Psychological Trauma: Theory, Research, Practice, and Policy © 2009 American Psychological Association 2009, Vol. 1, No. 1, 35–48 1942-9681/09/$12.00 DOI: 10.1037/a0015128 and Physical : A Psychoneuroimmunology Approach to Etiology of Negative Health Effects and Possible Interventions

Kathleen Kendall-Tackett Texas Tech University School of Medicine

People who have experienced traumatic events have higher rates than the general population of a wide range of serious and life-threatening illnesses including cardio- vascular disease, diabetes, gastrointestinal disorders, and cancer. An important ques- tion, for both researchers and clinicians, is why this occurs. Researchers have discov- ered that traumatic events dysregulate the hypothalamic-pituitary-adrenal axis and sympathetic nervous system. More recently, research from the field of psychoneuro- (PNI) suggests that traumatic life events can lead to health problems through dysregulation of another key system: the inflammatory response. Prior trauma “primes” the inflammatory response system so that it reacts more rapidly to subsequent life stressors. Elevated inflammation has an etiologic role in many chronic illnesses. Recent PNI studies also suggest some interventions that can serve as adjuncts to traditional trauma treatment. These treatments include long-chain omega-3 fatty acids, exercise, and sleep interventions. Each of these interventions downregulates inflam- mation, which will likely halt the progression to chronic disease for some trauma survivors.

Keywords: trauma, inflammation, depression, PTSD, omega-3s

For more than a decade, researchers have Trauma Increases the Risk of Health explored the impact of traumatic events on Problems physical health. Trauma survivors often have significant physical health problems that can The Adverse Childhood Experiences (ACE) linger for many years after the traumatic event study, with more than 17,000 participants from has ended and are often above and beyond the a large California health maintenance organiza- effects of direct physical injury (Kendall- tion (HMO), was the first large-scale study to Tackett, 2007; Kibler, Joshi, & Hughes, in demonstrate the link between childhood trauma press). The question of trauma’s impact and and organic health conditions. Felitti and col- health is at the junction of mind-body medicine. leagues (1998) found that patients who experi- It has important implications for not only health enced four or more types of adverse childhood care costs, but also for trauma practice, and how events had higher rates of ischemic heart dis- we can best meet the needs of people who have ease, cancer, stroke, chronic bronchitis, emphy- experienced trauma. In this article, I review sema, diabetes, skeletal fractures, and hepatitis research on psychological trauma and physical than their nontraumatized counterparts. The health using a psychoneuroimmunology (PNI) childhood events they studied included all types approach. PNI research can describe some of of maltreatment including psychological, phys- the possible mechanisms by which trauma af- ical, or contact sexual abuse and exposure to fects health. It also suggests some viable addi- parental substance abuse, mental illness, inti- tions to traditional trauma treatments that will mate partner violence, or criminal behavior. address the physical health sequelae. Since publication of that study, others have followed with similar findings. In the National Comorbidity Study, women who were mal- Correspondence concerning this article should be ad- treated as children had a ninefold increase in dressed to Kathleen Kendall-Tackett, Department of Pedi- atrics, Texas Tech University School of Medicine, 1400 N. cardiovascular disease compared with nonmal- Coulter Street, Rm 4301, Amarillo, TX 79106. E-mail: treated women (Batten, Aslan, Maciejewski, & [email protected] Mazure, 2004). In a sample of primary care

35 36 KENDALL-TACKETT patients, those who experienced childhood Leserman, and colleagues (Drossman, Li, abuse or partner violence in adolescence or Leserman, Toomey, & Hu, 1996; Leserman et adulthood reported twice as many symptoms on al., 1996) found that 60% of women in treat- a review of systems than their age-matched, ment for functional gastrointestinal (GI) illness nonabused counterparts. They were also more (such as irritable bowel syndrome) had a history likely to abuse substances and report a wide of abuse, as children or adults. In another study variety of chronic pain syndromes (Kendall- (Drossman, Leserman, Li, Keefe, Hu, & Tackett, Marshall, & Ness, 2000, 2003). Toomey, 2000), half of the women referred to a Patients with diagnoses of posttraumatic GI treatment center had been physically or sex- disorder (PTSD) report similar symp- ually abused. Fourteen percent reported severe toms. Data from the Canadian Community abuse. Patients who had been more severely ϭ Health Survey (N 36,984) indicated that par- abused had poorer health status relative to the ticipants with PTSD had significantly higher other patients. rates of cardiovascular disease, respiratory dis- Sachs-Ericsson, Kendall-Tackett, and Her- eases, chronic pain syndromes, gastrointestinal nandez (2007), using data from the National illnesses, and cancer (Sareen, Cox, Stein, Afifi, Comorbidity Study, noted that patients with a Fleet, & Asmundson, 2007). PTSD was also history of childhood or partner abuse reported strongly associated with chronic fatigue syn- more pain when describing their current health drome and multiple-chemical sensitivity, but symptoms. Van Houdenhove, Luyten, and Egle not diabetes. PTSD following a human-made disaster (2009) found that 64% of patients who had showed similar health effects (N ϭ 896; either fibromyalgia or chronic fatigue syndrome Dirkzwager, van der Velden, Grievink, & had experienced at least one type of either child Yzermans, 2007). In this study, PTSD was as- or adult trauma. More concerning was that 39% sociated with new vascular events as well as had experienced both, indicating a lifelong pat- physician-reported vascular, musculoskeletal, tern of revictimization and abuse. and dermatological problems. These problems appeared even after controlling for previous Possible Mechanisms: The health problem, smoking, and demographic characteristics. Psychoneuroimmunology of Trauma Not surprisingly, people with PTSD use more The negative health effects of trauma and health care services than those without PTSD. PTSD, at this point, are fairly well established. In a study of women seeking health care at A wide range of traumatic exposures and expe- Veterans’ Administration facilities (Dobie et riences can lead to poor health. Some of this is al., 2006), women with PTSD had more outpa- attributable directly to injury suffered during tient visits to the emergency department, pri- mary care, medical or surgery subspecialities, the trauma. Others result from the effects of the ancillary services, and diagnostic tests. They trauma. The question then becomes why does had higher rates of hospitalizations and surgical this occur? It’s an important question to ask procedures. Women with PTSD were signifi- because it has implications for not only treating cantly more likely to have a service-related dis- trauma survivors currently seeking care, but ability, have chronic pain, and to be obese. They also for prevention of these health problems in were also more likely to smoke and abuse alco- the future. hol. Seventy-five percent of the women with Research in the field of PNI is a useful PTSD also screened positive for depression. framework for understanding these effects. According to PNI research, many of these Trauma and Chronic Pain illnesses are because severe or overwhelming stress, and any resultant PTSD, alters and Researchers studying chronic pain syn- dysregulates the key systems that are part of dromes were among the first to note an associ- the stress response. To understand these find- ation between trauma and health (Gross, Doerr, ings, it’s helpful to review the three compo- Caldirola, Guzinski, & Ripley, 1980; Harrop- nents of the stress system. These are de- Griffiths et al., 1988). For example, Drossman, scribed in the following section. PSYCHOLOGICAL TRAUMA AND PHYSICAL HEALTH 37

