Introduction

Trauma exposure is highly prevalent in the United States and ranges from exposure to physical abuse, sexual trauma, intimate partner violence, elder abuse, and trauma related to warfare and combat exposure1,2,3,4.. There is a well-established association between trauma exposure and poor physical health outcomes, including cardiovascular and pulmonary disease, metabolic syndromes, and chronic pain5,6,7. Patients with trauma histories often carry a heavy disease burden and have a higher risk of mortality than their non-traumatized counterparts8,9,10,11,12,13. Moreover, patients with trauma histories, especially women, also report poorer Self-Rated Health (SRH), which is also linked to poor health outcomes and lower quality of life14,15,16,17. While the causal mechanisms between trauma exposure and health outcomes are not fully understood, one important pathway is through preventative health behaviors (PHB), including exercise, balanced nutrition, regular doctor and dentist visits, tooth brushing and flossing, getting recommended immunizations, adequate sleep, and not smoking18, p. 2019,20,. According to the Trauma-Informed Theory of Behavior (TTB), trauma exposure may hinder an individual’s ability to engage in positive health behaviors, and it may increase engagement in harmful health behaviors19. Consequently, researchers and policy makers have initiated screening for trauma exposure in primary care settings with the hope of identifying individuals at risk for poor outcomes21,22,23. However, the impact of trauma exposure on PHB does not exist in isolation,the trauma response, which includes physiological, psychological, and social responses to exposure also impact how an individual engages in PHB18. Given the ubiquity of trauma exposure in the general population, it is critical to disentangle the impact of trauma exposure versus trauma sequalae on PHBs. The following study seeks to examine the differential utility of trauma exposure and traumatic sequelae, specifically psychological symptoms and coping strategies in predicting PHB.

The Trauma Informed Theory of Individual Health Behaviors Model (TTB) provides a model to understand how trauma survivors may prioritize their health decisions19. Individuals may be exposed to acute trauma (e.g., assault, violence), trauma-replicating environments (e.g., poverty, homelessness, incarceration), and historical trauma (e.g., enslavement, genocide). Trauma exposure is generally characterized as direct (i.e., “happened to me”) or indirect (witnessing trauma, learning about trauma). Additionally, individuals may be exposed to trauma as part of their employment such as first responders and military personnel24. Trauma screening instruments proposed by policy markers commonly use measures such as the Adverse Childhood Experiences, which assess direct exposure in childhood including sexual, emotional, and physical abuse as well as neglect21,22,23,25. While there is a large literature dedicated to exploring the relationship between characteristics of trauma exposure and outcomes, there is not a consensus regarding the most appropriate trauma exposure screening measure for primary care25. Consequently, the current manuscript will specifically examine direct trauma exposure as well as total trauma exposure.

After exposure, the trauma response involves physiological and psychological reactions to trauma and may include dissociation and engagement in harmful coping mechanisms19. Resilience factors, such as positive coping styles and social support, may attenuate the impacts of trauma exposure and the trauma response on PHB and SRH19. The TTB model is consistent with research on post-traumatic stress disorder (PTSD) which suggests that over 70% of individuals who experience trauma will recover after six months, emphasizing the complex interplay of resilience factors, exposure, and responses to trauma26,27.

A systematic review examining PHB and trauma exposure and PTSD symptoms found that PHB, including physical activity, diet, and engagement in preventative health screenings each appeared associated with both trauma exposure and PTSD symptoms individually; however, the relationships were less apparent when controlling for trauma exposure or PTSD symptoms18. For instance, the unique impact of trauma exposure after controlling for PTSD symptoms was only significant for decreased engagement in preventative health screenings and not diet or physical activity18. This finding is consistent with other research that has found that some types of trauma (e.g., sexual violence) have been associated with avoidance of preventive services (e.g., mammograms, cervical cancer screenings) due to individuals’ fear of feeling out of control and anxious28,29. Previous research examining the impact of trauma exposure on PHB without controlling for psychological symptoms found that exposure was associated with risky behaviors such as smoking and substance abuse30,31,32.

Research has also found that PTSD symptoms as well as other sequelae of trauma such as depression and anxiety, are negatively associated with PHBs such as reduced physical activity, poor sleep quality, diet, substance use, and medication nonadherence18,33,34. Notably, many of these studies do not control for the effects of trauma exposure18,33,34,35. Moreover, some research has explored differential effects of depression and anxiety on PHBs33 but the specific effects of depression and anxiety on PHBs while controlling for trauma exposure is unknown.

