Skip to main content
. 2014 May 2;9(5):e96433. doi: 10.1371/journal.pone.0096433

Table 4. Smoking, drinking, diabetes, BMI<18.5 kg/m2 attributable new TB incident cases among men and women (18+ years) in 14 High-Burden Countries (HBCs), 2012.

Sex New cases in HBC Smoking*, % Drinking, % Diabetes ‡, % BMI<18.5 kg/m2 , %
Men 3,893,910 629,304 16 569,304 15 508,066 13 652,534 17
Women 2,045,090 102,076 5 65,111 3 221,011 11 303,614 15
M+W 5,939,000 731,380 12 634,415 11 729,077 12 956,148 16
*

Sources of prevalence for daily smoking were taken from Global Adult Tobacco Survey (GATS) [17], Demographic and Health Survey (DHS) [16], World Health Survey (WHS) [21]; RR men = 1·71 (99% CI:1·20–2·70) and RR women = 2·44 (99% CI:1·54–4·17) were taken from our continuous dose-response meta-regression model.

Prevalences for daily alcohol drinking (men: 40+ grams/day; women: 20+ grams/day) were taken from WHS [21] and where data on large nationally representative surveys were available [16], [18]; RR men = 1·84 (99% CI:1·32–2·83) and RR women = 1·61 (99% CI:1·11–2·49) were taken from our continuous dose-response meta-regression model.

Prevalence on diabetes were taken from International Diabetes Federation (2011) [19]; RR men = 2.87 (99%CI:1·44–5.69) and RR women = 2·68 (99% CI: 1·35–5.32) were taken from our D-L meta analysis.

Prevalence on low BMI (<18·5 kg/m2) were taken from DHS [16]; RR men = 2·12 (99% CI: 1·41–3.31) and RR women = 2·00 (99% CI: 1·43–3·24) were taken our continuous dose-response meta-regression model.