How Humans Respond to a Perceived interleukin 1-␤ (IL-1␤), interleukin-6 (IL-6), Threat -␣ (TNF-␣), and inter- feron-␥ (IFN-␥). These findings are an impor- Human bodies have a number of interdepen- tant part of the picture when studying the health dent mechanisms in place designed to preserve effects of traumatic events and are the focus of life in the face of danger: , hypo- this article. thalamic-pituitary-adrenal (HPA) axis, and im- The human stress response has a number of mune response. In response to threat, the sym- checks and balances built in to ensure that var- pathetic nervous system responds by releasing ious components do not become overactive. Un- the , epineph- fortunately, in the case of severe or overwhelm- rine, and . This is the fight-or-flight ing stress, the normal checks and balances fail, response. The HPA axis responds with a chem- causing inflammation levels to be abnormally ical cascade: the releases cortico- high. For example, (which is normally trophin-releasing (CRH), which antiinflammatory and keeps proinflammatory causes the pituitary to release adrenocortico- in check) can actually change func- tropin hormone (ACTH), which causes the ad- tion under severe stress and potentiate the ac- renal cortex to release cortisol, a . tions of IL-1 and IL-6 rather than inhibiting There have been numerous studies over the them (Dhabhar & McEwen, 2001). When there past 15 years that have examined changes to the are too many proinflammatory cytokines, or HPA axis and sympathetic nervous system that other inflammatory factors, humans become occur in the wake of traumatic events and/or vulnerable to disease (McEwen, 2003). PTSD (see Bremner, 2005; Kendall-Tackett, 2000, for reviews). Wilson, Friedman, and Inflammation in Trauma Survivors Lindy (2004) used McEwen’s (1998) model of allostasis to explain the effects of PTSD on Although a relatively new area of study, sev- these systems and subsequent health problems. eral researchers using a psychoneuroimmunol- Allostasis refers to the body’s attempt to main- ogy framework have found that traumatic tain stability through stress and change. When events increase levels of proinflammatory cyto- the body is overwhelmed by stressors, allostatic kines in trauma survivors. Elevated levels of load (or body “wear and tear”) occurs. Allo- proinflammatory cytokines or other inflamma- static load can threaten health if severe or tory markers (e.g., C-reactive protein, fibrino- chronic. Traumatic events can increase allo- gen) are associated with increased risk of a static load by chronically activating the HPA number of serious health problems including axis and sympathetic nervous system. This dys- coronary heart disease, myocardial infarction, regulation can be manifested with elevated or chronic pain syndromes, premature aging, im- abnormally low levels of cortisol, and low or paired immune function, impaired wound heal- elevated levels of norepinephrine (Kendall- ing, and Alzheimer’s disease (Frasure-Smith & Tackett, 2000). But the sympathetic nervous Lesperance, 2005; Kiecolt-Glaser et al., 2005; system and HPA axis are not the only systems Nivison, Guillozet-Bongaarts, & Montine, in involved. press; Pace, Hu, & Miller, 2007; Surtees et al., The third part of this stress system is the 2008; Wilson et al., 2002). immune response. The immune system re- Previous child maltreatment was related to sponds to threat by releasing proinflammatory high C-reactive protein levels when they were cytokines. These cytokines increase inflamma- assessed 20 years later (Danese, Pariante, Caspi, tion and serve the adaptive purpose of helping Taylor, & Poulton, 2007). The participants the body heal wounds and fight . Out- (N ϭ 1,037) were part of the Dunedin Multi- side of the trauma field, there is a large and disciplinary Health and Development Study, a growing literature on the impact of these proin- study of health behavior in a complete birth flammatory cytokines on health. A key finding cohort in New Zealand. Researchers assessed of PNI research is that both physical and psy- the participants every two to three years chological stress can trigger the inflammatory throughout childhood, and every five to six response (Robles, Glaser, & Kiecolt-Glaser, years through age 32. The impact of child mal- 2005). The cytokines studied most often are treatment on inflammation was independent of 38 KENDALL-TACKETT other factors that could have accounted for the the trauma survivor does not meet full criteria findings, such co-occurring life stresses in for PTSD. adulthood, early life risks, or adult health or health behavior. The effect was also dose- How Depression Influences Health responsive; the more severe the abuse, the more severe the inflammation. Depression is one of the most common of all A similar finding resulted from a study of mood disorders and is a common symptom women abused by intimate partners. In this among trauma survivors. It also has a well- study, 62 women who had had abusive part- documented negative impact on health—even ners 8 to 11 years previously had significantly beyond the effects of PTSD. In a large study of higher interferon-gamma (IFN-␥) levels than U.S. veterans (N ϭ 35,715), depressed patients nonabused women (Woods et al., 2005). The were at increased risk of dying over a 5-year women also had high rates of depression (52%) period (Kinder et al., 2008). Patients with PTSD and PTSD (39%). Even several years after their alone, or who had neither condition, did not abuse had ended, these women were still man- have this increased risk. ifesting significant physical symptomatology. Most of the research on depression and health Inflammation was also found to be elevated in a has examined the link between depression and study of 15 women who had been raped 24 to 72 cardiovascular disease. Depression is a robust hours after their assault, compared with 16 risk factor for cardiovascular disease, cardio- women who had not been sexually assaulted. vascular events, and cardiac-related mortality Sexually assaulted women had higher ACTH, (Frasure-Smith & Lesperance, 2005; Rutledge C-reactive protein, IL-6, IL-10, IFN-␥ than et al., 2006). Inflammation is a crucial factor in women in the nonassaulted group (Groe¨r, atherogenesis and the progression of coronary Thomas, Evans, Helton, & Weldon, 2006). artery disease (Zouridakis, Avanzas, Arroyo- In a sample of 14 otherwise healthy patients Espliguero, Fredericks, & Kaski, 2004). Inflam- with PTSD and a group of 14 patients without matory factors are not simply markers. They PTSD matched on age and gender of 14 patients have a pathogenic role to play. And these in- without PTSD, von Kanel and colleagues flammatory factors, including C-reactive pro- (2006) investigated PTSD and blood coagula- tein and proinflammatory cytokines, are often tion by measuring clotting factors, fibrinogen, elevated in depressed people. and D-dimer in the plasma. They found the Depression, inflammation, and heart disease. more severe the PTSD, the higher the levels of Several studies have found elevated C-reactive fibrinogen and the clotting factor FVIII:C. They protein in depressed patients with heart disease. concluded that PTSD may elicit hypercoagula- For example, in a study of men and women at bility, even at subthreshold levels, and this may risk for heart disease (N ϭ 68), Taylor and increase risk for cardiovascular disease in colleagues (2006) found that depressed patients trauma survivors. Fibrinogen is a soluble pro- (mean age ϭ 62) had elevated C-reactive pro- tein that aids in clotting. Because it increases tein compared to matched controls. Surtees and