As emphasized in the TTB model, an individual’s coping strategies, both positive and negative, may also impact their engagement in PHBs. Previous research has indicated that individuals with poorer coping strategies engage in less physical activity and more substance use36. Coping strategies may be related to PHBs by a variety of mechanisms. For instance, engagement in denial as a coping strategy may lead to avoidance of PHBs like cancer screenings or may reduce an individual’s ability to implement changes to their diet37,38. Positive coping strategies such as acceptance and use of emotional support may enable and motivate individuals to engage proactively in PHBs37,39. Additionally, engagement in coping strategies is significantly associated with symptoms of PTSD, depression, and anxiety, such that positive coping strategies may reinforce resilience and negative coping strategies may reinforce symptoms40,41,42. However, the relationships between trauma exposure, psychological symptoms, coping, and PHBs are less understood.

Like PHBs, the literature examining the impact of trauma on SRH does not distinguish between trauma exposure, psychological symptoms, and coping strategies. Previous research suggests that poorer SRH is positively associated with the number of total traumas experienced, PTSD, and depression16,43,44,45. While one previous study of survivors of intimate partner violence found that violence exposure was a significant predictor of SRH after controlling for demographic variables44, to our knowledge, there are no studies that examine the unique impact of trauma exposure after controlling for psychopathology. Moreover, while research indicates that engagement in positive coping strategies is associated with better SRH, studies have yet to examine the relationship between coping and SRH in the context of trauma exposure46,47. Given the growing interest in using trauma exposure as an indicator of individuals at risk of low engagement in PHBs and poor SRH, it is critical to disentangle the specific impact of trauma exposure, psychological responses to trauma, and coping styles in predicting PHBs and SRH. Consequently, this study was guided by the following hypotheses:

Hypothesis 1: Higher levels of trauma exposure will be related to lower levels of engagement in PHBs and poorer SRH.

Hypothesis 2: Trauma exposure will explain additional variance in PHBs and SRH above and beyond current levels of psychopathology measured using standard screening tools (depression, anxiety, and PTSD).

Hypothesis 3: Coping styles will explain additional unique variance in PHBs and SRH above and beyond trauma exposure and psychopathology.

Method

This project was approved by the University of Illinois Chicago Institutional Review Board (IRB). All methods were performed in accordance with relevant guidelines and regulations. Trained research assistants approached patients who were waiting for scheduled appointments in a primary care medical and dental outpatient clinic setting and offered them a flier describing the study. To prevent participants from completing the survey multiple times, each flier contained a unique ID number and link to the survey. In addition, we recruited patients through the [deleted to ensure blind review] online Heath Research Registry, where patients can voluntarily participate in research studies. Patients in clinic waiting rooms and those recruited online were screened for the following eligibility criteria: 1) at least 18 years old, 2) able to complete measures in English. Informed consent was obtained from all participants. During the process, participants were advised that: 1) participation was voluntary, 2) participation would not influence their services at the University clinics, and 3) all identifying information would be destroyed immediately after data collection was completed. All participants were emailed a $20 gift card upon completion of the online survey. One thousand one hundred and sixty-eight patients were approached in the waiting rooms of medical and dental outpatient clinics, prior to patients’ visit with the clinician, over a six-month time frame. Two-hundred and fifty-five patients accessed the survey link with a valid flier ID. Three participants completed the survey via the online Health Research Registry for a total sample of two-hundred and fifty-eight participants. Forty-three participants were excluded due to missing data, and over half of these participants (N = 31) completed less than 75% of the survey. Finally, three participants who completed the survey were excluded due to duplicate names/email addresses. Therefore, we had a final sample of 212 participants.

Measures

Participants provided standard demographic information, including age, race, gender identity, sexual orientation, income, and education. Participants were also asked about their health insurance status and how many times they had visited a primary care physician, specialist, emergency department, and dentist in the last year.

The 16-item Good Health Practices (GHP) subscale of the Health Behavior Checklist was used to measure engagement in preventative health behaviors (PHB). The GHP is a new subscale determined from a factor analysis of the Health Behaviors Checklist conducted from a large and diverse sample in the United States48. The items, which have responses from 1 (“Not true at all”) to 5 (“Very True”), cover common preventative health behaviors such as exercise, eating a balanced diet, refraining from smoking, brushing teeth, and getting enough sleep (see Table X for the complete list of items). In prior samples, the GHP scale had adequate reliability and validity; the GHP had high internal consistency in our sample (Cronbach’s alpha=.87).

