Table 4. Associations between drought and related characteristics and HIV prevalence in 15-to-24- and 25-to-59-year-olds, Lesotho 2016–2017.
| Characteristic | Ages 15–24 years N = 2,341 aOR (95% CI)a |
Ages 25–59 years N = 5,778 aOR (95% CI)a |
|---|---|---|
| Gender–drought interactionb | ||
| Males living in drought area | 0.35 (0.17–0.72) p = 0.006e | 0.72 (0.51–1.01) p = 0.06 |
| Females living in drought area | 1.80 (0.96–3.39) p = 0.07 | 1.18 (0.85–1.65) p = 0.31 |
| Age (per 1-year increase) | 1.16 (1.06–1.27) p = 0.002e | 1.02 (1.01–1.02) p < 0.001e |
| Lives in rural community | 0.64 (0.44–0.93) p = 0.02e | 0.80 (0.68–0.94) p = 0.007e |
| Household is impoverished | 0.68 (0.44–1.05) p = 0.08 | 0.96 (0.78–1.18) p = 0.70 |
| Household is food insecure | 1.24 (1.02–1.51) p = 0.03 | 1.22 (1.13–1.33) p < 0.001e |
| Lived outside Lesotho in past year | 2.06 (1.25–3.40) p = 0.006e | 1.12 (0.88–1.42) p = 0.34 |
| Currently or ever married | 1.74 (1.17–2.61) p = 0.009e | 1.48 (1.18–1.86) p = 0.002e |
| Secondary or higher educationc | 0.66 (0.47–0.92) p = 0.02e | 0.54 (0.46–0.62) p < 0.001e |
| Intergenerational sex in past yeard | 1.54 (0.97–2.44) p < 0.001e | 1.37 (1.10–1.71) p = 0.007e |
| Transactional or commercial sex in past year | 1.27 (0.63–2.56) p = 0.50 | 1.46 (1.10–1.93) p = 0.01e |
| Condom use at last sex | 1.49 (0.99–2.24) p = 0.05 | 2.42 (2.13–2.76) p < 0.001e |
All data are survey weighted.
a aOR calculated by logistic regression with Jackknife replicate variance estimates using weighted data.
b Reference group is anyone of either gender living in non-drought-affected areas.
c Defined as highest level attended.
d Defined as sex with a partner older by 10 or more years.
e Significant using Bonferroni-corrected threshold for two models of 0.025.
Abbreviations: aOR, adjusted odds ratio; CI, confidence interval.