Sex ratio in multiple sclerosis mortality over 65 years; an age-period-cohort analysis in Norway

Abstract

Increasing female: male ratio in multiple sclerosis (MS) has been assigned to cohort effects, with females in more recent birth cohorts possibly being more exposed or vulnerable to environmental risk factors than males. We collected MS mortality data in Norway from 1951 to 2015 from The Norwegian Cause of Death registry. Age-Period-Cohort analysis was conducted using log-linear Poisson models, including sex interaction terms. MS was registered as the underlying, contributing or direct cause in 6060 deaths. MS associated mortality remained stable with a slight preponderance among males until after 1980, and have since increased preferentially among females. Throughout the study period the mean annual increase was 1.25% for females and 0.3% for males (p < 0.0001). Age-period-cohort analysis revealed limited evidence of cohort effects for the gender differences; the best fitting model only included gender-age and gender-period interaction terms. The period effect evened out for males in the last three decades but increased for females, especially among the oldest age-groups. In conclusion, the increased female: male mortality ratio in MS associated mortality is driven mainly by increased mortality among females in the three last decades, particularly in the older age groups. It is best explained by disproportional period effects, providing evidence of time-varying external factors including improved access to diagnosis among females.

Keywords: Age-period-cohort model; Mortality; Multiple sclerosis; Neuro-epidemiology; Sex ratio.