Early first deployment and risk of suicide attempt among first-term enlisted soldiers in the U.S. Army

Abstract

Objective:

We examined early first deployment and subsequent suicide attempt among U.S. Army soldiers.

Method:

Using 2004-2009 administrative data and person-month records of first-term, Regular Army, enlisted soldiers with one deployment (89.2% male), we identified 1,704 soldiers with a documented suicide attempt during or after first deployment and an equal-probability control sample (n=25,861 person-months).

Results:

Logistic regression analyses indicated soldiers deployed within the first 12 months of service were more likely than later deployers to attempt suicide (OR=1.7 [95% CI=1.5-1.8]). Adjusting for socio-demographic characteristics, service-related characteristics, and previous mental health diagnosis slightly attenuated this association (OR=1.6 [95% CI=1.5-1.8]). Results were not modified by gender, deployment status, military occupation, or mental health diagnosis. The population-attributable risk proportion for deploying within the first 12 months of service was 17.8%. Linear spline models indicated similar risk patterns over time for early and later deployers, peaking at month 9 during deployment and month 5 post-deployment; however, monthly suicide attempt rates were consistently higher for early deployers.

Conclusions:

Enlisted soldiers deployed within the first 12 months of service have elevated risk of suicide attempt during and after first deployment. Improved understanding of why early deployment increases risk can inform development of policies and intervention programs.

Rates of suicidal behavior in the U.S. Army increased sharply during the wars in Iraq and Afghanistan (Black, Gallaway, Bell, & Ritchie, 2011; Schoenbaum et al., 2014; Ursano, Kessler, Heeringa, et al., 2015) and remain elevated (Gibson, Corrigan, Kateley, Youmans Watkins, & Pecko, 2017; Pruitt et al., 2017). Recent findings suggest that soldiers who deploy early in their career have elevated risk of attempting suicide. In a study of twice-deployed soldiers, those who deployed for the first time within 12 months of entering service were approximately twice as likely to attempt suicide during or after their second deployment (Ursano et al., 2018). While those findings have potential to inform the Army’s training and deployment schedules, it is important to determine whether early deployment is also associated with increased risk during or after first deployment. Soldiers deploying for the first time are less experienced and may be a distinct population relative to those with two or more deployments, as attrition from service and reenlistment create a selection bias in the composition and risk/resilience profile of those more experienced soldiers (Hoge, Auchterlonie, & Milliken, 2006; Ireland, Kress, & Frost, 2012; Larson, Highfill-McRoy, & Booth-Kewley, 2008; Wilson et al., 2009). Research on multiple deployments is also complicated by other important time-related factors, such as dwell time (i.e., length of time between deployments) (Bonds, Baiocchi, & McDonald, 2010; MacGregor, Han, Dougherty, & Galarneau, 2012; MacGregor, Heltemes, Clouser, Han, & Galarneau, 2014; Office of the Surgeon Multi-National Corps-Iraq & Office of the Surgeon General United States Army Medical Command, 2009; Ursano et al., 2018). It is therefore important to expand previous findings with research on less experienced soldiers, particularly those in their first term of enlistment (typically the first 4 years of service) who have deployed once.

Soldiers in their first term of enlistment are at substantially increased risk of attempting suicide compared to those who have served longer (Ursano, Kessler, Stein, et al., 2015; Ursano et al., 2016). Importantly, more than half of soldiers deployed in support of the wars in Iraq and Afghanistan had only one deployment (Institute of Medicine, 2013). Focusing on this large group of less experienced soldiers with a first-term deployment can inform a substantial proportion of the Army’s deployment-related suicide attempt risk. First-term soldiers with one deployment are also the most likely targets of interventions and policy changes aimed at reducing risks associated with early deployment. We are aware of only one study examining the proximal effects of early first deployment on suicide risk, finding that first-term soldiers who deployed during their first year of service had elevated risk of dying by suicide during that deployment (Gilman et al., 2014). Research on early first deployment and non-fatal suicide attempts among first-term soldiers is lacking.

