Identifying Key Correlates of Social Well-Being among Canadian Armed Forces Veterans: An Analysis of the 2016 Life after Service Study

Abstract

Many Veterans experience disruptions to their social connections during military to civilian transition. As low social support has been associated with difficulties adjusting to civilian life, there is value in better understanding social support in Veteran populations. The purpose of this study was to identify correlates of social support among recent Canadian Armed Forces (CAF) Veterans. Data were collected as part of the 2016 Life After Service Survey, which was administered to a sample of CAF Regular Force Veterans. This study focuses on more recently released Veterans, 5 years prior to the survey (n = 1,723) to better reflect the impact of transitioning to civilian life. Social support was measured using the 10-item Social Provision Scale. Regression models examined the relative associations of individual, release, and post-transition characteristics (i.e., family/household composition and main activity) with social support. Models explained up to 23% of the variance in social support. Lower social support was associated with being: male, older, of noncommissioned rank at release, and released involuntarily (p ≤ .05). Family/household composition, most notably living with a partner, was also associated with greater social support, especially among Veterans who were not in the workforce. Veterans’ main activity in the last year explained the most variance in social support, with a strong association noted for satisfaction with one’s main activity (p ≤ .001). Social support is an important and modifiable factor in the transition to civilian life. Results point to specific subgroups who may be at risk for low social support after military service.

Supplemental data for this article is available online at https://doi.org/10.1080/21635781.2021.2007182 .

Keywords:

• Social support
• veterans
• transition to civilian life
• life-satisfaction
• employment

Life after military service can be challenging, with over a third of Canadian Armed Forces (CAF) Veterans reporting difficulty adjusting to civilian life (Statistics Canada, 2020). While most Veterans successfully transition to civilian life (Vogt et al., 2020), those reporting unsuccessful transitions can face life-long issues (Castro et al., 2014). Veterans experiencing difficult adjustment report more physical health complaints (MacLean et al., 2014; Thompson et al., 2011; Van Til, 2014), poor mental health (Ahern et al., 2015; Bergman et al., 2016; Black & Papile, 2010; MacLean et al., 2014; Schinka et al., 2012; Thompson et al., 2011, 2013, 2019; Yeung et al., 2019), and reduced quality of life (Thompson et al., 2011; Yeung et al., 2019). In addition to physical and mental health issues, Veterans reporting difficult adjustment to civilian life can face economic insecurity (such as unemployment, dissatisfaction with their job, low income) (MacLean et al., 2019), and social problems (e.g., homelessness) (Ahern et al., 2015; MacLean et al., 2014; Schinka et al., 2012).

Acknowledging the range of issues that Veterans may face after service, researchers have underlined the importance of considering multiple domains of Veterans’ lives to better understand their well-being, including their employment or engagement in other meaningful activities, their finances, their health, their sense of community belonging, their life skills and resources, and their cultural environments (Thompson et al., 2016). Similarly, Vogt et al. (2019) proposed that comprehensive measures including military Veterans’ status, functioning, and satisfaction with respect to key aspects of their lives (e.g., satisfaction with work) offer a broad assessment of well-being while providing granular insight into specific factors that can contribute to well-being.

In recent years, an increasing number of studies on Veterans’ well-being have focused on social well-being (Hinojosa & Sberna Hinojosa, 2011; Kintzle et al., 2018; Lee et al., 2020). In the broader social psychological literature, social well-being has been defined as “the appraisal of one’s circumstances and functioning in society” (p. 122) and is proposed to encompass multiple dimensions, including the extent to which they are integrated in their society or community; accepting of other people; see themselves as a valuable member of society; see society as having a potential for progress or growth; and see society as a coherent and predictable entity (Keyes, 1998). In addition to viewing social well-being as being composed of similar elements as those listed above (i.e., satisfaction with relationships, performance in social roles and adjustment to one’s environment), others have considered the role of social support as yet another aspect of social well-being (Larson, 1993).

