Table 2.
Crude odds ratio | 95% CI | p value | Adjusted odds ratioa | 95% CI | p value | |
---|---|---|---|---|---|---|
Model 1: any violence victimization | ||||||
No violence (n = 72) | 1 | 1 | ||||
Any physical, psychological, or forced sex (n = 200) | 1.04 | (0.59, 1.82) | 0.89 | 1.09 | (0.61, 1.95) | 0.77 |
Model 2: frequency of any victimizationb | ||||||
No violence (scores of 0) (n = 72) | 1 | 1 | ||||
Single act of violence (scores of 1) (n = 31) | 0.72 | (0.29, 1.80) | 0.49 | 0.70 | (0.28, 1.79) | 0.46 |
Moderate frequency (scores of 2–11) (n = 150) | 0.99 | (0.55, 1.78) | 0.97 | 1.06 | (0.58, 1.96) | 0.84 |
High frequency (scores of 12–42) (n = 19) | 2.43 | (0.87, 6.81) | 0.09 | 3.58 | (1.14, 11.27) | 0.03 |
Model 3: polyvictimization | ||||||
No violence (n = 72) | 1 | 1 | ||||
1 type of violence (n = 93) | 0.93 | (0.49, 1.77) | 0.82 | 0.98 | (0.50, 1.90) | 0.94 |
2 or 3 types of violence (n = 107) | 1.14 | (0.62, 2.12) | 0.67 | 1.21 | (0.63, 2.29) | 0.57 |
Model 4: type of victimization | ||||||
Severity-times-frequency of physical violenceb | ||||||
No physical violence (scores of 0) (n = 144) | 1 | 1 | ||||
Single act of physical violence (scores of 1) (n = 34) | 1.16 | (0.54, 2.52) | 0.70 | 1.08 | (0.47, 2.47) | 0.85 |
Moderate severity-times-frequency (scores of 2–7) (n = 56) | 1.04 | (0.55, 1.99) | 0.90 | 0.93 | (0.45, 1.93) | 0.84 |
High severity-times-frequency (scores of 8–42) (n = 38) | 1.52 | (0.74, 3.14) | 0.26 | 1.18 | (0.49, 2.85) | 0.71 |
Frequency of psychological abuseb | ||||||
No psychological abuse (scores of 0) (n = 96) | 1 | 1 | ||||
Single act of psychological abuse (scores of 1) (n = 31) | 0.83 | (0.35, 1.96) | 0.67 | 0.84 | (0.34, 2.04) | 0.70 |
Moderate frequency (scores of 2–5) (n = 111) | 0.80 | (0.45, 1.42) | 0.46 | 0.81 | (0.42, 1.57) | 0.54 |
High frequency (scores of 6–18) (n = 34) | 2.49 | (1.12, 5.53) | 0.03 | 3.32 | (1.26, 8.70) | 0.01 |
Forced sex | ||||||
No forced sex (n = 258) | 1 | 1 | ||||
Any forced sex (n = 14) | 1.31 | (0.44, 3.88) | 0.63 | 1.19 | (0.35, 4.01) | 0.78 |
Model 5: perpetrator of physical violence | ||||||
Any vs. none from a parent/caregiver (n = 42 vs. 230) | 1.07 | (0.54, 2.10) | 0.85 | 0.95 | (0.46, 1.96) | 0.89 |
Any vs. none from another family member (n = 39 vs.233) | 2.27 | (1.15, 4.51) | 0.02 | 2.18 | (1.05, 4.54) | 0.04 |
Any vs. none from a romantic partner (n = 16 vs. 256) | 0.56 | (0.17, 1.77) | 0.32 | 0.77 | (0.21, 2.78) | 0.69 |
Any vs. none from a friend/peer (n = 44 vs. 228) | 2.40 | (1.25, 4.62) | 0.01 | 2.14 | (1.05, 4.36) | 0.04 |
Model 6: Perpetrator of psychological abuse | ||||||
Any vs. none from a parent/caregiver (n = 39 vs. 233) | 0.55 | (0.25, 1.18) | 0.12 | 0.48 | (0.21, 1.09) | 0.08 |
Any vs. none from another family member (n = 68 vs. 204) | 2.09 | (1.19, 3.67) | 0.01 | 2.50 | (1.37, 4.54) | 0.003 |
Any vs. none from a romantic partner (n = 24 vs. 248) | 0.85 | (0.35, 2.06) | 0.72 | 1.14 | (0.42, 3.10) | 0.80 |
Any vs. none from a friend/peer (n = 99 vs. 173) | 1.19 | (0.72, 1.99) | 0.50 | 1.18 | (0.68, 2.05) | 0.55 |
aAdjusted for age, sex, and study clinic (a priori), and the following: mode of HIV acquisition (all models), time on ART treatment (all models), orphan hood (all models except Model 3). Models 4–6 adjusted for the other violence variables in addition to covariates. p values are Wald tests
bFrequency of any victimization scores were generated by summing each act’s frequency (never-0, once-1, a few times-2, many times-3) across 14 acts of violence. Severity-times-frequency of physical violence scores were generated by multiplying the act’s severity level (moderate-1, severe-2) by its frequency (never-0, once-1, a few times-2, many times-3), and summing across 7 acts of physical violence. Frequency of psychological abuse scores were generated by summing each act’s frequency (never-0, once-1, a few times-2, many times-3) across 6 acts of psychological abuse. Categories for violence scores were determined based on locally-weighted scatterplot smoothing (lowess) plots for the association between the variable and viral load failure