Here, we used administrative data from the Army Study to Assess Risk and Resilience in Servicemembers (Army STARRS) (Ursano et al., 2014) to examine the association of early deployment with risk of suicide attempt among first-term, Regular Army, enlisted soldiers on their first deployment and after returning home. Analyses were conducted before and after adjusting for basic socio-demographic characteristics, other service-related variables, and history of mental health diagnosis. We also investigated whether the association of early first deployment with suicide attempt was modified by other important risk factors, including gender, deployment status (currently vs. previously deployed), combat vs. non-combat occupation, and pre-deployment mental health diagnosis. Finally, we examined whether the pattern of suicide attempt risk over time varied as a function of time in service before first deployment.

METHOD

Sample

This longitudinal, retrospective cohort study used data from the Army STARRS Historical Administrative Data Study (HADS), which integrates 38 Army/DoD administrative data systems, including every system in which suicidal events are medically documented. The HADS includes individual-level person-month records for all soldiers on active duty between January 1, 2004 and December 31, 2009 (n=1.66 million) (Kessler et al., 2013), with records for those soldiers available as far back as January 1, 2000. This component of Army STARRS was approved by the Institutional Review Boards of the Uniformed Services University of the Health Sciences, University of Michigan Institute for Social Research, University of California, San Diego, and Harvard Medical School, which determined that the present study did not constitute human participant research because it relies entirely on deidentified secondary data.

The HADS contains administrative records for the 975,057 Regular Army soldiers on active duty during the study period (excluding activated Army National Guard and Army Reserve), including 9,791 who had a documented suicide attempt. Enlisted soldiers comprised nearly 99% (n=9,650) of these suicide attempters, with officers comprising the other 1% (Ursano, Kessler, Stein, et al., 2015). The current study focused on Regular Army enlisted soldiers in their first term of enlistment (first 4 years of service) who were in-theater on their first deployment or had returned from their first deployment (n=1,969 suicide attempt cases). Within this subset of the population, we excluded soldiers with discontinuous Regular Army service (i.e., previous Army National Guard/Reserve service or ≥ 2 consecutive months when not on active duty), a deployment date that preceded the date they entered the Army (indicating prior non-Army military service), a first deployment longer than 24 months, or a Special Forces occupation (owing to their atypical deployment schedule). The final analytic sample included 1,704 soldiers whose first administratively recorded suicide attempt was during or after their first deployment, plus a 1:200 equal-probability sample of person-months from all other Regular Army enlisted soldiers with one deployment (n=25,861 person-months). Control person-months were selected after stratifying the population of enlisted soldiers by gender, rank, time in service, deployment status (currently or previously deployed), and historical time. The control sample excluded all person-months indicating a documented suicide attempt or other non-fatal suicidal event (e.g., suicide ideation) (Ursano, Kessler, Heeringa, et al., 2015), and person-months in which a soldier died. Data were analyzed using a discrete-time survival framework with person-month as the unit of analysis (Willett & Singer, 1993), such that each month a soldier’s career was treated as a separate observational record. Control person-months were assigned a weight of 200 to adjust for under-sampling.

Measures

Suicide attempt

Nonfatal suicide attempts were identified using Army/DoD administrative records from: the Department of Defense Suicide Event Report (DoDSER) (Gahm et al., 2012), a DoD-wide surveillance mechanism that aggregates information on suicidal behaviors via a standardized form completed by medical providers at DoD treatment facilities; and ICD-9-CM diagnostic codes E950-E958 (indicating self-inflicted poisoning or injury with suicidal intent) from the Military Health System Data Repository (MDR), Theater Medical Data Store (TMDS), and TRANSCOM (Transportation Command) Regulating and Command and Control Evacuating System (TRAC²ES), which together provide healthcare encounter information from military and civilian treatment facilities, combat operations, and aeromedical evacuations (eTable 1). Although use of the DoDSER began in 2008, the DoDSER database includes records from its Army-specific predecessor, the Army Suicide Event Report (ASER), which was active from 2004-2007. The E959 code (late effects of a self-inflicted injury) was excluded, as it confounds the temporal relationships between the predictor variables and the suicide attempt (Walkup, Townsend, Crystal, & Olfson, 2012). Records from different data systems were cross-referenced to ensure all cases represented unique soldiers. For soldiers with multiple suicide attempts, we selected the first attempt using a hierarchical classification scheme that prioritized DoDSER records (additional details available elsewhere (Ursano, Kessler, Heeringa, et al., 2015)).

