Introduction

Post-traumatic stress disorder (PTSD) is a syndrome that develops after exposure to traumatic event(s) such as threatened or actual death, serious injury, and/or sexual violence [1]. PTSD can develop from direct experience, witnessing of a traumatic event, learning of a traumatic event that happened to someone close, or from extreme exposure(s) to the details of a traumatic incident [2]. A wide range of experiences can result in PTSD including childhood maltreatment, accidents, exposure to war, violence, and conflict. PTSD results in clinically significant distress and/or impairment of daily functioning (i.e., capacity to work, social interaction). The lifetime prevalence estimates of PTSD have been estimated as 6.1% in the United States [3] and 9.2% in Canada [4]. Among groups that have had high exposure to violence and conflict, such as refugees, the prevalence of PTSD can be much higher with estimates close to 30% [5]. Other groups, particularly those who have endured exposure to combat and/or sexual abuse, are also more likely to experience PTSD (≥ 25–50%) than the general public [6]. PTSD can result in a wide range of negative outcomes including insomnia, greater difficulties in progressing in education and career [7, 8], increased risk for unemployment and homelessness [9], and problematic relationships [8]. PTSD is associated with substantial indirect costs to the economy due to lost work and health care-related needs [7]. For instance, in Canada there is an estimated $21 billion per year loss in productivity directly associated to PTSD [8].

Vulnerability to PTSD has been shown to differ by various social determinants of health. The lifetime prevalence of PTSD in women is twice that of men [6, 10,11,12]. Women are believed to be more prone to PTSD due to their heightened risk of exposure to sexual abuse. The age at which the peak prevalence of PTSD occurs varies by gender, with the prevalence of PTSD being highest among men in their early 40s and women in their early 50s [11]. The prevalence of PTSD is lower among individuals who are married [13, 14], those who are more educated [15, 16] and those who have higher socio-economic status [17]. Individuals with greater social support are less likely to develop PTSD after experiencing trauma [18]. Studies have shown a strong association between PTSD and comorbid conditions, including depression, anxiety, and substance use disorders (such as alcoholism) [6, 19,20,21,22,23]. Recovery from PTSD is more rapid among those who experience a non-intentional injury vs those who have had an intentional injury inflicted upon them during the traumatic event [24, 25].

PTSD has been reported to be associated with a wide range of health issues including neurological (e.g., epileptic seizures), gastrointestinal (e.g., ulcers), autoimmune, metabolic and/or joint/bone conditions [22, 26]. PTSD may also elevate the risk of subsequent suicidal ideation, suicidal attempt, and death by suicide [8, 27]. The prevalence of chronic pain is higher among those with PTSD than among those who do not have a diagnosis of PTSD [28, 29]. Chronic pain, particularly when it is a consequence of a traumatic event, exacerbates PTSD [30]. These reported relationships may be due to alterations in the gut microbiome, hypothalamic–pituitary–adrenal (HPA) axis, and sympathetic nervous system, as well as processes of inflammation and oxidative stress [31].

PTSD, and its associated health outcomes, may also be moderated or mediated by a wide range of health behaviors. PTSD and the use of substances, such as cigarettes [32] and alcohol [33], have been reported to be highly comorbid as individuals use these as mechanisms to self-medicate [34]. Some cross-sectional studies have found that PTSD is associated with obesity [35] and one prospective study reported that those with PTSD gained more weight than their peers who did not have PTSD [36]. The effects of trauma, such as sensory issues, hyperarousal, being easily startled, and feelings of numbness, have been shown to influence appetite and eating [37]. Potential mechanisms that may contribute to the relationship between PTSD and obesity include using food as a negative coping behavior [38], gut dysbiosis [39], HPA disruption [35, 40], cortisol stimulation [41], and inflammation [42]. Some of these associations may be influenced by PTSD-induced changes in sleep and metabolism; both are factors that are associated with obesity [38]. Specific studies examining the relationship between nutrition and PTSD are limited. Preliminary studies suggest that certain nutrients such as B vitamins (i.e., vitamin B12 and folic acid), antioxidants (i.e., vitamin C), essential fatty acids (i.e., omega-3 fatty acids), zinc, magnesium, curcumin, and probiotics have therapeutic potential in modulating various physiological alterations associated with PTSD [43,44,45,46].