the speed of platelet aggregation and thrombus colleagues (2008), in a sample of patients with formation, a high level of fibrinogen is a risk ischemic heart disease, found that major depres- factor for cardiovascular disease. sion in the previous year increased C-reactive protein levels. Those with higher C-reactive Emotional and Cognitive Factors protein had an odds ratio of 2.02 for incident ischemic heart disease. This was true even when The above-cited studies demonstrate that adjusting for other known cardiac risk factors trauma and PTSD may directly alter the stress- including cigarette smoking, diabetes, systolic response systems of trauma survivors. While , body mass index, and choles- these findings are interesting, they do not fully terol. explain the health effects of trauma. Addition- Coaguability may also be a factor in depres- ally relevant is the fact that psychological sion’s impact on heart disease. A study of pa- factors, such as depression and hostility, can tients 65 and older found that depression was trigger the same inflammatory response. Fur- related to both C-reactive protein and coagula- thermore, these effects can impair health even if tion factors (Kop et al., 2002). This study in- PSYCHOLOGICAL TRAUMA AND PHYSICAL HEALTH 39 cluded 4,268 patients who were free of cardio- come hypervigilant to rejection in social rela- vascular disease (mean age ϭ 72.4 years). tionships, and this world view has discernable, Depressed men and women had elevated C- physical sequelae (Kiecolt-Glaser & Newton, reactive protein, elevated white blood cell 2001). In their review, Smith and Ruiz (2002) count, and increased markers of coaguability. noted that people who are high in trait hostility Similarly, coagulation was also related to de- are more prone to ischemia and constriction of pression and cardiovascular risk in a study the coronary arteries during mental stress. Trait of 3,292 perimenopausal women (Matthews et hostility predicts new coronary events in previ- al., 2007). Over the five-year period, depressive ously healthy people. For patients who already symptoms were related to two indices of coag- have coronary heart disease, hostility speeds up ulation: fibrinogen and PAI-1. the progression of the disease. Depression predicted cardiac symptoms in a Unfortunately, hostility is common among sample of 750 women middle-aged women with trauma survivors. In a sample from primary suspected myocardial ischemia (Rutledge et al., care, 52% of female sexual abuse survivors 2006). The more severe the depression, the indicated that they could not trust others com- higher the mortality risk over the 2.3 year study pared with 17% of the nonabused women period. Each point increase on the Beck Depres- (Hulme, 2000). In a community sample, ap- sion Inventory was associated with a 3% in- proximately half of sexual abuse survivors in- crease in mortality risk even after adjusting for dicated that their views of themselves and oth- age, disease severity, and other coronary artery ers were very negative (Teegen, 1999). And in disease risk factors. a sample of 90 women veterans (Butterfield, Depression not only increases inflammation. Forneris, Feldman, & Beckham, 2000), women It also increases the incidence of health- with PTSD reported significantly higher levels compromising behavior (Kibler et al., in press). of hostility and had poorer overall health status In a study of 254 patients with major mood than women without PTSD. disorders (Lu, Mueser, Rosenberg, & Like depression, hostility also triggers the Jankowski, 2008), researchers found that expo- inflammatory response system. In a study of 44 sure to adverse childhood experiences was re- healthy, nonsmoking, premenopausal women, lated to a number of health compromising fac- hostility increased levels of proinflammatory tors. Those with increased exposure to ACEs ␣ ␤ ␣ had higher rates of risk-taking behaviors, diag- cytokines (IL-1 , IL-1 , IL-8 and TNF- ; noses of substance use disorders, trauma in Suarez et al., 2004). The combination of depres- adulthood, PTSD, suicide attempts, medical ser- sion and hostility was especially deleterious. vice utilization (including younger age at first The effect was dose-responsive: the more se- hospitalization), and homelessness. Thus, even vere the depression and hostility, the greater the within a population with severe mood disorders, production of cytokines. A study with men had adverse childhood experiences contribute to similar results (Suarez, 2003). Its author noted worse physical and mental health. Studies of that increased levels of IL-6 predicted both fu- other types of trauma have showed similar pat- ture risk of cardiac events and all-cause mortal- terns (Kibler et al., in press). ity, and hypothesized that IL-6 may mediate the relationship between hostility and these health Hostility problems. Kiecolt-Glaser et al. (2005) found that cou- Depression is not the only cognitive factor ples who were high in hostility had higher levels that can increase risk of death and disease. of circulating proinflammatory cytokines. As a Hostility has a similar effect. For people with a result, the rate of wound healing for the high- hostile world view, life is not benign. People hostility couples was 60% slower than low- high in trait hostility do not trust others, are hostility couples. High-hostility couples had suspicious and cynical about human nature, and fewer cytokines at the wound site, where they tend to interpret the actions of others as aggres- were supposed to be, and high levels circulating sive (Smith, 1992). This psychological state systemically, where they were more likely to also triggers the inflammatory response. Indi- impair health and increase the risk of age- viduals who expect the worst from people be- related diseases. 40 KENDALL-TACKETT

In his study of 135 healthy patients with no In a sample of battered women living in shel- symptoms of diabetes, Suarez (2006) found that ters (N ϭ 50), 70% reported poor sleep quality, women with higher levels of depression and 28% went to bed very fatigued, and 40% woke hostility had higher levels of fasting insulin, up feeling very fatigued (Humphreys, Lee, Ney- glucose, and insulin resistance. These findings lan, & Marmar, 1999). Moreover, 82% de- did not hold true for men and they were inde- scribed one or more of the following character- pendent of other risk factors for metabolic syn- istics of disturbed sleep: many wakings over the drome including BMI, age, fasting triglycerides, course of the night, restless sleep, and early exercise regularity, or ethnicity. A study morning waking. Six described vivid night- of 6,814 healthy men and women found that mares that included recent incidents of abuse. participants with higher levels of cynical dis- In a study of sleep disorders in sexual assault trust, or depression had higher survivors, 80% had either sleep-breathing or levels of C-reactive protein, IL-6, and fibrino- sleep-movement disorders. Both of these disor- gen (Ranjit et al., 2007)–all risk factors for heart ders were linked to higher levels of depression disease. and suicidality. Women who had both types of sleep disorders had the most severe symptoms. Sleep Disorders in PTSD For these women, fragmented sleep appeared to potentiate the symptoms they experienced after Cognitive factors are not the only triggers for their assault, making coping with the aftermath the inflammatory response system. Poor sleep is even more difficult (Krakow, Artar, et al., also associated with increased inflammation, in- 2000). cluding C-reactive protein, IL-6, IL-1, and TNF-␣ (Suarez & Goforth, in press). Unfortu- How Sleep Impacts Health nately, sleep disorders are also quite common among trauma survivors, especially those with Sleep problems impact health in a number of PTSD, and could be another way that trauma negative ways. They increase the risk of coro- impacts health. nary heart disease, Type-2 diabetes, and