Past trends of obesity attributable mortality in Europe: an application of age-period-cohort analysis

Nikoletta Vidra, University of Groningen
Maarten J. Bijlsma, Max-Planck Institute for Demographic Research
Sergi Trias-Llimós, University of Groningen
Fanny Janssen, Netherlands Interdisciplinary Demographic Institute (NIDI) and University of Groningen

Background: Obesity has dramatically increased over time and constitutes a major health burden which can be estimated by obesity-attributable mortality. Although there is evidence of age, period and cohort increments on obesity mortality association, previous studies in Europe did not account for the multiple dimensions of the obesity epidemic, namely age, period and cohort. Objective: To better capture the complexity of the obesity epidemic and its impact on mortality by assessing age, period and birth cohort effects and patterns in Europe, in the past. Data and Methods: We obtained the following data (by age and sex): Obesity prevalence by available sources, Relative Risks (RR) of dying from obesity from a recent meta-analysis and all-cause mortality by Human Mortality Database. We applied the standard Clayton & Schifflers age-period-cohort analysis. Results: Based on our preliminary results for the Netherlands, obesity-attributable mortality doubled in between 1981 and 2010; in Dutch men, the fraction of mortality due to obesity rose from 0.7 % to 1.3 % while in Dutch women from 1.0 to 2.0 %. The effect of birth cohort to obesity-attributable mortality was larger among Dutch women as compared to men. In specific, for women born after 1941-1945, obesity-attributable mortality is increasing with every next generation. Conclusions: Next to age and period a substantial effect of birth cohort on obesity-attributable mortality was shown for the Netherlands, especially in women. Future studies on obesity–attributable mortality should not ignore the multiple dimensions of obesity.