Mortality Among Male U.S. Army Soldiers Within 18 Months of Separation

ABSTRACT

Introduction

Understanding how soldiers die after separation from military service, particularly those who die shortly after separating from service, may help to identify opportunities to ease transitions to civilian life.

Materials and Methods

Mortality data were analyzed for male U.S. Army soldiers who died within 18 months of their separation from service (from 1999 to 2011). Descriptive statistics were calculated for natural, accidental, suicidal, homicidal, undetermined, and legal/operation of war deaths and were stratified by age, component, time in service, and type of discharge. Crude and age-adjusted mortality rates, standardized for all years using the 2004 Regular Army population, were also calculated. The Public Health Review Board of the U.S. Army Public Health Center approved this study as Public Health Practice.

Results

Of the 1,884,653 male soldiers who separated from service during the study period, 3,819 died within 18 months of separation. A majority of all separations were Reserve or National Guard (58%), and 62% of decedents were in the Reserve or National Guard. Deaths from natural causes (38%) were the most common, followed by accident deaths (34%), suicides (20%), homicides (6%), undetermined deaths (2%), and legal/operation of war deaths (<1%). For overall mortality, age-adjusted rates were higher among the male U.S. population when compared to male soldiers who had separated from the Army.

Conclusions

The time period immediately following separation from the Army presents a unique challenge for many soldiers. Developing more effective pre-separation prevention programs that target specific risks requires knowing the causes of death for natural deaths, suicides, and accidents. Over half of all deaths occurring shortly after separation from service are preventable. Continued surveillance of specific causes of preventable deaths following separation can improve pre-separation prevention programs and transitions to post-service care.

INTRODUCTION

Men and women in the U.S. Army have lower overall mortality rates because of accession standards, training that emphasizes health and physical fitness, and universal healthcare access for active duty soldiers.¹ However, much less is known about mortality rates, manners of death, and causes of death after separation from service (subsequently referred to as separation) because the Army does not routinely characterize mortality for separated soldiers. Understanding how soldiers die after separation, particularly those who die shortly after separation, presents a unique opportunity to identify soldiers at increased risk of early post-separation death, provide them with customized information regarding helping services in their respective community, and arrange a direct transition to United States Department of Veterans Affairs (VA) clinical care.

The time period after separation may pose an increased risk of death for soldiers for several reasons. First, soldiers with serious physical or behavioral health conditions are more likely to separate than healthy soldiers.^2,3 Second, soldiers exposed to combat are more likely to have a variety of mortality risk factors, such as increased risk-taking propensity, aggression, and alcohol misuse.⁴ Third, many separated soldiers experience major stressors, including difficulties obtaining employment, managing finances, maintaining strong social and family supports, and managing physical or psychological injuries associated with their service.⁵ Finally, soldiers are no longer held to the same medical and physical training (PT) standards in their civilian roles after separation. These factors, alone or in combination, may contribute to mortality risk from external (i.e., an unexpected and injurious event) and natural (i.e., death due to disease, illness, or the aging process) causes.

Research on mortality risk after separation for soldiers who served during Operation Enduring Freedom (OEF) (i.e., war in Afghanistan)/Operation Iraqi Freedom (OIF) has focused primarily on suicides, given the increasing trends in suicide rates seen among both active duty and veteran soldiers.^6,7 Large cohort studies of all U.S. military personnel who served in OEF/OIF have consistently shown an increase in suicide risk following separation, with rates most elevated shortly after separation and remaining higher than those observed in the U.S. civilian population.^8,9 Fewer studies have examined other external manners of death, but those studies have shown that U.S. military personnel are at increased risk of death from accidents and homicides after separation compared to their time in service.^10,11 Despite increases after separation, the overall mortality rates for accidents and homicides are in line with, or lower than, those for civilians.