Time in service before first deployment and covariates

Administrative records (eTable 1) were used to calculate time in service before first deployment (≤ 12 months vs. > 12 months) and all covariates, including: socio-demographic variables (gender, race/ethnicity, education, marital status, age at entry into the Army); service-related variables (deployment status [currently deployed vs. previously deployed], and military occupational specialty [MOS; combat arms vs. other; eTable 2]), and an indicator variable for any mental health diagnosis before first deployment. The diagnosis variable was created from administrative medical records by combining all categories of ICD-9-CM mental health diagnostic codes (e.g., major depression, bipolar disorder, posttraumatic stress disorder, personality disorders), excluding postconcussion syndrome, tobacco use disorder, and supplemental V-codes that are not disorders (e.g., stressors/adversities, marital problems) (eTable 3).

Analysis Methods

All analyses were conducted using SAS version 9.4 (SAS Institute Inc., 2013). Time in service before first deployment was first examined in a univariable logistic regression model, followed by a series of multivariable models that adjusted for socio-demographic variables, service-related variables, and previous mental health diagnosis. Adjusting for other covariates, we separately examined two-way interactions of time in service before first deployment with gender, deployment status, military occupation, and mental health diagnosis to determine whether the association of early deployment with suicide attempt was modified by other characteristics. Logistic regression coefficients were exponentiated to obtain odds-ratios (OR) and 95% confidence intervals (CI). All univariable and multivariable logistic regression models included a dummy predictor for calendar month and year to control for increasing rates of suicide attempt from 2004 through 2009 (Ursano, Kessler, Heeringa, et al., 2015). Coefficients of other predictors can consequently be interpreted as averaged within-month associations based on the assumption that effects of other predictors do not vary over time. In addition, a population-attributable risk proportion (PARP) (Rothman & Greenland, 1998) was calculated to estimate the proportion of observed suicide attempts that would not have occurred if all soldiers had more than 12 months in service before first deployment (assuming that coefficients in each model represent causal effects of the predictors).

To examine whether early first deployment influenced the timing of suicide attempt risk, we stratified the sample by deployment status and used discrete-time survival models to estimate risk (suicide attempters per 100,000 person-months) by (a) month since deployment began among those currently on their first deployment, and (b) month since returning from deployment among those who had completed their first deployment. Splines (piecewise linear functions) were examined using chi-square difference tests, deviance, and the Akaike Information Criterion to identify nonlinearities in changes in risk over time. We then examined whether the pattern of these nonlinearities differed by time in service before first deployment (≤ 12 months vs. > 12 months).

RESULTS

The majority of soldiers in this sample were male (89.2%), at least high school educated (80.0%), White Non-Hispanic (67.5%), unmarried (59.4%), 17-20 years old when they entered the Army (62.2%). Half were currently deployed (50.1%), just over one-third had a combat arms occupation (34.1%), and 11.4% received a mental health diagnosis before their first deployment. Approximately one-third of soldiers deployed within the first year of service (34.1%) (Table 1).

Table 1.

Socio-demographic and service-related characteristics among first-term, Regular Army, enlisted soldiers with one deployment.^a

	Unweighted suicide attempt cases	Unweighted control person-months	Weighted total person-months	Weighted total percentage
	(N)	(N)	(N)	(%)
Gender
Male	1,415	23,073	4,616,015	89.2
Female	289	2,788	557,889	10.8
Education
< High schoolb	602	5,164	1,033,402	20.0
High school or more	1,102	20,697	4,140,502	80.0
Race/ethnicity
White Non-Hispanic	1,222	17,457	3,492,622	67.5
Other	482	8,404	1,681,282	32.5
Marital status
Currently married	780	10,499	2,100,580	40.6
Never/previously married	924	15,362	3,073,324	59.4
Age at Army entry
17–20 years	1,142	16,085	3,218,142	62.2
21–24 years	390	6,507	1,301,790	25.2
≥ 25 years	172	3,269	653,972	12.6
Deployment status
Currently deployed	462	12,962	2,592,862	50.1
Previously deployed	1,242	12,899	2,581,042	49.9
Military Occupational Specialty
Combat Arms	636	8,829	1,766,436	34.1
Other	1,068	17,032	3,407,468	65.9
Mental health diagnosis before first deployment
Yes	493	3,102	620,893	12.0
No	1,211	22,759	4,553,011	88.0
Time in service before first deployment
≤ 12 months	789	8,806	1,761,989	34.1
> 12 months	915	17,055	3,411,915	65.9
Total	1,704	25,861	5,173,904	100.0

