To the Editor: As a former military psychiatrist and military psychiatric researcher with over 12 years of service, I read with interest the article by Dr. Hoge and colleagues. Their main conclusions 1) that mental disorders are common in military personnel as discharge diagnoses and 2) that far more attrition from military service occurs after mental disorder diagnoses than from physical disorders is far from "striking." What is surprising is the apparent lack of appreciation by the epidemiologist authors of the military medical regulations that specifically discriminate against service members with mental disorders to a far greater degree than those with primarily physical ICD-9 diagnoses. One could argue the merits of this discriminatory approach for some personnel in sensitive positions (a point not made in the article), but the fact remains that such regulations would readily account for this differential attrition rate. The authors cited the widespread use of antidepressants by military personnel as an example of why they appeared puzzled by their attrition findings; of greater concern is their false statement that the use of psychotropic medications is "not grounds for separation" (p. 1582). I respectfully beg to differ.

On many occasions I advocated for my patients on active duty who desperately wanted to remain in the military after recovery from an episode of mental illness. Patients taking medications who were in full remission from cyclothymia, bipolar disorder, major depressive disorder, obsessive-compulsive disorder, and other conditions that did not include psychotic symptoms were routinely discharged involuntarily in spite of expert military psychiatric opinion recommending retention. I would not deem this to be attrition; it was involuntary separation and, as such, should not be included in the attrition statistics. If an active duty member with an illness is not deemed "worldwide qualified," he or she is medically discharged, irrespective of his or her wishes and in spite of attaining full remission. This is true even if the service member’s job is clerical, secretarial, medical technical, or one of the majority of military careers that do not involve the use of—or access to—firearms, explosives, aircraft, nuclear weapons, or highly classified information. Another clear example of enforced involuntary separation—not attrition—is that any patient treated with lithium is not considered worldwide qualified, irrespective of diagnosis or clinical condition.

Given the serious institutional discrimination against service members with mental disorders, "equal access to ‘free’ medical care" (p. 1581) should not imply that those with a need or desire for treatment will necessarily receive this care. In fact, it is widely known by service members that reporting for mental health care may be a death sentence for a promising military career. The figures reported by Dr. Hoge et al. for mental disorder diagnoses (11.9% for any diagnostic position) should therefore be viewed as, at best, a minimum rate. Given the protections afforded most civilian employees who develop mental illness compared to the institutionalized discriminatory policies against retention of many of these service members who report or are forced to present for evaluation (commander-directed evaluations: members must report for psychiatric evaluation or face disciplinary action), it should not be surprising that the occupational attrition rate is high. This finding is not generalizable to most nonmilitary populations.

I do not quibble with the authors in their statement that the "U.S. military remains one of the most highly respected and effective military organizations in the world" (p. 1582), but to imply in their last paragraph that their data "do not suggest that the impact of mental disorders is greater among service members than in the general population" (p. 1582) suggests that the authors are unfamiliar not only with military medical regulations regarding the administrative management of personnel with mental disorders extant during their study period but also with the hard-won protections many Americans with psychiatric disorders enjoy in civilian venues. Their final conclusion that "mental disorders may have a greater adverse influence on occupational functioning than any other medical illness category" (p. 1582), is not supported by their data if one takes into account the administrative context outlined.