We found that none of the 24 genotypes which have a confirmed association to adult adiposity are associated with birth weight. The MTCH2-rs10838738 obesity risk variant was associated with a decreased ponderal index, however this association did not remain significant after accounting for multiple testing and it is therefore likely to be a chance finding. Moreover, an obesity risk score was strongly associated with adult BMI in the subsample of Inter99 with birth weight data available, but it did not associate with birth weight and ponderal index. We did not observe any interactions between risk variants and birth weight in the prediction of adult BMI.
Our results therefore suggest that the risk variants affecting adult adiposity do not have a strong impact on size at birth and therefore are not likely to explain the association between birth weight and adult BMI. Supporting this notion, previous studies have not found altered birth weight in carriers of mutations causing monogenic forms of obesity 
. The results of the present study are therefore in line with the majority of previous studies which have failed to show any associations between obesity risk variants and birth weight 
. However one recent well-powered study by Elks et al.
found a modest association between an obesity risk score comprising alleles from FTO, MC4R, TMEM18, GNPDA2, KCTD15, NEGR1, BDNF
and ETV5 (SFRS10
) loci and BMI at birth in 7,146 individuals. When constructing a similar risk score in 3,496 individuals from Inter99 with genotype and birth weight data available no significant association was found (data not shown). This may be due to a lower statistical power in the present study, but as discussed by Elks and colleagues the obesity risk variants seem to have a larger effect on weight gain during infancy and childhood.
It is well known that the central nervous system and especially areas of hypothalamus play key roles in whole body energy regulation. The majority of the genes, in which mutations cause monogenic obesity, are involved in the control of appetite and food intake through the hypothalamus 
. Similarly, most of the 24 loci analysed in the present study are suggested to include genes with hypothalamic expression, although the specific molecular mechanisms behind these links are not well characterised 
We hypothesized that the examined gene variants known to affect adult adiposity could be affecting the individual from as early as pre-natal life by facilitating energy uptake and storage and thus weight regulation of the fetus. In the present study we did, however, not find any genetic support for this hypothesis, suggesting that the genetic components of fetal and adult weight gain are largely independent. During pregnancy, multiple factors including nutrient supply, placental function and available space are determinants of fetal growth, indicating a major importance of external factors on the growth of the fetus. Thus, it is likely that the central appetite regulation is mainly activated or only matures in post natal life, when the external environment becomes a less marked limiting factor for the intake of nutrients and subsequent growth. Moreover, the lack of an interactive effect between genetic risk and birth weight suggests that these variables influence adult BMI and risk of obesity independently in this cohort.
We have recently reported that specific type 2 diabetes risk alleles influence birth weight in the Inter99 population 
and a large genome-wide association study on birth weight has recently reported two variants to associate with birth weight, one of which is a known type 2 diabetes locus 
. No obesity loci were identified by this powerful approach. The mechanism by which the type 2 diabetes loci influence birth weight is presumably by influencing fetal insulin availability. Although 8 of the 24 obesity gene variants investigated in the present study influence adult insulin secretion in the Inter99 population (unpublished data), they do not associate with birth weight, suggesting that their effect on insulin secretion may be predominantly observed later in life, presumably as a result of increased insulin resistance in carriers of the obesity risk alleles.
A total of 11 of the 24 investigated genes were significantly associated with adult BMI in the subsample of Inter99 with birth weight data available. For these variants, the effects of the risk alleles are therefore most likely to arise during post-natal life. The statistical power to detect minor changes in quantitative variables such as birth weight may be low for the remaining variants. In the Danish Inter99 study, we have more than 80% statistical power to detect effects in birth weight of 30 g, 35 g and 45 g assuming minor allele frequencies of 10%, 20% and 40%, respectively, and we can thus exclude effects on birth weight within this range.
It could have been of interest if the alleles with the strongest effect on adult BMI also had the strongest effect on birth weight although not reaching statistical significance in this study sample. We have illustrated this in by plotting the z-scores of BMI, birth weight and ponderal index. However, by visual comparison of the effect on adult BMI with the effect on birth weight and ponderal index from the figure, no tendency in this direction seems to exist.
A recent study comprising 20 of the variants investigated in the present study, found that the 20 variants explained only 4.5% of the obesity status in the Inter99 cohort 
. Therefore hypothetical, not yet identified risk variants affecting pathways important for both fetal and adult weight gain may still exist.
Lack of exact information regarding gestational age between week 37 and 42 is a limitation of this study. However, none of the investigated variants have to our knowledge been associated with gestational age and thus an even genotype distribution can be anticipated. Moreover, in a recent meta-analysis three studies without information of gestational age did not introduce heterogeneity in to the results 
It is known from previous studies that the maternal genotype can interact with the fetal genotype in the determination of birth weight suggesting a complex pattern for the regulation of fetal growth 
. We do not have information on parental genotypes at present but it could be of major interest to test if maternal obesity risk genotypes may affect fetal growth and thereby explain part of the association between high birth weight and subsequent risk of obesity.
In conclusion, 24 common variants associated with adult adiposity did not have a strong effect on birth weight in the Danish Inter99 population. Also no interactions between the genetic variants and birth weight in the prediction of adult BMI was observed.