Associations of infant milk feed type on early postnatal growth of offspring exposed and unexposed to gestational diabetes in utero

In this prospective Asian birth cohort study, we have demonstrated varied associations of breastmilk intake on early postnatal growth in offspring exposed and unexposed to GDM in utero. We noted that offspring of mothers without GDM with greater breastmilk intake (i.e. ≥4 months) were associated with decelerated growth compared to those who with no breastfeeding. However, the reduced weight gain in the first year conferred by greater breastmilk intake in non-GDM children was not observed in GDM children.

Our findings on observed slower growth from greater breastmilk intake amongst non-GDM mothers are a well-described phenomenon and have been widely documented in literature. Griffiths et al. [9] similarly reported that infants of mothers without GDM who did not receive breastmilk grew faster than those whose mothers initiated breastfeeding, as did those who breastfed for 4 months or longer. A cohort study of randomly selected healthy newborns in Denmark and Iceland showed that exclusive breastfeeding beyond 2 months of age was related to lower weight gain from 2 to 6 months as well as from 6 to 12 months [10]. A recent study on the Gemini cohort of 4680 infants also showed that infants breastfed for longer periods (>4 months) was independently associated with lower growth velocity by 6.8 % [27]. Taken together with our study findings, it supports the notion that periods of long and exclusive breastfeeding may have a protective effect on development of obesity later in life. Not all studies, however, have reported such similar results. A cluster randomized trial of a breastfeeding promotion intervention modelled on the WHO/UNICEF Baby-Friendly Hospital Initiative showed that prolonged and exclusive breastfeeding accelerated weight and length gain of the infants in the first few months with no detectable deficit by 12 months old [28].

Early animal studies have suggested that milk derived from mothers with GDM may impart metabolic consequences to their offspring. It has been reported that control offspring who were fed milk from dams with GDM showed complex “malprogramming” of hypothalamic neural circuits that are critically involved in the regulation of food intake, body weight, and metabolism [29]. Longitudinal studies in humans also echoed similar observations, where breastmilk feeds from diabetic mothers during the first week of life was associated with greater relative weight and risk of overweight at 2-years, compared to offspring of diabetic mothers who were fed banked donor breastmilk [17, 18]. Other studies, however, such as that by Crume et al. [19], have reported contrasting findings where it was shown that adequate breastfeeding (≥6 breastmilk months) reduces overall body size and BMI growth velocity in the first 9 months of life amongst offspring of diabetic pregnancies. Another study also reported that breastfeeding conferred similar protective effects against overweight at 9–14 years of age in offspring of both non-diabetic and diabetic women [20]. It is important to note, however, that these studies classified type-I diabetes and GDM into a single category, which may have explained the inconsistency in findings with our study due to the differences in aetiologies of both diabetic sub-types. Our findings thus provide critical insights into this area of research, given the lack of understanding of the biochemical impact of breastmilk from GDM mothers on infant growth.

The plausible mechanisms underlying the varied effects of breastmilk intake on early postnatal growth in offspring exposed and unexposed to GDM in utero are likely multiple. In GDM mothers, researchers have postulated that differences in breastmilk constituents of diabetic and non-diabetic mothers, such as increased glucose or insulin concentrations in breastmilk of diabetic mothers may contribute to increased growth rates during early infancy [30, 31]. Concentration of ghrelin in breastmilk of GDM-lactating women has also been reported to be lower when compared to non-diabetic control samples [32]. However, there is positive significant correlation between levels of active ghrelin in 4th month breastmilk and weight gain [33]; hence, the significance of this reduced ghrelin in breastmilk of GDM mothers in our observation is uncertain. Milk from diabetic mothers may also contain more inflammatory cytokines (e.g. TNF-α, IL-6) which mimic signalling pathways characteristic of dysfunctional adipocytes of metabolic syndrome [34]. Moreover, Kjos et al. [35] had earlier demonstrated that abnormal glucose metabolism still persists postpartum amongst women with GDM, further suggesting that continued exposure to altered fuels through breastmilk may bring about consequences to offspring growth. Plagemann et al. [36] had also proposed that milk originating from diabetic mothers may have an early obesogenic effect on infant weight gain that decreases with time. Taken together, these studies suggest that differences in breastmilk composition may explain our observations of increased early growth with greater breastmilk intake in offspring of GDM mothers. Despite this, it is important to note that there may still be positive effects of breastfeeding on reducing later adiposity in offspring of diabetic mothers, as highlighted by Crume et al. [19] and Mayer-Davis et al. [20]. It has been suggested that the benefits of breastfeeding in offspring of diabetic mothers may only be observed if breastfeeding is continued beyond a certain period where breastmilk composition would have normalized over time [36]. The Kaulsdorf Cohort Study in Germany had highlighted that intake of donor breastmilk from metabolically healthy mothers prevented increased early weight gain in offspring of diabetic mothers [37], further indicating that breastmilk of a normal composition, which would result from good metabolic control of women with diabetes [38], may be beneficial to prevent future obesity risk in offspring of diabetic women. This begets the question whether proactive intervention to achieve better glycemic control in the early postpartum period of GDM mothers would help to accelerate the normalization of the breastmilk and reduce the obesogenic effect, which poses an interesting hypothesis that should be examined properly in an interventional trial.

Strengths of our study include the prospective design with high follow-up rate, along with the study of Asian ethnic groups. Given the paucity of existing data on prenatal metabolic exposures and infant growth in Asian populations, our study thus provides useful and informative data on this relationship. This study, however, is with limitations. Due to the observational nature of our study, we cannot fully rule out the possibility that residual confounding by parental attributes or family environment may affect the observed associations, as breastfeeding is a behaviour that is self-selected and women are usually not randomized to breastfeed. As with most studies on breastfeeding and infant growth, it is largely observational and hence subject to potential confounding. Socio-economic status presents as an important confounder, as mothers who are more educated tend to be more “nutrition-conscious”, more likely to breastfeed and less likely to feed poor quality diets postweaning [39, 40]. Other important potential confounding factors include maternal BMI, which is generally associated with shorter durations of breastfeeding [41], as well as maternal age, which is generally associated with greater exclusive breastfeeding [42], all of which presents a potential bias. The associations observed in our study findings have been controlled for and are independent of these potential confounders.

In conclusion, our study findings have demonstrated varied associations of breastmilk intake on early postnatal growth in offspring exposed and unexposed to GDM in utero. Whilst offspring of non-GDM mothers exhibited reduced weight gain in the first year of life conferred by greater breastmilk intake, the same effect was not observed in GDM children. It remains to be seen if our observed associations of breastmilk intake on accelerated growth amongst offspring exposed to GDM in utero have independent long-term effects in adolescence and adulthood in our cohort.