MILITARY MEDICINE, 180, 4:419, 2015

Overview of Depression, Post-Traumatic Stress Disorder, and
Alcohol Misuse Among Active Duty Service Members Returning
From Iraq and Afghanistan, Self-Report and Diagnosis
Sarah A. Mustillo, PhD*; Ashleigh Kysar-Moon, M A f; Susan R. Douglas, PhDp, Ryan Hargraves, MS±;
Shelley Mac Derm id Wadsworth, PhD*; Melissa Fraine, MPH§; Nicole L. Frazer, PhD§

ABSTRACT Previous studies have found deployment to combat areas to be associated with an increased risk of posttraumatic stress disorder (PTSD), depression, and alcohol abuse, but many previous studies were limited by samples that were not representative of the deployed military as a whole. This study presents an overview of these three mental health problems associated with deployment among Air Force, Army, Marine Corp, and Navy service members returning from deployment to Iraq and Afghanistan between January 2007 and March 2008. With postdeployment health data on over
50,000 service men and women, including diagnostic information, we were able to estimate prevalence of those who screened positive for risk of each disorder in self-report data at two time points, as well as prevalence of diagnoses received during health care encounters within the military health care system. The prevalence ranges of the three disorders were consistent with previous studies using similar measures, but service members in the Navy had higher rates of screening positive for all three disorders and higher prevalence of depression and PTSD diagnoses compared to the other branches. Further, PTSD risk was higher for service members returning from Afghanistan compared to Iraq, in contrast to previous findings.

INTRODUCTION
The m ental health burden on service m em bers returning from
Iraq (O peration Iraqi Freedom [OIF]) and A fghanistan (O per­ ation Enduring Freedom [OEF]) is considerable. A recent study estim ated that 18.5% o f service m em bers returning from O IF/
O E F m eet the criteria for either post-traum atic stress disorder
(PTSD) or depression.1 In fact, research indicates sim ply being deployed to Iraq or A fghanistan substantially increases risk for PTSD, as well as having deployments longer than 180 days, or serving in m ultiple tours.2'3 T o date, m any studies o f the
O IF and O EF populations have been conducted on lim ited or unrepresentative sam ples, such as sam ples o f a small num ber o f m ilitary bases, a single geographic area, one m ilitary branch, m ost com m only the Arm y, convenience sam ples o f recently deployed service m em bers, or help-seeking sam ples.4 Even studies w ith large sam ples or sam ples with m ultiple branches have been lim ited by lack o f representativeness o f respondents, and representative surveillance data have lim ited information on correlates o f disorder. This study includes linked self-report and m edical records data from a population o f active duty
Arm y, A ir Force, N avy, and M arine Corps returning from

*Department of Sociology, University of Notre Dame, 810 Flanner Hall,
Notre Dame, IN 46556. fDepartment of Sociology, Purdue University, 700 West State Street,
West Lafayette, IN 47907. fVanderbilt University, Peabody No. 151,230 Appleton Place, Nashville,
TN 37203.
§Behavioral Health Branch, Clinical Support Directorate, Defense
Health Agency, 7700 Arlington Boulevard Falls Church, VA 22042.
The opinions, interpretations, conclusions and recommendations are those of the author and are not necessarily endorsed by the Department of Defense. doi: 10.7205/MILMED-D-14-00335

