Study design

We will conduct an individually randomized, quadruple-blind (blinded trial participants, investigators, data collectors and data analysts), superiority trial of daily antenatal MMS supplementation containing 60 mg elemental iron or 45 mg elemental iron as compared to MMS containing 30 mg elemental iron. This trial protocol was written in accordance with the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) checklist (see Additional file 1).

Study setting

In Tanzania, an estimated 56% of pregnant women are anemic based on the 2022 Demographic and Health Survey [3]. Two previous trials of MMS versus IFA, in which both groups received 60 mg iron, have been conducted in the country, and both indicated the beneficial effects of MMS [20, 21]. Tanzania is currently considering switching to MMS as the standard of care, but there are concerns about reducing the iron dose from 60 mg in the current IFA to 30 mg in MMS. The study will be conducted at selected public antenatal care clinics in Dar es Salaam, Tanzania. The study clinics currently provide all pregnant women with IFA containing 60 mg iron free of charge as the standard of care. The same study clinics were part of a recent non-inferiority trial of low-dose calcium supplementation in pregnant women [22].

Eligibility criteria and recruitment

The trial flow diagram is shown in Fig. 1. Research staff will assess the eligibility criteria for pregnant women who present for antenatal care at study clinics. The trial inclusion criteria are (i) pregnant women attending their first antenatal care visit, (ii) less than 20 weeks of gestation based on the last menstrual period (LMP), (iii) adult ≥ 18 years old, (iv) intending to stay in the study area until 6 weeks postpartum, and (v) provide written informed consent. Trial exclusion criteria are (i) severe anemia defined as a hemoglobin concentration < 8.5 g/dL per Tanzania standard of care; (ii) sickle cell disease including genotypes HbSS, HbSC or hemoglobin C disease (HbCC); and/or (iii) concurrent enrollment in another nutritional clinical trial. At the screening visit, participants who consent will have a finger prick for the collection of capillary blood that will be used to quantify hemoglobin concentration with the point-of-care HemoCue Hb 301 system (HemoCue AB, Ängelholm, Sweden). The HemoTypeSC test (Silver Lake Research Cooperation, Irwindale, USA) will be used for point-of-care hemoglobinopathy screening. Participants will also be asked to provide written consent for the storage of their data and blood specimens in future studies.

Interventions

Pregnant women will be randomized to one of three trial regimens (i) MMS that contains 15 micronutrients, including 30 mg elemental iron (standard UNIMMAP formulation), (ii) MMS that contains 45 mg elemental iron plus the standard UNIMMAP formulation for the other 14 micronutrients, or (iii) MMS that contains 60 mg elemental iron plus the standard UNIMMAP formulation for the other 14 micronutrients. The only difference between the randomized groups will be the amount of iron contained in the MMS supplements. Iron will be given in the form of ferrous sulfate for all three groups. To ensure blinding, the three MMS supplements will have the same appearance, color, odor, taste, size, and weight. The trial regimen will be manufactured by DSM Nutritional Products South Africa (Gauteng, South Africa).

All participants will be counseled to take one MMS (with IFA) tablet orally per day from randomization until delivery. Participants will receive a 35-day supply of MMS in blister packs at the randomization visit and at each subsequent monthly pregnancy visit (Additional file 2). Multiple strategies will be used to enhance participant adherence to the MMS regimens. At each pregnancy visit, research staff will take a pill count of tablets returned in the regimen blister packs. Study staff will then counsel participants on MMS and provide advice on overcoming side effects and other potential barriers. Text message reminders for adherence and upcoming study visits will also be sent to study participants. Participants who report they plan to travel outside of the study area may be given extra blister packs of the regimen to ensure daily regimen supply.

Concomitant care provided during the trial

At each trial site clinic, all participants will be provided with the standard of care throughout the study according to the national guidelines for antenatal care in Tanzania. Participants will have access to study clinics for post-trial care through the routine health system, and in the case diagnoses such as hypertensive disorders of pregnancy, red blood cell disorders, or depression are diagnosed through study procedures, we will refer participants to appropriate standard care. Any woman who is identified to be severely anemic (Hb < 8.5 g/dL) during the study will be discontinued from taking study supplements and referred for treatment according to the Tanzanian antenatal guideline for the management of severe anemia in pregnancy. We will continue to follow the women with severe anemia per the standard visit schedule according to the protocol, but they will not receive the trial regimen.

Assignment of interventions: allocation and blinding

Pregnant women will be randomized in a 1:1:1 ratio to the three trial groups. The allocation sequence will be generated by two non-study staff at George Washington University through computer-generated randomization lists that will be stratified by the study clinic and block randomization (block sizes of 9). The two non-study staff hold the randomization list codes until completion of the primary trial analysis or as requested by the Data and Safety Monitoring Board (DSMB). An independent study pharmacist will privately prepare regimen blister packs with participant IDs based on the randomization lists for each clinic. At the randomization visit, research staff will assign pregnant women to the next available participant ID, which corresponds to a set of pre-labeled blister packs. The trial statistician and data analysts will be blinded to the treatment allocation for the primary statistical analysis. Therefore, the trial is quadruple blind because all trial participants, investigators, outcome assessors, and trial data analysts will be unable to determine the randomized group for any individual participant and will not be able to determine participants who are in the same randomized group. Furthermore, the randomization procedures will ensure complete allocation concealment.

