Drinking patterns as predictors of alcohol withdrawal reactions in DBA/2J mice.

Individually housed DBA/2J mice were fed a liquid diet in which ethanol supplied 33% of the calories. The level of physical dependence that developed was estimated by scoring convulsions, elicited by handling the mice, after discontinuing the alcohol diet. The severity of the withdrawal reaction increased progressively with duration (5-12 days) of alcohol administration. A 2-day period on the diet produced no withdrawal reaction. Pretreatment of the mice with alcohol in their drinking water slightly increased the subsequent intake of the liquid diet. "Effective" alcohol intake was defined as uninterrupted alcohol consumption above 10 g/kg/day. Withdrawal scores correlated better with effective intake than with total intake under a variety of conditions. We interpret this to mean that brief interruptions in drinking (1 day) may allow the accrued physical dependence to disappear. On the basis of their effective alcohol intake, mice could be assigned to nondependent, moderately dependent or severely dependent groups for further study of the nature of physical dependence.