The relationship between immigrant status, ethnicity, and PTSD warrants greater research attention. Immigrants include refugees, displaced persons, uprooted people, and economic migrants [47]. From 2000 to 2017, the number of international migrants has increased by 85 million (49%) [48]. Regardless of the reason for migration, stressful life events often precede or co-occur with the process, such as pre-immigration exposure to stressful and traumatizing experiences [49, 50]. Migrant status is often associated with acculturative stress, discrimination, geographic segregation, reduced social networks and supports, urban violence, abuse by law enforcement officers, separation from families, food insecurity, poor mental health, and changes in diet quality [51,52,53]. Refugees have a much higher prevalence of developing PTSD compared to the general population [54,55,56]. Several studies indicate that the incidence of mental health outcomes such as psychological distress [57] and PTSD [58] among a diversity of ethnic immigrant groups has increased worldwide [59].

To better understand the mental health of immigrants, research is needed about PTSD and factors which contribute to its manifestation, particularly among racial/ethnic groups to develop an evidence base which can better contribute to health practice and policies [60, 61]. Using Canadian Longitudinal Study on Aging data, PTSD and its relationships with ethnicity and immigrant status were analyzed while controlling for a number of covariates that included socioeconomic status, physical health (e.g., multi-morbidity), health behaviors (e.g., substance use, physical activity), over-nutrition (e.g., disease risk), poor nutrition status (e.g., handgrip strength, nutrition screen scores, body composition, anemia screen), and dietary intake. Based on current understandings from the literature, the study was directed by three research questions:

  1. 1.

    Does the prevalence of PTSD vary by ethnicity and immigrant status among those aged 45–85 years?

  2. 2.

    Are the associations between ethnicity and immigrant status and PTSD attenuated by a wide range of health determinants?

  3. 3.

    Are there other determinants of health that are significantly associated with PTSD after controlling for ethnicity and immigrant status?

Materials and methods

Participants

Details about the participants have been discussed elsewhere [62,63,64,65,66]. For this investigation, the baseline (2012–2015) comprehensive data of the Canadian Longitudinal Study on Aging (CLSA) was utilized. This dataset includes biological, medical, psychological, social, lifestyle, and economic measures of Canadians aged 45–85 years [66]. The sampling framework for the comprehensive portion of the CLSA included a random selection of households within 50 km of 11 data collections sites [67,68,69]. The data collection sites were in cities in 7 of Canada’s 10 provinces. If the households included an eligible member between the age of 45–85 years, they were asked to participate. It is important to note that the response rate was very low (10%) and predominantly well-educated individuals, and therefore, the CLSA comprehensive sample may not be representative of the Canadian population.

Participants in the comprehensive data set (n = 30,097) completed in-home interviews and had a wide range of physical assessment measures taken at specific study sites [70]. The final sample size with complete data was 27,211. The study protocol of the CLSA has been approved by 13 research ethics boards across Canada. Details about the study are available at www.clsa-elcv.ca. The secondary analysis of CLSA data conducted in this paper was approved by the University of Toronto’s Health Sciences Research Ethics Board.

Measures

Dependent variable

Post-traumatic stress disorder (PTSD): The four-item Primary Care Post-Traumatic Stress Disorder (PC-PTSD) tool was used to screen for PTSD [71]. The PC-PTSD includes 4 questions with yes/no responses about DSM-IV PTSD symptoms and experiences in the last month. The screen begins with an introduction to orient respondents to traumatic events. If participants responded in the affirmative to at least three questions, they are considered to have screened positive for PTSD [72].

Independent variables

Ethnicity/Immigrant Status: A four-level variable was generated to assess differences in the prevalence of PTSD among those born in Canadians and immigrants: (1) Canadian-born white; (2) Canadian-born visible minority member; (3) immigrant white; and (4) immigrant visible minority member.

