Objective: To determine the association of initial tuberculin sensitivity, age and sex with the development of tuberculosis.
Methods: A 15-year follow-up of 280000 subjects in south India, where new cases of tuberculosis were detected mainly by periodic population surveys. Life-table technique was employed to estimate tuberculosis incidence and disease risk in survivors. The independent effect of tuberculin sensitivity, sex and age at intake was determined using Cox's proportional hazard model.
Results: Taking subjects with reaction size 0-7 mm to 3 IU PPD-S as reference group, the adjusted relative risk (RR) for developing culture-positive tuberculosis was 1.1, 1.9, 2.9, 3.6 and 3.3 for those with indurations of 8-11, 12-15, 16-19, 20-24 and > or = 25 mm (P < 0.01). Considering subjects aged 0-4 years as reference group, the adjusted RR for the other groups increased from 1.7 to 10.8 (P < 0.01). Males had a substantially higher incidence (adjusted RR 3.0, P < 0.001). The risk of culture-positive tuberculosis over 15 years in survivors was 3.3% (5.0% in males and 1.6% in females), and increased substantially with tuberculin sensitivity at intake. In those with > or = 12 mm at intake, the approximate lifetime risk was 6.1% (8.6% in males and 3.1% in females).
Conclusion: The incidence of tuberculosis increased steadily with tuberculin sensitivity to PPD-S and age at intake. Males had a significantly higher risk than females in every PPD-S group and the overall risk was three-fold higher.