: what the culturally informed evaluator U.S. Census Bureau. 2001a. The Hispanic Popu­ needs to know. Pp. 117-146 in Cultural Com­ lation: 2000. Washington, DC: USGPO. petence for Evaluators: A Guide for Alcohol ___. 2001 b. The Hispanic Population in the and Other Drug Abuse Prevention Practition­ United States: March 2000, Current Popula­ ers Working With Ethnic/Racial Communities. tion Reports, P20-535. Washington , DC: MA Orlandi, ed. Rockville, MD: US Department USGPO. of Health and Human Services. U.S. Department of Health and Human Services. Raffaelli M, G Carlo, M Carranza, & G Gonzalez­ 2001 . Mental Health: Culture, Race and Kruger. Forthcoming. Understanding Latino chil­ Ethnicity - Supplement to Mental Health: A dren and adolescents in the mainstream: put­ Report of the Surgeon General. Rockville, MD. ting culture at the center of developmental mod­ Williams JH, CD Ayers, RD Abbott, JD Hawkins, & els. In New Directions for Child and Adoles­ RF Catalano. 1996. Structural equivalence of cent Development. R Larson & L Jensen, eds. involvement in problem behavior by adolescents Raffaelli M, BL Zamboanga, & G Carlo. 2005. Ac­ across racial groups using multiple group con­ culturation status and sexuality among Cuban­ firmatory factor analysis. Social Work Res 20 American college students. J Amer College 168-177. Health 54 7-13. Wyatt JM & G Carlo. 2002. What will my parents Ramirez JR, WD Crano, R Quist, M Burgoon, EM th ink? Relations among adolescents' expected Alvaro, & J Grandpre. 2004. Acculturation, parental reactions, prosocial moral reasoning , familism, parental monitoring, and knowledge and prosocial and antisocial behaviors. J Ado­ as predictors of marijuana and inhalant use in lescent Res 17 646-666. adolescents. Psycho/ Addictive Behaviors 18 Youniss J. 1980. Parent and Peers in Social De­ 3-11 . velopment: A Sullivan-Piaget Perspective. Ch i­ Sabogal F, G Marin, R Otero-Sabogal, BV Marin, & cago: U Chicago Press. EJ Perez-Stable. 1987. Hispanic familism and Zamboanga BL, G Carlo, & M Raffaelli. 2004. Prob­ acculturation: what changes and what lem behavior theory: an examination of the be­ doesn't? Hispanic J Behavioral Sci 9 397-412. havior structure system in Latino and non­ Salzinger S, RS Feldman, T Stockhammer, & J Hood. Latina college students. lnteramer J Psycho/ 2002. An ecological framework for understand­ 38 253-262. ing risk for exposure to community violence Zapata JT & OS Katims. 1994. Antecedents of sub­ and the effects of exposure on children and stance use among Mexican American school­ adolescents. Aggression & Violent Behavior 7 age children. J Drug Education 24 233-251. 423-451 . Samaniego RY & NA Gonzales. 1999. Multiple medi­ ators of the effects of acculturation status on ACKNOWLEDGEMENTS delinquency for Mexican American adoles­ The authors thank Lizette Pabon and Sherry C. cents. Amer J Community Psycho/ 27 189- Wang for their editorial assistance on an earlier 209. draft of this paper and to Janine Olthuis for her Smith C & MD Krohn . 1995. Delinquency and family thoughtful comments and attention to detail in the life among male adolescents: the role of ethni­ final draft of this article. The second author was city. J Youth Adolescence 24 69-93. supported by a grant from the National Science Suarez L&AG Ramirez. 1999. Pp. 115-136 in Pro­ Foundation (BCS-01 32302). moting Health in Multicultural Populations: A Handbook for Practitioners. RM Huffer & MV Kline, eds. Texas Department of Health. Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 69 RUNAWAY YOUTH ADMITTED TO JUVENILE DETENTION: FACTORS ASSOCIATED WITH CIGARETTE, ALCOHOL, AND MARIJUANA USE

Sanna J. Thompson, Ph.D., University of Texas at Austin Liliane Cambraia Windsor, MSSW, University of Texas at Austin

ABSTRACT

The high rates of substance use among American adolescents are challenging, especially among runaway youth who are often identified as delinquent and find themselves in juvenile detention centers. Current services offered by juvenile detention centers focus on offering safe, short-term residential care; however, substance use issues are common among these youth. Although providing substance use treatment is an unlikely addition to these juvenile justice agencies, the authors posit that these facil ities are in a prime position to facil itate screening, assessment, and referral. Addressing issues of substance abuse among run­ away youth admitted to j uvenile detention must be a major objective for prevention and treatment for this high-risk population. Although interventions that address substance use among runaway, delinquent youth are limited, future research must implement and evaluate early intervention strategies aimed at addressing the complex and multifaceted challenges experienced by these youth. ADOLESCENT SUBSTANCE ABUSE Winters 1999). One group of youth at greater The rates of substance use and abuse risk for substance use and other high-risk among American high school students are behaviors are youth who have run away the highest in the industrialized world (Bach ­ (Kipke, Palmer, LaFrance, & O'Connor, 1997; man, Wallace, O'Malley, Johnston, Kurth, & Whitbeck, Hoyt, & Bao 2000). Neighbors 1991 ). Results from the National Survey on Drug Use and Health (NSDUH) Runaway/Homeless Youth (SAMHSA 2002) indicate that among youth Runaway youth have been defined as 12 to 17 years of age, 33.3 percent have those who stay away from home at least over­ smoked cigarettes, 43.4 percent drank alco­ night without the permission of a parent or hol, 20.6 percent have used marijuana, and guardian; they often live in unsupervised con­ 11 .6 percent have used illicit drugs during ditions and are in need of basic services, their lifetime. Between 1997 and 2002, illicit such as food and shelter (Farrow, Deisher, drug use among youth 12-13 years old in­ Brown, Kulig , & Kipke 1992). Between creased from 2.2 percent to 4.2 percent 310,000 and 1.6 million youth in the United (SAMHSA 2002). Although lifetime illicit drug States become homeless each year due to use has been on the increase, cigarette running away or being forced to leave their smoking significantly declined from 2001 to homes (Finkelhor 1995; Greene & Ringwalt 2002, down from 37.3 percent to 33.3 per­ 1997); one in seven adolescents run away cent (Johnston, O'Malley, & Bachman 2003). (News and Research 2003). Runaway be­ Although rates of substance use are sig­ havior implies a failure in the family relational nificant in general youth populations, the per­ system (Whitbeck, Hoyt, &Ackley 1997b) and vasiveness of substance abuse is higher runaway youth often describe family situa­ among youth that also engage in other high­ tions characterized by disorganization, inef­ risk behaviors. Adolescents who abuse al­ fective parenting behavior (including sub­ cohol or drugs frequently perform poorly in stance abuse by parents), violence, neglect, school, have been abused or neglected, and and physical and/or sexual abuse (Kipke et suffer from co-morbid psychiatric conditions, al 1996; Ringwalt, Greene, & Robertson especially depression and suicidality (Hawk­ 1998). When compared with parents of non­ ins, Catalano, & Miller 1992; Leslie, Stein, & runaway adolescent's, runaway youths' par­ Rotheram-Borus 2002; Rahdert & Czecho­ ents score lower on parental warmth, sup­ wicz 1995). These youth often engage in portiveness and monitoring, and higher on high-risk behaviors, such as illegal activity, parental rejection (Whitbeck, Hoyt, & Ackley homelessness, risky sexual behavior, and 1997a). Poor family environments, inconsis­ school truancy (Kipke, Unger, Palmer, & tent family practices, and adolescent-parent Edgington 1996; Smyth & Saulnier 1996; conflict have been shown to increase the 70 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology child's risk for drug abuse (Hawkins et al Table 1: Sample Characteristics 1992). Youth N=121 Characteristics n % Substance Use Among Runaway Youth Gender Male 53 43.8 Although research has documented the Female 68 56.2 particularly high risk for drug use during ado­ Ethnicity lescence, runaway/homeless youth have European American 45 37.3 even greater risk for substance use. Rates African American 49 40.5 of alcohol and other drug use are substan­ Hispanic/Latina 6 5.0 tially higher among runaways than their non­ American Indian 5 4.1 runaway counterparts (Greene et al 1997; Asian 0 0.0 Kipke, Montgomery, Simon , & Iverson 1997). Other 16 13.2 Recent national estimates of alcohol or drug Last grade completed 6th grade 2 1.7 use among runaway youths reveal more than 7th to 8th grade 44 36.3 90 percent of youth utilizing runaway shel­ 9th to 1Oth grade 70 57.8 ters report using substances during their life­ 11th to 12th grade 5 4.1 time and 77 percent report using during the Living situation at admission three months prior to running away (Thomp­ With Parent(s) 41 33 .9 son , Pollio, Constantine, Reid , & Nebbitt Other adult/friend 6 5.0 2002). Alarmingly, as many as 40 percent of Foster care 5 4.1 runaway and homeless youth have used in­ Institution 6 5.0 travenous drugs (Pennbridge, Freese, & On the Street/shelter 63 52 .1 Cigarette use (ever) 71 .1 MacKenzie 1992). Results of a study that 86 Alcohol use (ever) 78 64 .5 compared runaway and non-runaway youth Marij uana use (ever) 82 67 .8 showed that runaways are three times more Mean so likely to use marijuana (43% vs 15%), seven Youth 's age 14.6 1.0 times more likely to use crack/cocaine (19% Total times ran away 4.9 9.8 vs 2.6%), five times more likely to use hallu­ Cigarette use (days/month) 16.5 13.0 cinogens (1 4% vs. 3.3%), and four times Alcohol use (days/month) 6.1 9.4 more likely to use heroin (3% vs .7%) than Marijuana use (days/month) 9.8 11.9 their non-runaway counterparts (Forst 1994). The substances predominately used by Although high rates of substance use is runaway youth are cigarettes, alcohol, and found among runaway youth is clear, crisis marijuana. In the general adolescent popu­ services designed to meet the needs of these lation , 32 percent of twelfth grade students youth can seldom provide appropriate treat­ smoke cigarettes (National Institutes of ment for these problems. One service sec­ Health Report 2002). However, one study de­ tor that deals extensively with runaway youth termined that 37 percent of runaway youth is county detention centers. These facilities smoked cigarettes regularly compared to are frequently utilized by families, courts, and only 6.3 percent of adolescents in high police departments as short-term residen­ school ; runaway youth also smoked more tial housing for runaway youth with non-crimi­ heavily (Ensign & Santelli 1998). Similarly, nal behaviors. As the major focus of deten­ prevalence rates for alcohol use among run­ tion centers is locating suitable long-term away youth are extremely high. For example, housing for these youth , identification of sub­ a study of runaway/homeless youth in the stance use issues is limited. However, these Midwest found that 75 percent drank beer facilities are in a unique position to address and 66 percent reported drinking hard liquor the needs of youth by focusing service provi­ (Whitbeck et al1997b). Marijuana use is also sion efforts on substance-related issues. To consistently higher among runaway youth provide a more complete picture of sub­ than their non-runaway counterparts - 31 stance use among runaway youth admitted percent vs. 23 percent (Sherman 1992) and into juvenile detention, the following research 54 percent vs. 24 percent (Cohen, MacKenzie, questions were posed: 1) What are the demo· & Yates 1991 ). Adolescents that are home­ graphic and individual characteristics of run· less are more likely to have tried marijuana away youth using admitted to a juvenile de­ (43.1 %) than non-runaway peers (11.0%) tention center, 2) what demographic, indi· (Ensiqn & Santelli 1998). vidual characteristics and family factors pre· Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 71

