Suicide rates and alcohol consumption in the United States, 1970-89.

This study investigates the aggregate relationships between suicide rates and beverage specific measures of alcohol consumption for states in the United States over periods of from 14 to 20 years. Time series cross-sectional analyses of these aggregate state level data are presented which control for exogenous differences between states, time trends and covariations over time in nine measures; age composition, male population, non-white population, per capita land area, metropolitanism, income, unemployment, measures of religious preferences and divorce. After correcting for substantial autocorrelations in measurement error, the analyses revealed that suicide rates were specifically associated with spirits sales, age composition of state populations, per capita land area, unemployment and religious preferences over time. While suicide rates increased significantly as a function of increased spirits sales, beer and wine sales were not associated with suicide rates. These findings suggest that it is not the consumption of ethanol per se but rather the consumption of ethanol in the form of spirits that is related to suicides. Rather, it would appear that a population-based preference for the consumption of spirits is associated with suicide events.