Rachel Freathy and Mark McCarthy

Birth Weight

With colleagues in the EGG (Early Growth Genetics) Consortium, we have published a paper in Nature describing genetic analyses of birth weight. We have written about our main findings in a separate blog post. Here, we consider in more detail the genetic influences of mother and child, and how we can better understand their respective roles in early growth and later life disease.

Birth weight is influenced not only by the baby’s genes (inherited from mother and father), but also by the mother’s genes (that the baby may or may not not inherit), because those genes influence the womb environment.
In our study, we analysed information on birth weight in relation to genetic differences between more than 150,000 study participants. We showed that many of the same genetic differences that influence a person’s birth weight also influence their susceptibility to later life disease. Our initial calculations suggested that genetics makes a large contribution to the link between birth weight and adult diseases such as type 2 diabetes and high blood pressure. On the surface, that is a straightforward conclusion: a primary role for genetics might lead to a downplaying of the role of the maternal environment in programming the developing baby’s later risk of disease (see our other post). But the reality is more complex: to build a full picture of what is going on, we need also to consider the essential contribution made by the mother to the growth of the baby.

A baby’s growth is influenced by its own genetic profile, but crucially, it is heavily influenced by the gestational environment that the mother provides, and that maternal environment is in turn influenced by the mother’s genetic profile. Since each baby inherits half of its genes from its mother, the mother and baby’s genetic profiles will partly overlap with one another (see Figure 1 for summary of relationships). So, in our study, when we analysed the genetics of the study participants, we were also capturing some information about their mothers’ genetics. The key question for us was: what proportion of the genetic variation influencing birth weight (and later life disease) is acting directly, having been inherited by the child, and what proportion is acting indirectly, through the gestational environment, under the influence of the mother’s genes?

Figure 1: Summary of key relationships: Black arrows reflect purely genetic mechanisms, connecting parental to offspring genes, with shared impacts on fetal growth and later disease risk. Red arrows reflect the DoHAD programming model whereby intrauterine environment influences early growth and leads to long term programming effects on future health. The green arrows represent the impact of maternal genotype on intrauterine environment, and the blue arrow indicates that the baby’s birthweight is a readout of growth, but not of itself on the causal pathway to future disease risk.
Possible scenarios leading to an observed association between high/low birth weight and risk of later disease (these are not mutually exclusive):
(i) Developmental programming (red arrows)
(ii) Direct genetic influences on birth weight are the same as those influencing later disease (black arrows)
(iii) Maternal genetic effects influence baby’s growth via the womb environment, and those same genes, if inherited by the fetus, influence disease risk directly (thicker green, red and black arrows)

To unravel the genetic effects of child and mother on birth weight, we included mothers in our analyses and found evidence for a greater genetic contribution from the child.
In our study, we would ideally have had access to genetic information from the mothers of all study participants. However, studies on such a scale are not yet possible. We began to tackle the question using available resources.

First, we analysed genetic variations throughout the genomes of 4382 mother-child pairs from the UK-based Avon Longitudinal Study of Parents and Children (ALSPAC) using a “maternal-genome-wide complex trait analysis” (m-GCTA). This analysis enabled us to estimate the overall contribution to birth weight variation made by the mother’s vs. the baby’s genetics. We estimated that the contribution of direct genetic effects from the baby was larger (24% of overall variation in birth weight) than either the contribution of the mother’s genetic effects (4% of overall variation), or the joint contribution of mother and baby’s genetics (4% of overall variation). This is a useful first estimate, but it should be noted that the 4382 mother-child pairs are a relatively small sample in this context. So these results are preliminary estimates, which require confirmation in larger samples.

