Background: It has been well established that breastfeeding is beneficial for child health, however there has been debate regarding the effect of lactation on maternal health in the presence of HIV infection and the need for nutritional supplementation in HIV positive lactating mothers. Aims. To assess the effect of nutritional supplementation to HIV infected lactating mothers on nutritional and health status of mothers and their infants. Methods. A randomized controlled clinical trial to study the impact of nutritional supplementation on breastfeeding mothers. Measurements included anthropometry; body composition indicators; CD4 count, haemoglobin and albumin; as well as incidence rates of opportunistic infections; depression and quality of life scores. Infant measurements included anthropometry, development and rates of infections. Results: The supplement made no significant impact on any maternal or infant outcomes. However in the small group of mothers with low BMI, the intake of supplement was significantly associated with preventing loss of lean body mass (1.32 kg vs. 3.17 kg; p = 0.026). There was no significant impact of supplementation on the infants. Conclusions: A 50 g daily nutritional supplement to breastfeeding mothers had no or limited effect on mother and child health outcomes. Clinical trial registration. ISRCTN68128332 (). © 2011 Kindra et al; licensee BioMed Central Ltd.
The study sample consisted of HIV positive pregnant women of Zulu ethnicity attending the antenatal clinic (ANC) at Umkhumbane Community Health centre at Cato Manor. Breastfeeding mothers were randomized in a controlled clinical trial (RCT) to receive supplementation or non-nutritive household supplies. The supplement was a peanut/soya milk spread enriched with micronutrients and came packaged in plastic tubs. A daily serving of 50 g of the nutrition supplement provided 280 kcal energy and 8 g protein. The non-nutritive supplies included tea; shampoo and conditioner of equal monetary value and were therefore not considered to have an impact on nutrition. Randomisation took place by mothers removing a card pre-marked with group number from a slit in a closed box; this was done to enable them to be confident that the randomisation was unbiased. Randomisation was conducted by a study counsellor who maintained a separate record of the randomisation log. Outcomes were assessed by a separate clinician. To ensure that the clinician was blinded, the counsellors issued a brown bag containing either the nutritional supplement or the non-nutritive supplies. Both supplement and the non-nutritive supplies were packed in identical brown bags in an attempt to mask the allocation of group. The data on the randomization was collected using subject identification numbers only so as to reduce observer bias. To prevent dilution of the effect of the intervention and mothers discussing the content of their brown bags, separate clinic visit days were allocated for the two groups. To monitor adherence, a monthly register with mothers’ signatures was maintained as well as a questionnaire was administeredto the mothers at every visit discussing daily intake of the supplement and any possible reasons for not taking the supplement including information on any benefits or side-effects that they attributed to the supplement. The counselors also tried to ascertain that the mothers were not sharing the supplement with members of their household by interviewing the mothers. The study was designed to detect a difference of ≥ 4 kg between the two groups. It was determined that a sample size of 40 in each group would achieve 90% power using a one-sided, two-sample t-test. The true difference between the means was assumed to be 0.00. The significance level (alpha) of the test was 0.025. The data are drawn from populations with mean weights with standard deviations of 6.70 and 3.60 kg obtained from piloting the body composition methods in the same population. Allowing for a 30% loss to follow-up due to high mobility of the population post delivery, we estimated a sample size of 52 in each group. Then adding an additional 30% for loss of datadue to missed visits for body composition analyses or laboratory tests; we estimated a sample sizeof 64 in each group. HIV positive mothers attending the ANC were referred to the on-site MTCT Plus programme, an internationally funded programme that provided PMTCT services as well as comprehensive care and treatment for HIV infected mothers and their families. Mothers were routinely counselled on the feeding options and risks thereof were explained. Breastfeeding HIV positive mothers were referred to dedicated study counsellors who were housed in an independent building a few metres from the clinic. These study counsellors screened the mothers for study eligibility and counselled them on the study procedures. Only mothers planning to breastfeed for at least 6 months were eligible to enter the study. Mothers with advanced disease (CD4 < 200 or WHO stage 3 and 4) requiring ART and those not resident in the area were not eligible for recruitment. In addition, mothers of infants requiring specialized management and with gestation less than 36 weeks were not eligible for study enrolment. Interested mothers were pre-enrolled after a written informed consent was obtained at the next antenatal visit. Mothers were then seen at 2 weeks post-delivery and officially enrolled into the study. Study assessments for both mothers and their infants were done at 2 