Studies in a few countries (including the US) have reported that mortality rates in the population from psychiatric disorders are much higher when they are based on all causes of death ("multiple causes" or "mentions") coded on death certificates versus only the underlying cause. Studies appear to be lacking on geographic variation within the US in mortality rates from psychiatric disorders based on multiple causes of death. The present study examined the US age-standardized rate (ASR) for death with depression using multiple causes versus underlying cause alone in each of the Census Bureau's four regions and nine divisions. ASRs for schizophrenia were also examined for comparison. For the entire US, the ratio of the ASR based on multiple causes to the ASR based on underlying cause was 20.9 for depression and 9.2 for schizophrenia; in analyses by region and division, these ratios showed limited variation. The most consistent finding for both depression and schizophrenia was that ASRs, whether based on multiple causes or only on underlying cause, were highest in the Midwest region (especially the East North Central division) and lowest in the South (and in each of its three divisions). For ASRs (using multiple causes of death) from depression, these regional differences were evident within each of several levels of urbanization. For deaths with depression coded as other than the underlying cause, ASRs for each of the three most common underlying causes (cardiovascular diseases, intentional injuries, and neoplasms) were highest in the Midwest and lowest in the South. Studies are needed to determine if these regional differences in mortality from depression are due to regional differences in: certifier practices (i.e., in assigning causes of death among persons with psychiatric conditions); the prevalence (among persons with psychiatric disorders) of lifestyle-related factors (e.g., tobacco use and obesity) that mediate mortality risks; and/or in unmet need for psychiatric treatment and medical care for other chronic diseases in persons with psychiatric conditions. Similar studies are needed of regional variation within other countries.
