Birth cohort analysis of suicide mortality in Belgium 1954-1981 by a graphic and a quantitative method.

The effect of birth cohort upon the recent increase in Belgian suicide trends was analyzed by two techniques: a graphic display and an Age-Period-Cohort (APC) modelling procedure. Both suggest an analogous pattern: suicide mortality steadily increased within successive male cohorts born after 1920 and within female cohorts born even before 1900. Between 1970 and 1980, period related factors seem to have operated independently. Every technique of cohort analysis however has important interpretational limitations because of the variable interplay between age period and cohort factors. As a consequence, explanations remain very tentative, Notwithstanding its rather poor etiological power in suicide epidemiology, cohort analysis is very useful in the more accurate description and prediction of suicide trends.