Self-rated health (SRH) was measured with the first item of the SF-12 Health Survey: “In general, how would you say your health is” which was rated on a score of Poor (1), Fair (2), Good (3), Very Good (4), and Excellent (5) such that higher values indicate higher SRH. This item, which has been used in a variety of international and U.S. population studies, has been shown to be a reliable and valid indicator of health problems and functional impairment49,50,51.

The Centers for Epidemiologic Studies Depression Scale (CES-D) is a 20-item self-report scale used to screen for depression52. Items are scored on a scale of 0 (“Never) to 4 (“Most or all the time (5–7 days)”). The CES-D has shown appropriate reliability and validity in the general population53,in the current study, the CES-D had extremely high internal consistency (Cronbach’s alpha = 0.93. A score of 20 or higher was used to indicate moderate-high levels of depression53, compared to previous guidance (e.g., a score of 15) a score of 20 has been shown to be more specific and appropriate53.

The Generalized Anxiety Disorder-7 (GAD-7) is a seven-item self-report scale used to screen for anxiety. Items are scored on a scale from 0 (“Not sure at all”) to 3 (“Nearly Every Day:”). The GAD-7 has been shown to have appropriate reliability and validity as a screener for generalized anxiety54,55 internal consistency in our sample was high (Cronbach’s alpha = 0.93). A score of 10 or higher was used to indicate moderate levels of anxiety56.

Trauma exposure was measured using the Life Events Checklist (LEC). The LEC includes 16 different types of trauma exposure and an option for “any other stressful event or experience.” Exposure types include natural disasters, accidents, physical assault, sexual assault, and combat. The “other” trauma response was coded as a unique type of exposure. Response options included: “Happened to me,” “Witnessed it,” “Learned about it,” “Part of my job,” “Not sure,” and “Doesn’t apply”57. In the current study, we utilized total exposure and direct exposure. Direct exposure was defined as participants who reported that a traumatic exposure “Happened to me.” Direct exposure was quantified as the total number of exposures a participant reported “happened to me”57. Total exposure was quantified as the total number of exposures a participant endorsed across all categories (including learning about it, witnessing it, and engaging in it as part of their job).

The Posttraumatic Stress Disorder Checklist-5 (PCL-5) is a 20-item self-report scale used to assess PTSD symptoms58. Items are scored on a scale of 0 (“Not at all”) to 4 (“Extremely). The PCL-5 has established high reliability and validity in previous and current samples (Cronbach’s alpha = 0.9758). Participants with scores of 31 or higher on the PCL-5 were characterized as having symptoms indicative of probable PTSD59.

The Brief COPE is a 28-item self-report scale with responses rated on a scale of 0 (“not at all”) to 3 (“a lot”); the initial Brief COPE scale had fourteen subscales that corresponded to different types of coping60. There is currently a lack of consensus regarding the underlying factor structure and subscales of the Brief COPE, with some research suggesting that it is best represented with all fourteen subscales and other studies showing evidence for a three-factor solution61. In this study, we used a three-factor solution with the following factors: emotion-focused coping (including acceptance, using humor, emotional support, religious/spiritual support, acceptance,Cronbach’s alpha = 0.75, problem-focused coping (including instrumental support, planning, and active coping; Cronbach’s alpha = 0.81, and dysfunctional coping (including denial, self-distraction, substance use, and disengagement, self-blame, and venting; Cronbach’s alpha = 0.81.

Statistical analyses

The distribution of direct trauma exposure had a significant positive skew such that most participants reported zero types of trauma exposures (N = 81), and very few participants reported six or more types of trauma exposure (N = 27). Thus, for all analyses involving trauma exposure, we tested the uncorrected trauma exposure variable (i.e., raw) and a log-transformed trauma exposure variable to account for the skewness62. Any differences between uncorrected and log-transformed analyses are noted. To correct for type I error, we applied the Benjamini–Hochberg correction (using a false positive rate of 10%) to all analyses using the R2 change63.

For hypothesis 1, we analyzed the relationship between trauma exposure, preventive health behaviors, and SRH using a Pearson’s r correlation. Hypothesis 2 was examined using a hierarchical linear regression predicting preventive health behaviors and SRH from scores on the psychopathology screener in block one and trauma exposure in block two. Due to collinearity, we examined each psychopathology measure in separate models. Finally, for hypothesis 3, we added the subscales of the Brief-Cope in block three of the hierarchical regression model.

