Fortified Balanced Energy-Protein Supplementation, Maternal Anemia, and Gestational Weight Gain: A Randomized Controlled Efficacy Trial among Pregnant Women in Rural Burkina Faso

Background: Anemia and suboptimal gestational weight gain (GWG) are associated with adverse maternal and birth outcomes. Limited research indicates that balanced energy-protein (BEP) supplements reduce the incidence of inadequate GWG. Objectives: We assessed the efficacy of a micronutrient-fortified BEP supplement on the secondary outcomes of anemia, GWG, GWG rate, and GWG in relation to the Institute of Medicine (IOM)’s recommendations, as compared with an iron–folic acid (IFA) tablet. Methods: We conducted a randomized controlled trial in Burkina Faso, among pregnant women (15–40 y old) enrolled at <21 weeks of gestation. Women received either BEP and IFA (intervention) or IFA (control). Hemoglobin (g/dL) concentrations were measured at baseline and the third antenatal care visit (ANC), whereas maternal weight was measured at baseline and all subsequent ∼7-weekly ANCs. GWG (kg) was calculated as a woman's last weight measurement (at ∼36 weeks of gestation) minus weight at enrollment, whereas GWG rate (kg/wk) was GWG divided by the time between the first and last weight measurements. GWG adequacy (%) was computed as GWG divided by the IOM's recommendation. Binary outcomes included severely inadequate, inadequate, and excessive GWG. Statistical analyses followed the intention-to-treat principle. Linear regression and probability models were fitted for the continuous and binary outcomes, respectively, adjusting for baseline measurements. Results: Women in the BEP group tended to have higher, but nonsignificantly different, GWG (0.28 kg; 95% CI: −0.05, 0.58 kg; P = 0.099). Furthermore, there were no significant differences in prenatal anemia prevalence, GWG rate, GWG adequacy, or incidence of inadequate or excessive GWG. Findings were robust to model adjustments and complete case and per protocol analyses. Conclusions: This trial does not provide evidence that fortified BEP supplementation reduces maternal anemia or increases GWG, as compared with IFA. In conjunction with the small, but positive, effects of maternal BEP supplementation on birth outcomes, our findings warrant the investigation of additional biochemical and postnatal outcomes. This trial was registered at clinicaltrials.gov as NCT03533712. Our research was reported using the CONSORT 2010 checklist (22). The prenatal phase of the MISAME-III study ({"type":"clinical-trial","attrs":{"text":"NCT03533712","term_id":"NCT03533712"}}NCT03533712) was conducted between the first enrollment on 30 October, 2019 and the last delivery on 7 August, 2021 in the catchment areas of 6 rural health centers of the health district of Houndé, Tuy Province, in the Hauts-Bassins region of Burkina Faso. In the preceding MISAME-I (23) and MISAME-II (24) RCTs, 48.4% and 45.5% of pregnant women were anemic [hemoglobin (Hb) < 11 g/dL] at baseline, respectively. Malaria transmission is perennial, with seasonal variations. The usual diet during pregnancy is nondiverse (25), predominantly maize-based with a complement of leafy vegetables (26), and consequently dietary micronutrient intakes are inadequate to cover the Estimated Average Requirements (EARs) (27). Moreover, among a subsample of MISAME-III women, the mean daily energy intake of the base diet (i.e., excluding supplements) was estimated to be ∼1940 kcal in both trial arms at the end of the preharvest season (27). The MISAME-III protocol was published previously (20). In brief, the study was a community-based, nonblinded, individually randomized 2 × 2 factorial RCT, with directly observed daily supplement intake. Women aged between 15 and 40 y and living in the study villages were identified through a census in the study area (n = 10,165). A network of 142 locally trained community support staff visited all eligible women at their homes every 5 wk to identify pregnancy early, by screening for self-reported amenorrhea. Women suspected to be pregnant were referred to the health center for a urine pregnancy test. Once gestation was preliminarily confirmed, the MISAME-III study purpose and procedures were explained in the local language: Bwamu, Mooré, or Dioula. Before randomization, we excluded