M ILITARY MEDICINE, Vol. 180, April 2015

deploym ent betw een D ecem ber 2007 and M arch 2009 and focuses on branch- and deploym ent-related predictors of these disorders. The results provide the m ost com prehensive and representative exam ination to date o f patterns o f m ental health self-reports and diagnoses across the entire m ilitary popula­ tion with specific attention to differences by service branch.
O f the m any overview studies conducted with post­ deploym ent military personnel, m ost find an increased risk of m ental health sym ptom s or disorders, notably PTSD , depres­ sion, and alcohol m isuse.5~7 Factors associated w ith the developm ent o f m ental health issues include exposure to com bat, location o f deploym ent (e.g., Iraq vs. A fghanistan), and the length o f tim e since deploym ent.3,7 Research indi­ cates that in addition to PTSD, the prevalence o f depression am ong returning service m em bers is high, especially am ong those who served in OIF, with about 15% -20% o f post­ deploym ent service m em bers m eeting the criteria for m ajor depression diagnosis.s T here is some evidence that fem ales are at increased risk for depression in com parison to m ales follow ing deploym ent.3 H ow ever, a recent review o f m ilitary studies of PTSD following deploym ent to Iraq and Afghanistan asserts the findings o f sex differences for depression and
PTSD risk are not consistent across studies.9
A lcohol m isuse is a com m on problem experienced by returning O IF and OEF service members, with m ales being at greater risk com pared to fem ales.10 C onsistent with the find­ ings o f PTSD and depression, research describes that those service m em bers previously deployed to O IF report consid­ erably higher rates o f alcohol m isuse than those deployed to
O E F .10 H ow ever, M iliken et al report that w hile alcohol problem s are often reported am ong veterans, referrals for treatm ent were low. Because m uch o f this body o f know ledge is based on lim ited sam ples, this study exam ines a large

419

Overview of Depression, PTSD, and Alcohol Misuse in Active Duty Service Members sample of all four main branches of the military and seeks to provide new information on branch differences, deployment location differences, and discrepancies between service member self-reported symptoms and diagnoses by health care providers among formerly deployed OIF/OEF U.S.
Service men and women.
METHODS
This study utilizes data from the Defense Medical Surveillance
System, The Armed Forces Health Surveillance Center, U.S.
Department of Defense, Silver Spring, Maryland (January
2008 to March 2009; release date January 6, 2011). Service members returning from deployment complete the Post
Deployment Health Assessment (PDHA) within ±30 days of departure and the Post Deployment Health Reassessment
(PDHRA) between 90 and 180 days after redeployment or return from deployment. Both assessments include a selfreport instrument, typically completed online, and a structured assessment by a health care provider, completed face-to-face, or less often, by telephone to receive follow-up information and referrals if appropriate.1 Service members report symp­
1
toms related to their mental and physical health statuses and exposure concerns in both assessments.
To match the PDHA and PDHRA data, DoD linked assess­ ments completed between January 2008 and March 2009 and with the respective PDHA. In the event more than one PDHA existed for the same deployment, one was chosen at random for the matching process. A total of 143,248 matches were made. Explanations for nonlinkage ranged from the date of completion on the PDHRA being earlier than the date on the
PDHA, the PDHA and PDHRA corresponding to different deployments, and the completion of assessments outside the date range for this study.12 The sample was then further limited to active duty service members who served in Iraq or
Afghanistan only (n = 53,534).
To the PDHA/PRHRA dataset, we added Health Care
Encounter (HCE) data. HCE data are collected electronically on all active duty service members when they receive treat­ ment through the TRICARE system (included care from mil­ itary treatment facilities and outsourced facilities). Dates of service, diagnostic, and procedural information were included in these data. Approximately 33% of service members had at least one HCE for a mental health issue. Several demographic variables, including education level, component at the time of PDHA/PDHRA completion, race, and ethnicity, were extracted from the Defense Enrollment Eligibility Reporting
System database.
MEASURES
Dependent Variables
Depression
Risk of depression was assessed via self-report on the PDHA and PDHRA and depression diagnosis was determined via diagnostic codes in the HCE data. The PDHA and PDHRA