Reflecting one’s perception of being cared for, valued, and part of a mutually supportive social network (Taylor, 2011), social support has often been recognized for its potential to promote resilience and facilitate post-deployment adjustment among military personnel and Veterans (Cunningham et al., 2014; Pietrzak et al., 2009; Sippel et al., 2015). Similarly, many studies have reported an association between low social support and difficult adjustment to civilian life (MacLean et al., 2014; Pietrzak et al., 2009, 2010). As Veterans transition to civilian life, their access to social support may be especially vulnerable due to a disruption to social networks and disconnection (Ahern et al., 2015). Further, there is evidence to suggest that those who fail to develop social networks outside of the military (i.e., they continue to rely on social networks in which most members are still in the military) are at increased risk of negative transition outcomes (Hatch et al., 2013). Variation in social support has also been associated, at least in part, with measures of subjective well-being, such as affective and cognitive evaluations of one’s life (i.e., life satisfaction) and one’s individual circumstances in specific domains (e.g., work satisfaction; Diener, 2000). Veterans differ in how they perceive life events and the support they receive from others, and these subjective variations may shape well-being post-release. Consequently, there is value in understanding the factors associated with social support among recent Veterans to inform the development of interventions that can bolster social support and identify those who may benefit most from these interventions.

Despite the considerable number of studies having examined the association of social support with mental health and well-being in Veteran populations (e.g., Cunningham et al., 2014; Hachey et al., 2016; MacLean et al., 2014; Pietrzak et al., 2009, 2010; Sippel et al., 2015), comparatively little research has focused on the factors associated with social support itself. Nevertheless, a few observations are worth noting. For instance, some findings have suggested that Veterans who transitioned out of the military in an unexpected or unplanned manner, due to illness or injury, experience especially prominent feelings of social isolation or loss of connection (Lee et al., 2020; Williams, Skomorovsky, Wan, & Lee et al., 2020), indicating that transition characteristics may play a role. As well, Veterans in one study expressed an increasing need to turn to family and new relationships for social support as connections to their former military career fade during transition to civilian life (Ahern et al., 2015), suggesting that family characteristics play a role. Finally, participation in the workforce, or in other meaningful activities, has been posited to offer individuals who have retired a means to reshape their identity and forge new friendships, in addition to providing them with opportunities to both receive and give support (Knardahl et al., 2017; Steffens et al., 2016). Based on these observations, the purpose of this study was therefore to identify some of the key correlates of social support among recent CAF Regular Force Veterans. Specifically, an analysis was conducted on data collected as part of the 2016 Life After Service Survey (LASS) to examine the relative extent that Veterans’ release characteristics, family characteristics, and main activity are associated with their perceived social support within the first few years of their release.

Materials and methods

Participants

Survey data was collected between February and March 2016 by trained Statistics Canada interviewers using a Computer Assisted Telephone Interview. Veterans released from Regular Force service between 01 January 1998 and 31 August 2015 were contacted, using a cross-sectional sample across three strata: officers, senior noncommissioned members (NCMs), and junior NCMs. Statistics Canada interviewers obtained a 73% response rate on the entire population (Van Til et al., 2017). Those who released at entry ranks (e.g., Officer Cadet, Naval Cadet or Private Recruit), resided in the Territories or outside Canada, re-enrolled in the CAF, resided in an institution, or were no longer alive at the time of the study were excluded from the sample. Additional sampling frame details are outlined in Veterans Affairs Canada (VAC) technical reports (Van Til, 2014, 2016). To ensure that participants had relatively recent experiences with transition to civilian life, only Veterans having been released less than 5 years prior to the survey administration were included in the present analysis.

Measures

Individual characteristics

Individual factors, which included sex, age, environment, and rank at release were obtained from Department of National Defense human resources data. Deployment was assessed using a dichotomous questionnaire item reflecting whether the participant participated in an international or domestic deployment while in uniform.

Release characteristics

Type of release was dichotomized into two categories: “voluntary”, which included those who released after fixed service, those who reached retirement age, and those who released for other voluntary reasons; and “non-voluntary”, which included those who released under release codes such as medical, unsatisfactory service, or sentenced to dismissal (Van Til, 2014). Time since release was calculated for each participant using the most recent release date from human resources data, and categorized into those with less than one, one to three, and four (but less than five) full years since release at the time of the survey.

Post-transition characteristics

Family/household

Family/household composition was assessed using three LASS questionnaire items: marital status, household size, and number of children under 18 residing with the participants (Van Til, 2016). From these three variables, five categories of family/household composition were created: those living alone, those living with a partner without children, those living with a partner with children, those living with children (no partner), and those living with other adults (i.e., extended family or friends).