Covariates

Demographic, social, economic, health, and nutrition-related variables that could potentially attenuate the relationship between PTSD and immigration status were also examined in the analysis and are detailed in Supplementary Table 1.

Statistical analysis

All analyses were completed using SPSS Version 25. As a modification of the model-based approach, an adjusted weight variable was applied to provide estimates that are representative of the total population of Canadians between 45 and 85 years. In keeping with recommended strategies for complex national surveys [73], the adjusted weight was created by dividing the population-based weight by the average weight for the subsample of 27,211 respondents. Weighted percentages were used to describe the entire sample and unweighted sample sizes were also presented. Descriptions of the sample when stratified by presence and absence of PTSD were based on weighted means. Adjusted odds ratios were derived using binary logistic regression to examine associations between ethnicity and immigrant status and PTSD while adjusting for the covariates. An interaction term of gender by ethnicity/immigrant status was not significant in the regression analyses (p = 0.849), and therefore, logistic regression estimates are reported for both sexes together. Seven stepwise models were analyzed by entering seven clusters of factors, respectively, based on the core model with age, sex and ethnicity/immigrant status adjusted (Model 1): demographic and socioeconomic variables (Model 2), health characteristics (Model 3), health behavior variables (Model 4), over-nutrition and poor nutrition status indicators (Model 5), and dietary intake measures (Model 6). The final model was adjusted for all of the aforementioned variables (Model 7). For the final logistic regression model, diagnostics and model fit measures were applied that included the Omnibus Test of Model Coefficients and Nagelkerke’s R2.

Results

Description of sample

The sample mainly consisted of Canadian-born residents (n = 22,448; 82.5%), 45–65 years (n = 16,006; 58.8%), earning between C$20,000–$99,999/year (n = 15,977; 58.7%), in a relationship (n = 18,877; 69.4%), and who had earned a post-secondary diploma or degree or above (n = 21,148; 77.7%) (Table 1). The proportion of women is slightly higher than men (women: 50.6%; men: 49.4%).

Table 1 Description of sample characteristics by PTSD screen (n = 27,211)
Full size table

With reference to health-related characteristics, the majority of participants had at least one health condition (n = 22,302; 82.0%), a form of hypertension or were taking anti-hypertensive medication (n = 14,728; 54.1%), reported having no chronic pain (n = 20,596; 75.7%), did not drink excessive amounts of alcohol (n = 17,598; 64.7%), and indicated they had never or seldom engaged in light sports or recreational activities in the past 7 days (n = 23,289; 85.6%). A slightly higher proportion of participants smoked more than 100 cigarettes in their lifetime compared to those who did not (n = 14,507; 53.3%).

Based on the nutrition status prevalence estimates, most respondents tended to carry excess weight (BMI ≥ 25, n = 18,936; 69.6%) and, based on BMI and WHR measures, were at risk for developing chronic conditions (n = 19,239; 70.7%). Most of the sample did not report indications of poor nutrition status (> 58% across measures for normal/healthy status categories).

With reference to the dietary intake measurements, the majority of respondents tended to consume low amounts of fiber (≤ 2 servings/day; n = 21,656; 79.6%), pulses and nuts (< 1 serving/day; n = 15,048; 55.3%), fruits and vegetables (≤ 4 servings/day; n = 20,289; 74.6%), and calcium containing foods with high vitamin D content (≤ 2 servings/day; 18,376; 67.5%). All demographic, social, economic, health, nutrition status, and dietary intake measurements showed significant associations with PTSD (p’s < 0.05); the exception being binge drinking, bone density scores, and anemia screen, as well as intakes of omega-3 eggs, pure fruit juice, salty snack intakes, and calcium sources with low vitamin D content.

Multivariable analysis

Research Question #1: Does the prevalence of PTSD vary by ethnicity and immigrant status among those aged 45–85 years?