Table 2: Correlations Between Independent Variables and Cigarette, A lcohol, and Marijuana Use and Number of Days Used Independent Cigarettes Alcohol Marijuana Variables Ever Days Ever Days Ever Days Gender -.16 .1 8 .04 .27* .06 .21 Age .14 .28** .22** .33** .1 8* .31 ** European American .38* .26* .25** .12 .11 .10 Grade in school .1 0 .34** .19* .21 .13 .19 Total runaway episodes .06 .1 2 .05 .31** .07 .17 Lived with parent at admission .03 .27** -.02 .20 .03 .03 Combined sex and alcohol .1 9* .05 .32** .37** .34** .33** Worry about family relationships .16 .25* .17 .23* .15 .26* Ever smoked cigarettes .55** -.08 .48** .11 Days smoked cigarettes past month .1 8 .44** .09 .48** Ever drank alcohol .55** .18 .54** .1 0 Days drank alcohol past month -.08 .44** .15 .58** Ever smoked marijuana .48** .09 .54** .1 5 Days smoked marijuana past month .11 .47** .10 .58** *p < .05; **p < .01 diet cigarette, alcohol, and marijuana use student explained issues of confidentiality among this group of runaway youth and 3) and voluntary participation to the youth and what factors predict these youths' level of requested signed assent forms before they cigarette, alcohol, and marijuana use? were engaged in semi-structured interviews and standardized survey measures. METHODS Sample and Procedures Measures Between May and August 2001 , consecu­ The dependent variables used in the tive entrants to a juvenile detention center in analysis included: ever used cigarettes, al­ a mid-sized urban city in Western New York cohol or marijuana (coded 'ever used' = 1, were recruited for participation in the study. 'never used' = 0) and the frequency of sub­ Participants were typically admitted to the stance use as measured by how may days county detention center due to a mandate by during the past month the adolescent had family court. This facility was similar to other used cigarettes, alcohol, or marijuana. Inde­ juvenile detention centers as they provided pendent variables included demographic residential and custodial care for youth 11- and individual characteri stics of youth and 18 years of age who had committed a crimi­ their families. Demographic and individual nal offense or had been admitted due to non­ categorical variables included gender (' male' criminal behaviors, such as delinquency or = 1, 'female' = 2), ethnicity ('European Ameri­ running away (Dembo, Williams, Fagan, & can' = 1, 'African American' = 2, 'Hispanic/ Schmeidler 1993). Youth participants were Latino' = 3, 'American Indian' = 4, 'Asian' = 5, recruited from the detention center if they and 'other' = 6) , last grade completed (6th were between the ages of 11 -17 years, ad­ through 12th grade), and the last living situa­ mitted to the non-criminal juvenile offenders tion before admission to the detention cen­ unit, and reported a runaway episode during ter ('with parent(s)' = 1, 'with another adult, the previous six months. Nearly half of the friend , or relative' = 2, 'in foster care' = 3, 'in youth were admitted for a 'status offense', an institution, such as another residential typically running away. facility' = 4, 'on the street or in a temporary Parents of these youth had given tempo­ shelter' = 5) . Continuous variables included: rary custodial rights to the detention center age, total number of runaway episodes, and to act as their child's guardian; thus, the cen­ number of times the youth participant com­ ter provided consent to seek participation of bined sex and alcohol use in past month. the individual adolescent into the study. Of Family characteristics were evaluated the 171 youth that entered the detention cen­ usi ng the Family Functioning Scale (FFS) ter during the study period, 121 met inclu­ (Tavitian, Lubiner, Green, Grebstein, & Velicer sion criteria (admitted for non-criminal be­ 1987). The FFS consists of 40 items that havior and had runaway) and agreed to par­ measure five dimensions of family function­ ticipate. A Masters in Social Work graduate ing: positive family affect (i.e. "People in my 72 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Table 3: Logistic Regression Models of Cigarette, Alcohol, and Marijuana Use/Non-use Among Runaway Youth Utilizing Juvenile Detention Center Services Cigarette Use Predictor Characteristics B (SE) OR Ethnicity (European Am.) 2.85 (.88)** 17.30 Model: Age (years) -2.11 (.26) 0.73 X squared (5) = 45.95 Used alcohol (ever) 1.43 ( 63)* 4.16 p < .001 Used marijuana (ever) 1.86 (.71 )** 6.42 Cox & Snell R2 .34 Frequency of alcohol & sex 0.02 (.25) 1 02 Alcohol Use Predictor Characteristics B (SE) OR Ethnicity (European Am.) 0.67 (.63) 1.96 Model: Age (years) 0.11 (.26) 1.18 X squared (5) = 46.08 Used alcohol (ever) 1.50 (. 63)** 4.48 p < .001 Used marij uana (ever) 1. 91 (.61)** 6.75 Cox &Snell R2 .34 Frequency of alcohol & sex 0.44 (.27) 1. 55 Marijuana Use Predictor Characteristics B (SE) OR Ethnicity (European Am.) -1 .20 (70)* 0.30 Model: Age (years) 0.13 (. 27) 1.14 X squared (5) = 46.34 Used alcohol (ever) 1.86 ( 69)** 6.41 p < .001 Used marijuana (ever) 1.87 (.60)** 6.50 Cox & Snell R2 .34 Frequency of alcohol & sex 0.96 (.45)* 2.62 family listen when I speak"), rituals (i .e. "We 'elsewhere' = 0). Categorical variables yield pay attention to traditions in my family"), wor­ odds ratios (ORs) that reflect the likelihood ries (i.e. "I worry when I disagree with the of a positive response relative to a defined opinions of other family members"), conflicts reference category, after controlling for all the (i. e. "People in my family yell at each other"), other effects included in the model. Finally, and communication (i.e. "When I have ques­ OLS regression models were calculated to tions about personal relationships, I talk with evaluate predictors of the level of cigarette, my family member"). Respondents' rate alcohol and marijuana use (number of days items on a seven-point scale (1 = 'never' to 7 used). = 'always') and items are summed for the five subscales and a total score. Internal con­ Results sistency reliability ranged from alpha = .90 Analysis of the sample of adolescents in for positive fam ily affect to alpha = .74 for this study (see Table 1) revealed that 68 family conflict (Tavitian et al 1987). (56.2%) were female and averaged 14.6 (SO ± 1.0) years of age; most were in ninth and Method of Analysis tenth grade (57.8%). Youths' self-reported Descriptive analyses were followed by bi­ ethnicity indicated that most were African variate correlations to test for significant asso­ Ameri can (40.5%) or European American ciations between independent and depend­ (37.3%). Remarkably, the majority had been ent variables. A power analysis was also con­ living on the streets or in a temporary shelter ducted to determine whether the sample size before admission to the facility (52.1% ); how­ was sufficient to conduct multivariate analy­ ever, a large proportion reported living with ses (Faul & Erdfelder 1992). Given F2 = .25, their parent(s) (33.9%) . These youth had an alpha = .05, and 6 predictor variables in a average of five (SO ± 9.8) runaway episodes. model, power to detect an effect was 99 per­ Among youth participants, 71 .1 percent re­ cent; therefore, maximum likelihood logistic ported smoking cigarettes, 64.5 percent regression analyses were used to test pre­ drank alcohol, and 67.8 percent used mari­ dictor variables (youth and family character­ juana. Among those using substances, youth istics) on three dependent variables (ciga­ reported smoking an average of sixteen days rette, alcohol and marijuana use) . Nominal­ (SO± 13.0), used alcohol six days (SO± 94), level predictor variables with more than two and used marijuana nearly ten days (SO ± categories were transformed and assigned 11 .9) in the previous month. reference categories (e .g. last living situa­ Bivariate correlations were conducted to tion reference category: 'parent's home'= 1, test for significant relationships between the Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 73

Table 4: Regression Models To Predict Level of Cigarette, Alcohol, and Marijuana Use Among Runaway Youth Using Substances During Previous 30 Days DV = Days of Cigarette Use Predictor B (SE) p value Constant -42.39 (20.92) .05 Model: Ethnicity 5.49 (2.80) .05 F (5,73) = 4.01 , Number of runaway episodes 0.10 (0 12) .37 p < .001 Age 2.90 (1.39) .04 R squared .21 Gender 2.99 (2.82) .29 Worry about family relationships 0.31 (0.1 5) .03 DV = Days of Alcohol Use Predictor B (SE) p value Constant -52.31 (1 6.28) .002 Model: Ethnicity 1.66 (2.25) .46 F (5,63) = 5.53, Number of runaway episodes 0.20 (0.09) .02 p < .001 Age 3.08 (1.07) .006 R squared .30 Gender 5.17 (2.25) .02 Worry about family relationships 0.15 (0.11) .1 8 DV = Days of Marijuana Use Predictor B (SE) p value Constant -53. 38 (1 9 69) .008 Model: Ethnicity 1.27 (2 67) .64 F (5,71) = 3.57, Number of runaway episodes 0.12 (0.11) .27 p < .001 Age 3.34 (1.32) .01 R squared .20 Gender 3.82 (2.62) .15 Worry about family relationships 0.23 (0.13) .05 independent and dependent variables (see odds of using alcohol seven times (OR = Table 2). Age was a significant predictor of 6. 75). Smoking cigarettes also increased the every dependent variable, except lifetime use likelihood of alcohol use by nearly five times of cigarettes. Being European American was (OR= 4.48). significantly associated with smoking ciga­ Marijuana: The final model to predict rettes and ever drinking alcohol and com­ youths' marijuana use showed that being a bining sex and alcohol was associated with Eu ropean American adolescent decreased dependent variables measuring alcohol and the odds of marijuana use by 60 percent (OR marijuana use and level of use. Finally, many = .30). Greater frequency of combining sexual of the dependent variables (ever used sub­ activity and alcohol use nearly tripled the stance and number of days used in previous odds of marijuana use among youth in the month) were significantly related to each detention center (OR = 2.62). Drinking alco­ other. hol also increased the odds of using mari­ juana; those who reported drinking alcohol Predictors of Use were more than six times (OR = 6.41 ) more Cigarettes: As shown in Table 3, predic­ likely to use marijuana; those who have ever tors of cigarette use among the juvenile de­ smoked were also six times more likely to tainees revealed that being European Ameri­ use marijuana (OR = 6.50). can increased the odds of smoking ciga­ rettes more than seventeen times (OR = Predictors for Level of Substance Use 17.30) over that of other ethn ic groups. Also, Cigarettes: As shown in Ta ble 4, being those who reported ever drinking alcohol European American, older, and being wor­ were nearly four times more likely to smoke ried about family relationships predicted (OR = 4.16) than those not reporting alcohol greater cigarette use as measured by the use, and those who smoked marijuana were number of days the youth smoked in the pre­ more than six times more likely to smoke vious month (F(5, 73) = 4.01 , p < .01 ). This cigarettes (OR= 6.4) than those who did not model accounted for 21 percent of the vari­ report using marijuana. ance in cigarette use among those who re­ Alcohol: Youth that reported using mari­ ported smoking cigarettes in the month prior juana were significantly more likely to use to interview. alcohol than those who did not report mari­ Alcohol: Predictors of the level of alcohol juana use. Using marijuana increased the use among detained youth included a great- 74 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology er number of runaway episodes, being older, these adolescents. and female (F(5,63) = 5.53, p < .001). This The strongest predictor of alcohol and model accounted for 30 percent of the vari­ marij uana use among th is sample of de­ ance in alcohol use. tained adolescents was use of other sub­ Marijuana: Predictors for level of mari­ stances. Previous research has confirmed juana use among youth in juvenile detention that use of one type of drug often progresses included being older and reporting being to use of other drugs (Golub, Labouvie , & worried about family relationships (F(5 ,71) = Johnson 2000; Kandei ,Yamaguchi, & Chen 3.57, p < .001). This model accounted for 20 1992). These studies found that illicit drug percent of the variance in marijuana use. use among young men aged 15 to 25 was dependent on prior use of alcohol ; among Discussion young women either cigarettes or alcohol This study aimed to understand the risk was a sufficient condition for progression to factors in cigarette, alcohol, and marijuana marijuana (Kandel et al 1992). Others found use in youth admitted to juvenile detention that age of onset and frequency of use at a for non-criminal behaviors. Findings lower stage of drug use were strong predic­ showed that runaway youth admitted to juve­ tors of further progression (Golub et al 2000). nile detention services have significantly These studies support the findings of this higher levels of substance use than national study that use of one substance is often asso­ estimates of adolescent populations (John­ ciated with use of other substances as well. ston et al2003; SAMHSA 2002). Participants Ethnicity was significantly associated with in this study reported greater use of alcohol cigarette and alcohol use in this sample of (64.5%) and marijuana (67.8%) than those runaway youth admitted to juvenile detention identified in another study of non-runaway center services. This study confirmed previ­ youth that showed that 29 percent of eighth ous findings that being European American and tenth graders have used marijuana, 39.4 is a risk factor that increases the likelihood percent have smoked cigarettes, and 57 per­ for smoking cigarettes and drinking alcohol. cent drank alcohol (Johnston et al 2003). Feiglman & Lee (1995) found fewer African These findings confirm the magnitude of American youth smoke cigarettes when com­ substance use problems among runaway pared to European American teens, despite youth in general , but reveal greater preva­ a greater percentage of African American lence among youth who have been admitted adults who smoke. One area of difference to juvenile detention services. between European American and minority The unique characteristics of juvenile de­ teens may be the influences that initiate the tainees, such as multiple runaway episodes behavior. Some note that minority teens and more than half living on the street at the seem to be more influenced by family mem­ time of admission, suggest they li kely en­ bers who use cigarettes and European gaged in a variety of high-risk behaviors. As American teens are more influenced by their others have found strong correlations be­ substance using peers (Parker, Sussman, tween drug use and crime in samples of youth Crippens, & Scholl 1996). European Ameri­ entering the juvenile justice system (Dembo can youth are also more likely to drink alco­ et al 1993), it is likely that these youth also hol than are minority youth . According tc engaged in criminal activity, as well as sub­ Bachman et al (1991 ), drinking among most stance use. Living on the street also requires minority high school seniors is less than for survival skills necessary to cope with their their white counterparts. Some have posited often traumatic lifestyle. As alcohol and mari­ that alcohol use among African Americar juana have more of an anesthetizing affect youth is determined by family attitudes anc than do cigarettes, these substances may social support (Epstein, Botvin, Diaz, 8 be used to deal with abusive situations, feel­ Schinke 1995) and the perceived expecta· ings of detachment from others, and mental tions from their families that preclude drink· health symptoms (McMorris, Tyler, Whitbeck, ing alcohol. & Hoyt 2002; Whitbeck et al 2000). Thus, the Fam ilial factors also predicted the leve high occurrence of alcohol and marijuana of cigarette and marijuana use among these use among youth admitted to juvenile deten­ runaway youth . Others have reported that ris~ tion may reflect one high-risk, problem be­ factors for youth substance use include poor havior among many others experienced by parenting practices, family stress, and chile Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 75 victimization (Weinberg, Rahdert, Colliver, & residential care, substance use issues must Glantz 1998). In addition, the stressors high­ be acknowledged. Providing substance use risk families face may prevent their ability to treatment is an immense challenge for these deal with conflict (Pelton & Forehand 2001 ). agencies; however, they are in a prime posi­ However, in this study youth who worried tion to facilitate screening, assessment, and about family relationships used substances referral (Thornberry, To lnay, Flanagan, & more than other detained, runaway youth; Glynn 1991). issues of conflict and poor communication were not significant. Children are sensitive ACKNOWLEDGEMENT to family dynamics; thus, increased worry The authors wish to thank the staff and about poor family relationships and their in­ youth participants at the Erie County Deten­ ability to change the family environment may tion Center. The support of the director, Rich­ lead them to use substances to escape their ard Nelson, was crucial to the completion of problems and worries. Being worried about this project. family relationships may be an indicator that poor communication and conflict are the un­ REFERENCES derlying causes of this concern. Further re­ Bachman JG, JM Wallace, PM O'Malley, LD John­ search is needed to understand the path­ ston, CL Kurth, & HW Neighbors. 1991 . Racial/ ways through which family conflict and dis­ ethnic differences in smoking, drinking, and il­ licit drug use among American high school se­ organization might be indicative of adoles­ niors. Amer J Public Health 81 372-377. cents' perceptions of family relationships and Cohen E, RG MacKenzie, & GL Yates. 1991 . the association with substance use and run­ HEADSS, a psychosocial risk assessment in­ away behaviors (Johnson, Bryant, Collins, strument: implications for designing effective Noe, Strader, & Berbaum 1998). intervention programs for runaway youth. J Recognizing the inherent limitations of Adolescent Health 12 7 539-544. cross-sectional and self-repot data, the find­ Dembo R, L Williams, J Fagan, & J Schmeidler. ings of this study must be viewed as sugges­ 1993. Development and assessment of a clas­ tive rather than conclusive. Because the sification of high risk youths. J Drug Issues 24 sample size is relatively small, generalizing 1 25-53. Ensign J & J Santelli. 1998. Health status and ser­ the results to other runaway youth must be vice use. Comparison of adolescents at a made with caution. However, the power analy­ school-based health clinic with homeless ado­ sis suggested that there was sufficient power lescents. Archives Pediatrics & Adolescent to detect effects in the multivariate models. Medicine 152 1 20-24. In addition, it is likely that youth under-re­ Epstein JA, GJ Botvin, T Diaz, & SP Schinke. 1995. ported their level of substance use and other The role of social factors and individual char­ high-risk behaviors, making these behaviors acteristics in promoting alcohol use among in­ even more problematic than study results ner-city minority youths. J Stud Alcohol 56 1 39-46. demonstrate. Sensitive assessment of sub­ Farrow JA, RW Deisher, R Brown, JW Kulig, & MD stance use issues during admission to a Kipke. 1992. Health and health needs of home­ juvenile detention facility is an obvious re­ less and runaway youth. A position paper of quirement for appropriate service provision the Society for Adolescent Medicine. J Ado­ and further understanding of the challenges lescent Health 13 8 717-726. experienced by these youth. Additional re­ Faul F & E Erdfelder. 1992. Gpower: a priori, post­ search is needed to determine the extent to hoc, and compromise power analyses for MS­ which the predictor variables in this study DOS. Bonn, Germany: Bonn U Department of could be replicated by more rigorous meth­ Psychology. odological strategies. Feiglman W & JA Lee. 1995. Patterns of cigarette use among Black and White adolescents. Amer This study provides new information use­ J Addictions 4 3 215-225. ful to service providers and policy makers Finkelhor D. 1995. The victimization of children: a concerning the substance use of runaway developmental perspective. Amer J Orthopsy­ youth admitted into juvenile detention. It is chiatry 65 2 177-193. clear that intervention efforts are needed to Forst ML. 1994. A substance use profile of delin­ address substance use for these high-risk quent and homeless youths. J Drug Education runaway youth. Although the current services 24 3 219-231. offered by juvenile detention centers must Golub A, E Labouvie, & BD Johnson. 2000. Re­ sponse reliability and the study of adolescent continue to focus on offering safe, short-term 76 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

