Maternal vitamin D status in pregnancy is associated with adiposity in the offspring: prospective observational study

Abstract

Objective

To determine the relationship between maternal vitamin D status in pregnancy and body composition in the offspring.

Design

Prospective mother-offspring cohort study.

Setting

Southampton, UK.

Participants

977 pregnant women whose serum 25-hydroxyvitamin D concentration (25(OH)D) was measured in late pregnancy and their offspring, followed up within 3 weeks of birth, and at 4 and 6 years of age.

Main outcome measures

Offspring lean and fat mass assessed using Dual X-ray Absorptiometry.

Results

Median daily vitamin D intake (from food and supplements) in late pregnancy was 3.7μg/day (IQR 2.7 to 5.7). 22% of the women took vitamin D supplements in late pregnancy, but only 8.5% of the women complied with UK guidance to take 10μg per day. Median maternal serum 25(OH)D in late pregnancy was 62nmol/l (IQR 43-89); 35% of the women studied had values below 50 nmol/l. Lower vitamin D status was associated with lower fat mass in the offspring at birth, but with greater fat mass at 4 and 6 years. It was not associated with lean mass at any of the ages studied. The opposing associations seen between maternal 25(OH)D (SDs) and fat mass (SDs) in the offspring at birth and at 6 years were robust to adjustment for a range of confounding factors, including maternal BMI and weight gain in pregnancy (β (95% CI) 0.08 (0.02, 0.15) and −0.10 (−0.17, − 0.02 respectively). The key independent predictors of higher maternal vitamin D status were season of measurement and taking vitamin D in dietary supplements in late pregnancy.

Conclusion

These data suggest that insufficient maternal vitamin D status in pregnancy could result in programmed differences in offspring fat mass. The findings require replication but add to a growing evidence base for a role of vitamin D in the origins of adiposity.

INTRODUCTION

There is a growing recognition of the diversity of the metabolic roles played by vitamin D, beyond its effects on bone health. Low vitamin D status has been linked to impaired glucose tolerance, insulin resistance and the metabolic syndrome in adults¹. It has also been proposed that vitamin D insufficiency may be causally associated with adiposity in both adults and children^2-4; consistent with this proposition, vitamin D receptor gene polymorphisms have recently been linked to adiposity phenotypes⁵.

Vitamin D insufficiency is common in the UK⁶. A particular concern is that almost a third of young women have low vitamin D status, and are likely to start pregnancy with low maternal stores⁶. Whilst women are recommended to take an additional 10μg/day of vitamin D in pregnancy, supplementation is not a routine part of antenatal care in the UK, and low status is prevalent⁷. The long-term effects on children born to mothers who have a low vitamin D status in pregnancy have not been widely studied. Maternal vitamin D insufficiency has been linked to insulin resistance and low muscle mass⁸, and to poorer bone mineral accrual⁹ in the offspring; little is known about effects on adiposity¹⁰. The primary determinant of vitamin D status is sunshine exposure¹¹. Adult BMI has been shown to differ with season of birth¹² and it is possible that higher BMI, indicating greater adiposity, among adults born in winter and spring could be linked to lower maternal vitamin D status in the winter months. Recent findings from a follow-up study of Indian children support this possibility, as low maternal vitamin D status in pregnancy was associated with greater adiposity⁸. However, in an earlier study of UK children body composition was not related to maternal vitamin D status¹³.

In the context of current concerns about the high prevalence of vitamin D insufficiency in young women, and increasing rates of childhood obesity in the UK, the metabolic consequences for children born to mothers with low maternal status require further investigation. We therefore examined how maternal vitamin D status in late pregnancy related to the body composition of their children at birth, 4 and 6 years of age in a prospective cohort of 977 mothers who were studied in detail throughout pregnancy.

