Age-period-cohort analysis of tuberculosis notifications in Hong Kong from 1961 to 2005

Background: Despite its wealth, excellent vital indices and robust health care infrastructure, Hong Kong has a relatively high incidence of tuberculosis (TB) (85.4 per 100 000). Hong Kong residents have also experienced a very rapid and recent epidemiological transition; the population largely originated from migration by southern Chinese in the mid 20th century. Given the potentially long latency period of TB infection, an investigation was undertaken to determine the extent to which TB incidence rates reflect the population history and the impact of public health interventions. Methods: An age-period-cohort model was used to break down the Hong Kong TB notification rates from 1961 to 2005 into the effects of age, calendar period and birth cohort. Results: Analysis by age showed a consistent pattern across all the cohorts by year of birth, with a peak in the relative risk of TB at 20-24 years of age. Analysis by year of birth showed an increase in the relative risk of TB from 1880 to 1900, stable risk until 1910, then a linear rate of decline from 1910 with an inflection point at 1990 for a steeper rate of decline. Period effects yielded only one inflection during the calendar years 1971-5. Conclusions: Economic development, social change and the World Health Organisation's short-course directly observed therapy (DOTS) strategy have contributed to TB control in Hong Kong. The linear cohort effect until 1990 suggests that a relatively high, but slowly falling, incidence of TB in Hong Kong will continue into the next few decades.published_or_final_versio