substance use progression. J Drug Issues 30 vorce. J Family Violence 16 1 1-15. 1 103-118. Pennbridge JNF, TE Freese, & RG MacKenzie. 1992. Greene JM & CL Ringwalt. 1997. Substance use High-risk behaviors among male street youth in among runaway and homeless youth in three Hollywood, California. AIDS Education & Pre­ national samples. Amer J Public Health 87 2 vention Suppl 24-33. 229-236. Rahdert E & D Czechowicz, eds. 1995. Adoles­ Hawkins J, RF Catalano, & JY Miller. 1992. Risk cent drug abuse: clinical assessment and and protective factors for alcohol and other therapeutic interventions. Washington, DC : drug problems in adolescence and early adult­ NIDA Research Monograph 156. USGPO. hood : implications for substance abuse pre­ Ringwalt CL, JM Greene, & MJ Robertson. 1998. vention . Psychological Bu//112 1 64-105. Familial backgrounds and risk behaviors of Johnson K, DO Bryant, DA Collins, TO Noe, TN youth with thrownaway experiences. J Ado­ Strader, & M Berbaum. 1998. Preventing and lescence 21 3 241 -252 . reducing alcohol and other drug use among SAMHSA. 2002. National Survey on Drug Use and high-risk youths by increasing family resilience. Health . Retrieved November 25, 2003, from Social Work 43 4 297-308. http://www. samhsa. gov/oas/n hsda/2k2nsd u hi Johnston LD , PM O'Malley, & JG Bachman. 2003. Results/ 2k2Results.htm#toc. Monitoring the future national results on ado­ Sherman OJ . 1992. The neglected health care lescent drug use: overview of key findings , needs of street youth . Public Health Reports 2002. Retrieved November 20th , 2003, from 197 433-440. http://www. mon itoringthefuture .org / pubs/ Smyth NJ & CF Saulnier. 1996. Substance abuse monog ra phs/ove rview2002 . pdf prevention among high risk youth . J Preven­ Kandel DB, K Yamaguchi, & K Chen. 1992. Stages tion & Intervention in Community 14 1/2 61-79. of progression in drug involvement from ado­ Tavitian M, JL Lubiner, L Green, LC Grebstein, & lescence to adulthood : further evidence for WF Velicer. 1987. Dimensions of family func­ the gateway theory. J Studies Alcohol 53 5 tioning . J Social Behavior Personality 2 191- 447-457. 204. Kipke MD, SB Montgomery, TR Simon, & EF Iverson. Thompson SJ, DE Pollio, J Constantine, D Reid , & V 1997. "Substance abuse" disorders among Nebbitt. 2002. Short-term outcomes for youths runaway and homeless youth . Substance Use receiving runaway homeless shelter services. & Misuse 32 7-8 969-986. Res Social Work Practice 12 5 589-603. Kipke MD, RF Palmer, S LaFrance, & S O'Connor. Thornberry TP, SE Tolnay, T J Flanagan, & P Glynn. 1997. Homeless youths' descriptions of their 1991 . Children custody 1987: a comparison of parents' child-rearing practices. Youth & Soci­ public and private juvenile custody facilities. ety 28 4 415-431 . Washington, DC . Kipke MD, JB Unger, RF Palmer, & R Edgington. Weinberg NZ, E Rahdert, JD Colliver, & MD Glantz. 1996. Drug use, needle sha ring and HIV risk 1998. Adolescent substance abuse: a review among injection drug-using street youth . Sub­ of the past 10 years. J A mer Academy Child & stance Use & Misuse 3 1167-1187. Adolescent Psychiatry 37 3 252-262. Leslie MB, JA Stein , & MJ Rotheram-Borus. 2002 . Whitbeck LB, DR Hoyt, & KAAckley. 1997a. Fami­ Sex-specific predictors of suicidality among lies of homeless and runaway adolescents: a runaway youth . J Clinical Child & Adolescent comparison of parent/caretaker and adoles­ Psycho/ 31 1 27-40 . cent perspectives on parenting , family violence, McMorris BJ , KA Tyler, LB Whitbeck, & DR Hoyt. and adolescent conduct. Child Abuse & Ne­ 2002. Sex-specific predictors of suicidality glect 21 6 517-528. among runaway youth . J Clinical Child & Ado­ -.,...... ,.-· 1997b. Abusive family backgrounds and lescent Psychology 31 1 27-40. later victimization among runaway and home­ National Institutes of Health Report. 2002 . Recent less adolescents. J Res Adolescence 7 4 375- progress in decreasing youth tobacco use, but 392. much work remains. Vol. 2003. Bethesda, MD: Whitbeck LB, DR Hoyt, & WN Bao. 2000. Depres­ NIH. sive symptoms and co-occurring depressive News and Research . 2003. Retrieved November symptoms, substance abuse, and conduct 18 , 2003, from http://www.nrscrisisline.org/ problems among runaway and homeless ado­ news.asp . lescents. Child Development 71 3 721-732. Parker VC , S Sussman, DL Crippens, & D Scholl. Winters KC . 1999. Treatment of adolescents with 1996. Qualitative development of smoking pre­ substance use disorders. (SMA 99-3283). vention programming for minority youth . Ad­ Rockville MD: Center for Substance Abuse dictive Behaviors 21 4 521-525. Treatment. Pelton J & R Forehand . 2001 . Discrepancy be­ tween mother and child perceptions of their relationship: consequences for adolescents considered within the context of parental di- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 77 KEY RISK AND PROTECTIVE FACTORS AMONG MULTI-ETHNIC, ELEMENTARY AGED CHILDREN: FINDINGS FROM NEW MEXICO'S BEHAVIORAL HEALTH SERVICES PREVENTION BUREAU

Richard C. Cervantes, California State University, Long Beach Ann Del Vechio, Alpha Assessment Associates Jose Esquibel, New Mexico Department of Health, and Tony Rey, Behavioral Assessment, Inc.

INTRODUCTION total under the age of 5 years. Forty-two per­ More than at any time in recent history, cent of the population is Hispanic, 45 per­ many young school aged children are con­ cent Anglo, 10 percent Native American, and fronted with a multiplicity of factors that in­ 2 percent African American. The average crease their risk for behavioral problems household size in New Mexico is 2.63 per­ such as school failure, tobacco and other sons, with 3.18 persons per family (New drug use. Elementary aged youth face in­ Mexico Voices for Children 2003; New Mexico creased social pressures prompted by video Kids Count Data Book 2002). and media (TV) where peer conformity and Based on a number of other youth indica­ materialism are highly endorsed and valued tors, many children in New Mexico lag be­ by American society. In addition, the chang­ hind other elementary-aged youth nationally. ing nature of the family system, inadequate For example, 10.2 percent of the children and ineffectual parenting skills, and in some ages 5 to 17 have "difficulty speaking En­ cases economic hardship in single-parent glish", while the national average is at 6.6 headed homes, further exacerbate the risks percent (US Census 2000). Twenty six per­ that young children are expose to. Further, cent (135,428) of New Mexico's youth are liv­ multi-ethnic youth, including Native American ing in single-parent households, compared children also face a variety of cultural based to 23.3 percent of America's youth. Twelve risk factors such as acculturation stress (Cer­ percent (13,665) of New Mexico youth are vantes & Ortiz 2003), perceived discrimina­ high school dropouts, compared to 9.8 per­ tion , and other forms of conflict with majority cent nationally. Six percent (18,374) are liv­ culture youth. ing with one or more disabilities, compared Elementary aged youth in America in­ with 5.8 percent nationally. New Mexico's cludes over 72,293,812 children. Of this to­ 2000 census data show that in fiscal year tal, 5,274,343 children are living in high-pov­ 2001 , there were 62,025 juvenile offenses erty neighborhoods. Children who live in cen­ reported, and 30,032 referrals (6% of the ju­ sus tracts where 20 percent or more of the venile population) to the juvenile justice sys­ population is below poverty are classified as tem. In New Mexico's public schools, 56 per­ "high poverty" neighborhoods. Based on this cent of the students qualify for the free or criteria about 7.4 percent nationally live in reduced cost lunch program. In the area of high poverty neighborhoods. In comparison, prenatal care, in 2000, 12.9 percent of moth­ a higher percentage of New Mexico youth re­ ers received little or no prenatal care (New side in these high poverty neighborhoods. Mexico Voices for Children 2003 32). New Mexico leads the nation in childhood Data reported on the New Mexico Youth poverty. Of the 508,574 of New Mexico's youth, Risk and Resiliency (administered state­ 10.7 percent are living in high-risk poverty wide in the public schools by the Depart­ neighborhoods (US Census 2003). The New ments of Education and Health) for high Mexico median income in year 2000 was school students reflect that 50.1 percent re­ $39,425, as compared to $50,046 nation­ port having at least one drink on at least one ally, and 14 percent of New Mexico's families day during the past thirty days; 35.2 percent are food insecure, with 9.2 percent nation­ report having at least one drink on two to five ally (New Mexico Voices for Children 2003). days during the past thirty days (binge drink­ According to the 2000 U.S. Census, the ing); 30.2 percent report having five or more total New Mexico population is 1 ,819,046. drinks of alcohol within a couple of hours at Twenty-eight percent of the population is un­ least once during the past thirty days; 27 per­ der 18 years of age, with 7 percent of the cent report smoking cigarettes during the past 78 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology thirty days. Thirty percent reported using mari­ importance of specific risk and protective fac­ juana at least once during the past thirty days, tors within the domains of the individual, fam­ with 9 percent reporting marijuana usage on ily, peer group, school, and community. The school property during the past thirty days SAMSHA Center for Substance Abuse Pre­ (New Mexico Voices for Children 2003 32). vention has now articulated and tested pre­ New Mexico's youth are exposed to and vention interventions that are in many cases affected by substantial risk factors and prob­ life domain specific. While efforts are impres­ lem behaviors in their communities and in sive, most of this prevention research work the state as a whole , brought about by pov­ has focused on youth in the 12-17 age group. erty, lack of hope in the future, and an educa­ Very few studies exist on either the assess­ tional system that does not provide sufficient ment of risk and protective factors among social structures or bonding with positive younger school age children (K-6th grade) adults and their community. By the 4th grade, or on specific clinical trails that test the effi­ 49 percent of children in New Mexico score cacy of prevention programming for younger below the basic math level, compared with children. 33 percent nationally (The Annie E. Casey Recent advances have been made, for ex­ Foundation, Kids Count Data 2002 120). ample, Gensheimer, Roosa , and Ayers These data highlight the importance of de­ (1990) examine strategies for recruiting veloping systems for identifying youth who school aged youth into a school based pre­ experience multiple risks such that early in­ vention program. Among 4th thru 6th grad­ terventions and drug prevention programs ers exposed to a film about parental alcohol­ can be effectively implemented in the state ism , self-selected children tended to score of New Mexico. higher on emotional and behavioral symp­ In addition to the above, Native American toms. In another study, family risk in the form youth in New Mexico may represent a partic­ of parental alcohol dependence was found ularly high-risk population. For example, New to greatly increase a number of behavior and Mexico Kids Count Data 2002 shows that emotional risk factors in elementary aged 64,953 Native American children under the children (Dawson 1992). age of 18 who live in New Mexico's pueblos With respect to the research on drug pre­ and reservations are living in extreme high vention programming, a number of studies poverty neighborhoods. Close to 30 percent have focused on younger elementary aged live in single-family households, with 15 per­ children . Werch and colleagues (2003) tested cent linguistically isolated. There is a 20 per­ an alcohol prevention program, STARS, for cent high school dropout rate, and nearly 7 reducing alcohol risk. In that study, students percent live with at least one disability, which receiving the intervention had significantly is defined as a long-lasting physical, men­ less intentions to use alcohol in the future tal , or emotional condition determined for and less alcohol quantity. Dielman, Shope, non-institutionalized persons aged 5 and Leech , and Butchart (1989) tested a social over (Kids Count Census Data 2000). skills/peer resistance curriculum for alcohol misuse. Among 5th and 6th grade partici­ Substance Abuse Prevention for Young pants results indicated the intervention was Children effective in reducing the rate of increase of The field of Substance Abuse Prevention alcohol use and misuse among grade six has made significant advances over the past students who entered the study with prior decade. Scientifically based prevention ef­ unsupervised as well as supervised alcohol forts have resulted in a number of programs use. In one study related to peer pressure that reduce risks and improve resiliency among elementary school youth, Dielman among youth who are prone to drug use and and colleagues (1992) found that prevention abuse. The Center for Substance Abuse Pre­ programs had a differential effect to suscep­ vention along with other researchers tibility of peer pressure depending on the par­ (Hawkins, Catalano, & Miller 1992) have pro­ ticipants previous experience with alcohol moted a Web of Influence frame work for un­ experimentation. derstanding the causes and correlates of Shope, Copeland, Marcoux, and Kamp substance abuse and this model has been (1996) tested the Michigan Model for Com­ equally important in designing prevention prehensive School Health Education among programs. This framework emphasizes the 5th thru 8th grade students. Following these Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 79 youth across the three-year span, at the end The purpose of this research was to docu­ of grade seven, program students' rates of ment and evaluate those risk and protective substance use had increased significantly factors experienced in a multi-ethnic sample less and knowledge of alcohol pressures, of elementary aged youth across the state of effects, and skills to resist had increased New Mexico. This study was part of a larger significantly more than those of comparison research investigation aimed at identifying students. Another promising approach for effective drug prevention strategies for at risk ethnic youth, R.E.A.L. curriculum serves 7th youth in Pre-Kindergarten through sixth grade students and is currently in the pro­ grade. In this sub-study, we were particularly cess of being tested for positive program ef­ interested in examining how ethnicity pre­ fects (Gosin, Marsiglia, & Hecht 2003). One dicted various personal, family and school interesting tobacco prevention program was related risk factors in fifth and sixth grade assessed among American Indian Youth in students. New Mexico. This school based cancer based prevention project was implemented METHODOLOGY for 7th and 5th grade Navajo and Pueblo In­ The New Mexico State Incentive Grant drug dian children. Using the Pathways to Health prevention initiative established the states Curriculum, the researchers found that boys science based programming efforts. Con­ were more likely to use and intend to use tracted agencies implemented a variety of cigarettes more than girls. The use of smoke­ evidence-based prevention programs for less tobacco also increased with increasing youth in 5th and 6th grade. Most of the pro­ grade level, though this trend was less pro­ grams were directed primarily toward youth, nounced for girls. The researchers con­ with a few programs focusing on parents/ cluded that there is evidence of experimen­ families. A large number of 5th and 6th grad­ tation and regular use of tobacco products ers participated in the Life Skills Training qy both Navajo and Pueblo boys and girls. curriculum developed by Gilbert J. Botvin, Even more students' indicated intention to which was implemented mainly in rural use tobacco products in the future. These schools with high percentages of Hispanic data confirm the need for primary prevention and Native American youth, such as the com­ programs designed for this population of munities of Chama, Dulce (Jicarilla Apache), American Indians. Botvin, Griffin, Diaz, and Tierra Amarilla, Acoma Pueblo, Laguna Ifill-Williams (2001) tested a school-based Pueblo, the Pueblo of Zuni, Roswell and drug abuse preventive intervention in a Dexter. The single urban site for implementa­ sample of predominantly minority students tion of the Life Skills Training was in the Santa (N=3621) in 29 New York City schools. Re­ Fe Public Schools. Student participation in sults indicated that those who received the the curriculum was determined by accep­ program (n=2144) reported less smoking, tance and approval of the curriculum by drinking, drunkenness, inhalant use, and school administrative officials. As a univer­ poly-drug use relative to controls (n=1477). sal prevention approach, there were no spe­ The findings from this study show that a drug cific recruitment strategies for student par­ abuse prevention program originally de­ ticipation. In most cases, the entire student signed for White middle-class adolescent population of fifth and sixth graders of the populations is effective in a sample of mi­ various school sites participated in the cur­ nority, economically disadvantaged, inner-city riculum. adolescents. In addition, fifth and sixth grade students These studies highlight the efficacy of participated in the SMART Moves curriculum drug prevention programs and point out the in San Juan County. Specific schools within need for future research on identifying spe­ the county were identified as 'high risk' for cific, culturally based risk and protective fac­ substance abuse and related social prob­ tors and the need for testing appropriate drug lems based on a county needs assessment. prevention interventions for younger school­ School officials within the various school dis­ aged youth. In addition, studies that include tricts were approached about offering SMART multi-ethnic samples have great utility in iden­ Moves as part of classroom instruction, the tifying cross-cultural risk and protective fac­ first point of contact being district superin­ tors, as well as differential program effects tendents. Other school personal that as­ across cultural groups. sisted in identifying specific classrooms for 80 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