To date, no prior studies of OEF/OIF soldiers have characterized mortality for natural deaths following separation, reported specific causes of deaths (except Bullman et al.¹⁰ for motor vehicle accidents), or compared mortality outcomes among all three Army components (Regular Army, Army National Guard, and U.S. Army Reserve). Characterizing post-separation mortality for natural deaths is critical to contextualizing the overall mortality burden as compared to deaths from external causes (i.e., suicide, accident, homicide, and legal/operation of war deaths). Developing more effective pre-separation prevention programs that target specific risks (e.g., driving accidents vs. drug overdose) requires knowing the causes of death for natural deaths, suicides, and accidents.

Studies that have examined post-separation mortality differences between components combined Reserve and National Guard service members.¹² Within the Army, both the Reserve and National Guard components are large. The Army Reserve is similar in size to the entire U.S. Marine Corps, and the Army National Guard is larger than any of the other service branches. Differences in mortality outcomes between Army Reserve and Army National Guard personnel may inform component-specific prevention and transition efforts.

The aims of this study were to describe mortality rates among male soldiers who, from 1999 to 2011, died within 18 months of separation from service and to compare male soldier mortality rates during this period to male rates in the U.S. population. We include descriptions of specific causes of death for natural, accidental, suicidal, homicidal, undetermined, and legal/operation of war deaths and compare mortality rates for manners of death between the three Army components.

METHODS

Study Population

This study included 3,819 U.S. Army (Regular, National Guard, and Reserve components) male soldiers who died within 18 months of separation from service (from 1999 to 2011). The data analysis was conducted in 2019.

Measures

De-identified administrative data for male soldiers who died were obtained from the Joint DoD-Department of Veterans Affairs Suicide Data Repository (SDR). This repository contains records of all deaths (not limited to suicides) from January 1, 1979 to the present for anyone with prior military service as well as active duty service members. Data from seven different sources housed within the Suicide Data Repository were used to identify sex, date of birth, date of death, date the soldier entered service, most recent date of separation from service, underlying cause of death (as identified by the ICD-10 code from the National Death Index), Army component at the time of separation, and characterization of service code (i.e., the type of discharge a soldier received upon exiting the Army). Table S1 provides a definition for character of service.

Denominator data were obtained from the Defense Manpower Data Center for soldiers who separated from service due to returning to civilian life, retiring, serving in the Individual Ready Reserve or Inactive National Guard, or on standby. Variables included the year and month of separation, component, gender, age, and total length of military service (time in service) in months. For the calculation of U.S. civilian age-adjusted mortality rates, data were retrieved from the U.S. CDC WONDER online database.¹³

Mortality Classification

ICD-10 codes were classified into individual corresponding underlying causes of death using the National Center for Health Statistics “358 recodes” for causes of death.¹⁴ The causes of death categories corresponding to the ICD-10 codes were classified into six manners of death: natural, accidental, suicidal, homicidal, undetermined, and legal/operation of war. Table S1 provides a definition for each manner of death.

Statistical Analysis

Descriptive statistics were generated for age, component, time in service, and character of service for each manner of death. Crude and age-adjusted mortality rates were also calculated.

Crude rates were calculated using 12-month moving averages, allowing for smoothing of the rate curve over time when compared to single annual rates. To calculate rates, 12 months of denominator data were required. Therefore, the first crude rate was calculated for January 2000 and includes the sum of all deaths within 18 months of separation (numerator) that occurred during January to December 1999. The person-years denominator was calculated from the number of separations and the total months lived during that 12-month period.

Direct age-adjusted annual mortality rates were calculated for each category of death per 100,000 person-years to compare male soldiers and the U.S. population. Rates were standardized for all years using the 2004 active duty Army population.¹⁵ Age-adjusted rates were calculated using the Statistical Analysis System standard rate procedure, which provides directly standardized rates and risk estimates.¹⁶ If mortality rates are homogenous across age strata, this procedure also provides the Cochran-Mantel-Haenszel method to compute a pooled estimate of mortality that is based on stratum-specific mortality estimates. Age-adjusted overall mortality rates were also compared across years using a z-test statistic.