women who intended to leave the study area during their pregnancy, planned to deliver outside the study area, or mothers who had a peanut allergy. After written informed consent was obtained, women (pregnancy not yet confirmed by an ultrasound) were randomly assigned to receive either a daily fortified BEP supplement and IFA tablet (intervention group) or a daily IFA tablet alone (control group) during pregnancy. The stratified randomization scheme per health center was generated by an external research analyst before the start of the study with Stata version 15.1 (StataCorp), in permuted blocks of 8 (4 control, 4 intervention). The allocation was coded with the letter A for the prenatal control arm and the letter B for the prenatal intervention arm. Randomization codes were concealed in sequentially numbered sealed opaque envelopes by project employees, who were not in direct contact with enrolled women. The project midwives, who enrolled participants, assigned women to a trial arm by drawing a sealed envelope containing the A/B letter code. MISAME-III enrollment ran from 30 October, 2019 to 12 December, 2020. Within 14 d of enrollment, a woman's pregnancy was definitively confirmed by an ultrasound. Gestational age (GA) was estimated by measuring crown–rump length (7–13 weeks of gestation) or by calculating the mean of 3–4 measurements: biparietal diameter, head circumference, abdominal circumference, and femur length (12–26 weeks of gestation) (23). Postrandomization, we excluded nonpregnant women, mothers with a GA ≥ 21 completed weeks, and multifetal pregnancies (i.e., not meeting the a priori–defined study inclusion criteria) (28). Trained village-based project workers visited 10–25 pregnant women per day to ensure the directly observed intake of BEP supplements and IFA tablets. When women had a short and scheduled absence from home, BEP supplements and IFA tablets were given to the mother in advance (thus, counted as nonobserved intakes for the respective days). The home visitors also encouraged pregnant women to attend ≥4 scheduled ANC visits every ∼7 wk. All serious adverse events (e.g., miscarriage and stillbirth) were recorded on a case-by-case basis, and verbal autopsies were conducted by the MISAME-III physician for maternal or infant deaths that occurred outside a health center. Because the fortified BEP supplement and IFA tablet were identifiable, it was not possible to blind study mothers or community support staff. However, the RCT's physician and midwives responsible for measuring the prenatal secondary maternal outcomes might be deemed partially blinded (although access was permitted to the allocation code in the enrollment file). Researchers who managed, cleaned, and analyzed MISAME-III data were not blinded. In 2016, the Bill & Melinda Gates Foundation convened an expert group to recommend the optimal nutritional composition of the BEP supplement (29). In a formative study, the most preferred and suitable MMN-fortified BEP supplement was selected for administration in the MISAME-III efficacy trial (30, 31). The BEP supplement is an LNS in the form of an energy-dense peanut paste fortified with MMNs. The BEP is made by Nutriset and is ready-to-consume, does not require a cold chain, and has a long shelf life. On average, the 72-g fortified BEP provided 393 kcal and consisted of 36% lipids, 20% protein, and 32% carbohydrates. Furthermore, the MMN content alone covered at least the IOM's daily EARs of micronutrients for pregnant women, except for calcium, phosphorus, and magnesium (32). Supplemental Table 1 provides the complete nutritional composition of the MMN-fortified BEP. Women in the intervention group daily received a fortified BEP supplement and an IFA tablet {65 mg Fe (form: FeH2O5S) and 400 μg folic acid [form: C19H19N7O6; Tolerable Upper Intake Level from fortified food or supplements, not including folate from food: 1000 μg/d (33)]}, whereas women in the control group daily received an IFA tablet only (Sidhaant Life Sciences), in accordance with Burkina Faso's national health protocol (i.e., standard of care). Following Burkinabe guidelines, all enrolled women received malaria prophylaxis (3 oral doses of sulfadoxine-pyrimethamine) at the relevant ANC visits. At enrollment (i.e., first ANC visit), we measured maternal