Demographic Characteristics.
After excluding 12 active TB at baseline, a total of 40,311 were finally included in the baseline survey(Fig. 1). At enrollment, median age was 50 years old (IQR, 37–61) and only 16.7% (N = 6,722) of participants were older than 65 years of age; About half of participants were female (54.7%; N = 22,067); 15,661 (39.0%) were overweight or obesity; About a quarter of participants reported ever smoking (21.7%, N = 8,745) or use of alcohol (N = 6,394, 15.9%); 3,382 (8.4%) were diabetes. Diabetes were more likely to have high blood glucose level (43.1% Versus.26.0%) (Table 1).
Table 1
Demographic characteristics 40311 participants, overall and by test status.
Characteristic | All | Non-diabetes | Diabetes |
(N, %) | (N, %) | (N, %) |
Sex | | | |
Female | 22067(54.7) | 20258(54.9) | 1809(53.5) |
Male | 18244(45.3) | 16671(45.1) | 1573(46.5) |
Median Age (IQR) | 50(37–61) | 49(35–60) | 59(50–67) |
BMI | | | |
<18.5 | 1880(4.7) | 1738(4.7) | 142(4.2) |
≥18.5–<24 | 22566(56.3) | 20818(56.6) | 1748(52.1) |
≥24–<28 | 12449(31.0) | 11312(30.8) | 1137(33.9) |
≥28 | 3212(8.0) | 2884(7.8) | 328(9.8) |
Smoking status | | | |
Never smoked | 31566(78.3) | 28966(78.4) | 2600(76.9) |
Ever smoked | 8745(21.7) | 7963(21.6) | 782(23.1) |
Alcohol drinking | | | |
No | 33917(84.1) | 31092(84.2) | 2825(83.5) |
Yes | 6394(15.9) | 5837(15.8) | 557(16.5) |
High blood glucose level | | | |
No | 22577(72.4) | 20953(74.0) | 1624(56.9) |
Yes | 8606(27.6) | 7375(26.0) | 1231(43.1) |
Blood lipid level (IQR) | 5.1(4.6–5.6) | 5.0(4.6–5.4) | 7.2(6.1–8.5) |
BMI = Body Mass Index |
†We grouped diabetic by whether the participant with diabetes. Participants who self-reporting a clinical diagnosis, through medical records and for a diagnosis of diabetes and with a positive fasting blood glucose test ≥ 7.0 mmol/L were regarded as diagnosed diabetes. |
Risk of development of tuberculosis.
During ten years of follow-up with a median of 8.6 y, 143 individuals progressed to tuberculosis (incidence rate, 41.4 per 100,000 person-years, 95% CI, 35.0-48.6), among whom 19 (13.3%) were diabetes (incidence rate, 65.5 per 100,000 person-years, 95% CI, 40.6-100.4), and the 7 (4.9%) were with FBG ≥ 7 mmol/L(incidence rate, 94.7 per 100,000 person-years, 95% CI, 41.4-187.3). Diagnosed diabetics with FBG ≥ 7.0 mmol/L showed the highest incidence rate compared with nondiabetics diagnosed diabetics with FBG < 7.0 mmol/L and undiagnosed diabetics (Ptrend=0.092)(Fig. 2). Among BMI > 24, the overall incidence rate of TB were 26.8 per 100,000 person-years (95% CI: 18.3–38.0) and 93.8 per 100,000 person-years (95% CI: 41.0-185.6) in FBG < 7.0 mmol/L and FBG ≥ 7.0 mmol/L, respectively. And in individuals with high blood lipid level, the overall incidence rate of TB among FBG < 7.0 mmol/L and FBG ≥ 7.0 mmol/L were 31.8 per 100,000 person-years (95% CI: 20.2–47.7) and 94.7 per 100,000 person-years (95% CI: 41.4-187.3) (Table 2).