In addition to taking a global view, we were interested in the relative genetic contribution from mother and child to birth weight at the 60 specific regions of the genome identified in our study. For each region, we wanted to know whether the effect on birth weight was coming directly from the baby’s genetics, or whether it was in fact coming from the influence of the mother’s genetics on the gestational environment. Due to the overlap between the mother and baby’s genetic profiles, it was possible that our study of the baby’s genetics was in fact picking up a primary effect from the mother. We compared the size of the genetic effect on birth weight of the baby (from our study) with that of the mother (in a separate analysis of more than 68,000 women). We found that the baby’s genetic variation had a greater impact than the mother’s at the vast majority (93%) of those genetic regions.

Depending on the adult trait, we see different patterns of relative genetic contribution from mother and child.
The above analyses gave us some insight into the relative contributions made by the genetics of mother and baby to birth weight itself. But what about their relative contributions to the link between birth weight and adult diseases? It is possible that the pattern is different, depending on the adult disease in question. In Figure 1, we set out three scenarios (not mutually exclusive) by which an observed relationship between birth weight and later disease may arise. Each of these involves a different pattern of genetic contribution from mother and baby:
(i) if developmental programming is the only explanation, the mother’s genes would influence birth weight and later disease indirectly through the womb environment, with no direct effects of the baby’s genetics;
(ii) if birth weight and later disease are two “readouts” of the same genetic effects in the baby, there should be no indirect contribution of maternal genetics;
(iii) an association between birth weight and later disease could also arise if maternal genetic variations influence baby’s growth via the womb environment, and those same genetic variations, when inherited by the fetus, influence the disease risk directly.

We explored these scenarios further, selecting genetic variations involved in each of three adult characteristics/diseases: height, type 2 diabetes and blood pressure.

Our (and others’) analyses of height genetics show that only the baby’s inherited genes influence birth weight.
Taller parents tend to have longer babies, and we showed in our study that there is a strong overlap between the genetics of a baby’s size at birth and the genetics of their adult height. Using mother-child pairs from the ALSPAC study again, we selected 422 genetic variations known to influence adult height and compared the collective effects on birth weight of those genetic variations inherited by the child with those in the mother that were not inherited. We concluded, (as had others before, that only the inherited genetic variations influenced birth weight, and not the non-inherited ones. These results suggest it is unlikely that a mother’s adult height genetics have any influence on the weight of her baby other than through being inherited by the baby. (In theory, a taller mother could influence the size of her baby non-genetically by providing a larger space for growth, but we have no evidence to support this.) The link between birth weight and adult height is consistent with scenario (ii) above.

Analyses of type 2 diabetes genetics show that the baby’s inherited genes tend to reduce birth weight, while the non-inherited genes in the mother tend to increase birth weight.
Type 2 diabetes shows a U-shaped association with birth weight, in that both small and large babies are at greater risk of developing the disease in later life than those with birth weights close to the average. For some time, we and others have been trying to understand how genetics might contribute to these associations.

As mentioned in our other post, the idea that genetic differences could explain both lower birth weight and type 2 diabetes was first put forward in 1999, under the fetal insulin hypothesis. Since that time, our ability to perform larger and larger genetic association studies has strengthened evidence in support of that hypothesis, with genetic variations in particular regions of the baby’s DNA (for example, near the CKDAL1, HHEX-IDE and ADCY5 genes) explaining some of the link between lower birth weight and their later risk of type 2 diabetes, most likely through their effects on insulin: lower insulin production by a growing fetus results in reduced growth, while lower insulin production as an adult can predispose to diabetes. In our latest study, analyses in the ALSPAC mother-child pairs of 84 genetic variations that predispose to type 2 diabetes showed a collective effect on reduced birth weight of variations inherited by the child. Moreover, our analyses of the global genetic contribution to the link between lower birth weight and type 2 diabetes lent support for genetics contributing to a large proportion of this association (consistent with scenario (ii), above). However, our global analyses were preliminary and we were unable to model the U-shaped relationship or account for maternal genetics. Further, more detailed studies are needed to confirm our initial estimates.