^aThe sample of first-term enlisted soldiers with one deployment (n=1,704 suicide attempt cases, 25,861 control person-months) is a subset of the total sample of Regular Army soldiers (n=193,617 person-months) from the Army STARRS Historical Administrative Data Study (HADS). Control person-months were assigned a weight of 200 to adjust for the under-sampling.

^b< High School includes: General Educational Development credential (GED), home study diploma, occupational program certificate, correspondence school diploma, high school certificate of attendance, adult education diploma, and other non-traditional high school credentials.

In a univariable model that adjusted only for calendar time, soldiers who deployed within the first 12 months of service had higher odds of subsequently attempting suicide (OR=1.7 [95% CI=1.5-1.8]). The association of early deployment with suicide attempt persisted in a multivariable model that adjusted for socio-demographic variables (OR=1.7 [95% CI=1.5-1.8]). Including deployment status and military occupation (combat arms vs. other) resulted only in a slight attenuation of the odds associated with early deployment (OR=1.5 [95% CI=1.3-1.6]). When mental health diagnosis before first deployment was added as a sensitivity test, the elevated odds of suicide attempt associated with early deployment was largely unchanged (OR=1.6 [95% CI=1.5-1.8]) (Table 2). In separate analyses based on this full model, the two-way interactions of early deployment with gender, deployment status, military occupation, and mental health diagnosis were all nonsignificant. The PARP for deploying within the first 12 months of service (based on the full multivariable model, including socio-demographic variables, service-related variables, and mental health diagnosis) was 17.8%. Given that all combat arms soldiers were male during this time period, we repeated these analyses among males and found similar results.

Table 2.

Association of early first deployment with suicide attempt among first-term, Regular Army, enlisted soldiers with one deployment.^a

	Univariableb		Multivariable 1c		Multivariable 2d		Multivariable 3e
	OR	(95% CI)	OR	(95% CI)	OR	(95% CI)
Gender
Male	1.0	–	1.0	–	1.0	–	1.0	–
Female	1.7*	(1.5–2.0)	2.0*	(1.8–2.3)	2.1*	(1.8–2.4)	1.8*	(1.6–2.0)
χ 21	71.9* (p<0.0001)		113.7* (p<0.0001)		113.5* (p<0.0001)		66.4* (p<0.0001)
Education
< High schoolf	2.1*	(1.9–2.4)	2.1*	(1.9–2.3)	2.2*	(2.0–2.4)	2.1*	(1.9–2.2)
High school or more	1.0	–	1.0	–	1.0	–	1.0	–
χ 21	207.4* (p<0.0001)		203.6 (p<0.0001)		226.6* (p<0.0001)		190.0* (p<0.0001)
Race/ethnicity
White Non-Hispanic	1.0	–	1.0	–	1.0	–	1.0	–
Other	0.8*	(0.8–1.0)	0.8*	(0.7–0.9)	0.8*	(0.7–0.9)	0.9*	(0.8–1.0)
χ 21	9.8* (p=0.0017)		11.9* (p=0.0006)		11.4* (p=0.0007)		6.1* (p=0.0139)
Marital status
Currently married	1.0	–	1.0	–	1.0	–	1.0	–
Never or previously married	0.8*	(0.8–1.0)	0.8*	(0.7–0.8)	0.8*	(0.8–0.9)	0.9*	(0.8–1.0)
χ 21	12.5* (p=0.0004)		28.4* (p<0.0001)		12.3* (p=0.0004)		9.3* (p=0.0024)
Age at Army entry
17–20 years	1.2*	(1.0–1.4)	1.3*	(1.2–1.5)	1.3*	(1.2–1.5)	1.3*	(1.2–1.4)
21–24 years	1.0	–	1.0	–	1.0	–	1.0	–
≥ 25 years	0.9	(0.7–1.0)	0.8	(0.7–1.0)	0.9	(0.7–1.0)	0.9	(0.7–1.1)
χ 22	25.3* (p<0.0001)		43.1* (p<0.0001)		35.7* (p<0.0001)		36.2* (p<0.0001)
Deployment status
Currently deployed	1.0	–			1.0	–	1.0	–
Previously deployed	2.8*	(2.5–3.2)			2.7*	(2.4–3.0)	2.8*	(2.5–3.2)
χ 21	356.7* (p<0.0001)				323.8 (p<0.0001)		349.7* (p<0.0001)
Military Occupational Specialty
Combat Arms	1.1*	(1.0–1.2)			1.2*	(1.1–1.3)	1.2*	(1.1–1.4)
Other	1.0	–			1.0	–	1.0	–
χ 21	6.1* (p=0.0134)				8.7* (p=0.0031)		14.4* (p=0.0002)
Mental health diagnosis before first deployment
Yes	2.9*	(2.6–3.2)					3.1*	(2.8–3.4)
No	1.0	–					1.0	–
χ 21	388.2* (p<0.0001)						412.4* (p<0.0001)
Time in service before first deployment
≤ 12 months	1.7*	(1.5–1.8)	1.7*	(1.5–1.8)	1.5*	(1.3–1.6)	1.6*	(1.5–1.8)
> 12 months	1.0	–	1.0	–	1.0	–	1.0	–
χ 21	110.9* (p<0.0001)		107.1* (p<0.0001)		55.8* (p<0.0001)		99.6* (p<0.0001)

