Predictors of increased mortality in elderly depressed patients

Five‐year standardized mortality rates for a cohort of elderly depressed patients in Perth, Western Australia, were greater than expected (M = 2.047, F = 1.658). These results were very similar to the 4‐year rates in a London study (Murphy et al., 1988), as were the causes of death. There was a significantly greater chance of being alive at the end of 5 years if, on entry to the study, the patients were: female, less than 75 years of age, had no impairment of mobility, had a diagnosis of major depression with melancholia and psychosis, and had a good recovery from the depressive illness at the end of 12 months. Impairment of mobility was the best physical status indicator of subsequent mortality. Two measures of patient's self‐assessment of physical status were as good predictors of mortality as a physician assessment of health, other than impairment of mobility.