Nutritional Epidemiology and Research Methods

© PhotoAlto/Laurence Mouton/Getty Images© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION CHAPTER 3 © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NutritionalNOT FOR SALE OR DISTRIBUTION EpidemiologyNOT FOR SALE OR DISTRIBUTION © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALEand OR DISTRIBUTION Research MethodsNOT FOR SALE OR DISTRIBUTION

CHAPTER OUTLINE© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION ■ Introduction ■ Epidemiology in Community Health ■ Nutritional Epidemiology ■ Interpretation of Cause and Eﬀect in Nutritional Epidemiology © Jones■ Types & of Bartlett Public Health Learning, Nutritional LLCEpidemiology Research © Jones & Bartlett Learning, LLC NOT■ FORDescriptive SALE Measures OR DISTRIBUTION of Community Health NOT FOR SALE OR DISTRIBUTION ■ The Epidemiological Methods ■ Quantitative and Qualitative Methods ■ Types of Sampling ■ Concepts of Collaborative Research © Jones & Bartlett■ Reporting Learning, Research LLC Results © Jones & Bartlett Learning, LLC ■ Epidemiological Approaches to Community Health Assessment NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

LEARNING OBJECTIVES

■ Define nutritional© Jones epidemiology. & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC ■ Discuss theNOT different FOR types SALE of epidemiological OR DISTRIBUTION research. NOT FOR SALE OR DISTRIBUTION ■ Discuss the difference between quantitative and qualitative research methods.

83 © Jones & Bartlett Learning, LLC. NOT FOR SALE OR DISTRIBUTION.

9781284348286_CH03_083_104.indd 83 08/06/17 7:36 PM 84 Chapter 3 Nutritional Epidemiology and Research Methods

6 © Jones▸ & Bartlett Learning, LLC © Joneshealth. & e Bartlett science ofLearning, epidemiology LLC is important to Introduction community and public health nutritionists because it can NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Community and public health nutrition research has be utilized to assess and understand health, diseases, and increased over the years, but nutritionists must continue injury in a community or target population. Epidemi- to increase the scientic knowledge base that is unique ology is a population-focused applied science that uses to their practice to provide high-quality, scientically research and a statistical data collection methodology oriented, evidence-based© servicesJones and& Bartlett solutions Learning,to to- to LLC determine the following7: © Jones & Bartlett Learning, LLC day’s nutrition-related healthNOT conditions.FOR SALE Community OR DISTRIBUTION ■ e population that is aectedNOT by a FORdisease SALE or disorder OR DISTRIBUTION nutrition research is dened as the “organized study ■ e incidence or prevalence of a health problem of a trend at both the basic and more applied levels that in the community focuses on social, structural, and physical environmental ■ e likelihood the causes and risk factors contrib- inequities© Jones through & active Bartlett involvement Learning, of community LLC uting to the© Jones health problem& Bartlett may Learning, or may not LLCbe members,NOT organizational FOR SALE representatives, OR DISTRIBUTION and researchers determinedNOT FOR SALE OR DISTRIBUTION in all aspects of the research process with the specic topics varying between investigators and the public.”1 One example of epidemiological research in the Learning to conduct successful research and epidemi- community is that of the Racial and Ethnic Approaches ological studies is crucial to community nutrition. to Community Health Across the United States (REACH © Jones & EpidemiologyBartlett Learning, is the study LLC of the determinants, © JonesU.S.) Program. & Bartlett e Centers Learning, for Disease LLC Control and NOT FORoccurrence, SALE OR and distributionDISTRIBUTION of health and disease in a NOTPrevention FOR SALE (CDC) OR created DISTRIBUTION REACH U.S. to address the dened population.2,3 It studies scientically the factors problem of health disparities among dierent racial that aect the health of individuals in the community. It and ethnic groups and to show that the health status also serves as the foundation for intervention programs of groups most aected by health inequities can be improved. REACH U.S. supports strategies that ad- that use evidence-based information to address the pre- 8 © Jones2-4 & Bartlett Learning,dress LLC health disparities throughout© Jones the life & span. Bartlett One Learning, LLC vention of health conditions.NOT FOR Today, SALE epidemiologists OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION employ a range of study designs, from observational to of the success stories of the REACH U.S. program was experimental, with the purpose of revealing unbiased conducted at the Genesee County (Michigan) Health relationships between exposures (e.g., nutrition, bio- Department. e department coordinated a multifac- logical agents, stress, or chemicals) and outcomes (e.g., eted community eort to reduce the high death rates diseases,© wellness, Jones and & Bartletthealth indicators). Learning, Epidemiological LLC among African© AmericanJones & infants Bartlett born Learning, in and around LLC studiesNOT are usually FOR categorized SALE OR as DISTRIBUTION follows3,5: Flint, Michigan.NOT e FOR program’s SALE activities OR DISTRIBUTION included the following8: ■ Descriptive (organizing data by time, place, and ■ person) Case-management services designed to reach ■ Analytic (aiming to examine associations or pregnant women and new mothers in high-risk areas of the county © Jones & commonlyBartlett Learning,hypothesized LLC causal relationships, and © Jones & Bartlett Learning, LLC ■ Community dialogue sessions designed to educate NOT FOR SALEincorporating OR DISTRIBUTION a case-control or cohort study) NOT FOR SALE OR DISTRIBUTION ■ Experimental (clinical or community trials of residents about infant death rates and available treatments and other interventions) resources ■ Workshops conducted to help participants understand is chapter discusses the basic research process the connections between racism and health care and epidemiological studies© Jones in dietetic & practice.Bartlett A Learning, more LLC © Jones & Bartlett Learning, LLC advanced research process is presented in Appendix F. Social marketing eorts were used to communicate NOT FOR SALE OR DISTRIBUTIONimportant health messages aboutNOT how FOR to reduce SALE infant OR DISTRIBUTION deaths, and a medical services committee was formed to identify and promote best clinical practices. ▸ Epidemiology in Community Results showed that the death rate for African Health© Jones & Bartlett Learning, LLC American infants© Jones in Genesee & Bartlett County Learning, dropped from LLC NOT FOR SALE OR DISTRIBUTION a high of 23.5NOT deaths FOR per SALE1,000 live OR births DISTRIBUTION in 1999 to Epidemiological and demographic measures and research a low of 15.2 in 2005. In comparison, the rate among methods are used to study health-related conditions. Caucasian American infants in 2005 was 6.3 deaths per For example, they are used to investigate outbreaks 1,000 live births. e disparity ratio between African of food poisoning, chickenpox, measles, and acquired Americans and Caucasian Americans also dropped © Jonesimmunodeciency & Bartlett Learning, syndrome LLC (AIDS). ey are also © Jonesfrom a high & Bartlett of 3.6 African Learning, American LLC infant deaths for NOT FORused SALE to investigate OR DISTRIBUTION environmental conditions, lifestyles, NOTevery FOR 1 Caucasian SALE OR American DISTRIBUTION infant death in 2001 to health-promotion strategies, and other factors that aect 2.4:1 in 2005.8

9781284348286_CH03_083_104.indd 84 08/06/17 7:36 PM Nutritional Epidemiology 85

© Jones & Bartlett▸ Learning, LLC © Jonesbecause & Bartlett they may Learning, consume hundreds,LLC even thousands Nutritional Epidemiology of dierent foods during that period. In addition, the NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 9 Nutritional epidemiology is the study of dietary intake diculty may be due to the following : and the occurrence of disease in human populations.9 ■ Other people preparing the meals (e.g., restaurant, In nutritional epidemiology, accurate quantication friends, spouse, prepackaged foods) so that the of nutritional exposure is critical. Exposure is the client does not know what, or how much, he or characteristics or© agents Jones (e.g., & food, Bartlett medications, Learning, time LLC © Jones & Bartlett Learning, LLC she has consumed in the sun) that aNOT person FOR comes SALE in contact OR DISTRIBUTION with that NOT FOR SALE OR DISTRIBUTION ■ may be related to disease risk (see TABLE 31).10 It is Seasonal variations in food intake ■ complicated and challenging to evaluate and associate Life events (e.g., weekends, holidays, vacations, dietary intake with disease risk. For instance, cigarette birthdays) smoking© Jones can & be Bartlett accurately Learning, assessed as anLLC activity with Indeed,© Jones the daily & Bartlettvariability Learning, in food intake LLC makes aNOT simple FOR yes or SALE no question. OR DISTRIBUTION It has been reported that it dicultNOT to FORidentify SALE any criticallyOR DISTRIBUTION consistent pat- cigarette smoking is addictive, and smokers seem to tern.13 is means that researchers must depend on be consistent in their habits, instead of stopping and food composition databases to calculate the exposure starting. Most smokers seem to smoke about the same variable. ese issues have made it dicult to obtain number of cigarettes per day, and, because of the ex- consistent and strong evidence about how diet aects © Jones & Bartlettpense, most Learning, people know LLC how many cigarettes they© Jones disease & Bartlett risk. Learning, LLC NOT FOR SALEsmoke OR per DISTRIBUTION day.11,12 NOT FOR SALEIn the past OR 20DISTRIBUTION years, nutritional epidemiological In comparison, it is dicult for clients to report research has focused on nutrient-based analyses to help accurately the food consumed for more than 1 week guide the search for underlying causes of disease risk.

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION TABLE 31 Examples of Exposures Relevant to Nutritional Epidemiological Studies

Exposure Diet-Related Example Other Example © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Agent that may cause or protect Fruit and vegetable consumption Physical activity may provide NOTfrom diseaseFOR SALE OR DISTRIBUTIONmay provide protection againstNOT FORprotection SALE againstOR DISTRIBUTION colon cancer, colon cancer, heart disease, and obesity, and heart disease. stroke.

© Jones & BartlettConstitutional Learning, host factors LLC Persons can© have Jones a genetic & Bartlett Learning,Older adults LLC and low-income predisposition to nutrition-related individuals are more predisposed NOT FOR SALE OR DISTRIBUTION diseases (e.g.,NOT diabetes, FOR heart SALE disease, OR DISTRIBUTIONto chronic disease. cancer).

Other host factors Food preferences learned during More educated adults may have © Jones & Bartlettchildhood Learning, can determine LLC food better ©disease Jones screening. & Bartlett Learning, LLC choices. NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Agents that may confound the Correlation between dietary Alcohol and substance abusers and association between another agent constituents (e.g., a diet high in smokers are less likely to engage in and disease fruits and vegetables is usually physical activity. © Jones & Bartlett Learning, LLClow in calories and fat and high© inJones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTIONantioxidants). NOT FOR SALE OR DISTRIBUTION Agents that may modify the eﬀects Fruits and vegetables may protect Smoking causes an increased risk for of other agents against lung cancer among smokers. lung cancer.

© Jones & BartlettAgents Learning,that may determine LLC the Malnutrition.© Jones & Bartlett Learning,Medical treatment LLC and medical NOT FOR SALEoutcome OR DISTRIBUTIONof disease NOT FOR SALE OR DISTRIBUTIONnutrition therapy. Reprinted from Nutrition in the Prevention and Treatment of Disease, Coulston A, Rock C, Monsen E. (Eds), Copyright Academic Press (2001), with permission from Elsevier.

9781284348286_CH03_083_104.indd 85 08/06/17 7:36 PM 86 Chapter 3 Nutritional Epidemiology and Research Methods

■ © JonesHowever, & Bartlett food-based Learning, hypotheses LLC and analyses have © Jones e result& Bartlett correlates Learning, with other LLC criteria of causality been shown to complement this philosophy and can (given in Appendix G). NOT FOR SALE OR DISTRIBUTION NOT FORbe an SALE eective OR method DISTRIBUTION to link food intake to a health 14 It is not uncommon in the eld of nutritional epide- condition. For example, Ness and Powles conducted a miology for a cause-and-eect relationship to be beyond systematic review of the reported associations between doubt. However, research information should not be the consumption of fruits and vegetables and the risk ignored or discarded.17 One must exercise caution and © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC for cardiovascular diseases. Results showed that 9 of use thoughtful consideration before acting on epidemi- 10 ecological studies, 2NOT of 3 case-control FOR SALE studies, OR DISTRIBUTIONand ological evidence. is chapter’sNOT Successful FOR Community SALE OR DISTRIBUTION 6 of 16 cohort studies found a signicant protective Strategies feature presents a successful cohort research association between the consumption of fruits and study that veried the association between plant-based vegetables and coronary heart disease. For stroke, 3 of diets and the rate of mortality. 5 ecological© Jones studies & and Bartlett 6 of 8 cohortLearning, studies LLC found a © Jones & Bartlett Learning, LLC signicant protective association with consumption of NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION fruits and vegetables. For total circulatory disease, one of two cohort studies reported a signicant protective ▸ Types of Public Health association.15 Hence, the tools and methods of nutritional Nutritional Epidemiology epidemiology can be developed to deal with scientic © Jonesissues & Bartlett unique to Learning, the biology LLC of chronic diseases. Spe- © JonesResearch & Bartlett Learning, LLC NOT FORcically, SALE epidemiological OR DISTRIBUTION methods were designed to NOT FOR SALE OR DISTRIBUTION address the following9: Public health nutritional epidemiology research and research design can be divided into observational ■ e length of time for disease development and experimental/clinical categories. e dierence ■ e multifactorial nature of chronic diseases between an experimental and an observational study is ■ e fact that research© using Jones human & Bartlett beings prevents Learning, the LLC control the researcher may© have Jones over participants,& Bartlett Learning, LLC direct observations NOTof cause FOR and SALEeect OR DISTRIBUTIONthe methods and procedures,NOT and the FOR exposures SALE (see OR DISTRIBUTION FIGURE 31). In an experimental/clinical study, the researchers can control some of the clients’ entire dietary ▸ Interpretation of Cause intake. In an observational study, the researcher does © Jones & Bartlett Learning, LLC not intervene© inJones or manipulate & Bartlett the Learning,clients’ dietary LLC and Eﬀect in Nutritional 9,18,19 NOT FOR SALE OR DISTRIBUTION intake. WeNOT will FOR rst discussSALE observationalOR DISTRIBUTION stud- Epidemiology ies. e types of observational studies for individuals include cross-sectional, case-control, and cohort/ Epidemiology is the study of associations, and statistical prospective studies; observational studies of groups methods are used to test the strength of these associations. are referred to as ecological studies. Aer that, we will © JonesHowever, & Bartlett it is importantLearning, to LLCnote that the existence of © Jonesdiscuss experimental& Bartlett studies.Learning, Figure LLC 3-1 presents types NOT FORa statistically SALE OR signicant DISTRIBUTION association does not indicate NOTof epidemiological FOR SALE OR research DISTRIBUTION design. that the observed relationship is one of cause and eect. For any observed association, the following should be considered16,9: Observational Studies of Individuals In a cross-sectional study, sometimes described as ■ e observed association© Jones may be& dueBartlett to chance. Learning, LLC © Jones & Bartlett Learning, LLC prevalence or descriptive survey, nutrient intake and ■ e association mayNOT be dueFOR to SALEimproper OR study DISTRIBUTION NOT FOR SALE OR DISTRIBUTION design or implementation or inappropriate anal- outcome are both measured at the same point in time. ysis of data. ■ e association may be weak or strong. ■ A credible biological mechanism may exist. Analytical Descriptive ■ ere© Jones could be & anBartlett association Learning, between LLCexposure © Jones & Bartlett Learning, LLC andNOT the outcome. FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Experimental Nonexperimental/ Survey ■ The significance of the results depends on the Randomized Observational context of all the available evidence. Causality clinical trial Cohort/prospective is supported when studies that are conducted Nonrandomized Case control clinical trial Cross-sectional © Jones & atBartlett different Learning, times, using LLC different methods, and © Jones & Bartlett Learning, Ecological LLC NOT FOR SALEperformed OR amongDISTRIBUTION different populations show NOT FOR SALE OR DISTRIBUTION similar results. FIGURE 31 Types of epidemiological research design.