delivery of the curriculum included the school Table 1: Grade in School principals, the Safe and Drug-Free Schools Valid and Community Coordinators, and class­ Grade Number Percent Percent room teachers. The SMART Moves curricu­ 4th 141 13.3 13.5 lum was implemented in various schools in 5th 466 43.8 44.5 6th 438 41 .2 41 .8 the communities of Bloomfield, Aztec, Ship­ other 2 .2 .2 rock, and Farmington, which have a high per­ missing 16 1.5 centage of Native American youth, mainly total 1063 100 100 Navajo. Approximately 40 percent of the youth participants in the SMART Moves curricula in Table 2: Gender San Juan County were Native American, ap­ Valid proximately 30 percent were Anglo, and ap­ Number Percent Percent proximately 25 percent were Hispanic. In ad­ Male 550 51.7 51.0 dition, SMART Moves was implemented in Female 509 47.9 48.1 coordination with the local Boys and Girls Missing 4 .4 Club at local club settings with a small num­ Total 1059 100 ber of youth. Funds from the New Mexico State Incen­ Incentive Grant Project was Learning to tive Grant Project also supported locally de­ Lead, a program developed and imple­ veloped prevention programs with the po­ mented by Cornstalk Institute of Albuquer­ tential of becoming promising prevention que, New Mexico. The emphasis of the pro­ programs or model prevention programs. gram is on facilitating the transition from ele­ One of these programs, Project Venture, was mentary school to middle school and from developed by the National Indian Youth Lead­ middle school to high school by fostering ership Project in Gallup, New Mexico, and is self-efficacy in academics and social devel­ now being considered by the Center for Sub­ opment. The main components of the pro­ stance Abuse Prevention as a model preven­ gram include mentoring, tutoring, skills build­ tion program. Designed as an intervention ing, experiential education , leadership specifically for Native American youth, Project through service, as well as a family interac­ Venture is a positive youth development ap­ tion component. Services are delivered within proach that encourages youth to develop into the school setting and also have out-of capable, caring, and healthy individuals school activities. The schools selected for through adventure-based experiential edu­ the services are in high-risk urban neighbor­ cation and service learning in classroom and hoods in the City of Albuquerque. The staff of out-of-school settings and through intensive Cornstalk Institute negotiates with school summer camp and wilderness trek experi­ administrators and teachers to identify the ences. The approach incorporates elements classrooms for participation in the Learning of traditional wisdom shared by Native el­ to Lead Program. As such, the program main­ ders utilizing alternative methods--outdoor/ tains a high rate of retention. experiential education, servant leadership As mentioned, this study was part of a (services learning), reconnecting with tradi­ larger, state and federally funded drug pre­ tional culture and the natural world--as a vention initiative emphasizing "evidence means to assist youth develop in healthy and based" prevention programming. Local positive ways. Project Venture consists of a evaluation experts were instrumental in the universal prevention service component and coordination and collection of self report and a selective prevention service component. parent report data. The universal prevention service component was delivered in the school setting and was Participants negotiated between school officials and Youth in the NM K-6 programs were used project staff to determine which classrooms to develop the sample reported here. The are served and the fit of the services in re­ sample consists of children from some gard to the school curriculum. Of the identi­ twenty prevention programs throughout the fied classes, all students participated in Pro­ state. Programs were in both community­ ject Venture activities. based organizations and in prevention coali­ Another locally developed prevention pro­ tions and partnerships. Eighty percent of the gram funded through the New Mexico State children in this sample were ages ten and Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 81

Table 3: Ethnic Breakdown for the Sample tion strategy. This was accomplished through Group Frequency Percent the use of a standardized instrument devel­ Latino 388 36.5 oped in New Mexico that included self-re­ Native American 628 40.9 ported measures of substance use and re­ White/Anglo Saxon 194 18.3 lated behaviors. Known as the Pre K-6 Ver­ Total 1210 95.7 sion of "Strategies for Success" evaluation eleven. instrument, the survey asks a series of ques­ In terms of school grade, some 84 per­ tions about youth experimentation and pat­ cent of the sample were in grades 5 and 6. terns of current (past 30 day) alcohol, tobac­ Most of the remaining children, 13.3 percent co, and other illicit drug use. The data collec­ were in the fourth grade (Table 1) . The gen­ tion instrument was worded in simple, under­ der split for respondents was relatively even standable language for elementary aged with slightly more boys than girls answering youth and was made available in both En­ the survey (Table 2). glish and Spanish. This analysis is based upon pre-test, baseline survey responses. Ethnicity An initial set of evaluation input meetings A negligible number of African American were held with local evaluators in New Mexi­ (14 or 1.2%) and Asian American children (2 co, including those with considerable experi­ or .2%) responded to the survey. An additional ence in the assessment of young children. 67 students identified themselves as "other" The group was charged with developing a or a combination of more than one category. set of measures that would adequately as­ The majority of the survey respondents iden­ sess program effects in terms of youth and tified themselves as Anglo or White, Native family risk and protective factors, in addition American or Latino/Hispanic. The Native to the assessment of drug outcomes for American respondents could identify them­ young children. Through a series of advisory selves as Pueblo, Navajo, Apache, or other meetings, as well as a search of all aca­ Native American. For the purpose of this demic databases and literature sources, a analysis, Native American respondents were set of tools were evaluated for appropriate­ re-categorized to form one large Native Ameri­ ness and a final set of tools was selected. can group. Latino respondents were similarly This set of tools included components of the re-categorized from the various response following standardized tools: categories that included: Hispanic, Mexican, Central American, South American, Spanish, • Standardized Parent and Youth Demo- Puerto Rican, and Cuban. Respondents in graphic Information Sheets the other and mixed categories were elimi­ • The Conner's Rating Scales for Parent nated as outliers. Table 3 provides a break­ • The Conner's Rating Scales for Teachers down of the sample included in this study. • A Youth Risk Survey (Modeled from the All ethnic groups except for the three larg­ CSAP National High Risk Youth Study) est groups were eliminated from the sample. • The Family Cohesion and Adaptability The resulting total for ethnicity is larger than Scales (FACES Ill) the entire sample (1 063) as some students identified themselves in two or more of the Youth in the 5-6th grade age range were three major ethnic categories. A total of 65 administered the Strategies for Success­ percent of the respondents or 691 of the stu­ Youth Risk Survey which also included spe­ dents indicated that a language other than cific items on lifetime and current substance English was spoken at home. The largest use. No parent or teacher report data were number of these respondents spoke Navajo collected for this older aged sub-sample. or Spanish. Pueblo dialects (Tiwa, Tewa, Towa, and Keres) as well as Apache or "In­ Procedures dian" also were identified as the language Informed consent procedures were de­ spoken at home. veloped for each of the data collection sites, and in some cases, parental consent had Instruments been previously obtained as part of school The statewide evaluation of evidence­ district policy for extracurricular school and based outcomes consisted principally of data drug education activities. Local evaluators collection using a pre-test/post-test evalua- established data collection schedules and 82 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Table 4: Scale Reliabilities Alpha Scale Description Coefficient School disruptive behaviors 3 items including fighting , tagging, safe schools 0.63 violations School protective factors 11 items including teacher and staff care for child , 0.78 school is important to finish , try hard to do well in school, etc. Parent communication 3 items including talk to parents about your future, 0.32 how often do you talk to parents about your problems, etc. Family Bonding 5 items including get along well with parents, feel 0.42 safe with parents, have respect for parents, etc. 30 day tobacco use 2 items including used chewing tobacco, smoked a 0.55 cigarette 30 day illicit drug use 2 items including marijuana use and inhalant use 0.34 Attitude toward use 9 items that included how wrong wou ld adults think, 0.87 parent think, someone your age think it is for youth to use marijuana, smoke cigarettes, drink alcohol? Perceived availability 3 items including how easy to get cigarettes? 0.87 Marijuana? Alcohol? Perceived harm 3 items including how risky is it to tuse cigarettes? 0.82 Marijuana? Alcohol? worked closely with school officials and Levene's statistic to test for homogeneity teachers to gather data during regular school of variance was conducted for each scale. hours. Group administration of the research Several of the comparisons (usually when tools was conducted. the scale was unreliable) indicated that the These analyses focused on the survey assumption of homogeneity was violated. tool developed for fifth and sixth grade stu­ However, the AN OVA is not sensitive to viola­ dents. A total of nine scales were developed tions of the assumption of normality for an to reduce the number of items investigated independent variable with a fixed number of and to avoid inflating the probability of find­ levels (Shavelson 1988). Omega square was ing significant differences among the three conducted to test the strength of the asso­ ethnic groups. Table 4 illustrates sub-scales ciation for each of the analyses completed and reliability coefficients. Data were cleaned as the sample size was very large. Table 5 using the method prescribed by Tabachnick provides the results of the AN OVA tests. and Fidell (1989 67). Univariate outliers were Neither Disruptive School Behavior nor removed within plus or minus 3 standard Parent Communication yielded a significant deviations for each variable. Multivariate comparison indicating that the means for the analyses were not planned. A composite three groups on these scales are essential­ ethnicity variable was developed using the ly the same. However the reliability for parent three largest ethnic categories. This variable communication scale was extremely low. was used as the independent variable in one Usually when the scale was unreliable signif­ way analysis of variance in order to compare icant differences among the means were not the three different ethnic groups on the pro­ found. However, a number of the compari­ tective and risk factors measured with the sons were significant. Post hoc Sheffe tests outcome instrument. were conducted to identify which means were significantly different for each of the scale RESULTS comparisons. For the School Protective Fac­ Several of the sub-scales were shown to tors scale, Anglo students had the highest have low or very low alpha coefficient sub­ mean score followed by Native Americans. scales that were found to be particularly un­ Latino students scored the lowest on this reliable were PARENT COMMUNICATION, scale indicating the fewest protective factors. FAMILY BONDING and a two item composite Anglo students scored significantly higher score for illicit drug use. These low reliabil­ than Latino students with no significant differ­ ities should be considered in the following ence between the mean scores of Latino and analysis. Native American students on this scale. For Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 83

