Impact of preconception and antenatal supplementation with myo-inositol, probiotics, and micronutrients on offspring BMI and weight gain over the first 2 years

The first 1000 days from conception to 2 years is a critical window for influencing later growth and body composition, including the future risks of underweight and obesity [1, 2]. Increasing evidence also implicates preconception influences on adverse offspring health outcomes [3], leading to calls for new initiatives to improve preconception health and care [4]. While increasing observational data implicate important roles for maternal obesity, glycaemia, and micronutrient status before and during pregnancy in increasing the risk of child obesity [5‐14], there are few randomised trials of preconception and pregnancy interventions that examine outcomes in early childhood [15, 16].

Several micronutrients have been related to offspring adiposity, for example, increased adiposity has been observed among offspring whose mothers were vitamin D deficient during pregnancy [8, 12, 14]. Likewise, B-vitamin deficiencies have been observed among mothers with diabetes [10, 11, 17, 18], which may contribute to increased offspring adiposity [9, 13]. Similarly, myo-inositol, a non-essential sugar alcohol involved in regulating glucose and lipid metabolism, has been postulated to counteract the effects of maternal dysglycaemia and dyslipidaemia on offspring adiposity [19], and probiotics have been proposed as beneficial in preventing gestational dysglycemia [20].

The rate of offspring postnatal weight gain may mediate associations between maternal micronutrient deficiencies and later obesity [8, 12]. Though there are several criteria used to define rapid infant weight gain, it is most frequently defined as an increase in weight of 0.67 standard deviations (SD) or more, which is equivalent to the upward crossing of one or more major percentile lines on a growth chart [21]. Other studies have used posteriori methods to describe patterns of infant weight gain [22, 23]. Regardless of the definition used, rapid weight gain in infancy has consistently been associated with increased cardiometabolic risk [22‐26].

In this context, the Nutritional Intervention Preconception and During Pregnancy to Maintain Healthy Glucose Levels and Offspring Health (NiPPeR) study provides an opportunity to examine the impact of maternal preconception and antenatal nutritional supplementation on offspring outcomes [27]. Women were recruited prior to pregnancy from three study centres (the UK, Singapore, and New Zealand) and were randomly allocated to receive a twice-daily nutritional beverage containing myo-inositol, probiotics, and additional micronutrients or a control beverage containing standard pregnancy micronutrients. The primary outcome of the trial was maternal glycaemia at 28 weeks of gestation; however, there were no differences between the intervention and control groups, including in the incidence of gestational diabetes [28]. Offspring postnatal weight gain and early childhood obesity were pre-specified secondary outcomes of the NiPPeR trial [27]. We aimed to determine whether preconception and antenatal supplementation with myo-inositol, probiotics, and additional micronutrients would optimise offspring body size and growth in the first two years of life.

We have demonstrated that maternal preconception and antenatal supplementation with myo-inositol, probiotics, and additional micronutrients were associated with a lower risk of rapid infant weight gain and obesity among offspring at 2 years. A posteriori trajectory modelling supported these findings, with more offspring in the intervention group assigned to a normal weight gain trajectory. As previous research has associated rapid infant weight gain with later obesity [21, 24, 34, 35], these findings suggest a protective effect of preconception and antenatal nutritional supplementation, which may have long-term benefits to the offspring. For example, high childhood BMI has been associated with an increased risk of cardiovascular disease and subsequent obesity in adulthood [36‐38], which is consistent with reported findings for a tendency for obesity to track from childhood [24, 37].

Myo-inositol supplementation from early gestation has previously been associated with reductions in the incidence of gestational diabetes, preterm birth, and excessive foetal growth [39‐41]. In the NiPPeR study, despite no differences in maternal glycaemia and incidence of gestational diabetes, there was a reduction in late preterm birth [28]. In the current study, results of offspring auxology were unchanged following adjustment for gestational age at birth and after exclusion of preterm births. Thus, the observed lower risk of rapid infant weight gain appears to be independent of the protective effects of the intervention on preterm birth [28]. Furthermore, as there were no intervention effects on gestational glycaemia, our findings also cannot be attributed to improvements in maternal glycaemic regulation [28].

In contrast to previous studies of antenatal myo-inositol supplementation, the NiPPeR maternal population was a self-selected cross-section from the community rather than a selected high metabolic risk group, and there were differences in the timing of commencement (i.e. preconception vs early pregnancy) and formulation (i.e. single- vs multi-nutrient) of the intervention. Nonetheless, the risk for rapid weight gain and obesity at 2 years was similarly reduced among metabolic risk groups (e.g. maternal overweight/obesity and gestational diabetes) (Additional file 2: Table S4). To our knowledge, no study of antenatal myo-inositol supplementation has assessed offspring growth, and the role of inositols in foetal growth and fat accretion remains poorly understood. However, it has been postulated that inositol is involved in the cross-talk across the maternal-placental-foetal axis to regulate foetal growth and development [19].

