Aim: To study the effects of age, period and cohorts on alcohol-related mortality trends in Sweden.
Design: The study comprises an age-period-cohort analysis.
Setting and participants: The analysis was based on all deaths in the Swedish population between 1969 and 2002.
Measurements: Data on alcohol-related deaths in Sweden from 1969 to 2002 excluding accidental injury and homicide were used. The analysis covered 43 021 deaths.
Findings: Time period and birth cohort both influenced alcohol-related mortality. Male cohorts born in the 1930-40s exhibited the highest alcohol-related mortality, while for females those born in the 1940-50s had the highest alcohol-related mortality. For both men and women, those born in the 1960-70s had the lowest age-adjusted alcohol-related mortality. High-risk cohorts were young or in early adulthood during the periods that alcohol became more available in Sweden. The low-risk cohorts of the 1960-70s were brought up during a period when society was concerned with increasing alcohol problems and more emphasis was placed on issuing alcohol awareness information in schools.
Conclusions: Cohort effects were found suggesting that the link between alcohol consumption and non-accident alcohol-related mortality at the population level is dependent on other factors that may change over time. One such factor may be that restrictive alcohol policies have a greater effect on drinking in those who are younger at the time they are put into effect.