Contribution of the Mission in Afghanistan to the Burden of Past-Year Mental Disorders in Canadian Armed Forces Personnel,2013

Abstract

French

Objective:

The purpose of this study was to estimate the contribution of the mission in Afghanistan to the burden of mental health problems in the Canadian Armed Forces (CAF).

Methods:

Data were obtained from the 2013 Canadian Forces Mental Health Survey, which assessed mental disorders using the World Health Organization’s Composite International Diagnostic Interview. The sample consisted of 6696 Regular Force (RegF) personnel, 3384 of whom had deployed in support of the mission. We estimated the association of past-year mental health problems with Afghanistan deployment status, adjusting for covariates using logistic regression; population attributable fractions (PAFs) were also calculated.

Results:

Indication of a past-year mental disorder was identified in 18.4% (95% confidence interval [CI], 17.0% to 19.7%) of Afghanistan deployers compared with 14.6% (95% CI, 13.3% to 15.8%) in others. Afghanistan-related deployments contributed to the burden of a past-year disorder (PAF = 8.7%; 95% CI, 3.0% to 14.2%), with the highest PAFs being seen for panic disorder (34.7%) and posttraumatic stress disorder (32.1%). The PAFs for individual alcohol use disorders and suicide ideation were not different from zero. Child abuse, however, had a much greater PAF for any past-year disorder (28.7%; 95% CI, 23.4% to 33.7%) than did the Afghanistan mission.

Conclusions:

The mission in Afghanistan contributed significantly to the burden of mental disorders in the CAF RegF in 2013. However, the much stronger contribution of child abuse highlights the need for strong military mental health systems, even in peacetime, and the need to target the full range of determinants of mental health in prevention and control efforts.

Keywords

mental disorders military personnel Afghanistan service-related mental disorders population attributable fraction epidemiology

Clinical Implications

Prevention and control efforts in military populations need to target the full range of determinants of mental health, not just occupational trauma.

Military health systems need to be prepared to address the substantial burden of non-occupational mental disorders in military personnel.

Limitations

Calculation of the burden of mental disorders attributable to a risk factor (such as deployment in support of the mission in Afghanistan) using Population Attributable Fractions (PAFs) assumes that the risk association is causal.

Veteran benefit eligility determinations use a clinical, not epidemiological, approach to attribution, meaning that the PAFs presented in the present study will not be a good predictor of the future need for veterans’ benefits and services.

More than 40,000 Canadian Armed Forces (CAF) personnel have deployed in support of the mission in Afghanistan since it began in 2001. Data from Canada’s closest allies^1

–7 have shown that a substantial minority of their personnel who deployed to the conflicts in Southwest Asia have mental health problems, many apparently related to their deployment. Mental disorders have important implications for military organizations: they are a leading cause of impaired productivity,⁸ absenteeism,⁸ and turnover,^2,9 and mental health care represents a large and growing proportion of health services delivered by military organizations.^10
–12 Service-related mental health problems are also an important driver of benefits and services for veterans.¹³

The most recent complete and rigorous assessment of mental health in the CAF dates to 2002 with the Canadian Community Health Survey Cycle 1.2, Canadian Forces Supplement.¹⁴ At that time, approximately 15% of Regular Forces (RegF) personnel had 1 of 6 common past-year mental disorders,¹⁵ yet only a modest amount of the overall mental illness burden (9% in men, 6% in women) was attributable to deployment.¹⁶ In contrast, a recent study that was initiated while the CAF was still involved in the mission in Afghanistan indicated that 13.5% of those who deployed with this mission were later diagnosed with an Afghanistan-related mental disorder after a mean of just over 4 years of follow-up.¹⁷ In almost three-quarters of those with a mental disorder diagnosis, clinicians attributed 1 or more of their psychiatric diagnoses to the mission.¹⁷

Concurrent with the mission in Afghanistan, the CAF mental health system has undergone extensive renewal.^18,19 Initiatives have included doubling the number of mental health clinicians, establishing regional centres for diagnosis and treatment of posttraumatic stress disorder (PTSD) and other service-related mental disorders, initiation of postdeployment mental health screening, and initiation of mental health training and education programs to build resilience and enhance mental health literacy, specifically targeting deploying personnel. In their totality, these efforts may have had preventive effects, shortened the delay to first care, improved the effectiveness of care, and decreased the relapse rate of mental disorders. CAF mental health services renewal may thus have reduced the contribution of deployments to the overall burden of mental disorders relative to 2002.

A number of studies have quantified mental health problems among deployed cohorts,^{1

–7,17} but they generally lacked the ability to attribute the identified problem with a specific deployment and/or lacked a nondeployed comparison group. As such, it has been difficult to situate the burden of Afghanistan deployment-related mental health problems within a larger context of mental health problems in the military, including those related to other missions and those related to nonoccupational factors. Recently, analysis of data from the 2013 Canadian Forces Mental Health Survey (CFMHS) has shown a higher crude prevalence of most disorders in those with an Afghanistan-related deployment, including major depressive episode (MDE), PTSD, generalized anxiety disorder (GAD), and panic disorder (PD). In contrast, the crude prevalence of alcohol use disorders (AUDs) was lower, relative to those who had not deployed in support of the mission. Those comparisons did not, however, adjust for potentially important confounding factors such as differences in the sociodemographic characteristics of those who did and did not deploy or differences in other lifetime trauma exposure, such as child abuse. In addition, they did not specifically quantify the contribution of the mission to the burden of mental disorders.⁶

The purpose of this study was to 1) explore the association between Afghanistan-related deployment and mental disorders, adjusting for potential confounding factors, and 2) quantify the contribution of the Afghanistan mission to the burden of past-year mental disorders through calculation of a population attributable fraction (PAF).

Methods

Study Population and Sampling

The data source for this analysis was the 2013 CFMHS, a population-based survey of serving CAF personnel that was administered by Statistics Canada. The sampling frame consisted of 4857 Reserve Force personnel (only those with an Afghanistan-related deployment) and 67,776 Regular Force (RegF) personnel who were serving in September 2012, based on CAF administrative data. The analysis in this article was restricted to the data collected from RegF personnel: 6696 individuals (response rate = 80%), of whom close to half (3384) had an Afghanistan-related deployment. Individuals who agreed to participate in the survey were asked to provide their informed consent prior to being interviewed.

Survey Content

Mental Disorders and Suicide Ideation

The survey assessed past-year MDE, PTSD, GAD, PD, alcohol abuse, and alcohol dependence using the World Health Organization’s Composite International Diagnostic Interview (WHO-CIDI), version 3.0.²⁰ Three additional aggregate outcomes were used: 1) AUD, including alcohol abuse or dependence; 2) any anxiety disorder, including PTSD, GAD, and PD; and 3) any mental disorder. Past-year suicide ideation was determined based on whether respondents indicated having thought about committing suicide or taking their own life in the past 12 months.