420

used two questions from the Patient Health Questionnaire including, “Over the past month have you,” experienced,
“little interest or pleasure in doing things,” and/or been “feel­ ing down, depressed, or hopeless?13' 14 Per Patient Health
Questionnaire guidelines, service members were considered at risk for depression if they responded affirmatively to either question and indicated they felt this way on “more than half of the days,” in the past month. From the HCE data, if service members had an HCE in which they were diagnosed with depression (International Classification of Disease,
9th Edition (ICD-9) codes 296.20-296.36 and 311 )15 in the
6 months after returning from deployment, they were coded as having depression.
Post-Traumatic Stress Disorder
PTSD risk was assessed using the Primary Care 4-item post-traumatic stress disorder screen (PC-PTSD)16 on both the PDHA and PDHRA, and PTSD diagnoses were deter­ mined using ICD-9 diagnostic codes in the HCE data.1
5
Consistent with previous use of the PC-PTSD measure, if service members reported two or more symptoms, they were coded as 1, which indicated a positive screen for
PTSD risk. For the diagnostic measures, we created a vari­ able equal to 1 if the respondent received a PTSD diagnosis
(ICD-9 = 309.81 )15 at any HCE in the 6 months following return from deployment.
Alcohol Problems
Risk for alcohol issues was assessed with a modified Two-Item
Conjoint screen.17 Service members completed questions on alcohol use on each assessment. They were asked if they “use alcohol more than you meant to,” and/or “have felt that you wanted to or needed to cut down on your drinking.” The time frame on the PDHA was before or during deployment, whereas the time frame on the PDHRA was over the past month. Service members were coded 1 if they positively endorsed either item. For the diagnostic measure, we included any diagnosis on any alcohol abuse or dependence disorder
(291.81, 291.9, 303, 303.01, 303.9, 305, 305.02).15
Independent Variables
Branch and Deployment Characteristics
We included information about branch, location of deploy­ ment (Iraq or Afghanistan), duration of most recent deploy­ ment (measured in months), and self-reported exposure to traumatic combat stress (vs. not exposed). Exposure to trau­ matic combat stress was assessed with three questions on the
PDHA that asked if service members (a) saw dead bodies or saw people wounded or killed, (b) were engaged in direct combat in which they discharged a weapon, and (c) if they ever felt they were in great danger of being killed during the most recent deployment. Service members were coded 1 on this measure if they positively endorsed any of the 3 items.

MILITARY MEDICINE, Vol. 180, April 2015

Overview of Depression, PTSD, and Alcohol Misuse in Active Duty Service Members
C o n tro l V aria b les

Demographics
We included variables for pay grade, service member’s marital status, race/ethnicity, sex, age, and whether they had more than a high school diploma.
Health Variables
To control for the potential overlap of physical and mental health conditions, we included several self-rated health vari­ ables. Respondents were asked on both the PDHA and PDHRA to rate their health over the past month (past month health) and compare their current health status to the time period before deployment (current self-rated health) on a 5-point
Likert-type scale. Service members who indicated their health as either fair or poor were coded as 1 on past month health and who indicated worse or much worse were coded 1 on current self-rated health. In addition, during the clinical assessment with the provider, respondents were asked if they were under medical care for physical or mental health issues on both the PDHA and PDHRA. Those who indicated they were receiving any medical care were coded as 1.
Comorbidities
To isolate the predictors of each individual disorder, we con­ trolled for the other two mental health concerns (i.e.„ the two that are not the dependent variable in that model). We include all control variables in the tables, but do not present them in the text because of space issues.

ANALYSES
We used logistic regression to model the odds of service members reporting depression, PTSD, and alcohol problems on the PDHA, the PDHRA, and HCE data in separate models along with the demographic and military characteristics men­ tioned above. That is, we ran a total of 9 models because each model addressed a different aspect of mental health need.
Because the PDHA is administered shortly after return from deployment, those three models addressed a previous condi­ tion or a condition with an early postdeployment onset.
Because service members take the PDHRA 3 to 6 months after returning, this measure captured those who were still experiencing symptoms and also those who began experienc­ ing symptoms with a delayed onset. Finally, the HCE models measured diagnosis in the military health system with a men­ tal health disorder, rather than risk of mental health problems.
Because of the large sample size, we interpret odds and percentage change in the odds to assess the size of the effect.
Missing values were less than 6% for all variables except
PDHA depression, which was missing for 14% of cases.
Because of a data transcription issue in the raw data, more than half the data for depression questions were missing for
Marine Corp and Navy. Because this appears to be a random rather than systematic error, Marine Corp and Navy are under-represented in the sample with listwise deletion, but

