Diagnostic inflation in the DSM: A meta-analysis of changes in the stringency of psychiatric diagnosis from DSM-III to DSM-5

It is often argued that successive editions of the Diagnostic and Statistical Manual of Mental Disorders (DSM) have relaxed diagnostic criteria and thereby inflated rates of diagnosis. This claim has yet to be examined systematically. We quantitatively reviewed 123 studies in which one sample was concurrently diagnosed using two consecutive DSM editions (i.e., DSM-III & DSM-III-R, DSM-III-R & DSM-IV, or DSM-IV & DSM-5). Meta-analysis of 476 risk ratios enabled 123 comparisons of diagnostic rates for specific disorders. Comparisons demonstrating diagnostic inflation (i.e., higher diagnostic rate in the later edition) did not exceed those demonstrating deflation. The average risk ratio was 1.00, indicating no overall change in diagnostic stringency from DSM-III to DSM-5, and there were no reliable tendencies for particular DSM revisions to be more inflationary or deflationary than others. Specific disorders showing reliable evidence of diagnostic inflation or deflation were identified. Notable examples of diagnostic inflation included Attention-Deficit/Hyperactivity Disorder (ADHD), autism, eating disorders, and substance dependence. Although serious concerns have been raised about diagnostic inflation or "concept creep" in the DSM, these concerns may have been overstated.