Main activity

Main activity factors included two questionnaire items that have been used in previous LASS cycles: main activity in last 12 months and satisfaction with this main activity (Hachey et al., 2016; Van Til, 2014). Participants were classified into three mutually exclusive categories: those in the work force (worked at a job or ran a business), those who were retired (not looking for work), and those not in the work force (including those who attended school or training, looked for work, cared for family member, were disabled, or were on disability). This categorization separated those who were not active in the workforce for reason of circumstance from those who were working and those who had chosen to retire (i.e., willingly self-excluded from the workforce). Participants rated their satisfaction with their main activity in the last 12 months on a 5-point scale (1 = very satisfied to 5 = very dissatisfied). For the analysis, participants were classified based on whether they were satisfied (i.e., they reported being satisfied or very satisfied) or not satisfied (i.e., they reported being neutral, dissatisfied, or very dissatisfied).

Social support

The availability of social support was measured using the Social Provisions Scale (SPS), a validated 10-item scale which combined scores across five dimensions of social support: emotional attachment, social integration, reassurance of worth, material assistance, and advice/guidance (Caron, 2013). Participants rated their agreement with each of the items on a 4-point scale (1 = strongly agree to 4 = strongly disagree). A total score was re-calculated from these ratings, with higher scores indicative of greater social support. The psychometric properties of this scale in the weighted population are summarized in Table 1.

Table 1. Psychometric characteristics of Social Provisions Scale in sample.

# of items	Alpha (Weighted)	Correlation (Weighted)	CATEGORY & item description	M (Weighted)	Bootstrapped 95% CI
10	0.93	NA	SPSDCON- Social Provisions Scale (total)	34.05	33.65	34.45
2	0.80	0.67	SPSDATT - Attachment	6.91	6.82	7.01
			I have close relationships that provide me with a sense of emotional security and wellbeing.	1.61	1.56	1.67
			I feel a strong emotional bond with at least one other person.	1.47	1.43	1.52
2	0.84	0.72	SPSDGUI- Guidance	7.00	6.91	7.08
			There is someone I could talk to about important decisions in my life.	1.49	1.45	1.54
			There is a trustworthy person I could turn to for advice if I were having problems.	1.51	1.47	1.56
2	0.78	0.64	SPSDALL- Reliable Alliance	7.00	6.92	7.08
			There are people I can depend on to help me if I really need it.	1.52	1.48	1.57
			There are people I can count on in an emergency.	1.48	1.43	1.52
2	0.79	0.66	SPSDINT- Social Integration	6.48	6.38	6.58
			There are people who enjoy the same social activities I do.	1.67	1.61	1.72
			I feel part of a group of people who share my attitudes and beliefs.	1.86	1.80	1.92
2	0.80	0.68	SPSDWOR - Reassurance of Worth	6.58	6.49	6.68
			I have relationships where my competence and skill are recognized.	1.72	1.67	1.78
			There are people who admire my talents and abilities.	1.70	1.65	1.75

Note. CI = confidence interval.

Analysis

Statistical analyses were performed using Stata/SE Version 14.2 (StataCorp, 2015). All estimates were weighted using sample weights developed by Statistics Canada to incorporate the unequal probabilities of selection, eligibility, non-response, and sharing (Hachey et al., 2016; Van Til, 2014). Categorical variables were summarized with weighted frequencies and percentages, while summary statistics for continuous variables included weighted means and standard errors. The confidence intervals around estimates were bootstrapped to account for population weights and the stratified nature of the sample. Linear regression was used to test the association of individual, release, family/household, and main activity variables with social support using the “svy” function in Stata. Hierarchical regressions were used to examine the sequential contribution of variables. Four models were run, each adding additional variable groups (1- individual, 2- release, 3- family/household, and 4- main activity) to a model predicting social support. Variables with beta estimates with a p ≤ .05 were considered significantly associated with social support in the regression models.

Results

A total of 1,273 records were included in the sample, representing a population of 16,329 Regular Force Veterans who had released from the CAF between 2011 and 2015. There was very little missing data: Social Support was missing 3.5% of responses, family/household composition was missing 0.06% of responses, and main activity was missing 0.16% of responses.

The average age of participants was 45 years. The majority of participants were male, and had primarily released voluntarily, from the Army, and as junior NCMs. At the time of the survey, most participants were living with their partner only (no children), were working in the last 12 months, and were satisfied with their main activity). A summary of the characteristics of the sample is included in Table 2.

Table 2. Demographic characteristics of participants.

