Skip to main content

Table 5 Crude and adjusted odds ratios for determinants of childhood PTB in different logistic models for indoor air environment variables

From: Association of childhood pulmonary tuberculosis with exposure to indoor air pollution: a case control study

  Univariate Multivariate models
    Primary cooking fuel Secondhand smoke Dampness PM10 NO2
Variable OR (95% CI) aOR (95% CI) aOR (95% CI) aOR (95% CI) aOR (95% CI) aOR (95% CI)
Primary cooking fuel Clean ref ref     
  Mix + dirty 2.6 (1.1–6.4) 1.5 (0.5–4.6)     
#Secondhand smoke No ref   ref    
  Yes 1.7 (0.98–2.8)   1.2 (0.6–2.3)    
Visible dampness No room ref    ref   
  Some rooms 1.8 (1.01–3.1)    2.4 (1.1–5.0)   
PM10 concentrations Median + below ref     Ref  
  Above median 1.4 (0.8–2.3)     0.9 (0.5–1.8)  
NO2 concentrations Median + below ref      ref
  Above median 0.7 (0.4–1.2)      0.4 (0.2–0.8)
Age ≤2 ref ref ref ref ref ref
  > 2–14 0.3 (0.1–0.7) 0.2 (0.1–0.6) 0.2 (0.1–0.6) 0.2 (0.1–0.6) 0.2 (0.1–0.6) 0.2 (0.1–0.5)
Sex Female ref ref ref ref ref ref
  Male 1.3 (0.8–2.2) 1.2 (0.6–2.3) 1.2 (0.6–2.4) 1.3 (0.7–2.5) 1.2 (0.6–2.4) 1.3 (0.7–2.4)
**HIV + ve status No ref ref ref ref ref ref
  Yes 22 (7.4–63) 26 (8.4–83) 27 (8.6–84) 32 (9.7–104) 27 (8.7–87) 29 (9.0–93)
Household TB contact No ref ref ref ref ref ref
  Yes 5.1 (2.9–9.0) 6.2 (3.1–12) 6.3 (3.2–12) 6.2 (3.1–12) 6.5 (3.3–13) 7.8 (3.8–16)
Pseudo R2    0.31 032 0.33 0.31 0.33
  1. All multivariate models were adjusted for age and sex. All models included crowding, HIV positive status of the child, presence of household TB contact and caregiver education
  2. #Secondhand smoke assessed as presence of a smoker in the household
  3. **Univariate model for HIV status results was based on those that had tested (n = 143). The missing observations were substituted as negative before inclusion to multivariate models
  4. OR odds ratio
  5. aOR adjusted odds ratio