hyper- A number of studies have documented dis- tension. Sleep loss reduces count turbed sleep patterns in men and women who and the activity of the Natural Killer cells, mak- have experienced violence and resultant acute ing patients more vulnerable to infection. The stress disorder (ASD) or PTSD. For example, in link between sleep disturbances and inflamma- a European community sample, 68% of sexual tion appears to be bidirectional—with poor abuse survivors reported having sleep difficul- sleep increasing inflammation, and inflamma- ties, with 45% having repetitive nightmares tion decreasing the quality of sleep (Suarez & (Teegen, 1999). In a French sample, 33% of Goforth, in press). teens who had been raped reported poor quality The term “sleep disorders” cover a wide sleep compared with 16% of the nonassaulted range of problems. Some of these disorders can control group. Twenty-eight percent of the as- be measured via patient questionnaire. Other saulted teens had nightmares (compared with sleep difficulties need to be measured via poly- 11% of nonassaulted teens), and 56% woke somnographic studies. These are necessary to during the night (compared with 21%; Choquet, detect common problems, such as sleep- Darves-Bornoz, Ledoux, Manfredi, & Hassler, disordered breathing (e.g., sleep apnea), and 1997). sleep-movement disorders (e.g., restless leg Hulme (2000) found that sleep problems syndrome). Patients may not be aware of these were common among women primary-care pa- problems, and therefore, would not be able to tients who were sexual abuse survivors. Fifty- report them. Regardless of type of disorder, two percent of sexual abuse survivors reported poor sleep quality has a profound effect on that they could not sleep at night (compared health. It compromises immune, metabolic, and with 24% of the nonabused group), and 36% neuroendocrine function, chronically activates reported nightmares (compared with 13%). The the HPA axis, and even increases mortality risk most common symptom was intrusive thoughts, (Carmichael & Reis, 2005). reported by 53% of sexual abuse survivors Sleep may also be compromised by quality of (compared with 18% of the nonabused group). relationship with adult partner. This is of inter- PSYCHOLOGICAL TRAUMA AND PHYSICAL HEALTH 41 est because several studies have found trauma could be a health risk factor for women who are survivors have higher rates of unstable relation- not in stable, secure relationships. The deleteri- ships and are more likely to express dissatisfac- ous health effects of disturbed sleep have also tion with their current partners than the general been found in samples of trauma survivors. In a population (Kendall-Tackett, 2003). Two recent sample of female rape survivors with PTSD, studies have examined the relationship between trauma-related sleep disorders had an indepen- security of adult attachments and sleep quality. dent impact on health, even after controlling for Sleep is a physiologically vulnerable state. In both depression and PTSD (Clum, Nishith, & order to sleep soundly, people must feel suffi- Resick, 2001). ciently secure so they can downregulate vigi- In summary, sleep disturbance is another lance and alertness. And to do so, one must be common aftereffect of that trauma can nega- secure in social relationships (Troxel, Cyr- tively impact health. Unfortunately, sleep qual- anowski, Hall, Frank, & Buysee, 2007) or feel ity and disturbance are often not included in a relatively safe. normal trauma assessment, as they are also not In the first study, 78 married adults com- in most other psychological assessments. But pleted questionnaires about their sleep quality, quality of current partnership (secure vs. inse- given its large impact on health, it should be cure attachment), and depression (Carmichael assessed and treated before it can lead health & Reis, 2005). The sleep questionnaire asked problems. about seven aspects of sleep including per- ceived sleep quality, sleep latency (time to get What Trauma Survivors Can Do to to sleep), sleep duration, habitual sleep effi- ciency, sleep disturbances, and use of sleep Improve Their Health medications. Using structural equation model- Upon reviewing the literature on trauma ing, the researchers found that married partici- and health, many practitioners (and some sur- pants who were anxious about their current re- vivors) become concerned that nothing can be lationships reported poorer sleep quality, even done to reverse these effects; that nothing can after controlling for depression. Women with insecure attachments were concerned that their be done to halt their inevitable decline into partners were emotionally unavailable and not poor health. Fortunately, that is not the case. trustworthy. The researchers indicated that one In addition to standard treatments for PTSD, limitation to their study was that they used a PNI research suggests some other interven- self-report measure of sleep, rather than assess- tions that may help counter some of the phys- ing sleep directly. iological changes that occur in the wake of Troxel and colleagues (2007) addressed that traumatic events, in turn, lowering the risk of limitation by using polysomnographic studies to health problems. assess sleep directly. In a study of 107 women with recurrent major depression, marital status and security of that relationship predicted qual- Downregulate the Inflammatory ity and efficiency of sleep. If women had anx- Response System ious attachments, particularly if they were sep- arated or widowed, they had a significantly Kiecolt-Glaser and colleagues (2007) noted smaller percentage of stage 3–4 sleep than that prior trauma “primes” the inflammatory women who were currently partnered and who response system so that there is heightened and had secure attachments. The authors noted the more rapid rise in inflammation in response to importance of stage 3–4 sleep in protecting stress. One key to improving the health of individuals from cardiovascular and metabolic trauma survivors is downregulating this stress diseases. response and increasing resilience to stress. Re- McEwen (2003) reported that even short pe- search suggests two body-related adjuncts to riods of sleep deprivation can elevate cortisol traditional trauma treatment that specifically and glucose levels, and can increase both insu- downregulate the inflammatory response sys- lin and insulin resistance. Long-term sleep de- tem: Omega-3 fatty acids and exercise. These privation can seriously impair health, and this studies are summarized below. 42 KENDALL-TACKETT

Omega-3 Fatty Acids and the Stress Some parallel work has been done with re- Response gard to Type 2 diabetes. In a review, Delarue and colleagues (2004) found that in healthy Researchers have examined the impact of patients, fish oil reduced insulin resistance and EPA and DHA, the long-chain omega-3 fatty plasma triglycerides and increased resilience to acids, on stress. Several studies have suggested stress by decreasing the activity of the sympa- that EPA and DHA may have an adaptogenic thetic nervous system. Fish oil did not reduce role in stress by regulating and attenuating the insulin resistance in patients with diabetes, but stress response. For example, when college stu- did lower triglycerides. The authors concluded dents had a high ratio of omega-6s to omega-3s that fish oil showed promise in the prevention of (meaning low levels of omega-3s), they had insulin resistance and related health problems. more inflammation when exposed to a lab- Given that trauma survivors also appear to be at induced stressor. In contrast, students with increased risk of developing diabetes, the im- higher levels of EPA/DHA had a lower inflam- pact of EPA and DHA on insulin resistance matory response to stress (Maes, Christophe, could be another way that they positively im- Bosmans, Lin, & Neels, 2000). In other words, pact health. students were not deficient in