- Funding: NIAAA R01 AA021448
- Principal Investigator: William Kerr, PhD
Establishing causal relationships between alcohol consumption, drinking patterns and common health problems is important for health education of consumers, alcohol policy and regulation, prevention programming and cost estimates for health services. While a large body of literature has addressed this topic, the conclusions that can be drawn have been limited by shortcomings in three important and inter-related areas: poor measurement of alcohol intake and limited assessment of life-course drinking and abstention; lack of attention to the influence of health problems on subsequent alcohol patterns; and inadequate treatment of potentially confounding health risk factors, such as childhood adversity and economic hardship, and health behaviors such as tobacco use, drug use, and exercise. Moreover, there has been limited attention to racial/ethnic differences in the inter-relationships of alcohol and health, despite suggestive evidence of such. To address these issues and gaps in the extant literature, the study will conduct new analyses of the National Longitudinal Survey of Youth (NLSY) 1979 and 1997 cohorts and of the 2010 National Alcohol Survey (NAS).
We will examine: (1) effects of specific prevalent health conditions (heart disease/heart problems, hypertension, diabetes, cancers and stroke), alcohol-attributable health problems and self-reported health status on alcohol consumption over the life course, (2) influences of life-course alcohol use and childhood adversity (physical abuse, sexual abuse, parental alcoholism, economic hardship and serious illnesses) on the onset of specific health conditions, and (3) effects of life-course alcohol use patterns, childhood and adult adversity and economic impacts of the 2008-09 recession on self-reported general health status.
The study is informed by a life-course perspective, focusing on both the influences of drinking on health and the influences of health on drinking while attending to important alternative health risk factors. Both the NLSY and 2010 NAS studies over-sampled Hispanics and African Americans, allowing consideration of factors that may contribute to racial and ethnic disparities in health conditions and problems. This project uses innovative statistical methods for longitudinal data including marginal structural Cox proportional hazards survival models and latent class growth curve models as well as retrospective cohort design survival models and propensity score matching.
Results will provide new insights into inter-relationships between alcohol use and heath by avoiding common misclassification problems that undermine the validity of most previous research linking alcohol with health outcomes. Study findings will document the prevalence, severity and impact on future drinking of general and alcohol-attributed health problems, provide estimates of alcohol pattern risks for specific health conditions and overall health status, and will provide estimates of impacts of childhood adversity and economic losses in the recent recession on these health outcomes.