height, weight, midupper arm circumference (MUAC) in duplicate, and Hb concentration. Maternal weight and MUAC were measured again, in duplicate, at each subsequent ANC visit. Hb concentration was assessed again between 19 and 34 weeks of gestation (i.e., third ANC visit). Furthermore, a comprehensive socioeconomic and demographic questionnaire was administered at baseline (20). Maternal height was measured to the nearest 1 cm using a ShorrBoard® Infant/Child/Adult (Weigh and Measure) and weight to the nearest 100 g with a Seca 876 scale (Seca); and the accuracy of the scales was verified weekly. Maternal MUAC was measured to the nearest 1 mm using a Seca 212 measuring tape (Seca). Pregnant women's Hb concentration was measured by spectrophotometry with a HemoCue® Hb 201+ (HemoCue); and a weekly calibration check was made with the use of a HemoCue Control Cuvette. The study's physician performed transabdominal ultrasound fetal biometry within 2 wk of enrollment. Pregnancy was confirmed and GA was estimated using a portable diagnostic imaging and full-color, flow-mapping SonoSite M-Turbo (Fujifilm SonoSite Inc.). Concurrently, maternal subscapular and tricipital skinfold measurements were taken in triplicate using a Harpenden caliper. MISAME-III data were collected using SurveySolutions version 21.5 (The World Bank) on tablets by the project physician and midwives; these data were transferred to a central Ghent University server weekly. Questionnaire assignments were sent once a week to the field team including preloaded data collected at a previous ANC visit to lower the amount of incorrect data. Furthermore, we programmed generic validation codes to avoid the entry of implausible values and to improve the quality of data collection in the field. In addition, biweekly data quality checks were conducted and missing or inconsistent data were sent back to the field for revision. The quality of ultrasound images and estimation of GA were checked for >10% of the examinations regularly by an external gynecologist, using a quality checklist and scoring sheet. The trained project workers collected daily data on fortified BEP and IFA compliance in both prenatal study arms via smartphone-assisted personal interviewing programmed in CSPro version 7.3.1 (U.S. Census Bureau, ICF, and Serpro) . Six supervisors performed monthly lot quality assurance sampling schemes of each home visitor’s work on a random day (34). All data collection forms are available on the study’s website: https://misame3.ugent.be/resources. The study protocol was approved by the Ethics Committee of Ghent University Hospital in Belgium (B670201734334) and the ethics committee of Centre Muraz in Burkina Faso (N°2018–22/MS/SG/CM/CEI). An independent Data and Safety Monitoring Board (DSMB), comprising an endocrinologist, 2 pediatricians, a gynecologist, and an ethicist of both Belgian and Burkinabe nationalities, was established before the start of the efficacy trial. The DSMB conducted remote safety reviews for adverse and serious events at 9 and 20 mo after the start of enrollment. Analyses were recorded in the MISAME-III statistical analysis plan that was validated on 24 October, 2019 and published online on 3 November, 2020 on the study’s website: https://www.misame3.ugent.be/resource-files/MISAME-III_SAP_v1_102019.pdf. For consistency and comparability of study findings, the present analyses followed the analytical procedures used to assess the efficacy of the prenatal fortified BEP intervention on birth outcomes (21). The MISAME-III efficacy trial specified 1 primary prenatal outcome: SGA [<10th percentile of the International Fetal and Newborn Growth Consortium for the 21st Century (INTERGROWTH-21st) newborn size standards (35)]. Secondary maternal outcomes of the prenatal BEP intervention included anemia (Hb < 11 g/dL) at the third ANC visit and GWG (kg) between the first and last ANC visits (or before delivery). Furthermore, we estimated the GWG rate, defined as the absolute GWG divided by the time interval between the first and last maternal weight measurements, and the expected weight gain for each woman at the time of their last observed weight measurement using the following IOM 2009 formula (1): The expected total weight gain during the first trimester was assumed to be 2 kg for underweight (BMI < 