Selected socio-economic and behavioral characteristics of the resident population by HIV infection status.

Table 1 note: Row percentages shown, separately for men and women, respectively; percentages are weighted and frequencies are unweighted. Weights are propensity score weights based upon probability of participation in HIV testing. Statistically significant group differences (at p<.05) are highlighted in bold-print. Odds ratios are adjusted for age and age2 (due to a non-linear association between age and HIV status).
*Data are from the resident population who were eligible for testing on 01 June 2003 and not included in the non-resident sample selected for participation in HIV testing at that time. N = 11,677.
**The odds ratios given for the model regressing HIV test result on 10-year age group by sex has only one independent variable (i.e. unlike the other models shown in this table, single year of age and age2 not included as additional variables.).
†Women only up to age 49 were eligible for HIV testing; men up to age 54 were eligible.
††This category mainly comprised of those with a regular non-marital partner: a small minority (n = 407) reported a casual partner. HIV prevalence was lower among those reporting a casual partner (14.0%) than among those with a regular non-marital partner (37.1%), and closer to the level of those with a marital partner (18.8%).