The finding: More than 35,000 men aged 65-85 years living in a metropolitan Australian community were followed for a median of 12.7 years. 240 (0.6%) had a history of a prior suicide attempt and 443 (1.2%) attempted suicide during the follow up period, including 69 who died by suicide. Not surprisingly, past suicide attempt, bipolar disorder, alcohol use disorder and substance abuse disorder were associated with a new suicide attempt. However, past suicide attempt was not a robust predictor of future suicide completion. Rather, bipolar and depressive disorders, as well as men with multiple health systems affected by disease (defined as the presence or absence of: cancers, blood and immune diseases, endocrinological diseases, disease of the nervous system, diseases of the circulatory system, diseases of the respiratory system, diseases of the digestive system, and diseases of the genitourinary system) were associated with suicide completion. The sub-hazard ratios for these groups were: bipolar disorder (SHR=7.82), depressive disorders (SHR=2.26), 3-4 health systems affected by disease (SHR=6.02), and 5 or more health systems affected (SHR=11.18). The population fraction of suicides attributable to having 5 or more health systems affected by disease was approximately four times that for any mood disorder (79% vs 17%). A majority of men who died by suicide had no psychiatric diagnosis.

Strength and weaknesses: Screening tools and interventions designed to identify and avert suicide based on psychiatric diagnoses may miss important factors that contribute to suicide behavior; this study set out to shed light on whether the medical morbidity burden in older men contributes to the risk of suicide behavior. The recording of suicide relied on hospital morbidity data and may have failed to capture attempts or completions that did not result in hospital admission or those obscured by trauma, e.g., motor vehicle accidents. An arbitrary decision was made to rate disease burden by specifying eight common areas of health compromise or frailty rather than choosing specific diseases; there are pros and cons to each approach but the latter would not have enabled sufficient power to draw conclusions between specific diseases and suicide. We do not know if the same findings would hold in older women.

Relevance: The consultation psychiatrist regularly receives requests to assess older men who have made a despondent comment in the face of serious medical illness. This study reminds us to carefully consider the multiplicity of health morbidities a patient may be facing, and also to extend our attention beyond the diagnosis and management of a mood disorder (if present) and look for opportunities to decrease or postpone morbidity/frailty and strengthen or enhance constructive coping. It may be that helping the patient to lean effectively against the encroachment of their medical diseases will extend life and avert suicide more frequently than recommending another antidepressant.