On the other hand, evidence has been accumulating that genetic variations in mothers, at genomic regions known to predispose to diabetes (near the GCK and TCF7L2 genes) are associated with higher birth weight of their babies. This is likely to be because they put that mother at increased risk of elevated glucose levels during pregnancy, so that more sugar is transferred across the placenta, stimulating the fetus to make more insulin, and grow bigger as a result. The baby may be at higher risk of diabetes in later life due to inheriting the genetic risk from the mother (i.e. consistent with scenario (iii) above), or alternatively as a result of exposure to high glucose levels in the womb, as has been shown in other studies (consistent with scenario (i) above).

Our analyses of blood pressure genetics suggest that both the baby’s inherited genes and the non-inherited genes in the mother tend to reduce birth weight.
Women with higher blood pressure in pregnancy tend to have lower birth weight babies. Our analyses showed strong overlap between the genetics of blood pressure and birth weight, with those variations that predispose to high blood pressure being associated with lower birth weight. As with type 2 diabetes, our preliminary analyses suggested a large proportion (an estimated 85%) of the observed association between birth weight and blood pressure being attributable to genetics. When we attempted to separate the contribution of genetics of mother and baby using the ALSPAC mother-child pairs, we found evidence to support a contribution of both. Recent work in larger numbers of women also supports a strong contribution of the mother’s genetics. So, for blood pressure, current data are consistent with all three scenarios in Figure 1 being in play: (i) higher maternal blood pressure results in reduced growth and corresponding developmental changes that programme higher blood pressure in adulthood; (ii) genetic variation inherited by the baby predisposes both to reduced growth and to higher blood pressure in later life; (iii) genetic variation raising blood pressure in the mother causes reduced growth, possibly due to reduced placental function, while the same variations inherited by the child raise blood pressure in adulthood.

To conclude: it’s complex!
The results of our study suggest that genetics make a large contribution to the link between birth weight and adult diseases such as type 2 diabetes and high blood pressure. However, we hope to have illustrated in this blog post that this is just the beginning of the story, and we are certainly not saying “it’s all genetics”. The genetic contributions of mother and baby must be further unraveled, and the particular relationships between birth weight and individual adult diseases and traits should be considered separately, before we can build a clear picture. Ultimately, we aim to understand just how much of adult disease risk is under the influence of factors in the gestational environment that are potentially modifiable, because that will inform us how far it will be possible can prevent these diseases through improvements in antenatal care. It is worth clarifying that we are not saying that the value of antenatal care is in dispute. The case for good antenatal care is clear in terms of the impact on the immediate health of mother and baby. What is uncertain is the extent to which improvements in antenatal care that leave more babies appropriately nourished during pregnancy will feed forward into reduced rates of diabetes in the decades ahead, and it is crucial that we understand this.

These are the open questions we would like to tackle next:
• How much of the link between birth weight and later disease is due to genetic vs non-genetic factors? It looks, from our initial analyses, as if quite a lot is genetic, but we need larger samples and the ability to model non-linear (e.g. U-shaped) associations.
• How much of birth weight variation per se, and how much of the link between birth weight and later disease, is due to fetal vs. maternal genetic factors. We need larger studies of mothers and babies, along with information on adult disease outcomes in the offspring, to resolve this.
• Are the mother’s and father’s genomes equivalent in terms of their impact on the baby’s genetics in relation to birth weight? We know that at some genes (“imprinted genes”), there is a greater genetic contribution from one parent or the other, but to date, we do not know the extent of such imbalance.
• Are the observed relationships between aspects of the gestational environment (e.g. mother’s BMI) and birth weight causal? We have begun to use genetics to investigate this.

Excitingly, we have the prospect of more and larger resources that are becoming available to answer these questions. The next tranche of genotype data from the UK Biobank (see our other post) is coming soon, and within our EGG Consortium collaboration, there are more and more studies with genetic data available on both mother and child. We are therefore confident that we will have answers to all of the above questions in the months and years ahead.

Dr. Rachel Freathy, of the University of Exeter Medical School, and Prof. Mark McCarthy, of the University of Oxford, are part of the writing group who co-led a major paper on the genetics of birth weight, published in Nature on 28 September 2016.