Declines in mental illness over the adult years: an enduring finding or methodological artifact?

Epidemiological surveys have revealed dramatic declines in the prevalence of serious mental illness (SMI) as adults age. Analyses of these datasets have not determined whether this is attributable, on one hand, to declining incidence, shorter duration of condition, and lesser severity with age, or on the other hand, confounding variables and methodological biases. This study, thus, aims to test several such competing hypotheses.

This study employs a secondary analysis of data from the 2001/2002 US National Comorbidity Replication Survey of 9282 adults, 18 and older, living in the household population of the 48 contiguous states, as well as supplemental datasets from Israel, New Zealand, and other sources.

One-year SMI prevalence rates in the US drop from 8.0% of adults 18-29 to 1.4% of the 65+, and to similar degrees in Israel and New Zealand. The drop in the US can be explained by the early onset of most mental illnesses, and declines in both incidence and duration of condition with age. Comorbidity also drops with age; however, the remaining diagnoses show a gradually increasing severity. Institutionalization explains a small portion of the declines, as does premature mortality of the seriously mentally ill that accounts for 16.3% of the decline.

The results reveal that a substantial portion of the declines are explainable in terms of declining incidence and improving recovery rates, with some reduction also attributable to institutionalization and premature mortality which removes some older and more seriously disabled adults from the epidemiological survey populations.