Table 5: ANOVA Results Including Mean Scores and Omega Square Tests Omega Scale N Mean DF F Sig. Squared Disruptive School Anglo 169 2.17 Between 2 1.159 0.314 n/a Behavior Latino 266 1.79 Within 1013 Nat. Am. 581 1.86 Total1015 Protective School Anglo 169 37.68 Between 2 11 .674 0 0.0215 Factors Latino 266 39.19 Within 1014 Nat. Am. 582 37.41 Total1016 Parent Communica- Anglo 169 6.41 Between 2 2.299 0.101 n/a Latino 266 6.96 Within 1013 Nat. Am. 581 6.56 Total 1015 Family Bonding Anglo 169 4.66 Between 2 5.279 0.005 0.0083 Latino 266 4.61 Within 1013 Nat. Am. 581 4.48 Total1015 30 Day Tobacco Anglo 169 0.005 Between 2 13.863 0 0.0247 Use Latino 265 0.003 Within 1012 Nat. Am. 581 0.17 Total1014 30 Day Illicit Drug Anglo 169 0.008 Between 2 8.4216 0 0.0145 Use Latino 265 0.007 Within 101 1 Nat. Am. 581 0.18 Total1013 Attitude Toward Use Anglo 169 34.47 Between 2 4.21 6 0.015 0.0063 Latino 265 34.42 Within 1012 Nat. Am. 581 33.76 Total1014 Perceived Availability Anglo 168 6.34 Between 2 13.742 0 0.0246 Lation 262 5.85 Within 1007 Nat. Am. 580 5.01 Total1009 Perceived Harm Anglo 158 6.83 Between 2 4.894 0.008 0.0082 Latino 247 7.13 Within 936 Nat. Am. 534 6.47 Total 938 the Family Bonding scale, Anglo students scale that measured student perception of once again had the highest mean score and the harm from using drugs, alcohol, and to­ this was significantly different from the mean bacco, Latino students perceived substance score for Native Americans which were the use to be significantly more harmful than lowest on this scale. There was no signifi­ Native American and Anglo students did. cant difference between the mean score for Native Americans were found to to have sig­ Latino students and Native American stu­ nificantly lower scores when compared to dents. either Anglo or Latino students. There was For both 30-day tobacco use and 30-day no difference between Latino and Anglo stu­ illicit drug use, Native Americans reported dents on this scale. significantly higher use than the other two Omega square tests also were con­ groups. Latino and Anglo students did not ducted to identify how much of the variance differ in their drug or tobacco use as mea­ in the dependent variables was accounted sured by this instrument. In terms of the Atti­ for by the comparison among the three eth­ tudes Toward Drug and Tobacco Use scale, nic groups. The largest amount of variance Native Americans scored significantly lower accounted for was on the 30-day Tobacco than the other two groups indicating that they Use, Perceived Availability, and School Pro­ have more liberal attitudes toward drug use tective Factors scales. However the amount as compared to their non-Native peers and of variance accounted for on these scales would explain the higher rates of drug use in was between 2.15 percent and 2.4 7 percent, the Native American sample. There was no relatively small amounts. Overall, although difference between the Anglo and Latino stu­ there were statistically significant differences dents on this scale. Native Americans also among the three groups related to ethnicity, considered drug, alcohol, and tobacco more other factors are probably accounting for available than their non-Native peers did. more variance than this variable does. Their score on this measure was significantly lower than their peers who did not differ from DISCUSSION each on the Availability scale. Finally, on the A relatively large sample of elementary 84 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology school aged youth in grades four through six found both at the statewide and national lev­ took a standardized risk and resiliency sur­ els related to excessively high academic fail­ vey to assess baseline attitudes and behav­ ure rates. Drug prevention initiatives for La­ iors with regard to alcohol, tobacco, and other tinos must include some form of school en­ drug use. This sample included three large hancement strategies that could involve their ethnic groups with no prior exposure to sci­ English and non-English speaking parents. ence-based prevention programs in their In addition, our result suggests that there is schools and communities. First, our results a particular need for addressing attitudinal demonstrate the success of the New Mexico risks and perception of availability in tribal State Incentive Grant Initiative in recruiting a and other native American communities. uniquely multiethnic sample of children . Prevention and evaluation research on model Demographic information suggests a great programs implemented in tribal communi­ deal of cultural variation in the sample, with ties must ensure their success in strength­ a number of distinct languages spoken in ening these factors among Native American the homes of these children. In addition, this children and may require specific cultural ad­ sample was comprised of children residing aptation to existing science based preven­ in both rural and urban communities, and tion programs. Finally, more research is also included youth who reside in tribal com­ needed in the development of valid and reli­ munities. able measures for multiethnic children given There were significant differences be­ that existing standardized tool used in this tween the groups. Results of the eight com­ study were shown to have some weak­ parisons were significant indicating that the nesses, particularly Parent Communication groups differed on the risk and resiliency and Family Bonding scales. scales and drug use rates, although some of the Omega Square tests indicated that the REFERENCES significant differences did not account for a Botvin GJ, KW Griffin, T Diaz, & M lfiii-Williams. significant amount of variance in these mea­ 2001 . Drug abuse and prevention among mi­ sures. The strength of association was small. nority adolescents: posttest and one-year fol­ Of particular interest were the higher use low-up of a school-based preventive interven­ rates of alcohol and drugs among Native tion. Prevention Sci 2 1 1-13. Cervantes RC & F Ortiz. 2003. Substance abuse American children, and this coincided with among Chicanos and other Mexican groups. their more liberal attitudes toward drugs and Pp. 325-352 in The Handbook of Chicanalo lower perceptions of harm. Tribal and Pueblo Psychology and Mental Health. RJ Velasquez, communities may experience a number of LM Arellano, & BW McNeill, eds. Lawrence historical and current day trauma that could Erlbaum Associates, Inc. result in increased risk for drug use, although Dawson DA. 1992. The effect of parental alcohol it must be noted that this sample is not repre­ dependence on perceived children's behavior. sentative of all children statewide but only J Substance Abuse 4 4 329-340. those recruited into drug prevention services. Dielman TE, DO Klsoka , SL Leech, JE Schulenberg, & JT Shope. 1992. Susceptibility to peer pres­ Overall, Anglo students faired much better sure as an explanatory variable for the differen­ across all the eight risk and protective fac­ tial effectiveness of an alcohol misuse preven­ tors and had low drug, tobacco and alcohol tion program in elementary schools. J School use rates. Health 62 6 233-237. Future research must continue to focus Dielman TE, JT Shope, SL Leech, & AT Butchart. on ethnicity· and culture in explaining differ­ 1989. Differential effectiveness of an elemen­ ences in risk and protective factors. State­ tary school-based alcohol misuse prevention wide, normative samples using similar risk program. J School Health 59 6 255-263. Gensheimer LK, MW Roosa, & TS Ayers. 1990. and protective measures could shed addi­ Children's self-selection into prevention pro­ tional light on these issues. In addition, grams: evaluation of an innovative recruitment based on data from this study, drug preven­ strategy for children of alcoholics. Amer J tion strategies must be uniquely tailored to Community Psycho/ 18 5 707-723. meet the need of different groups. For ex­ Gosin M, FF Marsiglia, & ML Hecht. 2003. Keepin' it ample, our study suggests that Latino stu­ R.E.A.L. : a drug resistance curriculum tailored dents have few school related protective fac­ to the strengths and needs of pre-adolescents tors in the school domain. These are the of the southwest. J Drug Edu 33 2 119-142. same factors that other researchers have Hawkins JD, RF Catalano, & JY Miller. 1992. Risk Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 85

and protective factors for alcohol and other The Annie E. Casey Foundation . 2002. Kids Count drug problems in adolescent and early adult­ Data Book. pp. 120. hood: implications for substance abuse pre­ Tabachnick BG & LS Fidell. 1989. Using Multivari­ vention. Psychological Bu//112 64-105. ate Statistics. Second edition. California State Kids Count Census Data. 2003. Population refer­ University, Northridge: Harper Collins. ence bureau, analysis of data from US census Werch CE, OM Owen, JM Carlson, CC DiClemente, bureau. Retrieved November 28, 2003. Online. P Edgemon, & M Moore. 2003. One-year fol­ http://www. aecf. org. low-up results of the STARS for families alco­ New Mexico Voices for Children. 2003. Minority/ hol prevention program. Health Edu Res 18 1 Majority: A Profile of New Mexico's Children 74-87. 2003. A Kids Count Special Report. Pp 12-18, 32. ACKNOWLEDGMENT Shavelson RJ . 1988. Statistical Reasoning for the This research was made possible through fund­ Behavioral Sciences. Second edition. Boston, ing provided by the New Mexico Department of MA: Allyn and Bacon, Inc. Health/Behavioral Health Services Division. The Shope JT, LA Copeland, BC Marcoux, & ME Kamp. authors want to thank Cindy L. Keig and L. Monique 1996. Effectiveness of a school-based sub­ Lopez for their assistance in data managment and stance abuse prevention program. J Drug Edu preparation for this manuscript. 26 4 323-337. 86 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Sociology • the Social Sciences • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • I BIRDS IN THE HAND

If one bird in the hand is worth two in the bush ... Our two sources are invaluable ... and right at your fingertips.

For current thought and research in sociology and the social sciences, consult

sociological abstracts (sa) and Social Planning/Policy & Development Abstracts (SOPODA)

Abs~r ac:s of art·ces. books ~PC cor'e·e~ce par:ers 'ror rrore ~ han 2,500 J:::Jr"a s p ... o­ lished n 35 cour.tnes c1:a: ors oi relevan; J1sse-:a: ons a~::; book aro cthe· rre:Jta re.;1ews C mprehersrve, ccs:-efec: .e. - ret.

Ava lable n pr nt, on"r'e. Y t1e sociofile C:> ~Or,\ a~ a en rrag~et1c :aoe Our Web s1 te, http://www.socabs .org, 'ecL'eS r'le ,.,io~e Us reNs,ette• nformatron O" support serv1ces and don. me t oel1ver,. rKs tc re e•an: s tes. ar- d :he Sf I Web Search Ser 11ce offer ns ·ea­ so ably priced suoscr p: cr·s lc t~>c s~8se : s N·arr.age a'~a :aldy Issues & LaN, c~ine a"d Penology Ill P0 Box 22206, Sar, D1ego, CA 92192-0206 6191695-8803 • ~ax 6' 0 695-0416 • ematl soc o@cer"etcorr Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 87 TRAUMA REGISTRIES AS A POTENTIAL SOURCE FOR FAMILY VIOLENCE AND OTHER CASES OF INTIMATE PARTNER VIOLENCE FOR BORDER COMMUNITIES: INDICATOR DATA TRENDS FROM 2000-2002

Nickalos A. Rocha, Thomason General Hospital; Alberto G. Mata, Jr., University of Oklahoma; Alan H. Tyroch and Susan Mclean, Texas Tech University Health Science Center; and Lois Blough, Thomason General Hospital In developing a measured public health 1999). IPV and related family violence are response to Intimate Partner Violence (IPV) among the more common manifestations of and other manifestations of family violence violence. IPV incidents and their conse­ in U.S.-Mexico Border communities, there is quences are more commonly experienced a need for substantive, valid and reliable by border community's' residents than oth­ data. While in the past decade, border gate­ ers. Yet, border communities IPV issues and way cities have drawn national media and concerns are missing from major national policymakers' attention and interest; this at­ and state efforts (FBI 2001; US Department tention has been largely on narco-trafficking of Justice 1998; Center for Policy Research and drug related violence. In El Paso and Stalking in America 1997; Trapped by Abuse; Ciudad Juarez, attention has been drawn to The Taylor Institute 1997), especially data serial killings of young Mexican women. Yet, which could be used to inform preventive and little or next to nothing is reported in terms of intervention IPV services. These monitoring IPV and other manifestations of family vio­ and surveillance data could be used to in­ lence in these communities. There is a clear form preventive and intervention services for need for behavioral health data attending to border communities (Mata, Rocha, Blough, IPV and other manifestations of family vio­ & Lopez 1999). lence: its etiology, epidemiology and conse­ In Texas, family violence incidents have quences. not increased from 1997 to 2001 . But the For the last fifty years, the U.S.-Mexico bor­ number of women killed by an intimate part­ der region's communities have been seen ner has increased about 10 percent from largely as gateway cities for Mexican nation­ 1997 to 2001 . Most batterers were 20 to 24 als and Mexican-Americans to other parts of years of age followed by 25 to 29. Yet, in Texas, the U.S. The region's distinct communities 35 percent of female homicide victims are along the U.S.-Mexico border have experi­ murdered by an intimate partner. This is a enced great growth and change. These com­ rate that is substantially higher than reported munities have areas and families that may by the Federal Bureau of Investigations (FBI) be viewed as underserved. The border twin (Texas Department of Public Safety 2001). cities pose special political problems for To date, there are no major reports of border policymakers and those seeking to meet cities' IPV incidents and consequences simi­ public health and social service needs. These lar to those being reported to the state by border cities and communities have long­ non-border cities. standing and emerging problems. Yet, these There is a clear need for data and data problems are missing from national and systems that attend to major aspects of IPV state policymakers' agendas and priorities and related family violence manifestations (Texas Department of Human Services (Tjaden & Thoennes 2000; Gazmararian, 1998). This invisibility has left social services Petersen, Spitz, Goodwin, Saltzman, & Marks programmers and public health practitioners 2000; Straus & Gelles 1990; APA Taskforce with limited options or alternatives for ad­ on Violence and Family 1996; Tolman & dressing old, new and re-emerging public Raphael 2000). Medical centers' Trauma health agenda issues. Registries (TR) can play a pivotal role in iden­ Violence is one of those issues (Paul­ tifying key IPV and other related family vio­ ozzi, Saltzman, Thompson, & Holmgreen lence issues, vulnerability and other conse­ 2001 ; National Research Council 1996; quences (Wisner, Gilmer, Saltzman, & Zink Felitti, Anda, Nordenberg, Williamson, Spitz, 1999; National Research Council1996; Felitti Edwards. Koss. & Marks 1998: Wisner et ::~1 Pt :::.11 QQA· f'::<>?nl<:>r<:>ri<:>n ct <>I ')(\1'11'1· 1!. ~~II~~~~ ; 88 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology cal News 1992). ered ." Yet, the problems of IPV and related In this paper, we will explore the role of family violence largely remain outside the hospitals TRs, their use and limitations for scope and interest of national and state pub­ identifying family violence cases and the data lic health policy and programming efforts (US implications for border communities' policy Department of Justice 2001 ; FBI 2001) . and programming efforts. In short, what is Most concerns about violence remain link­ the role and nature of hospital TRs for bor­ ed to drug-related sexual assault and to drug­ der IPV surveillance & monitoring systems? related violence, thus placing them largely What are these TR data promises and limi­ within the criminal justice perspective. Bor­ tations? What is the overlap between "IPV" der communities IPV and related family vio­ cases to "drug" and "alcohol-related" trauma lence have been outside the scope of the and to "other trauma" in general? What alter­ Department of Justice's (DOJ) Victimization native perspective and implications do TR studies. They have also been outside CDC's monitoring systems pose? What is the pos­ special reports or other Department of Health sibility for border-wide IPV surveillance sys­ and Human Services (DHHS) IPV monitor­ tems? The data highlighted in this article is ing and surveillance efforts. In border com­ from the TR in El Paso, Texas at Thomason munities, comprehensive, coordinated and Hospital. effective IPV services are missing and unad­ Family Violence and other IPV issues re­ dressed (Trapped by Poverty Trapped by main a key concern among communities Abuse, Taylor Institute 1997). along both sides of the U.S .-Mexico border. While subject to occasional studies, Intimate partner violence-or IPV-is defined as these fail to adequately address IPV and re­ threatened physical or sexual violence or lated family violence issues and its public psychological and emotional abuse directed health consequences. Many health and hu­ toward a spouse, ex-spouse, current or form­ man service professionals in border com­ er boyfriend or girlfriend, or current or former munities recognize the serious, impacting dating partner. Intimate partners may be het­ nature and consequences of IPV and family erosexual or of the same sex. Some of the violence-yet there is little data or assess­ common terms used to describe intimate ment of the problem nor is there information partner violence are: domestic abuse, spou­ about how border communities compare to sal abuse, domestic violence, family vio­ non-border cities. While recognizing the po­ lence, courtship violence, battering, marital tential and limitations of border communi­ rape, and date rape (Saltzman, Fanslow, ties' IPV data, there is a clear need to expand McMahon, & Shelley 1999). The Centers for and enhance its use as well as measures of Disease Control and Prevention (CDC) use its health consequence (National Institute of the term intimate partner violence as it de­ Justice 1998; US Department of Justice scribes violence that occurs within a range 1994a, 1994b, 2001 ; and National Research of intimate partner relationships. Some of Council 1996). Border hospitals TR's are in the other terms are overlapping and may be unique positions to fill this IPV gap. used to mean other forms of violence includ­ We suggest that TRs have great potential ing abuse of elders, children, siblings, and for addressing IPV and related family violence other family members (CDC 2003). National, health consequences. Here we suggest the state and Third Sector effort's need to assist importance and limits of Texas-based TRs local communities seeking to provide IPV and for monitoring, surveillance and policy re­ related family violence services. The contin­ search. Second, we will discuss how utiliz­ ued use and improvement of TRs by state ing TRs in border communities will help re­ and national health and mental health ser­ define the U.S .-Mexico Border IPV and re­ vices could serve to give policymakers, re­ lated family violence issues. Then, we will searchers and practitioners the important present data that profiles key patterns and data they need to address the IPV problem. trends for major border gateway cities. These are data that need to be compared and con­ THE NEED TO REDEFINE IPV AND RELATED trasted to other border cities. Lastly, we dis­ FAMILY VIOLENCE IN BORDER cuss the potential of the borders' IPV data for COMMUNITIES monitoring, reporting and service planning. Periodically, the U.S.-Mexico border com­ Here we will suggest that TR data are avail­ munities' health problems are "rediscov- able for all border cities and can provide im- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 89

