Longitudinal analysis of adjustment to civilian life and self-rated mental health using the Life After Service Survey

Abstract

Introduction:

Transitioning from the military can be a stressful time in a military member’s life. Although evidence shows several factors are associated with a difficult transition, little is known about the long-term association of adjustment to civilian life and long-term mental health.

Methods:

Using the 2013 and 2019 longitudinal components of the Life After Service Survey, this study assessed the association between self-reported adjustment to civilian life in 2013 and self-reported mental health in 2019.

Results:

The logistic regression analysis showed that participants who reported a difficult transition in 2013 were 2.50 times more likely to report fair or poor mental health in 2019 (N = 2,015). In addition, participants who screened positive for posttraumatic stress disorder and those who reported fair or poor mental health in 2013 were 2.75 and 4.61 times more likely, respectively, to report fair or poor mental health in 2019.

Discussion:

The findings emphasize the importance of examining perceptions of the transition experience as a factor in long-term Veteran mental health.

Lay Summary

Transitioning from the military to civilian life can be a stressful time in a military member’s life. Although research has shown several reasons to be related to a difficult transition experience, little is known about what that means for long-term mental health. Using the 2013 and 2019 longitudinal versions of the Life After Service Survey, this study looked at the association between self-reported adjustment to civilian life in 2013 and self-reported mental health in 2019. Results showed that participants who reported a difficult transition in 2013 were 2.50 times more likely to report fair or poor mental health in 2019. In addition, participants who met the criteria for posttraumatic stress disorder and those who reported fair or poor mental health in 2013 were 2.75 and 4.61 times more likely, respectively, to report fair or poor mental health in 2019. Findings highlight the importance of considering how military members adjust to post-service life when examining long-term mental health.

Introduction

Transitioning out of the military represents a significant period in the careers and lives of military members. Many join the military at a young age and spend their emerging and young adult lives immersed in military culture. Military culture distinguishes itself from the civilian world with its strong emphasis on the power hierarchy and chain of command, collectivism, and the adoption of military norms and the warrior identity.^1,2 The military has been described as both a greedy institution and a total institution,^3–5 placing high demands on its members in terms of time commitment, social behaviour, responsibilities, and physical and psychological effort. Thus, for many, leaving the military is leaving not just an occupation but also a unique cultural environment and having to adjust to the civilian world.

The adjustment to post-service life may not always be easy, and in 2019, 39% of Regular Force Veterans of the Canadian Armed Forces (CAF) reported experiencing difficulty with adjusting to civilian life.⁶ This research has been based on the Life After Service Survey (LASS), which was specifically designed to assess the well-being of Veterans in Canada. To date, publications based on the cross-sectional component of the LASS and other studies of adjustment have shown that a difficult adjustment is associated with several different factors, including poor mental health, suicidal ideation and behaviour, and increased alcohol use.⁷ Furthermore, reporting low mastery — the extent to which one feels in control of their life choices — and low social support have been shown to be associated with increased odds of reporting a difficult adjustment to post-service life.^8,9 Veterans may also encounter economic difficulties during their transition to civilian life, including unemployment, job dissatisfaction, and low income.^10,11 Social issues, from establishing social connections and support to housing insecurity and homelessness, also affect their transition experiences.^12–14 Additionally, specific military factors, such as being medically or involuntarily released,¹² holding a lower rank at release,⁶ and serving in the army,¹² have been associated with a challenging transition, including the loss of military identity and difficulties with financial preparedness.

At present, few identified risk or protective factors are associated with long-term outcomes, including mental health. In recent years, however, Canadian researchers have adopted a holistic, lifespan perspective when conceptualizing the military-to-post-service transition.^15–17 This is critical because taking a lifespan perspective emphasizes the impact that specific life stages, events, or factors can have on later health and well-being.¹⁸ The duration and endpoints of the transition period may vary, but it is said to represent a defined life phase.¹⁸ Incorporating a lifespan perspective, it can be postulated that the transition or adjustment to post-service life is a significant event in a military member’s life, which has implications for future well-being. More specifically, taking a lifespan perspective suggests that a difficult transition or adjustment to civilian life has the potential to negatively affect future mental health.

