Small-quantity, lipid-based nutrient supplements provided to women during pregnancy and 6 mo postpartum and to their infants from 6 mo of age increase the mean attained length of 18-mo-old children in semi-urban Ghana: A randomized controlled trial

Background: Childhood stunting usually begins in utero and continues after birth; therefore, its reduction must involve actions across different stages of early life. Objective: We evaluated the efficacy of small-quantity, lipid-based nutrient supplements (SQ-LNSs) provided during pregnancy, lactation, and infancy on attained size by 18 mo of age. Design: In this partially double-blind, individually randomized trial, 1320 women at #20 wk of gestation received standard iron and folic acid (IFA group), multiple micronutrients (MMN group), or SQ-LNS (LNS group) daily until delivery, and then placebo, MMNs, or SQ-LNS, respectively, for 6 mo postpartum; infants in the LNS group received SQ-LNS formulated for infants from 6 to 18 mo of age (endline). The primary outcome was child length by 18 mo of age. Results: At endline, data were available for 85% of 1228 infants enrolled; overall mean length and length-for- Age z score (LAZ) were 79.3 cm and 20.83, respectively, and 12% of the children were stunted (LAZ ,22). In analysis based on the intended treatment, mean 6 SD length and LAZ for the LNS group (79.7 6 2.9 cm and 20.69 6 1.01, respectively) were significantly greater than for the IFA (79.1 6 2.9 cm and 20.87 6 0.99) and MMN (79.1 6 2.9 cm and 20.91 6 1.01) groups (P = 0.006 and P = 0.009, respectively). Differences were also significant for weight and weight-for- Age z score but not head or midupper arm circumference, and the prevalence of stunting in the LNS group was 8.9%, compared with 13.7% in the IFA group and 12.9% in the MMN group (P = 0.12). In analysis based on actual supplement provided at enrollment, stunting prevalences were 8.9% compared with 15.1% and 11.5%, respectively (P = 0.045). Conclusion: Provision of SQ-LNSs to women from pregnancy to 6 mo postpartum and to their infants from 6 to 18 mo of age may increase the child’s attained length by age 18 mo in similar settings. This trial was registered at clinicaltrials.gov as NCT00970866. Am J Clin Nutr 2016;104:797-808.

Details of the study setting, design, participants, and blinding schemes were reported previously (22). Briefly, the study was conducted in several adjoining semi-urban communities (Somany-Odumasi-Kpong area) in the Yilo Krobo and the Lower Manya Krobo districts ∼70 km north of Accra. It was designed as a partially double-blind, parallel, individually randomized, controlled trial with 3 equal-size groups. Pregnant women attending usual antenatal clinics in the 4 main health facilities in the area were included if they were ≥18 y old and at ≤20 wk of gestation based on the information available at the time. They were excluded if any of the following applied: not residing in the area, intention to move within the next 2 y, milk or peanut allergy, participation in another trial, HIV infection, asthma, epilepsy, tuberculosis, any malignancy, or unwillingness to sign or thumbprint the relevant consent forms, receive field workers, or take the study supplement. After baseline assessments, eligible women were randomly assigned to receive one of 3 treatments: 1) 60 mg Fe plus 400 μg folic acid during pregnancy and 200 mg Ca serving as a placebo for the first 6 mo postpartum (hereafter, IFA supplement or group); 2) MMN capsule containing 18 vitamins and minerals (including 20 mg Fe) during pregnancy and for the first 6 mo postpartum, (hereafter, MMN supplement or group); and 3) SQ-LNSs with similar micronutrients as the MMN supplement in addition to calcium, phosphorus, potassium, and magnesium as well as energy (118 kcal/d) and macronutrients (e.g., protein and EFAs) during pregnancy and for the first 6 mo postpartum, followed by supplementation for their offspring from 6 to 18 mo of age by using SQ-LNSs designed for infants (hereafter, LNS supplement or group). The IFA and MMN supplements served as controls because at the time of the study, IFA was the standard WHO nutritional supplementation for pregnant women (23), and available evidence suggested that prenatal supplementation with MMN supplements promoted fetal growth (24–26). The study statistician at University of California, Davis developed group allocations with the use of a computer-generated (SAS version 9.3; SAS Institute) randomization scheme in blocks of 9. At each enrollment, the study nurse offered sealed, opaque envelopes bearing group allocations, 9 envelopes at a time, and the woman picked one to reveal the allocation. Allocation information was kept securely by the field supervisor and the study statistician only. Enrollment was completed between December 2009 and December 