Paradoxical changes in alcohol consumption and CAGE ratings between 1996 and 2006 in the general population of Geneva (2011)

AIMS: To assess changes between 1996 and 2006 in alcohol use and dependence in the general population of Geneva, Switzerland. METHODS: Postal surveys of alcohol consumption patterns in representative samples of the general population were compared. The four-item CAGE (Cut down, Annoyed, Guilty, Eye-opener) screener was used to estimate alcohol dependence. RESULTS: There were 742 participants in 1996 (response rate 75%) and 1487 in 2006 (76%). The proportions of daily drinkers (1996: 18%; 2006: 10%; P < 0.001), and of heavy drinkers (>14 drinks/week) decreased between 1996 (13%) and 2006 (7%; P < 0.001). However, decreases in heavy drinking were limited to men aged 30+ and women aged 30-49. The average number of drinks/week decreased from 6.4 in 1996 to 5.0 in 2006 (P < 0.001). Perception of safe drinking levels decreased between 1996 (4 drinks/day) and 2006 (3 drinks/day, P < 0.001). The prevalence of alcohol dependence, defined as a CAGE score >= 2, increased among women (1996: 6%; 2006: 14%; P = 0.001), but not among men (1996: 18%, 2006: 19%). Two of the CAGE items were endorsed more frequently by women in 2006 than in 1996: 'guilty' (1996: 9%; 2006: 15%; P = 0.018) and 'should cut down' (1996: 12%; 2006: 18%; P = 0.04). Drinking and driving did not change significantly over time (2006: 21% of men, 9% of women). CONCLUSION: Alcohol consumption, heavy drinking and perceived levels of safe drinking decreased over these 10 years, but not drunk driving or alcohol dependence. Paradoxically, alcohol dependence even increased among women, driven by CAGE items 'guilty' and 'should cut down', suggesting that changes in CAGE scores reflected changes in the social acceptability of drinking rather than changes in dependence.