Factors Affecting Help Seeking for Mental Health Problems After Deployment to Iraq and Afghanistan

The sample was drawn from phase 2 of a cohort study conducted by the King’s Centre for Military Health Research (6,13). Phase 1 consisted of a randomly selected sample of personnel serving in the first phase of the Iraq war (January–June 2003). The sample included two groups: individuals deployed in the first phase of the Iraq war and individuals who served at that time but were not deployed to Iraq. Data were collected between June 2004 and March 2006 through a self-report questionnaire. Reservists and personnel who had left the military were included. Questionnaires were completed by 10,272 of the 17,499 phase 1 personnel contacted, for a response rate of 59% (13).

Phase 2, from which the study sample was taken, consisted of 9,395 individuals who participated in phase 1 and who consented to follow-up. The sample also included 1,825 regular military personnel and reservists deployed to Afghanistan between April 2006 and April 2007 and a replenishment group of 6,628 regular personnel and reservists, who joined and trained between April 2003 and April 2007 and who were added to ensure that the age and rank distribution remained representative of deployed members of the United Kingdom armed forces. A total of 9,990 regular personnel and reservists completed questionnaires between November 2007 and September 2009, for a response rate of 56% (6). Completion of the questionnaire at phase 2 was not associated with mental health status (6).

A total of 4,986 participants had been deployed to Iraq, Afghanistan, or both, and 4,725 (95%) responded to an item in the questionnaire asking about general medical problems, stress or emotional problems, and alcohol problems resulting from their last deployment to Iraq or Afghanistan. A total of 2,561 individuals were deployed most recently to Iraq, and 2,164 were deployed most recently to Afghanistan. For the 970 who had served in both Iraq and Afghanistan and who provided information on their health after both deployments, only data from the most recent deployment were used. The final sample comprised 4,246 regular military personnel and 479 reservists. Full details of the study were given to the participants, who then provided consent for future contact and access to medical and resettlement records. The study was reviewed and received ethical approval from the King’s College Hospital (London) research ethics committee and from the Ministry of Defence’s research ethics committee.

Participants who reported experiencing general medical problems, stress or emotional problems, or alcohol problems as a result of their last deployment were asked whether they had sought professional help from a general medical practitioner or medical officer, who provide both general medical and mental health care, along with referrals to specialist health providers, such as a psychologist or psychiatrist; a specialist, such as a hospital doctor, psychiatrist, community psychiatric nurse, or counselor; or a nonmedical professional, such as a military chaplain, welfare officer, or social worker (in the United Kingdom, a position that mainly involves helping individuals solve health, housing, or social welfare issues). Participants could report more than one problem and more than one source of help for each problem.

Help seeking comprised three categories. Medical help included seeking help from a general medical practitioner, medical officer, or specialist, even if additional help was sought from nonmedical professionals. Nonmedical help included seeking help only from nonmedical professionals. A third group comprised those who did not seek any help.

Both medical and nonmedical help seeking for alcohol problems were combined into one variable, given that the numbers reporting each type of help were small. Consequently, the variable for help seeking for alcohol problems consisted of two categories: seeking any help for alcohol problems—including help from a general practitioner, medical officer, specialist, and nonmedical professional—and not seeking help.

Mental health status in the few weeks prior to completing the questionnaire was also measured. Posttraumatic stress disorder (PTSD) was assessed by using the 17-item PTSD Checklist–Civilian Version (PCL-C), with a score of 50 or more indicating probable PTSD (14). The General Health Questionnaire−12 (GHQ-12) was used to identify a case of psychological distress (score of ≥4) (15). A 53-item Multiple Physical Symptom (MPS) scale, including symptoms of the Hopkins Symptom Checklist (16) and symptoms reported in studies of Gulf War veterans, was used to assess somatic symptoms that often co-occur with psychological problems but that are infrequently associated with physical problems unrelated to mental illness; having 18 or more symptoms indicates a case of somatic symptoms (13). To investigate whether help seeking was affected by number of mental health problems, we constructed a composite measure by combining the PCL-C, the GHQ-12, and the MPS. The measure identified persons who met case criteria for no, one, or two or more mental health problems (PTSD, psychological distress, or multiple physical symptoms).

The World Health Organization Alcohol Use Disorders Identification Test (AUDIT) was used to assess alcohol problems (17). The AUDIT score is split into three categories: scores below 16 indicate no serious alcohol misuse, scores of 16–19 indicate cases of serious alcohol misuse, and scores of 20 or more indicate cases of serious alcohol misuse that are likely to include functional impairment (18).