Table 1: Trauma Essential and Desired Reporting Data Items as Per The Texas Department of Health ESSENTIAL DESIRED Facility Number Trauma Number Medical Record Number Patient Name Race/Eth nicity Social Security Number Sex Pulse Date of Birth Revised Trauma Score Date of Injury Lenth of Stay County of Injury ICU Days Cause of Injury Five AIS Codes Time of Injury Five ICD9 Procedure Codes County of Residence Five ICD9 Pre-existing Condition Codes Place of Injury Body Region X Severity Date of Arrival Body Region X Surgery Time of Arrival Revised Trauma Score at Scene Alcohol Level Tested Ambulance Firm Number Alcohol Level Total Reimbursement Blood Pressure Vehicle Extrication Respiration Rate For First Hospital: Date of Arrival Glassgow Coma Scale Time of Arrival Discharge Destination Date of Departure Discharge Condition Time of Departure Date of Discharge or Death Time of Discharge or Death If Discharge to Facility, Facility Number Five ICD9 Diagnostic Codes Injury Severity Score Payor Category Billed Hospital Charges Systolic Blood Pressure at Scene Glassgow Coma Score at Scene Dispatch Time Arrival Time Leave Scene Time Protective Devices Transfer Referral Facility Source: Texas Department of Health, 2002 portant IPV health consequences data. State mandated reporting usually includes These data are useful not only for applied patient demographics, injury severity, medi­ and administrative research but they can cal care procedures, health outcomes, and serve as a basis for prospective basic and medical costs. In terms of injury surveillance policy research. and monitoring purposes for the U.S.-Mexico Border region, TRs have important potential TRAND FAMILY VIOLENCE AND IPVCASE for the highlighting of IPV and related family MONITORING: PROMISE AND LIMITATIONS violence issues. In Texas, TRs have been established for Since the first TR in Chicago in the nearly twelve years (Rocha, Mata, Tyroch, 1950's1, their role has been to monitor and McLean, & Blough 2005). While data for this evaluate trauma patient care for healthcare essay was generated from Thomason entities and the regional EMS systems that Hospital's TR which was initiated in 1994, they belong to; to identify and report major this paper only covers the year 2001. As in trauma injuries and outcomes; and to pro­ other parts of the country, hospitals used TRs vide a sense of how to prevent, treat and re­ to measure the quality of trauma care and to duce trauma costs. TRs are databases that evaluate the effectiveness on health out­ collect, archive and report information about comes. A second major utility of TRs is for patients that they receive through a trauma injury surveillance, patient care and patient care services system. Patient inclusion into cost. The data collected varies by local TR. a TR system generally requires that the pa- 90 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology tient population meet specific criteria: has continued since then to provide TDH­ mandated minimal trauma level data and •lCD Codes (967.0-967 .9) serve as the lead hospital for the Far West •Admission to ICU or hospital floor Texas and Southern New Mexico Regional •ICD9Code2 Advisory Council on Trauma (FWT & SNM •Injury Severity Score 3 RAC) . The state of Texas is divided into 11 RACs. The RAC for the Far West Texas and In 1989, the Texas legislature recognized Southern New Mexico region has eight hos­ the need and challenge that collecting stand­ pitals which participate in pooling TR data ardized trauma data from over 450 hospitals on an ongoing basis5. The FWT & SNM RAC would present and allowed reporting enti­ is unique, in that it covers four Texas coun­ ties to file their data electronically either on a ties and seven New Mexico counties6 . quarterly or annual basis. In Texas, four re­ Trauma care is provided through a four-tier gions ranging from El Paso to Brownsville system of providing care to acute and injured cover the Texas border (TDH 2001) . In 1990, patients. Level one trauma centers are ter­ the state legislature mandated the reporting tiary care facilities central to any Trauma of certain trauma cases. On August 31 , 1996, Care System (TCS). Level twos provide ini­ the Texas state legislature required the tial definitive care regardless of severity of state's Department of Health and hospital injury. They can be academic, community, trauma units to gather data about trauma in public or private facilities located in rural , Texas. One objective was to identify severely suburban and urban settings. The following injured trauma patients within each health­ describes each of the four tiers of this trauma care agency. Others monitoring patient care core system. within each hospital unit and regional emer­ gency medical services network were re­ Levell: quired to identify the total amount of uncom­ A Level I facility is a regional resource pensated trauma care delivered each fiscal trauma center serving as the area's tertiary year. All medical facilities needed to report to care facility. Tertiary Care Centers are cen­ the Texas Department of Health (TDH) Injury tral to the trauma care system. Each facility and Control Division. Minimal data sets con­ must have the capability of providing leader­ sist of TR data that is required by the Texas ship and total care for every aspect of injury, Department of Health (TDH). Due to the need from prevention through rehabilitation . In its for confidentiality, all public health reports of central role, a Level I center must have ad­ data are reported in the aggregate. Also , se­ equate Emergency Care facilities and per­ curity measures and guidelines were devel­ sonnel. Because of the large number of per­ oped to limit access to registry data. Gener­ sonnel and facility resources required for ally, TRs include all cases with lCD 9 codes patient care, education, and research, most of 800 to 959. It must be recognized that the Level I trauma centers are university-based collecting of standardized data is set by the teaching hospitals. Other comprehensive state legislature and corresponding state hospitals willing to commit these resources, agency(s).4 however, may meet the criteria for achieving The actual collection of TR data is guided a Level I certification recognition . by state and hospital reporting guidelines. In addition to acute care responsibilities, This mandate has allowed hospitals to re­ Level I trauma centers have the major re­ port required essential elements as well as sponsibility of providing leadership in educa­ desired optional elements. In a sense, there tion, research and system planning. This re­ are minimal required reportable and desired sponsibility extends to all hospitals caring data elements. Desired data elements are for injured patients in their regions. Medical variables, which state, professional, and education programs include residency pro­ some local agencies would like to see col­ gram support and postgraduate training in lected, but are not mandatory. In Texas, Table trauma for physicians, nurses, and pre-hos­ 1 lists the hospital data items and whether pital providers. Education can be accom­ their collection is essential or desirable. plished through a variety of mechanisms, Thomason Hospital is an American Col­ including classic continuing medical educa­ lege of Surgeons verified Level !-trauma fa­ tion (CME), training institutes, preceptor­ cility. Thomason initiated the TR in 1994 and ships, personnel exchanges, and other ap- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 91

Table 2: Number of Total Trauma Admissions by Year for Drugs and Alcohol for Thomason Hospital and the Far West Texas and Southern New Mexico Regional Advisory Council on Trauma Between 1996-2001 Number of Trauma Admissions By Year Far West Texas and Southern New Mexico Regional Advisory Council on Trauma Between Thomason Hospital 1996-2001 1996 1997 1998 1999 2000 2001 1996 1997 1998 1999 2000 2001 Total Admissions1031 1145 1496 1663 1595 1653 1046 1769 2031 2299 2789 2735 Drugs Only 151 200 264 219 248 207 151 270 333 265 248 208 Males 117 153 198 179 212 161 117 214 252 209 212 162 Females 34 4 7 66 40 36 46 34 56 81 56 36 46 Alcohol Only 439 330 420 413 382 338 439 504 629 510 484 441 Males 349 280 348 347 333 285 349 419 502 427 412 371 Females 90 20 72 66 49 53 90 85 127 83 72 70 Source: Thomason Hospital Trauma Registry preaches appropriate to the local situation. Level IV: Research and prevention programs are es­ Level IV trauma facilities provide trauma sential for a Level I trauma center. These life support prior to patient transfer in remote hospitals provide important services and areas where no higher level of care is avail­ data to community outreach and education able. Such a facility may be a clinic rather as it concerns serious injuries and trauma than a hospital and may or may not have a (CTACS 1999). physician available. Because of geographic isolation, however, the Level IV trauma facil­ Level II: ity should be an integral part of the inclusive The Level II trauma center is a hospital trauma care system. As at Level Ill trauma that is also expected to provide initial defini­ centers, treatment protocols for resuscita­ tive trauma care, regardless of the severity of tion, transfer protocols, data reporting, and injury. However, depending on geographic participation in system performance im­ location, patient volume, personnel, and re­ provement (PI) are essential. sources, the Level II trauma center may not A Level IV trauma facility generally has a be able to provide the same comprehensive good working relationship with the nearest care as a Level I center (for example, pa­ Level I, II, or Ill trauma center. This relation­ tients requiring extended surgical critical ship is vital to the development of a rural care). Level II trauma centers, however are trauma system in which realistic standards the most prevalent type of facility in a com­ must be based on available resources. Op­ munity that manages the majority of trauma timal care in rural areas can be provided by patients (CTACS 1999). skillful use of existing professional and in­ stitutional resources supplemented by guide­ Level Ill: lines that result in enhanced education, re­ The Level Ill trauma center serves com­ source allocation, and appropriate designa­ munities that do not have immediate access tion for all levels of providers. Also, it is es­ to a Level I or II institution. Level Ill trauma sential for the Level IV facility to have the in­ centers can provide prompt assessment, re­ volvement of a committed health care pro­ suscitation, emergency operations, and sta­ vider, who can provide leadership and sus­ bilization; and may also arrange for possible tain the affiliation with other centers. These transfer to a facility that can provide definitive facilities are key to providing critical care in trauma care. General surgeons are required many border communities (CTACS 1999). in a Level Ill facility. Planning for care of in­ Along the border, there are few of these jured patients in these hospitals requires key critical care institutions. Thus levels Ill & transfer agreements and standardized treat­ IV hospitals are the major community trauma ment protocols. Level Ill trauma centers are resource for their respective border commu­ generally not appropriate in an urban or sub­ nities. In most major cities substance abuse urban area with adequate Levell and/or Level injury surveillance is possible through TR, II resources (CTACS 1999). the state respective transportation depart- 92 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Table 3: Thomason Hospital Trauma Patient Profile for Selected Categories in 2001 2001 Characteristic IPV Drug Only Alcohol Only Other 26 201 25 207 Number (N) 25 25 25 25 Charges Average Charges $9 ,068 $36,555 $18 ,614 $12 ,247 Sum Charges $226,698 $913,886 $465,361 $306,183 Injury Severity Score Avg ISS 7 15 7 6 Range 25 37 18 23 Type of Injury Blunt 44% 84% 80 % 92% Penetrative 52% 12% 16% 4% Other 4% 4% 4% 4% Site of injury* Street/highways 12% 60% 56% 52% Home 64% 8% 20 % 16% Specified other 12% 28% 20% 12% *Top three site of injuries will not total to 100%. Source: Thomason Hospital Trauma Registry ment and criminal justice arrests and con­ Hospital for drug-related issues. There was viction reports. All of these data systems re­ a 3 percent increase in males being admit­ flect various aspects of substance abuse ted for alcohol-related cases between 1997- consequences. 2000. However, for females there was a 2 per­ BACKGROUND AND CONTEXT cent decrease for alcohol-related trauma and SUBSTANCEABUSEAMONGTRAUMA an 8 percent increase for drug-related cases. ADMISSIONS-THOMASON HOSPITAL: There was no significant age increase when A profile of the selected groups examining the data by individuals being ad­ During the 2001 calendar year, there were mitted since 1997. 1 ,653 trauma admission cases as com­ Yet , while blunt trauma (Table 3) has been pared to 1,031 in 1996 (Table 2) . Since 1996, increasing penetrative trauma has been de­ there has been a 35 percent increase in total creasing (39%) since 1997, and blunt has trauma admissions. During the 2001 calen­ increased by 3 percent. dar year, over three-fourths (81 %) were His­ These data suggest variations in cases panic, 2 percent were African-American, and presenting to Thomason's Trauma Center. 1 percent were members of other racial/eth­ Marked differences may be observed be­ nic groups. There was a 39 percent increase tween drug, alcohol and non-substance between 1996 to 2001 of drug-related ad­ abuse-related trauma care patients. It is un­ missions. In terms of alcohol-related admis­ clear how these patterns maybe related to sions, the number has decreased signifi­ IPV cases. However, it does raise the ques­ cantly each year from 1996 to 2000. In 1996, tion : how are IPV cases similar to other types there were 439 alcohol-related admissions, of trauma? We next will discuss the study's which decreased to 382 cases in 2000, rep­ methodology and its results. resenting a 13 percent decrease. Forty percent of these admissions had Methodology used drugs or alcohol, whereas 16 percent The cases examined in this pilot study of the admissions had used "drugs only" were derived from female trauma admissions (Table 2). Eighty-five percent were male. A at Thomason Hospital during 2001 . This majority (27%) of trauma patients in 2000 analysis only includes females between the were between the ages of 18-25 and male. ages of 18-60 years. The variables included Between 1997-2000, there was a 21 percent were as follows: category of subjects, age, increase in the total number of drug-related ICD9 codes (one to seven codes per patient), cases seen at Thomason Hospital. Also, Injury Severity Score, county of residency, there was an 8 percent increase in the num­ hospital charges, insurance payor (insurance ber of males being admitted to Thomason paying the hospital charges), date of admis- Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 93

Table 4: Injury Severity Score Calculation Region Injury Description AIS Square Top Three Head & Neck Cerebral Contusion 3 9 Face No Injury 0 Chest Flail Chest 4 16 Abdomen Minor Contusion of Liver 2 Complex Rupture Spleen 5 25 Extremity Fractured femur 3 External No Injury 0 Injury Severity Score: 50 Source: Center for Disease Control, 2003 sian, blood alcohol level, positive for alco­ form ; hol, positive for illicit drugs, number of illicit •an alphabetical index to the disease en­ drugs in system, medical record number, tries; and ethnicity, etiology, type of injury (blunt or pen­ •a classification system for surgical, diag­ etrating) and other recorded variables. This nostic and therapeutic procedures (al­ pilot study's total sample size is 100 trauma phabetic index and tabular list). cases (Table 3) . The design utilized a case control methodology. A case-control study can The process (Figure 1) begins by ab­ identify risks and trends, and can suggest stracting trauma patients, then entering them some possible causes for particular out­ into Thomason's TR. The TR utilizes the Col­ comes of a program. The trauma cases for lector (a TR software package) and is critical alcohol, drug and other trauma were ran­ to the development and maintenance of this domly selected. The IPV cases were se­ database. The Collector software version lected from an unscreened population. The utilized was 3.28, 1997-2002. Collector is cases were then followed-up by pulling the one of the commercial software packages trauma registry record and verifying these used in the state of Texas for trauma regis­ were documented IPV cases. tries. Then, after abstracting, we used the The ISS is an anatomical scoring system Statistical Package for Social Sciences that provides an overall score for patients (SPSS). Specifically, we selected a random with multiple injuries. Each injury is assigned sample using SPSS to compare to the IPV an Abbreviated Injury Scale (AIS) score and cases (SPSS 1999, Ver. 10). is allocated to one of six body regions (Table The sample was then categorized into 4) (Head, Face, Chest, Abdomen, Extremi­ four groups: 1) self reporting and/or ED staff ties, including Pelvis, External). ISS range identification of IPV cases; 2) female trauma from 1-75, with 1-14 being stable and 15 or admissions to Thomason Hospital who sub­ higher being critical. Only the highest AIS sequently tested positive for alcohol; 3) fe­ score in each body region is used. The 3 male trauma admissions to Thomason who most severely injured body regions have their were under the influence of illicit drugs and score squared and added together to pro­ finally 4) females who were admitted to duce the ISS score. We now will turn to spe­ Thomason for Trauma and who were not cific measures and protocols (CDC 2003). under the influence of any illicit drugs or al­ An example of the ISS calculation is cohol. These categories allow one to com­ shown below: pare and better understand how IPV cases differs from other types of trauma. Moreover ICD-9-CM is the official system of assign­ it also other similar types of trauma. ing codes to diagnoses and procedures as­ sociated with hospital utilization in the United FINDINGS States. The ICD-9 is used to code and clas­ Overall Sample Characteristics sify mortality data from death certificates Almost half of the TR's IPV cases are fe­ (CDC2003). males between the ages of 18-24 (26%) or 31 -40 (24 %). As this is a major Texas-Mexico The ICD-9-CM consists of: border community, a majority of the patients were Hispanic (79%). Followed by White, •a tabular list containing a numerical list of Non-Hispanic 19 percent. The two most com­ the disease code numbers in tabular mon sites of iniurv were motorwavs (45%) 94 Volume 34 No_ 1 Mav 2006 Free lnauirv In Creative Socioloav

Figure 1 f.'i!(. I Trauma l'atinll \flmi,.., ion l'ron·-.-.