Military and Sociodemographic Information

The primary covariate of interest was a previous deployment in support of the mission in Afghanistan. This information was obtained from administrative sources and linked to the survey data deterministically. The following sociodemographic and military characteristics were also available on the data set, either through survey questions or through linked administrative data: sex, age, military rank (junior noncommissioned member [JNCM], senior noncommissioned member [SNCM], and officer), service (Army, Navy, or Air Forces), marital status, highest education level attained, racial background (white, nonwhite, or multiple), household income, indication of having difficulty meeting basic expenses, and whether individuals had a previous deployment outside of North America, other than ones associated with the mission in Afghanistan.

Child Abuse

Physical abuse, sexual abuse, and exposure to intimate partner violence experienced before the age of 16 were assessed using items from the Childhood Experiences of Violence Questionnaire, a valid and reliable tool developed for assessing youth victimization.²¹ Information was collected from all respondents 18 years or older, and frequency thresholds were imposed, as per the guidelines of the Childhood Experiences of Violence Questionnaire items, to identify the presence of these 3 types of child abuse experiences. A count of the number of different types of child abuse experiences was also computed, varying from none to 3.

Data Analysis

The data were analyzed using SAS for Windows, version 9.3 (SAS Institute, Cary, NC); however, PAFs and associated 95% confidence intervals (CIs) were calculated using Stata for Windows, release 13 (StataCorp LP, College Station, TX). We applied the final survey weights provided by Statistics Canada to determine descriptive and regression statistics. Bootstrap methods were used to obtain standard errors for descriptive statistics, using replicate weights and a SAS macro (i.e., BOOTVAR) that were provided by Statistics Canada.²² The Taylor series linearization method²³ was used for all other standard error estimates. Listwise deletion was used for missing values, resulting in the exclusion of 0.9% to 3.6% of respondents.

Wald chi-square tests were used to assess associations between Afghanistan-related deployment and each variable. We also used a series of logistic regression models to assess the unadjusted and adjusted association of an Afghanistan deployment with the presence of a mental disorder and suicide ideation; Model 1 assessed the unadjusted association (expressed as an odds ratio, OR), and Model 2 assessed the association adjusted for military and sociodemographic variables, including number of different child abuse types.

PAF estimates quantify the fraction of disease or disorder cases in a population that are considered attributable to a particular population exposure and thus the proportion by which the outcome could have been reduced had the exposure not been present. PAF estimates assume that the relationship between an exposure and outcome is causal and, hence, rely on the removal of confounding between exposure and outcome. PAF estimates and associated 95% CIs were generated for an Afghanistan-related deployment and child abuse using the Stata punaf module that has been described in detail elsewhere.²⁴

Results

Survey Population Characteristics

As shown in Table 1, the survey population predominantly consisted of males. Individuals were largely younger than 45 years, were married or in a common-law relationship, were from a white racial background, were in the JNCM ranks and in the Army, and had a household income greater than $60,000. Deployments in this population were common; 45.1% had an Afghanistan-related deployment and 41.3% had other non-Afghanistan deployments outside of North America. Individuals with an Afghanistan-related deployment were more likely to be male, older, of white racial background, in higher household income groupings, in higher ranks, in the Army, in a married or common-law relationship, and previously deployed on non-Afghanistan-related missions outside North America, and they were less likely to be in higher education groupings (Wald chi-square test P ≤ 0.05).

Table 1.

Military and Sociodemographic Descriptives of Serving Regular Force Personnel with (Unweighted n = 3385; Weighted n = 29,040) and without (Unweighted n = 3,311; Weighted n = 35,340) an Afghanistan-Related Deployment in 2013.

	Regular Force Personnel
	Without Afghanistan-Related Deployments		With Afghanistan-Related Deployments		Overall
Characteristic	Weighted %	95% CI	Weighted %	95% CI	Weighted %	95% CI
Sex^a
Male	83.8	82.6 to 85.1	89.0	87.9 to 90.1	86.1	85.3 to 87.0
Female	16.2	14.9 to 17.4	11.0	9.9 to 12.1	13.9	13.0 to 14.7
Age, y^a
17-24	21.5	20.0 to 23.1	3.3	2.6 to 4.0	13.3	12.4 to 14.2
25-34	39.1	37.5 to 40.8	35.8	34.3 to 37.3	37.6	36.5 to 38.8
35-44	20.5	19.1 to 22.0	36.4	34.9 to 37.9	27.7	26.7 to 28.8
45-60	18.8	17.7 to 20.0	24.4	23.1 to 25.8	21.4	20.5 to 22.2
Rank^a
JNCM	60.7	60.3 to 61.0	48.2	47.7 to 48.7	55.0	54.8 to 55.2
SNCM	17.4	17.1 to 17.8	32.2	31.7 to 32.6	24.1	23.8 to 24.3
Officer	21.9	21.6 to 22.2	19.6	19.3 to 19.9	20.9	20.8 to 21.0
Service^a
Navy	20.0	18.7 to 21.4	13.8	12.5 to 15.0	17.2	16.3 to 18.2
Army	43.7	42.0 to 45.5	64.6	62.9 to 66.2	53.1	51.9 to 54.3
Air Force	36.2	34.6 to 37.9	21.7	20.2 to 23.1	29.6	28.6 to 30.7
Marital status^a
Married/common	58.1	56.3 to 59.8	74.7	73.3 to 76.2	65.6	64.5 to 66.7
Single	35.4	33.7 to 37.1	16.5	15.3 to 17.8	26.9	25.8 to 28.0
Widowed/separated/divorced	6.5	5.6 to 7.4	8.7	7.8 to 9.7	7.5	6.9 to 8.2
Education^a
< Secondary school graduation	3.5	2.7 to 4.2	4.9	4.2 to 5.6	4.1	3.6 to 4.6
Secondary school graduation	22.0	20.5 to 23.6	30.2	28.6 to 31.9	25.7	24.6 to 26.9
Some postsecondary	8.6	7.6 to 9.6	9.2	8.2 to 10.2	8.9	8.2 to 9.6
Postsecondary graduation	65.9	64.2 to 67.6	55.7	54.1 to 57.4	61.3	60.1 to 62.5
Racial background^a
White	87.7	86.3 to 89.0	93.1	92.2 to 94.0	90.1	89.3 to 90.9
Nonwhite	7.6	6.6 to 8.6	4.1	3.4 to 4.9	6.1	5.4 to 6.7
Multiple	4.7	3.9 to 5.5	2.8	2.2 to 3.3	3.8	3.3 to 4.3
Household income^a
<$40,000	5.9	5.0 to 6.8	0.8	0.5 to 1.1	3.6	3.1 to 4.1
$40,000-$59,999	13.8	12.4 to 15.1	4.1	3.4 to 4.9	9.4	8.6 to 10.2
$60,000-$79,999	23.1	21.5 to 24.8	22.1	20.7 to 23.5	22.7	21.6 to 23.8
$80,000-$99,999	16.0	14.6 to 17.3	18.9	17.5 to 20.2	17.3	16.3 to 18.2
$100,000 or more	41.3	39.5 to 43.0	54.2	52.6 to 55.8	47.1	45.9 to 48.3
Difficulty meeting basic expenses
Yes	6.1	5.1 to 7.0	6.1	5.3 to 7.0	6.1	5.5 to 6.7
No	93.9	93.0 to 94.9	93.9	93.0 to 94.7	93.9	93.2 to 94.5
Other deployments outside North America^a
Yes	28.7	27.1 to 30.4	56.6	54.9 to 58.2	41.3	40.2 to 42.4
No	71.3	69.6 to 72.9	43.4	41.8 to 45.1	58.7	57.6 to 59.8

CI, confidence interval; JNCM, junior noncommissioned member; SNCM, senior noncommissioned member.

