Skip to main content
. 2022 Aug 25;25(8):e25928. doi: 10.1002/jia2.25928

Table 3.

Multivariate multivariable associations between food security and HIV risk behaviours overall and stratified by adolescent motherhood, and both adolescent motherhood and HIV status

Multiple sexual partners Transactional sex Age‐disparate sex Condomless sex
AOR (95% CI) p‐value AOR (95% CI) p‐value AOR (95% CI) p‐value AOR (95% CI) p‐value
Food security
Overall 0.70 (0.52–0.93) 0.015 0.19 (0.12–0.31) <0.001 0.76 (0.56–1.04) 0.084 0.71 (0.53–0.94) 0.016
Among non‐mothers 0.45 (0.26–0.78) 0.073 0.32 (0.13–0.82) 0.23 1.60 (0.74–3.50) 0.037 0.63 (0.39–1.03) 0.597
Among adolescent mothers 0.81 (0.58–1.14) 0.17 (0.10–0.28) 0.66 (0.47–0.92) 0.74 (0.53–1.03)
Among HIV‐uninfected non‐mothers 0.77 (0.25–2.38) 0.079 0.18 (0.03–1.11) 0.68 3.64 (0.82–16.05) 0.258 0.84 (0.36–1.97) 0.403
Among non‐mothers living with HIV 0.36 (0.19–0.68) 0.40 (0.13–1.20) 1.08 (0.43–2.71) 0.54 (0.30–0.99)
Among HIV‐uninfected adolescent mothers 0.67 (0.46–1.00) 0.11 (0.06–0.21) 0.70 (0.46–1.06) 0.72 (0.48–1.07)
Among adolescent mothers living with HIV 1.21 (0.65–2.27) 0.42 (0.17–1.09) 0.61 (0.35–1.05) 0.79 (0.44–1.43)
Sex on substances Alcohol use Not in education/employment
AOR (95% CI) p‐value AOR (95% CI) p‐value AOR (95% CI) p‐value
Food security
Overall 0.57 (0.37–0.87) 0.010 0.65 (0.42–1.00) 0.048 0.53 (0.41–0.71) <0.001
Among non‐mothers 0.82 (0.32–2.08) 0.375 1.05 (0.54–2.07) 0.058 0.48 (0.29–0.77) 0.585
Among adolescent mothers 0.51 (0.32–0.82) 0.45 (0.25–0.79) 0.56 (0.40–0.78)
Among HIV‐uninfected non‐mothers 0.98 (0.20–4.86) 0.773 2.32 (0.55–9.70) 0.419 0.35 (0.15–0.83) 0.983
Among non‐mothers living with HIV 0.73 (0.22–2.41) 0.75 (0.35–1.61) 0.55 (0.30–1.00)
Among HIV‐uninfected adolescent mothers 0.51 (0.28–0.91) 0.51 (0.24–1.07) 0.51 (0.35–0.73)
Among adolescent mothers living with HIV 0.53 (0.24–1.18) 0.37 (0.15–0.93) 0.78 (0.40–1.51)

Note: n = 1638. For sub‐group analyses, we report Wald test p‐values for the interaction term. Multivariable models adjust for participant socio‐demographic characteristics: HIV status; age; relationship status; parental monitoring; rural/urban household location; informal housing type; number of people living in household; maternal orphanhood; and paternal orphanhood. Clustering of multiple outcomes within individuals is accounted for using the GEE method.

Abbreviations: AOR, adjusted odds ratio; CI, confidence interval; HIV, human immunodeficiency virus; OR, odds ratio.