The prevalence of PTSD among visible minority immigrants (7.5%) was more than double that of white immigrants (3.6%, p < 0.001), and approximately 50% higher than whites born in Canada (5.6%, p = 0.008), but did not differ significantly from that of Canadian-born visible minority members (4.9%, p = 0.19), possibly due to the small sample size of the latter. Canadian-born whites had significantly higher prevalence of PTSD than white immigrants (p < 0.001). There was no significant difference in prevalence of PTSD by visible minority status among those born in Canada (p = 0.62), nor between Canadian-born minorities and immigrant whites (p = 0.233).

In the final model of the logistic regression (Please see Table 2), in comparison to whites born in Canada, immigrants from visible minority groups (Table 2) had higher odds of PTSD (OR 1.47, 95% CI 1.14–1.90, p < 0.001). Conversely, white immigrants had 26% lower odds of PTSD (OR 0.74, 95% CI 0.61–0.90, p = 0.003) when compared to whites born in Canada. Canadian-born visible minority respondents did not differ significantly from whites born in Canada. In the base model (Model 1; variables ethnicity and immigrant status, age, sex), compared to Canadian-born whites, immigrant whites had lower odds of PTSD (OR 0.73, p < 0.05, 95% CI 0.61–0.88) (see Fig. 1b) while immigrant visible minorities had higher odds of PTSD (OR 1.26, p = 0.058, 95% CI 0.99–1.60) but this was not significant (see Fig. 1c). Supplemental analyses, not shown, indicate that compared to white immigrants, both visible minority immigrants (OR 1.72, p < 0.001, 95% CI 1.29–2.30) and Canadian-born whites (OR 1.37, p = 0.001, 95% CI 1.13–1.65) had higher odds of PTSD. Canadian-born visible minority respondents did not differ significantly from any of the other categories. Compared to visible minority immigrants only white immigrants had significantly lower odds of PTSD (OR 0.58, p < 0.001, 95% CI 0.43–0.78).

Research Question #2: Are the associations between ethnicity and immigrant status and PTSD attenuated by a wide range of health determinants?

Table 2 Adjusted odds ratios between PTSD and ethnicity and immigrant status, socio-demographic, physical health, health behavior, over- and under-nutrition, and dietary measures
Full size table

As shown in Fig. 1b, immigrant whites (ORs 0.73–0.75) had consistently significantly lower odds of PTSD when compared to whites born in Canada after adjusting for each model cluster (Models 1–7). Conversely, in all but the first model, immigrants from visible minority groups had consistently significantly higher odds of PTSD (ORs 1.47–1.56) than whites born in Canada (Fig. 1c: Models 2–7). Visible minority respondents born in Canada did not differ significantly in the odds of PTSD compared to whites born in Canada in any of the models. Research Question #3: Are there other determinants of health that are significantly associated with PTSD after controlling for ethnicity and immigrant status

Fig. 1
figure 1

Adjusted odds ratios for PTSD by ethnicity and immigrant status a Canadian-born visible minority vs Canadian-born white. b Immigrant white vs Canadian-born white. c Immigrant visible minority vs Canadian white. Model 1: Demographic, social, and economic characteristics: Immigrant status, sex, age, education, household income. Model 2: Model 1 + Physical Health: Co-morbidities, hypertension, chronic pain. Model 3: Model 2 + Health Behaviours: smoking, binge drinking, physical activity. Model 4: Model 3 + Over-nutrition: Disease risk, percent body fat, waist-to-hip ratio, waist-to-height ratio. Model 5: Model 4 + Poor nutrition: Grip strength, nutrition risk (SCREEN™), skeletal muscle index, T-scores, anemia screen, Model 6: Model 5 + Dietary Intakes: Fiber, pulses/nuts, fat sources, fish sources, omega-3 eggs, fruits and vegetables, pure fruit juice, salty snacks, calcium sources with high vitamin D content, calcium sources with low vitamin D content, pastries, chocolate bars. Model 7: Full Model: All variables in models 1–7

Full size image

After adjusting for ethnicity and immigrant status, many significant associations were found for demographic, social, and economic variables (Table 2). Compared to those aged 45–55 years, those in older groups had lower odds of PTSD (ORs 0.18–0.58, p’s < 0.001). For participant groups earning less than C$100,000 annually, there were higher odds of PTSD (ORs 1.39–2.58, p’s < 0.001) when compared to those earning more than C$150,000 annually. Those who were widowed, divorced, or separated had higher odds of PTSD (OR 1.39, 95% CI 1.20–1.60, p < 0.001) compared to those who were single.