9781284348286_CH03_083_104.indd 86 08/06/17 7:36 PM Types of Public Health Nutritional Epidemiology Research 87

© Jones & Bartlett e aim Learning,of the study LLCis to describe the relationship© Jones have & Bartlettnot been Learning,diagnosed with LLC the health condition? between diseases and dietary intake in a specied Or, do people diagnosed with hypertension smoke NOT FOR SALE OR DISTRIBUTION NOT FOR SALEcommunity OR DISTRIBUTION at a particular time. It can provide a snap- cigarettes and consume foods higher in sodium than shot of the frequency and characteristics of a disease persons without hypertension?18 For example, in 2004, in a population at a particular point in time.20 An Lubin et al.25 conducted a case-control study in Israel. example of a cross-sectional study question would be: e aim of the study was to determine the association What is the association© Jones between & Bartlett nutrient Learning, intake and LLCbetween body size and ©the Jones risk for & breast Bartlett cancer. Learning, e LLC disease outcome?NOT For example,FOR SALE the National OR DISTRIBUTION Health study used a population NOTof 1,065 FOR breast SALE cancer OR patients, DISTRIBUTION and Nutrition Examination Survey (NHANES) and 964 surgical controls, and 981 neighborhood controls. the International Study on Macronutrients and Blood Heights and weights at three periods throughout the Pressure (INTERMAP) are cross-sectional studies. participants’ lives (at age 18, for “most of adult life,” and INTERMAP© Jones & is Bartletta cross-sectional Learning, study LLCof associations most recent)© Jones were &ascertained. Bartlett Learning, e authors LLCanalyzed between macronutrients and blood pressure. In this these parameters and the body mass index (BMI) for NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION study, participants had their blood pressure measured each period, as well as BMI changes throughout life, twice on each of four occasions and completed a 24-hour controlling for age, menstrual status, and ethnic origin. recall on each day that their blood pressure was mea- Results showed an increase in risk for breast cancer with sured.18,11,21 Similarly, the National Center for Health greater recent BMI among postmenopausal women ages © Jones & BartlettStatistics Learning, conducts the LLC NHANES survey biannually© to Jones 60 & years Bartlett and older. Learning,25 LLC NOT FOR SALEmeasure OR the DISTRIBUTION health and nutritional condition of adultsNOT FOR SALEIn another OR retrospective DISTRIBUTION study, the Alaska Native and children living in the United States. e survey Medical Center diabetes program analyzed diabetes utilizes both questionnaire and examination techniques. care and outcomes audit data from 1994 to 2004 to In this study, a sample of 1,890 12- to 16-year-olds with evaluate the impact of the Special Diabetes Program measured height ©and Jones weight & were Bartlett obtained Learning, from the LLCfor Indians (SDPI) funding© Jones on process & Bartlett and interme Learning,- LLC NHANES III survey to identify factors to potentially diate outcomes.26 Congress established the SDPI grant NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION combat obesity in 12- to 16-year-olds with one or program in 1997 to respond to the diabetes epidemic two obese parents and those with no obese parents. among American Indians and Alaska Natives. e NHANES III surveys were conducted in two stages: 1) SDPI program provides funding for diabetes treatment an initial interview was conducted in the respondents’ and prevention services at 399 Indian Health Service homes,© Jones where & healthBartlett histories Learning, and sociodemographic LLC (IHS) and© Jones Tribal and & Bartlett Urban Indian Learning, Health ProgramsLLC characteristicsNOT FOR SALE were obtained,OR DISTRIBUTION and 2) 3 weeks later, in all 12NOT IHS FORareas acrossSALE the OR United DISTRIBUTION States. e IHS dietary intake assessments and physical examinations Division of Diabetes administers the program with were conducted on the participants in a mobile exam- guidance from the Tribal Leaders Diabetes Committee.27 ination center. e researchers conducted a retrospective analysis of Results show that eating breakfast every day or data28,29 from randomly selected standardized medical © Jones & Bartlettsome days Learning, was signicantly LLC protective against obesity© Jones records & Bartlett for trends Learning, from regional LLC sites in Alaska. Results NOT FOR SALEin adolescents OR DISTRIBUTION with obese parent(s), and this provedNOT to FORshow SALE that hemoglobin OR DISTRIBUTION A1c, total and low-density li- be the strongest protective factor in this group of chil- poprotein (LDL) cholesterol, triglycerides, and blood dren. Findings from the NHANES studies have been pressure signicantly improved from the pre-SDPI to used to create public health education and intervention the SDPI period. 22,15,23 programs. © Jones & Bartlett Learning, LLC In cohort studies, ©sometimes Jones &called Bartlett prospective, Learning, LLC In nutritionalNOT case-control FOR SALE studies OR, also DISTRIBUTION referred follow-up, and longitudinalNOT studies, FOR baseline SALE risk OR factors DISTRIBUTION to as retrospective and case-referent studies, persons re- are evaluated and participants are followed over time to cently diagnosed with a diet-related disease and a group monitor disease occurrence. It is the analytic method of persons without the disease (the control) from the of epidemiological studies. Participants are typically same population are interviewed concerning their dietary free of disease. A typical study question would be: Do habits.© Jones e dierences& Bartlett between Learning, both groups’LLC ndings people ©with Jones a higher & intakeBartlett of a nutrientLearning, develop LLC or die areNOT compared. FOR SALE Case-control OR DISTRIBUTION studies are a commonly from aNOT particular FOR disease SALE more OR frequentlyDISTRIBUTION or sooner used method of observational epidemiological studies. than persons with a lower intake? Cohort studies are e aim of this type of study is to identify the cause of a usually very large, exceeding 50,000 participants; may disease among a group of people, or the cause-and-eect take many years to be conducted; and are expensive.28 24 © Jones & Bartlettrelationships Learning, of the health LLC condition. Typical study© Jones An & example Bartlett of Learning,a cohort study LLC is the Women’s Health NOT FOR SALEquestions OR couldDISTRIBUTION be: Do people with the disease or healthNOT FORInitiative SALE (WHI), OR aDISTRIBUTION 15-year study investigating the degree condition have a dierent lifestyle than persons who to which diet, hormone replacement therapy, calcium,

9781284348286_CH03_083_104.indd 87 08/06/17 7:36 PM 88 Chapter 3 Nutritional Epidemiology and Research Methods

© Jones TABLE& Bartlett 32 Dietary Learning, Assessment LLC in Diﬀerent © JonesExperimental & Bartlett Studies Learning, LLC NOT FORStudy SALE Sessions OR DISTRIBUTION NOTExperimental FOR SALE studies OR involve DISTRIBUTION intentional alteration of or intervention in the course of a disease. Nutritional Study Situation Methods Commonly Used intervention studies include metabolic studies and randomized clinical trials.9,18 In metabolic or feeding Cross-sectional© 24-Hour Jones recall; & foodBartlett frequency Learning,studies, LLC the researcher develops© nutritionallyJones & Bartlett adequate Learning, LLC NOTquestionnaire FOR (FFQ);SALE brief OR DISTRIBUTIONdiets that vary on one or more components,NOT FOR to SALE investigate OR DISTRIBUTION methods the eect of this diet on internal physiological reactions (e.g., carbohydrate, lipid, and energy utilization and Case-control FFQ; diet history storage in relation to obesity and diabetes), metabolic (retrospective) processes (e.g., absorption, transport, and utilization of © Jones & Bartlett Learning, LLC minerals, vitamins,© Jones and protein),& Bartlett and specicLearning, enzyme LLC CohortNOT FOR SALEFFQ; dietOR history; DISTRIBUTION 24-hour systems in experimentalNOT FOR animalsSALE ORand DISTRIBUTIONhumans.14 An recall example of an experimental study is a metabolic ward study that reported changes in total cholesterol from Intervention FFQ; brief methods; 24-hour changes in intakes of saturated and polyunsaturated recall 18,35-37 © Jones & Bartlett Learning, LLC © Jonesfatty acids & Bartlettand dietary Learning, cholesterol. LLC Crossover research design is common in metabolic/feeding studies. NOT FORClinical SALE screening OR DISTRIBUTION 24-Hour recall; brief methods; NOT FOR SALE OR DISTRIBUTION diet history In this design, each volunteer participant serves as his or her own control. Surveillance 24-Hour recall; brief methods Another type of experimental study is the randomized clinical trial, which is a prospective study in which Reprinted from Nutrition in the Prevention© and Jones Treatment of &Disease Bartlett, Coulston A, Rock Learning, C, participants LLC are randomly assigned© Jones to intervention & Bartlett and Learning, LLC Monsen E. (Eds), Copyright Academic PressNOT (2001), FOR with permission SALE from Elsevier. OR DISTRIBUTIONcontrol groups. e participantsNOT are followed FOR SALE over time OR DISTRIBUTION to assess the eectiveness and safety of the intervention aer randomization.34 An example of a randomized study is the Fracture Intervention Trial, which was designed to and vitamin D might prevent heart disease, breast and © Jones & Bartlett Learning, LLC test the hypothesis© Jones that hormone & Bartlett replacement Learning, therapy LLC colorectal cancer, and osteoporosis in postmenopausal would reduce the rate of fracture in postmenopausal NOT FOR SALE OR DISTRIBUTION 38 NOT FOR SALE OR DISTRIBUTION women. e main advantage of cohort studies is that women. exposure to potential risk factors is assessed before the development of the disease. TABLE 32 shows examples Genetic Epidemiological Studies of cohort studies. Genetic epidemiological studies use genetic epidemi- © Jones & TABLEBartlett 33 presents Learning, examples LLC of cohort studies uti- © Jones & Bartlett Learning, LLC lizing nutrition assessments. ology and molecular epidemiology. e investigation NOT FOR SALE OR DISTRIBUTION NOTcould FOR be a SALEcross-sectional OR DISTRIBUTION and/or an intervention study. In genetic epidemiology, a question could be: Do genes Observational Studies of Groups determine eating behaviors, such as how much one eats, Ecological studies compare collective data that repre- preferences for certain types of foods, and frequency or 18,39 sent entire populations. ©An Jones ecological & studyBartlett focuses Learning, on patterns LLC of eating? Research© impliesJones that & Bartlettthere is a Learning, LLC the comparison of groups,NOT not individuals.FOR SALE e OR purpose DISTRIBUTION resemblance among family membersNOT for FOR these SALE behaviors, OR DISTRIBUTION of an ecological analysis may be to make biological im- but it is not clear whether genes, shared environments, plications about the association between exposure (e.g., or both are the cause.40,41 In molecular epidemiology, water contaminant) and disease outcome in various questions relate to the physical and hormonal mecha- communities within a population or to make ecological nisms leading to taste preferences, hunger, and satiety. inferences© Jones about eects & Bartlett on group rates.Learning,4 Ecological LLC studies For example,© taste Jones receptors & Bartlett are clearly Learning, genetically LLC may beNOT very usefulFOR SALEfor monitoring OR DISTRIBUTION national trends in determined, althoughNOT FOR the gene(s)SALE mayOR notDISTRIBUTION all be iden- health, social, cultural, economic, and environmental tied yet.35,42 For example, genes encoding hormones factors that inuence health that cannot be measured and proteins that regulate hunger and satiety have just at an individual level. For example, migration studies, been identied recently.36,43,44 It has been known for © Jonesa type& Bartlett of ecological Learning, study, show LLC a signicant increase © Jonessome time & Bartlett that certain Learning, enzyme deciencies,LLC which NOT FORin the SALE risk forOR several DISTRIBUTION chronic diseases, such as breast NOThave FOR a genetic SALE cause, OR giveDISTRIBUTION rise to specic nutritional cancer, when people move from an Eastern to a Western problems; examples of these are phenylketonuria and country.24,25,34 galactosemia.36,43

9781284348286_CH03_083_104.indd 88 08/06/17 7:36 PM NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning, LLC NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC 9781284348286_CH03_083_104.indd 89 NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC

30-33 NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC TABLE 33 Examples of Cohort Studies Utilizing Nutrition Assessments NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC

Years Nutritional Intake Cohort Study Conducted Sample Studied Purpose Assessment Results

Early risk factors for increased Baseline: 447 African The aim of this study was Baseline: Three variables measured at birth were NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC adiposity: A cohort study of BetweenNOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC American to identify risk factors, Anthropometric independently associated with adiposity African American subjects 1959 and pregnant women present at birth, for measurements and in young adulthood.