Table 2
Tuberculosis incidence among participants according to baseline diabetic, BMI and blood lipid status&.
| Events | Observation Time | Rate per 100 Thousand | 95% Confidence Interval |
(n) | (Person-Years) | Person-years |
All Participants | 143 | 345,793 | 41.4 | 35.0-48.6 |
Diabetic status | | | | |
Nondiabetics | 124 | 316,802 | 39.1 | 32.7–46.5 |
Diabetics | 19 | 28,991 | 65.5 | 40.6-100.4 |
Diabetic status | | | | |
Nondiabetics | 124 | 316,802 | 39.1 | 32.7–46.5 |
Diagnosed diabetics with FBG < 7.0 | 6 | 12,338 | 48.6 | 19.7-101.1 |
Undiagonsed diabetics | 6 | 8,919 | 67.3 | 27.3-139.9 |
Diagnosed diabetics with FBG ≥ 7.0 | 7 | 7,734 | 90.5 | 40.0-179.0 |
BMI ≤ 24 | | | | |
FBG < 7.0 | 76 | 160,442 | 47.4 | 37.6–59.0 |
FBG ≥ 7.0 | 6 | 9,062 | 66.2 | 48.4-182.3 |
BMI > 24 | | | | |
FBG < 7.0 | 29 | 108,272 | 26.8 | 18.3–38.0 |
FBG ≥ 7.0 | 7 | 7,462 | 93.8 | 41.0-185.6 |
Normal blood lipid level | | | | |
FBG < 7.0 | 77 | 186,562 | 41.3 | 32.8–51.3 |
FBG ≥ 7.0 | 3 | 7,732 | 38.8 | 9.9-105.6 |
High blood lipid level | | | | |
FBG < 7.0 | 21 | 66,059 | 31.8 | 20.2–47.7 |
FBG ≥ 7.0 | 7 | 7,394 | 94.7 | 41.4-187.3 |
BMI = Body Mass Index, kg/m2, FBG = fasting blood-glucose, mmol/L |
& Diagnosed diabetics with FBG < 7.0 mmol/L was defined as participants self-reporting or through medical records for a clinical diagnosis or confirmation of diabetes with this < 7 mmol/L, which was same diagnosed diabetics with FBG ≥ 7.0 mmol/L except for this fasting blood glucose. These diabetics were thought to have poor control of blood sugar; undiagnosed diabetics was defined as participants had a positive fasting plasma glucose test result without evidence of a prior diabetes. High blood lipid level was defined as participants with serum total cholesterol > 5.72 mmol/L or serum triglyceride > 1.70 mmol/L. |
Factors contributing to the risk of tuberculosis.
In univariate analysis (Table 3), we found that factors associated with the progression to TB were male (HR = 2.12, 95%CI: 1.51–2.98, P < 0.0001), increasing age (HR = 1.01, 95%CI = 1.00-1.02, P = 0.019), smoking (HR = 1.50, 95%CI = 1.05–2.16, P = 0.026), diabetics (HR = 1.67, 95%CI = 1.03–2.71, P = 0.037).
Table 3
Univariable Cox Proportional hazard model analysis of the risk of active tuberculosis.
Characteristic | Active tuberculosis |
N incident/N total (% in subcategory) | | Hazard Ratio (95% CI), P-value |
Sex | | | |
Female | 52/22067(0.2%) | | Reference |
Male | 91/18244(0.5%) | | 2.12(1.51–2.98), P < 0.0001 |
Age (Continuous) | | | 1.01(1.00-1.02), P = 0.019 |
BMI | | | |
<18.5 | 5/1880(0.3%) | | Reference |
≥18.5–<24 | 95/22566(0.4%) | | 1.58(0.64–3.89), P = 0.317 |
≥24–<28 | 35/12449(0.3%) | | 1.06(0.41–2.70), P = 0.910 |
≥28 | 8/3212(0.2%) | | 0.94(0.31–2.86), P = 0.908 |
Smoking status | | | |
Never smoked | 101/31566(0.3%) | | Reference |
Ever smoked | 42/8745(0.5%) | | 1.50(1.05–2.16), P = 0.026 |
Alcohol drinking | | | |
No | 119/33917(0.4%) | | Reference |
Yes | 24/6394(0.4%) | | 1.08(0.70–1.67), P = 0.736 |
Diabetic status | | | |
Nondiabetics | 124/36929(0.3%) | | Reference |
Diabetics | 19/3382(0.6%) | | 1.67(1.03–2.71), P = 0.037 |
Diabetic status | | | |
Nondiabetics | 124/36929(0.3%) | | Reference |
Diagnosed diabetics with FBG < 7.0 | 6/1434(0.4%) | | 1.24(0.55–2.81), P = 0.607 |
Undiagnosed diabetics | 6/1045(0.6%) | | 1.71(0.76–3.89), P = 0.198 |