Trauma UqJartnwnt

IJ .J lllll.! l';llJ~·nt

( ·.: llll'J ----- '-'------.. r r------~------~

:'-kdJ-- ,,-~il l

Table 5: Thomason Hospital Trauma Patient Profile for Selected Categories in 2001 2001 Characteristic IPV Drug Case Alcohol Case Other 26 201 338 207 Number (N) 25 25 25 25 Age 18-25 32% 32% 28% 32% 26-35 36% 24% 28% 20% 36-46 20% 40% 40% 12% 47+ 12% 4% 4% 36% Race/Ethnicity Hispanic 72% 88% 68% 88% White-non-Hispanic 24% 12% 28% 12% African-American 4% 0% 0% 0% Other 0% 0% 4% 0% Primary Payor* Bluecross/Biueshield 40% 36% 36% 40% Medicaid/Medicare 24% 36% 36% 36% Self pay 32% 16% 12% 8% *Will not total100% Source: Thomason Hospital Trauma Registry

Table 6: Site of Injury for Thomason Hospital Trauma Patients 2001 2001 Site IPV Drugs Alcohol Other Number (N) 25 25 25 25 Street/Highway 12% 60% 56% 52% Home 64% 8% 20% 16% Public Building /Residential Institution 0% 4% 4% 8% Other (specified & unspecified) 24% 28% 20% 12% Unspecified 0% 0% 0% 12% Source: Thomason Hosptial Trauma Registry azepines were the next most often used sub­ Trauma Center in 2001 . The racial/ethnic stances. composition was: 1) 72 percent Hispanic; 2) Trauma patients' payment of trauma care 24 percent White and 3) 4 percent African­ hospital charges were largely through pri­ American. While the average age was 32 vate health insurance (38%) followed by pub­ years of age; the ages ranged from 19 to 58 lic health insurance (30%). The remainders years of age. Among IPV cases the average were self-paying patients. Trauma care ISS was 7, and the scores ranged from 1 to (Table 5) is primarily paid for by the private 26. A majority of cases (88%) were from El and commercial health insurance compa­ Paso County with only 12 percent of cases nies. We will now turn to closer examination being out of state. of this overall sample by comparing IPV pa­ These cases were identified at admission tients to drug, alcohol and the other trauma with a trauma-related ISS that required fur­ cases. ther injury examination and treatment. While The only group whose site of trauma did blunt injuries comprised 44 percent, pene­ not occur most often on the motor-ways was trating cases consisted of 52 percent, with IPV cases which happened at home (64%). only 4 percent reporting unspecified type of The other three subgroups had the majority injuries. The most frequently reported cases of trauma case occurring on the motorways involved lacerations of the chest (3), followed ranging from 52 percent to 60 percent. For by laceration of the anterior abdomen and IPV cases the next most often site of occur­ traumatic hemothorax. A few cases involved rence was "Other'' (specified & unspecified) multiple blunt and penetrative injuries that (Table 6). made categorizing difficult. Among these lat­ ter cases they could have involved both blunt IPV Related Cases and penetrative injuries. The first sub-group consists of all IPV (but The majority of cases occurred at their one) cases presenting to the Thomason domicile (64%). followed bv 24 oercent ::~t 96 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Chart 1: Type of Illegal Drugs Used by a Sample of Trauma Admissions at Thomason Hospital in El Paso, Texas, 2001

60% 52% 50% 40% 40°

?8%

0 ~ ,..~ 0"' -<:o-0 # o

~ Alcohol Onl N=25) unspecified location(s) and 12 percent occur­ ing. The ethnic composition was (88%) His­ ring on motorways. There were more pene­ panic and (12%) White, Non-Hispanic. The trative injuries then blunt injuries among IPV most commonly reported type of trauma in­ cases. Among the IPV trauma cases only two volved blunt trauma- 84 percent. The remain­ tested positive for drugs. The first case in­ ing were penetrative injuries (12%) followed volved a female patient who had used co­ by 6 percent for unknown cases. caine. The second case involved a female Fifty-two percent of women admitted were who had used marijuana. found to be legally intoxicated. Thirty-eight The average charge for IPV-related inju­ percent of alcohol-related cases were found ries at Thomason was $12,247.31 . The to have used cocaine (Chart 1). Additionally, range was $1 ,160.61-$119,871 .10. Among thirty-two percent had used benzodiazepines. IPV cases, 40 percent were private health Slightly more than 72 percent had used only insurance, while 32 percent were self-pay­ one drug while 28 percent had two or more ing and 28 percent utilized public health in­ illicit substances. surance. A little less then a third of all IPV The average charge for a female who was cases were self-paying patients. admitted to Thomason Hospital for traumatic injuries while under the influence of illicit sub­ Illegal Substance Abuse Related Trauma stances was $35,499.99, with charges rang­ Cases ing from $2 ,023 to $182,649.33. A majority of Females admitted to Thomason for trau­ trauma care charges were covered by pri­ matic injuries while under the influence of vate health insurance (36%) and public health illicit drugs ranged in ages from 31-40 with insurance (26%) respectively. Only one per­ the average age of 32. Forty percent were cent of these cases was self-paying. Examin­ between 31-40, 28 percent were 18-24. The ing Table 7, the data details each of the average ISS was 14.84 with others reporting subgroup's use of various substances - co­ scores from 1 to 38. These patients gener­ caine was the most commonly used with 32 ally suffered from either a closed skull base percent. For the alcohol subgroup, the larg­ fracture or intra cranial injury or pelvic frac­ est drug use group was cocaine. For IPV ture . A majority of trauma patients were from cases the largest drug use cases were co­ El Paso County (68%) . Thirty-two percent of caine and marijuana. Yet even among alco­ substance abuse cases were from out of hol only cohorts who also tested positive for state. drugs-cocaine was the preferred sub­ Among these cases, 60 percent occurred stance. on motorways and 28 percent occurred in unspecified places, while the remaining 12 Alcohol-Related Trauma Cases oercent occurred at home or in a oublic build- The majority of the population (68%), were Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 97

Table 7: Type of Illegal Drugs Used by a Sample of Trauma Admissions at Thomason Hospital in El Paso, Texas, 2001 *Drugs IPV Drugs Alcohol Number (N) 25 25 25 Cocaine 4% 32% 20% Marijuana 4% 8% 4% Benzodiazepines 0% 16% 4% Opiates 0% 8% 12% Barbituates 0% 4% 0% Amphetamines 0% 4% 4% Other 0% 0% 0% Unknown•• 40% 0% 16% None 52% 0% 40% Multiple Drugs* 0% 28% 0% *Multiple drugs comprised of Cocaine+ THC, Cocaine+Opiates, Cocaine+ THC+Opiates, Opiates+Benzo, Opiates+ THC. **Not tested Source: Thomason Hosptial Trauma Registry Hispanic with 28 percent being White non­ involving alcohol were $18,096.21 . The costs Hispanics and other ethnic groups com­ of treatment ranged from $2,023.25 to prised of 4 percent. Their ages ranged from $93,044.62. Among alcohol-related trauma 21 to 52 years of age. Among alcohol-related patients, there were three major cost mo­ trauma cases, the average age was 34. The dalities. Sixty percent of patients' costs ISS ranged from 1 to 19 with the mean ISS ranged from $1-$10,000, 16 percent ranged being 6.76. Only 16 percent were penetrative from $10,001-$20,000, and 12 percent injuries. Nearly a quarter of cases (24%) had ranged from $20,001 -$40,000. Among alco­ an ISS above 15. Most ISS were 14 or below. hol-related trauma cases, most patients had Eighty percent of cases were from El Paso. private (36%) or public insurance (39%). The Sixteen percent came from New Mexico. Also, remainder (25%) was self-paying patients. there was 1 case (4%) from Webb County. Most alcohol-related trauma cases oc­ Other Trauma-Related Cases curred on motorways (56%), followed by 20 "Other" trauma (OT) is defined as any other percent occurring at home. The remainder trauma which entered the ED and did not occurred elsewhere. Eighty percent of the have alcohol or illegal substances in their blunt injuries tested positive for alcohol. system at the time of admission. This also Among alcohol patients, 56 percent of the excluded any IPV cases. A majority of these sample did not have any illicit drugs in their individuals were Hispanic (88%), 12 percent system. Moreover, 18 of these patients (72%) were of White, non-Hispanic origins. Forty­ were found to be over the legal limit of intoxi­ eight percent of the sample were between cation-- the lowest being .02 and the high­ the ages of 18-24 or 31-40 years of age. The est being .34 or four times over the legal limit. other half of the cases were between the The presence of illicit drugs in this sub-group ages of 25-30 or 31 -50. The average age for was evident upon subsequent toxicology ex­ other trauma cases was 36 years of age. aminations. Overall, 44 percent of this sub­ The youngest case was 18 and the oldest group had used illicit substances. Among being 57 years of age. While 80 percent of these cases, 28 percent tested positive for a this sub-group was from El Paso, 16 per­ drug. Cocaine was the most frequently used cent of all cases were from out of state. Only substance. It was followed by heroin, mari­ four percent were from Hidalgo County. juana, Benzodiazepines and amphetamines. Fifty-two percent of these trauma injuries The rest of these women did not test positive occurred on the motorways. Trauma injuries for illicit drugs. Eighty percent had used at at home comprised only 16 percent. A major­ least two illicit substances. Moreover, 8 per­ ity of these injuries were blunt (92%) with cent tested positive for having used three or only 8 percent being penetrative. The mean more illicit drugs. Again, among alcohol pa­ ISS was 6, although they ranged from 1 to tients 56 percent of the sample did test posi­ 24. The crucial distinction when examining tive for any illicit drugs. ISS is as follows: 1-14 considers the patient The average charges (Chart 2) for cases to be stable and 15 to 75 are critical. The 98 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

Chart 2: Thomason Hospital Trauma Patient Cost Indicator Data, 2001

$1 .000.000 -

$900.000 $845- ,664 $800.000

$700.000

$600.000

$500,000 $402,997 $400.000 $305,022 $300.000 .,...-- I $226,698 $182,650 ,.--- L $200.000 r-- $93,045 $119.871 q:~"' {)~? - $100.000 - $~ 7 $ .o6n $ ~0 ~ $0 l $ ~ Alcohol Cases IPV Drug Cases Other Trauma Catagory

n Sum of Charges IJ Average Charges

1 1 Largest Trauma Cost

focus on ISS is that they reflect the probabil­ per are: ity of survival. Twelve percent had an ISS Trauma data suggest that Border hos­ above 15. Eighty-eight percent had an ISS pitals' TR admissions can identify and pro­ between 1-9. vide substantive, reliable data about IPV & The average costs were $9,067.91 with other family violence issues. Since the prob­ charges ranging from $400 to $63,061 .72. lem is extensive, it is important to compare Among OT-related cases, (40%) relied on IPV cases to other types of injuries and private health insurance, 36 percent on pub­ trauma in these communities. IPV cases lic health insurance, the remainder was self­ seem to be distinct from these other types of paying. trauma. Yet, the IPV case sample is too lim­ ited to speak about trends. For example, the CONCLUSION patterns for drug abuse cases seem to be The current findings are tentative yet sug­ increasing and alcohol abuse cases are gestive. There is a need for further social decreasing. Yet marked profile differences science research on IPV cases in border may be observed when one compares IPV communities. Continuing and comparative cases to other trauma (OT), drug cases and studies need to be conducted. The criminal alcohol cases. IPV cases seem to be justice perspective needs to be augmented younger than OT, alcohol cases and drug by public health and social service perspec­ cases. IPV cases are more likely to reflect tives, as well. The main findings of this pa- OT cases and alcohol cases are more likely Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 99 to reflect drug use case patterns. While IPV IPV cases were lower than drug or alcohol ISS are comparable to OT and alcohol categories. It is unclear if this is the case case's, they are half of drug cases. In terms for other border communities or is the case of costs of trauma services, I PV average case for Hispanics throughout the state. costs are slightly less than OT, alcohol and drug trauma cases. In fact, IPV cases seem The data and approach presented herein to be almost half of these data. Alcohol cases needs to be compared to other data from cost a third of the drug cases. other border hospitals. Border cities research Only a few IPV patients had used either may suggest which strategy is more likely to drugs or alcohol. It is unclear if their partner's bear fruit and meet changing and future de­ had used drugs or alcohol. These data are mands. Some effort needs to be spent on outside the scope of trauma registry data col­ assessing and improving the quality of TR lection protocols. data as related to IPV, alcohol and drug abuse Yet among drug and alcohol cases use of cases. TRs hold a major promise for ascer­ other drugs is common, nonetheless one taining health consequences of IPV cases. finds distinct patterns. Moreover, Blood Alco­ We recognize the need not to oversell or over­ hol Concentration (BAC) levels seem higher extend Trauma Center programs, staff and among drug cases, than among alcohol capabilities. However, this is a key corner­ cases or IPV cases. Most injuries for IPV oc­ stone institution that allows us to measure cur at home or other locations. Yet for drug serious health consequences of IPV behav­ and alcohol cases the most common sites ior. are motorways and other specified and un­ specified places. SUMMARY: THE NEED FORA BORDER In short, there is tremendous potential in SUBSTANCE ABUSE IPVTR PROJECT using TRs for IPV and related family violence Just as there is a clear need to develop, research. Thomason's TR data allows for cultivate, and evaluate drug abuse monitor­ measuring IPV, as well as drug and alcohol­ ing systems in Border communities, there is related violence. Illicit substances and alco­ also a need for the data that will be gener­ hol use are captured in some TRs. This al­ ated by those systems to include Family Vio­ lows for measuring the extent of alcohol and lence and IPV incidence, vulnerability and drug use in TR cases. While tentative, this consequences. While this administration , TR data suggest IPV trauma cases are dis­ like past administrations' has expressed an tinct from illicit drug-related trauma, alcohol­ interest in Border drug abuse issues, public related trauma but also other trauma. As pi­ health issues have been left wanting. The lot data, these differences need to be further need for Border public health surveillance studied and examined. There are limitations and monitoring systems can be partially met to the TR, but as collected today, they still by taking advantage of Border TRs. could be used to help identify, profile and Thomason Hospital's TR has suggested serve as baseline data for prospective stud­ useful social indicators of drug abuse pat­ ies. terns and trends. This also applies to IPV and related manifestations of family vio­ •In Texas, TRs can be linked statewide to lence. Moreover, closer examination of these assess substance abuse if data is collected data also suggest that pilot studies need to on a continuous basis. Wh ile some RACs be undertaken to enhance and expand the collect substance abuse data, many have validity and reliability of I PV data. Efforts must yet to focus on IPV and family-related vio­ be undertaken to improve IPV TR case iden­ lence. These data are not currently being tification and followup. Perpetrator data has reported to state trauma reg istries. TR data been clearly established for ED prospective can capture the cost of trauma care ser­ studies, but have not been developed for vices, we strongly urge that TRs include border TR-wide systems. There is a need to IPV data. This would allow for establishing establish collaborative projects along the cost of trauma care services to IPV cases Border to collect and analyze trauma data on a statewide basis. related to substance abuse and violence. •Some RAC regions collect substance abuse These data can be useful in: data, others do not. •In th is pilot study substance abuse related •Developing area and regional surveillance 100 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

