Table 3.

ORs and 95% CIs of each variable versus respiratory symptoms, respiratory diseases and subject classification

	Cougha	Sputumb	Any respiratory symptomc	Normal lung function without symptomd
Former smoker	0.8 (0.3, 2.6)	1.7 (0.5, 5.3)	1.0 (0.3, 2.8)	0.9 (0.3, 2.5)
Current smoker	4.3 (1.5, 12.0)**	5.0 (1.9, 13.3) **	4.9 (2.0, 11.8) ***	0.3 (0.1, 0.6) **
Former marijuana user	0.7 (0.3, 2.1)	0.6 (0.2, 1.7)	-	-
Current marijuana user	1.5 (0.5, 4.6)	1.3 (0.5, 3.5)	-	-
White	1.1 (0.4, 3.5)	-	-	-
On HAART	5.5 (1.4, 21.5)*	-	2.7 (0.9, 8.1)	-
Undetectable viral load	0.3 (0.1, 0.9)*	-	-	-

All the results were from multiple logistic regression analysis. Non-smoker group was the reference group in all the analysis. Reference group for other variables was non-marijuana user, non-white subject, subject who was not on HAART and subject who had detectable viral load respectively.

- The variable had p > 0.2 in uni-variate analysis and was not selected into multiple logistic regression analysis.

*p < 0.05 for the OR. ** p < 0.01 for the OR. *** p < 0.001 for the OR.

^a Subject having no cough was reference group. Nagelkerke R² = 0.252.

^bSubject having no sputum was reference group. Nagelkerke R² = 0.177.

^cSubject having no respiratory symptom at all was reference group. Respiratory symptom was defined as cough, sputum or breathlessness. Nagelkerke R² = 0.194.

^d Subject having either abnormal lung function or any respiratory symptom was reference group. Abnormal lung function was defined as either FEV₁/FVC < 70% or %FEV₁<80% or %FVC < 80%, either pre- or post-salbutamol test. Nagelkerke R² = 0.128.