Interpretation
Contrary to our hypothesis, we found no additional effect of having two overweight or obese parents on infant birthweight outside what was seen if one parent was overweight or obese. Evidence from our rodent model of obesity also suggests that the effect on infant birthweight may unlikely be additive, but an accumulation of both the negative maternal and paternal phenotypes 20. This seems to be evident in our human cohort where infant birthweight increased from 3.13kg 95%CI=[3.03, 3.23] to 3.44kg 95%CI=[3.31, 3.56] in normal weight mothers compared with obese mothers when fathers were of a normal weight. When fathers were obese, this increase was much smaller (3.33kg 95%CI=[3.20, 3.45] to 3.41kg [3.32, 3.51]). This is likely because infants born to fathers who are obese already start out heavier (~200 g) and therefore, only require a small additional increase in size to match the effect of obese mothers. Whilst we saw no additive effect of combined parental BMI on infant birthweight, the effects maybe may be present as the infants grow. For instance, in another rodent model, insulin resistance and liver steatosis were greatest in offspring when both parents were fed a high fat diet prior to and during gestation, compared to just one parent21. In humans, Rath et al.,22, found that parental obesity was the strongest predictor of offspring adult BMI. These data suggest that the combined effect of having two obese parents on infant programming may manifest later in life.
There is a large body of literature demonstrating the impact of maternal BMI on infant birthweight including LGA 23, and there is some suggestion for paternal BMI also having an impact on infant birthweight 9. Unfortunately, much of the literature on paternal BMI included self-reported paternal height and weight from the mother, or, collection during pregnancy, at birth, or when the child was a toddler, rather than preconception 24, 25. Furthermore, studies that have assessed preconception paternal BMI have not always adequately controlled for maternal and other parental cofactors, and therefore, the results are currently conflicting. For example Chen et al., 26 found that paternal overweight and obesity only influence male infant birthweight, with a 1 unit increase in paternal BMI associated with a 19.5 g increase in infant birthweight, while Noor et al .,27 found that fathers with a BMI greater than 25kg/m2 increased infant birthweight in both sexes (z score, 0.38 [0.91] vs 0.11 [0.96]. In contrast, three other studies found no effect of paternal BMI on infant birthweight 28-30. Interestingly, when assessing the extreme ends of infant birthweight (SGA or LGA), McCownet al., 31 found that obese men were 1.5 times more likely to father SGA infants, while Yang et al.,32 found that overweight and obese men were 1.3 times and 1.9 times respectively more likely, to father an LGA infant. Similarly, in an ART cohort following frozen embryo transfer Ma et al., 19 found that men who were overweight or obese had an increased odds of having a LGA infant (OR=1.43; 95%CI=[1.27, 1.63] and OR=1.36; 95%CI=[1.04, 1.79] respectively). In our study, we found no evidence for an association between paternal overweight and obesity and SGA or LGA infants (<10th and >90th percentiles), although the median birthweight of infants were higher with increased paternal BMI (7.3 g for every 1 unit increase in paternal BMI). The lack of consensus in the reported effects of paternal overweight and obesity on infant birthweight highlights the necessity for further adequately controlled cohort studies. Nevertheless, animal models of male obesity support findings for increased infant birthweight 33-35.
The mechanism for transmission of altered infant birthweight from increasing paternal BMI is likely due to a combination of genetic and epigenetic factors delivered by sperm to the egg at fertilisation36, 37. A number of genes are known to play a part in the heritability of weight 38, 39, however these genetic loci do not fully account for the transmission. A number of studies in animal models and humans directly show a link between paternal obesity at conception, sperm epigenetic changes (non-coding RNAs and DNA and histone methylation) and altered fetal phenotypes27, 40-44, indicating that the paternal effect goes beyond that of a shared living environment, with preconception factors able to influence the health of subsequent offspring.
Our data shows that infants born from mothers or fathers of increasing BMI start their growth trajectory heavier than those infants born to normal weight mothers or fathers. This is of concern as birthweight has been reported to play an important role in the establishment of adolescent and early adulthood BMI 22, 45. For instance, evidence from the Early Childhood Longitudinal Study in the USA, found that LGA infants made up 1/3 (36%) of all children who were obese by age 14 years 45 and data from the RAINE cohort in Western Australia, Australia, found that both maternal and paternal preconception BMI were strong predictors of childhood, adolescent and adulthood obesity 22. If obesity aggregates within families, then a focus on preconception planning for family units is recommended. In Australia, there are no primary male preconception health-care initiatives 46. While Healthy Male (Andrology Australia) does provide education on the reproductive health of men, focusing on fertility, sexuality and fathering, and the Australian men’s health policy addresses various issues related to sexual problems, neither of these primarily focus on preconception health 46. Further, missing data for paternal preconception BMI in our study was nearly double that of missing maternal BMI (33% vs 18%). While some of this may be due to the lack of males in preconception care appointments, it highlights the dogma that mother’s preconception health is a key focus rather than fathers. Therefore, it is recommended that preconception health messages focus on ‘healthy couples’, emphasising the need to improve lifestyle for the family unit prior to pregnancy.