Objectives Alcohol use in Russia has a number of relatively distinct features which have been shown to be strongly associated with mortality. The aim of this paper is to investigate how specific aspects of Russian drinking behaviour and its consequences are related to scores obtained from the internationally recognised AUDIT instrument for detecting alcohol problems.

Design Population-based age-stratified cross sectional survey.

Setting Izhevsk, Russia.

Participants 925 men aged 25–59 years.

Main outcome measure Men and proxy respondents completed an interview including questions on alcohol consumption and socio-economic and demographic variables. These included questions on six drinking behaviours and their consequences. These were: going on zapoi (a period of continuous drunkenness of two or more days); drinking non- beverage alcohol; frequency of excessive drunkenness; frequency of hangover; drinking before noon and frequency of sleeping in clothes at night because of drunkenness. The men also completed a Self Completed Questionnaire containing questions constituting the alcohol use disorders identification test (AUDIT).

Results There was a strong association with AUDIT score and hazardous drinking (one or more episodes of zapoi in the past year or drinking non beverage alcohol or twice a week or more occurrence of excessive drunkenness, hangover or going to sleep at night clothed because of drunkenness) using both proxy and self report of drinking behaviours (self-report OR 5.25 95% CI 2.86 to to 9.65). The association was stronger when proxy report was used (proxy-report OR 6.66 95% CI 4.00 to to 11.07). The same pattern was seen for all the drinking behaviours individually. Confirmatory factor analysis supported a two factor structure for the AUDIT measuring alcohol consumption and alcohol related problems. Hazardous drinking behaviour specific to Russia was more strongly associated with AUDIT questions relating to alcohol related problems than to questions measuring consumption.

Conclusions The drinking behaviours used in this paper are unconventional measures of hazardous drinking. The strong association with AUDIT shows that both self and in particular proxy report of distinctive directly observable drinking behaviours provide useful data on alcohol consumption and abuse in Russian men.