Results

Demographic statistics for included participants are shown in Supplemental Table 1. Briefly, just under half (49%) of participants were white, and most were female (68%), heterosexual (86%), and on Medicaid or Medicare (75%). Most participants had scores on the CES-D indicative of moderate depression (85%, N = 181), and over a quarter of participants reported symptoms indicative of moderate generalized anxiety disorder (GAD; 26%, N = 54) and PTSD (33%, N = 70). Most participants reported at least one type of direct trauma exposure (N = 131, 62%). The median number of types of direct trauma exposures was two; 38% of participants reported no trauma exposures, 26% one-two exposures, 17% three-four exposures, and 19% reported five or more exposures. The median number of total exposures was 5 (M = 6.6, SD = 6.7). The distribution of trauma exposures is presented in Supplemental Table 2. Participants reported moderately high levels of engagement in preventative health behaviors (M = 3.28, SD = 0.74, scale of 1–5). Descriptive statistics for individual health behaviors are presented in Supplemental Table 3. Most participants reported that their health was “good” (Score = 3, 42%) with an average rating of 3.23 (out of 5).

Descriptive statistics and Pearson’s correlations for all variables are shown in Supplemental Table 4. Engagement in PHBs was negatively associated with depression (r = −0.33, p < 0.000), anxiety (r = −0.37, p < 0.001), and PTSD (r = −0.34, p < 0.000). Number of types of direct trauma exposure (r = 0.06, p = 0.38) and total exposure (r = −0.01, p = 0.84) was not significantly associated with PHBs. Log transforming trauma exposure to correct for the positive skew did not impact the association between trauma exposure and engagement in PHBs. Direct and total trauma exposure, both in the raw and log transformed, was associated with higher levels of depression, anxiety, PTSD and the three subscales of the Brief COPE. Lower rating of engagement in preventative health behaviors was associated with poorer SRH (r = 0.18, p = 0.01). Poorer SRH was also associated with more direct exposure and higher levels of depression, anxiety, and PTSD; however, it was not significantly associated with total exposure.

Number of types of direct trauma exposure did not explain additional variance in health behaviors above and beyond scores on the depression screener (CESD; R2 change = 0.01, p = 0.06; Supplemental Table 5); similar results were shown in the analysis using total traumatic exposures and the log-transformed analyses (Supplemental Tables 6, 7, 8). Direct trauma exposure explained a small but not statistically significant portion of variance in health behaviors over and above scores on an anxiety screener (GAD7; R2 change = 0.02, p = 0.04; Benjamini–Hochberg correction applied; Supplemental Table 5). Similarly, total traumatic exposures or the log-transformed variables were not significant (Supplemental Tables 6, 7, 8). Trauma exposure explained 5% additional variance in health behaviors over and above PTSD scores on the PTSD screener (PCL-5; R2 change = 0.05, p = 0.001 Supplemental Table 5); results were similar with the log-transformed direct trauma exposure (see Supplemental Table 6). However, when total trauma exposures were used, the results were not significant (Table 7, 8).

When analyses were repeated with SRH as the outcome, direct trauma exposure explained a small portion of variance in SRH over and above scores on the depression screener (R2 change = 0.03;), and anxiety screener (R2 change = 0.03), but not the PTSD screener (Supplemental Table 5). When total trauma exposure was used, exposure did not predict SRH over and above the depression, anxiety, or PTSD screeners (Table 7). Results were not impacted by the use of the log-transformed exposure variables (Supplemental Table 6 & 8).

The dysfunctional coping subscale of the Brief COPE was negatively associated with engagement in preventative health behaviors (r = −0.232 p = 0.002) and was not significantly associated with emotion or problem-focused coping. When predicting PHB with the Brief-COPE, direct trauma exposure, and psychopathology, the Brief-COPE explained an additional 7% of variance above depression and trauma (p = 0.001), 5% above anxiety and trauma (p = 0.003), and 5% above PTSD and trauma (p = 0.004). All results were similar in the log-transformed models and the models using total trauma exposure (Supplemental Table 5, 6, 7, 8). The only coping style that was uniquely predictive of PHBs above trauma exposure and psychopathology was dysfunctional coping in the depression and anxiety models (Supplemental Table 5). Scores on the Brief COPE subscales were not significantly associated with SRH, and the subscales did not explain significant variance in SRH over and above direct exposure, total exposure, and psychopathology (Supplemental Table 5, 6, 7, 8).

