Skip to main content
. 2022 Oct 15;19:115. doi: 10.1186/s12954-022-00697-3

Table 2.

Unadjusted and adjusted prevalence ratios (aPR) and 95% confidence intervals (95%CI) of recent transactional sex obtained from log-binomial regression (N = 412)

Covariates PR (95%CI) p value aPR (95%CI) p value
Age, in years 0.96 (0.94–0.98) 0.001 0.98 (0.97–0.99)  < 0.001
Sexual orientation
Heterosexual/straight 1.00 Ref 1.00 Ref
Gay or bisexual 6.43 (4.52–9.15)  < 0.001 5.30 (3.81–7.37)  < 0.001
Arrested, past 12 months
No 1.00 Ref 1.00 Ref
Yes 2.48 (1.50–4.12)  < 0.001 1.49 (0.87–2.56) 0.151
Currently homeless*
No 1.00 Ref
Yes 2.37 (1.07–5.23) 0.032
Food insecurity, past month
Went to sleep hungry < 1 × weekly 1.00 Ref 1.00 Ref
Went to sleep hungry ≥ 1 × weekly 2.50 (1.45–4.33) 0.001 1.77 (1.05–3.00) 0.032
Made, sold, or traded drugs, past 3 months
No 1.00 Ref 1.00 Ref
Yes 2.03 (0.91–4.52) 0.085 1.51 (0.70–3.27) 0.296
Injection drug use, past 3 months
None 1.00 Ref 1.00 Ref
Any 1.83 (1.02–3.31) 0.044 1.75 (1.02–3.01) 0.043

Multivariable log-binomial model adjusted for all covariates presented in the table. Cluster-robust standard errors were implemented at the recruitment zone level

Bolded values indicate statistically significant (p < 0.05) differences

*Covariate excluded from multivariable analysis due to considerable multicollinearity (VIF ≥ 2.5) with food insecurity