Murphy et al. (1988) found that, when the effect of physical illness was controlled, depressed elderly patients (particularly men) had a significantly higher four-year mortality, suggesting that the greater mortality in the depressed group was not due to differences of pre-depression physical health alone.