This is a nationwide cohort study of Finnish adults aged 25 and older, who are followed up for the incidence of an alcohol-attributable hospitalization or death. The primary objective is to examine the relative and absolute socioeconomic differences in cause-specific alcohol-attributable hospital admissions and deaths.

Background

Alcohol consumption is a leading risk factor for death and disability, resulting in substantial societal costs and social and community harm. Alcohol-related harm has been consistently shown to be higher among people of lower socioeconomic status, despite the fact they often report similar or lower levels of alcohol use. This apparent contrast has been called the alcohol harm paradox.

Understanding what explains the alcohol harm paradox is crucial for the design and implementation of population health interventions to reduce socioeconomic differences in alcohol harm. Explanations for the paradox can be broadly categorised into three groups: (i) differential exposure to alcohol, drinking patterns and trajectories and to other behavioural risk factors and joint effects among them, (ii) differential vulnerability resulting from individual factors, such as biological characteristics, psychological traits or stress, cumulative disadvantage and broader community and societal upstream factors, and (iii) differential biases in the measurement of alcohol exposure.

A crucial weakness of existing empirical studies comes from the operationalization of alcohol harm. With few exceptions, a vast majority of studies has used a composite endpoint combining several causes of alcohol-attributable deaths or hospitalizations or even merging alcohol-attributable hospital admissions with deaths in a single outcome. While this strategy increases the statistical power to analyse rare events, composite endpoints are prone to misclassification bias by masking divergent underlying patterns and associations.

Cause-specific analyses might shed light on different mechanisms driving the socioeconomic differences in overall alcohol-attributable harm, as well as opening potential avenues for policy interventions to reduce them. The study will take advantage of a recently formed dataset covering the total Finnish population to explore associations between socioeconomic status and cause-specific alcohol-attributable events. An additional contribution to the literature will be reporting both relative and absolute inequalities in alcohol-attributable harm. Absolute differences have been rarely reported even though might be as relevant as relative differences because, from a pragmatic standpoint, they are more feasible to reduce.

The study will aim to (1) examine the relative and absolute socioeconomic differences in cause-specific alcohol-attributable hospital admissions and deaths; (2) describe the geographical differences in cause-specific alcohol-attributable hospital admissions and deaths; and (3) quantify the relative contribution of each specific alcohol-attributable cause to the overall alcohol-attributable harm.

Inclusion Criteria:

• Permanent residents (defined by Statistics Finland as those Finnish and foreign nationals who have a legal domicile in Finland and intend to stay (or have stayed) for at least one year) alive in Finland by December 31, 2015
• Age 25 years or older by January 1, 2016
• Not having an alcohol-attributable hospitalization during the past three years ((January 2013-December 2015

Exclusion Criteria:

• Temporary residents, including foreign nationals living in Finland for less than a year, asylum seekers who have not been granted a legal domicile, temporary migrant workers and Finnish nationals living temporarily abroad by December 31, 2015

All permanent residents in Finland older than 25 years old and without an alcohol-attributable hospital admission during the past three years

The investigators will share the harmonization protocol to define the variables and the statistical code. Individual participant data is available upon request at FinData, following national data protection procedures.