MILITARY MEDICINE, Vol. 180, April 2015

the results are still representative of the Marine Corp and
Navy as a whole. Because listwise deletion caused a loss of nearly 20% of cases when missing on all variables was com ­ bined, we performed multiple imputation with 10 imputed datasets. The multiple imputation results were extremely close to the unimputed models, and so we present the unimputed models for ease of interpretation and because efficiency is not an issue with the large sample size ( n = 41,351).

RESULTS
D e s c rip tiv e S ta tis tic s

Table I presents demographic and deployment characteristics of the sample. These figures demonstrate that the sample over represents soldiers and under represents seaman and Marines, but resembles the active duty military as a whole on race/ ethnicity, gender, and paygrade.18 As shown in Table II, mental health risk and diagnoses varied greatly by service branch. On the PDHA, Air Force had the lowest percentage of service members screening positive for depression risk
(3%), whereas Marine Corps has the highest (27.4%). For
PTSD risk, 4.8% of Air Force, 11.9% of Army, 12.9% of
Marines Corps, and 18.2% of Navy screened positive. Air
Force had the lowest percentage of service members screen­ ing positive for alcohol issues with less than 1%. About 6% of Army, 10% of Marine Corps, and 6% of Navy screened positive. On the PDHRA, prevalence of mental health risk
TABLE I.
Demographic and Deployment Characteristics of
Active Duty Service Members Returning From Iraq or Afghanistan
(A/ = 41,351)
M ean or %

SD

R ange

SM D e m ographic C h aracteristics
M ilitary B ranch
A ir Force

27.2

A rm y

68.7

0 -1

M arine C orps

3.1

0 -1

N avy

1.0

0 -1

Single

33.4

0 -1

D iv o rced /S ep arated /W id o w ed
M arried

10.2

0 -1

56.4

0 -1

0 -1

M arital S tatus

R ace/E thnicity
A sian

4.2

0 -1

B lack

15.0

0 -1

H ispanic

9.9

0 -1

O ther

1.3

0-1

69.6

0 -1

W hite
SM Sex— M ale
Pay G rade

89.5
2.032

0 -1
1.307

1 -6

H ealth S tatus
W orse H ealth

22.6

R eceiv in g M edical C are— PD H A

13.1

0-1

R eceiv in g M edical C are— PD H R A

25.4

0 -1

0 -1

D eploym ent C haracteristics
Iraq

81.6

0 -1

T rau m atic C o m b at E xposure

46.9

0 -1

T im e D ep lo y ed (M onths)

10.799

4.907

1^10

421

Overview o f Depression, PTSD, and Alcohol Misuse in Active Duty Sendee Members
TABLE II.
Mental Health-Related Outcomes Reported by Active Duty Service Members Returning From Iraq or Afghanistan on the
Post-Deployment Health Assessment (PDHA), the Post-Deployment Health Re-Assessment (PDHRA), and Healthcare Encounters (HCE),
N = 41,351
PDHA

PDHRA

Mean
SM Reported Outcomes
A ir Force (n = 11,264)
Depression Risk
PTSD Risk
Alcohol Risk
Army (n = 28,426)
Depression Risk
PTSD Risk
Alcohol Risk
M arines ( n = 1,265)
Depression Risk
PTSD Risk
Alcohol Risk
Navy (396)
Depression Risk
PTSD Risk
Alcohol Risk

