Table 1.

Associations between maternal HIV biomarkers in the second trimester and infant mortality through 18 months

Maternal biomarker	Child mortality n/N (%)	Hazard ratio	P	Adjusted hazard ratio	P
HIV viral load undetectable	11/260 (4.2%)	Reference		Reference
Detectable ≥150 copies/mL	27/330 (8.2%)	1.96 (0.89, 4.34)	0.097	1.95 (0.82, 4.63)a	0.130
C-reactive protein <4.72 mg/L	9/318 (2.8%)	Reference		Reference
C-reactive protein ≥4.72 mg/L	32/318 (10.1%)	3.64 (1.86, 7.13)	<0.001	3.53 (1.73, 7.19)b	0.001
Soluble CD14 < 1.36 mg/L	20/318 (6.3%)	Reference		Reference
Soluble CD14 ≥ 1.36 mg/L	21/318 (6.6%)	1.06 (0.56, 2.04)	0.851	1.04 (0.53, 2.05)c	0.914
Log rise in HIV viral load	N/A	1.40 (1.05, 1.86)	0.022	1.41 (1.02, 1.94)a	0.035
Log rise in C-reactive protein	N/A	2.31 (1.52, 3.52)	<0.001	2.09 (1.33, 3.27)b	0.001
Log rise in soluble CD14	N/A	1.65 (0.25, 10.75)	0.599	1.65 (0.25, 10.80)c	0.600

Cox proportional hazard models were used to evaluate associations. Presented P-values have not been adjusted for multiple comparisons.

^aAdjusted for maternal age and education, household socioeconomic status, and gestational age at blood sampling and randomised trial arm.

^bAdjusted for maternal HIV viral load, CD4 count, CMV co-infection, gestational age at blood sampling and randomised trial arm.

^cAdjusted for maternal HIV viral load, CD4 count, gestational age at blood sampling and randomised trial arm.