METHODS

Study sample

The Southampton Women’s Survey (SWS) is a prospective cohort study that has assessed the diet, body composition, physical activity and social circumstances of a large group of non-pregnant women aged 20 to 34 years living in the city of Southampton, UK. Comprehensive details of the study have been published previously¹⁴. Women were recruited through General Practices across the city between April 1998 and December 2002. Each woman was invited to take part by letter, followed by a telephone call when an interview date was arranged; 12,583 women agreed to take part, 75% of all women contacted. Trained research nurses visited the women at home and collected information about their health, diet and lifestyles, as well as taking anthropometric measurements. Women who subsequently became pregnant were followed up at 11, 19 and 34 weeks gestation and their offspring were studied in infancy and childhood.

Maternal data

Details of mothers’ parity, educational attainment (defined in 6 groups according to highest academic qualification) and social class were obtained at the pre-pregnancy interview, and height and weight were measured. Amongst women who became pregnant, smoking status in pregnancy was ascertained at the 11 and 34 week interviews. At 34 weeks a research nurse obtained a venous blood sample and an aliquot of maternal serum was frozen at −80°C. Serum 25-hydroxyvitamin D concentrations were analyzed by radioimmunoassay (Diasorin, Stillwater, Minnesota, USA). This assay measures both vitamin D2 and vitamin D3. The assay met the requirements of the UK National Vitamin D External Quality Assurance Scheme (DEQAS), and intraassay and interassay coefficients of variation were less than 10%. At 34 weeks the research nurses weighed the women again; pregnancy weight gain from before to late pregnancy was defined as inadequate, adequate or excessive according to the Institute of Medicine (IOM) 2009 recommendations¹⁵, as described previously¹⁶. In late pregnancy, diet during the preceding 3 months was assessed using a 100-item validated food frequency questionnaire (FFQ)¹⁷ and supplement use was ascertained. Dietary vitamin D intake was calculated by multiplying the frequency of consumption of a portion of each food by its vitamin D content according to the UK food composition tables or manufacturers’ composition data. Vitamin D intake from dietary supplements was calculated from the frequency and dose of individual supplements reported by the participant using composition data obtained from the manufacturer.

Infancy and childhood data

The infants were visited at the ages of 6, 12 months and 24 months, and the dates of the last breastfeed as recorded at these visits were used to define duration of breastfeeding and categorised into 6 groups. Crown-heel length at birth was measured using a neonatometer (CMS Ltd, UK) and childhood height was measured at age 4 and 6 years using a portable stadiometer (Leicester height measurer). Within 3 weeks of birth and at 4 and 6 years of age subsets of children attended for an assessment of body composition by Dual X-ray Absorptiometry (DXA) using a Lunar DPX-L instrument (GE Corp, Madison, WI) in infancy and, in childhood, a Hologic Discovery instrument (Hologic Inc., Bedford, MA, USA), both cross-calibrated. Fat mass and fat-free mass were derived from the whole body scan, using paediatric software. The total X-ray dose for the whole body scans was approximately 10.5 microsieverts (paediatric scan mode), equivalent to around 1-2 days background radiation. At 3 years of age dietary vitamin D intake was ascertained from an 80-item FFQ¹⁸ and supplementary vitamin D intake from reported supplement use. Physical activity was assessed as the reported time the child was actively on the move each day.

Sample size

A total of 1981 women became pregnant and delivered a live-born singleton infant before the end of 2003. Six infants died in the neonatal period and 2 had major congenital growth abnormalities, which left 1973 mother-offspring pairs. 121 mothers who delivered before 37 weeks gestation and a further 156 mothers who did not have measures of 25(OH)D were not included in the analysis. Of these, 977 children who had whole body DXA scans at birth, 4 years or 6 years were included in the analysis sample: 574 with DXA data at birth, 565 at 4 years and 447 at 6 years. 116 children had DXA scans at all three time points.