			Weighted count	Weighted %
			n = 16,328.81	Estimate	Bootstrapped 95% CI
Individual	Sex	Female	2155.403	13.2	11.0	15.7
		Male	14174.28	86.8	84.3	89.0
	Age	<30	3674.432	22.5	19.5	25.7
		30-39	3527.671	21.6	18.7	24.9
		40-49	3968.873	24.3	21.6	27.2
		50-59	4344.793	26.6	23.9	29.5
		60+	833.0753	5.1	4.1	6.3
	Environment at release	Sea	3218.155	19.7	17.1	22.6
		Land	9671.392	59.2	55.7	62.6
		Air	3463.616	21.2	18.5	24.0
	Rank at release	Officer	2516.177	15.4	14.4	16.5
		Senior NCM	4607.826	28.2	26.5	30.0
		Junior NCM	9216.217	56.4	54.1	58.6
	Deployment	Not deployed	4526.681	27.7	24.6	31.1
		Deployed	11815.85	72.3	68.9	75.4
Transition	Type of release	Involuntary	6668.274	40.8	38.3	43.5
		Voluntary	9676.128	59.2	56.5	61.7
	Years since release	≤1	2909.554	17.8	15.5	20.4
		1 to <4	11540.85	70.6	66.9	74.1
		4 to <5	1896.346	11.6	8.6	15.3
Family	Family/household composition	Partner only	4405.698	27.0	23.9	30.2
		Children only	620.0612	3.8	2.5	5.7
		Partner & children	4944.172	30.3	27.0	33.9
		Living alone	2643.419	16.2	13.3	19.6
		Living with other adults	3720.367	22.8	20.0	25.9
Main activity	Activity in last year	Working	9175.048	56.1	52.6	59.5
		Retired	2175.323	13.3	11.5	15.5
		Not in workforce	5005.184	30.6	27.5	33.8
	Satisfaction with main activity	Unsatisfied/Neutral	5496.224	33.6	30.4	36.9
		Satisfied	10862.25	66.4	63.1	69.6

Note. Based on a sample of n = 1,273 Veterans who released less than five years prior to completing the survey. CI = confidence interval; NCM = noncommissioned member.

Table 3 summarizes the results of the SPS by individual, release, and post-transition characteristics. As shown, the weighted total social support score had a mean of 34.1 out of 40. When weighted, the mean score for each of the five dimensions of social support (summed to create the overall score) ranged from 6.5 to 7.0 out of 8.00.

Table 3. Weighted Social Provisions Scale scoring across demographic variables.

