Alcohol consumption and community hospital admissions in the United States: a dynamic regression analysis, 1950-1992.

The annual time-series analysis examines the impact of changes in per capita alcohol consumption (NIAAA,AEDS) on changes in community hospital admission rates (AHA) in the United States from 1950 to 1992 (n = 43). Increases in per capita alcohol consumption were expected to increase hospital admission rates contemporaneously and several years thereafter following an exponential risk function. Distributed lag models based on differenced data controlling for changes in: (1) per capita cigarette consumption; (2) private hospital insurance coverage; (3) the drinking age population; (4) per capita disposable personal income; and (5) health care regulatory interventions show a contemporaneous effect of per capita alcohol consumption on hospital admission rates. The time-series analyses imply that between 22-26% of US community hospital admissions are alcohol related. A comparable analysis indicates that per capita alcohol and tobacco expenditures contribute to approximately 28% of US community hospital admissions. The absence of statistically significant lagged effects is inconsistent with an exponentially declining risk functions. However, the contemporaneous effects of per capita alcohol and tobacco consumption suggest that a reduction in smoking and drinking will produce quick reductions in morbidity and hospitalizations.