Gender Differences in Posttraumatic Stress Disorder and Help Seeking in the U.S. Army

Abstract

Background:

Inconsistent findings between studies of gender differences in mental health outcomes in military samples have left open questions of differential prevalence in posttraumatic stress disorder (PTSD) among all United States Army soldiers and in differential psychosocial and comorbid risk and protective factor profiles and their association with receipt of treatment.

Methods:

This study assesses the prevalence and risk factors of screening positive for PTSD for men and women based on two large, population-based Army samples obtained as part of the 2005 and 2008 U.S. Department of Defense Surveys of Health Related Behaviors among Active Duty Military Personnel.

Results:

The study showed that overall rates of PTSD, as measured by several cutoffs of the PTSD Checklist, are similar between active duty men and women, with rates increasing in both men and women between the two study time points. Depression and problem alcohol use were strongly associated with a positive PTSD screen in both genders, and combat exposure was significantly associated with a positive PTSD screen in men. Overall, active duty men and women who met criteria for PTSD were equally likely to receive mental health counseling or treatment, though gender differences in treatment receipt varied by age, race, social support (presence of spouse at duty station), history of sexual abuse, illness, depression, alcohol use, and combat exposure.

Conclusions:

The study demonstrates that the prevalence of PTSD as well as the overall utilization of mental health services is similar for active duty men compared with women. However, there are significant gender differences in predictors of positive PTSD screens and receipt of PTSD treatment.

Introduction

The increase in military women participating in combat-related roles and deployment provides an opportunity to better understand gender differences in posttraumatic stress disorder (PTSD), especially in relation to several important correlates such as sexual trauma, injury/illness, and comorbid conditions. Although these factors have been individually associated with gender differences among military personnel,^1

–6 few studies have examined the relationship of these factors in combination with combat exposure in both men and women. For example, recent evidence shows that military sexual trauma interacts with combat exposure to increase risk for PTSD symptoms in female veterans of Iraq and Afghanistan,¹ but it is unknown to what extent these factors may interact in men and which of these confer the largest risks in the military population as a whole.

Prevalence of PTSD

Most studies that have examined gender differences in PTSD among military personnel have found that women were at equal or higher risk than men.^2

–8 Most studies have focused on selected Operation Iraqi Freedom/Operation Enduring Freedom (OIF/OEF) veteran populations surveyed postdeployment, representing only a portion of the military population, or have measured only military-related PTSD or military sexual trauma.^{2,6,9

–15} Bray and colleagues,¹⁶ however, found that 35% of active duty personnel reported some type of physical or sexual abuse before age 18. Premilitary abuse has been shown to be an independent predictor of military-related PTSD,^1,7 increasing the risk of PTSD to those individuals who might not have been deployed as well as to those deployed to combat zones. To assess the total burden of PTSD in the military, a representative sample of all active duty personnel is needed.

Deployment and comorbidity

Results of studies among deployed women veterans have been mixed. No gender differences were found in perceived war zone threat¹² or in PTSD symptoms identified through a postdeployment screening program.⁶ In contrast, Skopp and colleagues¹³ found that overall risk for presumed PTSD after deployment was nearly 2.5 times greater for women than for men. One study found that combat exposure was more strongly associated with postdeployment depression and PTSD symptoms in women than in men¹¹ and there is evidence of a stronger association between injury and PTSD symptoms for women than for men.⁶ To clarify these relationships, the present study examines gender differences in PTSD relative to risk factors (combat exposure, sexual abuse, injury) and comorbid health and mental health conditions (depression and problematic alcohol use).

PTSD and predictors of help seeking

Similarly, inconclusive evidence exists with regard to gender differences in help seeking and health care utilization for PTSD. Several studies suggest that among civilians, women are more likely than men to seek treatment after exposure to a traumatic event;^14,15 women veterans of OIF/OEF with PTSD also had higher use of U.S. Department of Veterans' Affairs (VA) health care than men,¹⁷ and military women had higher rates of hospitalization for mental disorders before the wars.¹⁸ However, at least one study of women veterans with military sexual assault and PTSD suggested that they receive fewer health care services.⁹ Although recent studies have identified an increase in PTSD in service members—presumably resulting from the cumulative effects of over a decade of combat operations—it is unknown if this increase is comparable between men and women. Therefore, the present study also examines a wide range of sociodemographic, health, and psychosocial correlates of potential help seeking including illness, family stress, and social support among men and women with PTSD.

Taken together, the above studies suggest that the relationship between gender and PTSD is complex, and a more comprehensive approach is required to understand correlates of PTSD and help-seeking behaviors in men and women. Small sample sizes of women and the use of different PTSD screening measures have also contributed to the uncertainty of results. Even the most frequently used PTSD screening instrument in the military, the PTSD Checklist (PCL), has been validated primarily in men, and gender differences in possible cutpoints have not been examined.^19,20 The present study will provide population-based estimates of PTSD among soldiers by gender and various PTSD definitions (screening criteria) and assess the stability of prevalence in a population-based representative sample of Army personnel across two survey periods.

The 2005 and 2008 U.S. Department of Defense (DoD) Surveys of Health Related Behaviors among Active Duty Military Personnel (HRB Surveys)^16,21 offer the opportunity to address many of the limitations of the current gender and PTSD literature. These studies use a representative sample of the entire Army that is not restricted to combat veterans, oversamples of women, well-validated mental health measures, and anonymous data-collection methods, yielding optimal samples for better understanding the relationship between gender and PTSD. Specific aims will (1) examine gender differences in PTSD prevalence across 2 study years, using different PCL cutoff criteria to assess differences in screening criteria; (2) examine the impact of risk factors, comorbid health and mental health conditions, and other sociodemographic and psychosocial variables on PTSD classification including potential interactions with combat exposure; and (3) assess the impact of the above variables associated with mental health help seeking and service utilization in men and women soldiers meeting screening criteria for PTSD. Given literature suggesting a higher prevalence of physical/sexual abuse among women,²² the greater role of interpersonal relationships in women's perceived well-being¹³ and the greater potential for threat appraisal among women suggested by the cognitive model of gender differences in trauma processing,²³ it may be hypothesized that sexual abuse and family stress will be stronger predictors of PTSD among women than among men; however, greater prevalence of alcohol abuse among men may be a stronger predictor of PTSD for men than for women. It is expected that these factors will also be associated with gender differences in mental health service utilization.