Other key micronutrients in the intervention have been associated with offspring adiposity. Low vitamin D status in late pregnancy has been associated with lower fat mass in offspring at birth, but subsequent elevated fat mass at 6 years, which is suggestive of rapid postnatal weight gain [12]. Similarly, in a study of over 60,000 Norwegian children, increasing maternal vitamin D intake was associated with lower weight gain trajectories and reduced odds of rapid weight gain and childhood overweight [8]. Pre-pregnancy maternal BMI was a modulating factor, and contrasting trends were observed among children of mothers with pre-pregnancy overweight compared with pre-pregnancy normal weight [8]. Furthermore, there is evidence to suggest a plateauing of the effect, where serum vitamin D was inversely associated with offspring adiposity, plateauing above serum levels of approximately 64 nmol/L [12]. While serum levels of vitamin D were increased among intervention mothers in the NiPPeR trial (manuscript in preparation), a previous study [42] found that supplementation with vitamin D from mid-pregnancy, when combined with calcium, iron, and folic acid (vs control supplement without vitamin D), did not reduce the incidence of gestational diabetes, preterm birth, or offspring anthropometry at birth or 1 year. The population studied had a high prevalence of vitamin D insufficiency, but the lack of an observed difference might have been related to the multiple-micronutrient format or the timing of the intervention [42]. B-group vitamins have also been related with maternal glycaemia [11, 17, 18, 43], with maternal vitamin B12 deficiency associated with increased insulin resistance and adiposity among offspring [9, 13, 44]. Thus, the beneficial effects of supplementation on infant weight gain among our cohort could be related to improvements in nutritional sufficiency, though the mechanisms are unclear.

In our cohort, rapid weight gain was relatively common, with 26% of infants experiencing rapid weight gain (> 0.67 SD) in the first year of life. Previously, a meta-analysis of 17 studies reported incidences of rapid infant weight gain in the range of 12 to 54%. Rapid weight gain was associated with 3.7 times increased odds of later overweight or obesity, though there was substantial heterogeneity in the time period examined and the age at outcome assessment [21]. Similar associations have been found when analysing a posteriori weight gain trajectories, with classes characterised by rapid weight gain having the highest risk of increased adiposity and cardiometabolic perturbations later in life [22‐26].

The weight gain trajectories between birth and 2 years identified in our cohort are similar to those previously described in contemporary paediatric cohorts [22, 23, 25]. In The Applied Research Group for Kids (TARget Kids!) study, the “Rapid Accelerating” class (characterised by increasing BMI SDS from 6 months) was associated with increased age- and sex-standardised cardiometabolic risk scores at 3 to 5 years, but few children (1%) were assigned to this class [22]. Similarly, among children in the Growing Up in Singapore Towards healthy Outcomes (GUSTO) study, four BMI SDS trajectories were identified in the first 2 years of life—“stable low”, “normal”, “stable high”, and “rapid BMI SDS gain after 3 months”—equivalent to our low, normal, high, and accelerating high classes. The stable high and rapid BMI SDS gain groups were associated with increased obesity at 5 years [23]. More recently, trajectories have been developed among the GUSTO cohort incorporating data from birth to 6 years [25]. Five BMI SDS trajectories were identified, with three stable trajectories (equivalent to the trajectories previously described), as well as two accelerating trajectories, with elevated foetal abdominal circumference and BMI acceleration immediately after birth or normal foetal growth with BMI acceleration after infancy. These accelerating trajectories were associated with increased abdominal fat, liver fat, insulin resistance, and hypertension at 6 years [25]. In GUSTO, ethnic differences were apparent between growth trajectories, with Malay and Indian offspring being more likely to be in the accelerating BMI gain trajectories [23]. We similarly observed increased assignment to the ascending classes among Singaporean offspring, though the NiPPeR study was not powered to explore ethnic differences and thus could not determine if patterns differed between Chinese, South Asian, and Malay Singaporean offspring. Together, these findings suggest that offspring in our trial assigned to the two ascending classes may be at an increased risk of later cardiometabolic perturbations, and this may differ by ethnicity.

There are several strengths to this study. The NiPPeR study is a multinational, randomised controlled trial with extensive ongoing prospective data collection commencing from the preconception period. Anthropometric data were collected frequently in the first 2 years starting from birth, reducing the risk of spurious findings. Cross-sectional and longitudinal data were considered in the analyses, including multiple approaches for modelling nonlinear growth trajectories [25, 32]. Although each statistical method has its limitations, for example, LCGA requires subjective decisions about the number and placement of knots and the optimal number of classes to be retained [32], results were consistent across the statistical methods employed.

Limitations to the current study include that anthropometric measurements cannot distinguish between fat- and fat-free masses; therefore, it is unclear if the observed differences in weight gain are attributable to increased fat mass deposition or higher fat-free mass. Nonetheless, a study among Ethiopian infants attributed catch-up weight gain to fat mass [45], with higher fat mass accretion in infancy associated with increased adiposity at 5 years [46]. Further work is required to establish if the weight gain trajectories observed in the NiPPeR cohort are associated with fat or fat-free masses and if the intervention effects on early postnatal weight gain are associated with reduced adiposity later in life, as well as independent validation of the findings. Furthermore, the intervention’s multinutrient formulation limits the ability to determine the effects of specific nutrients and, thus, potential mechanisms.