Mortality trends of multiple sclerosis in Spain, 1951-1997: an age-period-cohort analysis.

Trends of multiple sclerosis mortality in Spain from 1951 to 1997 are examined. Age-adjusted mortality strongly decreased from 3.08 per 100,000 women in 1951-1967 to 0.59 in 1968-1980 (similar figures were obtained for men). This decrease seems to be due to an artifact produced by changes in codification of causes of death. An age-period-cohort analysis, limited to the period 1968-1997, showed that the mortality trend in Spain cannot be fully explained by year of death (period effect), but that a cohort (year of birth) effect is also necessary. An increase in mortality related with the cohort of birth was detected: people born after 1953 had double the risk of those born between 1938 and 1947, and four times the risk of those born between 1923 and 1932. Regarding the period effect, there is a decrease in mortality, probably due to improvements in life expectancy of multiple sclerosis patients.