Methods

Participants

Data were drawn from the 2005 and 2008 HRB Surveys.^16,21 These cross-sectional surveys consisted of a randomly selected representative sample of active duty military personnel from the Army, Navy, Marine Corps, Air Force, and Coast Guard. A two-stage replacement cluster sample proportional to size was employed in which geographic areas were clustered and randomly selected in the first stage, and individuals within the clusters were randomly selected in the second stage. All active duty members were eligible except for recruits, academy cadets, and persons who were absent without leave or incarcerated. Women and officers were oversampled to ensure adequate representation. The final sample consisted of 16,146 personnel in 2005 and 28,546 military personnel from all branches of service in 2008 who completed self-administered questionnaires anonymously. The overall response rate was 51.8% in 2005 and 70.6% in 2008. The response rates were high relative to most military surveys, and most items showed less than a 5% missing rate. Data were weighted to represent all active duty personnel, meaning that the results of the survey represent population estimates of the entire active force. The current study included only respondents from the Army to ensure a more homogeneous population that had the opportunity for combat duty in Iraq or Afghanistan, for a study total of 3,639 in 2005 (2,818 men; 821 women) and 5,927 in 2008 (4,320 men; 1,607 women).

Survey procedures

The majority (97%) of the 32-page anonymous self-report questionnaires were obtained during on-site visits to 64 military installations worldwide by the study team. The remainder was obtained from questionnaires mailed to respondents who were unable to attend group sessions. On average, the questionnaires required about an hour to complete. Institutional Review Board approvals were obtained from RTI International and DoD. Additional sampling and methodological details have been reported and published elsewhere.²⁴

Initial sample weights were constructed using probability of selection at each stage of the study design. Variables used in sampling included service gender, pay grade, and location (continental United States [CONUS] or outside the continental United States [OCONUS]). These weights were adjusted for survey eligibility and nonresponse after data collection was completed. The weights were also poststratified and included age, race, and ethnicity.

Key measures

Questionnaires for the 2005 and 2008 surveys used the same or similar items for all constructs. In addition to demographic items, surveys included questions assessing the mental health, health behaviors, and psychosocial variables discussed below.

PTSD

PTSD symptom severity was assessed using the PCL, Specific Stressor version (PCL-S),¹⁹ a 17-item self-report instrument that asks respondents to rate the extent to which they have been bothered by PTSD symptoms during the previous 30 days using a 5-point scale (1, not at all; 5, extremely). PCL-S items parallel Diagnostic and Statistical Manual of Mental Disorders, fourth edition²⁵ PTSD symptom criteria B, C, and D.^26,27 The HRB Survey and other DoD studies have most often used the PCL-S or PCL-Civilian version over the military version of the PCL to ensure comparability of results at any time in relation to deployment; also, the PCL military version can miss symptoms from nonmilitary traumatic experiences, noncombat-related PTSD occurring during deployment (e.g., sexual assault), or deployment-related exacerbations of PTSD from prior traumas.^24,26
–28 A standard cutoff for population-level studies was used such that if the sum was greater than or equal to 50, the participant was classified as positive for PTSD; those with scores less than 50 were considered not to have PTSD.^19,26,27,29 Alternative lower cutoffs validated for identifying PTSD in various clinical populations (though not validated in women), including 33 and 44, were also reviewed based on those most frequently applied in the literature.^20,29–30

Depression

Depression was assessed using the 10-item short version of the Center for Epidemiologic Studies Depression Scale (CESD-10). Cutoff scores for depressive symptoms were identified as ≥10. Andresen and colleagues³¹ found good predictive accuracy when they compared the CESD-10 with the full-length 20-item version of the CESD.

Problematic alcohol use

This measure used the Alcohol Use Disorders Identification Test (AUDIT),³² which was developed by the World Health Organization as a simple method of screening for excessive drinking and of assisting in brief assessment. The AUDIT consists of 10 questions, each scored 0–4, which are summed to a total score between 0 and 40. A standard cutoff of 8 was used to indicate problematic alcohol use.

Combat exposure

Exposure to combat and related circumstances was measured using a modified version of the 17-item Combat Experiences Scale of the Deployment Risk and Resilience Inventory.^33,34 These items assess exposure to incoming fire, mines, improvised explosive devices, firing on the enemy, viewing dead bodies or human remains, interacting with enemy prisoners of war, and similar circumstances that may be relevant. Each item asked how many times the respondent was exposed, and response options were 0 (0 times), 1 (1 to 3 times), 2 (4 to 12 times), 3 (13 to 50 times), and 4 (51 or more times). For our analyses, items were summed and the sum score was used to create a categorical combat exposure item where a score equal to zero was considered “deployed but no exposure to combat,” a score from 1 to 9 was considered “low/moderate combat exposure,” and a score of 10 or greater was considered “moderate/high combat exposure.” A fourth category was added to capture personnel who had not been deployed. These cutoffs were subsequently examined with factor analysis; item-scoring methods suggested that these categories captured meaningful distinctions between groups of scores.

History of physical and sexual abuse

The 2005 and 2008 surveys included three items from the Brief Trauma Questionnaire³⁵ to assess physical and sexual trauma or abuse. The first item inquired whether respondents were ever physically punished or beaten by a parent, caretaker, or teacher so that they were very frightened, injured, or thought they would be injured. A second item inquired whether they had ever been attacked, beaten, or mugged. A third item inquired whether anyone had ever forced or pressured the respondent into having some type of unwanted sexual contact. Response items inquired whether the trauma happened before age 18, between age 18 and the time they entered the service, or since entering the service. The first two items were combined to form a dichotomous measure of a lifetime history of physical abuse.