			SPSDCON			SPSDATT			SPSDGUI			SPSDALL			SPSDINT			SPSDWOR
			M	Bootstrapped 95% CI		M	Bootstrapped 95% CI		M	Bootstrapped 95% CI		M	Bootstrapped 95% CI		M	Bootstrapped 95% CI		M	Bootstrapped 95% CI
	All		34.05	33.65	34.45	6.91	6.82	7.00	7.00	6.91	7.08	7.00	6.92	7.08	6.48	6.38	6.58	6.58	6.49	6.68
Individual	Sex	Female	35.01	34.00	36.03	7.17	6.99	7.36	7.18	6.99	7.38	7.16	6.96	7.36	6.72	6.44	7.00	6.72	6.42	7.02
		Male	33.90	33.46	34.34	6.87	6.77	6.98	6.97	6.87	7.06	6.98	6.89	7.07	6.44	6.33	6.56	6.56	6.46	6.67
	Age	<30	35.14	34.24	36.03	7.15	6.98	7.33	7.17	6.98	7.37	7.26	7.09	7.43	6.65	6.38	6.93	6.80	6.56	7.04
		30-39	33.79	32.75	34.83	6.78	6.53	7.03	7.01	6.82	7.20	6.98	6.77	7.20	6.46	6.21	6.71	6.48	6.23	6.73
		40-49	33.52	32.71	34.33	6.80	6.62	6.98	6.86	6.68	7.03	6.82	6.64	7.00	6.39	6.20	6.58	6.52	6.34	6.70
		50-59	33.82	33.22	34.42	6.92	6.79	7.05	6.97	6.85	7.10	6.96	6.84	7.08	6.42	6.25	6.58	6.55	6.39	6.71
		60+	34.06	32.97	35.15	6.94	6.70	7.19	6.93	6.70	7.16	7.05	6.81	7.29	6.54	6.26	6.81	6.52	6.29	6.75
	Environment at release	Sea	34.62	33.72	35.51	7.01	6.81	7.21	7.07	6.87	7.27	7.09	6.90	7.28	6.65	6.44	6.87	6.73	6.52	6.93
		Land	33.94	33.37	34.52	6.91	6.78	7.04	7.00	6.88	7.11	6.99	6.87	7.11	6.42	6.27	6.57	6.54	6.39	6.69
		Air	33.82	33.05	34.58	6.83	6.67	7.00	6.92	6.75	7.09	6.95	6.80	7.10	6.48	6.29	6.66	6.57	6.41	6.73
	Rank at release	Officer	35.68	35.12	36.24	7.19	7.07	7.32	7.24	7.12	7.37	7.30	7.18	7.41	6.89	6.75	7.03	7.01	6.88	7.13
		Senior. NCM	33.99	33.36	34.61	6.92	6.77	7.06	6.97	6.82	7.12	6.94	6.81	7.07	6.50	6.35	6.66	6.63	6.50	6.77
		Junior NCM	33.63	33.01	34.25	6.84	6.70	6.97	6.94	6.82	7.07	6.95	6.83	7.08	6.36	6.19	6.52	6.44	6.28	6.60
	Deployment	Not deployed	35.13	34.38	35.89	7.16	7.01	7.32	7.16	7.00	7.33	7.23	7.08	7.37	6.66	6.44	6.88	6.75	6.54	6.96
		Deployed	33.64	33.19	34.08	6.82	6.71	6.93	6.93	6.84	7.02	6.92	6.82	7.01	6.41	6.30	6.52	6.52	6.41	6.63
Release	Type of release	Involuntary	32.55	31.92	33.17	6.62	6.48	6.77	6.75	6.61	6.88	6.74	6.62	6.87	6.11	5.96	6.27	6.25	6.09	6.40
		Voluntary	35.07	34.52	35.61	7.12	7.00	7.23	7.17	7.06	7.28	7.18	7.07	7.29	6.73	6.59	6.87	6.81	6.68	6.95
	Years since release	<1	34.64	33.88	35.40	7.05	6.89	7.21	7.10	6.93	7.27	7.12	6.96	7.27	6.63	6.42	6.83	6.70	6.53	6.87
		1 to <4	33.82	33.37	34.26	6.86	6.76	6.96	6.95	6.86	7.04	6.93	6.83	7.03	6.45	6.35	6.56	6.55	6.45	6.66
		4 to <5	34.56	32.75	36.36	7.03	6.61	7.44	7.13	6.75	7.50	7.26	6.93	7.59	6.41	5.87	6.95	6.60	6.08	7.13
Family	Family/household composition	Partner only	34.80	34.08	35.52	7.15	7.01	7.29	7.11	6.96	7.26	7.16	7.03	7.29	6.56	6.33	6.80	6.73	6.54	6.91
		Children only	32.36	29.77	34.96	6.41	5.72	7.09	6.51	5.82	7.20	6.69	6.22	7.15	6.36	5.80	6.93	6.41	5.92	6.90
		Partner & children	34.51	33.82	35.20	7.07	6.93	7.21	7.14	7.00	7.28	7.04	6.89	7.19	6.53	6.36	6.70	6.69	6.52	6.85
		Living alone	32.98	31.72	34.23	6.57	6.31	6.83	6.73	6.50	6.95	6.83	6.59	7.06	6.33	6.02	6.64	6.37	6.04	6.70
		Living with other adults	33.58	32.75	34.42	6.75	6.52	6.99	6.94	6.76	7.12	6.94	6.76	7.13	6.44	6.27	6.61	6.46	6.26	6.65
Mainactivity	Activity in last year	Working	35.02	34.44	35.59	7.10	6.98	7.22	7.16	7.05	7.28	7.12	7.00	7.24	6.72	6.56	6.87	6.87	6.74	7.01
		Retired	34.01	33.11	34.90	6.95	6.78	7.12	6.96	6.77	7.16	7.00	6.79	7.21	6.57	6.36	6.79	6.54	6.33	6.74
		Not in work force	32.24	31.51	32.98	6.56	6.35	6.76	6.70	6.54	6.87	6.78	6.63	6.93	5.99	5.82	6.17	6.07	5.87	6.26
	Satisfaction with main activity	Unsatisfied/Neutral	31.47	30.70	32.24	6.40	6.21	6.59	6.63	6.47	6.78	6.59	6.42	6.76	5.84	5.66	6.03	5.96	5.78	6.14
		Satisfied	35.32	34.90	35.73	7.17	7.08	7.26	7.18	7.09	7.28	7.21	7.12	7.29	6.80	6.68	6.92	6.89	6.78	7.00