The impact of ill health on functioning was assessed by a self-rating of the extent to which general medical or emotional problems interfered with normal social activities, which could range from not at all to extremely. A binary variable was constructed, with responses of “quite a bit” or “extremely” indicating functional impairment and responses of “moderately,” “slightly,” or “not at all” indicating no functional impairment.

Multinomial regression was used to investigate factors (demographic, military, social functioning, and self-reported current health) associated with help seeking for stress or emotional problems resulting from deployment. No help seeking, nonmedical help seeking, and medical help seeking were the dependent variables. Multinomial odds ratios (MORs) with 95% confidence intervals (CIs) are reported. Adjusted analyses were conducted by using variables identified by the unadjusted analysis and in the literature as possible confounders (age, sex, marital status, type of engagement, and rank).

Logistic regression was used to investigate factors (demographic, military, social functioning, and self-reported current health) associated with any help seeking (medical or nonmedical) for alcohol problems. The analyses were adjusted for demographic variables (age, sex, marital status, engagement type, and rank). Odds ratios (ORs) with CIs are reported for both analyses.

Given that previous research within the military indicates that there is less recognition of alcohol problems than of other mental health problems (10) and that help seeking for these problems may be affected by severity of the problem, we report, by AUDIT score (<16, 16–19, and ≥20), the proportion of those seeking help for alcohol problems.

Weights were created to account for sample fractions, given that the analyses combined three independent randomly selected samples, and for the lower response rate at phase 2 among younger service personnel, members of lower ranks, and males (6). All analyses were conducted by using the statistical software package Stata, version 10. Unweighted frequencies, weighted percentages, MORs, and ORs are reported.

A majority of respondents were male, married or in a long-term relationship, and regular military personnel. The weighted mean±SD age was 32.7±7.6. For 61%, the last deployment was to Iraq, and for 39%, the last deployment was to Afghanistan (Table 1). Altogether, 25% reported any health problems; 19%, stress or an emotional problem; 8%, a general medical problem, and 6%, an alcohol problem (Table 1). A total of 152 (4%) reported a general medical problem and stress or an emotional problem. In addition, 47% (N=827) of those with a general medical problem reported stress or an emotional problem, but only 18% (N=309) of the group with stress or an emotional problem reported a general medical problem.

Of the 888 service members reporting stress or an emotional problem, 42% reported seeking help, with medical help (29%) reported most commonly (Table 2). Thirty-one percent of those with alcohol problems reported seeking help, with medical help reported most commonly (17%). In contrast, 85% of those with a general medical problem reported seeking medical help.

Adjusted analyses indicated that medical help seeking for stress or emotional problems was associated with being female (MOR=2.73), having current functional impairment (MOR=3.22), meeting criteria for two or more mental health problems (MOR=2.68), and an AUDIT score of 20 or more (MOR=1.79). The odds of seeking medical help were reduced for individuals of officer rank (MOR=.40) (Table 3).

Nonmedical help seeking for stress or emotional problems was positively associated with being divorced or separated (MOR=2.25) and negatively associated with last deploying to Afghanistan (MOR=.53).

There was an increasing trend in the association between seeking any help for alcohol problems and number of current mental health problems (Table 4). AUDIT score was not associated with help seeking for alcohol problems, and no other significant associations were found for help seeking for perceived alcohol problems (data available on request).

Forty-one percent of respondents scoring 20 or more on the AUDIT reported having an alcohol problem after deployment, compared with only 14% of those with an AUDIT score of 16–19 and 3% of those with a score below 16 (data not shown), indicating that most participants who recognized alcohol use as a problem had a very high AUDIT score.

Our study demonstrates the reluctance to seek medical help for stress or emotional problems or alcohol problems compared with the willingness to seek help for a general medical problem and confirms the findings of other reports (2,10,19–22). Given that health care is free in the United Kingdom, the ability to pay is not a barrier to treatment among active-duty personnel and veterans, but other barriers must be considered. Previous research has found that military populations have concerns about taking leave from work for treatment (3,23). However, only 15% of those reporting a general medical problem did not seek medical help, indicating that taking time off to attend a medical consultation is not an overriding concern. It may be the case that personnel view general medical problems as having a more direct impact on their service preparedness and fitness, but research also suggests that they are concerned that seeking help for mental health problems could damage military careers (24) and cause others in the military to lose confidence in them (3). In addition, the low percentage seeking professional help for mental health problems may reflect a willingness to deal with the problems personally (25) and an unwillingness to admit help seeking.