Statistical analysis

All children’s fat mass variables were positively skewed and thus were transformed using Fisher-Yates normal scores¹⁹ to a normally distributed variable with a mean of 0 and a standard deviation of 1. Whilst fat-free mass variables were normally distributed, they were similarly Fisher-Yates transformed so that all outcome variables were on the same scale of measurement. Maternal serum 25(OH)D concentrations were not normally distributed and were transformed using Fisher-Yates normal scores, as were other variables that were non-normal (maternal pre-pregnancy BMI, 3 year vitamin D intake and 3 year physical activity). Linear regression models were fitted with body composition variables as the outcomes and maternal serum 25(OH)D concentration as the predictor. Outcomes at birth were adjusted for sex, gestation, age at measurement, age squared and crown-heel length. All outcomes at 4 and 6 years were adjusted for sex, age at measurement and childhood height; by including length or height as covariates the effects of pregnancy weight gain on body composition were independent of child’s stature. Additional adjustments were made for potential confounding factors: maternal educational attainment, smoking in pregnancy, pre-pregnancy BMI, height, parity, social class, IOM weight gain category, breastfeeding duration, vitamin D intake at 3 years, and physical activity at 3 years (the latter two only for models with outcomes at 4 or 6 years). A cubic spline model was fitted to describe the association between vitamin D status and 6 year fat mass in more detail.

A form of Fourier analysis was used to model the association between vitamin D status and date of sample. A variable was derived that described the number of years the sample was taken after the first sample in the study, and multiplied by 2π to give a value θ, in radians. Maternal serum 25(OH)D concentration was regressed on cos θ and sin θ (representing one cycle per annum), cos 2θ and sin 2θ (representing two cycles per annum), and cos 3θ and sin 3θ (representing three cycles per annum). The most parsimonious model was the regression of maternal serum 25(OH)D on cos θ, sin θ, cos 2θ and sin 2θ.

Maternal height, age, parity, educational qualifications, social class, smoking in pregnancy, IOM weight gain category, pre-pregnancy BMI, supplementary vitamin D intake in late pregnancy and vitamin D intake from food in late pregnancy were considered as predictors of late pregnancy vitamin D status. The predicted seasonal component of vitamin D as described by the Fourier analysis model was also considered as a predictor of vitamin D status. Predictors with a P-value of < 0.2 in univariate linear regression models were entered into a multivariate linear regression model and a backwards stepwise procedure was employed to identify significant predictors at the 5% level. Statistical analysis was performed using Stata 11.1²⁰.

RESULTS

The characteristics of the mothers studied and their children are given in Table 1. Vitamin D status was measured in late pregnancy at a mean (SD) of 34.6 (0.7) weeks gestation. It was not associated with gestational age at measurement (P = 0.51).

Table 1. Descriptive statistics.

Characteristic	n	Value
Mother
Pre-pregnancy BMI (kg/m2)a	968	24.3 (22.2 to 27.6)
Height (m)b	973	1.64 (0.64)
Degree qualification or above (%)	975	23.8
Smoked in pregnancy (%)	977	14.6
Age at child’s birth (years)b	977	30.4 (3.8)
Primiparous (%)	977	45.9
Total vitamin D intake in late preg. (μg/day)a	969	3.9 (2.7 to 5.7)
Use vit. D supplements in late preg. (%)	969	22.2
Use > 10μg vit. D supplements in late preg. (%)	969	8.5
Late pregnancy vitamin D status (nmol/l)a	977	62 (43 to 89)
− 50 nmol/l (%)		35.1
− 75 nmol/l (%)		28.3
> 75 nmol/l (%)		36.6
Child
Male (%)	977	52.3
Birth fat mass (kg)a	574	0.53 (0.40 to 0.68)
Birth fat-free mass (kg)b	574	3.01 (0.35)
4 year fat mass (kg)a	565	4.65 (4.07 to 5.55)
4 year fat-free mass (kg)b	565	13.20 (1.59)
6 year fat mass (kg)a	447	5.21 (4.26 to 6.53)
6 year fat-free mass (kg)b	447	18.07 (2.20)

^aMedian (Interquartile range)

^bMean (Standard deviation)

Compared with the 875 participants who were eligible for the analyses but who did not have a measure of maternal vitamin D status or did not have DXA scans at any time point, the 977 participants in the current study were better educated (P < 0.0001), less likely to smoke during pregnancy (P=0.0004), older (P=0.0006), slightly taller (P=0.02), and more likely to be from a white ethnic group (P=0.0002). However, there were no differences in pre-pregnancy BMI between the two groups (P=0.86).