Illness in past 12 months

A single item asked respondents how many times they had had an illness in the past 12 months that kept them from duty for a week or longer. Responses were dichotomized to indicate one or more illnesses versus none.

Injury in past 12 months

Three items assessed how often respondents were injured from three potential sources. The first asked about the frequency of injuries or pain that restricted respondents' duty or physical activity for a week or longer. The second asked about injuries sustained during the course of physical training, whether through accident or overuse. The third item assessed the occurrence of injury, by overuse or accident, during activities other than physical training. A dichotomous injury indicator was created, in which any occurrence of any of the three types of injury yielded a positive indication of past 12-month injury.

Family stress

A single item asked “During the past 12 months, how much stress did you experience in your family life or in a relationship with your spouse, live-in fiancé, boyfriend or girlfriend, or the person you date seriously?” Response options ranged from “a lot” (1) to “I had no stress” (5) on a 5-point scale.

Social support

Two separate items were used as proxy social support measures. One came from a coping scale and asked respondents to indicate how often they “talk to a friend or family member” when they feel pressured, stressed, depressed, or anxious, on a 4-point scale from “never” to “frequently.” A second item asked how many times respondents attended religious/spiritual services (not including special occasions such as weddings, christenings, funerals, or other special events) in the past 12 months. These items were adapted from those used in the National Survey on Drug Use and Health.³⁶

Mental health service utilization

Respondents were asked if at any time in the past 12 months they felt they “needed counseling or therapy from a mental health professional (either military or civilian).” Respondents were also asked whether they had received counseling or therapy in the past 12 months for mental health or substance abuse from a variety of sources, including a mental health professional at a military facility, a general medical doctor at a military facility, a military chaplain, a civilian mental health professional, a general medical doctor at a civilian facility, a civilian pastor, rabbi, or other pastoral counselor; or a self-help group. A dichotomous variable was created, indicating whether personnel had received any counseling. Additional items asked for what concerns they sought counseling or therapy, whether they had “been prescribed medication for depression, anxiety, or sleeping problem by a doctor or other health professional,” and whether they had “been prescribed medication to relieve pain or discomfort by a doctor or other health professional.”

Statistical analyses

Analyses were conducted using SAS version 9.1³⁷ and SUDAAN version 9.0³⁸ to account for the complex sampling design of the HRB Surveys. Weighted prevalence rates of demographics, service-related characteristics, and select physical and mental health indicators were estimated for the total Army sample, as well as by gender. Differences in prevalence rates for these constructs were examined with logistic regression models with gender as the predictor. Similar procedures examined the prevalence of PTSD (PCL >50) within each category of the demographic and other variables, as well as differences in PTSD by gender for each construct. To estimate the unique contribution of sociodemographic, comorbid, and psychosocial variables on the likelihood of PTSD, a multiple regression model was estimated with PTSD as the outcome and the full set of variables hypothesized to relate to PTSD entered as predictors. A second set of prevalence estimates and regression analyses were conducted to investigate mental health service need and utilization.

Results

To assess the first aim, Table 1 shows the weighted prevalence of PTSD in the Army for each of the two comparison survey years overall and by gender based on the three most frequently used PCL cutoffs—50, 44, and 33. PTSD prevalence increased in the Army approximately four percentage points from 9.05% in 2005 to 13.07% in 2008 based on the most conservative screening cutoff of 50. The prevalence of PTSD significantly increased in women over the 3-year period for two of the cutoff points and the large increase in men was nearly significant (p = 0.06). Gender differences by screening cutpoints within the two survey years however were not significant and therefore further comparisons by survey year were not made.

Table 1.

Gender Differences in Posttraumatic Stress Disorder Prevalence Among Active Duty Army Personnel by Year and Cutoff Score

	PCL score ≥50				PCL score ≥44+				PCL score ≥33
	2005		2008		2005		2008		2005		2008
Predictor	% (SE)	n	% (SE)	n	% (SE)	n	% (SE)	n	% (SE)	n	% (SE)	n
Male	9.40 (1.27)	226	12.93 (1.41)	492	13.59 (1.41)	311	16.66 (1.52)	631	25.47 (2.37)	611	28.51 (1.53)	1,088
Female	6.96 (1.36)	60	13.94^* (1.77)	218	10.97 (1.38)	94	17.86^* (2.24)	274	26.50 (2.14)	199	32.08^* (3.23)	487
Total	9.05 (1.10)	286	13.07^* (1.31)	710	13.22 (1.26)	405	16.82 (1.44)	905	25.62 (1.89)	810	29.00 (1.51)	1,575

Weighted percent (standard error [SE]) meeting PTSD criteria.

Significant difference between 2005 and 2008. No gender differences were significant.

PCL, Posttraumatic Stress Disorder Checklist; PTSD, posttraumatic stress disorder; SE, standard error.

To assess the second aim, Table 2 presents the distribution of demographic and military characteristics as well as potential risk and protective factors for PTSD. As shown, a higher proportion of women than men were black and unmarried, or married but whose spouses were not present at their duty stations. A lower percentage of women than men had been deployed in their lifetime or exposed to high combat levels, but more had been exposed to moderate levels of combat. Although history of physical abuse was similar among men and women, almost four times more women than men (39.35% vs. 10.06%) reported a history of sexual abuse in 2008. Women were significantly more likely to report illness or injury in the past 12 months, as well as current depression symptoms. They were also more likely to report protective factors including attending religious services and talking about their problems when stressed and reported less problem alcohol use.

Table 2.