EPA/DHA had an Sources of EPA and DHA. The chief di- attenuated inflammatory response. etary source of EPA and DHA is fish. Accord- Similarly, Kiecolt-Glaser and colleagues ing to population studies, people must eat 1 (2007) found that EPA/DHA supplementation to 1.5 pounds of fish per week to achieve the downregulated the stress response and its mental health effects (Noaghiul & Hibbeln, maladaptive impact on mood. In their study 2003). However, eating that much fish may not of 43 older adults, they noted that “even mod- be safe because of contaminants in seafood— est supplementation with n-3 PUFAs [Ome- and not everyone likes fish. Fish oil is the best ga-3 polyunsaturated fatty acids] reduces source of EPA and DHA in supplements. For- plasma norepinephrine, an important link to tunately, finding safe sources of fish oil is now the stress response” (p. 221). Patients who relatively easy. The U.S. Pharmacoepia is an were both depressed and had low levels of independent, not-for-profit organization that omega-3s had the highest levels of proinflam- tests fish oil for contaminants and lists specific matory cytokines. The authors noted that a brands that they verify on their Web site (ww- diet that is low in EPA/DHA increases the w.USP.org). The minimum dosage of DHA for risk of both depression and other diseases possible prevention of depression is 200–400 related to chronic inflammation. mg. The effective dosage for treatment of de- EPA and DHA also specifically lower proin- pression and other affective disorders is 1,000 flammatory cytokines, which helps with depres- mg of EPA. The FDA GRAS (generally recog- sion and health problems associated with it. In a nized as safe) levels are 3000 mg of EPA/DHA large population study (N ϭ 1,123), high levels together, and 1500 mg of DHA. of EPA and DHA were related to lower levels of proinflammatory cytokines (IL-1␣, IL-1␤, IL-6, Exercise and the Inflammatory Response and TNF-␣) and higher levels of anti-inflamma- tory cytokines, such as IL-10. For people with Exercise is another intervention with stress- low levels of EPA and DHA, the opposite was lowering capabilities. Two clinical trials have true: they had high levels of proinflammatory found exercise as effective as sertraline (Zoloft) cytokines and low levels of anti-inflammatory for the treatment of even major depression cytokines (Ferrucci et al., 2006). This study had (Babyak et al., 2000; Blumenthal et al., 2007). several advantages over previous research. The Exercise also downregulates the inflammatory sample in this study was representative of the response system. Initially, exercise acts as an population—not a specific subgroup. The fatty acute physical stressor and raises IL-6 and acids were directly measured in the plasma TNF-␣. An initial burst of these cytokines does rather than estimated from patient dietary re- not appear to be harmful. Indeed, and as ports. Moreover, it was the first study that spe- Goebels, Mills, Irwin, and Ziegler (2000) point cifically examined the relationship between out, in a normally functioning immune system, fatty acids and cytokines. high levels of cytokines trigger the body’s anti- PSYCHOLOGICAL TRAUMA AND PHYSICAL HEALTH 43 inflammatory mechanisms to keep inflamma- flexibility/resistance program. These effects tion in check. were independent of depression, optimism or Over a longer period of time, however, espe- sense of coherence. cially in people with chronically elevated proin- Exercise also had a positive effect on wound flammatory cytokines, exercise lowers inflam- healing, and this is an indirect measure of sys- mation. Older adults, for example, are one temic inflammation (Emery, Kiecolt-Glaser, group with higher levels of proinflammatory Glaser, Malarky, & Frid, 2006). In this study, cytokines since levels naturally increase with participants were randomized into exercise and aging. Because of this increased vulnerability in control conditions, and were then brought into older adults, they are frequently the population the laboratory and given a punch biopsy. The of choice for studies on exercise, depression, researchers then monitored participants’ rate of and inflammation. The results of these studies wound healing. The average number of days for are helpful for understanding the mechanism for the wound to heal in the exercise group was 29 exercise’s impact on the stress response. days. In the control group, it was 38 days. A study of adults ages 60 to 90 tested the Exercise one hour a day, three days a week, effect of physical activity on perceived stress, lowered perceived stress and improved wound mood, and quality of life. The researchers also healing. This study is of interest because we assessed serum IL-6 and cortisol. The patients know from these researchers’ other studies that (N ϭ 10) assigned to the exercise group were wound healing is impaired when stress or hos- instructed to walk for 30 minutes, at a rate that tility levels are high (e.g., Kiecolt-Glaser et al., would elevate their to 60% of its 2005). Stress and hostility both increase sys- maximal capacity, five times a week for the temic inflammation. When systemic inflamma- 10-week study. The control group was 10 older tion is high, wound healing is impaired because adults who were not engaging in physical ac- proinflammatory cytokines are in the blood tivity. After the 10-week exercise intervention, stream and not at the wound site where they the subjects had significantly lower stress on the belong. The Emery et al. (2006) study indicates Perceived Stress Scale, and improved mood and that exercise improves wound healing by low- quality of life on the SF-36 Health Question- ering levels of circulating systemic cytokines, naire. They reported better physical function- and diverting cytokines to where they belong. ing, more vitality, better mental health, and less Overall level of fitness was also related to in- bodily pain. They also had a significant de- flammation in another recent study (Hamer & crease in serum IL-6, an effect independent of Steptoe, 2007). The sample consisted 207 men an association between psychological variables and women from London who had no history or and IL-6 (Starkweather, 2007). symptoms of heart disease, and were not being Another study of older adults compared two treated for hypertension, inflammatory disease, or types of workouts to see if either type lowered . Participants were given one of two men- inflammation (Kohut et al., 2006). In this tally stressful tasks in the laboratory (a computer- study, 83 adults, ages 64 to 87, were random- ized Stroop test or mirror tracing task). Research- ized to either cardio or flexibility/resistance ers drew blood and measured heart rate via a conditions. The cardio workouts were 45 min- submaximal exercise test. A high-systolic blood utes, at 60% to 80% of maximal cardiac effort, pressure indicated a low level of fitness. Partici- three times a week, for 10 months. The flexi- pants who responded with higher systolic blood bility/resistance workouts were 45 minutes of pressure to stress also had a higher IL-6 and resistance and flexibility training, three times a TNF-␣ response. The TNF-␣ response to stress week, also for 10 months. Both types of exer- was five times greater in the low-fitness group cise led to improved levels of depression, opti- compared to the high-fitness group. The authors mism, and sense of coherence, the three psy- concluded that participants who were physically chological measures the researchers assessed. fit had a lower inflammation response when under At the end of 10 months, the cardio workout had stress. They believed that this was another way the most impact on inflammation. Participants that exercise protected individuals from heart dis- in the cardio condition had significant reduc- ease and other conditions. tions in C-reactive protein, IL-6, and IL-18. When starting an exercise program for TNF-␣ levels improved with both cardio and trauma survivors, some cautions are in order. 44 KENDALL-TACKETT