Diagnosed diabetics with FBG ≥ 7.0 | 7/903(0.8%) | | 2.31(1.08–4.94), P = 0.032 |
Blood glucose level | | | |
FBG < 7.0 | 105/31436(0.3%) | | Reference |
FBG ≥ 7.0 | 13/1948(0.7%) | | 2.00(1.12–3.56), P = 0.018 |
High blood lipid level | | | |
No | 80/22577(0.4%) | | Reference |
Yes | 28/8606(0.3%) | | 0.92(0.60–1.42), P = 0.709 |
BMI = Body Mass Index, kg/m2, FBG = fasting blood-glucose, mmol/L |
Table 4. Multivariable Cox Proportional hazard model analysis of risk of active tuberculosis. | | |
Characteristic | Active tuberculosis |
Hazard Ratio (95% CI), P-value | | Hazard Ratio (95% CI), P-value | | Hazard Ratio (95% CI), P-value | | Hazard Ratio (95% CI), P-value |
Sex | | | | | | | |
Female | Reference | | Reference | | Reference | | Reference |
Male | 2.26(1.53–3.33), P < 0.0001 | | 2.27(1.54–3.35), P < 0.0001 | | 2.00(1.30–3.09), P = 0.002 | | 2.00(1.30–3.09), P = 0.002 |
Age (Continuous) | 1.01(1.00-1.02), P = 0.034 | | 1.01(1.00-1.02), P = 0.033 | | 1.01(1.00-1.02), P = 0.110 | | 1.01(1.00-1.02), P = 0.095 |
BMI | | | | | | | |
<18.5 | Reference | | Reference | | Reference | | Reference |
≥18.5–<24 | 1.45(0.59–3.56), P = 0.423 | | 1.44(0.59–3.55), P = 0.425 | | 2.95(0.72-12.00), P = 0.131 | | 2.99(0.73–12.17), P = 0.126 |
≥24–<28 | 0.89(0.35–2.27), P = 0.802 | | 0.89(0.35–2.26), P = 0.799 | | 1.73(0.41–7.26), P = 0.454 | | 1.75(0.42–7.34), P = 0.445 |
≥28 | 0.82(0.27–2.50), P = 0.724 | | 0.82(0.27–2.49), P = 0.720 | | 1.64(0.34–7.91), P = 0.536 | | 1.66(0.34–7.98), P = 0.528 |
Smoking status | | | | | | | |
Never smoked | Reference | | Reference | | Reference | | Reference |
Ever smoked | 1.03(0.68–1.56), P = 0.903 | | 1.03(0.68–1.56), P = 0.909 | | 1.18(0.75–1.87), P = 0.480 | | 1.18(0.75–1.87), P = 0.479 |
Alcohol drinking | | | | | | | |
No | Reference | | Reference | | Reference | | Reference |
Yes | 0.78(0.49–1.25), P = 0.305 | | 0.78(0.48–1.24), P = 0.288 | | 0.72(0.42–1.23), P = 0.228 | | 0.71(0.42–1.22), P = 0.211 |
Diabetic status | | | | | | | |
Nondiabetics | Reference | | | | | | |
Diabetics | 1.51(0.92–2.47), P = 0.104 | | | | | | |
Diabetic status | | | | | | | |
Nondiabetics | | | Reference | | | | |
Diagnosed diabetics with FBG < 7.0 | | | 1.07(0.47–2.45), P = 0.871 | | | | |
Undiagnosed diabetics | | | 1.67(0.73–3.81), P = 0.222 | | | | |
Diagnosed diabetics with FBG ≥ 7.0 | | | 2.04(0.94–4.40), P = 0.070 | | | | |
Blood glucose level | | | | | | | |
FBG < 7.0 | | | | | Reference | | |
FBG ≥ 7.0 | | | | | 1.91(1.07–3.42), P = 0.030 | | |
Blood glucose level(Continuious) | | | | | | | 1.10(1.02–1.20), P = 0.020 |
BMI = Body Mass Index, kg/m2, FBG = fasting blood-glucose, mmol/L |
In multivariable analysis stratifying diabetic status as nondiabetics, diagnosed diabetics with FBG < 7.0 mmol/L, undiagnosed diabetics, and diagnosed diabetics with FBG ≥ 7.0, there was a nonsignificant trend towards a risk association with diagnosed diabetics with FBG ≥ 7.0 mmol/L but it was near the borderline (P = 0.070). When grouping blood glucose level with the cutoff of 7.0 mmol/L, compared with FBG < 7.0 mmol/L, participants with FBG ≥ 7.0 mmol/L had 91% increase in tuberculosis hazard (HR = 1.91, 95%CI: 1.07–3.42, P = 0.030). With a 1-unit (1 mmol/L) increase in blood glucose, the hazard of tuberculosis increased by 10% (95% CI: 2–20%, P = 0.020)(Table 4).