Discussion

This study examined the relationship between direct and total trauma exposure, psychopathology, and coping strategies in a diverse urban sample. Overall, our results showed that trauma exposure explains a very small portion of the variance in PHBs and SRH compared to psychopathology and coping strategies. Contrary to our hypothesis and prior studies, higher levels of direct trauma and total trauma exposure were not significantly associated with lower engagement in PHBs18,30,31,32. The measure of PHBs in this study, the GHP subscale, averages engagement in various PHBs,this measure captures a participant’s overall attitude towards engagement in PHBs rather than an individual’s specific inclination to engage in a given behavior (such as diet, smoking, or getting a mammogram48;). It is possible that trauma exposure may have a different relationship with specific health behaviors, particularly procedures that may involve physical exams that re-trigger trauma. Additionally, while our sample included participants who endorsed a range of trauma exposures, we did not examine the relationship between specific exposures (i.e., sexual assault, transportation accident, physical assault) and PHBs. Future research should explore possible connections between specific types of exposures and PHBs. However, given that the majority of individuals who experience trauma do not struggle with long-term psychological difficulties, the nonsignificant relationship may reflect the resilience of the general population26,27.

The impact of resilience and current mental health status is further supported by the examining the impact of psychopathology on PHBs. We found significant negative relationships between PHBs and symptoms of depression, anxiety, and PTSD, suggesting that those individuals who had higher symptom burden reported lower engagement in PHBs. Moreover, trauma exposure explained a small portion of variance above and beyond PTSD symptoms (5%), no variance above and beyond depression symptoms, and no statistically significant portion of variance and beyond anxiety symptoms. These results were largely similar when total traumatic exposures were used. Overall, these results indicate that after accounting for psychopathology using standard screening instruments, trauma exposure explains a negligible portion of the variance in PHBs. This finding is supported by prior literature that found that the unique impact of trauma after controlling for PTSD was not significant in predicting diet or physical activity18. Further, we found that coping strategies explained a significant portion of additional variance in PHBs, over and above trauma and psychopathology. These findings suggest that assessing for current symptoms and functioning is more useful than assessing for trauma exposure alone25.

Consistent with our hypotheses and previous literature, higher levels of trauma exposure were associated with poorer SRH and higher levels of SRH were associated with lower levels of depression, anxiety and PTSD symptoms16,43,44,45. Direct trauma exposure was significant in predicting a small portion of the variance in SRH above and beyond depression and anxiety symptoms, although exposure did not predict SRH above and beyond PTSD symptoms. Given that SRH is a significant indicator of overall health status, this result may reflect the impact of trauma on physical health20. Alternatively, survivors of traumas such as serious motor vehicle accidents or intimate partner violence may struggle with chronic pain resulting from traumatic injuries, leading to poorer SRH64.

Unexpectedly, SRH was not related to any of the subscales of the Brief COPE, and the subscales did not explain significant variance in SRH over and above direct exposure and psychopathology. Overall, engagement in the three coping strategies were evenly distributed across the responses for SRH. While SRH was normally distributed in our sample, only eight participants (4%) reported that their health was “poor.” Consequently, limited range restriction at the negative side of SRH may have limited our ability to find a significant relationship between SRH and coping strategies.

Overall, our results suggest that trauma-informed programs in healthcare settings may consider screening for current psychopathology rather than implementing universal trauma screening tools. Patients who endorse high levels of depression, anxiety, and PTSD symptoms may be less likely to be engaging in PHBs and therefore more likely to have poor health outcomes18,19. Patient education can be targeted toward those who are at higher risk of negative health outcomes. Similarly, our results suggest that rather than screening for trauma history, health programs may consider screening for current coping strategies. Dysfunctional coping (for example, denying problems, or completely disengaging from challenging situations) was uniquely associated with lower PHB engagement, Thus, patients with dysfunctional coping styles may benefit from health education that emphasizes the importance of emotional or problem-focused coping strategies.

In conclusion, while trauma exposure has a nuanced relationship with preventive health behaviors and self-rated health, our findings highlight the dominant role that current psychopathology and coping strategies play in shaping health outcomes. The results suggest that trauma-informed interventions should focus on mental health screening and the assessment of coping strategies, rather than solely on trauma history. This approach could provide more immediate insight into individuals' vulnerability to poor health behaviors and outcomes. Additionally, these findings emphasize the importance of personalized healthcare that addresses both psychological distress and coping mechanisms, fostering resilience, and promoting better overall health management. Future research should explore how different types of trauma exposure moderate or mediate specific preventive health behaviors and coping styles. Additionally, future research may explore how interventions can be used to reduce the impact of trauma exposure, psychological symptoms, and negative coping styles on outcomes.