Weighted Proportions (and Standard Errors) of Demographic and Risk/Protective Factors by Gender

	All	Males	Females
Characteristic	(n = 5927)	(n = 4320)	(n = 1607)
Age (years)
17–20	14.01 (2.13)	13.58 (2.09)	16.72 (3.9)
21–25	31.61 (3.19)	31.32 (3.08)	33.42 (4.92)
26–34	30.46 (1.36)	31.15 (1.61)	26.05 (1.21)
35 or older	23.92 (3.3)	23.94 (3.01)	23.81 (6.8)
Race/ethnicity
Non-Hispanic white	62.7 (1.97)	65.52 (1.48)	44.67 (5.39)
Non-Hispanic black	19.88 (1.52)	17.69 (0.86)	33.89 (5.22)
Hispanic	10.74 (0.62)	10.48 (0.63)	12.38 (1.18)
Non-Hispanic other	6.68 (0.59)	6.3 (0.58)	9.06 (0.85)
Family status
Not married	43.74 (2.02)	41.64 (2.19)	57.16 (2.94)
Married, spouse not present	11.43 (1.59)	11.31 (1.64)	12.22 (1.44)
Married, spouse present	44.83 (2.41)	47.05 (2.4)	30.63 (3.27)
Rank
Enlisted	82.69 (4.54)	83.06 (4.18)	80.31 (7.2)
Officer	17.31 (4.54)	16.94 (4.18)	19.69 (7.2)
History of physical abuse	39.82 (0.85)	40.3 (0.85)	36.83 (3.28)
History of sexual abuse	14.12 (0.38)	10.06 (0.61)	39.35 (2.97)
Illness in the past 12 months	16.69 (1.21)	16.01 (1.22)	21.06 (1.45)
Injury in the past 12 months	57.96 (1.8)	57.06 (1.82)	63.73 (2.85)
Met criteria for depression	23.74 (1.06)	23.18 (1.18)	27.22 (1.9)
Problem alcohol use (AUDIT 8+)	28.08 (2.2)	30.33 (2.2)	13.7 (1.35)
High family stress	19.5 (1.01)	19.22 (1.05)	21.25 (1.9)
Frequently talk when stressed	38.11 (0.69)	34.94 (0.56)	58.1 (1.63)
≥6 Religious services past year	32.65 (2.3)	31.53 (2.05)	39.61 (3.68)
Not deployed	30.3 (3.34)	27.36 (3.19)	48.3 (4.31)
Deployed, no combat exposure	8.44 (0.77)	8.22 (0.78)	9.78 (0.96)
Deployed, moderate combat exposure	19.78 (2.31)	19.11 (2.36)	23.88 (2.67)
Deployed, high combat exposure	41.48 (3.86)	45.31 (3.87)	18.04 (2.27)

Unweighted sample size.

Italics indicates that estimate for females differs at p < .05 from estimate for males.

AUDIT, Alcohol Use Disorders Identification Test.

Table 3 shows the percentage of men and women meeting PTSD screening criteria by independent risk factors. Overall, the prevalence of PTSD was strongly associated with depression symptoms, combat exposure, alcohol problems, history of physical and sexual abuse, illness in the past 12 months, and high family stress. The youngest age category (age 17–20 years) showed a significantly higher positive screen prevalence in women than in men. There were no gender differences by ethnicity. Fewer married women whose spouses were present at their duty stations met criteria for PTSD than married men whose spouses were with them. Men who reported sexual abuse were more likely than women with sexual abuse to meet criteria for PTSD, although prevalence was high in both men and women. Notably, there were no gender differences in positive screen for PTSD prevalence by deployment or combat exposure.

Table 3.

Multiple Logistic Regression Model of Demographic and Psychosocial Factors Predicting Positive Classification of PTSD by PCL-C Cutoff ≥50 Among Army Personnel

	All		Males		Females
Predictor	OR (CI)	%	OR (CI)	%	OR (CI)	%
17–20 years old	1.97 (1.05–3.71)^*	13.74	1.86 (0.91–3.79)	11.00	2.75 (1.20–6.33)^*	11.43
21–25 years	1.44 (1.07–1.94)^*	16.29	1.49 (1.05–2.13)^*	16.09	1.34 (1.05–1.70)^*	17.29
26–34 years	1.43 (1.03–1.99)^*	13.11	1.43 (1.03–1.98)^*	8.82	1.74 (1.01–3.00)^*	8.15
35 years or older		8.57		42.24		40.99
Non-Hispanic white	0.88 (0.58–1.32)	13.12	0.88 (0.52–1.47)	4.26	1.00 (0.59–1.68)	4.02
Non-Hispanic black		10.36		23.03		30.90
Hispanic	0.90 (0.66–1.23)	13.82	0.86 (0.54–1.36)	8.67	1.12 (0.53–2.37)	11.23
Non-Hispanic other	1.44 (0.81–2.58)	19.20	1.48 (0.69–3.19)	30.47	1.18 (0.83–1.70)	29.58
Not married	0.85 (0.67–1.09)	14.52	0.77 (0.60–0.98)^*	8.79	1.75 (1.16–2.63)^*	9.92
Married, spouse not present	0.68 (0.53–0.88)^*	11.86	0.64 (0.47–0.86)^*	13.23	1.40 (0.74–2.65)	11.35
Married, spouse present		12.00		12.60		17.64
Enlisted	2.32 (1.25–4.30)^*	14.98	2.66 (1.32–5.36)^*	10.81	1.22 (0.68–2.20)	11.31
Officer		4.55		13.80		15.39
History of physical abuse	1.30 (0.90–1.89)	20.09	1.24 (0.81–1.91)	9.68	1.58 (1.13–2.21)^*	14.28
No history of physical abuse		8.47		6.66		10.18
History of sexual abuse	2.55 (1.90–3.42)^*	30.71	3.00 (2.00–4.48)^*	5.79	1.56 (1.17–2.08)^*	9.49
No history of sexual abuse		10.22		19.19		22.71
Illness in the last 12 months	1.67 (1.17–2.39)^*	23.75	1.71 (1.19–2.46)^*	11.00	1.48 (0.76–2.87)	11.43
No illness in last 12 months		11.06		16.09		17.29
Injury in the last 12 months	1.19 (0.87–1.64)	16.27	1.13 (0.82–1.55)	8.82	1.58 (1.01–2.49)^*	8.15
No injury in last 12 months		8.74		42.24		40.99
Depression symptoms	11.06 (8.77–13.95)^*	42.04	11.24 (8.96–14.11)^*	4.26	12.00 (7.22–19.96)^*	4.02
No depression		4.23		23.03		30.90
Problem alcohol use (AUDIT8+)	2.12 (1.63–2.76)^*	23.57	2.10 (1.61–2.73)^*	8.67	2.28 (1.34–3.87)^*	11.23
No problem alcohol use		9.09		30.47		29.58
High family stress	1.72 (1.35–2.20)^*	30.34	1.70 (1.30–2.23)^*	8.79	2.03 (1.27–3.24)^*	9.92
Some–no family stress		8.94		13.23		11.35
Frequently talk when stressed	0.89 (0.67–1.17)	12.84	0.96 (0.73–1.27)	12.60	0.60 (0.46–0.78)^*	17.64
Sometimes–never talk when stressed		13.07		10.81		11.31
≥6 Religious services past year	0.99 (0.73–1.33)	10.90	1.02 (0.75–1.38)	13.80	0.81 (0.54–1.22)	15.39
≤5 Religious services past year		14.00		9.68		14.28
Not deployed	1.22 (0.74–2.01)	10.70	1.34 (0.63–2.87)	6.66	0.89 (0.40–1.96)	10.18
Not exposed to combat		7.21		5.79		9.49
Moderate combat exposure	0.82 (0.35–1.93)	6.42	0.86 (0.29–2.56)	19.19	0.74 (0.29–1.91)	22.71
High combat exposure	2.71 (1.58–4.65)^*	19.40	2.98 (1.37–6.46)^*	11.00	1.53 (0.48–4.82)	11.43