Note. All values are weighted. SPSDCON = Social Support overall score; SPSDATT = Attachment; SPSDGUI = Guidance; SPSDALL = Reliable Alliance; SPSDINT = Social Integration; SPSDWOR = Reassurance of Worth; NCM = noncommissioned member.

Table 4 presents the results of the hierarchical regressions. The addition of each group of variables to the hierarchical regression models explained increasingly more model variance (R²): the model containing individual variables explained 7% of the variance in social support, while the models with release variables, family/household variables, and main activity variables added in sequence explained 10%, 12%, and 23% of the variance in social support, respectively.

Table 4. Results of hierarchical regression models of social support.

		Model 1(β)	Model 2(β)	Model 3(β)	Model 4(β)
Individual	Sex (reference Female)
	Male	−1.11*	−1.280*	−1.42**	−1.40**
	Age	−0.08***	−0.053*	−0.06*	−0.07**
	Environment at release (reference Sea)
	Land	−0.52	−0.315	−0.45	−0.54
	Air	−0.68	−0.65	−0.71	−0.68
	Rank at release (reference Officer)
	Senior NCM	−1.42**	−1.077*	−0.99*	−0.96*
	Junior NCM	−3.59***	−2.730***	−2.50***	−1.62**
	Deployment history (reference Not deployed)
	Deployed	−1.43**	−1.228*	−1.439**	−0.80
Release	Type of release (reference Involuntary)
	Voluntary		1.959***	1.89***	0.96*
	Years since release (reference <1)
	1 to <4		−0.405	−0.46	−0.43
	4 to <5		0.343	0.29	0.20
Family	Family/household composition (reference Living alone)
	Partner only			1.71*	1.71**
	Children only			−0.75	−0.56
	Partner & children			1.68*	1.65**
	Living with other adults			0.27	0.71
Main Activity	Activity in last year (reference Working)
	Retired				−0.16
	Not in work force				−0.63
	Satisfaction (reference Unsatisfied/Neutral)
	Satisfied				1.54***
Model	Model constant	42.35***	39.47***	39.03***	33.51***
	# Records in analysis	1,228	1,228	1,227	1,227
	R-squared	0.07	0.10	0.12	0.23

Note.

* p < .05.

** p < .01.

***p < .001. NCM = noncommissioned member.

In all four models, sex, age, and rank were significantly associated with social support. Specifically, being male, older, and released as an NCM (as opposed to an officer) were associated with lower social support. Voluntary release was associated with higher social support in models 2, 3, and 4, Living with one’s partner or with one’s partner and children, as opposed to living alone, were associated with significantly higher social support in models 3 and 4 (p < .001). While main activity variables accounted for the greatest proportion of variance in social support, satisfaction with one’s main activity, but not the nature of the activity itself, was significantly associated with higher social support in model 4 (p < .001). To further explore this observation, separate regressions of the full model were conducted based on Veterans’ main activity (i.e., working, retired, and not in the work force). While environment at release and time since release were not associated with social support in any of the models, these variables were retained in the models for consistency.

Table 5 presents the results of each regression. Satisfaction with one’s main activity was associated with higher social support in all subpopulations. Sex and age were only significantly associated with lower social support among working Veterans. NCM rank was associated with lower social support among Veterans who were working or retired, but not among those who were not in the workforce. Living with a partner or other adults was associated with higher social support among those who were not in the work force, but not among working or retired Veterans.

Table 5. Modeling factors across main activity subpopulations: predictor βs and model descriptors.