It’s important to start this program slowly. Ex- A brief overview of interventions for sleep ercise that causes too rapid a rise in heart rate is difficulties. Interventions for sleep disorders similar to hypertension responses and may include cognitive–behavioral approaches, assis- cause trauma survivors distress, including an tive devices, and medications. Cognitive– increased PTSD response. Implementing an ex- behavioral approaches are the most common ercise program that builds up slowly allows form of sleep treatments. In one recent review, patients to adapt to it psychologically and phys- it was effective for 70% to 80% of patients, and ically over time (C. Courtois, personal commu- was comparable to sleep medications (Morin, nication, January 3, 2009). 2004; Stepanski & Perlis, 2000). Cognitive– In summary, omega-3s and exercise both behavioral interventions help with sleep be- downregulate the stress response, particularly cause they produce changes in REM sleep. Cog- lowering levels of proinflammatory cyto- nitive approaches can also address worrying and kines. Trauma and resultant PTSD can cause rumination that may be at the base of primary or alterations in the stress response system that secondary insomnia (Morin & Ware, 1996). leave trauma survivors more vulnerable to Cognitive therapy for insomnia includes three components: behavioral, cognitive, and subsequent life stress. These interventions educational, and can be most easily incorpo- give survivors the tools they need to lessen rated into standard trauma care. These include their heightened stress response and increase establishing regular bedtimes, not using the bed their resilience to stress. Further, chronically for anything but sleeping and sex, getting out of elevated inflammatory cytokines are related bed when unable to sleep, and eliminating naps to an increased risk of disease. By lowering during the day. There is also a relaxation com- these through fish-oil supplements, exercise, ponent that focuses on both on autonomic re- or both, trauma survivors can also lessen their laxation techniques (e.g., progressive muscle risk of disease. relaxation), and cognitive techniques that stops the worrying that keeps people from sleeping. Treatment of Sleep Disorders Finally, sleep-hygiene education helps people minimize behaviors that might interfere with As described earlier, sleep is often compro- sleep. This might include eliminating caffeine, mised in trauma survivors. This too increases exercise, upsetting reading or media, alcohol their risk of health problems because dis- and smoking too close to bedtime (Morin, 2004; rupted sleep upregulates the stress response, Stepanski & Perlis, 2000). A combination of including the level of proinflammatory cyto- these approaches is effective for most patients with sleep disorders. kines (Suarez & Goforth, in press). Krakow, For sleep apnea, and other types of sleep- Lowry, et al. (2000) noted that since sleep breathing disorders, a CPAP (Continuous Posi- medicine is not well-integrated into trauma tive Airway Pressure) device can be helpful. treatment, practitioners are often less effec- This is a small unit includes a facial mask, tive than they could be if they also treated which provides pressurized air to prevent the underlying sleep disorders. airway from collapsing while the patient sleeps. To lessen the negative health effects of sleep This ensures a better night’s sleep. Some dental problems, the first step is to assess trauma sur- devices, such as a Mandibular Repositioning vivors for possible sleep disorders. Clinicians Device, can also help keep the airway open should ask about the onset of the sleep disorder. during sleep. A sleep-medicine specialist can If clients have any psychiatric disorders, they help determine is a patient needs devices like should ask about the temporal sequence of when these. The physician and therapist should be the sleep disorder and psychiatric disorders aware that devices such as these might be espe- manifested. Did the symptoms of the psychiat- cially difficult for some trauma survivors be- ric disorder predate the onset of sleep problems cause they have elements reminiscent of the or vice versa? Polysomnigraphic studies can trauma (the assault/domestic violence survivor also reveal whether there are any sleep- who was choked, the sexual abuse survivor who breathing or sleep-movement disorders that was forced to perform oral sex and who felt might also be treated. smothered during the experience). They should, PSYCHOLOGICAL TRAUMA AND PHYSICAL HEALTH 45 therefore, work carefully with the survivor to References implement the use of these devices, or in some cases, may need to find alternative approaches Babyak, M., Blumenthal, J. A., Herman, S., Khatri, that are easier to tolerate (C. Courtois, personal P., Doraiswamy, M., Moore, K., et al. (2000). communication, January 3, 2009). Exercise treatment for major depression: Mainte- nance of therapeutic benefit at 10 months. Psycho- Although medications are not always part of somatic Medicine, 62, 633–638. trauma treatment, for patients with sleep disor- Batten, S. V., Aslan, M., Maciejewski, P. K., & ders, the -2 Antagonists/Reuptake In- Mazure, C. M. (2004). Childhood maltreatment as hibitors (SARIs), such as trazadone (Desyrel) a risk factor for adult cardiovascular disease and and nefazadone (Serzone), can be helpful ad- depression. Journal of Clinical , 65, juncts. When they are administered with a Se- 249–254. lective Serotonin Reuptake Inhibitor (SSRI), Bezchilibnzyk-Butler, K. Z., & Jeffries, J. J. (1999). SARIs boost the actions of these drugs, and Clinical handbook of psychotropic drugs (9th ed.). reverses medication-induced insomnia. Traza- Seattle, WA: Hogrefe & Huber. Blumenthal, J. A., Babyak, M. A., Doraiswamy, done suppresses REM sleep, which can reduce P. M., Watkins, L., Hoffman, B. M., Barbour, the number of nightmares patients experience K. A., et al. (2007). Exercise and pharmacotherapy (Lange et al., 1999). In contrast, nefazadone in the treatment of major depressive disorder. Psy- increases REM sleep and improves overall sleep chosomatic Medicine, 69, 587–596. quality (Bezchilibnzyk-Butler & Jeffries, 1999). Bremner, J. D. (2005). The neurobiology of child- Both of these drugs may be too sedating for hood sexual abuse in women with posttraumatic some patients, however (Friedman, 2001). For a stress disorder. In K. A. Kendall-Tackett (Ed.), more detailed discussion of medications for Handbook of women, stress and trauma (pp. 181– PTSD, see Friedman (2004). 203). New York: Taylor & Francis. Butterfield, M. I., Forneris, C. A., Feldman, M. E., & Beckham, J. C. (2000). Hostility and functional Conclusion health status in women veterans with and without posttraumatic stress disorder: A preliminary study. Until recently in the trauma field, researchers Journal of Traumatic Stress, 13, 735–741. and clinicians have tended to focus predomi- Carmichael, C. L., & Reis, H. T. (2005). Attachment, sleep quality, and depressed affect. Health Psy- nantly on mental health sequelae, largely leav- chology, 24, 526–531. ing physical health sequelae for other health Choquet, M., Darves-Bornoz, J-M., Ledoux, S., care providers to deal with. This oversight in the Manfredi, R., & Hassler, C. (1997). Self-reported field has been much more recognized in recent health and behavioral problems among adolescent years and the physical/medical effects associ- victims of rape in France: Results of a cross- ated with trauma and its aftermath are now well sectional survey. Child Abuse and Neglect, 21, researched and well recognized (e.g., Schnurr & 823–832. Green, 2003). For more than 20 years now, PNI Clum, G. A., Nishith, P., & Resick, P. A. (2001). researchers have demonstrated that what indi- Trauma-related sleep disturbance and self-reported physical health symptoms in treatment-seeking