Sample size

We will perform three statistical tests of superiority for the primary outcome: (i) MMS 60 mg iron versus MMS 30 mg iron; (ii) MMS 45 mg iron versus MMS 30 mg iron; (iii) MMS 60 mg iron versus MMS 45 mg iron. Therefore, to account for multiple testing, the α was set to 0.01667 (0.05/3). Table 2 presents a summary of the outcome prevalence assumptions, relative risks, and power calculations for the trial. Based on data from a prior calcium trial conducted in Tanzania, we expect the prevalence of the primary outcome of maternal third-trimester moderate or severe anemia to be 25% in the MMS 45 mg iron group. A meta-analysis of iron trials and cohort studies suggests that there is a 0.88 relative risk for anemia per 10 mg increase in iron dose [12]. Therefore, we assumed a relative risk of 0.82 for comparisons with a 15 mg difference in iron (MMS 45 mg iron vs MMS 30 mg iron and MMS 60 mg iron vs 45 mg iron) and a relative risk of 0.67 for the comparison with a difference of 30 mg iron (MMS 60 mg iron versus MMS 30 mg iron). Based on these assumptions, we require third-trimester hemoglobin data from 1,808 women per arm. Assuming a conservative 15% of pregnant women will have missing third-trimester hemoglobin data (including 5% fetal loss and 10% missing blood samples), the total sample size for the trial will be 6381 (2,127 participants in each of the three randomization groups). R pwr package version 1.3 was used for sample size calculations [23].

Participant timeline

An overview of screening, randomization, and follow-up is presented in Fig. 2. Study visits include a randomization visit and pregnancy study visits every 4 weeks until delivery. Regardless of pregnancy outcome, we will also conduct a visit around the end of the pregnancy and a postnatal visit scheduled after 42 days postpartum. Participant retention will be promoted at clinic visits and through phone calls and home visits.

Pregnancy study visits

Pregnant participants will have a study visit at the time of randomization and monthly follow-up visits until delivery. Starting at 32 weeks of gestation, all participants will also receive weekly phone calls to check their pregnancy status. Participants will be asked to contact the research team at the time of labor onset. All women will receive an ultrasound assessment of fetal biometry around the time of randomization. All ultrasound images will be electronically stored, and a random 10% will be assessed for quality by the study obstetrician and the ultrasound quality control team. Participants will also receive a malaria rapid diagnostic test (RDT) at the first antenatal care visit, the time of randomization, as the standard of care. At pregnancy follow-up visits, participants who report a history of fever during the last month or have a body temperature of > 37.5 ℃ measured at the study visit will also have a malaria RDT performed. At all pregnancy visits, study physicians will perform a clinical exam and treat all comorbidities per Tanzanian standard of care. Blood pressure will be assessed with electronic blood pressure monitors and standardized procedures. Participants will also provide urine samples at each pregnancy visit, and dipsticks will be used to assess proteinuria. Nurses will measure maternal height with stadiometers at the randomization visit. Maternal weight will be taken on digital weighing scales, and the mid-upper arm circumference (MUAC) will be taken with a measuring tape at randomization and each pregnancy visit. A 24-h diet recall will be conducted at the first pregnancy follow-up visit and at 32 weeks of gestation. The Patient Health Questionnaire (PHQ)−9 and FACIT Fatigue Scale [24, 25] will be administered to assess symptoms of depression and symptoms of anxiety and fatigue, respectively, at randomization and at 32 weeks of gestation.

Labor and delivery study visit

Clinic records and interviews with the clinic staff and participants will be used to ascertain labor and delivery information. Participants who deliver outside of the study area will be reached by phone by the research staff to obtain relevant information from the mother and/or facility records. Maternal blood pressure and proteinuria will be assessed by clinic or study staff or collected from clinic records. Infant length and weight will also be taken by study staff or recorded from facility records.

Postpartum study visit

Women or women/infant pairs will have a study visit at 6 weeks postpartum (42 days) and will be discharged from the trial after completion of this visit. At the postpartum visit, research staff will assess maternal and infant morbidity and hospitalization history. Nurses will also collect weight and blood pressure from mothers. Study nurses will assess infant feeding, infant weight with a digital infant balance scale to the nearest gram, and infant length to 1-mm precision with a rigid length board with an adjustable footpiece. Study nurses will measure infant head circumference and MUAC with a flexible measuring tape. All infant anthropometric measurements will be recorded in duplicates. A verbal autopsy will be conducted to ascertain the cause of infant death [21].

Data management

All data will be entered into an electronic data capture system developed in Tanzania, and the program will have in-built skip patterns and range check validations for each variable. All identifiable electronic data will remain in Tanzania and will be stored on secure local servers that are accessible only by the respective study data teams and investigators. Data from the trial will be stored indefinitely on servers in Tanzania and the USA.