Italics indicate female estimate differs from male estimate at p < 0.05.

Significant predictor of positive PTSD screen.

CI, confidence interval; OR, odds ratio; PCL-C, PTSD Checklist–Civilian.

Table 3 also gives odds ratios and confidence intervals for the demographic and service-related factors predicting a positive classification of PTSD. Depression was strongly associated with PTSD for both men and women. Combat exposure was the second strongest risk factor in which men with high combat exposure had an odds ratio of PTSD that was three times that of those deployed but not exposed to combat, and women had an odds ratio of 1.5. A history of sexual abuse also was a strong risk factor, as was problem alcohol use, high family stress, and illness in the past 12 months for men and injury in the past 12 months for women. Further analysis of potential interactions showed that the combination of high combat exposure and high family stress significantly increased the risk of a positive classification of PTSD (odds ratio 9.59; 95% confidence interval 1.76–52.19) among women only such that in the absence of high combat exposure, family stress did not significantly impact the risk of PTSD.

To assess the third aim, Table 4 examines gender differences in help seeking and service utilization behaviors among Army personnel meeting PTSD criteria. Overall, utilization of mental health services was similar for men and women screening positive for PTSD. Military mental health providers were most commonly utilized, followed by military medical providers and chaplains. Examining comorbidity in help seeking, Table 4 also shows that among those meeting criteria for PTSD, significantly more women sought help for depression and significantly more men sought help for substance use problems, although gender differences in overall need or receipt of mental health services were not significant. Well over half of those meeting criteria for PTSD had been prescribed pain medications in the past 12 months.

Table 4.

Weighted Prevalence of Mental Health Service Utilization Correlates of Meeting Screening Criteria for PTSD Among U.S. Army Personnel by Gender

Predictor	Men	Women	Total
Felt needed mental health counseling, past 12 months	68.51	73.87	69.32
Received any mental health counseling, past 12 months	60.03	63.80	60.57
Received from military mental health provider	43.85	46.14	44.18
Received from military medical provider	30.91	34.48	31.42
Received from military chaplain	26.96	27.33	27.01
Received from civilian mental health provider	23.29	24.54	23.47
Received from civilian medical provider	14.85	13.67	14.69
Received from civilian religious provider	12.71	10.22	12.35
Received from self-help group (e.g., Alcoholics Anonymous)	9.90	9.31	9.82
Sought help for depression	40.43	50.97	42.00
Sought help for anxiety	34.13	42.63	35.40
Sought help for family problems	23.58	28.36	24.29
Sought help for substance use problems	7.68	3.41	7.04
Prescribed mental health medications, past 12 months	32.66	39.36	33.63
Prescribed mental health medications, past month	19.95	23.14	20.41
Prescribed pain medications, past 12 months	57.82	66.48	59.08
Prescribed pain medications, past month	30.99	31.39	31.05

Data presented is weighted prevalence/percentage of those meeting PCL-C ≥50 screening criteria that are positive of type/source of counseling or substance use.

Italics indicate estimate for females differed from males within same year at p < 0.05.

Table 5 shows the multivariate model of predictors for receipt of any mental health service in the past 12 months among men and women who met screening criteria for PTSD. Predictors for receipt of mental health services varied considerably between men and women. Whereas moderate and high combat exposures had the highest odds for receipt of services among men meeting screening criteria for PTSD, combat exposure was not a significant factor in service utilization among women. This was also the case for illness in the past 12 months and history of sexual abuse. In contrast, the highest odds of receipt of mental health services among women meeting criteria for PTSD were being married with a spouse not present at their duty station, depressive symptoms, and problem alcohol use, although these were not significant predictors of service utilization among men.

Table 5.

Multiple Logistic Regression Model of Demographic and Psychosocial Factors Predicting Receipt of Mental Health Services Among Army Personnel Reporting Likely PTSD, 2008 Data