fe- viduals think and feel has a measurable effect male rape victims. Journal of Nervous & Mental on the immune system and response to stress, Disease, 189, 618–622. demonstrating that the line between physical Danese, A., Pariante, C. M., Caspi, A., Taylor, A., & and mental health is not as solid as we once Poulton, R. (2007). Childhood maltreatment pre- believed. Indeed, mental health sequelae, such dicts adult inflammation in a life-course study. as depression, are both outcomes and mecha- Proceedings of the National Academy of Sciences nisms that lead to poor health. One of the more USA, 104, 1319–1324. hopeful findings of this research is that these Delarue, J., LeFoll, C., Corporeau, C., & Lucas, D. trauma-related effects can be managed and pos- (2004). N-3 long-chain polyunsaturated fatty ac- sibly reversed. Trauma survivors can take a ids: A nutritional tool to prevent insulin resistance associated to type 2 diabetes and obesity? Repro- number of specific actions to downregulate their ductive Nutrition & Development, 44, 289–299. stress response and halt their decline into seri- Dhabhar, F. D., & McEwen, B. S. (2001). Bidirec- ous health problems. For trauma survivors tional effects of stress and glucocorticoid hor- working to regain control of their lives, that is mones on immune function: Possible explanations good news indeed. for paradoxical observations. In R. Ader, D. L. 46 KENDALL-TACKETT

Felten, & N. Cohen (Eds.), Psychoneuroimmunol- Groe¨r, M. W., Thomas, S. P., Evans, G. W., Helton, ogy (3rd ed., Vol. 1, pp. 301–338). New York: S., & Weldon, A. (2006). Inflammatory effects and Academic Press. immune system correlates of rape. Violence and Dirkzwager, A. J., van der Velden, P. G., Grievink, Victims, 21, 796–808. L., & Yzermans, C. J. (2007). Disaster-related Gross, R. J., Doerr, H., Caldirola, D., Guzinski, posttraumatic stress disorder and physical health. G. M., & Ripley, H. S. (1980). Borderline syn- , 69, 435–440. drome and incest in chronic pelvic pain patients. Dobie, D. J., Maynard, C., Kivlahan, D. R., Johnson, International Journal of Psychiatry in Medi- K. M., Simpson, T., David, A. C., et al. (2006). cine, 10, 79–96. Posttraumatic stress disorder screening status is Hamer, M., & Steptoe, A. (2007). Association be- associated with increased VA medical and surgical tween physical fitness, parasympathetic control, utilization in women. Journal of General Internal and proinflammatory responses to mental stress. Medicine, 21, S58–64. Psychosomatic Medicine, 69, 660–666. Drossman, D. A., Leserman, J., Li, Z., Keefe, F., Hu, Harrop-Griffiths, J., Katon, W., Walker, E., Holm, L., Y. J. B., & Toomey, T. C. (2000). Effects of Russo, J., & Hickok, L. (1988). The association coping on health outcome among women with between chronic pelvic pain, psychiatric diag- gastrointestinal disorders. Psychosomatic Medi- noses, and childhood sexual abuse. Obstetrics and cine, 62, 309–317. Gynecology, 71, 589–593. Drossman, D. A., Li, Z., Leserman, J., Toomey, Hulme, P. A. (2000). Symptomatology and health T. C., & Hu, Y. J. B. (1996). Health status by care utilization of women primary care patients gastrointestinal diagnosis and abuse history. Gas- who experienced childhood sexual abuse. Child troenterology, 110, 999–1007. Abuse and Neglect, 24, 1471–1484. Emery, C. F., Kiecolt-Glaser, J. K., Glaser, R., Ma- Humphreys, J. C., Lee, K. A., Neylan, T. C., & larky, W. B., & Frid, D. J. (2006). Exercise accel- Marmar, C. R. (1999). Sleep patterns of sheltered erates wound healing among health older adults: A battered women. Journal of Nursing Scholar- ship, 31, 139–143. preliminary investigation. The Journals of Geron- Kendall-Tackett, K. A. (2000). Physiological corre- tology Series A, Biological Sciences and Medical lates of childhood abuse: Chronic hyperarousal in Sciences, 60, 1432–1436. PTSD, depression, and irritable bowel syndrome. Felitti, V. J., Anda, R. F., Nordenberg, D., William- Invited review. Child Abuse & Neglect, 24, 799– son, D. F., Spitz, A. M., Edwards, V. J., et al. 810. (1998). Relationship of childhood abuse and Kendall-Tackett, K. A. (2003). Treating the lifetime household dysfunction to many of the leading health effects of childhood victimization. Kingston, causes of death in adults. American Journal of NJ: Civic Research Institute. Preventive Medicine, 14, 245–258. Kendall-Tackett, K. A. (2007). Cardiovascular dis- Ferrucci, L., Cherubini, A., Bandinelli, S., Bartali, B., ease and metabolic syndrome as sequelae of vio- Corsi, A., Lauretani, F., et al. (2006). Relationship lence against women: A psychoneuroimmunology of plasma polyunsaturated fatty acids to circulating approach. Trauma, Violence and Abuse, 8, 117– inflammatory markers. Journal of Clinical Endo- 126. crinology & : Clinical and Experimen- Kendall-Tackett, K. A., Marshall, R., & Ness, K. E. tal, 91, 439–446. (2000). Victimization, healthcare use, and health Frasure-Smith, N., & Lesperance, F. (2005). Reflec- maintenance. Family Violence & Sexual Assault tions on depression as a cardiac risk factor. Psy- Bulletin, 16, 18–21. chosomatic Medicine, 67, S19–S25. Kendall-Tackett, K. A., Marshall, R., & Ness, K. E. Friedman, M. J. (2001). Posttraumatic stress disor- (2003). Chronic pain syndromes and violence der: The latest assessment and treatment strate- against women. Women and Therapy, 26, 45–56. gies. Kansas City, MO: Compact Clinicals. Kibler, J. L., Joshi, K., & Hughes, E. E. (in press). Friedman, M. J. (2004). Allostatic vs. empirical per- Cognitive and behavioral reactions to stress among spectives on pharmacotherapy for PTSD. In J. P. adults with PTSD: Implications for and Wilson, M. J. Friedman, & J. D. Lindy (Eds.), health. In The psychoneuroimmunology of chronic Treating psychological trauma and PTSD (pp. 94– disease: Exploring the links between inflammation, 124). New York: Guilford Press. stress, and illness. Washington, DC: American Goebel, M. U., Mills, P. J., Irwin, M. R., & Ziegler, Psychological Association. M. G. (2000). Interleukin-6 and tumor necrosis Kiecolt-Glaser, J. K., Belury, M. A., Porter, K., Bev- factor-alpha production after acute psychological ersdoft, D., Lemeshow, S., & Glaser, R. (2007). stress, exercise, and infused isoproterenol: Differ- Depressive symptoms, omega-6: Omega-3 fatty ential effects and pathways. Psychosomatic Re- acids, and inflammation in older adults. Psychoso- search, 62, 591–598. matic Medicine, 69, 217–224. PSYCHOLOGICAL TRAUMA AND PHYSICAL HEALTH 47

Kiecolt-Glaser, J. K., Loving, T. J., Stowell, J. R., Matthews, K. A., Schott, L. L., Bromberger, J., Cyr- Malarky, W. B., Lemeshow, S., Dickinson, S. L., anowski, J., Everson-Rose, S. A., & Sowers, M. F. et al. (2005). Hostile marital interactions, proin- (2007). Associations between depressive symp- flammatory production, and wound heal- toms and inflammatory/hemostatic markers in ing. Archives of General Psychiatry, 62, 1377– women during the menopausal transition. Psycho- 1384. somatic Medicine, 69, 124–130. Kiecolt-Glaser, J. K., & Newton, T. L. (2001). Mar- McEwen, B. S. (1998). Protective and damaging ef- riage and health: His and hers. Psychological Bul- fects of stress mediators: Allostasis and allostatic letin, 127, 472–503. load. New England Journal of Medicine, 338, 171– Kinder, L. S., Bradley, K. A., Katon, W. J., Ludman, 179. E., McDonell, M. B., & Bryson, C. L. (2008). McEwen, B. S. (2003). Mood disorders and allostatic Depression, posttraumatic stress disorder, and load. , 54, 200–207. mortality. Psychosomatic Medicine, 70, 20–26. Morin, C. M. (2004). Cognitive-behavioral ap- Kohut, M. L., McCann, D. A., Konopka, D. W. R., proaches to the treatment of insomnia. Journal of Cunnick, J. E., Franke, W. D., Castillo, M. C., et Clinical Psychiatry, 65, 33–40. al. (2006). Aerobic exercise, but not flexibility/ Morin, C. M., & Ware, J. C. (1996). Sleep and resistance exercise, reduces serum IL-18, CRP, psychopathology. Applied and Preventive Psy- and IL-6 