^aThe sample of first-term enlisted soldiers with one deployment (n=1,704 suicide attempt cases, 25,861 control person-months) is a subset of the total sample of Regular Army soldiers (n=193,617 person-months) from the Army STARRS Historical Administrative Data Study (HADS). Control person-months were assigned a weight of 200 to adjust for the under-sampling.

^bEach characteristic was examined separately in a logistic regression model that also included a dummy predictor variable for calendar month and year to control for secular trends.

^cMultivariable 1: Logistic regression that included socio-demographic characteristics (gender, education, race/ethnicity, marital status, age at Army entry), time in service before first deployment, and a dummy predictor variable for calendar month and year to control for secular trends.

^dMultivariable 2: Logistic regression that included all variables in the Multivariable 1 model, plus service-related characteristics (deployment status, military occupational specialty).

^eMultivariable 3: Logistic regression that included all variables in the Multivariable 1 and Multivariable 2 models, plus mental health diagnosis before first deployment.

^f< High School includes: General Educational Development credential (GED), home study diploma, occupational program certificate, correspondence school diploma, high school certificate of attendance, adult education diploma, and other non-traditional high school credentials.

To examine the possibility that the pattern of suicide attempt risk over time differed for those who deployed during vs. after their first 12 months of service, we first stratified the sample by deployment status in recognition of the distinct environmental context of soldiers who are still in-theater vs. those who have returned home. Spline models identified a single nonlinearity (knot) among those currently deployed, with risk of suicide attempt increasing from the start of deployment to a peak at month 9, and decreasing thereafter. The month of peak risk did not differ significantly for those who had deployed during vs. after the first 12 months of service. We therefore present splines for both groups with a knot at month 9, demonstrating the elevated risk over time among early deployers in-theater (Figure 1). Analysis of previously deployed soldiers also identified a single knot, with risk increasing to a peak in the 5^th month post-deployment, followed by a gradual decrease in risk over time. Month of peak risk similarly did not differ based on early vs. later deployment, so Figure 2 presents both groups with a knot at month 5 and demonstrates the persistently elevated post-deployment risk for soldiers who deployed within the first 12 months of service.

Figure 1. Risk of suicide attempt by month into first deployment.^{a, b}.

^aThe sample of first-term, currently deployed, enlisted soldiers with one deployment (n=462 suicide attempt cases, 12,962 control person-months) is a subset of the total sample of Regular Army soldiers (n=193,617 person-months) from the Army STARRS Historical Administrative Data Study (HADS). Control person-months were assigned a weight of 200 to adjust for the under-sampling.