	All		Males		Females
Predictor	OR (CI)	Estimate (SE)	OR (CI)	Estimate (SE)	OR (CI)	Estimate (SE)
17–20 years of age	1.15 (0.47–2.78)	0.6 (0.08)	1.42 (0.49–4.07)	0.63 (0.1)	0.29 (0.10–0.84)^*	0.54 (0.11)
21–25 years of age	1.59 (0.91–2.80)	0.68 (0.04)	1.67 (0.85–3.28)	0.67 (0.05)	0.70 (0.25–1.96)	0.74 (0.05)
26–34 years of age	0.86 (0.54–1.37)	0.53 (0.04)	0.92 (0.48–1.74)	0.52 (0.04)	0.39 (0.20–0.77)^*	0.61 (0.08)
35 years or older		0.57 (0.06)		0.55 (0.08)		0.8 (0.05)
Non-Hispanic white	0.42 (0.19–0.93)^*	0.55 (0.04)	0.34 (0.13–0.89)^*	0.53 (0.04)	0.61 (0.16–2.25)	0.67 (0.05)
Non-Hispanic black		0.74 (0.07)		0.77 (0.08)		0.77 (0.09)
Hispanic	0.61 (0.25–1.46)	0.64 (0.04)	0.56 (0.17–1.85)	0.65 (0.06)	0.41 (0.09–1.81)	0.58 (0.1)
Non-Hispanic other	0.68 (0.21–2.27)	0.66 (0.09)	0.70 (0.15–3.19)	0.7 (0.12)	0.37 (0.12–1.13)	0.55 (0.09)
Not married	0.83 (0.56–1.23)	0.58 (0.03)	0.81 (0.50–1.32)	0.58 (0.04)	0.99 (0.46–2.14)	0.63 (0.05)
Married, spouse not present	0.77 (0.38–1.55)	0.57 (0.07)	0.62 (0.26–1.51)	0.51 (0.1)	4.26 (2.22–8.18)^*	0.88 (0.04)
Married, spouse present		0.63 (0.05)		0.63 (0.05)		0.63 (0.06)
Enlisted	1.01 (0.41–2.49)	0.6 (0.03)	1.10 (0.37–3.25)	0.6 (0.04)	0.64 (0.08–4.88)	0.66 (0.03)
Officer		0.6 (0.1)		0.57 (0.13)		0.75 (0.18)
History of physical abuse	1.12 (0.78–1.62)	0.61 (0.03)	1.05 (0.71–1.54)	0.6 (0.04)	1.49 (0.62–3.61)	0.7 (0.05)
No history of physical abuse		0.58 (0.04)		0.59 (0.05)		0.61 (0.06)
History of sexual abuse	1.90 (1.28–2.83)^*	0.7 (0.04)	2.08 (1.37–3.17)^*	0.72 (0.05)	1.54 (0.74–3.23)	0.7 (0.05)
No history of sexual abuse		0.55 (0.03)		0.55 (0.04)		0.61 (0.05)
Illness in the last 12 months	1.89 (1.16–3.06)^*	0.7 (0.04)	2.06 (1.19–3.56)^*	0.71 (0.05)	1.66 (0.81–3.43)	0.74 (0.06)
No illness in last 12 months		0.56 (0.04)		0.55 (0.04)		0.63 (0.04)
Injury in the last 12 months	1.34 (1.03–1.74)^*	0.62 (0.03)	1.30 (0.93–1.81)	0.61 (0.04)	1.57 (0.53–4.69)	0.69 (0.03)
No injury in last 12 months		0.55 (0.04)		0.55 (0.04)		0.59 (0.12)
Depression symptoms	1.63 (0.70–3.76)	0.63 (0.04)	1.56 (0.66–3.66)	0.62 (0.05)	2.77 (1.00–7.61)^*	0.71 (0.04)
No depression		0.51 (0.07)		0.51 (0.07)		0.47 (0.1)
Problem alcohol use (AUDIT8+)	0.88 (0.52–1.48)	0.58 (0.05)	0.73 (0.42–1.28)	0.56 (0.06)	2.55 (1.39–4.66)^*	0.79 (0.03)
No problem alcohol use		0.62 (0.04)		0.63 (0.04)		0.6 (0.05)
High family stress	1.10 (0.70–1.73)	0.61 (0.04)	1.08 (0.64–1.82)	0.61 (0.04)	1.40 (0.51–3.87)	0.71 (0.05)
Some-no family stress		0.59 (0.05)		0.59 (0.05)		0.63 (0.07)
Frequently talk when stressed	1.48 (0.93–2.35)	0.66 (0.04)	1.72 (0.98–3.02)	0.68 (0.04)	0.85 (0.43–1.67)	0.65 (0.04)
Sometimes-never talk when stressed		0.57 (0.04)		0.55 (0.05)		0.68 (0.05)
≥6 Religious services past year	1.13 (0.73–1.75)	0.62 (0.04)	1.19 (0.69–2.04)	0.63 (0.05)	0.81 (0.48–1.36)	0.64 (0.06)
≤5 Religious services past year		0.59 (0.04)		0.58 (0.04)		0.68 (0.03)
Not deployed	1.99 (0.94–4.18)	0.53 (0.05)	2.38 (0.87–6.53)	0.52 (0.05)	1.59 (0.51–4.95)	0.65 (0.07)
Not exposed to combat		0.36 (0.1)		0.31 (0.12)		0.53 (0.1)
Moderate combat exposure	3.96 (1.63–9.64)^*	0.69 (0.09)	5.64 (1.71– 18.60)^*	0.72 (0.12)	1.61 (0.43–6.12)	0.65 (0.11)
High combat exposure	3.10 (1.37–7.01)^*	0.63 (0.03)	3.63 (1.14– 11.54)^*	0.62 (0.03)	2.56 (0.94–6.91)	0.75 (0.07)

Female estimate differs from male estimate at p < 0.05.

Discussion and Conclusions

This study examined the prevalence of positive PTSD screens for men and women who participated in the 2005 and 2008 HRB surveys and evaluated associations between gender and other key potential risk factors with PTSD among military personnel. Our findings are based on the largest and most representative samples of active duty soldiers that had a high response rate and provided the ability to look for consistency across years.

Notably, there were no significant gender differences in the prevalence of positive screens for PTSD or in symptom severity, with rates increasing in both men and women between 2005 and 2008. A higher proportion of women than men reported lifetime history of physical and sexual abuse as well as past year illness and injury, potential risks which may have been offset by their greater social support (talking when stressed, attendance at religious services) compared to men. Although active duty men who reported sexual abuse were more likely than women to meet criteria for PTSD, rates were high for both genders.