^aSignificant at P < .01 when using a Wald chi-square test comparing Regular Force personnel who deployed in support of the mission in Afghanistan with those who did not.

As shown in Table 2, a past-year mental disorder was identified in 16.5% (95% CI, 15.6% to 17.4%) of the survey population, 18.4% (95% CI, 17.0% to 19.7%) among those with an Afghanistan deployment and 14.6% (95% CI, 13.3% to 15.8%) in those without; 51.7% of those with any mental disorders were among those with an Afghanistan deployment. The median time from most recent Afghanistan-related deployment return to interview was 5.3 years (minimum = 0.35 years, maximum = 12.6 years; mean = 5.4 years) among those with a deployment. Past-year PTSD, GAD, PD, and MDE were all more likely among personnel who deployed in support of the mission, but AUD was less likely (Wald chi-square test P ≤ 0.05); differences in past-year suicide ideation were statistically nonsignificant. Personnel with an Afghanistan deployment were more likely to have experienced physical abuse as a child (P ≤ 0.05) but not the other child abuse experiences; the number of experience types did not differ significantly (P ≤ 0.05), and few individuals had all 3 types.

Table 2.

Past-Year Mental Disorders, Suicide Ideation, and Child Abuse Experiences Identified among Serving Regular Force Personnel with (Unweighted n = 3385; Weighted n = 29,040) and without (Unweighted n = 3311; Weighted n = 35,340) an Afghanistan-related Deployment in 2013.

	Regular Force Personnel
	Without Afghanistan-Related Deployments		With Afghanistan-Related Deployments		Overall
Characteristic	Weighted %	95% CI	Weighted %	95% CI	Weighted %	95% CI
Past-year mental disorders
Any mental disorder^a	14.6	13.3 to 15.8	18.4	17.0 to 19.7	16.5	15.6 to 17.4
Any anxiety disorder^a	7.1	6.1 to 8.0	12.9	11.8 to 14.0	9.7	8.9 to 10.5
GAD^a	3.8	3.1 to 4.5	5.8	5.0 to 6.5	4.7	4.2 to 5.2
PD^a	2.0	1.4 to 2.5	5.1	4.3 to 5.9	3.4	2.9 to 3.9
PTSD^a	3.2	2.5 to 3.9	7.7	6.8 to 8.6	5.3	4.6 to 5.9
MDE^a	6.8	5.9 to 7.8	9.3	8.2 to 10.4	8.0	7.3 to 8.6
Alcohol abuse^a	3.0	2.3 to 3.7	1.9	1.5 to 2.4	2.5	2.1 to 2.9
Alcohol dependence^b	2.4	1.8 to 3.0	1.4	1.1 to 1.8	2.0	1.6 to 2.3
AUD^a	5.4	4.6 to 6.2	3.4	2.8 to 4.0	4.5	3.9 to 5.0
Suicide ideation^c	4.0	3.2 to 4.8	4.6	3.9 to 5.4	4.3	3.7 to 4.8
Child abuse types
Physical abuse^b	42.6	40.7 to 44.6	45.0	43.3 to 46.8	43.7	42.4 to 45.1
Sexual abuse^c	7.6	6.6 to 8.6	7.6	6.7 to 8.5	7.6	6.9 to 8.3
Exposure to intimate partner violence^c	10.3	9.2 to 11.5	10.4	9.3 to 11.4	10.3	9.5 to 11.1
Number of child abuse types^c
0	53.2	51.2 to 55.1	51.3	49.5 to 53.1	52.3	50.9 to 53.7
1	35.1	33.3 to 36.9	36.0	34.3 to 37.8	35.5	34.2 to 36.8
2	9.9	8.9 to 11.0	11.1	10.0 to 12.2	10.5	9.7 to 11.2
3	1.8	1.2 to 2.3	1.5	1.1 to 1.9	1.7	1.3 to 2.0

AUD, alcohol use disorders (includes alcohol abuse or dependence); CI, confidence interval; GAD, generalized anxiety disorder; MDE, major depressive episode; PD, panic disorder; PTSD, posttraumatic stress disorder.

^aSignificant at P < .01 when using a Wald chi-square test comparing Regular Forces personnel who deployed in support of the mission in Afghanistan with those who did not.

^bSignificant at a P ≤ .05.

^cNot significant; P > .05 level.

Logistic Regression Results

After adjusting for potential confounding (Model 2, Table 3), the Afghanistan deployment ORs remained statistically significant for the grouped outcomes, any mental disorder, any anxiety disorder, and AUD and remained nonsignificant for suicide ideation. While this deployment exposure was associated with an increased odds for any mental disorder (adjusted OR, 1.27; 95% CI, 1.09 to 1.49) and anxiety disorders (adjusted OR, 1.61; 95% CI, 1.32 to 1.96), it was associated with a decreased odds for AUD (adjusted OR, 0.73; 95% CI, 0.54 to 0.98).

Table 3.

Adjusted Odds Ratios (AORs) for the Association of Afghanistan-Related Deployments with Grouped Past-Year Mental Health Problems and Suicide Ideation Identified among Serving Regular Force Personnel in 2013, Adjusting for Military and Sociodemographic Characteristics.