© Jones & Bartlett Learning, LLC. NOT FORSALEORDISTRIBUTION. © Jones&Bartlett Learning, followed from birth to young 1965 increased adiposity in socioeconomic factors adulthood adulthood in an African American population

Framingham Study: Heart and 1947–1972 1,423 women in To investigate the Ultrasound at 12-year Women who ate a heart-healthy diet Vascular Disease Program the population- relationship between follow-up at baseline had less cardiovascular Dietary patterns, smoking, based a heart-healthy dietary Food Frequency disease risk factor proﬁles. Women and subclinical heart disease Framingham pattern and subclinical Questionnaire and who had heart-healthy eating pattern, in women: opportunities for Oﬀspring/Spouse heart disease in women, 24-hour recall plus avoidance of smoking, had lower primary prevention from the group and to identify potential chances of subclinical heart disease. NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning, LLC NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC Framingham nutrition studies opportunities for primary prevention Types ofPublic HealthNutritionalEpidemiologyResearch

Family Structure and 1998–1999 Kindergarten This study exam ines Face-to-face Children living with single mothers Childhood Obesity, Early (n = 14,493), the eﬀect of number assessments or had more tendency to obesity by ﬁfth

Childhood LongitudinalNOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC third grade of parents and number interviews, telephone grade than were children living with two NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC Study—Kindergarten Cohort (n = 11,855), of siblings on children’s interviews, and parents. Children with siblings had lower and ﬁfth grade body mass index and risk questionnaires, height BMI and were less likely to be obese than (n = 10,036) of obesity and weight, body children without siblings. Also, living mass index (BMI) with a single mother or no sibling was associated with increases in BMI from kindergarten through ﬁfth grade. NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC NOT FORSALEORDISTRIBUTION © Jones&BartlettLearning,LLC

Nurses’ Health Study: Waist 1986 47,382 U.S. The associations of waist Waist circumference, Higher waist circumference is associated circumference, waist-to- through registered nurses circumference and waist- hip circumference, with the risk for breast cancer, especially hip ratio, and risk for breast May 1994 to-hip circumference and waist-to-hip ratio. among postmenopausal women. cancer in the Nurses’ Health ratio with risk for breast Questionnaire Study. cancer

Sources: Chen AY, José JE. Family structure and childhood obesity: early childhood longitudinal study—kindergarten cohort. Prev Chronic Dis. 2010;7(3):A50-A60. Huang Z, Willett WC, Colditz GA, Hunter DJ, et al. Waist circumference, waist:hip ratio, and risk of breast cancer in the Nurses’ Health Study. Am J Epidemiol. 1999;150(12):1316-1324. 89

08/06/17 7:36 PM Millen BE, Quatromoni PA, Nam BH, O’Horo CE, et al. Dietary patterns, smoking, and subclinical heart disease in women: opportunities for primary prevention from the Framingham Nutrition Studies. J Am Diet Assoc. 2004;104(2):208-214. Stettler N, Tershakovec T, Zemel B, Leonard MB, et al. Early risk factors for increased adiposity: A cohort study of African American subjects followed from birth to young adulthood. Am J Clin Nutr. 2000;72:378-383. 90 Chapter 3 Nutritional Epidemiology and Research Methods

may be calculated in a “one-shot” cross-sectional (“slice © Jones▸ & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC 6 Descriptive Measures of of time”) or retrospective (backward-looking) study. NOT FOR SALECommunity OR DISTRIBUTION Health NOT FOR SALE OR DISTRIBUTION Interpretation of Incidence and Prevalence. In Descriptive epidemiological studies explain the oc- both incidence and prevalence, it is important to clearly currence of disease and are collected using various dene the condition being studied; otherwise, scientists methods. e data are then© Jones assembled & Bartlett based on Learning, time may LLC make a mistake in dierentiating© Jones between & Bartlett inci- Learning, LLC (e.g., changes of eating habitsNOT over FOR a long SALE period), OR placeDISTRIBUTION dence (new cases) and prevalence.NOT49,50 FOR Scientists SALE must OR DISTRIBUTION (geographic areas), and person (e.g., age, gender, eth- 5,45 allow enough time to identify new cases to determine nicity, lifestyle). ese data can be used to quantify incidence and prevalence. is period must be long the extent and location of nutrition problems within a enough to adequately detect the condition, especially population and to suggest associations between diet and in the case of rare diseases or diseases with a low rate of disease© that Jones can be & evaluated Bartlett in Learning,analytic research. LLC39,46 © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION diagnosis. However,NOT FORthe period SALE cannot OR be DISTRIBUTION so long that it will be aected by mortality and other follow-up losses.51 Demographic Measures Measures of incidence and prevalence provide dierent information and have dierent implications. Demography is an analytic tool used to measure a For example, an increase in the prevalence of diabetes population by recording births, deaths, age distribution, © Jones & Bartlett Learning, LLC © Jonesor endometrial & Bartlett cancer Learning, means that thereLLC are more per- and other vital statistics.6,7 Some human characteristics, NOT FOR SALE OR DISTRIBUTION NOTsons FOR with SALEthese health OR DISTRIBUTION conditions in the population. or demographics, may be associated with wellness or is may be due to more new cases (in other words, illness. Age, race, gender, ethnicity, income, and educa- increased incidence) or because individuals with these tional level are important demographics that may aect health conditions are living longer. In either case, the health outcomes. For example, men are more likely than community may need to direct resources toward treat- women to develop certain© Jonesheart diseases, & Bartlett and African Learning, LLC © Jones & Bartlett Learning, LLC ing the problem. However, if knowledge of incidence is Americans are more likelyNOT than FOR Caucasian SALE Americans OR DISTRIBUTION to NOT FOR SALE OR DISTRIBUTION lacking, it will be dicult to decide whether to target experience hypertension.23,29,47 Community nutritionists the resources toward primary prevention or toward must be familiar with the demographic characteristics of secondary and tertiary treatment services. the community they serve and with the health problems Here is a specic example of dierentiating incidence associated with that community before developing a © Jones & Bartlett Learning, LLC from prevalence.© InJones a community & Bartlett health center,Learning, 120 infants LLC health promotion program. NOT FOR SALE OR DISTRIBUTION free from measlesNOT are FOR followed SALE for 3 weeks.OR DISTRIBUTION In this period of time, 45 developed measles. Remember, incidence Morbidity and Mortality gives an approximation of the likelihood (or risk) that a Epidemiology studies both wellness and illness. Wellness client will develop a particular health condition during a is hard to measure; therefore, the measures of health are specied period. erefore, there is an incidence rate of © Jones & Bartlett Learning, LLC © Jones45 per 120 & (orBartlett 37.5 percent) Learning, during LLC the 3-week period. NOT FORexpressed SALE in OR terms DISTRIBUTION of morbidity (illness) and mortality NOT FOR SALE OR DISTRIBUTION (death).6 e CDC’s website (http://www.cdc.gov) is However, prevalence computes the proportion of a given an excellent source of morbidity and mortality data population with the problem. So at the same community organized by states and for select cities.48 health center, 150 infants were screened for the presence of measles and 30 infants were aected. is results in a Incidence and Prevalence© Jones & Bartlett Learning,prevalence LLC of measles of 20 percent© Jones (30 ÷ &150). Bartlett In this Learning, LLC NOT FOR SALE OR DISTRIBUTIONcase, prevalence means the proportionNOT FOR of infants SALE with OR DISTRIBUTION Two types of disease frequency measures are used to the problem (e.g., measles) at the specied period.52 e determine the morbidity rate in a dened population— formulas for calculating both are in BOX 31. incidence and prevalence. Incidence is the rate of acquisition of a new health condition or the number of persons© Jones in a dened & Bartlett population Learning, who developed LLC the Rates © Jones & Bartlett Learning, LLC conditionNOT during FOR a SALEspecied OR period. DISTRIBUTION6 e calculation e mathematicalNOT measures FOR SALE used to OR articulate DISTRIBUTION incidence of incidence, therefore, requires that a population and prevalence are known as rates. Epidemiological be followed over a period in a prospective or cohort studies must relate the occurrence of a health condition (forward-looking) study.38,45 to the population base. Rates express a mathematical © Jones & OnBartlett the other Learning, hand, prevalence LLC is the total number of © Jonesrelationship & Bartlett in which Learning,the numerator LLC is the number of NOT FORpersons SALE in the OR dened DISTRIBUTION population who are a ected by a NOTpersons FOR experiencing SALE OR the DISTRIBUTION condition, and the denominator certain disease or condition at a specied time. is includes is the population at risk, or the total number of persons both new and existing cases.39,38 erefore, prevalence who have the possibility of experiencing the condition.

9781284348286_CH03_083_104.indd 90 08/06/17 7:36 PM The Epidemiological Methods 91

Interpretation of Rates. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning,International LLC comparisons BOX 31 The Formulas for Both Prevalence and rankings of infant mortality should be interpreted NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION and Incidence with caution. For example, a study investigated how infant mortality rates are reported in the United States, Incidence rate (IR) = Number of new cases (people Norway, and Israel. Results showed that the United who developed the disease) States reported live births less than 750 g, whereas © Jonesduring a& specified Bartlett period Learning, LLCother countries did not ©report Jones live births& Bartlett less than Learning, 750 LLC IR NOT FOR SALE OR DISTRIBUTIONg. Disparities among countriesNOT resultedFOR SALE from dierences OR DISTRIBUTION Total number of a defined population at risk at the in birth and death registration practices. 48 time of getting the disease Also, to determine if the population in a specic community is at greater or lesser risk for a condition, Prevalence = Number of both existing and new cases © Jones & Bartlett Learning, LLC the research© Jones must compare & Bartlett the rates Learning, for the community LLC during a specified period with rates from similar communities, the state, or the NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION P country. Most oen, rates are based on data from a Total number of a defined population at risk calendar year, which may create some problems. When calculating an infant mortality rate, it is important to recognize that some of the infants who die during a given © Jones & Bartlett Learning, LLC © Jonescalendar & Bartlett year may Learning, not be a partLLC of the demographics, NOT FOR SALE e formula OR DISTRIBUTION for calculating rate is shown in the followNOT- FORsuch SALE as those OR who DISTRIBUTION died in 2005 but who were actually 6 ing section. Epidemiologists use rates to examine the born in 2004 and thus were not part of the 2005 popu- experiences of particular groups of people at specied lation at risk. Also, some of the infants who were born times, in dierent cities or countries. Another important in 2005 might die in 2006 and not be reected in the use of rates is in calculating the risks to individuals and 6 © Jones & Bartlett Learning, LLC2005 infant mortality rate.© Jones A cohort & Bartlett research Learning,study LLC groups of experiencing an event such as a heart attack (discussed earlier) can help overcome the limitations NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION or an occurrence such as cancer or obesity. of the conventionally calculated calendar year rate.48

Example of a Rate. In the United States, an acceptable denition for infant mortality rate is the number of deaths© Jones during & theBartlett rst year Learning, of life (i.e., LLC the number of Think© Jones About & Bartlett It Learning, LLC bothNOT neonatal FOR SALE and postneonatal OR DISTRIBUTION deaths) per 1,000 live NOT FOR SALE OR DISTRIBUTION births. To calculate infant mortality rate, add the number What type of research methodology did Peter use for of neonatal (the rst 28 days of life) and postneonatal his case-control study? (between the 28th and 365th days of life) deaths and Peter, a nutrition professor, received funding from divide by the number of live births; then, multiply the the Special Diabetes Program for Indians to conduct © Jones & Bartlettresult by Learning,1,000. e numerator LLC represents the number© Jones further& Bartlett analysis Learning, on the data he LLC collected from NOT FOR SALEof infants OR experiencingDISTRIBUTION the “condition” of dying inNOT the FORdiabetes SALE care OR and DISTRIBUTION nutrition education program. The rst year of life, and the denominator represents the information he collected included the participants’ population of infants at risk of dying in that year.3,53 dietary preferences, knowledge, and awareness of diabetes care, attitude about nutrition education, and If 500,000 live births and 10,000 infant deaths were management of diabetes. reported in the United States for a given year, the rate © Jones6 & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC can be calculated NOTas follows FOR: SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 1. Infant mortality rate = Number of infant deaths ▸ The Epidemiological Methods (age < 1 year during time interval) × 1,000 Epidemiological research methods can be descriptive, Total live births during time arrival © Jones & Bartlett Learning, LLC analytic,© orJones experimental. & Bartlett Although Learning, all can be LLC used in NOT2. FOR Divide SALE 10,000 OR byDISTRIBUTION 500,000 = 0.020 of the studyingNOT the FORoccurrence SALE of ORdisease, DISTRIBUTION the method used infants died during the rst year of life. most oen is descriptive epidemiology. Once the basic 3. Because it is dicult to relate to 0.020 infants, epidemiology of disease has been described, analytic multiply 0.020 by a constant, in this case 1,000, methods can be used to study the disease further, and © Jones & Bartlett Learning,resulting in LLC20 infants who died during ©the Jones an & experimental Bartlett Learning, approach can LLC be developed to test a 3 NOT FOR SALE OR rstDISTRIBUTION year of life per 1,000 live births. isNOT FORhypothesis. SALE OR DISTRIBUTION means that the infant mortality rate was In epidemiological studies, the investigator at- 20 infant deaths per 1,000 live births for that year. tempts to identify risk factors for particular diseases,

9781284348286_CH03_083_104.indd 91 08/06/17 7:36 PM 92 Chapter 3 Nutritional Epidemiology and Research Methods