^bMonthly risk estimates based on hazard rates and linear spline models.

Figure 2. Risk of suicide attempt by month since returning from first deployment.^{a, b}.

^aThe sample of first-term, previously deployed, enlisted soldiers with one deployment (n=1,242 suicide attempt cases, 12,899 control person-months) is a subset of the total sample of Regular Army soldiers (n=193,617 person-months) from the Army STARRS Historical Administrative Data Study (HADS). Control person-months were assigned a weight of 200 to adjust for the under-sampling.

^bMonthly risk estimates based on hazard rates and linear spline models

DISCUSSION

Among first-term enlisted soldiers, odds of suicide attempt during and after first deployment was approximately 70% higher for those who deployed within the first 12 months of service. This association persisted after adjusting for socio-demographic and service-related characteristics known to be associated with suicide attempts, including combat arms occupation (Ursano, Kessler, Naifeh, et al., 2017; Ursano, Kessler, Stein, et al., 2015; Ursano et al., 2016). Importantly, this association remained, only slightly attenuated, after adjusting for previous mental health diagnosis, indicating that this risk was not explained by mental health problems identified prior to deployment. Furthermore, the association of early deployment with suicide attempt did not differ based on important soldier characteristics, including gender (indicating the increased risk associated with early deployment was similar for males and females); having a combat occupation (indicating that additional combat-specific training neither increased nor decreased the risk associated with early deployment); deployment status (indicating that the association of early deployment with attempted suicide is similar whether soldiers are in-theater or have returned home); and previous mental health diagnosis (indicating that early deployment similarly increases risk for those who deploy with and without an identified mental health problem, although it is notable that early deployment allows less time for such problems to be recognized and diagnosed). Analyses based on the full model suggest that suicide attempts in this population could be reduced by nearly 18% if all soldiers were in service for more than 12 months before deploying.

These results expand upon previous research that found early first deployment increased the odds of suicide attempt among soldiers with two deployments (Ursano et al., 2018). Importantly, this study of first-term soldiers speaks to the large segment of the Army that is most likely to experience a first deployment and most directly affected by the timing of that first deployment. Overall, these two studies suggest that soldiers who deploy within the first 12 months of service have elevated risk of suicide attempt that persists across the first two deployments.

In addition to the overall increase in suicide attempt risk among early deployers, analysis of monthly attempt rates indicated that this heightened risk continued over time both during and after deployment. It is notable that the pattern of monthly risk (i.e., the timing of increasing risk, peak risk, and decreasing risk) was similar for soldiers who deployed during and after the first 12 months of service. Specifically, risk peaked around mid-deployment for both early and later deployers in-theater, but monthly suicide attempt rates were consistently elevated for those who deployed within the first 12 months of service. Similarly, among soldiers who had returned home, suicide attempt rates during the first 18 months post-deployment were consistently higher for those who deployed within the first 12 months of service. The point of greatest risk was similar for early and later deployers, peaking around the 5^th month after return. These findings suggest that although earlier deployment carries increased risk during and after deployment, the timing of the risk – which is similar for both early and later deployers – may be more related to the cycles of stressors that soldiers experience in-theater and after returning home. These patterns of suicide attempt risk during and after first deployment are generally similar to previous findings based on the full population of enlisted soldiers (Ursano et al., 2016). They are also consistent with patterns of self-reported mental health outcomes, such as depression (Mental Health Advisory Team (MHAT-V), 2008; Milliken, Auchterlonie, & Hoge, 2007). It is important to identify and understand what it is about the deployment cycle that creates these consistent patterns of increasing and decreasing risk. Although combat stressors accumulate throughout deployment, mid-deployment may represent a turning point when soldiers begin to become optimistic about returning home (Mental Health Advisory Team (MHAT-V), 2008). Returning from deployment is a challenging readjustment for both soldiers and their families (Institute of Medicine, 2013). This readjustment phase may last several months, as indicated by the gradual decrease in suicide attempt risk after the first half-year or so. Improved understanding of the stressors and readjustment challenges that soldiers face before, during, and after these points of peak risk can further identify targets for preventive intervention.