18.5) and normal-weight women (BMI = 18.5–24.9), 1 kg for overweight women (BMI = 25–29.9), and 0.5 kg for obese women (BMI ≥ 30); and the recommended rates of GWG for the second and third trimesters were 0.51, 0.42, 0.28, and 0.22 kg/wk for underweight, normal-weight, overweight, and obese women, respectively (1). Finally, the percentage adequacy of GWG was calculated by dividing the actual GWG by the expected GWG at the last observed weight measurement, multiplied by 100. This is a continuous measure that has been used in previous GWG studies in Africa (36, 37). Following Liu et al. (37), we further categorized the percentage adequacy of GWG into binary outcome measures. Inadequate GWG was defined as a percentage adequacy of GWG < 90%, severely inadequate GWG as a percentage adequacy of GWG < 70%, and excessive GWG as a percentage adequacy of GWG ≥ 125%. The cutoffs 90% and 125% correspond to the lower and upper limits of the recommended total weight gain during pregnancy by the IOM's guideline. The recommended range is 12.5–18 kg for women who are underweight (BMI < 18.5), 11.5–16 kg for normal weight (BMI = 18.5–24.9), 7–11.5 kg for overweight (BMI = 25–29.9), and 5–9 kg for obese (BMI ≥ 30) (1). Only singleton pregnancies were included in the analysis. All analyses were conducted by the intention-to-treat (ITT) principle to reduce potential bias arising from missing data. Therefore, before analyses, we performed multiple imputation of missing maternal Hb concentration (g/dL), maternal weight before delivery (kg), and gestational duration (wk) under the “missing at random” assumption. Fifty imputations of missing continuous outcome data for cases lost to follow-up were run to estimate the regression coefficients, based on the following predictors at baseline: maternal height (cm), maternal weight (kg), MUAC (mm), Hb (g/dL), and age (y) at inclusion; GA at baseline (wk); primiparity; and month of inclusion. Anemia and GWG adequacy variables were calculated from the imputed continuous data. Descriptive data are presented as percentages or means ± SDs. Unadjusted and adjusted group differences were estimated by fitting linear regression models for the continuous outcomes, to estimate the mean group difference, and using linear probability models with robust variance estimators for the binary outcomes, to estimate risk differences in percentage points (pp). All models were adjusted for the baseline value of the outcome of interest [i.e., either Hb or maternal weight at study enrollment, thus an ANCOVA (38, 39)] and contained health center and randomization block as fixed effects to account for clustering by the study design. Adjusted models in addition contained potential baseline prognostic factors of maternal outcomes, including maternal height (cm), MUAC (mm), and age (y) at inclusion; GA at baseline (wk); and primiparity. We did not adjust for any other sociodemographic variables, owing to balanced baseline characteristics across the prenatal study groups (i.e., < |2.5| pp difference). To assess the robustness of the primary findings, we conducted the following sensitivity analyses: 1) complete case analysis (i.e., excluding women who were lost to follow-up for birth outcomes); and 2) per protocol analysis restricting the intervention sample to women with fortified BEP compliance of ≥75%. The strict compliance rate was calculated by dividing the total number of BEP supplements effectively taken under direct observation of a trained home visitor by the theoretical maximum number of prenatal BEP supplements allowed (i.e., the number of days between study inclusion and delivery). Moreover, to assess the potential underestimation of absolute GWG, we replicated our complete case analysis among mothers who had a baseline weight measurement taken at <14 weeks of gestation (i.e., first trimester). In addition, for complete cases, we used the INTERGROWTH-21st GWG standards to derive GWG z scores (40). Because the currently available GWG equation for recommended weight gain is for normal-weight women with a GA between 14 and 40 wk, we restricted this outcome to normal-weight women with a GA of ≤40 completed weeks at their last ANC visit. Statistical significance was set at P < 0.05 for all 2-sided tests. All analyses were conducted with Stata version 16.1 (StataCorp).