Characterizing midlife-onset alcohol dependence: Implications for etiology, prevention, and healthy aging

We evaluated the developmental epidemiology of midlife-onset alcohol dependence (AD) in the Dunedin Study (N=1,037), a population-representative cohort followed across five decades. At ages 18, 21, 26, 32, 38, and 45, past-year AD prevalence was 11.0%, 18.4%, 13.6%, 8.1%, 9.6%, and 11.3%, respectively. As expected, relative to never-diagnosed individuals, those with early-onset AD (first diagnosis: age-18 or age-21, prevalence=22.9%) were distinguished by a range of early-life and adult correlates. Individuals with midlife-onset AD (first diagnosis: age-38 or age-45, prevalence=5.6%) were distinguished by fewer early-life correlates, but exhibited a family history of AD, and adolescent dysregulation and marijuana-use. They were characterized by an array of adult correlates, including internalizing disorders, mental-health treatment-contact, criminal-behavior, perceived-stress, coping-by-drinking, lower likelihood of marriage and parenthood, and reduced preparedness for old age. They also experienced more adult alcohol-related impairment than the early-onset group. Results can guide efforts to reduce midlife alcohol-related problems and support healthy aging.