A recent report found that patients with serious mental illness treated in eight state public mental health systems between 1997 and 2000 died on average 25 years earlier than the U.S. general population (1,2). The primary reasons attributed to the substantial years of potential life lost (YPLLs) include heart disease and the fragmentation of medical and mental health care (1). However, the extent to which this disparity is seen across other mental disorder diagnoses or in more integrated health care systems, notably the Veterans Health Administration (VHA), has not been determined. We assessed all-cause and heart disease-specific mortality risk for VHA patients with a diagnosis of serious mental illness, posttraumatic stress disorder (PTSD), or depression and for VHA patients without a mental disorder diagnosis and compared the YPLLs between these groups.

Data were from the VA Serious Mental Illness Treatment Research and Evaluation Center's National Psychosis Registry (2) and the Centers for Disease Control and Prevention's National Death Index.

In fiscal year 2005, 4.2% of VHA patients with serious mental illness, 1.4% with PTSD, 2.8% with depression, and 2.7% without a mental disorder died from all causes. Overall, 1.0% of patients with serious mental illness, .4% with PTSD, .7% with depression, and .8% without a mental disorder died from heart disease.

Age- and sex-adjusted mean YPLLs from all-cause and heart disease-specific mortality (Table 1) were similar in magnitude for VHA patients with serious mental illness, depression, and no mental disorder, but they were higher for those with PTSD. Nonetheless, for all mental disorder diagnoses, YPLL estimates were lower than those reported for patients without a mental disorder outside of the VHA (25 years) (1). It appears that observed disparities in YPLLs are less severe in the VHA patient population.

This research was supported by grant IIR 07-115 from the Health Services Research and Development Service, VHA, and by the VA Office of Mental Health Services.

The authors report no competing interests.