There were also many significant associations between PTSD and different health determinants such as physical health and health behaviors. Those who reported having at least two health conditions (ORs 1.76–3.34, p’s < 0.001), having chronic pain (OR 1.64, p < 0.001), or smoking (OR 1.22, p < 0.001), had higher odds of reporting PTSD.

Some indicators of over-nutrition and poor nutritional status were also associated with PTSD. Those with a high waist-to-height ratio (OR 0.82, 95% CI 0.69–0.97, p = 0.018) had lower odds of PTSD; conversely, those who were assessed to be at high nutritional risk (OR 1.60, 95% CI 1.41–1.80, p < 0.001) had higher odds of PTSD.

Finally, some interesting associations were found between PTSD and the consumption of certain food components. For those who on average consumed fiber sources less than three times daily, lower odds of PTSD were indicated (ORs 0.83–0.86, p’s 0.018–0.049) compared to those consuming fewer than one source daily. Conversely, those who consumed on average one or more sources of pulses and nuts had higher odds of PTSD (ORs 1.50–1.16, p’s 0.029 to  < 0.001) compared to those who consumed half or less daily. Those who consumed on average one or more pastries or at least two-thirds of a chocolate bar or more daily had higher odds of PTSD (ORs 1.35–1.37, p’s 0.016–0.025) compared to those who reported no consumption of these types of foods.

Assessment of model fit

The results of the Omnibus Test of Model Coefficients were highly significant (χ2 (79) = 1316.9, p < 0.001) indicating the final model is significantly better than the baseline model. The Nagelkerke’s R2 suggests that the model explains about 13.8% of the variation in PTSD outcome.

Discussion

This study which examined PTSD in a national sample of mid-age and older adults found that compared to white Canadians, those who were immigrant and of visible minority groups had higher odds of PTSD, whereas Canadians who were from visible minority groups were less likely to have PTSD. Furthermore, these associations were robust to adjustments for a number of demographic, social, economic, health status, health behaviors, nutrition status, and dietary intake indicators. After adjusting for ethnicity and immigrant status, those in older groups (> 55 years) had lower odds of PTSD relative to those between 45 and 55 years, whereas lower income earners had higher odds of PTSD when compared to high income earners (≥ C$150,000 annually). Participants who were widowed, divorced, or separated had higher odds of PTSD compared to those who were single. When assessing health status indicators, those who reported having at least two health conditions, having chronic pain, or ever smoking, had higher odds of reporting PTSD. An indicator of over-nutrition, high waist-to-height ratio, was associated with lower odds of PTSD, whereas an indicator of poor nutrition status, high nutrition risk, was associated with increased odds of PTSD. In comparison to low levels of consumption, those with moderate intakes of fiber sources (average two to three times daily) had lower odds of PTSD, while consumption of one or more sources of pulses and nuts on average was associated with higher odds of PTSD compared to those who did not consume pulses or nuts daily. The intake of pastries and chocolate bars was associated with higher odds of PTSD.

Research Question #1: Does the prevalence of PTSD vary by ethnicity and immigrantstatus among those aged 45–85 years? and

Research Question #2: Are the associations between ethnicity and immigrant status and PTSD attenuated by a wide range of health determinants?

Interestingly, associations with PTSD varied according to ethnicity and immigrant status which, to the best of our knowledge, has not been reported elsewhere. While higher incidence of mental health outcomes, such as PTSD, among immigrants have been previously reported, this study concluded that this was only true among visible minority immigrants, with the unexpected finding that white immigrants had significantly lower odds of PTSD compared to Canadian-born whites. Despite controlling for a broad range of demographic, social, economic, health, and nutrition related factors, the associations between PTSD by ethnicity and immigrant status remained consistent. These findings suggest that the lower odds of PTSD among white immigrants and the higher odds of PTSD among visible minority immigrants is not due to the various factors investigated in the fully adjusted model. This suggests that the relationship may be driven by other factors not examined in this study.