SD

Range

Mean

SD

Range

0.032
0.048
0.008

0.177
0.213
0.090

0-1
0-1
0-1

0.034
0.046
0.024

0.182
0.209
0.154

0-1
0-1
0-1

0.122
0.119
0.058

0.327
0.324
0.233

0-1
0-1
0-1

0.099
0.124
0.102

0.299
0.330
0.303

0-1
0-1
0-1

0.274
0.129
0.099

0.446
0.335
0.299

0-1
0-1
0-1

0.231
0.173
0.172

0.422
0.379
0.378

0-1
0-1
0-1

0.207
0.182
0.063

0.406
0.386
0.244

0-1
0-1
0-1

0.189
0.222
0.141

0.392
0.416
0.349

0-1
0-1
0-1

Diagnosis
Mean
Air Force (n = 11,264)
Depression
PTSD
Alcohol Abuse
Army (« = 28,426)
Depression
PTSD
Alcohol Abuse
M arines (n = 1,265)
Depression
PTSD
Alcohol Abuse
Navy (396)
Depression
PTSD
Alcohol Abuse

SD

Range

0.042
0.016
0.005

0.200
0.125
0.067

0-1
0-1
0-1

0.058
0.029
0.012

0.234
0.168
0.107

0-1
0-1
0-1

0.066
0.057
0.003

0.248
0.232
0.056

0-1
0-1
0-1

0.088
0.096
0.015

0.284
0.295
0.122

0-1
0-1
0-1

was similar across disorder and branch, with the exception of alcohol issues, which was higher among soldiers, Marines, and seaman. W ith respect to clinical diagnosis at an HCE, depression was diagnosed in 4%—
9% of service members,
PTSD in 2% -10% , and less than 2% of service members in any branch received an alcohol-related diagnosis at an HCE.
Table III presents the results of the three models for depression risk or diagnosis. Model 1 examines the predictors of risk of pre- or early deployment depression. Personnel serving in Air Force had lower odds of depression compared to those in the Army (Odds ratio [OR] = 0.49), whereas
Marines and members of the Navy had higher odds (OR =
3.07 and 2.24, respectively). Serving in Iraq was associated with 33% higher odds of depression and traumatic combat exposure with 29% higher odds. Each additional month deployed was associated with a 2% increase in the odds of depression as well.

422

The results in the PDHRA model (Table III, Model 2,) were largely consistent with the PDHA model; however, unlike the PDHA model, there was no difference in risk of depression on the PDHRA based on location of deployment or time deployed and the effect of traumatic combat exposure was smaller (OR = 1.16). Table III, Model 3 presents results from the model predicting depression diagnosis at an HCE.
Service members in the Air Force had 22% lower odds of a diagnosis compared to those in the Army, but the odds for
Navy and Marine Corp were not significantly different than
Army. There was no significant difference based on location of deployment or traumatic combat exposure, but greater length of time deployed was associated with lower odds of depression diagnosis (OR = 0.99).
Table IV, Model 1 examined the predictors for selfreported PTSD risk in the PDHA among active duty service members. For branch of the military, the odds of PTSD risk

MILITARY MEDICINE, Vol. 180, April 2015

Oveiyiew o f Depression, PTSD, and Alcohol Misuse in Active Duty Service Members
TABLE III.

Predictors of Depression Risk and Depression Diagnosis among Active Duty Service Members Returning From Iraq or Afghanistan (n = 41,351)
Model 1 (PDHA)

Model 2 (PDHRA)

Model 3 (HCE)

OR
Demographics
Branch (Compared to Army)
Air Force
Marines
Navy
Marital Status (Compared to Married)
Single
Divorced/Separated/Widowed
Race/Ethnicity (Compared to White)
Asian
Black
Hispanic
Other
Male
Pay Grade
Health Status
Worse Health
Receiving Care
Deployment Characteristics
Iraq
Traumatic Combat Exposure
Time Deployed
Comorbidities
PTSD
Alcohol Problems
Drug Abuse