	AOR (95% CI)
	Any Mental Disorder^a	Any Anxiety Disorder^b	AUD	Suicide Ideation
Model 1: Afghanistan-related deployments
Afghanistan deployment^c	1.37 (1.19 to 1.56)	1.94 (1.63 to 2.31)	0.61 (0.48 to 0.78)	1.16 (0.90 to 1.48)
Model 2: Adjusting for military and sociodemographic characteristics
Afghanistan deployment	1.27 (1.09 to 1.49)	1.61 (1.32 to 1.96)	0.73 (0.54 to 0.98)	1.12 (0.84 to 1.48)
Other international deployment	1.07 (0.91 to 1.26)	1.23 (1.00 to 1.50)	1.03 (0.73 to 1.46)	0.87 (0.66 to 1.15)
Rank
JCNM	Reference	Reference	Reference	Reference
SNCM	1.02 (0.85 to 1.22)	0.98 (0.79 to 1.23)	0.86 (0.61 to 1.22)	1.14 (0.82 to 1.58)
Officer	0.66 (0.54 to 0.81)	0.66 (0.51 to 0.85)	0.73 (0.50 to 1.09)	0.88 (0.61 to 1.26)
Service
Navy	0.76 (0.62 to 0.92)	0.79 (0.62 to 1.02)	0.79 (0.55 to 1.15)	0.85 (0.59 to 1.21)
Army	Reference	Reference	Reference	Reference
Air Force	0.68 (0.57 to 0.81)	0.74 (0.60 to 0.92)	0.50 (0.35 to 0.72)	0.78 (0.56 to 1.09)
Sex
Male	Reference	Reference	Reference	Reference
Female	1.52 (1.25 to 1.85)	2.15 (1.71 to 2.71)	0.50 (0.30 to 0.85)	0.88 (0.59 to 1.30)
Age, y
17-24	1.24 (0.89 to 1.72)	0.63 (0.39 to 1.04)	3.70 (2.04 to 6.74)	1.19 (0.67 to 2.11)
25-34	1.29 (1.04 to 1.61)	1.08 (0.83 to 1.41)	2.83 (1.75 to 4.58)	1.27 (0.87 to 1.86)
35-44	1.35 (1.11 to 1.63)	1.38 (1.11 to 1.72)	1.63 (1.02 to 2.61)	1.20 (0.85 to 1.70)
45-60	Reference	Reference	Reference	Reference
Marital status
Married or common law	Reference	Reference	Reference	Reference
Single	1.37 (1.12 to 1.68)	1.03 (0.80 to 1.33)	1.80 (1.28 to 2.51)	1.18 (0.82 to 1.70)
Widowed/separated/divorced	1.83 (1.44 to 2.33)	1.44 (1.07 to 1.93)	1.86 (1.15 to 2.99)	1.49 (0.98 to 2.26)
Education
< Secondary school graduation	1.23 (0.87 to 1.74)	1.65 (1.10 to 2.47)	1.16 (0.66 to 2.06)	1.67 (0.93 to 2.97)
Secondary school graduation	1.13 (0.95 to 1.34)	1.09 (0.88 to 1.35)	1.22 (0.88 to 1.69)	1.03 (0.74 to 1.44)
Some postsecondary	1.14 (0.89 to 1.47)	0.99 (0.71 to 1.38)	0.88 (0.56 to 1.40)	1.22 (0.79 to 1.89)
Postsecondary graduation	Reference	Reference	Reference	Reference
Racial background
White	Reference	Reference	Reference	Reference
Nonwhite	0.77 (0.56 to 1.05)	0.69 (0.46 to 1.01)	0.77 (0.45 to 1.32)	0.59 (0.35 to 1.02)
Multiple	0.84 (0.58 to 1.22)	0.89 (0.57 to 1.38)	0.90 (0.46 to 1.77)	0.34 (0.13 to 0.92)
Household income
<$40,000	0.35 (0.20 to 0.60)	0.32 (0.14 to 0.73)	0.52 (0.25 to 1.11)	0.23 (0.07 to 0.78)
$40,000-$59,999	1.32 (0.99 to 1.76)	1.40 (0.97 to 2.04)	1.84 (1.21 to 2.81)	1.88 (1.17 to 3.01)
$60,000-$79,999	1.14 (0.93 to 1.39)	1.38 (1.07 to 1.76)	1.00 (0.69 to 1.44)	1.65 (1.16 to 2.36)
$80,000-$99,999	0.95 (0.78 to 1.17)	0.84 (0.65 to 1.09)	1.30 (0.89 to 1.91)	1.52 (1.07 to 2.17)
$100,000 or more	Reference	Reference	Reference	Reference
Difficulty meeting basic expenses
Yes	2.41 (1.89 to 3.07)	2.43 (1.84 to 3.20)	1.65 (1.07 to 2.55)	1.98 (1.35 to 2.92)
No	Reference	Reference	Reference	Reference
Number of child abuse types
0	Reference	Reference	Reference	Reference
1	1.80 (1.54 to 2.10)	1.63 (1.34 to 1.98)	1.93 (1.45 to 2.58)	1.77 (1.32 to 2.37)
2	3.12 (2.52 to 3.87)	3.33 (2.60 to 4.27)	3.07 (2.05 to 4.60)	2.52 (1.71 to 3.73)
3	4.67 (3.01 to 7.24)	4.67 (2.81 to 7.75)	1.16 (0.25 to 5.27)	10.25 (5.58 to 18.82)

AOR, adjusted odds ratio; AUD, alcohol use disorders (includes alcohol abuse or dependence); CI, confidence interval; JNCM, junior noncommissioned member; SNCM, senior noncommissioned member.

^aAny of the following: major depressive episode, generalized anxiety disorder, panic disorder, posttraumatic stress disorder, alcohol abuse, or alcohol dependence.

^bAny of the following: generalized anxiety disorder, panic disorder, or posttraumatic stress disorder.

^cAn unadjusted odds ratio was initially computed for the Afghanistan deployment variable (Model 1).

Some covariates had significant independent associations with these outcomes (Table 3). Notably, lower ranks; Army service; females; being widowed, separated, or divorced; having difficulty with basic expenses; and child abuse were associated with a higher odds of any mental disorder and any anxiety disorder. Being in the lowest education category, being single, and having other non-Afghanistan deployments were also associated with a higher odds of an anxiety disorder. Air Force service, females, younger age, not being married, having difficulty with basic expenses, and child abuse were associated with a lower AUD odds. Moreover, nonmultiple racial backgrounds, certain household income categories, having difficulty with basic expenses, and child abuse were associated with a higher odds of suicide ideation. With the exception of AUD, a dose-response relationship was observed for the association of child abuse types with these outcomes.

The association of an Afghanistan deployment with individual disorders was also assessed after adjustment for potential confounding (Model 2, Table 4). This deployment exposure was associated with a statistically significant increased odds for MDE (adjusted OR, 1.25; 95% CI, 1.01 to 1.54), PTSD (adjusted OR, 2.09; 95% CI, 1.57 to 2.78), and PD (adjusted OR, 2.16; 95% CI, 1.56 to 2.99) but not the other disorders. A number of the covariates had an independent association with the disorders, but the significance of specific covariates varied across disorders (Table 4).

Table 4.

Adjusted Odds Ratios (AORs) for the Association of Afghanistan-Related Deployments with Individual Past-Year Mental Health Problems Identified among Serving Regular Force Personnel in 2013, Adjusting for Military and Sociodemographic Characteristics.