Joinpoint software was used to test for differences in trends for age-adjusted mortality rates among male soldiers in the Army and U.S. male population.¹⁷ For separated soldiers, the December crude 12-month moving average rates for each age-at-death strata were used to compute the age-adjusted mortality rates. U.S. age-adjusted mortality rates were calculated for each category of death as the total number U.S. deaths (numerator) divided by the annual U.S. population from CDC WONDER (denominator).

Ethics

The Public Health Review Board of the U.S. Army Public Health Center approved this study as Public Health Practice (#16-500.M1).

RESULTS

A total of 1,884,653 male U.S. Army soldiers separated from service from January 1999 to December 2011. A majority were 20 to 34 years of age (66%); 42% were active duty, 21% served in the Reserve, and 37% served in the National Guard.

Among these soldiers, 3,819 (0.2%) died within 18 months of separating from service. The majority of those who died were less than 30 years of age (52%). Sixty-two percent of decedents were in the Reserve or National Guard, 40% served from 1 to 4 years, and 25% of soldiers served over 20 years (Table I). All-cause age-adjusted mortality rates were highest among Reservists, followed by Regular Army and National Guard soldiers (Table II). Deaths from natural causes (38%) were the most common, followed by accidental deaths (34%), suicides (20%), homicides (6%) undetermined deaths (2%), and legal/operations of war deaths (<1%). Notably, a greater proportion of accidental deaths (5%) and suicides (5%) had a character of service as other than honorable when compared to natural deaths (1%).

TABLE I.

Characteristics of Male U.S. Army Soldiers^ab by Manner of Death, 1999-2011

Demographic characteristic	Total(n = 3819)		Natural(n = 1445)		Accident(n = 1299)		Suicide(n = 780)		Homicide(n = 219)		Legal/war e(n = 15)		Undetermined(n = 61)
	n	%	n	%	n	%	n	%	n	%	n	%	n	%
Age group (years)
17-19	202	5	11	1	95	7	77	10	18	8	0	0	1	2
20-24	1105	29	127	9	551	42	295	38	99	45	6	40	27	44
25-29	683	18	119	8	294	23	192	25	59	27	6	40	13	21
30-34	343	9	100	7	131	10	82	11	17	8	1	7	12	20
35-39	289	8	131	9	77	6	67	9	11	5	0	0	3	5
40-44	315	8	214	15	59	5	32	4	4	2	1	7	5	8
45-64	877	23	740	51	91	7	34	4	11	5	1	7	0	0
Component
Active	1433	38	404	28	530	41	359	46	100	46	6	40	34	56
National	1159	30	528	37	367	28	190	24	56	26	5	33	13	77
Reserve	1222	32	510	35	401	31	230	30	63	29	4	27	14	23
Time in service (months)
1-12	535	14	53	4	245	19	176	23	49	23	4	27	8	13
13-48	993	26	138	10	476	37	262	34	91	42	5	33	21	34
49-120	768	20	183	13	320	25	200	26	43	20	4	27	18	30
121-240	546	14	302	21	135	10	80	10	18	8	0	0	11	18
>241	955	25	763	53	118	9	55	7	14	7	2	13	3	5
Character of service
Honorable	1499	39	470	34	560	45	341	46	88	42	8	57	32	54
Other c	133	3	20	1	63	5	36	5	9	4	0	0	5	8
Uncharacterized	386	10	50	4	166	13	129	17	35	17	3	21	3	5
Unknown d	1628	43	836	61	454	37	236	32	80	38	3	21	19	32

Sources: National Death Index, Defense Casualty Analysis System, Active Duty Military Personnel Master, Reserve Components Common Personnel Data System (RCCPDS) Master, Active Duty Military Personnel Transaction, RCCPDS Transaction, and Contingency Tracking System Activation OCO files.