independent of ␤-blockers, BMI, and chology, 5, 211–224. psychosocial factors in older adults. , Behav- Nivison, M., Guillozet-Bongaarts, A. L., & Montine, ior, & Immunity, 20, 201–209. T. J. (in press). Inflammation, fatty acid oxidation Kop, W. J., Gottdiener, J. S., Tangen, C. M., Fried, and neurodegenerative disease. In K. A. Kendall- L. P., McBurnie, M. A., Walston, J., et al. (2002). Tackett (Ed.), The psychoneuroimmunology of Inflammation and coagulation factors in persons chronic disease: Exploring the links between in- Ͼ65 years of age with symptoms of depression but flammation, stress, and illness. Washington, DC: without evidence of myocardial ischemia. Ameri- American Psychological Association. Noaghiul, S., & Hibbeln, J. R. (2003). Cross-national can Journal of Cardiology, 89, 419–424. comparisons of seafood consumption and rates of Krakow, B., Artar, A., Warner, T. D., Melendez, D., bipolar disorders. American Journal of Psychiatry, Johnston, L., Hollifield, M., et al. (2000). Sleep 160, 2222–2227. disorder, depression, and suicidality in female sex- Pace, T. W., Hu, F., & Miller, A. H. (2007). Cyto- ual assault survivors. Crisis, 21, 163–170. kine-effects on glucocorticoid receptor function: Krakow, B., Lowry, C., Germain, A., Gaddy, L., Relevance to glucocorticoid resistance and the Hollifield, M., Koss, M., et al. (2000). A retrospec- pathophysiology and treatment of major depres- tive study on improvements in nightmares and sion. Brain, Behavior and Immunity, 21, 9–19. post-traumatic stress disorder following treatment Ranjit, N., Diez-Roux, A. V., Shea, S., Cushman, M., for co-morbid sleep-disordered breathing. Journal Seeman, T., Jackson, S. A., et al. (2007). Psycho- of Psychosomatic Research, 49, 291–298. social factors and inflammation in the multi-ethnic Lange, A., deBeurs, E., Dolan, C., Lachnit, T., Sjol- study of atherosclerosis. Archives of Internal Med- lema, S., & Hanewald, G. (1999). Long-term ef- icine, 167, 174–181. fects of childhood sexual abuse: Objective and Robles, T. F., Glaser, R., & Kiecolt-Glaser, J. K. subjective characteristics of the abuse and psycho- (2005). Out of balance: A new look at chronic pathology in later life. Journal of Nervous and stress, depression, and immunity. Current Direc- Mental Disease, 187, 150–158. tions in Psychological Science, 14, 111–115. Leserman, J., Drossman, D. A., Li, Z., Toomey, Rutledge, T., Reis, S. E., Olson, M., Owens, J., T. C., Nachman, G., & Glogau, L. (1996). Sexual Kelsey, S. F., Pepine, C. J., et al. (2006). Depres- and physical abuse history in gastroenterology sion is associated with cardiac symptoms, mortal- practice: How types of abuse impact health status. ity risk, and hospitalization among women with Psychosomatic Medicine, 58, 4–15. suspected coronary disease: The NHLBI-spon- Lu, W., Mueser, K. T., Rosenberg, S. D., & sored WISE study. Psychosomatic Medicine, 68, Jankowski, M. K. (2008). Correlates of adverse 217–223. childhood experiences among adults with severe Sachs-Ericsson, N., Kendall-Tackett, K. A., & Her- mood disorders. Psychiatric Services, 59, 1018– nandez, A. (2007). Childhood abuse and pain in 1026. the National Comorbidity Study. Child Abuse & Maes, M., Christophe, A., Bosmans, E., Lin, A., & Neglect, 31, 531–547. Neels, H. (2000). In humans, serum polyunsatu- Sareen, J., Cox, B. J., Stein, M. B., Afifi, T. O., Fleet, rated fatty acid levels predict the response of C., & Asmundson, G. J. (2007). Physical and men- proinflammatory cytokines to psychologic stress. tal comorbidity, disability, and suicidal behavior Biological Psychiatry, 47, 910–920. associated with posttraumatic stress disorder in a 48 KENDALL-TACKETT

large community sample. Psychosomatic Medi- chophysiological and cortisol responses to psycho- cine, 69, 242–248. logical stress in depressed and nondepressed older Schnurr, P. P., & Green, B. L. (2003). Trauma and men and women with elevated cardiovascular dis- health: Physical health consequences of exposure ease risk. Psychosomatic Medicine, 68, 538–546. to extreme stress. Washington, DC: American Psy- Teegen, F. (1999). Childhood sexual abuse and long- chological Association. term sequelae. In A. Maercker, M. Schutzwohl, & Smith, T. W. (1992). Hostility and health: Current Z. Solomon (Eds.), Posttraumatic stress disorder: status of a psychosomatic hypothesis. Health Psy- A lifespan developmental perspective (pp. 97– chology, 11, 139–150. 112). Seattle: Hogrefe & Huber. Smith, T. W., & Ruiz, J. M. (2002). Psychosocial Troxel, W. M., Cyranowski, J. M., Hall, M., Frank, influences on the development and course of cor- E., & Buysee, D. J. (2007). Attachment , onary heart disease: Current status and implica- relationship context, and sleep in women with re- tions for research and practice. Journal of Consult- current major depression. Psychosomatic Medi- ing and Clinical , 70, 548–568. cine, 69, 692–699. Starkweather, A. R. (2007). The effects of exercise Van Houdenhove, B., Luyten, P., & Egle, U. T. on perceived stress and IL-6 levels among older (2009). The role of childhood trauma in chronic adults. Biological Nursing Research, 8, 1–9. pain and fatigue. In V. L. Banyard, V. J. Edwards, Stepanski, E. J., & Perlis, M. L. (2000). Behavioral & K. A. Kendall-Tackett (Eds.) Trauma and phys- sleep medicine: An emerging subspecialty in ical health: Understanding the effects of extreme and sleep medicine. Journal of stress and of psychological harm. London: Rout- Psychosomatic Research, 49, 343–347. ledge. Suarez, E. C. (2003). Joint effect of hostility and von Kanel, R., Hepp, U., Buddeberg, C., Keel, M., severity of depressive symptoms on plasma Inter- Mica, L., Aschbacher, K., et al. (2006). Altered leukin-6 concentration. Psychosomatic Medi- blood coagulation in patients with posttraumatic cine, 65, 523–527. stress disorder. Psychosomatic Medicine, 68, 598– Suarez, E. C. (2006). Sex differences in the relation 604. of depressive symptoms, hostility, and ex- Wilson, C. J., Finch, C. E., & Cohen, H. J. (2002). pression to indices of glucose metabolism in non- Cytokines and cognition: The case for a head-to- diabetic adults. Health Psychology, 25, 484–492. toe inflammatory paradigm. Journal of the Amer- Suarez, E. C., & Goforth, H. (in press). Sleep and ican Geriatrics Society, 50, 2041–2056. inflammation: A potential link to chronic diseases. Wilson, J. P., Friedman, M. J., & Lindy, J. D. (2004). In K. A. Kendall-Tackett (Ed.), The psychoneuro- Treatment goals for PTSD. In J. P. Wilson, M. J. immunology of chronic disease: Exploring the Friedman, & J. D. Lindy (Eds.), Treating psycho- links between inflammation, stress, and illness. logical trauma and PTSD (pp. 3–27). New York: Washington, DC: American Psychological Associ- Guilford Press. ation. Woods, A. B., Page, G. G., O’Campo, P., Pugh, Suarez, E. C., Lewis, J. G., Krishnan, R. R., & L. C., Ford, D., & Campbell, J. C. (2005). The Young, K. H. (2004). Enhanced expression of cy- mediation effect of posttraumatic stress disorder tokines and by blood monocytes to in symptoms on the relationship of intimate partner vitro lipopolysaccharide stimulation are associated violence and IFN-gamma levels. American Jour- with hostility and severity of depressive symptoms nal of Community Psychology, 36, 159–175. in healthy women. , 29, Zouridakis, E., Avanzas, P., Arroyo-Espliguero, R., 1119–1128. Fredericks, S., & Kaski, J. C. (2004). Markers of Surtees, P. G., Wainwright, N. W. J., Bockholdt, inflammation and rapid coronary artery disease S. M., Luben, R. N., Warcham, N. J., & Khaw, progression in patients with stable angina pectoris. K.-T. (2008). Major depression, C-reactive pro- Circulation, 110, 1747–1753. tein, and incident ischemic heart disease in health men and women. Psychosomatic Medicine, 70, 850–855. Received January 1, 2009 Taylor, C. B., Conrad, A., Wilhelm, F. H., Neri, E., Revision received January 8, 2009 DeLorenzo, A., Kramer, M. A., et al. (2006). Psy- Accepted January 8, 2009 Ⅲ