	AOR (95% CI)
	MDE	GAD	PTSD	PD	Alcohol Abuse	Alcohol Dependence
Model 1: Afghanistan-related deployments
Afghanistan deployment^a	1.40 (1.16 to 1.68)	1.54 (1.22 to 1.94)	2.54 (1.98 to 3.26)	2.64 (1.94 to 3.60)	0.63 (0.45 to 0.87)	0.61 (0.43 to 0.88)
Model 2: Adjusting for military and sociodemographic characteristics
Afghanistan deployment	1.25 (1.01 to 1.54)	1.21 (0.93 to 1.57)	2.09 (1.57 to 2.78)	2.16 (1.56 to 2.99)	0.80 (0.56 to 1.16)	0.66 (0.41 to 1.04)
Other international deployment	0.99 (0.79 to 1.22)	1.08 (0.82 to 1.42)	1.38 (1.06 to 1.80)	1.01 (0.72 to 1.41)	1.08 (0.68 to 1.71)	1.00 (0.59 to 1.68)
Rank
JCNM	Reference	Reference	Reference	Reference	Reference	Reference
SNCM	1.10 (0.87to 1.40)	1.07 (0.79 to 1.46)	0.94 (0.70 to 1.27)	0.94 (0.66 to 1.35)	0.75 (0.48 to 1.19)	1.05 (0.63 to 1.75)
Officer	0.70 (0.53 to 0.93)	0.83 (0.59 to 1.18)	0.59 (0.41 to 0.85)	0.38 (0.23 to 0.61)	0.58 (0.34 to 0.99)	1.09 (0.62 to 1.94)
Service
Navy	0.82 (0.63 to 1.06)	0.81 (0.58 to 1.12)	0.57 (0.41 to 0.81)	0.86 (0.57 to 1.32)	1.12 (0.70 to 1.79)	0.45 (0.24 to 0.83)
Army	Reference	Reference	Reference	Reference	Reference	Reference
Air Force	0.76 (0.60 to 0.96)	0.92 (0.69 to 1.22)	0.63 (0.48 to 0.84)	0.53 (0.36 to 0.78)	0.53 (0.33 to 0.87)	0.50 (0.30 to 0.83)
Sex
Male	Reference	Reference	Reference	Reference	Reference	Reference
Female	1.45 (1.12 to 1.89)	1.89 (1.40 to 2.56)	2.10 (1.56 to 2.82)	1.61 (1.07 to 2.41)	0.48 (0.24 to 0.99)	0.54 (0.25 to 1.15)
Age, y
17-24	0.85 (0.53 to 1.36)	0.49 (0.25 to 0.96)	0.68 (0.35 to 1.35)	0.66 (0.28 to 1.56)	3.49 (1.70 to 7.18)	4.40 (1.61 to 12.00)
25-34	1.13 (0.84 to 1.51)	1.07 (0.75 to 1.53)	1.05 (0.74 to 1.47)	1.04 (0.68 to 1.60)	1.56 (0.85 to 2.85)	6.88 (3.17 to 14.97)
35-44	1.24 (0.97 to 1.59)	1.32 (0.98 to 1.77)	1.47 (1.11 to 1.96)	1.25 (0.87 to 1.80)	1.51 (0.86 to 2.66)	2.04 (0.92 to 4.52)
45-60	Reference	Reference	Reference	Reference	Reference	Reference
Marital status
Married or common law	Reference	Reference	Reference	Reference	Reference	Reference
Single	1.42 (1.07 to 1.88)	0.86 (0.60 to 1.23)	1.08 (0.76 to 1.54)	0.79 (0.53 to 1.20)	1.70 (1.12 to 2.59)	1.81 (1.06 to 3.09)
Widowed/separated/divorced	2.09 (1.56 to 2.80)	1.51 (1.03 to 2.22)	1.33 (0.90 to 1.95)	1.36 (0.88 to 2.11)	1.73 (0.94 to 3.18)	1.98 (0.97 to 4.06)
Education
< Secondary school graduation	0.93 (0.55 to 1.58)	1.74 (1.00 to 3.02)	0.95 (0.54 to 1.68)	1.54 (0.89 to 2.67)	0.91 (0.39 to 2.09)	1.57 (0.74 to 3.36)
Secondary school graduation	1.22 (0.97 to 1.54)	1.21 (0.91 to 1.62)	1.17 (0.89 to 1.55)	0.92 (0.66 to 1.29)	1.23 (0.82 to 1.84)	1.20 (0.72 to 2.00)
Some postsecondary	1.51 (1.10 to 2.07)	1.24 (0.79 to 1.94)	1.08 (0.70 to 1.65)	0.78 (0.45 to 1.36)	0.76 (0.40 to 1.43)	1.08 (0.56 to 2.06)
Postsecondary graduation	Reference	Reference	Reference	Reference	Reference	Reference
Racial background
White	Reference	Reference	Reference	Reference	Reference	Reference
Nonwhite	0.76 (0.50 to 1.16)	0.51 (0.28 to 0.91)	0.76 (0.46 to 1.25)	1.05 (0.60 to 1.85)	1.25 (0.67 to 2.32)	0.35 (0.13 to 0.95)
Multiple	0.51 (0.28 to 0.93)	0.24 (0.09 to 0.67)	1.08 (0.64 to 1.82)	1.34 (0.71 to 2.55)	1.60 (0.76 to 3.36)	0.22 (0.06 to 0.78)
Household income
<$40,000	0.33 (0.14 to 0.77)	0.36 (0.12 to 1.08)	0.47 (0.17 to 1.31)	0.12 (0.02 to 0.72)	0.41 (0.15 to 1.15)	0.76 (0.26 to 2.25)
$40,000-$59,999	1.39 (0.96 to 2.03)	1.28 (0.76 to 2.16)	1.79 (1.08 to 2.97)	1.33 (0.72 to 2.44)	1.45 (0.82 to 2.57)	2.18 (1.19 to 4.02)
$60,000-$79,999	1.20 (0.92 to 1.57)	1.41 (1.01 to 1.96)	1.36 (0.96 to 1.92)	1.16 (0.82 to 1.65)	1.16 (0.72 to 1.86)	0.83 (0.47 to 1.45)
$80,000-$99,999	0.99 (0.75 to 1.30)	0.81 (0.56 to 1.16)	0.90 (0.64 to 1.27)	0.84 (0.57 to 1.25)	1.75 (1.10 to 2.78)	0.78 (0.40 to 1.51)
$100,000 or more	Reference	Reference	Reference	Reference	Reference	Reference
Difficulty meeting basic expenses
Yes	2.26 (1.67 to 3.04)	1.94 (1.34 to 2.82)	2.33 (1.64 to 3.29)	2.46 (1.64 to 3.71)	0.72 (0.37 to 1.41)	3.04 (1.73 to 5.32)
No	Reference	Reference	Reference	Reference	Reference	Reference
Number of child abuse types
0	Reference	Reference	Reference	Reference	Reference	Reference
1	1.76 (1.43 to 2.17)	1.38 (1.07 to 1.77)	1.82 (1.40 to 2.36)	1.37 (0.99 to 1.88)	1.30 (0.89 to 1.90)	3.15 (2.01 to 4.93)
2	2.53 (1.89 to 3.38)	2.16 (1.53 to 3.03)	3.34 (2.43 to 4.57)	2.70 (1.84 to 3.98)	2.37 (1.42 to 3.97)	4.15 (2.21 to 7.77)
3	4.45 (2.49 to 7.95)	1.36 (0.48 to 3.86)	7.13 (4.03 to 12.60)	1.95 (0.84 to 4.53)	1.39 (0.22 to 8.69)	0.74 (0.11 to 4.92)

AOR, adjusted odds ratio; CI, confidence interval; GAD, generalized anxiety disorder; JNCM, junior noncommissioned member; MDE, major depressive episode; PD, panic disorder; PTSD, posttraumatic stress disorder; SNCM, senior noncommissioned member.