Cl

OR

Cl

0.489***
3.066***
2.235***

0.423-0.564
2.622-3.584
1.695-2.948

0.655***
2.482***
1.841***

0.562-0.764
2.078-2.964
1.354-2.504

0.778**
0.894
1.081

0.665-0.911
0.693-1.155
0.734-1.591

1.070
1.234***

0.990-1.156
1.104-1.380

0.976
1.229***

0.893-1.067
1.089-1.388

0.811***
1.585***

0.730-0.901
1.397-1.798

1.059
1.071
0.970
1.166
1.004
0.754***

0.888-1.263
0.972-1.181
0.866-1.086
0.874-1.556
0.889-1.134
0.728-0.782

0.957
1.046
0.952
1.243
0.874*
0.756***

0.778-1.176
0.937-1.166
0.840-1.080
0.910-1.698
0.768-0.995
0.726-0.787

0.618***
0.641***
0.834*
0.917
0.346***
0.785***

0.474-0.806
0.558-0.735
0.718-0.968
0.638-1.319
0.308-0.388
0.752-0.820

2.739***
1.411***

2.548-2.945
1.293-1.538

2.890***
1.610***

2.663-3.137
1.483-1.748

1.456***
2.283***

1.319-1.608
2.077-2.509

1.332***
1 293***
1.020***

1.202-1.477
1.197-1.396
1.010-1.031

0.986
1.158***
1.003

0.887-1.096
1.065-1.259
0.991-1.015

0.934
1.063
0.985*

0.830-1.051
0.967-1.167
0.971-0.998

3.541***
2.070***

3.249-3.859
1.840-2.329

4.875***
2.701***

4.464-5.324
2.447-2.980

6.129***
2.877***
2.021***

5.299-7.089
2.169-3.815
1.402-2.913

OR

CI

*p < 0.05; **p< 0.01 ;***/> < 0.001. PDHA, Post-Deployment Health; PDHRS, Post-Deployment Health Re-Assessment; HCE, Health Care Encounter.

were not significantly different for service members in the
Air Force compared to Army, whereas Marine Corp and
Navy had higher odds. Those who served in Iraq had lower odds of reporting PTSD than did those who served in
Afghanistan and those who were exposed to combat trauma were more than 7 times more likely to report PTSD symp­ toms than those who did not experience combat trauma. The amount of time deployed was also associated with higher odds of PTSD.
Table IV, Model 2 shows the predictors for PTSD based on the PDHRA. The findings for risk of PTSD on the PDHRA were similar to the PDHA PTSD model with respect to serv­ ing in Iraq and time deployed, but were different for branch and traumatic combat exposure. In the PDHRA model,
Marine Corp and Navy had greater risk of PTSD as they did in the PDHA model, but Air Force members had lower risk of screening positive for PTSD compared to members of the
Army (OR = 0.87). Further, traumatic combat exposure was associated with about a 3-fold increase in the odds of PTSD, which was considerably smaller than the association in the
PDHA model.
Table IV, Model 3 examined the predictors of PTSD diagnosis at an HCE. Service members in the Air Force had about the same odds of receiving a diagnosis compared to
Army, whereas Marines had almost 3 times greater odds and seaman had almost 4 times the odds. Serving in Iraq was

MILITARY MEDICINE, Vol. 180. April 2015

associated with 27% lower odds and traumatic combat expo­ sure with over 2.5 times the odds. Unlike the PDHA and
PDHRA models, time deployed did not significantly predict
PTSD diagnosis.
Table V, Model 1 presents the predictors of alcohol prob­ lems on the PDHA among service members. Service members in the Air Force had 76% lower odds of alcohol problems than those in the Army, whereas Marines had almost two times high odds compared to those in the Army. There was no difference in the odds for alcohol problems between those in the Navy and Army. Traumatic combat exposure was associated with
40% higher odds and time deployed with a 3% increase for every additional month deployed.
The second model in Table V presents the predictors of alcohol problems among active duty service members returning from Iraq and Afghanistan between 3 and 6 months postdeployment. The results were largely consistent with the
PDHA model, but the effect of traumatic combat exposure was smaller, such that it was associated with 26% higher odds on this assessment compared to 40% at the previous assessment.
The last model in Table V examined the predictors for an alcohol-related diagnosis from an HCE (Model 3). Those serving in the Air Force had 48% lower odds of an alcohol abuse diagnosis and those in the Marine Corp had 77% lower odds compared to those in the Army, but no differ­ ences were observed between Army and Navy. None of the

423

Overview o f Depression, PTSD, and Alcohol Misuse in Active Duty Service Members
TABLE IV.

