Skip to main content
. Author manuscript; available in PMC: 2020 Aug 1.
Published in final edited form as: AIDS Behav. 2019 Aug;23(8):2025–2036. doi: 10.1007/s10461-018-2301-4

Appendix B.

Sensitivity analysis comparing direct and indirect effects of HIV-related stigma on viral suppression among two samples of African-American women living with HIV enrolled in the Unity Study, differing by viral load collection period

Primary sample (8-week viral load) Sensitivity sample (24-week viral load)
N = 100 N = 184

β (95% CI) β (95% CI)
Primary model*

Total effect (difference in log-odds) −0.069 (−0.117, −0.021) −0.052 (−0.084, −0.020)
Direct effect −0.085 (−0.145, −0.025) −0.043 (−0.082, −0.004)
Indirect effects
 Stigma > depression > nonadherence > viral suppression −0.006 (−0.045, 0.029) −0.003 (−0.010, 0.002)
 Stigma > depression > viral suppression 0.010 (−0.046, 0.054) 0.010 (−0.037, 0.014)
 Stigma > nonadherence > viral suppression 0.002 (−0.087, 0.051) 0.001 (−0.012, 0.010)

Secondary model**
Total effect (difference in log-odds) −0.096 (−0.016, −0.032) −0.056 (−0.089, −0.022)
Direct effect −0.238 (−0.385, −0.091) −0.052 (−0.092, −0.011)
Indirect effects
 Stigma > depression > nonadherence > viral suppression −0.010 (−0.072, 0.054) −0.002 (−0.011, 0.003)
 Stigma > depression > viral suppression 0.064 (−0.140, 0.206) 0.006 (−0.034, 0.021)
 Stigma > nonadherence > viral suppression 0.009 (−0.105, 0.107) 0.001 (−0.012, 0.011)
*

Adjusted for treatment, site, age, education, time living with HIV

**

Adjusted for treatment, site, age, education, time living with HIV, alcohol use severity