To date, however, there has been little to no research on how the adjustment to post-service life may influence or relate to the mental health and well-being of Canadian Veterans over time. Most studies examining the adjustment of Canadian Veterans to civilian life are correlational or have examined adjustment as the dependent variable in logistic regression models (i.e., what factors are associated with increased odds of reporting a difficult transition). For this reason, the mental health outcomes associated with transition and adjustment to post-service life are not well understood. Therefore, the purpose of this study was to begin to address this gap by assessing the association between perceived adjustment to post-service life and self-rated mental health six years later. We hypothesized that participants who reported a difficult adjustment to civilian life in 2013 would be more likely to report fair or poor mental health in 2019. We further hypothesized that, consistent with the literature, poor mental health in 2013, physical health, low rank, and not being partnered would be associated with increased odds of reporting fair or poor mental health in 2019.

Methods

Study design and sample size

This study was a secondary data analysis using data from the LASS. The LASS was designed to capture the mental and physical health of Regular Force military Veterans in Canada and was administered by Statistics Canada in 2013, 2016, and 2019. Data collected as part of the 2013 and 2019 longitudinal components were included in this study (N = 2,015). Statistics Canada granted permission to access and use the data. The study received approval from the Royal Ottawa Health Care Group Research Ethics Board. Additional information about the LASS can be found on the Statistics Canada website.¹⁹

Participants in the longitudinal version of the LASS are Regular Force Veterans who released from the military between 1998 and 2012. For summary statistics, the data were weighted to be representative of the Veteran population in Canada, using weights provided by Statistics Canada (N_weighted = 48,700). To facilitate analysis and interpretation, and consistent with previous research, mental health, physical health, experience of pain, adjustment to civilian life, and marital status variables were dichotomized.

Measures

Mental and physical health

Self-rated mental and physical health were measured using single items, which asked participants to rate their mental and physical health on a five-point scale (poor, fair, good, very good, excellent). To facilitate analysis and interpretation, self-rated mental and physical health were dichotomized into positive (good, very good, excellent) and negative (fair or poor). Self-rated mental health reported in 2019 was included as the outcome of interest, and self-rated mental health reported in 2013 was included as a control variable. The presence of posttraumatic stress disorder (PTSD) was assessed using a derived dichotomous variable indicating whether participants screened positive for PTSD (yes or no; see Statistics Canada¹⁹). Experience of pain was assessed by asking participants to rate how often (sometimes, often, or never) pain limited certain activities while they were at work, school, or home. Response options were grouped so that participants who selected sometimes or often were coded as having pain that limits activity. Heavy episodic drinking was measured by asking participants to identify how frequently they consumed four (females) or five (males) alcoholic beverages on one occasion; these responses were grouped into weekly, monthly, or never.

Transition

The military-to-civilian transition was captured using a single item asking participants to rate their level of adjustment to civilian life since release on a five-point scale (very difficult, moderately difficult, neither difficult nor easy, easy, very easy). Response options were grouped into positive (neither difficult nor easy, easy, very easy) and negative (very difficult, moderately difficult) categories.

Socio-demographics

Socio-demographic variables of interest, including sex, marital status, rank, and element of service, were included in the analyses. Response options for marital status were grouped into married or common-law and widowed, divorced, or single categories.

Analyses

The longitudinal data set was prepared for analysis according to the guidelines provided by Statistics Canada. This included applying weights to the data set, using bootstrapping, and rounding counts to base 10. Prevalence estimates and logistic regression models were performed in SAS Enterprise Guide, version 7.1 (SAS Institute Inc., Cary, NC). Significance was determined using both confidence intervals and p < 0.05. To ensure that results were reliable, coefficients of variation were calculated, and all results presented in this article are within the acceptable range.²⁰ Refusal to answer or non-responses were treated as missing and not included in analyses.

Sex and rank at release were not collected in 2013, so values from 2019 have been included because these are constructs that would not have changed across time. For parsimony, and to focus on military factors, income, education, and age were not included in the logistic regression model because preliminary analysis indicated no change in estimates when these variables were removed.

Logistic regression analysis with sex-stratified data was conducted, but the small number of female participants (n = 241) resulted in low cell counts and unreliable estimates (i.e., wide confidence intervals, coefficient-of-variation values exceeding guidelines).

Results

Sample characteristics

The frequency of general and military-specific characteristics is presented in Table 1. The majority of the sample were male and were either married or in a common-law relationship. Almost half of the sample were senior non-commissioned members (NCMs) and had served in the army. A very small proportion served in the Gulf War, Afghanistan War, or both.