Predictors of PTSD Risk and PTSD Among Active Duty Service Members Returning From Iraq or Afghanistan
(n = 41,351)
Model 1 (PDHA)

Model 2 (PDHRA)

Model 3 (HCE)

OR
Demographics
Branch (Compared to Army)
Air Force
Marines
Navy
Marital Status (Compared to Married)
Single
Divorced/Separated/Widowed
Race/Ethnicity (Compared to White)
Asian
Black
Hispanic
Other
Male
Pay Grade
Health Status
Worse Health
Receiving Care
Deployment Characteristics
Iraq
Traumatic Combat Exposure
Time Deployed
Comorbidities
Depression Problems
Alcohol Problems
Drug Abuse

Cl

OR

Cl

OR

Cl

0.976
1.408**
2.038***

0.852-1.118
1.148-1.727
1.501-2.766

0.867*
1.408***
2.156***

0.755-0.995
1.160-1.709
1.613-2.883

1.110
2.738***
3 819***

0.880-1.400
2.050-3.657
2.589-5.635

0.807***
1.080

0.743-0.876
0.962-1.212

0.818***
1.153*

0.753-0.889
1.029-1.291

0.647***
0.929

0.556-0.753
0.765-1.128

0.880
1.007
0.971
0.965
0.732***
0.888***

0.729-1.062
0.908-1.116
0.864-1.092
0.710-1.313
0.647-0.828
0.861-0.916

0.897
0.936
0.982
1.296
0.749***
0.897***

0.741-1.086
0.844-1.038
0.874-1.103
0.970-1.733
0.664-0.845
0.869-0.926

0.773
0.771*
1.062
1.495
1.012
0.843***

0.538-1.110
0.632-0.940
0.873-1.291
0.970-2.304
0.822-1.245
0.795-0.895

2.220***
1.611***

2.060-2.392
1.474-1.760

2.435***
1.883***

2.258-2.627
1.746-2.030

1.496***
3.104***

1.310-1.708
2.718-3.545

0.840***
7 657***
1.027***

0.766-0.921
6.944-8.444
1.016-1.038

0.893*
3.380***
1.031***

0.813-0.982
3.116-3.666
1.020-1.042

0.730***
2.851***
1.017

0.626-0.852
2.467-3.295
0.998-1.036

3.575***
2.314***

3.281-3.895
2.048-2.614

4.939***
2.585***

4.522-5.394
2.349-2.845

6.166***
2.684***
1.961**

5.326-7.139
1.866-3.861
1.225-3.140

*p < 0.05; **p < 0.01; ***p < 0.001. PDHA, Post-Deployment Health; PDHRS, Post-Deployment Health Re-Assessment; HCE, Health Care Encounter.

deployment characteristics were significant predictors of an alcohol diagnosis.
D ISC U SS IO N

Thousands of American troops continue to be deployed to areas of conflict, and many of those deployed return with mental health issues. Previous studies have found deploy­ ment to combat areas to be associated with an increased risk of PTSD, depression, and alcohol abuse, but many previous studies were based on samples that were not representative of the deployed military as a whole. In this study, we attempted to gain a comprehensive overview of these three major mental health problems associated with battle among all four main branches. With postdeployment data on over 40,000 active duty Service men and women, including diagnostic informa­ tion, we were able to estimate prevalence of those who screen positive for risk of each disorder in self-report data at two time points in the deployment cycle, as well as prevalence of diagnoses received during HCEs within the military healthcare system. Further, we were able to examine factors associated with positive screens and diagnoses.
In terms of key findings, the overall prevalence estimates of those who screen positive for risk of depression, PTSD, and alcohol problems are within the range found in previous studies.4'7 Further, the substantial amount of comorbidity, particularly between depression and PTSD fits with previous