^a Deceased soldiers represent those who died within 18 months of separating from service.

^b Missing observations for deceased soldiers: age (n = 5), time in service (n = 17), and character of service (n = 168). Time in service for the deceased soldiers does not include those with less than 1 month of service.

^c Other includes under other than honorable, bad conduct, and dishonorable-dismissal.

^d Ninety-seven percent of unknown discharge types were Guard and Reserve.

^e Legal intervention includes injuries inflicted by the police or other law-enforcing agents, including military on duty, in the course of arresting or attempting to arrest lawbreakers, suppressing disturbances, maintaining order, and other legal actions, and excludes injuries caused by civil insurrections. Operation of war includes injuries to military personnel or civilians caused by war or civil insurrection, including those occurring during the time of war or insurrection and after cessation of hostilities. The number of deaths in this category (i.e., ICD-9 codes E990-E999 for 1981 through 1998 and ICD-10 codes Y36.0-0.9 for 1999 to present) is small (e.g., 14 in 2003). These operation of war deaths are included in total but cannot be reported separately using the Web-based Injury Statistics Query and Reporting System (WISQARS).²¹

TABLE II.

Age-adjusted Rates (Males) for All-cause, Accident, Natural, and Suicide Deaths, by U.S. Army Component and U.S. Population, 2000-2011

	All-cause					Accident
Year	U.S.	All Army	Regular	Guard	Reserve	U.S.	All Army	Regular	Guard	Reserve
2000	202.9	170.8	172.0	123.5	219.0	51.3	46.8	42.7	31.0	75.6
2001	206.1	196.3	183.6	157.4	252.2	52.4	55.9	43.0	57.1	83.6
2002	206.0	172.5	160.1	151.1	209.7	54.6	55.7	54.8	43.8	76.9
2003	205.7	170.8	164.1	150.1	245.5	54.7	49.8	48.1	44.1	56.2
2004	201.0	196.3	201.6	117.8	262.0	55.1	59.4	63.4	47.5	99.1
2005	204.7	172.5	147.6	168.4	293.6	58.4	69.8	48.8	70.2	133.6
2006	206.2	165.3	178.3	114.4	212.9	61.3	61.4	58.4	32.7	116.5
2007	201.7	162.9	144.4	156.5	220.5	60.0	67.3	56.4	72.8	98.8
2008	196.7	192.3	174.1	171.2	249.8	56.5	83.4	79.2	81.1	89.9
2009	191.2	173.8	173.4	133.0	222.7	51.7	65.2	70.5	52.7	76.7
2010	186.6	206.1	192.3	139.1	378.2	51.4	80.4	79.0	61.0	129.6
2011	187.7	195.6	170.8	143.2	373.6	52.9	80.9	69.7	62.3	166.6
	Natural					Suicide
Year	U.S.	All Army	Regular	Guard	Reserve	U.S.	All Army	Regular	Guard	Reserve
2000	109.9	79.3	84.8	65.9	87.6	20.7	31.7	33.3	22.1	36.7
2001	108.3	87.6	94.1	66.3	93.8	21.1	39.7	36.5	23.1	64.4
2002	107.5	68.8	61.0	62.0	79.0	21.3	34.5	29.7	35.3	38.1
2003	106.8	72.8	49.5	70.6	97.1	21.1	41.6	48.0	23.5	68.0
2004	102.6	65.3	76.9	41.4	90.4	21.3	38.6	45.1	21.8	59.2
2005	102.7	61.2	47.6	51.0	93.1	21.0	37.8	37.9	35.2	48.7
2006	100.4	56.6	65.3	43.4	45.4	21.2	32.4	33.9	27.4	44.5
2007	97.2	47.8	38.1	38.2	71.0	21.6	39.0	40.5	34.9	46.4
2008	96.7	50.0	45.6	33.5	80.0	21.8	44.4	34.6	39.7	69.4
2009	86.6	45.5	31.9	35.5	80.0	21.8	42.2	48.5	33.9	43.3
2010	93.2	60.9	46.9	53.1	95.2	22.8	50.3	52.3	20.3	116.1
2011	92.6	42.4	38.6	31.7	60.4	23.1	58.8	55.0	38.2	108.8

