Table 3.

Relationship between HIV and socioeconomic status from multivariate analysis

Variable	Western Cape Province, aOR (95% CI)	Free State Province, aOR (95% CI)
Age	1.03 (0.91 to 1.17)	1.07 (1.02 to 1.12)
Setting
Urban informal	1.0	1.0
Urban formal	0.12 (0.01 to 1.28)	NA
Rural formal	NA	1.03 (0.93 to 1.14)
Marital status
Never married	1.0	1.0
Currently married	0.77 (0.33 to 1.80)	0.98 (0.20 to 4.90)
Divorced	0.76 (0.00 to 719.64)	0.44 (0.14 to 1.40)
Separated	0.51 (0.01 to 23.71)	2.48 (0.21 to 29.71)
Widowed	1.57 (0.03 to 90.49)	4.22 (0.51 to 34.81)
Years of formal education	0.90 (0.85 to 0.96)	0.99 (0.89 to 1.11)
Employment (employed vs unemployed)	1.54 (0.84 to 2.84)	0.96 (0.71 to 1.30)
Wealth quintile*
Lowest quintile (poorest)	1.0	1.0
Second lowest quintile	1.39 (0.29 to 6.56)	0.88 (0.67 to 1.16)
Middle quintile	0.90 (0.08 to 10.57)	0.91 (0.72 to 1.16)
Second highest quintile	0.77 (0.02 to 29.80)	0.93 (0.41 to 2.08)
Highest quintile (wealthiest)	0.62 (0.00 to 250.83)	0.76 (0.16 to 3.71)

*Household wealth quintiles were generated separately for each province using data for the specified province only. Separate regression models were run for each province using household wealth quintiles generated separately for each province. Point estimates and their CIs were adjusted for the two-stage stratified cluster random sampling and design weights.

aOR, adjusted OR; NA, not applicable.