© Jonesconditions, & Bartlett behaviors, Learning, or risks LLC that result from particular © Joneshypertension, & Bartlett 5 studies Learning, were prospective LLC and 48 were causes, such as environmental or industrial agents.50,54 cross-sectional, of which 4 also contained a longitudinal NOT FOR SALE OR DISTRIBUTION NOT FOREpidemiology SALE OR uses DISTRIBUTION dierent methods—such as statistical, component. Various dietary methodologies were used; pathological, clinical, demographic, microbiological, and the most common was the 24-hour dietary recall. Most sociological—to study disease processes. ese are not studies controlled for age, sex, and BMI, with variable exclusive to epidemiology; however, the ways in which control for other confounding variables.56 Together, these they are used distinctively© Jones dene the & Bartlettepidemiological Learning, studies LLC showed only modest associations© Jones between & Bartlett calcium Learning, LLC method.39 e followingNOT examples FOR use SALE the investigation OR DISTRIBUTION and blood pressure. But in theNOT beginning, FOR SALEthe higher OR DISTRIBUTION of the diet–disease hypothesis to show the scientic calcium and magnesium content of “hard” water, and its approach of epidemiologists. relation to cardiovascular mortality, triggered epidemiological inquiry into the relationship of both minerals 57 Observation© Jones & Bartlett Learning, LLC to blood pressure.© Jones & Bartlett Learning, LLC ere areNOT many FOR instances SALE of OR disease DISTRIBUTION control based on NOT FOR SALE OR DISTRIBUTION epidemiological observations. In an observational Making Comparisons study (descriptive method), the investigators observe Another scientic approach epidemiologists use is the behaviors of participants without interfering. For making comparisons. e Dietary Approaches to © Jonesexample, & Bartlett epidemiological Learning, researchers LLC observed consis- © JonesStop Hypertension & Bartlett (DASH) Learning, sodium LLC multicenter trial, NOT FORtently SALE that vegetarianOR DISTRIBUTION groups in the United States and NOTfor FORexample, SALE was ORcreated DISTRIBUTION to resolve the controversy overseas have lower blood pressure than nonvegetarians surrounding sodium intake and blood pressure.58 e in most studies. e term vegetarian includes several eect of dietary composition on blood pressure is an heterogeneous groups, but in general, vegetarian diets important public health topic. e National Heart Lung are usually high in whole grains, beans, fruits, and © Jones & Bartlett Learning,and LLC Blood Institute funded the© JonesDASH study,& Bartlett which Learning, LLC vegetables, and sometimes sh, dairy products, and compared the eects of three sodium levels on blood NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION eggs. e aspects of the vegetarian diet that have been pressure versus a control. e subjects were asked to observed to be protective against hypertension include eat either a control diet typical of intake in the United low intake of animal products and high intake of po- States or the DASH diet, a diet rich in fruits, vegetables, 52 tassium, magnesium, ber, and (sometimes) calcium. and low-fat dairy foods; low in saturated fat, total fat, © Jones & Bartlett Learning, LLC and cholesterol;© Jones high in &dietary Bartlett ber, Learning,potassium, calLLC- CountingNOT Cases FOR orSALE Events OR DISTRIBUTION cium, and magnesium;NOT FOR and SALE moderately OR highDISTRIBUTION in protein. Within each diet (control and DASH), subjects ate foods Counting cases is a descriptive epidemiological method with high, intermediate, and low levels of sodium. e sometimes useful for health policy. It can be used to know investigators found that reducing the sodium intake the number of people with a particular characteristic from the high to the intermediate level reduced systolic © Jonesor & number Bartlett of eventsLearning, that occurred LLC in a given region © Jonesblood pressure & Bartlett by 2.1 mmLearning, Hg.58,59 Reducing LLC the sodium NOT FORor community.SALE OR For DISTRIBUTION example, the migration studies of NOTintake FOR from SALE the intermediate OR DISTRIBUTION to the low level caused indigenous populations showed that the prevalence of additional reductions of 4.6 mm Hg on the control hypertension increased because of urbanization. With diet and 1.7 mm Hg on the DASH diet. e eects of urbanization, people consumed more processed foods sodium intake were observed in those subjects with and less fresh foods (that were previously available). In © Jones & Bartlett Learning,and LLC without hypertension, all ©races, Jones and both& Bartlett women Learning, LLC addition, researchers observed lifestyle changes (e.g., lack NOT FOR SALE OR DISTRIBUTIONand men. Compared to the controlNOT diet FOR with SALEthe highest OR DISTRIBUTION of physical activity, overeating) that caused increases in sodium level, the DASH diet with the lowest sodium body weight, sodium intake, and dietary fat during the 55 level reduced systolic blood pressure by 8.9 mm Hg in process of acculturation. subjects with hypertension and 7.1 mm Hg in those without hypertension.56 Relating© Jones Cases & or Bartlett Events toLearning, Population LLC In another© comparativeJones & Bartlett study, the Learning, Caucasian Pro LLC- NOT FOR SALE OR DISTRIBUTION spective InvestigationNOT FOR into CancerSALE and OR Nutrition DISTRIBUTION (EPIC) at Risk study recruited 20,343 participants for a cohort study. Inherent in the epidemiological method is the need for e participants had no hypertension. e investigators measuring the number of disease cases or events in a pop- examined whether the Mediterranean diet (see the Suc- © Jonesulation & Bartlett and relating Learning, the number LLC of cases to a population © Jonescessful Community & Bartlett Strategies Learning, feature LLC in this chapter), NOT FORbase. SALE Cases ORthat tDISTRIBUTION the case description are identied, NOTand FOR olive oilSALE in particular, OR DISTRIBUTION can reduce blood pressure. counted, and correlated to time, place, and individuals. Results showed that intakes of olive oil, vegetables, and In reviewing 53 observational studies of calcium and fruit were associated with a reduction in blood pressure

9781284348286_CH03_083_104.indd 92 08/06/17 7:36 PM The Epidemiological Methods 93

© Jones & Bartlettwhereas intakesLearning, of cereals, LLC meat and meat products, ©and Jones those & Bartlett with higher Learning, baseline blood LLC pressures, but diet and alcohol were associated with higher blood pressure.60 other variables were not assessed in a standard manner NOT FOR SALE OR DISTRIBUTION NOT FOR SALEIn OR addition, DISTRIBUTION another study conducted a 1-year across studies.9 randomized controlled trial to determine the degree to which dietary change is inuenced by providing seven Drawing Scientiﬁc Inferences home-delivered therapeutic meals weekly to adults age e data showing that diet modication can prevent and 60 years or older with© Jones hyperlipidemia & Bartlett or hypertension. Learning, LLC © Jones & Bartlett Learning, LLC lower blood pressure are signicant. e intervention Fiy percent of participantsNOT FOR received SALE seven OR therapeuticDISTRIBUTION NOT FOR SALE OR DISTRIBUTION approaches and methods involved in some data col- meals per week for 12 months. e nutrients that make lection are still being claried. Because of a variety of up the DASH diet were measured using 24-hour food research design limitations, inadequate statistical power, recalls at baseline, 6 months, and 12 months. Results and measurement issues, studies of single nutrients showed that delivery of seven DASH meals per week © Jones & Bartlett Learning, LLC (except ©for Jones potassium) & Bartlett have mainly Learning, been inconsistent. LLC 65 increased compliance with dietary recommendations NOT FOR SALE OR DISTRIBUTION However,NOT when FOR several SALE nutrients, OR DISTRIBUTION lifestyle, or dietary among noncompliant older adults with cardiovascular factors are combined in the same intervention studies, disease.61 as in the DASH study, blood pressure was eectively reduced.66 It has been shown that nutrients have in- Developing the Hypothesis © Jones & Bartlett Learning, LLC © Jonesteractive & Bartlett eects Learning, when they are LLC consumed together in a e hypothesis or research question is an assumption diet.62 In addition to the DASH dietary pattern, other NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION written in a clear, concise manner about what the inves- factors such as reduced sodium intake, weight loss, and tigator thinks will happen in the research project. For moderate alcohol intake have been shown to reduce example, aer reviewing the results of the migration blood pressure. Simultaneous observance of several studies and the vegetarian group’s studies in the United recommendations (as with the DASH diet) is likely to States, researchers© developed Jones & the Bartlett hypothesis Learning, that diets LLCbe the best strategy for preventing© Jones and& Bartlett lowering bloodLearning, LLC low in sodium andNOT high FOR in fruits, SALE vegetables, OR DISTRIBUTION and ber pressure. Future researchNOT should FOR center SALE on methods OR DISTRIBUTION to can reduce blood pressure. motivate and maintain dietary changes for controlling blood pressure, plus address the unanswered nutritional Testing the Hypothesis hypotheses. e implication of these ndings is that the suc- In© testingJones the & hypothesis,Bartlett Learning, research observations LLC have © Jones & Bartlett Learning, LLC cess of dietary intervention depends on support from shown a direct relationship between higher sodium NOT FOR SALE OR DISTRIBUTION clinicians,NOT government FOR SALE agencies, OR privateDISTRIBUTION institutes, and intake and higher blood pressure across population industries at both the population and individual levels. groups.9,13 e INTERSALT study (International Study In particular, industries can improve the nutritional of Salt) measured the relationship between 24-hour value of the food supply, such as reducing the sodium urinary sodium excretion and blood pressure in 10,079 and fat content of processed foods; this is a very critical © Jones & Bartlettmen and Learning,women from LLC 52 centers around the world.© 13Jones & Bartlett Learning, LLC role in implementing dietary changes. Dietetic and other NOT FOR SALEResults OR showed DISTRIBUTION a positive relationship between meanNOT FOR SALE OR DISTRIBUTION healthcare professionals must also play a signicant role urinary sodium excretion and blood pressure; that is, the in educating the public and promoting adherence to participants’ blood pressure increased with an increase nutritional guidelines for the prevention of hypertension in sodium excretion. A slight positive relationship also and other health conditions.16,67 was observed when© theJones researchers & Bartlett adjusted Learning, for alcohol LLC © Jones & Bartlett Learning, LLC and BMI. e researchersNOT FOR also reported SALE aOR strong DISTRIBUTION positive NOT FOR SALE OR DISTRIBUTION relationship between sodium intake and the slope of Conducting Experimental Studies blood pressure increase with age across populations, Another scientic approach that epidemiologists use indicating a role for sodium in age-related blood pres- is conducting experimental studies. One such study sure increase. examined the eects of garden-based nutrition edu- © JonesIn a recent & Bartlett reanalysis Learning, of the original LLC INTERSALT cation on© Jonesadolescent & fruitBartlett and vegetable Learning, consumption. LLC data,NOT corrected FOR SALE for measurement OR DISTRIBUTION error due to the use Sixth-gradeNOT studentsFOR SALE at three OR dierent DISTRIBUTION elementary of single 24-hour urine collections, results were stron- schools were used as a control group and two treatment ger: a 100 mmol/day (2,300 mg sodium) increase in groups. Students in the treatment groups participated urinary sodium was associated with an increase of 3 to in a 12-week nutrition education program, and one © Jones & Bartlett6 mm Hg Learning, in systolic blood LLC pressure and of 0 to 3 mm© Jones treatment & Bartlett group Learning, also participated LLC in garden-based 62,63 NOT FOR SALEHg diastolic OR DISTRIBUTION blood pressure. In a meta-analysisNOT of FORactivities. SALE Students OR DISTRIBUTION in all three groups were required observational studies, Law et al.64 reported stronger to complete three 24-hour food-recall workbooks results than INTERSALT, especially in the elderly and before and aer the intervention. Results showed that

9781284348286_CH03_083_104.indd 93 08/06/17 7:36 PM 94 Chapter 3 Nutritional Epidemiology and Research Methods

© Jonesadolescents & Bartlett who Learning, participated LLCin the garden-based nutri- © Jones & Bartlett Learning,Observing LLC tion intervention increased their servings of fruits and NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION vegetables more than students in the control groups. Counting cases or events Also, the experimental group’s vitamin A, vitamin C, 68 and ber intake increased signicantly. Relating cases or events to populations at risk In an analytical research design, the aim is to evaluate hypotheses based© Jones on existing & Bartlett knowledge Learning, LLC Making comparisons© Jones & Bartlett Learning, LLC or ndings. e DASHNOT trial, FORdiscussed SALE earlier, OR was DISTRIBUTION a NOT FOR SALE OR DISTRIBUTION randomized, multicenter, controlled feeding study to Developing the hypothesis compare the eects of three dietary patterns on blood pressure. e control group consumed fruits and veg- Testing the hypothesis etables ©and Jones a combination & Bartlett diet Learning, (having less LLCtotal fat, © Jones & Bartlett Learning, LLC less saturated fat, and less cholesterol than the fruit NOT FOR SALE OR DISTRIBUTION NOTDrawing FOR scientific SALE inferences OR DISTRIBUTION and vegetable diet or the control diet). e dietary patterns diered in selected nutrients hypothesized to Conducting experimental studies alter blood pressure. is study examined the food- group structure and nutrient composition of the study Intervening and evaluating © Jonesdiets & Bartlett and participant Learning, nutrient LLC consumption during © Jones & Bartlett Learning, LLC NOT FORthe SALEintervention. OR DISTRIBUTION Participants consumed the control NOTFIGURE FOR 32 SALEExamples OR of the DISTRIBUTION scientiﬁc method of dietary pattern during a 3-week start-up period. ey epidemiologists. were then randomized either to continue on the control diet or to change to the fruits and vegetables or nutrition research because they depend on isolation of the combination diet for© 8Jones weeks. Sodium& Bartlett intake Learning, and one LLC or more variables that can be© measured Jones &and Bartlett analyzed Learning, LLC body weight were constant during the entire feeding statistically. In quantitative (objective) research, random NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION period. Analysis of variance models compared the sample surveys and structured interviews are used to nutrient content of the three diets. collect mostly quantiable data that are analyzed using Targeting a few nutrients thought to inuence blood statistical techniques. is method promotes gathering pressure resulted in diets that were very dierent in their data that can be veried by another researcher and food-group© Jones and nutrient & Bartlett composition. Learning, e control LLC and generalized to© other Jones populations. & Bartlett Deductive Learning, reasoning, LLC fruits andNOT vegetables FOR SALE diets containedOR DISTRIBUTION more oils, table objectivity, quasi-experiments,NOT FOR SALE statistical OR DISTRIBUTION techniques, fats, salad dressings, and red meats and were higher and control characterize these methods.70 in saturated fat, total fat, and cholesterol than was the Although many nutrition research questions are open combination diet. e fruits and vegetables and combi- to traditional methodologies, others require a dierent nation diets contained relatively more servings of fruits, approach. ere is a growing acceptance of nutrition © Jonesjuices, & Bartlett vegetables, Learning, and nuts/seeds, LLC and were higher in © Jonesscience as& a Bartlett composite Learning, of dierent viewpointsLLC and var- NOT FORmagnesium, SALE OR potassium, DISTRIBUTION and ber than was the control NOTied FORresearch SALE methodologies. OR DISTRIBUTION70,71 Although recognition diet. Both the fruits and vegetables and combination of qualitative (subjective) methods for scientic study diets were low in sweets and sugar-containing drinks. is relatively recent in nutrition literature, they began e combination diet contained a greater variety of to be used in the United States during the late 1800s.5 fruits, and its high calcium© Jonescontent was& Bartlett obtained byLearning, in- Qualitative LLC strategies were used© Jonesto reveal & the Bartlett rapidly Learning, LLC creasing low-fat dairy products.NOT FOR In addition, SALE the OR distinct DISTRIBUTION developing social problems in citiesNOT at FOR that time SALE caused OR DISTRIBUTION food grouping pattern across the three diets resulted in by industrialization, urbanization, and mass immigration. 72 substantial dierences in the levels of vitamins A, B6, Munhall and Boyd stated that qualitative descriptions C, and E; folate; and zinc.69 encouraged social change by making urban problems FIGURE 32 presents examples of the scientic method visible to the public. used by© epidemiologists. Jones & Bartlett Learning, LLC Qualitative© Jonesmethods & may Bartlett be more Learning, benecial for LLC a NOT FOR SALE OR DISTRIBUTION particular incidentNOT FORbeing studied.SALE ORFor example,DISTRIBUTION qualitative approaches are rarely appropriate in genetic ▸ Quantitative and Qualitative engineering but may be appropriate when analyzing a Methods social behavior, such as food choice and dieting. ey © Jones & Bartlett Learning, LLC © Jonesuse semi-structured & Bartlett or Learning, interactive interviewsLLC to collect NOT FOR e SALE methodologies OR DISTRIBUTION appropriate for answering the NOTdata FOR relating SALE to people’s OR DISTRIBUTION judgments, attitudes, prefer- research question need to be considered. Quantitative ences, priorities, and/or perceptions about a nutrition methods have been the standard methodology used in issue; the data are then analyzed through sociological