Reluctance to seek help for alcohol problems was compounded by low recognition of having a problem. Only 6% of participants reported an alcohol problem after deployment, whereas a report based on validated measures indicated that 13% of the military in the United Kingdom engage in alcohol misuse (6). In addition, no association was found between serious alcohol misuse and help seeking for alcohol problems. Instead, our study suggests that alcohol misuse was recognized as a problem if it was accompanied by other mental disorders or if the participant had an AUDIT score of 20 or more, a score associated with functional impairment (18). This is consistent with previous findings (10).

The military mirrors the general population in that women are more likely to seek medical help (11,26–28), a pattern that is often attributed to social norms related to masculine identity (29). Within this sample, the military population differed from the general population in that age was not found to be associated with help seeking, possibly because of the relatively young age of active-duty personnel. Although the age distribution of our sample was comparable with that of the United Kingdom armed forces, the mean age of the sample was higher than the mean age in the U.S. military population, which is approximately 29 years (30).

Participants who were no longer in a long-term relationship were more likely to seek nonmedical help for stress or an emotional problem than participants who were single or in a relationship. Persons who are in or have previously been in a long-term relationship access mental health services more often (26), and encouragement from friends or family may help to overcome barriers to care (31).

As indicated by previous research (2), we found that participants returning from Afghanistan reported seeking help less than those who deployed to Iraq, but we did not find an association between combat role and help-seeking behavior, as has been reported (22). There was also no association between engagement type and help seeking.

Commissioned officers were less likely than other ranks to seek medical help for stress or an emotional problem. Personnel of higher rank are less likely to seek help for mental health problems (26,32), and the rate of mental health problems is lower among commissioned officers (33). However, our study went a step further by demonstrating that commissioned officers were less likely to seek medical help for stress or an emotional problem. This may be related to concerns about stigma and damage to their military career (34,35).

It is not surprising that meeting criteria for two or more mental health problems was associated with medical help seeking, given that this is consistent with previous research (11,16,26,32). Our study suggests that experiencing more than one mental health problem was a trigger to seeking medical help for both those who endorsed stress or an emotional problem and those with an alcohol problem. Our study adds to the literature by demonstrating that although meeting criteria for one mental health problem was not associated with seeking help, participants who met criteria for two or more mental health problems were more inclined to seek medical help.

Our study is unique in demonstrating that functional impairment is associated with seeking medical rather than nonmedical help. Indeed, it was the factor most strongly associated with seeking medical help for stress or emotional problems, indicating that personnel commonly do not seek help until problems are severe. There are opposing views about the desirability of help seeking for mild psychological morbidity (27,36,37), and consideration should be given to the importance of treating less severe mental health problems within a military population.

This was a large study based on a representative sample of the armed forces of the United Kingdom and included a large number of covariates to assess factors affecting the use of health and welfare facilities. One limitation of this study was that the use of self-report may have led to underreporting or overreporting of help seeking. The findings for prevalence of health problems may have been affected by the study’s focus on problems resulting from deployment, given that some participants may not have acknowledged problems that they did not perceive to be deployment related. We were able to assess the association between reporting stress or emotional problems and alcohol problems after deployment and mental health at the time of questionnaire completion. Although nonparticipation was a possible source of potential bias, nonresponse in phase 2 among those who participated in phase 1 was unrelated to mental health status (6). Given that this study was based on a cross-sectional analysis, care should be taken to avoid interpreting associations as causal relationships.

The most striking finding is that a high percentage of military personnel reported stress or an emotional problem yet did not seek help. This was the case despite implementation by the armed forces of the United Kingdom of psychoeducational interventions that address deployment stress, depression, PTSD, alcohol use, relationship problems, and ways to cope with postdeployment problems (38) and the use of a trauma risk management tool (39). Given the association between help seeking and symptom severity, a proportion of those who self-reported stress or an emotional problem may have believed that the issue would be of short duration and that professional help would not be required.

Commissioned officers may need to be targeted in relation to stress or emotional problems. Our study showed that officers who recognize stress or an emotional problem were less likely than other ranks to seek help. It is possible that commissioned officers do not take part in programs for coping with stress that are provided by the armed forces and that they may feel isolated because they are unwilling to share personal problems with their peer group.

Alcohol misuse is common in the United Kingdom armed forces but is not usually recognized as a problem, and few of those who recognize their problem seek help. Although stress or emotional problems were the most common health problem after deployment, help seeking for these problems was low. In addition, medical help seeking was less likely among those who met criteria for only one mental health problem and those who were less likely to have functional impairment, indicating lower severity of stress or emotional problems. It is unclear whether refraining from help seeking for less severe disorders was a response to the belief that the problem would be short lived or a balancing act between seeking help and concerns about the negative effects of help seeking on service members’ reputation and career prospects.