Exploring the dynamic relationship between alcohol consumption trajectories and different aspects of health

Katalin Gémes, Ph.D., Karolinska Institutet, Institute of Environmental Medicine, Unit of Translational Epidemiology, Stockholm, Sweden

Background and motivation:
While heavy drinking, i.e., over 60 gr/ day, increases overall mortality and morbidity, moderate alcohol consumption, i.e., 10-40 gr daily alcohol is associated with lower overall mortality and lower risk for cardiovascular and other chronic diseases when compared to non-drinking. While several biological mechanisms have been proposed to explain these protective associations, there is an ongoing debate and some contradictory evidence, which questions the validity and the exact public health implications of these findings. One of the primary concern is related to the dynamic relationship between alcohol consumption and health. Individuals may decrease their alcohol consumption or quit drinking due to a decline in their health status. Therefore studiesmeasuring drinking habits at one point in time may overestimate the preventive effect of moderate alcohol consumption. Evidence on how individuals may change their drinking habits due to declining health and which sociodemographic, lifestyle factors may predict this change are contradictory. Therefore, the project aims to understand how the changes and stability of alcohol consumption over time are associated with different aspects of health and which sociodemographic and lifestyle factors may explain these associations.

How do different alcohol consumption trajectories over time are associated with general and mental health? Which social and lifestyle factors may explain the association between alcohol consumption trajectories over time and general and mental health? 

I will use data from the Stockholm Public Health Cohort (SPHC). The SPHCis a prospective population-based study conducted in the Stockholm County in four waves: in 2002, 2006, 2010 and 2014, among individuals between 18 and 84 years of age. In each time a new randomly selected sub-cohort was enrolled and were followed-up every four years. Overall, 50,157 individuals participated in the different waves. Amount and frequency of alcohol consumption, hazardous drinking habits as well as self-rated health, psychological distress, chronic diseases, lifestyle and psychosocial factors were assessed by questionnaires. Information on sociodemographic was obtained through register linkage. For this study, I will use the 2002 and 2006 sub-cohorts with follow-up surveys in 2010 and 2014 (Table1).

Table 1: Number of participants and follow-up for wave1 and wave2 that will be used in this study from the Stockholm Public Health Cohort.

Survey waves (W) of the SPHC





W1 (2002)





W2 (2006)





Alcohol consumption trajectories over time will be constructed, based on longitudinal alcohol measures and logistic regression will be used to investigate the association between the different alcohol trajectory membership and self-rated health and psychological distress.