9781284348286_CH03_083_104.indd 94 08/06/17 7:36 PM Types of Sampling 95

© Jones & Bartlettor anthropological Learning, research LLC techniques. Inductive© Jones North & Bartlett Carolina Learning,county. e purpose LLC of the study was to reasoning, subjectivity, discovery, description, and the describe an approach to conducting a community audit NOT FOR SALE OR DISTRIBUTION NOT FOR SALEmeaning OR of DISTRIBUTION an experience to an individual or groups (consisting of windshield tours and ground truthing) to typify qualitative methods.73 compute resources, to assess community characteristics, Both quantitative and qualitative data collection and to inform revisions to a community guide on nutrition methods can be combined in a single research study. and physical activity resources. Community audits have is is known as ©a mixed Jones method & Bartlett study. For Learning, example, LLCbeen used for epidemiological© Jones studies & Bartlettand in program Learning, LLC adding subjectiveNOT and open-ended FOR SALE questions OR DISTRIBUTION to a stan- planning for health-promotionNOT FOR interventions. SALE OR77 In DISTRIBUTIONthis dard household survey could provide a better approach study, the researchers used Google Earth mapping soware to dening and measuring poverty.74 e use of mixed to examine commercial and residential districts for the methods will increase as the ability to acquire and presence of sidewalks using both aerial and street-level analyze© Jones large & amounts Bartlett of Learning, quantitative LLC data improves views. ey© Jones also determined & Bartlett the Learning, number of fast-foodLLC along with improvements in electronic technologies. restaurants through the Reference USA business data- NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION For example, it is now possible to measure and record base. e researchers compared their observations on in an accessible database every step and movement community characteristics with available secondary data made by an individual, as well as every word he or she sources to examine the level of agreement. e initial speaks and hears.74 resource guide included 42 resources; the community © Jones & BartlettAn example Learning, of research LLC that utilized both quantitative© Jones audits & Bartlett identied Learning, 38 additional LLC resources. Results showed NOT FOR SALEand qualitative OR DISTRIBUTION methods was a study that examinedNOT the FORthat SALE there was OR moderate DISTRIBUTION to high agreement between awareness and acceptability of brown rice in Chinese windshield tour observations and secondary data sources adults, and the possibility of introducing brown rice into for several community characteristics, such as number their diet. is was part of a large, long-term random- of fast-food restaurants (67 percent agreement) and ex- ized clinical trial to© lowerJones the & risk Bartlett for acquiring Learning, type 2 LLCistence of sidewalks (100© percent Jones agreement). & Bartlett Moreover, Learning, LLC diabetes. e authors used questionnaires (quantitative) the audit resulted in an enhanced understanding of the NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION and focus-group discussions (qualitative) to collect data related barriers and facilitators to lifestyle change. e from participants, who resided in Shanghai. e focus techniques presented in this article may serve as a group discussion was about their dietary preferences, model for health-promotion professionals in other rural purchasing practices for specic varieties of rice, and communities.76 their© Jones knowledge & Bartlett and awareness Learning, of brown rice.LLC Ninety-four © Jones & Bartlett Learning, LLC percentNOT FORof the participantsSALE OR consumed DISTRIBUTION white rice daily; only NOT FOR SALE OR DISTRIBUTION 8 percent of the participants had previously consumed ▸ Types of Sampling brown rice. Before tasting, most participants considered brown rice inferior to white rice in terms of taste and Regardless of the type of research methodology, the quality. However, aer tasting brown rice and learning research questions and methodologies chosen must © Jones & Bartlettabout its nutritionalLearning, value, LLC the majority indicated greater© Jones guide & Bartlett the plan forLearning, data analysis. LLC It is helpful to identify NOT FOR SALEwillingness OR DISTRIBUTION to consume brown rice. e main barriersNOT FORthe computerSALE OR soware DISTRIBUTION that will assist with data analysis to acceptance were the perception of rough texture and and decide how results will be presented. For example, unpalatable taste, as well as higher price. All participants the researcher may want to prepare tables for the data suggested that large-scale promotion was needed to once they are collected. ese tables will help ensure all change societal attitudes© Jones toward & brownBartlett rice. Learning, In addition, LLCthe data necessary for answering© Jones the & research Bartlett questions Learning, LLC most of the participantsNOT articulatedFOR SALE their OR willingness DISTRIBUTION to have been collected. NOT FOR SALE OR DISTRIBUTION participate in a future long-term brown rice intervention e research questions and plan for data analysis will study. ese results provided valuable information for guide sample selection and size. ere are many methods the design of the future brown rice intervention trial of sample selection, but two will be considered here: and highlighted the importance of increasing awareness random and deliberate sampling. Random sampling about© Jones the nutritional & Bartlett value Learning, of brown rice. LLC75 means ©that Jones every case& Bartlett or participant Learning, has an equalLLC op- NOTAnother FOR example SALE is OR a community DISTRIBUTION audit qualitative and portunityNOT to beFOR included SALE in OR the DISTRIBUTIONstudy.5 For example, quantitative research method in which researchers drive rst the target group or community is informed about through a community to observe its physical and social the research study’s purpose/aim, as well as its risks, and attributes, mainly through windshield tours and ground is asked to participate. en, individuals consenting to © Jones & Bartletttruthing Learning,. Ground truthing LLC is a verication process that© Jones participate & Bartlett in the Learning, study are assigned LLC randomly to one, NOT FOR SALEuses data OR gathered DISTRIBUTION by direct observation to validate dataNOT FORtwo, SALE or more OR treatment DISTRIBUTION or intervention groups. e gathered from secondary sources.76 is method was used individuals randomized to receive the standard inter- to conduct research in 10 communities in a rural eastern vention treatment serve as the control group. e main

9781284348286_CH03_083_104.indd 95 08/06/17 7:36 PM 96 Chapter 3 Nutritional Epidemiology and Research Methods

Decide the information required. Dietary behavior, cultural inﬂuence, physical activity, food insecurity © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Describe the target participants.NOT FOR SALE OR DISTRIBUTIONPregnant women, teenagers, children, olderNOT adults, FOR households, SALE OR DISTRIBUTION college students

Select the research methodology appropriate for Food frequency questionnaire, 24-hour questionnaire, the target participants. and others © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC DecideNOT on question FOR SALEcontent. OR DISTRIBUTION Nutrient intake, physicalNOT activity FOR levels, SALE food availability, OR DISTRIBUTION demographics

Develop the question wording. Avoid technical terms and jargon, be clear and concise, measure one thing per question, avoid leading questions © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FORPut SALE questions OR into DISTRIBUTION a meaningful order and format. AdministerNOT FOR by telephone, SALE in-person, OR DISTRIBUTION online, mail.

Check the length of the questionnaire. Keep the questionnaire short

Pretest the questionnaire. Administer the questionnaire to a few participants for feedback © Jones & Bartlett Learning,and modiﬁcation LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Develop the ﬁnal survey form. Getting the permission from participants. If it is online, how it looks, consider the readability

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC ■ factor isNOT random FOR assignment; SALE OR that DISTRIBUTIONis, chance determines Must meanNOT the FORsame thing SALE to allOR respondents DISTRIBUTION the intervention assignment. Deliberate sampling ■ Must determine the hypotheses around which the means that specic target groups or communities are questionnaire is to be designed 5 invited to participate in the study. e choice of research TABLE 34 presents nine steps involved in the devel- methodologies will determine the characteristics of the opment of a questionnaire.79,78 © Jonessample & Bartlett population. Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALEAer consultation OR DISTRIBUTION with a statistician, the researcher NOT FOR SALE OR DISTRIBUTION determines the factors, such as the sample size, gender, ▸ Concepts of Collaborative and age range. e researcher may use power analysis soware to help determine sample size. Most of the time, Research it is impossible to know the exact number of people living © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC in the community being studied, so the nutritionist may Collaboration is very important in research and sometime estimate the total populationNOT based FOR on SALE census ORdata. DISTRIBUTIONis required for community typesNOT of grants. FOR Collaboration SALE OR DISTRIBUTION involves mutual participation in decision making and working with others toward a common goal. It is a process General Rules for Writing Survey of joint decision making in an atmosphere of shared respect 73 Questions© Jones & Bartlett Learning, LLC and cooperation.© Jones Collaboration & Bartlett should Learning, always involve LLC NOT FOR SALE OR DISTRIBUTION interaction amongNOT community FOR SALE nutritionists, OR DISTRIBUTION clients, and In addition, the nutritionist may need to write or mod- other healthcare practitioners. Nutrition professionals ify an existing questionnaire for collecting data from work as members of a team in all healthcare settings. In the targeted community. e general rules for writing community nutrition settings, the nutritionist is part of a survey questionnaire include, but are not limited to, 78 a multidisciplinary team that includes other healthcare © Jonesthe & following Bartlett :Learning, LLC © Jonesworkers, as& wellBartlett as community Learning, organizations, LLC social service NOT FOR■ SALEMeasuring OR the DISTRIBUTION fundamental concept it is intended NOTagencies, FOR andSALE the judicial OR DISTRIBUTION system. e multidisciplinary to obtain team can conduct research studies that can benet the ■ Must not measure other concepts public. e need for collaborative research is especially © Jones & Bartlett Learning, LLC. NOT FOR SALE OR DISTRIBUTION.

9781284348286_CH03_083_104.indd 96 08/06/17 7:36 PM Epidemiological Approaches to Community Health Assessment 97

© Jones & Bartlettstrong now Learning, due to the prevalence LLC of health disparities© in Jones & Bartlett5. Recognize Learning, that other LLC grassroots community the United States. In addition, there is a belief that part- organizations are potential partners. NOT FOR SALE OR DISTRIBUTION NOT FOR SALEnerships OR between DISTRIBUTION researchers and community members 6. Recruit community-based sta. can contribute not only to the ability of communities to 7. Engage in strategic planning. eectively address local issues, but also to the development 8. Meet training needs. of new knowledge.80,81 9. Involve other relevant community members, Successful collaboration© Jones requires & Bartlett that the Learning, multidisci- LLC public institutions,© Jones and organizations. & Bartlett Learning, LLC plinary team developNOT a common FOR SALE purpose, OR communicate DISTRIBUTION 10. CommunicateNOT and pursue FOR research SALE and OR service DISTRIBUTION to coordinate eorts, and recognize the unique and com- agendas at the community level. plementary skills possessed by each team member. Each 11. Facilitate the transfer of knowledge. team member brings special abilities and expertise to the 12. Maintain trusting and collaborative collaborative© Jones & process. Bartlett Collaboration Learning, does LLC not work when © Jonesrelationships. & Bartlett Learning, LLC one team member designs a research project or nutrition NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION program and then “coordinates” by informing others of the plan. For a nutrition program or research to be suc- ▸ Reporting Research Results cessful, collaboration must be a joint eort on the part of the clients and all team members to identify mutual Reporting research ndings is an essential aspect of © Jones & Bartlettgoals and Learning, acceptable means LLC for meeting those goals.©82 Jonesconducting & Bartlett research Learning, activities. LLC Reporting the results NOT FOR SALEAlthough OR DISTRIBUTION they do share some similarities, collaboNOT- FORof research SALE ORoutcomes DISTRIBUTION in professional journals and at ration and coordination are dierent. Coordination is professional meetings facilitates the growth and use the act of managing interdependencies among activi- of nutrition knowledge. By applying research results, ties. It is the ability to manage services without gaps or nutritionists can institute social change. Policy makers, 80 other professionals, and the community can then learn overlaps. e nutritionist,© Jones as & coordinator, Bartlett Learning, may or may LLC © Jones & Bartlett Learning, LLC not consult others when carrying out a management that the research ndings are relevant and applicable. NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION function. Collaboration, on the other hand, involves e results of a study involving the community could joint decision making between two or more people. support expansion of community or public health nu- Collaboration also can enhance the opportunity for trition services. e nutrition profession is in urgent obtaining grant money. Many granting agencies encourage need of research documenting the cost eectiveness collaborative© Jones & eorts. Bartlett A community Learning, nutritionist LLC can col- of nutrition© Jones care. Eectiveness,& Bartlett Learning,however, can LLC be mea- laborateNOT FOR with statisticians, SALE OR epidemiologists, DISTRIBUTION nurses, social sured inNOT many FOR ways SALEbeyond ORcost, DISTRIBUTION including quality of care, client satisfaction, and the ability for a person to workers, physicians, and colleagues to complement his 83 or her nutrition expertise.39 Collaboration is especially maintain independence. helpful for a new university faculty member, rst-time grant seeker, or nutritionist who wants to change his or © Jones & Bartletther focus Learning, area of research. LLC Community nutritionists ©can Jones ▸ &Epidemiological Bartlett Learning, LLC Approaches NOT FOR SALEnd collaborators OR DISTRIBUTION by networking with colleagues NOTand FOR SALEto Community OR DISTRIBUTION Health community members and communicating with other professionals or experts in the area of interest. Assessment Amuwo and Jenkins81 outlined four distinct phases e changing disease patterns throughout the world are necessary for successful© Jones partnership & Bartlett buildingLearning, in LLC © Jones & Bartlett Learning, LLC linked to changing lifestyles. One of the functions of community-basedNOT research, FOR based SALE on their OR experiences DISTRIBUTION NOT FOR SALE OR DISTRIBUTION with collaborative work: networking, cooperation, col- epidemiological approaches is to determine the etiology laboration, and partnership. Within these four phases, or cause of the disease or risk factors for the disease. they identied 12 concrete, systematic steps that a One way that epidemiologists study the cause and eect of disease is through a three-sided conceptual model partnership-building eort must complete in a rather 5 sequential© Jones order, & Bartlett as follows Learning,81: LLC known© as Jones the epidemiological & Bartlett Learning,triangle. is LLC model NOT FOR SALE OR DISTRIBUTION explainsNOT how FORchanges SALE in one OR element DISTRIBUTION can inuence the 1. Involve the community early so they can see occurrence of a disease by increasing or decreasing an the importance of the program. individual’s risk for the health condition. 2. Be aware that partnership building is a non- linear procedure. © Jones & Bartlett3. Learning,Recruit and empowerLLC important community-© Jones The & BartlettEpidemiological Learning, Triangle LLC NOT FOR SALE OR basedDISTRIBUTION personnel. NOT FOR e SALEepidemiological OR DISTRIBUTION triangle explores health and disease 4. Conduct a community needs assessment using three elements: an agent, a host, and an envi- (both human and capital resources). ronment (each of which is discussed in this section). © Jones & Bartlett Learning, LLC. NOT FOR SALE OR DISTRIBUTION.