We examined associations between maternal 25(OH)D in late pregnancy and offspring fat mass and fat-free mass at birth, 4 years and 6 years (Table 2). The unadjusted results indicate associations between lower vitamin D status and greater fat mass at both 4 and 6 years. After adjustment for maternal educational attainment, smoking in pregnancy, pre-pregnancy BMI, height, parity, social class, IOM weight gain category, child’s breastfeeding duration, and vitamin D intake and physical activity at 3 years, the association between maternal 25(OH)D and 4 year fat mass became non-significant. After adjustment for confounders there was a positive association between maternal 25(OH)D and fat mass at birth (P=0.02); also the inverse association with 6 year fat mass remained (P=0.01). There was no association between maternal 25(OH)D and fat-free mass at birth, 4 or 6 years, either before or after adjustment for confounders.

Table 2. Late pregnancy vitamin D status (SD) as a predictor of offspring body composition.

	Unadjusted		Adjusted for confoundersb
Outcomea	β (95% CI)	P-value	β (95% CI)	P-value
Birth fat mass (SD)	0.06 (−0.01, 0.12)	0.09	0.08 (0.02, 0.15)	0.02
Birth fat-free mass (SD)	0.02 (−0.03, 0.07)	0.44	0.04 (−0.02, 0.09)	0.17
4 year fat mass (SD)	−0.09 (−0.16, −0.02)	0.02	−0.01 (−0.08, 0.07)	0.81
4 year fat-free mass (SD)	0.03 (−0.02, 0.08)	0.21	0.03 (−0.02, 0.08)	0.30
6 year fat mass (SD)	−0.16 (−0.23, −0.08)	< 0.001	−0.10 (−0.17, −0.02)	0.01
6 year fat-free mass (SD)	0.01 (−0.04, 0.06)	0.65	0.02 (−0.03, 0.07)	0.43

^aBirth outcomes adjusted for sex, gestation, age at measurement, age squared and length; childhood outcomes adjusted for sex, age and height

^bBirth confounders were maternal educational attainment, smoking in pregnancy, pre-pregnancy BMI, height, parity, social class and IOM weight gain category; childhood confounders were maternal educational attainment, smoking in pregnancy, pre-pregnancy BMI, height, parity, social class, IOM weight gain category, breastfeeding duration, vitamin D intake at 3 years, and physical activity at 3 years

The associations between maternal vitamin D status in late pregnancy and offspring’s fat mass at birth, 4 and 6 years are illustrated in Figure 1. Compared with the offspring of mothers with serum 25(OH)D concentrations of up to 50 nmol/l, fat mass at birth was 8% (95% CI 1, 16) greater amongst those born to mothers whose 25(OH)D was between 50 and 75 nmol/l, and 10% (95% CI 3, 17) greater amongst those whose mothers had a 25(OH)D above 75 nmol/l. Compared with mothers who had a vitamin D status of up to 50 nmol/l, 6 year fat mass was 8% (95% CI 2, 14) lower amongst mothers who had a vitamin D status between 50 and 75 nmol/l, and 6% (95% CI 0, 12) lower amongst mothers who had a vitamin D status above 75 nmol/l.

Figure 1. Mean (95% CI) fat mass at birth, 4 years and 6 years according to maternal vitamin D status in late pregnancy.

