Menu
Menu
Menu
Menu
Children of HIV-infected mothers experience poor growth, but not much is understood about the extent to which such children are affected. The Research to Improve Infant Nutrition and Growth (RIING) Project used a longitudinal study design to investigate the association between maternal HIV status and growth among Ghanaian infants in the first year of life. Pregnant women in their third trimester were enrolled into three groups: HIV-negative (HIV-N, n=185), HIV-positive (HIV-P, n=190) and HIV-unknown (HIV-U, n=177). Socioeconomic data were collected. Infant weight and length were measured at birth and every month until 12 months of age. Weight-for-age (WAZ), weight-for-length (WLZ) and length-for-age (LAZ) z-scores were compared using analysis of covariance. Infant HIV status was not known as most mothers declined to test their children’s status at 12 months. Adjusted mean WAZ and LAZ at birth were significantly higher for infants of HIV-N compared with infants of HIV-P mothers. The prevalence of underweight at 12 months in the HIV-N, HIV-P and HIV-U were 6.6%, 27.5% and 9.9% (P<0.05), respectively. By 12 months, the prevalence of stunting was significantly different (HIV-N=6.0%, HIV-P=26.5% and HIV-U=5.0%, P<0.05). The adjusted mean±SE LAZ (0.57±0.11 vs. -0.95±0.12; P<0.005) was significantly greater for infants of HIV-N mothers than infants of HIV-P mothers. Maternal HIV is associated with reduce infant growth in weight and length throughout the first year of life. Children of HIV-P mothers living in socioeconomically deprived communities need special support to mitigate any negative effect on growth performance.
The study was conducted from 2003 to 2008 in the Yilo Krobo (population 93 586) and Manya Krobo (population 165 409) districts located about 80 km north of Accra, in the Eastern region of Ghana. The main adult occupations are trading, fishing, pottery and farming. The districts are served by three hospitals. The Eastern region where the two districts are located had an HIV prevalence of 3.7% compared with the national prevalence of 2.2% at the time of the study (GSS 2004). The prevalence of stunting among children under 5 years in the Eastern region was 27% (GSS 2004). Literacy rates among men and women in the region were 81% and 64%, respectively (GSS 2004). Most households in the districts had access to electricity and public pipe‐borne water within the community. The study was a prospective cohort involving pregnant women (n = 552) in their third trimester attending prenatal clinics in three hospitals in the districts. To be eligible, the participant had to be (1) pregnant; (2) at least 18 years of age; (3) undergo voluntary pre‐test counselling and if tested agree to have her HIV test results released to the project supervisor; (4) available for the entire duration of the study; and (5) free from clinical and physical conditions that would limit her ability to care for the infant. The women went through the regular Ghana Health Service (GHS) antenatal clinic procedures, which included voluntary pre‐test counselling to offer HIV testing (Okronipa et al. 2012). Women who agreed to be tested were identified as HIV‐P or HIV‐N those who refused testing were identified as HIV‐U. Recruitment into the study was done in partnership with the hospital nursing staff responsible for voluntary counselling and testing (VCT) in the three participating hospitals. After testing (or after pre‐test counselling if testing was refused), the GHS nurse informed women about the study. Recruitment of HIV‐N and HIV‐U (because refusal to be tested) women followed the identification of HIV‐P women to assure similar seasonal enrolment in the three groups. All HIV‐P women who consented were enrolled. If more HIV‐N and HIV‐U women were available at the time, participants were randomly chosen from those available that day. The GHS nurse approached 692 pregnant, of whom 653 expressed interest in the study. HIV status was released only to the project supervisor who personally visited each woman at her home to further explain the study. Informed written consent was obtained from 552 women: HIV‐N (n = 185), HIV‐P (n = 190) and HIV‐U (n = 177). HIV tests were routinely done by the recruiting hospitals using the Rapid Test Abbott Determine HIV‐1/2 (Abbott Laboratories, Abbott Park, IL, USA). At the time of the study, the administration of nevirapine to the pregnant woman at labour and to the infant at birth was the national protocol for the prevention of mother‐to‐child transmission (PMTCT). Sample size was determined based on estimates of effect sizes and individual‐level variability documented for anthropometric, breast milk intake and morbidity data from previous studies in similar low‐income communities (Marquis et al. 2002). Calculations used an effect ratio of 1, a one‐tailed test, a significance level of 0.05 and a power of 80% (Kelsey et al. 1986). Morbidity rates based on 20% and mean difference of 15% yielded