Regression analyses showed that although depression symptoms were the strongest predictors of positive PTSD screens in both genders, men and women soldiers varied in their odds of PTSD by several predictors. Specifically, the highest odds of PTSD among men were history of sexual abuse, combat exposure, problem alcohol use, and enlisted rank; the highest odds of PTSD among women were problem alcohol use, age 17–20 years, and high family stress, though history of sexual abuse, physical abuse, and injury in the past 12 months were also strongly predictive as well. The fact that combat exposure was not significantly associated with PTSD in women (though the trend was in the predicted direction) may have been related to insufficient power to detect a difference, given the smaller sample size of women than men and small percentage of the total population reporting high combat exposure. Furthermore, it is possible that the inclusion of depression in the regression model, a highly comorbid outcome that also can result from combat, may have dampened other associations.

Our hypothesis that interpersonal relations had a stronger moderating effect on PTSD in women than in men was supported, in that women who reported sometimes or never talking with friends and family (as opposed to frequently talking) were more likely than men to meet criteria for PTSD. Also, family stress was a strong moderator in the relationship between high combat exposure and PTSD among women only. This suggests that the negative effects of limited social support may be greater for women than men and is consistent with the finding of Skopp et al.¹³ that intimate relationship issues were positively associated with presumed PTSD at higher levels of combat only in women.

Although alcohol abuse did not interact with combat in predicting positive PTSD screens in men, it was a significant predictor of PTSD in both men and women, and also predicted receipt of mental health services among women. History of sexual abuse strongly increased the odds of a positive classification of PTSD for both men and women, and also increased the odds of mental health services receipt among men. Also, illness but not injury in the past 12 months was both a significant predictor of a positive screen for PTSD and of receipt of mental health services among men but not women. This extends the findings of Sripada et al.,³⁹ who found higher levels of past-year medical illness associated with increased odds of receiving past-year treatment for PTSD among a population-based survey of civilians though they did not evaluate gender differences.

This study also found that contrary to civilian studies,^14,15 active duty women who met criteria for PTSD were not more likely to receive mental health counseling or treatment than their male counterparts. However, if there was comorbid depression or problem alcohol use with PTSD, women were more likely to receive services than were men. The reasons for the lack of differences between men and women in utilizing health services for PTSD may have to do with similarities across the military, regardless of gender, in perceived stigma or access to care. The military population has universal access to health care, and there is routine screening for PTSD and other mental health problems throughout a soldier's career. The strong association between receipt of PTSD treatment and the spouse's absence at the duty station among married female personnel suggests that providing friends and family members with information about PTSD symptom detection and referral services may help increase service utilization when needed, especially among women. In addition, the associations between receipt of treatment and combat exposure among men but not women and between depression symptoms and problematic alcohol use among women but not men suggests that personnel with PTSD are receiving mental health treatment for different reasons. Why white/non-Hispanic soldiers had significantly lower utilization of mental health services is unclear, and findings are inconsistent with one study from the VA that showed that African American veterans attending a VA facility were less likely to receive treatment after receiving a PTSD diagnosis.⁴⁰ These associations deserve further study.

The most important limitations of this study include the self-report and cross-sectional nature of the surveys. However, this study represents one of the most rigorous examinations of gender differences in PTSD among military personnel to date. The study also involved a sophisticated sampling approach and oversampled women. The study suggests that despite similar prevalence rates, there are significant gender differences in both predictors of positive PTSD screens and receipt of PTSD treatment, especially with regard to combat exposure. As combat roles for women continue to open, it will be important to monitor their mental and physical health across time, including their health care utilization patterns.

Footnotes

Acknowledgments

This research was supported by contract number W81XWH-10-F-0444 from the U.S. Army. This article represents the views of the authors and is not an official position of the U.S. Army. The authors thank Mr. Justin Faerber and Anne Gering for their editorial assistance.

Author Disclosure Statement

No competing financial interests exist.

References

Cobb Scott

, Pietrzak

, Southwick

, et al. Military sexual trauma interacts with combat exposure to increase risk for posttraumatic stress symptomatology in female Iraq and Afghanistan veterans. J Clin Psychiatry, 2014; 75:637–643.

Street

, Vogt

, Dutra

. A new generation of women veterans: Stressors faced by women deployed to Iraq and Afghanistan. Clin Psychol Rev, 2009; 29:685–694.

Crum-Cianflone

, Jacobson

. Gender differences of postdeployment post-traumatic stress disorder among service members and veterans of the Iraq and Afghanistan conflicts. Epidemiol Rev, 2014; 36:5–18.

Vogt

Research on gender differences in deployment stress and postdeployment mental health: Addressing conceptual and methodological challenges and expanding our research agenda. J Womens Health (Larchmt), 2014; 23:485–487.

Kline

, Ciccone

, Weiner

, et al. Gender differences in the risk and protective factors associated with PTSD: A prospective study of National Guard troops deployed to Iraq. Psychiatry, 2013; 76:256–272.

Maguen

, Luxton

, Skopp

, Madden

. Gender differences in traumatic experiences and mental health in active duty soldiers redeployed from Iraq and Afghanistan. J Psychiatr Res, 2012; 46:311–316.

Hourani

, Williams

, Bray

, Kandel

. Gender differences in the expression of PTSD symptoms among active duty military personnel. J Anxiety Disord, 2015; 29:101–108.

Hourani

, Williams

, Bray

, Kandel

. Posttraumatic stress disorder, substance abuse, and other behavioral health indicators among active duty military men and women. J Trauma Stress Disord Treat. 2014; 3. https://dx-doi-org.web.bisu.edu.cn/10.4172/2324-8947.1000124.

Suris

, Lind

, Kashner

, Borman

, Petty

. Sexual assault in women veterans: An examination of PTSD risk, health care utilization, and cost of care. Psychosom Med, 2004; 66:749–756.

10.

Himmelfarb

, Yaeger

, Mintz

. Posttraumatic stress disorder in female veterans with military and civilian sexual trauma. J Trauma Stress, 2006; 19:837–846.

11.

Luxton

, Skopp

, Maguen

. Gender differences in depression and PTSD symptoms following combat exposure. Depress Anxiety, 2010; 27:1027–1033.

12.