^aAn unadjusted odds ratio was initially computed for the Afghanistan deployment variable (Model 1).

The possibility that child abuse experiences act differently among those with and without Afghanistan-related deployments was evaluated via a 2-way interaction. This interaction was only statistically significant for MDE. Among Afghanistan deployers, the dose response for child abuse types on MDE was less pronounced and not as significant relative to nondeployers. This interaction was not included in the final model.

PAFs

PAF estimates that quantify the proportion of the burden of mental disorders and suicide ideation identified in 2013 that were attributable to the Afghanistan mission are provided in Table 5. These estimates are dependent on the adjustment for confounding variables, but not mediator variables, in the relationship between Afghanistan-related deployments and each outcome (Model 2). This model yielded PAFs that were significant for any mental disorder, any anxiety disorder, and, specifically, for MDE, PTSD, PD, and AUD. Overall, the mission in Afghanistan accounted for 8.7% (95% CI, 3.0 to 14.2) of the burden of the 6 past-year disorders measured by the survey in 2013. PAFs were highest and comparable for PD (34.7%; 95% CI, 20.0 to 46.7) and PTSD (32.1%; 95% CI, 19.9 to 42.3). Although the PAF CI for AUD was broad, it suggests that deployment contributed to an 11.6% (95% CI, 1.3 to 23.0) reduction in AUD. A negative PAF is suggestive of either protective factors that were associated with the deployment environment or deployment selection factors (deployers being disproportionately healthy). The PAF for past-year suicide ideation was not statistically significant.

Table 5.

Population Attributable Fractions for the Burden of Mental Disorders and Suicide Ideation Identified among Serving Regular Force Personnel in 2013 That Were Attributable to an Afghanistan-Related Deployment Exposure and Child Abuse Experiences.

	PAF (95% CI)
	Any Mental Disorder^a	Any Anxiety Disorder^b	AUD	Suicide Ideation	MDE	GAD	PTSD	PD	Alcohol Abuse	Alcohol Dependence
Model 1: Afghanistan-related deployment
Afghanistan deployment^c	11.8 (6.5 to 16.8)	27.0 (19.7 to 33.6)	–20.2 (–30.2 to –11.0)	6.3^d (–5.1 to 16.6)	13.9 (5.9 to 21.3)	18.6 (7.9 to 28.0)	39.2 (28.7 to 48.1)	41.3 (29.0 to 52.1)	–19.7 (–33.5 to –7.4)	–20.7 (–36.2 to –7.0)
Model 2: Adjusting for military and sociodemographic characteristics
Afghanistan deployment	8.7 (3.0 to 14.2)	19.6 (11.3 to 27.1)	–11.6 (–23.0 to –1.3)	4.7^d (–8.2 to 16.0)	9.3 (0.3 to 17.5)	8.8^d (–4.1 to 20.2)	32.1 (19.9 to 42.3)	34.7 (20.0 to 46.7)	–8.2^d (–22.7 to 4.6)	–16.1^d (–35.6 to 0.6)
Any child abuse	28.7 (23.4 to 33.7)	30.6 (23.3 to 37.1)	31.4 (19.8 to 41.3)	33.8 (21.7 to 44.0)	30.2 (21.8 to 37.6)	20.0 (8.9 to 29.8)	37.1 (27.1 to 45.7)	24.7 (10.7 to 36.5)	18.2 (1.7 to 32.0)	48.4 (30.6 to 61.6)
Afghanistan deployment and any child abuse^e	37.7 (28.4 to 45.9)	45.3 (36.4 to 52.9)	23.3 (7.9 to 36.2)	NA	37.0 (26.7 to 45.8)	NA	58.5 (48.3 to 66.6)	51.4 (35.0 to 63.7)	NA	NA

AUD, alcohol use disorders (includes alcohol abuse or dependence); CI, confidence interval; GAD, generalized anxiety disorder; MDE, major depressive episode; NA, not applicable; PAF, population attributable fraction; PD, panic disorder; PTSD, posttraumatic stress disorder.

^aAny of the following: major depressive episode, generalized anxiety disorder, panic disorder, posttraumatic stress disorder, alcohol abuse, or alcohol dependence.

^bAny of the following: generalized anxiety disorder, panic disorder, or posttraumatic stress disorder.

^cThese PAFs (Model 1) were unadjusted.

^dNot statistically significant at α ≤ 0.05 level.

^eThese PAFs were calculated for an exposure of both Afghanistan-related deployments and child abuse victimization experiences together. These were not the sum of the individual PAFs for the separate exposures.

The estimated PAFs for the contribution of child abuse to suicide ideation and each mental disorder, other than PD, were larger than PAFs for an Afghanistan-related deployment. Notably, child abuse accounted for 28.7% (95% CI, 23.4 to 33.7) of the burden associated with the 6 past-year disorders, while the mission accounted for only 8.7% of this burden. The child abuse–associated PAF was highest for alcohol dependence (48.4%; 95% CI, 30.6 to 61.6), followed by PTSD (37.1%; 95% CI, 27.1 to 45.7) and suicide ideation (33.8%; 95% CI, 21.7 to 44.0).

Together, both exposures (Afghanistan-related deployments and child abuse experiences) accounted for a substantial percentage of the burden of common mental disorders: 38.0% of any past-year disorder, 58.5% of PTSD, 51.4% of PD, and 37.0% MDE (Table 4).

Discussion

Key Findings

Personnel with an Afghanistan-related deployment had a significantly higher crude prevalence of past-year MDE, PTSD, GAD, and PD, relative to those who did not deploy. Except for GAD, these differences in risk persisted after adjustment for potential confounders. Notwithstanding this higher prevalence of mood and anxiety disorders, there were no differences in the risk of past-year suicide ideation. Those deployed in support of the mission actually had a lower crude risk of AUD, although this difference was eliminated (or almost so) after adjustment for potential confounders.

The Afghanistan mission had a small but statistically significant contribution to the overall burden of past-year mental disorders in the CAF RegF in 2013, accounting for 8.7% of the burden of the 6 past-year disorders assessed in the survey. The greatest contribution was seen for PD (34.7%) and for PTSD (32.1%). A smaller contribution was seen for past-year MDE (9.3%). The PAF for Afghanistan deployment on GAD was not significant; neither was, notably, the PAF for suicide ideation. After adjustment, there was no association between Afghanistan deployment status and the individual alcohol abuse and alcohol dependent disorders, leading to a nonsignificant PAF. In contrast, child abuse accounted for a much larger proportion of the burden associated with any of the 6 past-year mental disorders (28.7%) than did the mission in Afghanistan (8.7%).