In the trend analysis, there was no statistically significant change in the age-adjusted death rate for the U.S. population or U.S. Army soldiers between the years 2000 and 2011 (Figure S1). However, all-cause age-adjusted death rates for males in the U.S. population decreased from 202.9 (95% CI, 201.8-203.9) per 100,000 population in 2000 to 187.6 (95% CI, 186.7-188.6) in 2011 (Table II). All-cause adjusted death rates for male U.S. Army soldiers increased from 170.8 (95% CI, 150.5-191.1) per 100,000 soldiers in 2000 to 196 (95% CI, 172.1-220.5) in 2004, then dipped to 162.9 (95% CI, 142.3-183.7) in 2007 and began to rise from 192.3 (95% CI, 169.6 to 214.9) in 2008 to 195.6 (95% CI, 170.8-220.3) in 2011 (Table II).

The plurality of natural deaths included malignant neoplasms (44%), followed by major cardiovascular disease (28%) (Table III). Crude mortality rates (data not shown) for natural deaths were higher among Reserve and National Guard soldiers when compared to Regular Army soldiers. However, when compared to the male U.S. population, age-adjusted rates for natural deaths were consistently lower for the Army as a whole (Fig. 1). Moreover, age-adjusted rates for natural death decreased over time among all male soldier groups. A more marked decrease occurred among National Guard and Regular Army soldiers when compared to the U.S. population. Age-adjusted rates of major cardiovascular disease were consistently lower among male U.S. Army soldiers when compared to the male U.S. population (Figure S2). Cancer age-adjusted rates for male soldiers were higher than the U.S. male age-adjusted rate from 2000 to 2006, and then fell below the male U.S. population age-adjusted rates in 2007 (Figure S2). There was a substantial decrease in the age-adjusted rates for cancer deaths among male soldiers from a peak rate of 48.6 deaths per 100,000 in 2001 to a low rate of 14.8 deaths per 100,000 in 2011.

TABLE III.

Underlying Cause of Natural Deaths among Recently Separated Male U.S. Army Soldiers^a, 1999-2011

Disease	n (%)
All	1,445 (100)
Malignant neoplasmb	640 (44)
Other malignant neoplasmsc	261
Trachea, bronchus and lung	126
Colon, rectum, anus	96
Leukemia	58
Breast	0
Pancreas	36
Urinary tract	28
Stomach	23
Prostate	12
Major cardiovascular diseased	401 (28)
Ischemic heart disease	194
Other diseases of the hearte	115
Cerebrovascular diseases	42
Hypertensive heart disease with or without renal disease	33
Other diseases of the circulatory system f	11
Hypertension and hypertensive renal disease	6
Atherosclerosis	0
All other diseases	221 (15)
Symptoms, signs, and lab findings	73 (5)
Chronic liver disease and cirrhosis	39 (3)
Diabetes mellitus	21 (1)
Influenza and pneumonia	14 (1)
Chronic lower respiratory diseases	10 (1)
Human immunodeficiency virus	12 (1)
Congenital and chromosomal abnormalities	9 (1)
Nephritis, nephrotic syndrome, and nephrosis	5 (<1)
Pregnancy, childbirth, and puerperium	0 (0)

Sources: U.S. Army separated soldier deaths are from the National Death Index and Defense Casualty Analysis System.

^a Deceased soldiers represent those who died within 18 months of separating from service.

^b Malignant neoplasm deaths are identified with the ICD-10 underlying causes of death codes C00-C97 and exclude underlying death codes D00-D48.