2011. The study protocol was approved by the ethics committees of the University of California, Davis; the Ghana Health Service; and the University of Ghana Noguchi Memorial Institute for Medical Research, and was registered on clinicaltrials.gov as {“type”:”clinical-trial”,”attrs”:{“text”:”NCT00970866″,”term_id”:”NCT00970866″}}NCT00970866. A 5-member independent Data and Safety Monitoring Board monitored the incidence of serious adverse events (SAEs). The micronutrient contents of the maternal supplements are reported elsewhere (22) as was the rationale underlying the concentrations of the nutrients used in all the supplements, including SQ-LNSs for infants (15). Apart from iron, which was kept at 20 mg/d in the MMNs and SQ-LNSs for women, the vitamin and mineral contents were either 1 time or 2 times the Recommended Dietary Allowance for pregnancy or, in a few cases, the maximum amount that could be included in the supplement given technical and organoleptic constraints. The nutrients in the SQ-LNS for infants (20 g/d) were energy (118 kcal), protein (2.6 g), fat (9.6 g), linoleic acid (4.46 g), α-linolenic acid (0.58 g), vitamin A (400 μg retinol equivalents), thiamin (0.3 mg), riboflavin (0.4 mg), niacin (4 mg), vitamin B-6 (0.3 mg), vitamin B-12 (0.5 μg), vitamin C (30 mg), vitamin D (5 μg), vitamin E (6 mg), vitamin K (30 μg), folic acid (80 μg), pantothenic acid (1.8 mg), iron (6 mg), zinc (8 mg), copper (0.34 mg), calcium (280 mg), phosphorus (190 mg), potassium (200 mg), magnesium (40 mg), selenium (20 μg), iodine (90 μg), and manganese (1.2 mg). The IFA and MMN supplements were provided as capsules in blister packs, and the LNS supplements for women were in 20-g sachets and those for infants in 10-g sachets (2 given/d). All supplements were intended for daily consumption, the IFA and MMN with water after a meal, 1 capsule/d. The LNS supplements for women were mixed with any prepared food, 1 sachet/d, and LNS supplements for infants were mixed with complementary foods, 1 sachet at a time, on 2 different occasions during the day. To maintain blinding, 2 individuals independent of the study placed color-coded stickers (3 different colors for IFA and 3 for MMN supplements), which also bore the letters P or L (indicating consumption during pregnancy or lactation, respectively) behind the blister packs, so that the capsules were known to the study team and participants only by the colors of the stickers. It was not possible to blind study workers and participants to the capsules (IFA and MMN supplements) compared with the LNS supplements because of their apparent differences, but laboratory staff, anthropometrists, and data analysts had no knowledge of group assignment until all preliminary analyses had been completed. At baseline, we collected women’s sociodemographic information, determined gestational age (GA) by ultrasound biometry (Aloka SSD 500; Hitachi), except for a few who had ultrasound information at the time of screening, and assessed women’s anthropometric status (standard procedures), hemoglobin (HemoCue AG; Wetzikon) and malaria parasitemia (Vision Biotech; South Africa) (22). Immediately after enrollment, the study nurse gave each woman a 2-wk supply of assigned supplement (IFA or MMN, bearing the letter P, or LNS for women), with a description of mode of consumption and a standard nutrition message (“Do not forget to eat meat, fish, eggs, fruits, and vegetables whenever you can; you still need these foods even as you take the supplement we have given you.”). Thereafter, field workers visited women in their homes biweekly until delivery and delivered fresh supplies of supplements as well as monitored morbidity and supplement intakes during each visit (22). At delivery field workers immediately delivered a new supply of LNSs or color-coded capsules bearing the letter L and thereafter visited women and their infants each week. All live-born singleton infants were automatically enrolled. For twins (n = 22), only one infant was randomly selected, and the selected infant was enrolled if he or she was born alive. Stillborn infants were not enrolled. Within 48 h after delivery or between 3 and 14 d after delivery for 105 infants (9%), trained anthropometrists recorded the infants’ date of birth and measured the infants’ birth weight to the nearest 20 g (Seca 383; Seca), length to the nearest 0.1 cm (Seca 416; Seca), and head and midupper arm circumferences (MUAC) to the nearest 0.1 cm (Shorr Productions) by using procedures described by WHO (27). These measurements were performed in duplicate, unless the second measurement differed from the first by more than a predefined tolerable amount (0.1 kg for weight and 0.5 cm for length, head circumference, and MUAC), in which case a third measurement was taken. At every other weekly home visit until 6 mo