The mechanisms through which early first deployment increases risk of suicide attempt are not yet known. Studies indicate that rates of documented suicide ideation and attempts are highest during the first year of service (Ursano, Kessler, Stein, et al., 2015; Ursano et al., 2016; Ursano, Kessler, Stein, et al., 2017), suggesting that this initial period of intensive training and Army acculturation is particularly challenging for some soldiers (Klein, Hawes-Dawson, & Martin, 1991). For those who deploy during this early career phase, the stress associated with adapting to the Army may be compounded by numerous pre-deployment and peri-deployment individual and familial stressors (La Bash, Vogt, King, & King, 2008; Maguen et al., 2008; Troxel, Trail, Jaycox, & Chandra, 2016; Watkins, 2014). Early deployment also reduces the amount of time soldiers have to train and prepare, an important consideration in light of evidence that lower perceived preparedness for deployment is associated with increased psychological distress (Franz et al., 2013; Mott, Graham, & Teng, 2012; Renshaw, 2011). It is also possible that certain groups of soldiers (e.g., specific occupations) are independently at higher risk for both suicide attempt and early deployment, leading to a gross association between early deployment and suicide attempt that is inflated due to those soldiers’ mutual risk for both events. Although our findings did not identify group differences based on having a combat occupation, it may be possible to identify other relevant occupations and subgroups that explain the risk associated with early deployment.

It is also possible that early deployment increases suicide attempt risk for different reasons in different soldiers depending on their personal characteristics and experiences. For example, early deployers whose increased risk persists through their second deployment (Ursano et al., 2018) may be a distinct population with intervention needs that differ from those who recently deployed for the first time. There is a complex relationship between deployment history and population risk/resilience. On one hand, those with mental health problems are less likely to deploy (Ireland et al., 2012; Larson et al., 2008; Wilson et al., 2009), suggesting that soldiers who are repeatedly selected to deploy may be more resilient than those who deploy only once. On the other hand, repeated combat deployments are associated with increased risk of posttraumatic stress disorder compared to a single deployment (MacGregor et al., 2012; Reger, Gahm, Swanson, & Duma, 2009), highlighting the cumulative stressors experienced by multiple deployers. Thus, the characteristics (e.g., personality traits, mental health history, life stressors) of those with one deployment may differ from those of soldiers with multiple deployments. Owing to these differences, early deployment may affect single deployers and multiple deployers for different reasons. Similar diversity likely exists within the respective populations of single and multiple deployers. Identification of these various mechanisms can enhance pre-deployment screening (Warner, Appenzeller, Parker, Warner, & Hoge, 2011) and allow matching of high-risk early deployers to appropriate preventive interventions. This kind of textured analysis would mostly likely require focused survey research. Administrative data have limited capacity to elucidate why factors such as early deployment influence risk, which is a crucial step in advancing suicide research (Nock, Ramirez, & Rankin, 2019).

There are some noteworthy limitations to this study. First, administrative data may be incomplete and/or inaccurate. Medical records are subject to errors in clinician diagnosis and administrative/medical coding, and they are unlikely to capture all suicide attempts and mental disorders. In addition, our records do not capture mental health history prior to Army service, an important consideration given the substantial number of soldiers with a pre-enlistment history of mental disorders and suicidal thoughts and behaviors (Rosellini et al., 2015; Ursano, Heeringa, et al., 2015). Second, these results are specific to enlisted soldiers in their first four years of service with exactly one deployment. They are also specific to the 2004-2009 period of the wars in Iraq and Afghanistan. As such, the findings may not generalize to other service members, other phases of those wars, or other U.S. military conflicts.

With these limitations in mind, our findings indicate that first-term enlisted soldiers who deploy within the first 12 months of service have elevated risk of suicide attempt during and after first deployment, even after adjusting for other important variables. Although time in service before first deployment is a potentially modifiable risk factor, the elements of training and experience that are critical to altering risk are not yet known. Personnel management, including training and acculturation, plays an important role during the first year of service, and it may also be an important component of the Army’s efforts to prevent suicidal behavior. While extending the initial training period may increase soldiers’ perceived preparedness, such policy changes raise a number of important considerations (e.g., Myers, 2018) that must be addressed in order to balance the Army’s operational needs with its training and preparedness goals for new soldiers.