A study that examined the psychological effects of trauma exposure among Asian Americans and European American survivors of the 1994 Northridge, California earthquake found Asian American participants reported more psychiatric distress based on the Brief Symptom Inventory [74]. Ethnic differences remained after accounting for age, immigrant status, and exposure to the earthquake. Furthermore, moderator analyses showed that ethnic differences were explained by the interaction of ethnicity and having a foreign-born parent. These findings may be due to heterogeneity of those reporting immigrant status, which can include refugees, displaced persons, uprooted people, and economic migrants [47]. Regardless, they challenge beliefs about immigration and PTSD and suggest there is a heightened need for greater mental health resources for visible minority immigrant groups.

Previous research suggests that first-generation immigrants are physically healthier than individuals of the same ethnic background who are born in the receiving country [75]; a phenomenon referred to as the ‘healthy immigrant effect’. The evidence for the healthy immigrant effect is particularly robust in a 15-year follow-up study based upon the 1991 Census [75]. This analysis included a representative sample of more than 500,000 immigrants aged 25 to 74 and 2 million adults born in Canada, In comparison to those born in Canada, immigrants to Canada had more than a 30% lower risk of death from preventable causes and approximately 20% lower risk of death from treatable causes [76]. However, the physical and mental health of immigrants are known to decline over time [77]. Visible minority immigrants in Canada are largely from South Asia, China and the Middle East source regions, where groups of individuals have experienced political conflict and/or disruption during the previous 60 years [78, 79]. Immigrants from these regions are also more likely to have experienced traumatic incidents such as natural disasters and are at greater risk of PTSD as a result [80,81,82]. Post-migratory challenges such as difficulties with applications, employment, discrimination, and social isolation are also known as contributing factors to the onset or worsening of PTSD [80].

Research Question #3: Are there other determinants of health that are significantly associated with PTSD after controlling for ethnicity and immigrant status?

The lower odds of PTSD among those over 55 years, compared to younger respondents, is consistent with other findings which have shown that PTSD tends to be highest among men in their early 40s and women in their early 50s [11]. Our findings that widowed, divorced, and/or separated individuals and those with low household incomes had higher odds of PTSD than married respondents and those with higher incomes, respectively, are similar to the findings of a study of survivors of an earthquake [83]. The association between PTSD and multiple morbidities is of little surprise based on prior findings showing increased risks of cardiovascular, metabolic, and musculoskeletal disorders among individuals with PTSD; these associations may be due to alterations in the HPA axis, sympathetic nervous system, inflammation, and/or health behaviors that could increase the risk of illness [31].

The association between PTSD and chronic pain is also consistent with prior findings [28, 29]. Previous research has shown that both smoking and binge drinking are associated with higher levels of PTSD [84]. In our study, the former, but not the latter, relationship was significant. The current study found that those with higher waist-to-height ratio had lower odds of PTSD. This association has not been extensively studied in the extant literature, although one study indicated higher waist-to-height ratio is associated with anxiety in middle-aged Latin American women [85]. Another investigation suggests that high waist-to-height ratio is positively associated with somatic but not the cognitive-affective symptoms of depression [86, 87]. Finally, based on longitudinal research [38], individuals with new onset, persistent, or resolved PTSD had higher odds of at least 10% body weight gain compared to those without PTSD. In addition, it was indicated that those with new onset or persistent PTSD had a higher incidence of at least 3% body weight loss. Evidently, more work is needed that explores the relationships between PTSD and body anthropometrics.

