BRAzil magnesium (BRAMAG) trial: a double-masked randomized clinical trial of oral magnesium supplementation in pregnancy

The World Health Organization (WHO) estimates that each year 15 million children are born preterm before 37 weeks’ gestation [1]. Prematurity is a leading cause of death in the neonatal period, and the second most common cause in children under age five years [1‐4]. Preterm has immediate and long-term implications for the child, including cerebral palsy, learning difficulties and visual and hearing impairment [5‐8]. The ensuing social and economic costs are great [9‐11].

Brazil is among the top-10 countries with the highest number of premature births [1]. For example, of the nearly 3 million births registered in the country in 2015, about 10.8% were preterm [12]. In that same year, prematurity was the main cause of death in children under the age of five years [13]. Furthermore, in 2016, a national study identified that the North and Northeast regions of Brazil had the highest rates of preterm birth, at 13.0 and 12.9%, respectively [14]. Factors associated with prematurity in Brazil are higher social vulnerability -- low income, adolescent pregnancy and limited high school education – as well as inadequate prenatal care, a high rate of cesarean delivery and preeclampsia [15].

Magnesium deficiency has been implicated in adverse maternal and perinatal outcomes. Moreover, magnesium deficiency has been associated with an increased risk for gestational and adverse perinatal outcomes, such as gestational hypertensive syndromes, leg cramps and preterm birth [16‐21]. Currently, there is no consensus about the daily requirements of magnesium during pregnancy [21, 22], with recommendations vary between 220 mg to 500 mg per day [22]. A recent meta-analysis of 10 randomized clinical trials (RCT), comprising 9090 pregnancies, evaluated oral magnesium supplementation during pregnancy [23]. Therein, oral magnesium did not significantly reduce the risk of preeclampsia, perinatal mortality or small for gestational age birthweight (SGA). Magnesium was associated with higher newborn Apgar scores and less hypoxic-ischemic encephalopathy. However, as only two of the 10 RCTs were considered to be of higher quality, the authors concluded that there is insufficient evidence as to whether oral magnesium supplementation in pregnancy is beneficial to mother or fetus [23]. Additionally, the effect of magnesium supplementation in the prevention of prematurity has had conflicting results, perhaps partly due to different the different forms of magnesium assessed in those trials [24‐32].

Magnesium citrate has a high solubility in water, favoring greater intestinal tract absorption and bioavailability compared to other preparations [33]. Accordingly, the current study compared magnesium citrate with placebo in the prevention of adverse perinatal and maternal outcomes among women at higher risk.

Due to a regional Zika virus outbreak, the intended recruitment of 2000 women in the magnesium arm and 1000 women in the placebo arm was not achieved.

Of the 1031 women who were eligible according to study entry criteria, 108 declined to participate and 12 were excluded for other reasons (Fig. 1). In total, 911 women were randomized: 449 to the magnesium citrate arm and 462 to the placebo arm. Of these, 31 (6.9%) and 25 (5.4%), respectively, did not complete the trial, because they voluntary abandoned prenatal care at IMIP (19 in the magnesium group and 21 in the placebo group), or moved to other services (12 in magnesium group and 11 in placebo group). Another 11 pregnant women in the intervention group and 15 in the placebo group had symptoms that led them to cease taking the study medication (Fig. 1).

Analyses comprised 407 women who received magnesium citrate and 422 who receivedplacebo. Intention-to-treat principle was used. The mean (SD) gestational age at study entry was 15.1 (3.6) and 15.6 (3.8) weeks, respectively. The two groups were similar according to their socio-demographic and clinical characteristics (Table 1). At trial entry, 194 women in the magnesium arm (47.9%), and 219 women in the placebo arm (52.1%) had a low serum magnesium concentration below 1.8 mg/dL.

The perinatal composite outcome occurred among 75 (18.4%) and 76 (18.0%) pregnancies, respectively (adjusted odds ratio [aOR] 1.10, 95% CI 0.72–1.68) (Table 2). The most common event within the perinatal composite outcome was preterm birth < 37 weeks’ gestation (9.3% vs. 9.0%, respectively) (adjusted OR 1.02, 95% CI 0.64 to 1.65).