and 6 weeks post-delivery and monthly thereafter till 6 months of age with a final study visit at 9 months. At each visit, a clinical examination and anthropometric measurements were done; along with a developmental assessment on the infants and Karnofsky scoring on the mothers. Body composition and SRQ 20 assessments were done on the mother at 2 weeks and then three monthly till 9 months. Mothers and the positive infants were assessed for any signs of disease progression; opportunistic infections; WHO disease staging; and CD4 counts. When indicated they were started on ART as per national guidelines [34]. A 24-hour dietary recall was done at 3, 6 and 9 months to assess for similarity of dietary intake in the groups to rule out any confounding for the outcome of supplementation. Infant feeding history was taken at each visit. Infants received multivitamins and cotrimoxazole for Pneumocystis jiroveci pneumonia (PCP) prophylaxis from 6 weeks that was discontinued when the infant was confirmed HIV DNA PCR negative. Mothers received PCP prophylaxis if indicated. Anthropometric measurements included weight; height; length; mid-upper arm circumference (MUAC); and triceps skinfold thickness (TSF) on both mother and infant. In addition head circumference measurements were taken on the infants. All measurements were performed in duplicate as per ISAK (International Society for the Advancement of Kinanthropometry) methods by either the principal investigator or the dietitian to reduce inter-observer variability and enhance reliability [35]. Body Mass Index (BMI) was calculated by dividing the weight of the subject by their height squared (kg/m2). All women in the study were counseled and supported to practice exclusive breastfeeding for 6 months. Mothers whose infants were asymptomatic and who tested HIV negative at 6 weeks were counseled on either discontinuing breastfeeding at 6 months, or heat-treating expressed breast milk [36,37]. Saliva samples from mothers were collected to measure Deuterium enrichment using standardized IAEA (International Atomic Energy Agency) methodology [38]. These samples were analysed for total body water (TBW) using an FTIR (Fourier transform infra-red spectrophotometer). LBM was calculated from the TBW (73.2% of TBW). Fat mass (FM) was calculated as the difference between the body weight and the LBM. Maternal bloods were done at 2 weeks and at 6 months and tests included haemoglobin; haematocrit; mean cell volume; platelet count; white cell count and lymphocyte count; total protein and albumin; and CD4 counts. Infants had a HIV DNA PCR done routinely at 6 weeks and the infants who tested positive had a repeat PCR at 10 weeks. A DNA PCR was repeated at 9 months in infants who tested negative at 6 weeks. All investigations were routine tests done at the accredited National Health Laboratory Services. Data was entered into an MS Access database (©MS Office 2003). It was rechecked for any missing information and entry errors. A unique subject identification number to maintain confidentiality identified mothers and infants. To study associations/trends within categories: BMI was categorized as ≤ 24.9 kg/m2 and ≥ 25 kg/m2. Disease progression was categorized as progression to WHO stage 3 or higher; a poor Karnofsky score was categorized as 80% or below (80% and above is associated with ability to do normal activity); and the WHO cutoff of an SRQ20 score of 8 and above (associated with depression) was used as an indicator of depression [30,31]. SPSS 15.0 for windows was used for analysis (© SPSS Inc.). The two groups were compared for baseline socio-economic as well as clinical criteria to assess for similarity. Continuous variables were checked for normality of distribution. Independent t tests and chi-squared tests were done for the continuous and categorical variables respectively. Non-parametric tests were conducted for the skewed continuous variables and medians with their inter-quartile range are reported. Paired T tests for normally distributed variables and Wilcoxon signed rank test for skewed variables were performed to study changes over time in the laboratory parameters, anthropometric and body composition measurements. Incidence rates and incidence rate ratios (IRR) were calculated for clinical events, opportunistic infections, Karnofsky score and SRQ 20 score using STATA 9.2 (©StataCorp). As low BMI is considered to be an important prognostic marker for breastfeeding HIV positive women, all anthropometric, body composition, laboratory and disease progression parameters were analysed according to BMI category. As BMI ≥ 25 kg/m2 was being used by the province as a cutoff for not being eligible for supplements in the HIV positive population and as our median BMI was 26 kg/m2, we categorized the BMI in our sample using the same categorization as the province: ≥ 25 kg/m2 and < 25 kg/m2. An intention to treat analysis was conducted using generalized estimating equations (GEE) with robust estimators and an exchangeable covariance matrix to analyse the continuous variables over time controlling for socio-economic status; baseline CD4 and BMI. Generalised linear models (GLM) with Poisson distribution and log function were used to analyse the incidence rates of the opportunistic infections. The WHO igrowup tool using the 2007 WHO growth standards was used to assess the growth parameters [39]. Ethical approval for the study was obtained from the Bioethics Committee of the University of KwaZulu-Natal (H081/05).
N/A