To our knowledge, the relationship we report between nutritional risk and PTSD is novel. Nutritional risk is measured in the CLSA using the AB SCREEN™ II (Abbreviated Seniors in the Community Risk Evaluation for Eating and Nutrition II) which consisted of eight questions related to appetite, frequency of eating, motivation to cook, ability to shop and prepare food, weight changes, isolation and loneliness, chewing and swallowing, digestion, and food restrictions, to assess eating habits on a typical day. There are many pathways which may explain the relationship between nutritional risk and PTSD. First, oxidative stress and high homocysteine levels are associated with PTSD and with increased mitochondrial damage and neurotransmitter signaling dysfunction in the brain [45, 86, 87]. Oxidative stress occurs when there is an imbalance between the amount of free radicals produced by the body and the body’s antioxidant defenses [8]. A source of oxidative stress includes exposure to certain nutrition-related components (e.g., high fat, high sugar, and salt) [87,88,89]. A second explanatory pathway is that of the gut microbiome which may be altered due to nutritional status. The microbiota of the gut communicates with the brain through the ‘microbiota-gut-brain axis’; a pathway involved in the pathophysiology of mental health conditions such as PTSD [90]. Finally, some of the manifestations of PTSD, such as severe anxiety or depression, may interfere with food skills and eating behaviors [91] and lead to heightened nutrition risk.

Based on a review of the dietary intake results and PTSD, it was not a surprise that fiber intakes were associated with lower odds of PTSD; however, this benefit was not significant beyond average intakes of three or more fiber sources daily. Because high fiber foods contain phytates which bind with trace minerals such as iron, zinc, and manganese and reduce their bioavailability [92], higher intakes may not impart the same mental health benefits. The findings that indicated higher odds of PTSD with increased intakes of pulses and nuts were surprising. The food frequency questions included sources such as peanut butter and also did not distinguish if the sources of pulses and nuts included less healthy components such as salt (e.g., salted nuts), fat (e.g., sauces with legumes such as pork and beans), and sugar (e.g., candied nuts) which may have impacted these findings. PTSD has been associated with overeating [91], which may suggest consumption of high amounts of these food items may be indicative of self-soothing behavior. The association of PTSD with intake of pastries and chocolate is not surprising as consumptions of these foods may be due to heightened appetite and eating psychopathologies associated with PTSD [91].

Limitations

The results of this study must be interpreted with caution due to several important limitations. Firstly, our sample was restricted to those aged 55 and older. Consequently, our findings cannot be generalized to younger populations. The use of cross-sectional data makes it impossible to infer causality. PTSD is a multi-factorial condition and many of the factors that may contribute to it, such as the nature and frequency of traumatic experience(s), were not available in the CLSA data set. We also did not have information on any therapy that the respondents had undergone. The measure of PTSD was based upon the DSM-IV, not the more current DSM-5. The latter reflects more up-to-date knowledge about the PTSD diagnosis. The use of self-report measures for many of the variables made misreporting and misclassification possible. We also did not control for other mental health conditions such as bipolar disorder [93], eating disorders [94], and personality disorders [95] which have been associated with PTSD but do not have prior traumatic experience as part of their diagnostic profiles. Lastly, the very low response rate and disproportionate number of participants with a university degree in the CLSA sample limits generalizability of the findings.

Despite the investigative limitations noted here, this study had a very large sample size that enabled a comprehensive assessment of associations between PTSD and various factors, including immigration status, ethnicity, and nutritional intake. Suggestions for future investigations include oversampling different ethnic groups to allow for ethnic-specific analyses, examining longitudinal data to explore relationship trajectories between PTSD and health determinants post-settlement, and conducting qualitative work to gain important insights into the experience of PTSD among visible minority immigrants.

Conclusion

PTSD contributes to significant morbidity and is associated with substantial indirect costs to the economy due to lost work and extensive health care-related needs [7]. The current study found important differences by immigrant and visible minority status, with those who were immigrants from visible minority groups having a higher prevalence of PTSD than both white immigrants and white and visible minority respondents born in Canada. This association was robust across various analyses that accounted for various demographic, social, economic, health status, health behaviors, nutrition status, and dietary intake indicators. This suggests that immigrant groups may differ in their experience of and/or responses and susceptibility to traumatic events. This investigation provides important insights for policy and program development to mitigate PTSD among mid-age and older adults, particularly for marginalized groups such as visible minority immigrants. Future investigations that use estimate models such as these as well as longitudinal analyses may better inform mental health practice and policies to both prevent and treat PTSD.