The maternal composite outcome occurred among 49 (12.0%) and 41 (9.7%) pregnancies, respectively (aOR 1.29, 95% CI 0.83–2.00) (Table 2). The most common events within the maternal composite outcome were preeclampsia < 37 weeks’ gestation (5.9% vs. 4.7%) and severe gestational hypertension < 37 weeks’ gestation (4.9% vs. 4.5%) (Table 2). Of note, the risk of placental abruption was lower in the magnesium group (9 events [2.2%]) compared to the placebo arm (21 events [5.0%]), equivalent to an adjusted OR of 0.43 (95% CI 0.20 to 0.95).

Apparently, in this study, oral magnesium citrate supplementation was not superior to placebo in the prevention of an array of adverse perinatal or maternal outcomes.

The 300 mg daily dose of magnesium citrate used herein approximated that recommended in pregnancy [22, 34]. A notable proportion of pregnant women (around 50%) had low serum levels of magnesium. Studies in India, Nigeria and Iran identified hypomagnesemia in pregnancy, at respective rates of 43.6, 16.2 and 13% [20, 35, 36]. Among 52 healthy pregnant women living in São Paulo, Brazil, despite having serum and red cell magnesium concentrations within normal limits, dietary intake was inadequate, and 39% had low urinary magnesium excretion [37]. An association between low maternal serum magnesium levels and preterm labour has been described previously [38].

The current findings are in keeping with some previously completed randomized clinical trials of oral magnesium supplementation for the prevention of preterm birth. Martin et al. studied 54 women at high risk of preterm birth, and observed that 1 g of magnesium gluconate, administered four times daily, was not effective for preventing preterm birth [27]. A double-masked RCT by Sibai et al. evaluated 374 high- risk pregnant women, and also did not show a significant difference in the risk of preterm birth with oral magnesium supplementation [25]. Arikan et al. studied 530 low risk pregnant women, and did not observe any reduction in the risk of preterm labour with oral magnesium citrate, 365 mg daily, started before 18 weeks’ gestation [39]. In a quasi-randomized trial of 568 pregnant women receiving magnesium aspartate or placebo, there was a significant reduction in preterm labor in the magnesium arm [24]. Another recent study of 180 pregnant women also found a significant reduction in preterm labor with magnesium in women with hypomagnesemia, using women with a normal serum magnesium concentration as controls [40].

The rate of preterm birth in Northeast Brazil is 12.9%, with a national rate of 11.1% [41]. Herein, the 9.2% rate of preterm birth was a little lower than expected, perhaps due to enhanced prenatal care offered within the context of a clinical trial, and the selection of more engaged study participants. There was only one stillbirth observed herein, and 10 neonatal deaths among controls vs. 4 deaths in the magnesium group. While a non-significant finding, future studies might consider the potential importance of this outcome related to magnesium supplementation. One study evaluated the risk of newborn hypoxic ischemic encephalopathy among 4082 pregnant mothers assigned oral magnesium stearate or placebo [30]. Although the rate of medication adherence was about 25%, no reduction in hypoxic ischemic encephalopathy was seen therein. Perinatal mortality was also higher in the treatment group (17 deaths) compared to the placebo arm (7 cases), but most deaths were attributed to congenital anomalies [30]. A recent meta-analysis of magnesium supplementation in pregnant women included 10 trials, with two judged to be of high quality [23]. Analysis restricted to these two trials showed no significant difference between the oral magnesium and control groups for perinatal mortality, SGA or preeclampsia [23].

A major study limitation herein was the lack of achievement of the target sample size. One major reason was an outbreak of microcephaly associated with Zika virus infection [42], which posed a major distraction to participant recruitment and the availability of resources. For example, there was concern among women about the use of medications in pregnancy, given that the etiology of the microcephaly outbreak was unclear at that time. In a post hoc calculation, using the currently achieved study sample size, to detect the expected 22% relative risk reduction in the primary perinatal composite outcome (i.e., an event rate of 14% in the magnesium group and 18% in the placebo group), our study had only 34.9% statistical power. Even so, the current data do provide important information about the efficacy of magnesium citrate supplementation in pregnancy, and the need to conduct further research in this area.

Oral magnesium supplementation in pregnancy among higher-risk women does not seen to reduce preterm birth, or adverse perinatal or maternal outcomes.