Table 1.

Frequencies of socio-demographic variables

Variable	Weighted counts	Frequency (%)	95% CIs
Sex 2019
Male	42,300	86.86	85.02−88.56
Female	6,400	13.14	11.44−14.97
Marital status 2013
Married or common−law	38,500	79.06	76.72−81.33
Widowed, divorced, or single	10,200	20.94	18.67−23.28
Military rank 2019
Officer	6,100	19.18	18.17−20.30
Junior NCM	10,500	33.02	31.27−34.49
Senior NCM	15,200	47.80	45.73−50.03
Environment 2013
Navy	15,500	31.83	29.51−34.01
Army	25,200	51.73	49.11−54.33
Air force	8,000	16.43	14.62−18.42
Gulf War Veteran 2013
Yes	2,000	4.11	3.11−5.11
No	46,700	95.86	94.84−96.89
Afghan War Veteran 2013
Yes	3,000	6.17	5.05−7.41
No	45,600	93.83	92.59−94.95

CI = confidence interval; NCM = non-commissioned member.

The frequency of health and well-being variables is presented in Table 2. Self-reported negative mental health increased from 2013 to 2019, with 14.99% reporting fair or poor mental health in 2013 and 24.78% reporting fair or poor mental health in 2019. Furthermore, 30.25% reported a difficult adjustment to civilian life. More than one-third (36.21%) reported that pain limited their activity.

Table 2.

Frequency of health variables

Variable	Weighted counts	Frequency (%)	95% CIs
Self-rated mental health 2019
Positive	26,100	75.22	72.42−78.04
Negative	8,600	24.78	21.95−27.58
Self-rated mental health 2013
Positive	41,400	85.01	82.96−87.15
Negative	7,300	14.99	12.85−17.04
Adjustment to civilian life 2013
Easy	33,900	69.75	67.31−72.30
Difficult	14,700	30.25	27.70−32.69
Self-rated physical health 2013
Positive	39,800	81.89	79.59−84.11
Negative	8,800	18.11	15.89−20.41
Experience of pain 2013
No pain	31,000	63.79	61.09−66.39
Pain limits activity	17,600	36.21	33.61−38.91
PTSD 2013
Yes	7,200	14.97	12.74−17.10
No	40,900	85.03	82.90−87.26
Heavy episodic drinking 2013
Never	31,000	72.26	69.50−74.91
Monthly	7,600	17.72	15.47−20.19
Weekly	4,300	10.02	8.23−11.79

CI = confidence interval; PTSD = posttraumatic stress disorder.

Logistic regression

A single logistic regression was conducted with self-rated mental health in 2019 entered as the outcome variable. Adjustment to civilian life, self-rated mental health in 2013, self-rated physical health, PTSD, activity-limiting pain, marital status, sex, exposure to the Afghanistan or Gulf War, rank, environment, and heavy episodic alcohol use were entered into the regression model as predictor variables.

The results are presented in Table 3. As can be seen, the most important predictor of mental health in 2019 was self-rated mental health in 2013, with those reporting negative mental health in 2013 being 4.61 times more likely to report negative mental health in 2019. Furthermore, participants with a difficult transition were 2.52 times more likely to report negative mental health in 2019. Having a PTSD diagnosis, pain that limits activity, and being single, separated, or divorced were all associated with increased odds of having negative mental health in 2019. Conversely, being a senior NCM at the time of release was associated with decreased odds of reporting negative mental health in 2019. No other variables were significantly associated with self-rated mental health in 2019.

Table 3.

Logistic regression predicting reported fair or poor mental health in 2019

Variable	aOR	95% CI	p-value
Sex 2019
Male (ref.)
Female	0.86	0.41−1.78	0.68
Marital status 2013
Single, separated, divorced, widowed	1.98	1.07−3.69	0.03
Married or common-law (ref.)
Rank at release 2019
Officer (ref.)
Junior NCM	0.49	0.29−0.82	0.40
Senior NCM	0.34	0.19−0.61	<0.01
Environment 2013
Army (ref.)
Air force	0.74	0.38−1.45	0.23
Navy	1.14	0.66−1.97	0.26
Gulf War Veteran 2013
Yes	0.71	0.25−1.96	0.07
No (ref.)
Afghanistan War Veteran 2013
Yes	1.98	0.93−4.21	0.50
No (ref.)
Self-rated mental health 2013
Positive (ref.)
Negative	4.61	2.16−9.85	<0.001
Adjustment to civilian life 2013
Easy (ref.)
Difficult	2.52	1.55−4.08	<0.001
Self-rated physical health 2013
Positive (ref.)
Negative	1.85	0.94−3.78	0.07
Pain 2013
Limits activity	1.75	1.09−2.18	<0.01
Does not limit activity (ref.)
PTSD diagnosis 2013
Yes	2.75	1.44−5.27	<0.01
No (ref.)
Heavy episodic alcohol use 2013
Weekly (ref.)
Monthly	1.30	0.50−3.39	0.62
Never	1.16	0.55−2.47	0.95

