Shifting Cohort Patterns in the Use of Drugs with Elevated Overdose Risk in the United States.

Rising drug overdose rates are a major social problem, but understanding of trends in the use of high-risk drugs is limited. The increasingly addictive potential of high-risk drugs, broader social changes, and the importance of peers and social contexts in shaping use may create conditions where some cohorts have elevated use further into adulthood than others. We use an age-period-cohort model that defines cohort effects as the differential influences of social events for individuals of different ages. We analyze cross-sectional nationally representative data from the 1979-2018 National Survey on Drug Use and Health (NSDUH) to study past-year (mis)use of prescription analgesics, heroin, cocaine, methamphetamine, and any of these drugs. Pre-1990 cohorts had either consistently lower-than-average odds or decreasing odds of use as they aged. The 1990s cohorts had higher-than-average odds of use, which increased as they aged. Early-2000s cohorts had increasing odds of use with age despite low odds in adolescence. High-risk drug use appears to be an important cohort differentiating mechanism, with implications for theories of deviance and stratification and for policy. The typical focus on period trends obscures the elevated odds of high-risk drug use in certain cohorts, whose members are entering the ages when overdose is most likely and consequences of substance use become most pronounced.