Table 4.

Multivariable models for geometric mean viral load (VL) at two overlapping intervals of early HIV-1 infection

Factors in model	Analyses of VLa in the 2–36 months interval					Analyses of VLa in the 3–24 months intervalb
	n	Δβ	SE	Adjusted P	R 2	n	Δβ	SE	Adjusted P	R 2
Age (if >40)c	80	−0.05	0.10	0.585	0.000	75	−0.11	0.10	0.276	0.002
Southern Africa (Zambia)	206	0.35	0.07	<0.0001	0.038	199	0.36	0.08	<0.0001	0.040
HLA-B*18	34	0.35	0.14	0.015	0.010	34	0.35	0.15	0.018	0.010
HLA-B*45	83	0.29	0.10	<0.001	0.014	82	0.27	0.10	0.008	0.012
HLA-B*53	101	0.27	0.09	0.003	0.014	93	0.22	0.10	0.021	0.009
HLA-B*57	48	−0.49	0.12	<0.0001	0.026	46	−0.46	0.13	<0.001	0.022
Female sex	194	−0.28	0.08	<0.001	NA	187	−0.29	0.08	<0.001	NA
Female sex × HLA-A*03	22	−0.72	0.25	0.004	0.014	22	−0.77	0.26	0.003	0.016
Overall	521	NA	NA	<0.0001	0.170	503	NA	NA	<0.0001	0.162

^aWith log₁₀-transformation before analysis. Summary statistics: β regression beta (mean deviation, Δ, from the reference group), SE standard error of the mean (Δ); R ² proportion of VL variance attributable to each factor

^bAs part of the sensitivity analyses

^cFor consistency with earlier work, age is retained as a covariate regardless of its statistical significance