424

studies as well.19-21 One contribution of this study, however, is detecting the notable service differences in prevalence of risk and disorder. Our descriptive findings demonstrate siz­ able risk among Navy personnel, similar to the increased risk in PTSD found by Shen et al. While Air Force personnel had lower prevalence of all three mental health issues, Navy personnel had surprisingly high rates of depression risk and depression diagnoses, PTSD risk and PTSD diagnoses, and potential alcohol problems—higher than the Army on all measures and higher than the Marine Corp on PTSD risk and
PTSD diagnosis. Given that the diagnosis rates only capture individuals who sought treatment and are thus likely an undercount of those who could qualify for a diagnosis, the fact that 10% of the Navy subsample received a PTSD diag­ nosis and 9% received a depression diagnosis is striking.
Navy was the smallest subgroup in the sample (n = 396), but these findings suggest that this branch may warrant increased attention to potential mental health issues.
As to why Navy personnel may be at an increased risk for mental health issues, some Navy personnel served in Iraq and Afghanistan with the Army as individual augmentees
(IA), which could partially explain the unexpectedly high rates of disorder. Though 1 study found no difference in mental health problems between soldiers and Navy IA,22 our results suggest that a more careful examination of the mental health of IA is warranted. These sailors receive only 3 weeks

MILITARY MEDICINE, Vol. 180, April 2015

Overview o f Depression, PTSD, and Alcohol Misuse in Active Duty Service Members
TABLE V.

Predictors of Alcohol Misuse and Abuse Among Active Duty Service Members Returning From Iraq or Afghanistan
(n = 41,351)
Model 1 (PDHA)

Model 2 (PDHRA)

Model 3 (HCE)

OR
Demographics
Branch (Compared to Army)
Air Force
Marines
Navy
Marital Status (Compared to Married)
Single
Divorced/Separated/Widowed
Race/Ethnicity (Compared to White)
Asian
Black
Hispanic
Other
Male
Pay Grade
Health Status
Worse Health
Receiving Care
Deployment Characteristics
Iraq
Traumatic Combat Exposure
Time Deployed
Comorbidities
Depression Problems
PTSD
Drug Abuse

CI

OR

Cl

OR

Cl

0.238***
1.848***
1.073

0.187-0.304
1.486-2.297
0.697-1.651

0.345***
1.542***
1.227

0.295-0.404
1.291-1.841
0.896-1.680

0.518***
0.233**
1.272

0.352-0.760
0.085-0.638
0.545-2.969

1 292***
1.117

1.164-1.435
0.951-1.312

1.607***
1.467***

1.481-1.744
1.303-1.653

1.483***
2.010***

1.181-1.862
1.486-2.719

1.274*
1.128
1.146
1.852***
1.492***
0.846***

1.013-1.603
0.985-1.291
0.988-1.331
1.304-2.631
1.224-1.819
0.807-0.887

1.281**
1.249***
1.251***
1.120
1.576***
0.881***

1.070-1.535
1.127-1.385
1.115-1.403
0.812-1.544
1.362-1.823
0.850-0.912

0.522
0.861
1.064
0.725
1.923**
0.594***

0.257-1.060
0.630-1.176
0.776-1.461
0.265-1.983
1.261-2.934
0.514-0.686

1.349***
0.969

1.214-1.499
0.853-1.100

1.597***
0.955

1.468-1.736
0.878-1.038

0.759*
1.299*

0.590-0.977
1.035-1.632

0.997
1.400***
1.026***

0.874-1.137
1.257-1.558
1.012-1.041

0.919
1.261***
1.008

0.832-1.014
1.165-1.365
0.997-1.019

0.787
1.010
1.002

0.610-1.015
0.816-1.250
0.973-1.032

2.076***
2.289***

1.846-2.333
2.030-2.582

2.623***
2.517***

2.378-2.894
2.289-2.769

2.809***
2.749***
9.214***

2.116-3.729
1.924-3.927
6.117-13.88

*p < 0.05; **p