Note: Boldface indicates statistical significance.

aOR = adjusted odds ratio; CI = confidence interval; ref. = reference category; NCM = non-commissioned member; PTSD = posttraumatic stress disorder.

Discussion

The purpose of this study was to examine the longitudinal association between adjustment from military to civilian life and self-reported mental health. The results supported the hypothesis that a perceived difficult adjustment to civilian life would be associated with future negative mental health. Specifically, those who perceived their military to civilian adjustment to be difficult in 2013 were 2.50 times more likely to state that their mental health was fair or poor six years later. This extends the previous understanding of self-reported adjustment to indicate that it is a risk factor for reporting poor mental health several years later. Understanding the potential for long-term negative outcomes associated with a difficult transition is an important contribution to the understanding of Veteran health and well-being. Recent research on the mental health of British Veterans who served in the war in Afghanistan led Brewer and Herron to suggest that Veterans can become stuck in a negative transition phase, which then leads to a more permanent negative transition experience.²¹ Taken with the findings of the present study, support during the transition phase to mitigate difficulty may be essential in preventing long-term mental health challenges.

The results further indicate the significant roles of mental and physical health in the long-term perception of mental health of CAF Regular Force Veterans. The most important predictor of mental health across time was mental health reported in 2013. Participants who rated their mental health as fair or poor in 2013 were more than 4.61 times more likely to report fair or poor mental health six years later in 2019. A recent study by Vogt and colleagues found an overall decline in mental health and well-being for a sample of U.S. Veterans across a three-year period.²² Of note were an increase in reported mental health diagnoses and a decrease in community involvement, with more pronounced effects in these areas for female Veterans. A decline in Veteran mental health has been reported in Canada as well, including in other studies using LASS data. For example, 17.0% of participants in the 2013 LASS cross-sectional component reported fair or poor mental health, and that number increased to 28.2% of participants in the 2019 LASS.⁶ The present study’s results begin to elucidate potential reasons why Regular Force Veterans report fair or poor mental health, such as experiencing a difficult transition.

Findings also demonstrate specific risk factors for reporting poor mental health in later years. Both activity-limiting pain and diagnosed PTSD in 2013 were associated with almost two and three times the odds, respectively, of reporting fair or poor mental health in 2019. Previous research indicates a positive association among chronic pain, diagnosed PTSD, and poor mental health.²³ However, the findings of the present study show an independent contribution of each variable to the prediction of increased likelihood of reporting fair or poor mental health across a six-year period. In other words, while controlling for self-reported mental and physical health, activity-limiting pain, diagnosed PTSD, and additional socio-demographic factors in 2013, a self-rated difficult adjustment to civilian life in 2013 still increased the likelihood of reporting fair or poor mental health in 2019. This finding suggests a need to understand adjustment to civilian life and ways to either prevent difficult adjustment or mitigate negative outcomes for those who report a difficult adjustment.

Identity factors and intersectionality

Identity factors, such as marital status, have been recognized as important aspects of military members’ and Veterans’ lived experience and health outcomes. In particular, being divorced or separated was associated with higher levels of post-deployment stress in a sample of U.S. military Veterans.²⁴ In the present study, those who were not partnered in 2013 were twice as likely to report fair or poor mental health in 2019. One potential reason for this outcome may be the social support and caregiving that partners often provide to Veterans. In support of this contention, Hachey and colleagues reported lower odds of reporting an easier adjustment in the cross-sectional LASS for those who reported being dissatisfied with social relationships, including those with family members.⁹

Furthermore, the present study found that senior NCMs who reported a difficult transition were less likely to report fair or poor mental health in 2019 compared with officers. On the basis of work with the cross-sectional 2019 LASS,²⁵ previous transition studies with CAF Veterans,²⁶ and international research,²⁷ lower rank is generally associated with more adverse outcomes, including reporting fair or poor mental health. Interestingly, McCuaig Edge and colleagues found that being a junior NCM was uniquely associated with transition difficulties for those who were medically released.²⁸ This would suggest that there is a more nuanced association among rank, adjustment to civilian life, and health that needs further investigation. Future research should endeavour to capture changes across time to assess factors that interact with rank to alter the association between adjustment and perceived mental health.