		Working (β)	Retired (β)	Not in work force (β)
Individual	Sex (reference Female)
	Male	−1.62*	−1.46	−1.35
	Age	−0.07*	−0.12	−0.05
	Environment at release (reference Sea)
	Land	−0.62	0.03	−0.88
	Air	−1.32	0.57	−0.77
	Rank at release (reference Officer)
	Senior NCM	−0.78	−0.78	−1.59
	Junior NCM	−1.62*	−2.62*	−1.46
	Deployment history (reference Not deployed)
	Deployed	−0.69	−0.97	−0.93
Release	Type of release (reference Involuntary)
	Voluntary	0.52	1.29	1.10
	Years since release (reference <1)
	1 to <4	−0.49	−1.81*	−0.05
	4 to <5	0.21	−2.39	0.671
Family	Family/household composition (reference Living alone)
	Partner only	1.54	−0.62	3.29***
	Children only	−1.03	−1.28	0.72
	Partner & children	1.33	0.34	3.00***
	Living with other adults	0.35	0.44	1.91*
Main Activity	Satisfaction (reference Unsatisfied/Neutral)
	Satisfied	1.49***	2.47***	1.43***
Model	Model constant	34.68***	34.14***	31.29***
	# Records in analysis	1,257	1,266	1,250
	R-squared	0.16	0.38	0.22

Note:.

* p < .05.

** p < .01.

*** p < .001. NCM = noncommissioned member.

Discussion

The purpose of this study was to examine the relative associations of individual characteristics, release characteristics, and post-transition characteristics related to the family/household and main activity with perceived levels of social support among CAF Veterans. In the final model, results pointed to lower social support among those who were male, older, released as NCMs, transitioned out of the military involuntarily, were living alone, and were not satisfied with their main activity. Relative to all other factors, aspects of Veterans’ main activity explained the greatest proportion of the variance in social support. Specifically, this appeared to be related to whether or not Veterans were satisfied with their main activity more so than to the actual nature of their main activity. In a post-hoc analysis performed to explore the nuances of these associations, satisfaction with main activity retained its strong association with social support, regardless of whether Veterans were working, retired, or not in the work force. The importance of work satisfaction was noted in previous cycles of the LASS, as MacLean and colleagues found that work satisfaction was associated with less difficulty adjusting to civilian life (MacLean et al., 2014). Past analyses have also pointed to groups of Veterans who are at risk of having lower levels of work satisfaction, such as those who were released with a lower rank (MacLean et al., 2018). However, less has been documented about non-working Veterans’ satisfaction with their main activity.

Similar to our findings, past research has pointed to sex differences in social support, with sources of support differing across genders. For instance, being widowed has been found to be more detrimental to men, whereas women have been found to be more likely to benefit from friendships (Gurung et al., 2003). This is also the case for age, where studies have found evidence of weaker social relationships among older adults, including a greater risk of loneliness, social isolation, and lower social support (Valtorta et al., 2018). Likewise, lower rank has been similarly found in other studies to be associated with lower reported support (Pietrzak & Cook, 2013).

Of note, having a history of deployment was associated with lower social support in all models but the final one, in which adjustment was made for main activity variables. It may be that both deployment history and engagement in a satisfying meaningful activity are correlates of social support, but the relatively stronger association of the latter washes out that of the former. Alternatively, this finding may indicate that the association between Veterans’ history of deployment and their social support is confounded by their ability to engage in a satisfying main activity. Specifically, Veterans with a deployment history may experience lower levels of social support in part due to their lower likelihood of being engaged in a meaningful and satisfying activity following their transition out of the military. Additional longitudinal research is required to better understand the pathways involved in these associations.

The circumstances around Veterans’ transition to civilian life, including the reasons for their release, have been found to be associated with transition outcomes. For example, Veterans who involuntary released from the military service mid-career, or who were released medically, have been found to report more difficulty adjusting to civilian life and experience more transition challenges (Lee et al., 2020; MacLean et al., 2016). Similarly, we found that being released involuntarily (including being medically released) was associated with lower social support.

Family/household characteristics emerged as important correlates of social support in our study, with findings generally providing converging evidence on the benefits of living with a partner (Iversen et al., 2005; Pietrzak & Cook, 2013). Similarly, Therrien et al. (2016) found in their study of CAF Regular Force members that being married or in a common law partnership and living with others were associated with positive mental health, and this could be attributed to higher levels of perceived social support. As well, as Veterans adjust to civilian life, many credit their family or their spouse as the most important aspect of their successful transition (Black & Papile, 2010). In fact, CAF Veterans who were satisfied with their family relationships have reported easier transitions to civilian life (Hachey et al., 2016). Notwithstanding evidence pointing to the overall benefits of living with a partner on Veterans’ social support in this study, results of post-hoc analyses indicated that these are especially prominent for those who are not in the workforce and that living with any other adult may also be beneficial for these Veterans. Veterans who are not in the workforce may be particularly reliant on their partners or other adults living in their home for social support, as being away from the workforce can limit their opportunities to form new friendships and engage with other potential sources of support outside of the home. Past research has underlined the role that spouses/partners play as a primary source of social support for CAF personnel with illness or injury (Lee et al., 2016).