Comparison with Other Findings

Other findings on the risk of mental disorders in those who deployed to Southwest Asia since 2001 are mixed.^{1,25

–30} Slight differences in PTSD prevalence have been noted in UK personnel, relative to nondeployed personnel.^26,27 A 2010 Australian survey found no difference in past-year mental disorders (other than obsessive-compulsive disorder) between ever and never deployers, even when further categorized into warlike and non-warlike deployments.²⁸ US studies have, however, consistently reported higher PTSD prevalence among deployers to Iraq or Afghanistan.^1,25 These studies used different methods and instruments to identify disorders that at least partially account for country-specific differences; these have been highlighted by others.^27,29,30 Combat exposure differences also contribute to these divergent findings.³¹ In addition, our finding of a variation in risk for certain mental disorders by rank, service, sex, age, and marital status has been identified in other studies.^{3,17,32
–34}

Our PAFs are similar to those found in a comparable 2002 survey of the CAF Regular and Reserve Forces, where lifetime trauma exposure questions determined any exposure to combat or peacekeeping operations.¹⁶ Adjusting for military and sociodemographic variables, these authors reported statistically significant PAFs for this exposure on past-year PTSD (46.6% and 23.6% for men and women, respectively) and panic disorder (27.0% for women only); it was nonsignificant for MDE. Our PAFs were a little higher for PD (34.7%) and MDE (9.3%) and a little lower for PTSD (32.1%) relative to a combat or peacekeeping exposure in 2002. For any past-year mental disorder, the PAFs in 2002 were quite similar to our own (9.3% in men, 6.1% in women). Using the same 2002 survey data, the PAF for adverse childhood experiences (including but not limited to physical abuse, sexual abuse, and exposure to intimate partner violence) in relation to any past-year mood or anxiety disorder was lower than our own findings (16.4% in men and 19.5% in women),³⁵ although consistent with our observed pattern of higher PAFs for child abuse than for deployment.

Superficially, our findings conflict with earlier ones showing that in personnel with an Afghanistan-related deployment who were later diagnosed with a mental disorder, 71% of these diagnoses were attributed to this mission by clinicians.¹⁷ However, these 2 studies used very different approaches to attribution.³⁶ The present study assessed past-year disorders among personnel still in-service in 2013 and estimated the fraction of disorders epidemiologically attributed to an Afghanistan deployment, while the earlier study identified all new clinically diagnosed mental disorders during a follow-up period that were clinically attributed, partially or wholly, to an Afghanistan deployment. The present study used an epidemiological approach that explicitly takes into account other contributing factors, while the estimate from the clinical study does not. In other words, the estimate from this earlier clinical study does not assess the underpinnings of the mixture of cause or risk but merely that the disorders were precipitated, at least in part, by the deployment. Determinations of veteran entitlements use a clinical approach, and as such, the earlier study is more relevant when estimating the future need for veterans’ benefits and services.

Our findings that an Afghanistan-related deployment was associated with a lower crude risk of AUD relative to those who did not deploy contrasts with the large body of evidence from other nations showing a strong and consistent relationship between trauma-laden deployments and substance use disorders.^37,38 Deployment selection effects and deployment environmental characteristics may have had an influence in addition to any from trauma exposure: while deployers are at risk of experiencing psychological trauma, they also receive additional mental health training, and those who deploy are disproportionately healthy.³⁹ Their alcohol use status after deployment reflects the net result of their vulnerability, risks, and experiences, the balance of which happens to be largely neutral in our study population. Differences in systems of mental health care may also be an important factor.

Limitations

Aside from the generic limitations of cross-sectional mental health surveys, the most important limitation relates to the PAF estimates: their calculation from cross-sectional data assumes the relationship between an exposure and outcome is causal. The abundance of longitudinal data on the contribution of both deployment-related trauma and child abuse to mental health outcomes supports such a causal relationship. We were able to identify and adjust for many potential confounders, but residual confounding (e.g., by other sources of adult trauma exposure) is always a possibility.

Personnel with mental disorders have a significantly elevated risk of release from military service.⁴⁰ Hence, the failure to include the more than 10,000 personnel with an Afghanistan-related deployment but had left service prior to the survey¹⁹ means that the present findings presumably represent a systematic underestimate of overall contribution of the mission in Afghanistan to the mental health of those who served. We also limited our analysis to past-year disorders, so those whose health had been affected by the mission (or by child abuse experiences) but had recovered were not captured in our PAF estimates. This analysis also excluded Reserve Force personnel, due to the lack of a nondeployed control group in the Reserve sample of the 2013 CFMHS. The present findings are, however, an accurate picture of the contribution of the mission to the burden of common past-year mental disorders in RegF personnel in service at the time of the survey.

Implications

The CAF mission in Afghanistan had a detectable contribution to the burden of mental health problems among the CAF RegF in 2013, and this was greatest for PD and PTSD. The primary practical implication of this finding is that while the CAF mental health system needs to be capable of responding to the increase in burden of mental disorders that predictably results from traumatogenic military operations, it must also be ready to respond to the baseline burden of mental disorders that will manifest, with similar implications, even during periods of relative operational calm. The stronger contribution of child abuse experiences to the burden of mental disorders in the CAF speaks to the potential benefits of interventions to attenuate their linkage with later mental health outcomes.

The lack of association between Afghanistan-related deployments and suicide ideation, even in the face of an increased risk of mood and anxiety disorders that contribute to it, may speak to the potential of systems of care to interrupt the link between psychological trauma—specifically deployment-related trauma—and later suicidality. This hypothesis is testable using the CFMHS and similar general population data. The complexities of the association of deployment, deployment-related trauma, and AUD also merit exploration.

Conclusion

Our findings indicate that a significant proportion of past-year mental disorders in the CAF, particularly PTSD and PD, were attributed to Afghanistan-related deployments. However, the contribution from child abuse experiences was much larger. These findings speak to the need for robust mental health systems that target the full range of determinants of mental health in military personnel, both in times of war and in times of peace.

Footnotes

Disclosures

Both authors are employees of the Canadian Department of National Defence and funding for this research came via this Department. No other disclosures reported.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by funding from the Canadian Armed Forces Surgeon General’s Medical Research Program.

References

Hoge

Castro

Messer

. Combat duty in Iraq and Afghanistan, mental health problems, and barriers to care. N Engl J Med. 2004;351(1):13–22.

Hoge

Auchterlonie

Milliken

. Mental health problems, use of mental health services, and attrition from military service after returning from deployment to Iraq or Afghanistan. JAMA. 2006;295(9):1023–1032.

Wells

Leardmann

Fortuna

. A prospective study of depression following combat deployment in support of the wars in Iraq and Afghanistan. Am J Public Health. 2010;100(1):90–99.

Smith

Ryan

Wingard

. New onset and persistent symptoms of post-traumatic stress disorder self reported after deployment and combat exposures: prospective population based US military cohort study. BMJ. 2008;336(7640):366–371.

Fear

Jones

Murphy

. What are the consequences of deployment to Iraq and Afghanistan on the mental health of the UK armed forces? A cohort study. Lancet. 2010;375(9728):1783–1797.