^c Other malignant neoplasms are identified with ICD-10 underlying cause of death codes C02-C15, C17, C22, C31-C32, C37, C41, C43-C45, C47, C49, C53-C56, C62, C71-C74, C78-C79, C80-C81, C82-C85, C90, and C97.

^d Major cardiovascular disease deaths are identified with the ICD-10 underlying causes of death codes I00-I78.

^e Other diseases of the heart deaths are identified with ICD-10 codes I00-I09 and I26-I51.

^f Other diseases of the circulatory system deaths are identified with ICD-10 underlying causes of diseases codes I71-I78 and exclude codes I80-I81 and I83-I99.

FIGURE 1.

Age-Adjusted Mortality Rates per 100,000 Person-Years for Natural and Accidental Deaths by U.S. Army Component, 2000-2011

The majority of accidental deaths were by motor vehicle (59%), followed by drug/alcohol overdose (25%), and motorcycle (10%). While the number of motor vehicle deaths among soldiers decreased from 1999 to 2011, the number of drug-related deaths increased during the same time period. From 2009 to 2011, the number of drug-related deaths was higher than motor vehicle, motorcycle, or alcohol-related deaths. Crude mortality accidental death rates ranged from 47.0 deaths per 100,000 in 2000 to 81.8 deaths per 100,000 in 2008, the highest rate within the time frame of this study. Across most years, Army age-adjusted rates for accidental deaths among male soldiers were higher than the male U.S. population, with the largest differences occurring among male Reserve soldiers (Fig. 1).

Gunshot wound was the most common (66%) method of suicide, followed by hanging/asphyxiation (22%), other (e.g., jumping from a high place) (4%), and drug/alcohol overdose (8%). Crude rates of death by suicide for male soldiers increased from 30.7 deaths per 100,000 in 2000 to 57.3 deaths per 100,000 in 2011. From 2000 to 2011, age-adjusted rates for death by suicide were consistently higher among male soldiers, as a whole, when compared to the U.S. population. Age-adjusted suicide rates also increased, most notably in 2010 and 2011 among the male Army population, when compared to the male U.S. population.

For homicides, 77% were due to gunshot wound, 11% because of a sharp object (e.g., stabbing), and 12% due to other means. Of deaths in the legal/operation of war category, which were rare, 93% were due to other legal intervention and their sequelae (e.g., injuries inflicted by the police or other law-enforcing agents, including military on duty) and 7% were due to operations of war and their sequelae (e.g., injuries to military personnel or civilians caused by war or civil insurrection, including those occurring during the time of war or insurrection and after cessation of hostilities). Cause codes for undetermined deaths were not provided. Rates for these categories were not calculated due to the small numbers (e.g., <20) in each year.

DISCUSSION

This is the first study to characterize all-cause mortality for a cohort of U.S. Army soldiers 18 months immediately following their separation from the military. We compared trends with the overall U.S. population, elucidated deaths attributable to natural causes and accidents, and confirmed the previously reported high rate of suicide following military service.¹⁸

Sixty percent of the deaths were considered behavioral or potentially preventable (i.e., accidents, homicides, suicides). From 1999 to 2011, age-adjusted rates of both accidental death and suicide were significantly higher among separated male soldiers when compared to the male U.S. population. In contrast, the age-adjusted rates of natural deaths (e.g., cancer, cardiovascular disease) among separated male soldiers were considerably lower than the general male U.S. population.

In this study, the second highest cause of accidental death was drug overdose/alcohol poisoning. Moreover, the number of drug-related deaths increased from 1999 to 2011. However, similar trends have been observed in the overall population of the United States, especially regarding death from opioid overdoses.¹⁹ This is concerning given that soldiers have a number of options for addressing substance use disorders while in the military, such as substance use screening in primary care, evidence-based treatment programs for soldiers diagnosed with substance use disorders, and monitoring of patient outcomes.²⁰ Wraparound services during the transition from military to civilian life are critical to ensuring soldiers do not fall into treatment gaps.

