Parents were classified into four different categories: "abstainers", "light drinkers", "at risk drinkers" and "alcohol abusers". Abstainers were categorized as a separate group since they differ in some respects from people with very low consumption [10,47]. Parents were classified as alcohol abusers if they were among the top 10% consumers within their gender, together with having scored at least 1 on the collapsed CAGE questionnaire. Parents who either had a positive score on the collapsed CAGE or who were among the top 10% consumers were coded into the "at risk" category. The remaining responders were categorized as light drinkers and used as reference group. This classification rendered 2.2% (135) of participating mothers as alcohol abusers and 12.8% (781) as being at risk, while 4.5% (219) of fathers were alcohol abusers and 16.5% (807) at risk. Mothers and fathers classified as abusers scored on average 16.8 and 26.7 on the summative index combining frequency and amount mentioned above, which is 3.4 and 3.9 times as high as the sample means. The test-retest reliability was measured among 8298 parents who participated both
