Population-level relationships between alcohol consumption measures and Ischemic Heart Disease mortality in U.S. time-series.

BACKGROUND: Individual-level studies indicate the possibility of both protective and harmful effects of alcohol consumption on Ischemic Heart Disease (IHD) mortality depending on the pattern of consumption. Population-level relationships could be in either direction and previous studies have found mixed results.
METHODS: Population-level relationships between IHD mortality rates and per capita consumption of alcoholic beverages, cirrhosis mortality rates, cigarettes, and sugar sweetened soda for the period from 1950 to 2002 are modeled using autoregressive integrated moving average (ARIMA) and vector error correction methods.
RESULTS: In multivariate ARIMA models controlling for accumulated heavy drinking as represented by cirrhosis mortality, a protective effect of 4%/l was found for total alcohol consumption while cirrhosis mortality rates had significant positive effects on IHD rates. Beverage-specific models found no effect for wine, positive risks for spirits, and significant protective effects for beer. The protective effects for both total alcohol and beer were also found in vector error correction models. Significant positive effects of cigarette sales on IHD rates were also found in both types of models.
CONCLUSIONS: The complexity of alcohol's relationship with IHD is highlighted. Aspects of pattern represented by beverage-specific consumption and cirrhosis mortality indicate potential protective effects from moderate drinking and harmful effects from heavy drinking in accord with individual-level findings.