In the stratified analysis, in participants with BMI ≤ 24 kg/m2, there was no statistically significant with diabetics or blood glucose and tuberculosis (P = 0.940, 0.610, 0.690 and 0.505, respectively). In participants with BMI > 24 kg/m2, diagnosed diabetics with FBG ≥ 7.0 mmol/L showed nearly three-fold increased risk of active TB (HR = 3.78, 95%CI: 1.32–10.79, P = 0.013), and FBG ≥ 7.0 mmol/L was associated with more than three-fold higher risk of active TB(HR = 3.16, 95%CI:1.37–7.28, P = 0.007)(Table 5) (Fig. 3).
Table 5
Multivariable Cox Proportional hazard model analysis of risk of active tuberculosis by MBI status
Characteristic | Active tuberculosis |
BMI ≤ 24 | | BMI > 24 |
Hazard Ratio (95% CI), P-value | Hazard Ratio (95% CI), P-value | | Hazard Ratio (95% CI), P-value | Hazard Ratio (95% CI), P-value |
Sex | | | | | |
Female | Reference | Reference | | Reference | Reference |
Male | 2.04(1.29–3.21), P = 0.002 | 1.84(1.11–3.07), P = 0.019 | | 3.30(1.50–7.23), P = 0.003 | 2.64(1.14–6.11), P = 0.023 |
Age (Continuous) | 1.01(1.00-1.02), P = 0.175 | 1.01(0.99–1.02), P = 0.370 | | 1.02(1.00-1.04), P = 0.038 | 1.02(1.00-1.05), P = 0.088 |
Smoking status | | | | | |
Never smoked | Reference | Reference | | Reference | Reference |
Ever smoked | 1.06(0.63–1.78), P = 0.836 | 1.17(0.66–2.07), P = 0.593 | | 0.99(0.49–1.99), P = 0.979 | 1.25(0.57–2.73), P = 0.574 |
Alcohol drinking | | | | | |
No | Reference | Reference | | Reference | Reference |
Yes | 0.57(0.30–1.08), P = 0.086 | 0.61(0.30–1.23), P = 0.167 | | 1.24(0.60–2.56), P = 0.554 | 0.95(0.41–2.21), P = 0.909 |
Diabetic status | | | | | |
Nondiabetics | Reference | | | Reference | |
Diagnosed diabetics with FBG < 7.0 | 1.04(0.38–2.86), P = 0.940 | | | 1.14(0.27–4.81), P = 0.858 | |
Undiagnosed diabetics | 1.35(0.43–4.29), P = 0.610 | | | 2.35(0.72–7.69), P = 0.157 | |
Diagnosed diabetics with FBG ≥ 7.0 | 1.27(0.40–4.04), P = 0.690 | | | 3.78(1.32–10.79), P = 0.013 | |
Blood glucose level | | | | | |
FBG < 7.0 | | Reference | | | Reference |
FBG ≥ 7.0 | | 1.33(0.57–3.09), P = 0.505 | | | 3.16(1.37–7.28), P = 0.007 |
BMI = Body Mass Index, kg/m2, FBG = fasting blood-glucose, mmol/L |
In another stratified analysis by high blood lipid level, in normal high blood lipid level, there was no statistic difference between the diabetics and TB, however, among high blood lipid levels undiagnosed diabetics was related to increase the high risk of TB (HR = 3.04, 95%CI: 1.03–8.95, P = 0.044) and FBG ≥ 7.0 mmol/L nearly than two-fold higher risk of TB (HR = 2.66, 95%CI: 1.13–6.30, P = 0.026)(Table 6) (Fig. 3). In the linear dose-response analysis, the hazard of TB increased with FBG (with a 1-unit (1-mmol/L) increase in FBG, the hazard of TB increased by 15% (95% CI, 3–29%)(Table 7)(Fig. 3).