In past studies, employment has not only been found to be important for health and well-being; it has also been identified as a key predictor of successful adjustment to civilian life (MacLean et al., 2016; Resnik et al., 2012; Sayer et al., 2011). A review of the published evidence by Waddell and Burton concluded that working is generally good for one’s health and well-being, even among sick and disabled populations (Waddell & Burton, 2006). Work provides economic stability and opportunities for achievement, fulfillment, and friendship (Knardahl et al., 2017). Steffens et al. proposed that participation in the workplace offers retirees opportunities to receive and give meaningful support to others, adding to self-worth (Steffens et al., 2016). Fortunately, as in past research, the majority of CAF Veterans in this study reported that they had been working in the past year (Statistics Canada, 2017, 2020). Interestingly, however, our findings indicated that being satisfied with one’s main activity, rather than the actual nature of the main activity, was among the strongest and most consistent correlates of social support. These findings lend support to the importance of ensuring that Veterans prioritize the extent to which they enjoy or derive meaning from an activity when exploring their options for main activities after their transition out of the military.

Factors that influence well-being exist across the life course of a service member, including pre-service, in-service, and post-service factors. The main strength of this study is that it describes the relative association of social support with a breadth of factors spanning the military career and beyond. This study explored a diverse set of factors in a large, nationally representative sample of CAF Veterans and identified satisfaction with one’s main activity as the strongest predictor of social support. Past research has not only demonstrated that social support is related to more positive experiences in life after military service; social support is also a well-established predictor of morbidity and mortality (Ahern et al., 2015; Black & Papile, 2010; MacLean et al., 2014; Pietrzak et al., 2009, 2010; Ren et al., 1999; Uchino, 2006). Moreover, evidence suggests that interventions are successful at increasing social support in populations of retiree individuals (Heaven et al., 2013) and at improving transitions. For example, bridge employment can mitigate the negative effect of involuntary retirement (Dingemans & Henkens, 2014). Therefore, our findings highlight important and modifiable factors to ameliorate Veterans’ transition to civilian life.

Limitations

While our findings help shed light onto particular groups of Veterans who may benefit from targeted interventions aimed at enhancing social well-being, some limitations must be acknowledged. Aside from limitations inherent to our use of self-reported data, which include the potential for social desirability and recall biases, it is important to note that data were also in part drawn from administrative databases, which were developed for human resource management purposes, and were not validated for quality or appropriateness for use in research. Moreover, although the LASS survey is comprehensive, only a minority of the variance in social support was explained by our models. This suggests that other key factors that contribute to social support may have been overlooked, such as participation in extra-curricular activities or hobbies in addition to one’s main activity, from which Veterans may derive meaning or a sense of purpose as well as social support. Our findings also have limited generalizability, representing only CAF Regular Force Veterans who were released less than 5 years before the survey. Furthermore, the survey method did not include some of the most vulnerable individuals, such as Veterans who were living in an institution or those who are homeless. The cross-sectional nature of the data also limits our ability to draw any conclusions based on the results with respect to causality or the temporality of the associations. As such, it is not possible to discern whether any of the variables examined led to lower social support or vice versa. Finally, the models presented in this study included variables categorizing military service and post-release circumstances, but they excluded pre-military factors (such as childhood adversity) and health issues (such as mental health diagnosis). Though the impact of these factors on Veteran well-being is well documented (Fortier et al., 2019; Vogt et al., 2020), their direct effect on post-release social support is less clear and beyond the scope of this study. Nonetheless, we acknowledge that it is possible that some of the associations observed could be the result of confounding from these or other omitted variables.

Conclusions

We identified a number of Veteran subgroups who may be at risk of low social support, including males, older Veterans, those who released as NCMs, those who live alone, those who are not working, and those who are not satisfied with their main activity. Combined with research that consistently demonstrates that social support has a positive effect on mental health in military populations (Elbogen et al., 2012; Hourani et al., 2012; Therrien et al., 2016), we hypothesize that increasing social support in these subgroups may lead to sizable improvements in Veterans’ well-being.