the largest sample size of 151 per group. Assuming a loss to follow‐up of 25%, a total sample of 189 per group was considered to be adequate. Data were collected on socio‐demographic information (age, education, marital status), occupation, household characteristics (size and composition), food production, economic activities, household FI, maternal post‐natal depression and maternal stress. These data were collected at baseline, birth, 3, 6, 9 and 12 months after birth. Infant anthropometric measurements (weight and length) were taken within 24 h of birth, and then monthly thereafter until the infant reached 12 months. Infants were weighed naked to the nearest 100 g (Tanita Corporation of America Inc., Arlington Heights, IL, USA), and recumbent length was measured to the nearest 0.1 cm using an infant stadiometer at home (Shorr Productions, Olney, MD, USA). Maternal post‐natal depression was measured at birth and at 6 months after birth using the Edinburgh Post‐natal Depression Scale as described elsewhere (Okronipa et al. 2012). Maternal stress was measured at baseline, birth, 3, 6, 9 and 12 months using the Perceived Stress Scale as described elsewhere (Okronipa et al. 2012). Other data on infant morbidity and feeding, and maternal anthropometry, morbidity and social capital not reported in this paper were also collected. At 12 months of age, all mothers were given the opportunity through a separate informed consent process to have their infant tested for HIV. Those who agreed to testing (n = 81) had a finger prick blood taken on to a filter paper to determine the child's HIV status by DNA polymerase chain reaction (PCR) analysis (Fransen et al. 1994, 1998). Analyses were done using SAS v. 9.2 (SAS Institute, Cary, NC, USA). Background characteristics were assessed by using chi‐squared tests for categorical variables and analysis of variance (with Ryan–Einot–Gabriel–Weich for post hoc pair‐wise comparison) for continuous variables. As a proxy for household socioeconomic status, we created an ‘amenities’ factor from a set of 18 socioeconomic variables (house‐building materials, location of household water, toilet, access to electricity, cooking fuel, ownership of appliances) using factor analysis with varimax rotation. Lower values for amenities were assumed to indicate poorer households. Household‐level FI at birth, 3, 6, 9 and 12 months after birth was determined using a 14‐item scale derived from the US Household Food Security Survey Module. Rasch analyses confirmed the psychometric validity of the scale (R. Perez‐Escamilla, ‘unpublished observations’). Based on these analyses, households were classified as food secure if none of the questions were affirmed. The cut‐off points for classifying households into different FI levels based on adding the number of questions affirmed were: mild (0), moderate (1–6) and severe (7–14). Questions were asked in reference to the month preceding the survey. Eight questions were asked in reference to adult(s)/household and six in reference to children who were defined as individuals under 16 years of age living in the household. At each time point, household FI was assigned a score of 0 if they were food secure, 1 if they were mildly, 2 if they were moderately or 3 if they were severely food insecure. A mean FI score was calculated across time to obtain the final mean FI level for each household throughout the duration of the study. The mean FI level was considered as a continuous variable, and was used as a covariate in the analysis. Child anthropometric measurements were converted to weight‐for‐age (WAZ), length‐for‐age (LAZ) and weight‐for‐length z‐scores (WLZ) using the WHO Child Growth Standards (WHO Multicenter Growth Reference Study Group 2006). Mean WAZ, LAZ and WLZ for each group at each month from 1 to 12 months were calculated and compared using ANCOVA, adjusting for child sex and birthweight, and maternal age, education, marital status, household amenities (as a proxy for socioeconomic status) and mean household FI. These variables were selected because they were either different among the three groups at baseline, or were related to growth. We used a repeated measures analysis (SAS Proc Mixed) to determine whether the growth of children over time differed for children of mothers in the three HIV status categories with Tukey–Kramer post hoc test. A three‐way HIV–age–sex interaction term reflecting sex differences in the association between HIV and growth across the first year was tested and found to be not significant and therefore was not included in the final model. In all analyses involving growth, we controlled for the child and maternal characteristics mentioned earlier. The percentage of children with z‐scores <−2 standard deviation at 6 and 12 months was determined. Ethical approval for the study was obtained from the Institutional Review Boards of the University of Ghana, Iowa State University, University of Connecticut and McGill University.
N/A