postpartum, women received a fresh supply of supplements, and their supplement consumption was monitored. Data on morbidity and SAEs were collected for women biweekly until 6 mo postpartum and for infants weekly until 18 mo of age. Women (and their husbands, if present) were told to call the project office or field supervisor whenever there was any or suspected SAE (e.g., death or hospitalization) in a study woman or infant, and therefore field workers following-up on such calls also collected data on SAEs between the weekly visits. At 3, 6, 12, and 18 mo of age, children were brought to the laboratory, and the anthropometric measurements carried out at birth were repeated. All anthropometrists were standardized according to WHO standards (27) shortly before data collection began and then every 6 mo thereafter. At the weekly home visit after the anthropometric measurements at 6 mo of age, field workers delivered the following standard nutrition message to all mothers, which was also repeated by the study nurse during the anthropometric measurements at 12 mo of age: “breastfeed your baby as you did before 6 mo of age; do not forget to give your baby other foods such as meat, fish, eggs, fruits, and vegetables whenever you can because your baby still needs these foods.” If a mother was in the LNS group, the fieldworker modified the last part of the message to say, “Your baby still needs these foods even if you give him/her the infants’ LNSs.” Mothers of infants who had not yet received any complementary foods were told to “start giving complementary foods to the infants as soon as possible, because breast milk alone is not enough for the baby after he or she is 6 mo old.” To mothers in the LNS group, fieldworkers gave a 1-wk supply of infants’ LNSs along with the instruction to mix the entire content of one sachet with 2–3 tablespoons (30–45 mL) of any food for the infants before feeding additional foods if the infants desired, 2 times/d. Mothers in the LNS group who had not yet started feeding complementary foods to their infants were told to do so for at ≥3 d before feeding LNSs to the infants. Mothers in the LNS group received a fresh supply of infants’ LNSs every week thereafter until the infants exited the study at 18 mo of age. Primary outcomes were infants’ length (cm) and length-for-age z score (LAZ) at 18 mo of age. Secondary outcomes were infants’ weight (kg); head circumference (cm); MUAC (cm); z scores for weight-for-age (WAZ), weight-for-length (WLZ), and head circumference-for-age (HCZ); stunting; underweight; wasting; small head circumference at 18 mo of age; growth from 0 (birth) to 18 mo of age; and incidence of SAEs from 0 to 18 mo of age. An effect size (Cohen’s d) of 0.3 (small-to-moderate effect size) (28) was the basis for all sample size calculations in the iLiNS-DYAD-Ghana study. This produces the same target sample size regardless of the outcome being considered, because it is independent of the units of measurement. Thus, our sample size was based on detecting an effect size of 0.3 between any 2 groups for any continuous variable at any time point (e.g., 18 mo of age), with a 2-sided 5% test and 80% power. We enrolled 1320 pregnant women into the study, as described previously (22). Because of the temporary mislabeling of IFA and MMN capsules described previously (22), 170 women initially assigned to the IFA group inadvertently received the MMN capsule either throughout pregnancy (n = 85) or during part of pregnancy (n = 85) before receiving the intended IFA capsule, and another 170 women initially assigned to the MMN group received the IFA capsule either throughout pregnancy (n = 78) or during part of pregnancy (n = 92) before receiving the intended MMN capsule. In this current analysis, we elected to include all of the children enrolled (including those whose mothers temporarily received the unintended capsule) because no children received any unintended supplements themselves, and the consumption of the unintended supplements by the relatively small number of women during pregnancy likely had little if any impact on child growth in the entire sample by 18 mo of age. With a total sample size of ∼1043 children (∼348/group) for whom anthropometric data were available at 18 mo of age, we had 97% power to detect an effect size of 0.3 between any 2 groups for any continuous outcome, and 80% power to detect an RR of >2.1 for stunting, >2.4 for underweight, >4.0 for wasting, and >2.0 for small head circumference. We posted the statistical analysis plan (www.ilins.org) before analysis. Statistical analysis was performed by using SAS version 9.3 on an intention-to-treat basis. That is, children were included in the analysis regardless of adherence to treatment. To address the protocol violation associated with the consumption of mislabeled capsules by