9781284348286_CH03_083_104.indd 97 08/06/17 7:36 PM 98 Chapter 3 Nutritional Epidemiology and Research Methods

© JonesAll & three Bartlett of the Learning,components LLCof this model are of equal © Jonesthe Midwestern & Bartlett United Learning, States). An environmental LLC factor strength. FIGURE 33 shows the model in its normal state also may have multiple levels, such as low, middle, and NOT FOR SALE OR DISTRIBUTION NOT FORof equilibrium. SALE OR EquilibriumDISTRIBUTION does not mean optimal high socioeconomic status. Low socioeconomic status health, but it does signal the usual patterns of illness is associated with the reduced availability of health and and health in a population. Any change in one of the social services.84 High socioeconomic status is associ- components (agent, host, or environment) will result ated with excessive behaviors, such as heavy drinking, in disequilibrium (i.e., a© change Jones in the& Bartlettusual patterns). Learning,6 overeating, LLC and risk taking.85,86© Jones & Bartlett Learning, LLC e nutritionist usingNOT the epidemiologicalFOR SALE OR triangle DISTRIBUTION e analysis of these threeNOT areas—agent, FOR SALE host, andOR DISTRIBUTION in a situation such as teenage obesity might focus on diet environment—should provide information regarding as the agent causing teenage obesity. However, behavior groups at risk for increased rates of obesity and may pro- also could be a factor. An example of a behavioral factor vide direction toward a program aimed at reducing the is a sedentary© Jones lifestyle & Bartlett that promotes Learning, weight gain.LLC e risks. erefore,© theJones epidemiological & Bartlett triangle Learning, can provide LLC nutritionist would strive to determine whether there a useful guide for investigating dierent health problems NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION has been a change in any of the possible agents. of teenage obesity, as well as other health problems. e host factors are personal characteristics of the teenage population who are at risk for obesity and other Host health-related conditions, such as diabetes. e personal © Jonescharacteristics & Bartlett includeLearning, age, ethnicity, LLC gender, socioeco- © Jones & Bartlett Learning, LLC NOT FORnomic SALE status, OR eating DISTRIBUTION patterns, exercise behavior, and NOT FOR SALE OR DISTRIBUTION lifestyle. Some personal characteristics, such as gender, age, and ethnicity, are not modiable, but others are. Characteristics that can be changed include nutritional status, physical activity, and© Jones income &level. Bartlett By evaluating Learning, LLCAgent © JonesEnvironment & Bartlett Learning, LLC these factors, it may be possible to identify groups of NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION teenagers who are at an increased risk for dying from obesity-related health conditions. Finally, the environment must be assessed. Environ- mental factors are external factors that encompass an agent and© Jonesa host. ey & Bartlett may be exclusively Learning, physical, LLC such FIGURE 33 Epidemiological© Jones & triangle Bartlett model. Learning, LLC as climateNOT or surroundings FOR SALE (e.g., OR an DISTRIBUTIONurban housing area in Modiﬁed from: Stanhope M, LancasterNOT J. Community FOR and Public SALEHealth Nursing. 5th OR ed. St. Louis, DISTRIBUTION MO: Mosby; 2000.

Successful Community Strategies © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOREPIC SALE Elderly OR Cohort DISTRIBUTION Research Study87-97 NOT FOR SALE OR DISTRIBUTION The Mediterranean diet has been reviewed in many studies because several of its components have been related to decreases in common chronic diseases. Ecological evidence suggests that such a diet may be beneficial to health, and variants of this diet have improved the prognosis of patients with coronary heart disease. The Mediterranean diet is characterized by a high© intakeJones of vegetables, & Bartlett legumes, Learning, fruits, and LLC cereals; a moderate to high intake© Jones of fish; a& low Bartlett Learning, LLC intake of saturated lipidsNOT but a highFOR intake SALE of unsaturated OR DISTRIBUTION lipids, particularly olive oil; a low to moderateNOT FORintake ofSALE dairy OR DISTRIBUTION products, mostly cheese and yogurt; a low intake of meat; and a modest intake of ethanol, mostly as wine. Several studies have reported that a diet that adheres to the principles of the traditional Mediterranean diet is associated with longer survival. Trichopoulou and Vasilopoulou87 used a 10-unit dietary score to determine whether adherence to a Mediterranean diet was associated with a longer life expectancy among elderly Caucasians. However, several studies have ©used Jones variations & Bartlettof this score Learning, and reported LLC opposite associations with overall© Jones mortality. & These Bartlett studies, Learning, however, LLC reliedNOT on small FOR samples SALE of mostly OR DISTRIBUTIONelderly participants or only on the Greek population.NOT FOR SALE OR DISTRIBUTION The Caucasian Prospective Investigation into Cancer and Nutrition (EPIC) study used a multicenter, prospective cohort study to investigate the role of biological, dietary, lifestyle, and environmental factors in cancer and other chronic diseases and to examine whether adherence to the modified Mediterranean diet, in which unsaturated fats were substituted for monounsaturated fats, is associated with longer life expectancy among elderly persons. © Jones & BartlettBetween 1992 Learning, and 2000, 519,978LLC healthy volunteers were© recruited Jones in & 23 Bartlett centers from Learning, 10 European LLC countries NOT FOR(Denmark, SALE ORFrance, DISTRIBUTION Germany, Greece, Italy, the Netherlands, Norway,NOT FORSpain, Sweden,SALE andOR the DISTRIBUTION United Kingdom). After initial (continues)

9781284348286_CH03_083_104.indd 98 08/06/17 7:36 PM Chapter Summary 99

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALESuccessful OR DISTRIBUTION Community StrategiesNOT FOR SALE OR DISTRIBUTION (continued) screening, only 74,607 individuals were included in the study. Data were collected from participants ages 60 or over. Information on the vital status of participants was obtained from mortality registries and by active follow-up. The earliest and latest years of follow-up were 1999 (some participants in the Netherlands) and December 2003 (most of the centers). Participants were© classified Jones as &alive Bartlett at last follow-up, Learning, dead, emigrated,LLC refused to participate© further, Jones or unknown.& Bartlett Learning, LLC Dietary intakesNOT were FORassessed SALE through OR compatible DISTRIBUTION instruments (food frequency questionnairesNOT FOR and, SALEin some OR DISTRIBUTION centers, records of intake over 7 or 14 days) that had been developed and validated within each center. In addition, a computerized instrument for recall of dietary intake over 24 hours was developed to collect information from a standardized random sample of the aggregate cohort. The aim was to calibrate the measurements across countries. Nutrient intakes were calculated using food composition tables specific to the country. Fourteen food groups and ©nutrients Jones were & Bartlett considered: Learning, potatoes, vegetables, LLC legumes, fruits, dairy products,© Jones cereals, & Bartlett meat and Learning,meat products, LLC fish NOTand seafood, FOR SALEeggs, monounsaturated OR DISTRIBUTION lipids, polyunsaturated lipids, saturatedNOT lipids,FOR sugar SALE and confectioneryOR DISTRIBUTION (candy, pastry, etc.), and nonalcoholic beverages. Results showed that an increase in the modified Mediterranean diet score was associated with lower overall mortality. When dietary exposures were calibrated across countries, the reduction in mortality was 7 percent (ranging from 1 to 12 percent). The study concluded that the Mediterranean diet, modified so as to apply across Europe, was © Jones & Bartlettassociated Learning, with increased LLC survival among older people.© Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Children and Adolescents, National Health and Nutrition Examination Survey98 The National Health and Nutrition Examination Survey (NHANES) is designed to assess the health and nutritional status of adults and children in the United States. The survey is a combination of interviews and physical examinations. The objective of this NHANES study was to describe perceptions of child weight status among U.S. children, adolescents, and their parents© and Jones to examine & Bartlett the degree Learning, in which personal LLC and parental perception of© weight Jones status & Bartlettis associated Learning, LLC with self-reportedNOT attempted FOR weight SALE loss. OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION The sample included 2,613 participants aged 8–15 years in the NHANES from the two most recent consecutive cycles (2007–2008 and 2009–2010). Classifications of weight perception were developed by comparing measured to perceived weight status. Multivariable logistic regression analyses were used to examine the association between weight misperception and self-reported attempted weight loss. © JonesResults showed& Bartlett that among Learning, children LLC and adolescents, 27.3 percent ©underestimated Jones & Bartlett and 2.8 percent Learning, overestimated LLC NOTtheir weightFOR status.SALE Among OR DISTRIBUTION parents, 25.2 percent underestimated and 1.1NOT percent FOR overestimated SALE OR their DISTRIBUTION child’s weight status. Logistic regression analyses showed that the odds of self-reported attempted weight loss was 9.5 times as high among healthy-weight children and adolescents who overestimated their weight status as among those who perceived their weight status accurately; the odds of self-reported attempted weight loss were 3.9 and 2.9 times as high among overweight and obese children and adolescents, respectively, who accurately perceived their weight status than among © Jones & Bartlettthose who Learning, underestimated LLC their weight status. Parental© Jonesmisperception & Bartlett of weight Learning, was not significantly LLC associated with NOT FOR SALEself-reported OR DISTRIBUTION attempted weight loss among children NOTand adolescents FOR SALE who were OR overweight DISTRIBUTION or obese. The authors recommended that to prevent childhood obesity both children and parents should be included in the education regarding the appropriate identification and interpretation of actual weight status. Interventions for appropriate weight loss can target children directly because of the child’s perception of his or her weight status.

Chapter© Jones Summary & Bartlett Learning, LLC instance,© Jones an increase & Bartlett in the Learning,prevalence of LLC diabetes or breast cancer means that there are more persons NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION ■ A community nutritionist is a member of a multi- with these health conditions in the population. disciplinary team and can conduct research studies ■ Successful collaboration requires that the multidisci- that can benet the public. plinary team develop a common purpose, communicate ■ Epidemiological and demographic measures and to coordinate eorts, and recognize the unique and © Jones & Bartlettresearch Learning, methods LLCare used to study nutrition ©and Jones &complementary Bartlett Learning, skills possessed LLC by each team member. NOT FOR SALEhealth-related OR DISTRIBUTION conditions. NOT FOR■ SALE e three OR components DISTRIBUTION of the epidemiological triangle ■ Measures of incidence and prevalence provide dier- model are agent, host, and environment, and they ent information and have dierent implications. For are used to explore health and disease. © Jones & Bartlett Learning, LLC. NOT FOR SALE OR DISTRIBUTION.

9781284348286_CH03_083_104.indd 99 08/06/17 7:36 PM 100 Chapter 3 Nutritional Epidemiology and Research Methods

■ © Jones &In Bartlett experimental Learning, studies, theLLC investigator has control © JonesCritical & Thinking Bartlett ActivitiesLearning, LLC over participants, procedures, and exposures. NOT FOR SALE OR DISTRIBUTION NOT■ FORIdentify SALE a research OR DISTRIBUTION study in the literature rele- ■ Epidemiology is the study of associations, and sta- vant to community and public health nutrition. tistical methods are used to test the associations. Then identify the strengths and limitations of e existence of a statistically signicant association the research. does not indicate that the observed relationship is ■ Communicate/present the research findings one of cause and eect.© Jones Causality & is Bartlett supported Learning, when LLC © Jones & Bartlett Learning, LLC in class. a number of studiesNOT that are FOR conducted SALE at OR dierent DISTRIBUTION NOT FOR SALE OR DISTRIBUTION ■ Meet with the librarian as a class/group and learn times, using dierent methods, and performed how to use the Medline database; ERIC database; among dierent populations show similar results. CINAHL database; Ingenta: e Global Research ■ In quantitative research, random sample surveys and Gateway (http://www.ingenta.com), a portal to structured interviews are used to collect quantiable © Jones & Bartlett Learning, LLC online information© Jones around& Bartlett the world; Learning, and other LLC data; qualitative research analyzes social behavior NOT FOR SALE OR DISTRIBUTION search enginesNOT forFOR nding SALE grant OR sources. DISTRIBUTION such as food choice and dieting.