Vogt

, Vaughn

, Glickman

, et al. Gender differences in combat-related stressors and their association with postdeployment mental health in a nationally representative sample of U.S. OEF/OIF veterans. J Abnorm Psychol, 2011; 120:797–806.

13.

Skopp

, Reger

, Mishkind

, Raskind

, Gahm

. The role of intimate relationships, appraisals of military service, and gender on the development of posttraumatic stress symptoms following Iraq deployment. J Trauma Stress, 2011; 24:277–286.

14.

Foa

, Keane

, Friedman

. Effective treatments for PTSD: Practice guideline from the international Society for Traumatic Stress Studies. New York, NY: Guilford Press, 2000.

15.

Tarrier

, Pilgrim

, Sommerfield

, et al. A randomized trial of cognitive therapy and imaginal exposure in the treatment of chronic posttraumatic stress disorder. J Consult Clin Psychol, 1999; 67:13–18.

16.

Bray

, Pemberton

, Hourani

, et al. 2008. Department of Defense Survey of Health Related Behaviors among Active Duty Military Personnel. Research Triangle Park, NC: RTI International, 2009.

17.

Maguen

, Cohen

, Madden

, Bertenthal

, Seal

. Gender differences in health service utilization among Iraq and Afghanistan veterans with posttraumatic stress disorder. J Womens Health (Larchmt), 2012; 21:666–673.

18.

Hoge

, Lesikar

, Guevara

, et al. Mental disorders among U.S. military personnel in the 1990s: Association with high levels of health care utilization and early military attrition. Am J Psychiatry, 2002; 159:1576–1583.

19.

Weathers

, Litz

, Herman

, Huska

, Keane

. The PTSD Checklist (PCL): Reliability, validity, and diagnostic utility. Paper presentation at Ninth Annual Conference of the ISTSS. San Antonio, TX, 1993.

20.

Bliese

, Wright

, Adler

, Cabrera

, Castro

, Hoge

. Validating the primary care posttraumatic stress disorder screen and the posttraumatic stress disorder checklist with soldiers returning from combat. J Consult Clin Psychol, 2008; 76:272–281.

21.

Bray

, Hourani

, Rae Olmstead

, et al. 2005. Department of Defense Survey of Health Related Behaviors among Active Duty Military Personnel. Research Triangle Park, NC: RTI International, 2006.

22.

Norris

, Foster

, Weisshaar

. The epidemiology of sex differences in PTSD across developmental, societal, and research contexts. In: Kimerling

, Ouimette

, Wolfe

, eds. Gender and PTSD. New York, NY: Guilford Press, 2002:3–42.

23.

Tolin

, Foa

. Gender and PTSD: A cognitive model. In: Kimerling

, Ouimette

, Wolfe

, eds. Gender and PTSD. New York, NY: Guilford Press, 2002:76–97.

24.

Bray

, Pemberton

, Lane

, Hourani

, Mattiko

, Babeu

. Substance use and mental health trends among U.S. military active duty personnel: Key findings from the 2008 DoD Health Behavior Survey. Mil Med, 2010; 175:390–399.

25.

American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. (DSM-IV). Washington, DC: American Psychiatric Association, 1994.

26.

Hoge

, Castro

, Messer

, McGurk

, Cotting

, Koffman

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

27.

Thomas

, Wilk

, Riviere

, McGurk

, Castro

, Hoge

. Prevalence of mental health problems and functional impairment among active component and National Guard soldiers 3 and 12 months following combat in Iraq. Arch Gen Psychiatry, 2010; 67:614–623.

28.

Kok

, Herrell

, Thomas

, Hoge

. Posttraumatic stress disorder associated with combat service in Iraq or Afghanistan: Reconciling prevalence differences between studies. J Nerv Ment Dis, 2012; 200:444–450.

29.

Terhakopian

, Sinaii

, Engel

, Schnurr

, Hoge

. Estimating population prevalence of posttraumatic stress disorder: An example using the PTSD checklist. J Trauma Stress, 2008; 21:290–300.

30.

Yeager

, Magruder

, Knapp

, Nicholas

, Frueh

. Performance characteristics of the posttraumatic stress disorder checklist and SPAN in Veterans Affairs primary care settings. Gen Hosp Psychiatry, 2007; 29:294–301.

31.

Andresen

, Malmgren

, Carter

, Patrick

. Screening for depression in well older adults: Evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). Am J Prev Med, 1994; 10:77–84.

32.

Babor

, Higgins-Biddle

, Saunders

, Monteiro

. The Alcohol Use Disorders Identification Test: Guidelines for use in primary care, 2nd ed. Geneva, Switzerland: World Health Organization, Department of Mental Health and Substance Dependence, 2001.

33.

U.S. Department of Veterans Affairs. PTSD: National Center for PTSD. Available at: www.ptsd.va.gov/professional/assessment/deployment/combat-experiences-scale.asp Accessed September 16, 2015 .

34.

Vogt

, Smith

, King

, Knight

, Vasterling

. Deployment risk and resilience inventory-2 (DRRI-2): An updated tool for assessing psychosocial risk and resilience factors among service members and veterans. J Trauma Stress, 2013; 26:710–717.

35.

Schnurr

, Vielhauer

, Weather

, Findler

The brief trauma questionnaire. White River Junction, VT: National Center for PTSD, 1999.

36.

Center for Behavioral Health Statistics and Quality. Results from the 2010 National Survey on Drug Use and Health: National findings. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2011.

37.

SAS Institute, Inc.. SAS, Version 9.1. Cary, NC: SAS Institute, Inc., 2003.

38.

RTI International.. SUDAAN, Version 9.0. Research Triangle Park, NC: RTI International, Inc., 2007.

39.

Sripada

, Pfeiffer

, Valenstein

, Bohnert

. Medical illness burden is associated with greater PTSD service utilization in a nationally representative survey. Gen Hosp Psychiatry, 2014; 36:589–593.

40.

Spoont

, Hodges

, Murdoch

, Nugent

. Race and ethnicity as factors in mental health service use among veterans with PTSD. J Trauma Stress, 2009; 22:648–653.