Wells

Miller

Adler

. Mental health impact of the Iraq and Afghanistan conflicts: a review of US research, service provision, and programmatic responses. Int Rev Psychiatry. 2011;23(2):144–152.

Cohen

Gima

Bertenthal

. Mental health diagnoses and utilization of VA non-mental health medical services among returning Iraq and Afghanistan veterans. J Gen Intern Med. 2010;25(1):18–24.

Rost

Smith

Dickinson

. The effect of improving primary care depression management on employee absenteeism and productivity. A randomized trial. Med Care. 2004;42(12):1202–1210.

Wang

. Perceived work stress, imbalance between work and family/personal lives, and mental disorders. Soc Psychiatry Psychiatr Epidemiol. 2006;41(7):541–548.

10.

Armed Forces Health Surveillance Center. Hospitalizations among members of the active component, U.S. Armed Forces, 2009. MSMR. 2010;17(4):3–9.

11.

Armed Forces Health Surveillance Center. Ambulatory visits among members of the active component, U.S. Armed Forces, 2009. MSMR. 2010;17(4):10–15.

12.

Armed Forces Health Surveillance Center. Absolute and relative morbidity burdens attributable to various illnesses and injuries, U.S. Armed Forces, 2009. MSMR. 2010;17(4):16–21.

13.

Levy

Sidel

. Health effects of combat: a life-course perspective. Annu Rev Public Health. 2009;30:123–136.

14.

Statistics Canada. Canadian Community Health Survey Cycle 1.2: Mental Health and Well-Being—Canadian Forces: Master file documentation. 2004 [cited 2012 Jul 5]. Available from: http://www.statcan.ca/english/sdds/document/5084_D1_T1_V1_E.pdf.

15.

Sareen

Cox

Afifi

. Combat and peacekeeping operations in relation to prevalence of mental disorders and perceived need for mental health care: findings from a large representative sample of military personnel. Arch Gen Psychiatry. 2007;64(7):843–852.

16.

Sareen

Belik

Afifi

. Canadian military personnel’s population attributable fractions of mental disorders and mental health service use associated with combat and peacekeeping operations. Am J Public Health. 2008;98(12):2191–2198.

17.

Boulos

Zamorski

. Deployment-related mental disorders in Canadian Forces personnel deployed in support of the mission in Afghanistan, 2001-2008. CMAJ. 2013;185(11):E545–E552.

18.

Department of National Defence and Canadian Armed Forces. Surgeon General’s mental health strategy: Canadian Forces Health Services Group—an evolution of excellence [cited 2014 Oct 3]. Available from: http://www.forces.gc.ca/assets/FORCES_Internet/docs/en/about-reports-pubs-health/surg-gen-mental-health-strategy.pdf.

19.

Zamorski

Bennett

Boulos

. The 2013 Canadian Forces Mental Health Survey: background and methods. Can J Psychiatry. 2016;61 Suppl 1:10S–25S.

20.

Haro

Arbabzadeh-Bouchez

Brugha

. Concordance of the Composite International Diagnostic Interview Version 3.0 (CIDI 3.0) with standardized clinical assessments in the WHO World Mental Health surveys. Int J Methods Psychiatr Res. 2006;15(4):167–180.

21.

Walsh

MacMillan

Trocme

. Measurement of victimization in adolescence: development and validation of the Childhood Experiences of Violence Questionnaire. Child Abuse Negl. 2008;32(11):1037–1057.

22.

Bootvar. Ottawa (ON): Statistics Canada; 2015 [cited 2014 Nov 3]. Available from: http://www.statcan.gc.ca/eng/rdc/bootvar_sas.

23.

Wolter

. Introduction to variance estimation. New York (NY): Springer-Verlag; 1985.

24.

Newson

. Attributable and unattributable risks and fractions and other scenario comparisons. Stata J. 2013;13(4):672–698.

25.

Health Program Analysis & Evaluation. 2003-2004 Active Duty Health Study. Revised by Health Program Analysis & Evaluation (HPA&E) [cited 2015 Aug 21]. Available from: http://www.tricare.mil/ocfo/_docs/ADHS%20Final%20Report_Dec%2006.pdf.

26.

Hotopf

Hull

Fear

. The health of UK military personnel who deployed to the 2003 Iraq war: a cohort study. Lancet. 2006;367(9524):1731–1741.

27.

Hunt

Wessely

Jones

. The mental health of the UK Armed Forces: where facts meet fiction. Eur J Psychotraumatol. 2014;5. doi: 10.3402/ejpt.v5.23617.

28.

McFarlane

Hodson

Van Hooff

. Mental Health in the Australian Defence Force: 2010 ADF Mental Health Prevalence and Wellbeing Study [cited 2015 Aug 21]. Available from: http://www.defence.gov.au/Health/DMH/Docs/MHPWSReport-FullReport.pdf.

29.

Sundin

Fear

Iversen

. PTSD after deployment to Iraq: conflicting rates, conflicting claims. Psychol Med. 2010;40(3):367–382.

30.

Sundin

Herrell

Hoge

. Mental health outcomes in US and UK military personnel returning from Iraq. Br J Psychiatry. 2014;204(3):200–207.

31.

Hoge

Castro

. Post-traumatic stress disorder in UK and US forces deployed to Iraq. Lancet. 2006;368(9538):837.

32.

Bean-Mayberry

Yano

Washington

. Systematic review of women veterans’ health: update on successes and gaps. Womens Health Issues. 2011;21(Suppl):S84–S97.

33.

Riddle

Smith

. Millennium Cohort: the 2001-2003 baseline prevalence of mental disorders in the U.S. military. J Clin Epidemiol. 2007;60(2):192–201.

34.

Smith

Ryan

Wingard

35.

Sareen

Henriksen

Bolton

. Adverse childhood experiences in relation to mood and anxiety disorders in a population-based sample of active military personnel. Psychol Med. 2013;43(1):73–84.

36.

Zamorski

Boulos

. The impact of the military mission in Afghanistan on mental health in the Canadian Armed Forces: a summary of research findings. Eur J Psychotraumatol. 2014;5. doi: 10.3402/ejpt.v5.23822.

37.

Armed Forces Health Surveillance Center. Associations between repeated deployments to Iraq (OIF/OND) and Afghanistan (OEF) and post-deployment illnesses and injuries, active component, U.S. Armed Forces, 2003-2010. Part II. Mental disorders, by gender, age group, military occupation, and “dwell times” prior to repeat (second through fifth) deployments. MSMR. 2011;18(9):2–11.

38.

Hooper

Rona

Jones

. Cigarette and alcohol use in the UK Armed Forces, and their association with combat exposures: a prospective study. Addict Behav. 2008;33(8):1067–1071.

39.

Larson

Highfill-McRoy

Booth-Kewley

. Psychiatric diagnoses in historic and contemporary military cohorts: combat deployment and the healthy warrior effect. Am J Epidemiol. 2008;167(11):1269–1276.

40.

Boulos

Zamorski

. Military occupational outcomes in Canadian Armed Forces personnel with and without deployment-related mental disorders. Can J Psychiatry. Forthcoming 2016.