Antenatal Diet Quality Modifies the Effect of Multiple Micronutrients (MMS) vs Iron-Folic Acid (IFA) on Risk of Low Birth Weight (LBW) in Bangladesh

Objectives: In a large cluster-randomized trial in rural area in North-West Bangladesh, MMS vs IFA reduced risk of LBW by 12%. In meta-analyses, pregnant women (PW) with anemia appear to benefit more from MMS than non-anemic PW, suggesting diet may modify MMS's effect on birthweight. However, this has not previously been studied. Methods: Seven-day food frequency (FFQ) data from 19,136 PW in late pregnancy with singleton live births participating in the JiVitA-3 randomized controlled trial of MMS vs. IFA supplementation in Bangladesh were analyzed. A dietary diversity score (DDS) was created using the Minimum Dietary Diversity-Women (MDD-W) food groups. Because 7-day FFQ data were used, each food category score used consumption of ≥ 4 times in the past week rather than ≥ 1 time. Risk ratios and 95% confidence intervals for the effect of MMS on LBW in the presence of the continuous DDS and with the DDS x MMS interaction were estimated using log-binomial regression. Point estimates of the effect of MMS vs IFA on LBW across all 10 levels of the DDS were generated and plotted. A sub-analysis examined whether intake of component food groups also showed effect modification. Multivariable-adjusted models included maternal age, parity, BMI, SES, education, infant sex, and Ramadan fasting data collection. The likelihood ratio test was used to test the statistical significance of interactions. Results: The median DDS score was 3 (IQR: 2-4), and 18.6 percent of PW had a score ≥ 5. In the fully adjusted model, MMS's effect on LBW had a linear, inverse relationship with dietary diversity (p for interaction = 0.048) – the effect was greatest at the lowest levels of dietary diversity [DDS= 1, RR LBW: 0.83 (0.78-0.88)] and decreased at the highest levels [DDS = 10, RR: 1.03 (0.97-1.10)]. Dairy intake (median: 0 servings/week, IQR: 0-4) revealed a similar interaction with MMS [at 0 servings/week, RR LBW: 0.86 (0.83-0.89), at 21 servings/week, RR: 1.08 (1.05-1.11), p for interaction = 0.06]. Conclusions: We tested whether the effect of MMS on LBW varied by diet quality during pregnancy in rural Bangladesh and confirmed that the effect size was highest among those with low dietary diversity; dairy consumption among all food groups showed the strongest effect modification. Pregnant women in low dietary diversity settings may benefit more from supplementation with MMS. Funding Sources: Bill & Melinda Gates Foundation.