Growth of children according to maternal and child HIV, immunological and disease characteristics: a prospective cohort study in Kinshasa, Democratic Republic of Congo

Background Most HIV-infection in children occurs in sub-Saharan Africa where antiretroviral therapy is seldom available. This study compares the growth progression and retardation of HIV-infected and uninfected children in the Democratic Republic of Congo (formerly Zaire). It estimates the risk for child growth retardation according to child and maternal immunological factors, severity of maternal and child illness, and maternal socioeconomic and marital status. Methods In a prospective cohort study of 258 children born to HIV seropositive mothers and 256 children of seronegative mothers in Kinshasa, Congo, the growth in length, weight, and weight-for-length of infected children (n = 68), uninfected children born to seropositive mothers (n = 190), and uninfected children born to uninfected mothers (n = 256) was compared. Serological, anthropometric and other clinical measures were collected monthly from 3-12 months and bi-monthly during the second year of life. Polymerase chain reaction for HIV was performed on bloods drawn at 2 days and 3 months post partum. Length-for-age, weight-for-age, and weight-for-length mean z-scores against National Center for Health Statistics (NCHS) reference data were calculated, and Cox proportional hazards models were used to estimate the risk of falling below -2.00 z-scores as a function of child and maternal immunological. clinical and sociodemographic variables. Results There was no difference in mean length-for-age at birth between HIV-infected (Group I) children, uninfected children of infected mothers (Group 2) or Control children, but by 3 months old, HIV-infected children were shorter than both Group 2 and Controls. In weight-for-age and weight-for-length, Group 1 infants were lighter and more wasted at birth and onwards. Group 2 newborns were lighter than Controls at birth, but by three months they had caught up to Controls in both length and weight and remained the same as Controls thereafter. The odds of falling below -2.00 z-scores by 20 months for length, weight, and weight-for-length for HIV-infected children compared to uninfected children were 2.10, 2.84, and 2.56 respectively. Both HIV-infection and associated illnesses were factors associated with child stunting, underweight and wasting. The mother's age, socioeconomic status, presence of father, stage of illness and immune status had no detectable effect on the child's growth in the first two years of life. Conclusion The HIV-infected children in Congo with no access to antiretroviral therapy were stunted, underweight, and wasted compared to same age uninfected children. Both HIV infection and HIV-associated signs and symptoms, not maternal immunological or socioeconomic circumstances, placed children at risk for growth retardation.