Limitations and future directions

This study provides valuable insight into factors affecting mental health. Yet, as with all research, this study had limitations. The primary limitation is that the study relied on secondary data, which can be a valuable method to answer research questions expeditiously and without placing additional strain on participants. However, it limited what questions could be answered and what could be analyzed on the basis of the variables included in the dataset. An important caveat is the self-report and imprecise nature of the question assessing transition. Although research lacks a standardized approach to defining and measuring successful transition, there is some consensus that transition involves multiple dimensions, covering various domains of well-being, such as physical health, mental wellness, social integration, disability management, family, and community.^29,30 Future research should develop a more comprehensive measure of transition to capture the multi-dimensional nature of the transition experience.

As has been pointed out in previous publications using the LASS, the sample captures only a segment of the Veteran population (e.g., only those released between 1998 and 2018) and cannot be generalized to the entire Canadian Veteran population. Of importance regarding the purposes and outcomes of the present study, it does not include those who transitioned under recently introduced changes to the process within the CAF (i.e., the CAF Transition Group), which may change the transition experience and associated outcomes. The sample also does not include Royal Canadian Mounted Police Veterans, who fall under the jurisdiction of Veterans Affairs Canada.

Despite application of sex- and gender-based analy­sis plus being a now-fundamental component of the collection of government-produced data, the small number of women who participated in the LASS makes it difficult to undertake more complex analyses. This is a significant limitation of any work conducted with the LASS and of this study. Depending on the variables of interest, cell counts can become quite low when stratifying and disaggregating the data, and caution is needed to ensure that low cell counts and unreliable estimates (e.g., those with wide confidence intervals) do not lead to misinterpretation of the results or perpetuate misinformation. Intersectionality and identity factors are an important part of understanding the well-being of Veterans in Canada, particularly for women Veterans. For example, in a recent study of U.S. military members, women (n = 666) were almost one-and-a-half times more likely to screen positive for PTSD compared with men.³¹ If screening positive for PTSD is an indicator of later mental health in the adjustment-to-civilian-life phase, as suggested by the results of the present study, then women may need additional, targeted supports. Consequently, oversampling and targeted recruitment should be a priority in study designs that include marginalized populations.

Conclusion

Leaving the military is more than just leaving an occupation, and adjusting to post-service life has its challenges for some Veterans in Canada. This study provides important evidence for a longitudinal association between a difficult adjustment to post-service life and mental health. Although additional research is needed, particularly research using longitudinal designs, this study highlights the importance of considering the transition experience in mental health outcomes and points to a need for mitigation strategies to better support positive mental health among Canadian Veterans.

Biographies

Kate Hill MacEachern, PhD, is a senior research associate at the Atlas Institute for Veterans and Families. Her work is anchored in social determinants of health with a particular interest in how sex- and gender-specific issues relate to mental health and well-being. She is a co-founder of the Athena Project, a community-engaged research initiative conducting applied research in support of the mental health of women Veterans. Hill MacEachern received her PhD in psychology from Carleton University.

Adelina Gregory, MS, is a research associate at the Atlas Institute for Veterans and Families. She completed her master’s in neuroscience in 2019, where she focused on clinical research with a particular focus on Veterans experiencing psychiatric conditions, such as posttraumatic stress disorder. This work sparked a deep passion for supporting Veterans and their families, a commitment she continues through her role at the Atlas Institute.

Sara Rodrigues, PhD, is Director, Applied Research, Atlas Institute for Veterans and Families, and Adjunct Professor, School of Psychology, University of Ottawa. Her work has been published in the European Journal of Psychotraumatology, Human Studies, and International Journal of Feminist Approaches to Bioethics, among others, and she is co-editor of On the Politics of Ugliness. She is a Co-Chair of the Canadian Moral Injury Research Community of Practice. She received her PhD in social and political thought from York University.

Funding Statement

No funding was received for this article.