Table 5. Summary of studies included in the IPD meta-analysis.
| Study [Reference] | Country (Latitude) | Study Design and Study Population | Total Number of Participants | Number of TB Cases (%) | Median Age, Years (IQR) | Female, N (%) | HIV-Positive Cases, N (%) | Method of Measuring Vitamin D | Median Baseline 25-OH Vitamin D, nmol/L (IQR) | Length of Follow-up, Yearsa | TB Disease Definition | Adjusted Effect Estimate (95% CI) Reported From Original Study |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Arnedo-Pena et al., 2015 [28] | Spain (40.4637° N, 3.7492° W) | Prospective cohort study of household and community contacts of TB cases | 523 | 3 (0.6) | 37.0 (28.0–46.0) | 255 (48.8) | NAb | ECLIAs and CLIAs | 60.0 (42.5–79.3) | Mean 1.6 (± 0.9) | Smear or culture positive | aHR for continuous vitamin D and microbiologically confirmed TB: 0.88 (0.80–0.97) |
| Gupta et al., 2016 [29] | South Africa (30.5595° S, 22.9375° E) | Prospective case-cohort study of HIV-positive and HIV-exposed infants (no previous known TB exposure) | 366 | 100 (27.3) | 0.7 (0.5–0.7) | 196 (53.6) | 193 (52.7) | Immunoassay | 90.8 (75.8–109.0) | 3.7 | 2004 South African NTP criteria for definite, probable, or possible TBc | aHR for vitamin D < 80 nmol/L and any TB: 1.76 (1.01–3.05) |
| Mave et al., 2015 [30] | India (20.5937° N, 78.9629° E) | Nested case-control study of HIV-positive breastfeeding mothers (TB exposure status not specified) | 120 | 33 (27.5) | 23.0 (21.0–25.0) | 120 (100.0) | 120 (100.0) | Radioimmunoassay | 39.4 (24.4–47.5) | 1.0 | Culture confirmed OR Probable TB: (1) smear positive; and (2) histological and clinical features suggestive of TB and response to anti-TB treatment |
aOR for vitamin D < 50 nmol/L and any TB: 1.57 (0.49–4.98) |
| Owolabi et al., 2016 [31]d | The Gambia (13.4432° N, 15.3101° W) | Prospective cohort study of household contacts of TB cases | 139 | 12 (8.6) | 24.0 (20.0–37.0) | 72 (51.8) | 0 (0.0) | ELISA | 47.4 (37.4–56.4) | 2.0 | smear/culture positive | Adjusted linear regression estimate for continuous vitamin D and microbiologically confirmed TB: 3.65 (0.59–6.71) |
| Sudfeld et al., 2013 [32] | Tanzania (6.3690° S, 34.8888° E) | Prospective cohort study of HIV- positive patients initiating ART (TB exposure status not specified) | 1,092 | 50 (4.6) | 37.0 (32.0–42.8) | 752 (68.9) | 1,092 (100.0) | High-performance liquid chromatography | 73.5 (60.3–86.8) | Median 1.7 (IQR 0.7–2.8) | smear positive or chest radiograph | aHR for vitamin D < 50 nmol and microbiologically confirmed TB: 2.89 (1.31–7.41) |
| Talat et al., 2010 [33] | Pakistan (30.3753° N, 69.3451° E) | Prospective cohort study of household contacts of TB cases | 109 | 8 (7.3) | 20.0 (15.0–35.0) | 59 (54.1) | NA | ELISA | 23.5 (13.8–43.5) | 4.0 | smear positive or chest radiograph | aHR for 1-log decrement in continuous vitamin D and any TB: 5.1 (1.2–21.3) |
| Tenforde et al., 2017 [34] | Brazil (14.2350° S, 51.9253° W), Haiti (18.9712° N, 72.2852° W), India (20.5937° N, 78.9629° E), Malawi (13.2543° S, 34.3015° E), Peru (9.1900° S, 75.0152° W), South Africa (30.5595° S, 22.9375° E), Thailand (15.8700° N, 100.9925° E), US (37.0902° N, 95.7129° W), Zimbabwe (19.0154° S, 29.1549° E) |
Prospective case-cohort study of HIV-positive patients initiating ART (TB exposure status specified by history of TB disease) | 306 | 70 (22.9) | 35.0 (29.0–41.0) | 141 (46.1) | 306 (100.0) | Immunoassay | 80.0 (57.5–97.5) | 1.8 | ACTG criteria for confirmed, probable, or clinical TBe | aHR for vitamin D < 50 nmol and any TB: 3.66 (1.16–11.51) |
| Lima cohort study | Peru (9.1900° S, 75.0152° W) | Nested case-control study of HIV-negative household contacts of TB cases | 889 | 180 (20.2) | 24.0 (18.0–37.0) | 429 (48.3) | 0 (0.0) | Immunoassay | 54.5 (44.1–66.8) | 1.0 | Peru’s NTP criteria for TB diagnosis [15] | aOR for vitamin D < 50 nmol and any TB: 1.70 (0.84–3.46) aOR for vitamin D < 50 nmol and microbiologically confirmed TB: 1.78 (0.79–4.03) |
a Data as reported from original study either in total length or median/mean length.
b Three incident TB cases were HIV-negative; otherwise, HIV status is not available for remaining study participants.
c The South African National Tuberculosis Control Program: Practical Guidelines 2004. Available from: http://www.kznhealth.gov.za/chrp/documents/Guidelines/Guidelines%20National/Tuberculosis/SA%20TB%20Guidelines%202004.pdf
d Original study reported results from case-control analysis but baseline vitamin D levels for entire cohort of household contacts provided for individual participant data meta-analysis.
e Campbell TB, Smeaton LM, Kumarasamy N, Flanigan T, Klingman KL, Firnhaber C, et al. Efficacy and safety of three antiretroviral regimens for initial treatment of HIV-1: a randomized clinical trial in diverse multinational settings. PLoS Med. 2012;9(8): e1001290.
Abbreviations: ACTG, AIDS Clinical Trial Group; aHR, adjusted hazard ratio; aOR, adjusted OR; ART, antiretroviral therapy; CLIA, chemiluminescence immunoassay; ECLIA, electrochemiluminescence immunoassay; IPD, individual-patient data; IQR, interquartile range; NTP; OR, odds ratio; TB, tuberculosis