PREDICTING ALCOHOL-WITHDRAWAL SEVERITY - SUPPORT FOR THE ROLE OF EXPECTATIONS AND ANXIETY

Seventy clients of a community based district alcohol service particip ated in this study to test hypotheses about predicting the severity of the alcohol withdrawal syndrome. All participants undertook home deto xification and were supported by a member of the alcohol service. A to tal of 64 participants completed the required five days of withdrawal. Two models of withdrawal were tested. Recent drinking and withdrawal history predictors (the DW model) consisted of: the total amount of al cohol consumed in the 7 days prior to stopping, the length of the most recent period of uninterrupted daily drinking, age, the number of pre vious occasions withdrawal symptoms had been experienced, and medical condition. The expectations and anxiety predictors under test (the EA model) consisted of: expectations of physical or psychological symptom s, trait anxiety, and state anxiety. Each model was tested on two sets of symptoms experienced on day one of withdrawal: physical and psycho logical. For both sets of symptoms, the results of multiple regression analysis showed the EA model variables to be significantly more predi ctive of withdrawal symptom severity than the DW model variables. The strongest EA predictor for both physical and psychological symptoms ex perienced was symptom expectancy. Further analyses were conducted to a ttempt to trace the origins of higher withdrawal expectations in, for example, higher levels of concerns over stopping drinking and remainin g stopped. A new model is proposed combining psychological and drinkin g/withdrawal history predictors.