Evolution of the levels of human leukocyte antigen G (HLA-G) in Beninese infant during the first year of life in a malaria endemic area: using latent class analysis

Table 2 Risk factors of soluble HLA-G evolution from 3 to 12 months in 165 newborns at Tori Bossito, 2007–2010: latent class analysis

Covariates	Univariate estimation^a	Adjusted estimation^b	95 % CI	p value
Intercept	–	0.58	0.28–0.88	<10⁻³*
HLA-G in cord blood
No	–	–
Yes	0.60	0.42	0.15–0.70	0.002^c*
Low birthweight (<2500 g)
No	–	–
Yes	0.62	0.86	0.42–1.29	<10⁻³*
Malaria infections
≤1	–	–
>1	0.20	0.33	0.11–0.54	0.003*
Malaria exposure (quartiles)
Low	–	–
Middle	0.33	0.43	0.19–0.68	0.001*
High and very high	0.32	0.35	0.13–0.56	0.002*
Placental malaria
No	–	–
Yes	–0.07	–0.12	–0.35–0.11	0.36

^aUnivariate estimation represents the coefficient of the univariate regression
^bAdjusted estimation is obtained after the multivariate analysis
^cResults were the same, when “HLA-G in cord blood” was replaced by “HLA-G in maternal peripheral blood”. Presence of sHLA-G in maternal blood is highly significantly related to high level of HLA-G in newborns during the first 12 months. We did not introduce together the two covariates in the model because they are highly correlated [24]

ISSN: 1475-2875