Skip to main content
. 2021 Oct 1;16(10):e0258139. doi: 10.1371/journal.pone.0258139

Table 3. Multinomial models of HIV specific predictors of RLS status in men with HIV.

HIV variable Model A Model B Model C Model D
OR 95% CI OR 95% CI OR 95% CI OR 95% CI
Definite RLS (N = 47)
Disease duration (years) 1.05 1.01–1.10 * 1.05 0.99–1.12 1.06 1.00–1.13 * 1.08 1.02–1.16 *
CD4 count > 500 1.64 0.73–3.67 1.61 0.71–3.63 1.94 0.83–4.54 1.53 0.65–3.60
Undetectable viral load 1.70 0.33–3.95 1.61 0.69–3.78 1.59 0.67–3.77 1.75 0.69–4.45
Indeterminate RLS (N = 90)
Disease duration (yrs) 1.00 0.97–1.04 1.03 0.99–1.07 1.03 0.99–1.07 1.03 0.99–1.08
CD4 count > 500 0.76 0.45–1.28 0.69 0.41–1.19 0.70 0.43–1.20 0.64 0.36–1.15
Undetectable viral load 0.96 0.56–1.65 0.99 0.57–1.73 1.00 0.57–1.74 0.93 0.50–1.72

Multinomial logistic regression results (referenced to “no RLS”; n = 367).

Significance level is labeled as * p < 0.05, ** p < 0.01.

Model A: Unadjusted; Model B: demographics (age, race).

Model C: Model B + Comorbidities (anemia, depression).

Model D: Model C + medications (antidepressants).