Table 2.
Adjusted hazard ratios of tuberculosis disease in early childhood by the timing of cytomegalovirus acquisition
| Cytomegalovirus acquisition by age 3 weeks and younger* | Cytomegalovirus acquisition by age 6 weeks and younger | Cytomegalovirus acquisition by age 3 months and younger | Cytomegalovirus acquisition by age 6 months and younger | Cytomegalovirus acquisition by age 12 months and younger | Cytomegalovirus acquisition by age 24 months and younger | ||
|---|---|---|---|---|---|---|---|
| Primary analyses: follow-up for tuberculosis disease | |||||||
| After age 6 months | 3·7 (1·1–11·9) | 3·3 (1·2–9·3) | 2·3 (1·3–4·1) | 3·1 (1·8–5·5) | 2·7 (1·6–4·6) | 3·6 (2·0–6·4) | |
| Tuberculosis episodes | 43 | 44 | 50 | 52 | 57 | 58 | |
| Follow-up, person–years | 5334 | 5382 | 5734 | 5904 | 6037 | 6197 | |
| After age 1 year | 3·9 (0·9–16·4) | 4·1 (1·2–13·8) | 2·8 (1·4–5·8) | 3·6 (1·7–7·3) | 3·2 (1·6–6·4) | 4·2 (2·0–8·8) | |
| Tuberculosis episodes | 27 | 27 | 31 | 33 | 37 | 38 | |
| Follow-up, person–years | 4544 | 4584 | 4882 | 5028 | 5138 | 5276 | |
| Secondary analyses: follow-up for tuberculosis disease | |||||||
| Age 6 months to 1 year | 3·2 (0·4–24·6) | 2·0 (0·3–15·3) | 1·6 (0·6–4·2) | 2·5 (1·0–6·2) | 2·0 (0·8–4·9) | 2·6 (1·0–6·8) | |
| Tuberculosis episodes | 16 | 17 | 19 | 19 | 20 | 20 | |
| Follow-up, person–years | 791 | 797 | 851 | 877 | 898 | 921 | |
| Age 1–2 years | 6·8 (0·8–56·8) | 10·9 (2·1–55·3) | 3·8 (1·1–13·2) | 4·0 (1·2–13·4) | 5·6 (1·6–20·2) | 8·4 (1·9–37·4) | |
| Tuberculosis episodes | 8 | 8 | 10 | 12 | 14 | 15 | |
| Follow-up, person–years | 1548 | 1560 | 1664 | 1713 | 1753 | 1799 | |
| Age 2–9 years | 2·7 (0·4–20·5) | 1·8 (0·2–13·9) | 2·5 (1·0–6·0) | 3·3 (1·4–8·1) | 2·4 (1·0–5·5) | 3·0 (1·2–7·2) | |
| Tuberculosis episodes | 19 | 19 | 21 | 21 | 23 | 23 | |
| Follow-up, person–years | 3776 | 3811 | 4058 | 4179 | 4270 | 4386 | |
| At any timepoint | 2·8 (0·9–9·1) | 3·1 (1·2–7·8) | 2·1 (1·3–3·4) | 2·3 (1·4–3·7) | 2·3 (1·4–3·7) | 2·9 (1·7–4·7) | |
| Tuberculosis episodes | 55 | 58 | 66 | 69 | 74 | 75 | |
| Follow-up, person–years | 5339 | 5386 | 5739 | 5910 | 6042 | 6203 | |
Data are adjusted hazard ratio (95% CI) or n. All models are adjusted for child sex, study site, and maternal HIV. The reference values for each row are children that are cytomegalovirus negative at that timepoint. Cox regression models were performed with differing timing of cytomegalovirus positivity and follow-up time for tuberculosis disease. Sensitivity analysis adjusting for further risk factors (including socioeconomic status, birthweight, household exposure, tuberculin conversion, and number of cytomegalovirus tests) can be seen in the appendix (pp 12–13). Excluding congenital cytomegalovirus infections did not affect these results and can also be seen in the appendix (pp 14–15). Cytomegalovirus positivity is cumulative and therefore, these cutoffs include all positive tests before that age.
Defined as congenital cytomegalovirus.