some women during pregnancy, we analyzed our data by using 2 scenarios: in the first, intervention groups were based on the supplement that women were intended to receive when they were enrolled, and in the second, intervention groups were based on the supplement that women actually received when they were enrolled. The latter scenario is consistent with our previous publication reporting the birth outcomes of this trial (22). In addition, we performed a secondary analysis by using a 2-group comparison in which the IFA and MMN groups were combined. We summarized the background characteristics of women at enrollment as means ± SDs for continuous variables, or number of participants and percentages for categorical variables by using the assignment of women based on supplements received at enrollment to be consistent with our previous publication (22). We calculated a household assets index and a housing index as proxy indictors for household socioeconomic status as well as the household food insecurity access score (22). Adherence to maternal supplement intake during pregnancy and postpartum was calculated as the percentage of days from enrollment to delivery and from delivery to 6 mo postpartum, respectively, when the supplements were reportedly consumed by women. We calculated adherence during pregnancy and postpartum separately in case these differed, and compared them between the 3 groups by using assignment based on supplements received at enrollment. For infants in the LNS group, adherence was calculated as the percentage of days from 6 to 18 mo of age when LNSs were reportedly added to the child’s food. For growth measurements, we used the mean of duplicate measurements, or in the case of triplicate measurements, we used the mean of the 2 closest values. We determined LAZs, WAZs, WLZs, and HCZs as described by WHO (29), and considered values <−2.0 indicative of stunting, underweight, wasting, and small head circumference, respectively. We evaluated the impact of the intervention by comparing continuous and binary outcomes by 18 mo of age between the 3 groups for each of the 2 analysis scenarios mentioned above. In a secondary analysis to examine postnatal growth only, we compared the change in continuous growth outcomes from birth to 18 mo of age (defined as the difference between the 18-mo and birth values of the outcome). These analyses were accomplished by using linear (continuous outcomes) and logistic (binary) regression models (SAS; PROC GLIMMIX), with Tukey-Kramer adjustment for multiple comparisons. Along with the group comparisons, we estimated all pairwise differences in means (continuous outcomes) and RRs (binary outcomes) with their 95% CIs and P values. RRs were calculated by using Poisson regression (30). We performed these analyses first without any covariates (unadjusted) and then with prespecified covariates (adjusted) if the covariates were significantly associated with the outcome at a 10% level of significance in a correlation analysis. The potential covariates were maternal age, height, BMI, education, GA at enrollment, primiparity, season at enrollment, and baseline anemia status; proxy indicators for household socioeconomic status, namely assets, food insecurity, and housing scores; and child sex. For the binary outcomes, the covariate-adjusted percentages were generated by using the technique described by Kleinman and Norton (31). In a separate exploratory analysis, we examined the mean LAZ at 0 (birth), 3, 6, 12, and 18 mo (SAS) to determine whether the group differences that were present at birth persisted during the postnatal period. We evaluated potential interaction of intervention group with the prespecified maternal, household, and child variables listed above. When an interaction was significant (P < 0.10), we performed subgroup analysis by including an interaction term between treatment and the effect modifier in the ANCOVA or logistic regression model. For continuous effect modifiers, we created, using data from all participants, a linear regression model with which we predicted the values of the outcome at the 10th and 90th percentiles of the effect modifier distribution. Each effect modifier was considered separately in the models to reduce collinearity. Finally, we used logistic regression (binary variables) and ANOVA (continuous variables) models to evaluate the occurrence of SAEs in women during pregnancy and lactation and in infants from birth until exit at 18 mo of age. These SAEs were defined to include deaths, hospitalizations (at least overnight stay in the hospital because of illness), congenital abnormalities, and life-threatening conditions requiring an immediate hospital visit. The SAEs recorded from enrollment to delivery (including miscarriage and stillbirths) were previously reported (22).

N/A

• Dima Health