CASE STUDY 31: Collaboration Eﬀorts Between Nutrition © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALEStudents OR DISTRIBUTION and Business NOT FOR SALE OR DISTRIBUTION

Damian and Monica, two senior baccalaureate nutrition students in a community nutrition course, had their community outreach experience at a rural community clinic. They were involved with the owner of a local candy factory in planning for disease prevention and health promotion activities for the employees. After consulting with the faculty member responsible for the course© andJones the director & Bartlett of the community Learning, clinic, LLC they designed a survey to distribute© Jones to employees & Bartlett Learning, LLC to assess their needs forNOT health FOR education SALE programming. OR DISTRIBUTION The students submitted the survey, whichNOT included FOR blood SALE OR DISTRIBUTION pressure screening, to the nutrition professor for approval before distributing it to the employees. The students also reviewed the medical records of the participants, with their consent. Some of the employee priorities identified from the survey data that need to be addressed were the following problems: © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC ■ High blood pressure (160/90 on average) ■ NOTSmoking FOR cessation SALE (2 packsOR DISTRIBUTIONa day on average) NOT FOR SALE OR DISTRIBUTION ■ Weight loss (average body fat percentage of 49 percent) ■ High stress level The students met with the factory owner, the faculty member, and a representative of the local chapter of the © Jones &American Bartlett Lung Learning, Association to LLC plan for implementation of this© initiative.Jones A& timeline Bartlett was Learning,established to LLCnotify employees NOT FORof SALE an opportunity OR DISTRIBUTION to participate in the weight loss, stress reduction,NOT FOR and smoking SALE cessation OR DISTRIBUTION program. This included notices distributed in employee paychecks and signs posted in prominent locations throughout the factory. The students discussed plans for conducting a research study related to implementation and evaluation of the weight loss, stress reduction, and smoking cessation program with the faculty member. The students collected a 3-day dietary record and observed that the participants consumed high-fat, high-sodium snacks and plenty of candy, mostly from the vending machine located© inJones their break & Bartlett rooms. After Learning, reviewing their LLC medical records with permission,© Jones data showed & Bartlett a Learning, LLC fasting blood glucose levelNOT of 150/100 FOR SALE mg/dl. OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Questions 1. What is community nutrition research? 2. What is the definition of epidemiology? Describe the methods of epidemiological studies. What kind of ©epidemiological Jones & Bartlett study could Learning, Damian and LLC Monica use for their proposed© research Jones study? & Bartlett Learning, LLC 3. NOT Describe FOR each SALE of the stages OR DISTRIBUTIONof the research process that Damian and MonicaNOT must FOR understand. SALE OR DISTRIBUTION 4. What are the differences between collaboration and coordination? Who could Damian and Monica collaborate with (other than their current collaborators) on their new research study? 5. What is the epidemiological triangle? Describe a situation in which Damian and Monica could utilize the epidemiological triangle. © Jones & 6.Bartlett What is theLearning, difference betweenLLC the quantitative and qualitative© Jones methods & Bartlett that Damian Learning, and Monica LLC must know NOT FOR SALEbefore OR they DISTRIBUTION can carry out the project? NOT FOR SALE OR DISTRIBUTION 7. What are some of the negative implications of cigarette smoking?

9781284348286_CH03_083_104.indd 100 08/06/17 7:36 PM References 101

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR CASE DISTRIBUTION STUDY 31: CollaborationNOT Eﬀorts FOR SALE Between OR DISTRIBUTION Nutrition Students and Business (continued)

8. What types of dietary assessments could be used in a cross-sectional research study? What type of dietary assessment© would Jones be useful & Bartlett for Damian Learning, and Monica’s LLC research study? © Jones & Bartlett Learning, LLC 9. Describe the diﬀerence between incidence and prevalence that Damian and Monica must know before carrying out the project.NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 10. Work in small groups or individually to discuss the case study and practice using the Nutrition Care Process chart provided on the companion website. You can also add other nutrition and health-related conditions or assessments to the case study to make it more challenging and interesting. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC ThinkNOT FORAbout SALE It OR DISTRIBUTION nutritionalNOT epidemiology: FOR SALE OR Study DISTRIBUTION of dietary intake and the occurrence of disease in human populations. Answer: Qualitative research method random sampling: When every case or participant has an equal opportunity to be included in a study. © Jones & BartlettKey Terms Learning, LLC © Jonesground & Bartlett truthing: Learning, A verication LLC process that uses data NOT FOR SALEcase-control OR DISTRIBUTION study: Exposure and other characterNOT- FORgathered SALE by directOR DISTRIBUTION observation to corroborate data gath- istics of cases of the disease under investigation are ered from secondary sources. compared with a control group of persons unaected by the disease. e results are analyzed to acquire eect References estimates. © Jones & Bartlett Learning, LLC 1. Israel BA, Schulz AJ,© Parker Jones EA, & Becker Bartlett AB. Review Learning, LLC cohort study: GroupsNOT ofFOR individuals, SALE denedOR DISTRIBUTION in terms of community-basedNOT research: FOR assessing SALE partnership OR DISTRIBUTION of their exposures, are followed over time to see if there approaches to improve public health. Annu Rev are dierences in the development of new cases of the Public Health. 1998;19:173-202. disease of interest (or other health outcome) between the 2. Frank-Spohrer G. Community Nutrition: Applying groups with and without exposure. Epidemiology to Contemporary Practice. Rockville, collaboration:© Jones & Bartlett Working Learning, with others LLCtoward a com- MD:© Jones Aspen; 2007.& Bartlett Learning, LLC monNOT goal. FOR SALE OR DISTRIBUTION 3. BrachmanNOT FOR PS. SALE Epidemiology OR DISTRIBUTION General Concepts. community nutrition research: Organized study of 4 ed. Galveston, TX: e University of Texas a trend at both the basic and more applied levels that fo- Medical Branch; 1996. cuses on social, structural, and physical environmental 4. Morgan R, Jain M, Miller A. A comparison of inequities through active involvement of community dietary methods in epidemiologic studies. Am © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC members. J Epidemiol. 2006;107(6):488-498. NOT FOR SALEcoordination: OR DISTRIBUTION Ecient management of servicesNOT FOR5. SALEStanhope OR M,DISTRIBUTION Lancaster J. Community and Pub- without gaps or overlaps. lic Health Nursing: Population-Centered Health cross-sectional study: Measures the prevalence of dis- Care in the Community. 9th ed. New York, NY: ease and measures exposure and eect at the same time. Elsevier; 2015. deliberate sampling:© Jones Inviting & Bartlett specic Learning, people to LLC6. Anderson E, McFarlane,© Jones J. Community & Bartlett as Partner: Learning, LLC participate in a study.NOT FOR SALE OR DISTRIBUTION eory and PracticeNOT in Nursing. FOR 3rd SALE ed. Philadelphia OR DISTRIBUTION demography: An analytic tool used to measure a pop- PA: Lippincott; 2010. ulation by recording births, deaths, age distribution, and 7. Hitchcock J, Schubert, PE, omas, SA. Commu nity other vital statistics. Health Nursing. 2nd ed. United States: omson ecological study: Involves the investigation of a group Delmar Learning; 2003. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC of people, such as those living within a geographic area 8. Center for Disease Control and Prevention. suchNOT as FORa region SALE or state. OR DISTRIBUTION REACHNOT FOR U.S. ndingSALE solutionsOR DISTRIBUTION to health dispar- epidemiology: e study of the determinants, oc- ities: at a glance 2010. www.cdc.gov. Accessed currence, and distribution of health and disease in a March 12, 2016. dened population. It studies scientically the factors 9. Berdanier CD. Handbook of Nutrition and Food. © Jones & Bartlettthat aect Learning, the health of LLC individuals in the community.© Jones & BartlettNew York Learning, NY: CRC Press; LLC 2013. NOT FOR SALEexposure: OR DISTRIBUTION Characteristics or agents (e.g., food, medNOT- FOR10. SALEAbramson OR DISTRIBUTIONJH, Abramson ZH. Survey Methods in ications, sunlight) that the participant comes in con- Community Medicine. 5th ed: Edinburgh & London: tact with that may be related to disease risk. Livingstone; 2000. © Jones & Bartlett Learning, LLC. NOT FOR SALE OR DISTRIBUTION.

9781284348286_CH03_083_104.indd 101 08/06/17 7:36 PM 102 Chapter 3 Nutritional Epidemiology and Research Methods

© Jones11. & BartlettStamler Learning,J, Elliott P, DennisLLC B. INTERMAP: © Jones25. Grant & Bartlett W. A multicountry Learning, ecologic LLC study of risk background aims, design methods and descrip- and risk reduction factors for prostate cancer NOT FOR SALE OR DISTRIBUTION NOT FOR SALEtive statistics OR DISTRIBUTION (nondietary). J Hum Hypertens. mortality. Eur Urol. 2004;45(3):271-279. 2003;17:591–608. 26. Dreon DM, Fernstrom HA, Campos H, et. al. Change 12. Zimlichman E, Kochba I, Mimouni FB, Shochat T, in dietary saturated fat intake is correlated with et al. Smoking habits and obesity in young adults. change in mass of large low-density-lipoprotein Addiction. 2005;100(7):1021-1025.© Jones & Bartlett Learning, LLCparticles in men. Am J Clin© Nutr.Jones 1998;67:828-836. & Bartlett Learning, LLC 13. Tande DDL. e associationsNOT FOR between SALE blood OR lipids DISTRIBUTION 27. Sukhinder KC, Luis BA.NOT Metabolism FOR SALE of choles OR- DISTRIBUTION and the Food Guide Pyramid: ndings from the terol is altered in the liver of C3H mice fed fats ird National Health and Nutrition Examination enriched with dierent C-18 fatty acids. J Nutr. Survey. Prev Med. 2004;38(4):452-457. 1999;129:1718-1724. 14. Ness© Jones AR, Powles & Bartlett JW. Fruit Learning, and vegetables LLC and 28. Greenspan© Jones SL, Resnick & Bartlett NM, Parker Learning, RA. Combi LLC- cardiovascular disease: a review. Int J Epidemiol. nation therapy with hormone replacement and NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 1997;26:1-13. alendronate for prevention of bone loss in elderly 15. Fiore H, Travis S, Whalen A, Auinger P, et al. women a randomized controlled trial. JAMA. Potentially protective factors associated with 2003;289:2525-2533. healthful body mass index in adolescents with 29. Perusse LB, Bouchard CE. e genetics of obesity. © Jones & Bartlettobese and Learning, nonobese parents:LLC a secondary data © JonesIn: & Bouchard Bartlett CE, Learning, ed. Family StudiesLLC . Boca Raton, NOT FOR SALEanalysis OR of DISTRIBUTIONthe ird National Health and Nutri- NOT FORFL: SALECRC Press; OR 1994:125. DISTRIBUTION tion Examination Survey 1988–1994. J Am Diet 30. Stettler N, Tershakovec T, Zemel B, Leonard M, Assoc. 2006;106(1):55-64. et al. Early risk factors for increased adiposity: a 16. Coulston A, Rock C, Monsen E. Nutrition in the cohort study of African American subjects followed Prevention and Treatment© Jones of &Disease. Bartlett New Learning, York LLCfrom birth to young adulthood.© Jones Am & J ClinBartlett Nutr. Learning, LLC NY: Academic Press; 2001. 2000;72:378–383. NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 17. Eugene W, Kaplan H, James S. e epidemiologic 31. Chen AY, José JE. Family structure and child- evidence for a relationship between social support hood obesity, early childhood longitudinal and health. Am J Epidemiol. 2005;17(5):521-537. study—kindergarten cohort. Prev Chronic Dis 18. Lubin F, Ruder M, Wax Y, Modan B. Over- 2010;7(3):A50-A60. weight© Jones and changes& Bartlett in weight Learning, throughout LLC adult 32. Huang ©Z, JonesWillett WC,& Bartlett Colditz Learning,GA, Hunter DJ,LLC lifeNOT in breastFOR SALEcancer etiology.OR DISTRIBUTION Am J Epidemiol. et al. WaistNOT circumference, FOR SALE waist:hip OR DISTRIBUTION ratio, and risk 2004;122(4):579-588. of breast cancer in the Nurses’ Health Study. Am 19. Berdanier CD, Feldman EB, Flatt WP, St. Jeor ST. J Epidemiol. 1999;150(12):1316-1324. Handbook of Nutrition and Food. New York, NY: 33. Millen BE, Quatromoni PA, Nam BH, O’Horo CE, CRC Press LLC; 2007. et al. Dietary patterns, smoking, and subclinical © Jones20. & BartlettRamesh M,Learning, Schraer C, LLC Mayer AM, Asay E, et al. © Jonesheart & Bartlett disease in Learning, women: opportunities LLC for pri- NOT FOR SALEEect OR of special DISTRIBUTION diabetes program for Indians’ NOT FORmary SALE prevention OR fromDISTRIBUTION the Framingham Nutrition funding on system changes in diabetes care and Studies. J Am Diet Assoc. 2004;104(2):208-214. outcomes among American Indian/Alaska Native 34. Winitz M, Seedman DA, Gra J. Studies in met- people 1994-2004. Int J Circumpolar Health. abolic nutrition employing chemically dened 2008;67:203–212.© Jones & Bartlett Learning, LLCdiets. I. Extended feeding© of Jones normal human& Bartlett adult Learning, LLC 21. Indian Health Service.NOT Special FOR Diabetes SALE Program OR DISTRIBUTION for males. Am J Clin Nutr. 1970;23(5):525-545.NOT FOR SALE OR DISTRIBUTION Indians. http://www.ihs.gov/. Accessed March 12, 35. Cummings DE, Schwartz MW. Genetics and 2016. pathophysiology of human obesity. Annu Rev 22. Cade JE. e UK Women’s Cohort Study: compar- Med. 2003;54:453-471. ison of vegetarians, sh-eaters and meat-eaters. 36. Behrens MM. e human taste receptor hTAS2R14 Public© Jones Health & Nutr. Bartlett 2004;7(7):871-878. Learning, LLC responds© toJones a variety & of Bartlett dierent bitter Learning, compounds. LLC 23. FrancoNOT FOR O. Blood SALE pressure OR inDISTRIBUTION adulthood and life BiochemNOT Biophys FOR Res Commun. SALE OR2004;319(2):479-485. DISTRIBUTION expectancy with cardiovascular disease in men 37. Smith J, Lensing S, Horton JA, Lovejoy J, et al. and women: life course analysis. Hypertension. Prevalence of self-reported nutrition-related 2005;46(2):280-286. health problems in the Lower Mississippi Delta. © Jones24. & BartlettKant A, BlockLearning, G, Schatzkin LLC A, Ziegler RG, et al. © JonesAm & J Bartlett Public Health. Learning, 1999;89:1418-1421. LLC NOT FOR SALEDietary OR diversity DISTRIBUTION in the U.S. population NHANES NOT38. FORSchneider SALE MJ. OR Introduction DISTRIBUTION to Public Health. Boston II 1976-1980. J Am Diet Assoc. 1991;91:34-38. MA: Jone and Bartlett; 2006.