Table 6
Multivariable Cox Proportional hazard model analysis of risk of active tuberculosis by blood lipid status.
Characteristic | Active tuberculosis |
Normal high blood lipid level | | High blood lipid level$ |
Hazard Ratio (95% CI), P-value | Hazard Ratio (95% CI), P-value | | Hazard Ratio (95% CI), P-value | Hazard Ratio (95% CI), P-value |
Sex | | | | | |
Female | Reference | Reference | | Reference | Reference |
Male | 1.89(1.12–3.18), P = 0.017 | 1.89(1.12–3.19), P = 0.017 | | 2.03(0.84–4.94), P = 0.117 | 2.05(0.84–4.97), P = 0.114 |
Age (Continuous) | 1.00(0.99–1.02), P = 0.819 | 1.00(0.99–1.02), P = 0.728 | | 1.04(1.01–1.07), P = 0.011 | 1.03(1.01–1.07), P = 0.009 |
Smoking status | | | | | |
Never smoked | Reference | Reference | | Reference | Reference |
Ever smoked | 1.10(0.62–1.94), P = 0.743 | 1.10(0.62–1.93), P = 0.748 | | 1.42(0.56–3.63), P = 0.459 | 1.42(0.56–3.62), P = 0.464 |
Alcohol drinking | | | | | |
No | Reference | Reference | | Reference | Reference |
Yes | 1.07(0.57–1.98), P = 0.843 | 1.06(0.57–1.97), P = 0.863 | | 0.25(0.06–1.09), P = 0.064 | 0.25(0.06–1.09), P = 0.064 |
Diabetic status | | | | | |
Nondiabetics | Reference | | | Reference | |
Diagnosed diabetics with FBG < 7.0 | 1.43(0.52–3.96), P = 0.491 | | | 1.47(0.34–6.41), P = 0.609 | |
Undiagnosed diabetics | 0.65(0.09–4.70), P = 0.670 | | | 3.04(1.03–8.95), P = 0.044 | |
Diagnosed diabetics with FBG ≥ 7.0 | 1.30(0.32–5.36), P = 0.716 | | | 2.46(0.72–8.40), P = 0.150 | |
Blood glucose level | | | | | |
FBG < 7.0 | | Reference | | | Reference |
FBG ≥ 7.0 | | 0.95(0.30–3.04), P = 0.934 | | | 2.66(1.13–6.30), P = 0.026 |
BMI = Body Mass Index, kg/m2, FBG = fasting blood-glucose, mmol/L; $ High blood lipid level was defined as participants with serum total cholesterol > 5.72 mmol/L or serum triglyceride > 1.70 mmol/L. |
Table 7
Multivariable Cox Proportional hazard model analysis of risk of active tuberculosis by blood lipid status.
Characteristic | Active tuberculosis |
Normal high blood lipid level | | High blood lipid level‡ |
Hazard Ratio (95% CI), P-value | | Hazard Ratio (95% CI), P-value |
Sex | | | |
Female | Reference | | Reference |
Male | 1.87(1.11–3.16), P = 0.019 | | 1.74(0.72–4.20), P = 0.217 |
Age (Continuous) | 1.00(0.99–1.02), P = 0.532 | | 1.04(1.02–1.07), P = 0.002 |
Smoking status | | | |
Never smoked | Reference | | Reference |
Ever smoked | 1.10(0.62–1.94), P = 0.744 | | 1.19(0.47–3.02), P = 0.712 |
Alcohol drinking | | | |
No | Reference | | |
Yes | 1.09(0.59–2.04), P = 0.781 | | |
Blood glucose level (Continuous) | 0.84(0.64–1.10), P = 0.202 | | 1.15(1.03–1.29), P = 0.011 |
‡High blood lipid level was defined as participants with serum total cholesterol > 5.72 mmol/L or serum triglyceride > 1.70 mmol/L. |
Another factor that increased the hazard of TB were male sex and increasing age. Male sex had two to three-fold higher risk of TB and every unit increase in age was associated with 1–4% increase in TB risk.