Crude mortality rates for natural deaths were higher among male Reserve and National Guard soldiers when compared to male Regular Army soldiers. The differences in rates between these groups could be attributable to differences in the intensity and duration of service requirements, structural support, and access to health care. Although all soldiers, regardless of service component, are required to maintain Army physical fitness standards, structural support barriers may inhibit routine PT for Reserve and National Guard soldiers. For example, PT is built into the workday duties of most Regular Army soldiers. In contrast, time allotted for PT may not be allowed or prioritized in the civilian jobs in which Reserve and National Guard soldiers are employed. Additionally, after being activated, Reserve and National Guard soldiers have a limited time period (i.e., 180 days) in which to access comprehensive healthcare services included in their military benefits. Conversely, Regular Army soldiers are consistently eligible for universal healthcare which may allow for medical interventions for medical conditions that cause natural death. In this study, 44% of natural deaths were due to malignant neoplasms (cancer). Early identification and intervention and treatment for cancer are critical to long-term success, and not having consistent access to health care could be a barrier to minimizing cancer mortality.

There are numerous hypotheses that may explain the higher proportion of age-adjusted accidental and suicidal deaths among male military separations when compared to the U.S. male population. For accidental deaths, soldiers with combat experience may be more likely to take risks due to combat exposures. Suicide following separation from the military may be explained by (1) a lack of social support and thus difficulty adapting to civilian life, (2) difficulty securing employment, and (3) barriers to accessing medical care (e.g., absent or limited health insurance). Of note, for all-cause mortality, some soldiers may have been at an increased risk of death before separation for a variety for reasons, such as behavioral health problems, including but not limited to PTSD, past suicide attempt, substance abuse, personality disorders, and criminal behavior; or physical problems/diseases, such as cancer. There is no reason to expect that these risks would vanish upon separation, resulting in increased risk of accident, suicide, and undetermined deaths and, perhaps, natural deaths or homicide. Furthermore, there is a possibility that certain problems develop as a result of separation. For example, soldiers may develop depression after separating from service provided social support systems are not present in the civilian sector, thus increasing the risk for suicidal behavior. However, the reasons for these higher mortality rates have not been sufficiently explored and should serve as a priority for future study.

Limitations

Several important limitations must be considered in this study. First, military medical records were not available for this study nor were medical records for care received after separation. Therefore, it is unclear whether the military was notified of the causes of natural deaths, whether medical conditions associated with natural deaths had contributed to separation, and if so, whether the affected soldiers received VA care for those conditions. Second, key demographic variables such as race, education, and rank were not used to compare specific mortality estimates. Future studies should consider analyzing these potential determinants of mortality to determine if there is a relationship between these demographic covariates and cause-specific mortality. Third, although characterization-of-service data were available (e.g., type of discharge), there were high proportions of missing values (21%-61%) for all manners of death, thus preventing a comprehensive analysis of the relationship between characterization of service and death. Fourth, there is a possibility that there were soldiers who left the service, rejoined, and then left the service again within the study period. This may lead to an inadvertent double-counting of individuals for the denominators. Provided this bias is present, the results downwardly bias mortality estimates following military separation. Fifth, this analysis was only conducted for male soldiers due to low counts of deaths across years and manners for female soldiers. Last, rates for deaths by homicide, cause-undetermined deaths, and legal/operation of war deaths were excluded because they were too unstable. The strengths of the study include the long analysis period (1999-2011); the inclusion of the Regular Army, Reserve Army, and the National Guard; and the inclusion of all-cause mortality instead of a focus on suicide and accidental death.

CONCLUSION

The time period immediately following separation from the military presents a unique challenge for many former soldiers. Studies that analyze the reasons for all-cause mortality are needed to inform the development of tailored programs that can ease transition for these individuals. Formative research and subsequent programmatic and policy changes are critical to addressing the increased rate of death in the period following military separation.

FUNDING

None declared.

CONFLICT OF INTEREST STATEMENT

None declared.