systems underestimate the resources needed to initiate •Establishing the need for and the range of and maintain a registry. Herein, we describe services required for Border communi­ the purposes, resource requirements, and limi­ ties tations of trauma registries. 2The International Classification of Diseases (lCD) •Demonstrating the ways drug problems is designed to promote international compara­ impact Border communities bility in the collection , processing , classifica­ •Demonstrating the ways IPV and family vio­ tion , and presentation of mortality statistics. lence impact Border communities This includes providing a format for reporting •Demonstrating the costs related to sub­ causes of death on the death certificate. The stance abuse and IPV problems reported conditions are then translated into •Serving as a baseline for prevention and medical codes through use of the classifica­ allowing for specialized stud ies of Bor­ tion structure and the selection and modifica­ der communities tion rules conta ined in the applicable revision of the lCD, published by the World Health Orga­ nization . These coding rules improve the use­ Border Epidemiology Health Data Work­ fulness of mortality statistics by giving prefer­ groups would benefit greatly by involving ence to certain categories, by consolidating trauma registry programs and staffs in their conditions, and by systematically selecting a efforts. While border-wide monitoring and single cause of death from a reported sequence surveillance projects have long been touted , of conditions. The single selected cause for Trauma Centers and TRs represent an im­ tabulation is called the underlying cause of proved operating vehicle to provide impor­ death , and the other reported causes are the tant monitoring and surveillance data. IPV non-underlying causes of death . The combina­ baseline and trend data is lacking from the tion of underlying and non-underlying causes is the multiple causes of death. The lCD has Criminal Justice System, Public Health Sys­ been revised periodically to incorporate tem and social service agencies. changes in the medical field . To date, there Most attention to violence in Border com­ have been 10 revisions of the lCD. munities remains focused on drug-related, 3The Injury Severity Score (ISS) takes values from including narco-trafficking violence. In Bor­ 0 to 75. If an injury is assigned an AIS of 6 der communities, public health concerns (unsurvivable injury), the ISS score is auto­ about IPV and related family violence issues matically assigned to 75 . The ISS score is vir­ have emerged as important state and fed­ tually the only anatomical scoring system in use and correlates linearly with mortality, mor­ eral public health policy and programming bidity, hospital stay and other measures of se­ issues. To date, most programming has verity. It's weaknesses are that any error in been limited to cursory outreach and educa­ AIS scoring increases the ISS error, many dif­ tional campaigns. The seriousness of fam­ ferent injury patterns can yield the same ISS ily violence in Border communities has yet to score and injuries to different body regions be adequately recognized by either the fed­ are not weighted. Also, as a full description of eral or state governments as a major public patient injuries is not known prior to full inves­ health initiative. Social Services and public tigation & operation , the ISS (along with other health researchers have yet to conduct seri­ anatomical scoring systems) is not useful as a triage tool. (CDC 2003) ous continuing and systematic IPV research •Pollock, D. and P. McClain. 1989. Trauma regis­ as it concerns border communities and popu­ tries. Current status and future prospects. lations. JAMA 262 16: 2280-3. As of August 31 , 1996, Section of 157.129 of the state trauma registry ENDNOTES rule established Texas hospital standard data 'Pollock, D. and P. McClain . 1989. Trauma reg is­ set requirements, TR case inclusion , and what tries. Current status and future prospects. constituted major trauma. JAMA 262 16: 2280-3. Hospital trauma regis­ 5These hospitals are William Beaumont Army Medi­ tries are evolving rapidly as a result of a re­ cal Center (WBAMC), Providence, Sierra, Las newed focus on trauma care evaluation and Palmas, Cu lberson , Del Sol Medical Center, recent advances in microcomputer technology. Southwestern General and Thomason Hospi­ In theory, trauma registries can serve as the tal. principal tool for the systematic audit of the 6The Texas counties are Hudspeth, Culberson, quality of patient care provided by a hospital or Presidio, and El Paso. The New Mexico Coun­ a trauma system and as a potential source of ties are Hidalgo, Luna , Grant, Dona Ana, Sierra part of the data needed for injury surveillance. and Otero. In practice, however, there is a tendency to Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 101

REFERENCES Straus MA & RJ Gelles. 1990. Editors. Physical American Medical Association Medical News. Violence in American Families: Risk Factors 1992. Domestic Violence and Health Care: and Adaptations to Violence in 8, 145 Fami­ What Every Professional Needs to Know. Thou­ lies. New Brunswick, NJ : Transaction Books. sand Oaks, CA: Sage Publications. Taylor Institute. 1997. Trapped by Poverty Trapped American Psychological Association Presidential by Abuse. Poverty, Welfare and Battered Task Force on Violence and the Family. 1996. Women: What does the research tell us ? Violence and the Family. Washington, DC: Eleanor Lyon. American Psychological Association. Texas Department of Health. 2001 . Regional Advi­ Centers for Disease Control and Prevention (CDC). sory Council Handbook & Texas Department 2003. Costs of Intimate Partner Violence of Health Trauma Registry Available at: http:// Against Women in the United States. Atlanta, www. td h. state.tx . us/defa u II. htm. GA: CDC, National Center for Injury Prevention Tjaden P & N Thoennes. 2000. Extent, nature, and and Control. Available from URL: http://www. consequences of intimate partner violence: find­ cdc.gov/ncipc/pub-res/ipv_ cost/ipv. htm. ings from the National Violence Against Women Committee on Trauma American College of Sur­ Survey. Washington, DC: Department of Jus­ geons. 1999. Resources for Optimal Care of tice. Publication No. NCJ 181867. Available from the Injured Patient. URL: http://www.ojp.usdoj.gov/nij/pubs-sum/ FBI. 2001. Crime in the United States-2001 . Uni­ 181867.htm. form Crime Reports. Tolman RM & J Raphael. 2000. A review of re­ Felitti V, RAnda, D Nordenberg, D Williamson, A search on welfare and domestic violence. J Spitz, V Edwards, M Koss, & J Marks. 1998. Social Issues 56 4 655. Relationship of childhood abuse and house­ Trapped in Poverty/Trapped by Abuse: New Evi­ hold dysfunction to many of the leading causes dence Documenting the Relationship Between of death in adults. Amer J Preventive Medicine Domestic Violence and Welfare. 1997. J 14 4 245-58. Raphael & RT Tolman. Taylor Institute and the Gazmararian JA, R Petersen, AM Spitz, MM Good­ University of Michigan Research Development win, LE Saltzman, & JS Marks. 2000. Violence Center on Poverty, Risk and Mental Health. and reproductive health: current knowledge Available from URL: http://www.ssw.umich. and future research directions. Maternal & ed ultra ppedlp ubs_trapped . pdf Child Health J 4 2 79-84. US Department of Health and Human Services. Mala A , NA Rocha, L Blough, & L Lopez. 1999. 1998. Healthy People 2010. Washington, DC: Trauma registries as a potential source of bor­ Public Health Service. der epidemiology workgroup indictor data. Bor­ US Department of Justice. 2001 . Violence by Inti­ der Epidemiology Work Group J. mates: Analysis of Data on Crimes by Current National Institute of Justice and Centers for Dis­ or Former Spouses, Boyfriends, and Girl­ ease Control and Prevention. 1998. Prevalence, friends. March. Incidence, and Consequences of Violence US Department of Justice, Bureau of Justice Sta­ Against Women: Findings from the National tistics. 1994a. National Crime Victimization Violence Against Women Survey. November. Survey, Crime Data Brief Violence and Theft National Research Council. 1996. Understanding in the Workplace. NCJ-148199. Violence Against Women. Washington, DC: US Department of Justice Bureau of Justice Sta­ National Academy Press. tistics. 1994b. Sex Differences in Violent Vic­ Paulozzi LJ, LA Saltzman, MJ Thompson, & P Holm­ timization. Available from: URL: http:// green. 2001 . Surveillance for homicide among www. oj p. usdoj .g ov/bjs/abstract/sdvv. htm. intimate partners - United States, 1981- 1998. US Department of Justice, Bureau of Justice Sta­ CDC Surveillance Summaries 50 (SS-3) 1-16. tistics. 2000. Intimate Partner Violence. CM Pollock D. & P. McClain. 1989. Trauma registries. Rennison & S Welchans. Available from URL: current status and future prospects. JAMA 262 http :1 lwww. ojp. u sdoj .gov/bjs/pu b/pdflipv. pdf. 16: 2280-3. Wisner CL, TP Gilmer, LE Saltzman, & TM Zink. Rocha NA, AG Mala, AH Tyroch, S Mclean, & L 1999. Intimate partner violence against women: Blough. 2005. Trauma registries as a potential do victims cost health plans more? J Family source of border epidemiology work group in­ Practice 48 6 439-43. dicator data: trends from 1996-2000. Free lnq Great Social 33 2 143-1 51. Saltzman LE, JL Fanslow, PM McMahon, & GA Shelley. 1999. Intimate Partner Violence Sur­ veillance: Uniform Definitions and Recom­ mended Data Elements. Atlanta: National Cen­ ter for Injury Prevention and Control. 102 Volume 34 No. 1 May 2006 Free Inquiry In Creative Sociology

WORD OF MOUTH

Communication ·s power An 1dea, passed from person to person, and village to vii age, ca1 transform the worlo. I

Start w1th he n:rt 1dea

Linguistics and Language Behavi or Abstracts offen'1g • abstrac s of scno1ar l'r a1.1c 1es and books

• b !b l1 ograp 1C entries for subje:::t soec1 1c Glsserta: 10'IS and book and other med1a rew~vvs

LLBA's t1me ly ana compref-)ens ve coverage speaks vo Jmes on current 1deas 1n l,ngcJ·stlcs and language research.

Ava ,l able n pnm, or 'le or CD-RQf, l &, o ....., S lvecP ia:te' a 'Ia NIS() ana en magnetiC tape

Vis1t our Web s1te wvvv,_socabs org 'or product 1nformat1on, l1nks to re evant Sites, ana suoscr,ptlon-based access to the LLBA Speech Lang_;cge ar,o Hear,rg Patno logv subset ILBa Linguistics and Language Beharior Abstracts P0 Box 22206 Sar, D1eg o, CA 92192-0206 619 '695-8803 • Fax 619 605-04 16 • e'Tla I soc ,o@cerfnetc m Free Inquiry In Creative Sociology Volume 34 No. 1 May 2006 103

AUTHOR'S MANUSCRIPT SUBMISSION FORM

MAIL TO: EDITOR, FREE INQUIRY IN CREATIVE SOCIOLOGY OKLAHOMA STATE UNIVERSITY DEPARTMENT OF SOCIOLOGY 006 CLASSROOM BLDG STILLWATER, OK 74078-4062

Phone: 405-744-6126 Fax: 405-744-5780 Email: [email protected]

SUBSCRIBERS: Subscription on a current calendar year basis only, May and November. Fees of $30 (US resident) or $35 (US $) for individuals outside the U.S. Institutional rates are $40 in the U.S. and $45 (US $) outside the U.S. Send check or mony order only, made payable to Free Inquiry in Creative Sociology.

AUTHORS: Enclose 3 manuscript copies and I manuscript copy on a 3 1/2" computer disk, IBM com pat­ ible, saved in rich text/ASCII format.

MANUSCRIPT FEES: Authors who are current subscribers submit a $25 (US $) manuscript fee. Authors who are not current subscribers submit a $40 (US $) fee which includes a current subscription. Send check or money order onl y, made payable to Free Inquiry in Creative Sociology. MANUSCRIPT TITLE: ------

ADDRESS (Print clearly) Nrune: ______

University: ------

Department: ------

Building & Rm#/ StreetAddress/PO Box: ------

City, State, Zip+4: ------

Phone: ------F~ : ------Email: ------

STATEMENT OF MISSION FREE INQUIRY IN CREATIVE SOCIOLOGY: Sociologists convey discoveries, ideas & research fi ndings to students, colleagues and lay readers interested in social processes. Policy obj ectives: readability, creati vity, diversity; economize space and costs. READABILITY demands persuasive, interesting & lucid writing understandable to the literate lay person. It is lively enough to sustain interest & sufficiently supported with evidence & logic to convince the skeptical reader. CREATIVITY demands inventive, innovative & venturesome approaches to problems, propositions, ques­ tions, structures, data sets, theory, & social phenomena. Strange creatures are welcome if their outlines are clear. DI VERSITY call s for a persistent spread of materials in all areas of sociology, giving voice to all fields, positions & technologies. We encourage sociologists in micro-, meso-, & macro-sociology. C ross-cultural studies & manuscripts from foreign authors are most welcome. ECONOMY demands a rich product at least cost, with thin margins & small print.