9781284348286_CH03_083_104.indd 102 08/06/17 7:36 PM References 103

© Jones & Bartlett39. Monsen Learning, ER. Research: LLC Successful Approaches. 2nd© ed. Jones 56. & BartlettINTERSALT Learning, Cooperative LLC Research Group. So- Chicago, IL: American Dietetic Association; 2003. dium, potassium, body mass, alcohol, and blood NOT FOR SALE OR DISTRIBUTION NOT FOR SALE40. ORSheats DISTRIBUTION NN, Sheats NN. Prevalence, treatment, and pressure: the INTERSALT study. J Hypertens. control of hypertension among African Americans 1988;6(4):S584-S586. and Caucasians at primary care sites for medically 57. Crawford MD, Gardner MJ, Morris JN. Mortal- under-served patients. Ethn Dis. 2005;15(1):25-32. ity and hardness of local water supplies. Lancet. 41. Prutkin J, Fisher© Jones EM, Etter & L,Bartlett et al. Genetic Learning, variation LLC 1968:827-831. © Jones & Bartlett Learning, LLC and inferencesNOT about FOR perceived SALE taste OR intensityDISTRIBUTION in 58. Sacks FFM. EectsNOT on blood FOR pressure SALE of OR reduced DISTRIBUTION mice and men Physiol Behav. 2000;69:161-173. dietary sodium and the Dietary Approaches to Stop 42. Arnold A, Arnold AM. Incidence of cardiovascular Hypertension (DASH) diet. DASH-Sodium Collab- disease in older Americans: the cardiovascular orative Research Group. N Engl ed. 2001;344(1):3-10. © Joneshealth &study. Bartlett J Am Geriatr Learning, Soc. 2005;53(2):211-218. LLC 59. Sacks© Jones FM, Appel & Bartlett LJ, Moore Learning, TJ, Obarzanek LLC E. 43. National Center for Health Statistics. National A dietary approach to prevent hypertension: NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION vital statistics reports. http://www.cdc.gov/nchs/. a review of the Dietary Approaches to Stop Accessed March 12, 2016. Hypertension (DASH) Study. Clin Cardiol. 1999;22 44. Freeman J, Hutchison, G. Prevalence, incidence and (7 Suppl):1116-1110. duration. Am J Epidemiol. 2005;112(5):707–723. 60. Trichopoulou A, Orfanos P, Norat T, Bueno-de- © Jones & Bartlett45. Flanders Learning, WD, O'Brien LLC TR. Inappropriate compari© Jones- & BartlettMesquita Learning, B, et al. Modied LLC Mediterranean diet NOT FOR SALE ORsons DISTRIBUTION of incidence and prevalence in epidemiologicNOT FOR SALEand survival: OR DISTRIBUTION EPIC elderly prospective cohort research. Am J Public Health. 1989;79(9):1301-1303. study. BMJ. 2005;330:7498. 46. Robinson K. Family care giving: who provides the 61. Alpert JE, Fava, M. Nutrition and depression: the care, and at what cost? Nurs Econ. 1997;15(2):243. role of folate. Nutr Rev. 1997;55:145. 47. Paulette GB.© Populations Jones & at Bartlett risk across Learning, the lifespan LLC62. Sacks FM, Willett ©WC, Jones Smith &A, BartlettBrown LE, Learning, et al. LLC reducing infant mortality rates using the perinatal Eect on blood pressure of potassium, calcium, and NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION periods of risk model. Public Health Nurs. 22(1):2-7. magnesium in women with low habitual intake. 48. Kramer MS, Platt RW, Yang H, Haglund B, et al. Hypertension. 1998;31:131. Registration artifacts in international comparisons 63. Beevers DG. e epidemiology of salt and hy- of infant mortality. Paediatr Perinat Epidemiol. pertension. Clin Auton Res. 2002;12(5):353-357. © Jones2002;16(1):16-22. & Bartlett Learning, LLC 64. Law© Jones MR, Frost & CD,Bartlett Wald NJ. Learning, Analysis of dataLLC from 49.NOT Silva FOR AA, SALE Barbieri OR MA, DISTRIBUTION Gomes UA. Trends in low trialsNOT of FORsalt reduction. SALE BrOR Med DISTRIBUTION J. 1991;302:819-824. birth weight: a comparison of two birth cohorts 65. Haddy FJ, Vanhoutte PM, Feletou M. Role of separated by a 15-year interval in Ribeirao Preto, potassium in regulating blood ow and blood Brazil. Bull World Health Organ. 1998;76(1):73-84. pressure. Am J Physiol Regul Integr Comp Physiol. 50. Wakeford R, McElvenny, D. From epidemio- 2006;290:R548-R552. © Jones & Bartlettlogical Learning, association LLC to causation. Occup Med.© Jones 66. & BartlettLawrence Learning, J. Lifestyle modication LLC as a means to NOT FOR SALE OR2007;57(7):464-465. DISTRIBUTION NOT FOR SALEprevent OR and DISTRIBUTION treat high blood pressure. J Am Soc 51. McDonald JC, McDonald AD, Hughes JM, Rando RJ, Nephrol. 2003;14:S99-S102. et al. Mortality from lung and kidney disease in a 67. Persky VP. e eect of dietary sucrose of blood cohort of North American industrial sand workers: pressure in spontaneously hypertensive rats. Nutr an update. Ann© Jones Occup Hyg. & Bartlett 2005;49(5):367-373. Learning, LLC Res. 1986;6(9):1111-1115.© Jones & Bartlett Learning, LLC 52. Alexander GL.NOT Dietary FOR sodium SALE intake OR and DISTRIBUTION its rela- 68. McAleese JD, RankinNOT LL. FOR Garden-based SALE ORnutrition DISTRIBUTION tion to human health: a summary of the evidence. education aects fruit and vegetable consump- J Am Coll Nutr. 2006;25(3):165-169. tion in sixth-grade adolescents. J Am Diet Assoc. 53. Barry MP. Urbanization lifestyle changes and the 2007;107(4):662–665. nutrition transition. World Dev. 1999;27:1905-1916. 69. Karanja N. Descriptive characteristics of the dietary 54.© JonesPryer &J, CappuccioBartlett Learning,FP, Elliott P. DietaryLLC calcium patterns© Jones used & in Bartlett the Dietary Learning, Approaches LLC to Stop NOTand FOR blood SALE pressure: OR aDISTRIBUTION review of the observational HypertensionNOT FOR Trial.SALE DASH OR Collaborative DISTRIBUTION Research studies. J Hum Hypertens. 1995;9(8):597-604. Group. J Am Diet Assoc. 1999;99(8):S19-27. 55. Troyer JL, Racine EF, Ngugi GW, McAuley WJ. 70. Robinson JR, Young TK, Roos LL, Gelskey DE. e eect of home-delivered Dietary Approach Estimating the burden of disease. Med Care. © Jones & Bartlettto StopLearning, Hypertension LLC (DASH) meals on the diets© Jones & Bartlett1997;35:932-947. Learning, LLC NOT FOR SALE ORof older DISTRIBUTION adults with cardiovascular disease. NOTAm FOR71. SALEGeng OR Z, Vasanti DISTRIBUTION SM, Shuba K. Substituting J Clin Nutr. 2010;91(5):1204-1212. brown rice for white rice to lower diabetes risk:

9781284348286_CH03_083_104.indd 103 08/06/17 7:36 PM 104 Chapter 3 Nutritional Epidemiology and Research Methods

© Jones & Bartletta focus-group Learning, study in LLCChinese adults. J Am Diet © Jones86. Hill & AB. Bartlett e environment Learning, and disease:LLC association or Assoc. 2010;110(8):1216-1221. causation? Proc Roy Soc Med. 1965;65(58):295–300. NOT FOR SALE OR DISTRIBUTION NOT FOR72. SALEMunhall OR PL, DISTRIBUTION Boyd, CO. Nursing Research: A Qual- 87. Trichopoulou A, Vasilopoulou E. Mediterranean itative Perspective. New York, NY: National League diet and longevity. Br J Nutr. 2000;84(2):S205-209. for Nursing; 1993. 88. De Groot LC, Staveren WA, Burema J. Survival 73. Sullivan MB, Camer L. Benets, barriers and beyond age 70 in relation to diet. Nutr Rev. 1996; enablers of university-community© Jones & Bartlett research: Learning, input LLC54:211-212. © Jones & Bartlett Learning, LLC from health and socialNOT service FOR agencies. SALE ORSoc Mark DISTRIBUTION 89. Keys A. Seven Countries.NOT A Multivariate FOR SALE Analysis OR DISTRIBUTION Q. 2000;4:9-25. of Death and Coronary Heart Disease. Cambridge, 74. National Institute of Medicine. e Future of Public MA: Harvard University Press; 1980. Health. Washington, DC: National Academy 90. De Lorgeril M, Renaud S, Mamelle N, Salen P, et al. Press;© Jones 1998. & Bartlett Learning, LLC Mediterranean© Jones alpha-linolenic & Bartlett acid-richLearning, diet LLC in 75. Vega WA. eoretical and pragmatic implications secondary prevention of coronary heart disease. NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION of cultural diversity for community research. Am Lancet. 1994;343:1454-1459. J Community Psychol. 1992;20:375-391. 91. Singh RB, Dubnov G, Niaz MA. Eect of an 76. McGuirt JT, Jilcott SB, Keyserlingm TC. Conducting Indo-Mediterranean diet on progression of cor- community audits to evaluate community resources onary artery disease in high risk patients (Indo- © Jones & Bartlettfor healthful Learning, lifestyle LLCbehaviors: an illustration © JonesMediterranean & Bartlett Learning,diet heart study): LLC a randomized NOT FOR SALEfrom ORrural DISTRIBUTION eastern north carolina. Prev Chronic NOT FORsingle-blind SALE ORtrial. DISTRIBUTION Lancet. 2002;360:1455-1461. Dis. 2011;8(6):A149-A160. 92. Osler M, Schroll M. Diet and mortality in a cohort 77. Valdés Hernández MC, Morris Z, Dickie DA, Royle of elderly people in a north European community. NA, et al. Close correlation between quantitative Int J Epidemiol. 1997;26:155-159. and qualitative assessments© Jones of white& Bartlett matter lesions. Learning, 93. LLC Lasheras C, Fernandez S,© Patterson Jones AM. & Bartlett Mediter- Learning, LLC Neuroepidemiology. 2013;40:13-22. ranean diet and age with respect to overall survival NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 78. Research HUPS. Questionnaire design tip sheet. http:// in institutionalised, nonsmoking elderly people. psr.iq.harvard.edu/les/psr/les/PSRQuestionnaire Am J Clin Nutr. 2000;71(4):987-992. TipSheet_0.pdf?m=1357530492. Accessed March 10, 94. Knoops KT, de Groot LC, Kromhout D. Mediter- 2016. ranean diet, lifestyle factors, and 10-year mortality 79. Agriculture© Jones and& Bartlett Consumer Learning, Protection. LLC e steps in elderly© EuropeanJones & men Bartlett and women: Learning, the HALE LLC precedingNOT FOR questionnaire SALE OR design. DISTRIBUTION http://www.fao. project.NOT JAMA. FOR 2004;292:1433-1439. SALE OR DISTRIBUTION org/docrep/w3241e/w3241e05.htm. Accessed 95. Slimani N, Kaaks R, Ferrari P. European pro- March 10, 2016. spective investigation into cancer and nutrition 80. Malone TW, Crowston K. Toward an Interdis- (EPIC) calibration study: rationale, design and ciplinary Study of Coordination. Cambridge, MA: population characteristics. Public Health Nutr. © Jones & BartlettCenter for Learning, Coordination LLC Science, MIT; 1991. © Jones2002;5:1125-1145. & Bartlett Learning, LLC NOT FOR81. SALEAmuwo OR S, JenkinsDISTRIBUTION E. True partnership evolves over NOT96. FORDeharveng SALE G,OR Charrondiere DISTRIBUTION UR. Comparison of time. In: Sullivan M, James K, eds. Collaborative nutrients in the food composition tables available Research: University and Community Partnership. in the nine European countries participating Washington, DC: American Public Health Asso- in EPIC. European prospective investigation ciation; 2001:25-43.© Jones & Bartlett Learning, LLCinto cancer and nutrition.© Jones Eur J& Clin Bartlett Nutr. Learning, LLC 82. Clark MJ. NursingNOT in the FOR Community. SALE OR 3rd DISTRIBUTION ed. 1999;53:60-79. NOT FOR SALE OR DISTRIBUTION Stamford, CT: Appleton & Lange; 1999. 97. Kaaks R, Riboli E. Validation and calibration of 83. Hinke M, Maurits W, Seidell C, ijs A, et al. dietary intake measurements in the EPIC project: Eectiveness and cost-eectiveness of early screen- methodological considerations. European pro- ing and treatment of malnourished patients. Am spective investigation into cancer and nutrition. J© Clin Jones Nutr. 2005;82:1082-1089.& Bartlett Learning, LLC Int J Epidemiol.© Jones 1997;26(Suppl & Bartlett 1):S15-25.Learning, LLC 84. MelloraNOT FOR JM. Exploring SALE OR the relationshipsDISTRIBUTION between 98. Chen HL,NOT Pagoto FOR SL, SALE Barton BA,OR Lapane DISTRIBUTION KL, et al. income inequality, socioeconomic status and health: Personal and parental weight misperception and a self-guided tour? Int J Epidemiol. 2002;31:685-687. self-reported attempted weight loss in US children 85. Frank-Spohrer G. Community Nutrition: Applying and adolescents, National Health and Nutrition © Jones & BartlettEpidemiology Learning, to Contemporary LLC Practice. 2nd ed. © JonesExamination & Bartlett Survey, Learning, 2007–2008 LLC and 2009–2010. NOT FOR SALESudbury, OR MA: DISTRIBUTION Jones and Bartlet; 2008. NOT FORPrev SALE Chronic OR Dis. DISTRIBUTION2014;11:140123-1401143.

9781284348286_CH03_083_104.indd 104 08/06/17 7:36 PM