Age-period-cohort analysis of pulmonary tuberculosis mortality in Taiwan: 1961 to 1990.

The specific aim of this study was to examine the effects of age, calendar period of death, and birth cohort in pulmonary tuberculosis (TB) mortality in Taiwan during the period 1961 to 1990. A log-linear Poisson regression model modified from the method of Osmond and Gardner was used and 79,881 deaths (58,025 males and 21,856 females) were included in the analysis. Birth cohort is the most significant predictor of pulmonary TB mortality according to the model. The earliest birth cohort from 1891 had a pulmonary TB mortality 17,327 and 6,186 times those born from 1986 for males and females, respectively. There was also a significant age effect. The youngest age group of zero to four years had a pulmonary TB mortality 7.10 and 5.87 times those for the age group of five to nine years for males and females, respectively. The oldest age group of 70 to 74 years had a risk of pulmonary TB mortality 2.89 and 1.88 times those for the five- to nine-year age groups for males and females, respectively. Parameters of the period factor showed a decreasing pulmonary TB mortality from 1961 to 1990 that was less significant than those of age or cohort. In addition to the improvement in medical measures that influenced the effect of calendar year on TB mortality, year of birth is an important determinant in the trend of TB mortality in Taiwan. The result shows that the major focus for TB mortality in Taiwan is in the age groups born between 1891 and 1921.