^aAdjusted for sex, gestation, age at measurement, age squared, length, maternal educational attainment, smoking in pregnancy, pre-pregnancy BMI, height, parity, social class and IOM weight gain categories

^bAdjusted for sex, age, height, maternal educational attainment, smoking in pregnancy, pre-pregnancy BMI, height, parity, social class, IOM weight gain categories, breastfeeding duration, vitamin D intake at 3 years, and physical activity at 3 years

It appeared there might be a threshold effect of vitamin D status on 6 year fat mass. A cubic spline model fitted to the data indicated a plateauing of the effect at around 64 nmol/l. Amongst mothers whose serum 25(OH)D concentrations were below 64 nmol/l there was a significant negative association between vitamin D status and 6 year fat mass (β = −0.24 (95% CI −0.41 to −0.07), P = 0.006), whilst amongst mothers with higher 25(OH)D there was no association (β = 0.00 (95% CI −0.19 to 0.18), P = 0.99), after adjustment for confounders. This interaction was statistically significant (P = 0.03).

We examined a number of potential influences on vitamin D status in pregnancy. The multivariate linear regression model of predictors of maternal 25(OH)D is presented in Table 3. The most important predictor of higher vitamin D status was season of measurement, as described by the pattern in Figure 2. Amongst women whose samples were taken in July-September, 84% had 25(OH)D greater than 64 nmol/l, whereas amongst women whose samples were taken in December-April, only 23% had values greater than 64 nmol/l. The second most important predictor was taking vitamin D in dietary supplements in late pregnancy. Of the 8.5% of women taking the recommended 10μg of vitamin D, 94% had serum 25(OH)D concentrations greater than 64 nmol/l, whereas amongst the rest of the cohort only 44% had values greater than 64 nmol/l. Higher late pregnancy vitamin D intake from food, and not smoking in pregnancy, were less important predictors of higher maternal vitamin D status. When maternal pre-pregnancy BMI was added to this multivariate model it was not statistically significant (P = 0.11).

Table 3. Predictors of maternal vitamin D status in late pregnancy (SD).

Predictor	Coefficient (95% CI)	t-statistic	P-value
Seasona	1.51 (1.37, 1.65)	21.6	< 0.001	5
Taking vitamin D supplements in late pregnancy (Yes/No)	0.67 (0.55, 0.79)	11.3	< 0.001
Late pregnancy vitamin D food intake (SDs)	0.08 (0.04, 0.13)	3.4	0.001
Smoking in pregnancy (No/Yes)	−0.18 (−0.32, −0.04)	−2.6	0.01

^aSeason defined as the predicted seasonal component of vitamin D as described by the Fourier analysis model and scaled so that a one unit change describes the difference between the summer maximum and the winter minimum.

Figure 2. Fourier analysis of maternal vitamin D status in late pregnancy.

DISCUSSION

Statement of principal findings

We have demonstrated that maternal vitamin D status measured in late pregnancy is associated with differences in DXA-assessed adiposity in the offspring. However the pattern of association differed at birth and in later childhood; lower vitamin D status was associated with lower fat mass assessed at birth, and with higher fat mass assessed at 6 years. These associations were robust to adjustment for a range of confounding factors, including childhood vitamin D intake and known predictors of adiposity in children. To our knowledge these associations between maternal vitamin D status and adiposity in the offspring at birth and in later childhood, have not been described before. In contrast to the relationships observed with adiposity, we found no association between maternal vitamin D status and fat-free mass in these children.

Comparison with other studies

Although the potential effects of maternal vitamin D deficiency on offspring body composition have been recognised¹⁰, there are few published studies with which we can compare our findings directly. We found that lower vitamin D status was associated with lower adiposity at birth (Figure 1); the significance of variations in fat mass at birth measured directly using DXA, as in the current study, are not yet fully understood²¹. We have previously described limited ‘tracking’ of DXA-assessed fat mass between birth and 4 years in the SWS cohort (r = 0.24), whereas the correspondence between 4 and 6 years is much stronger (r = 0.86)¹⁶. This suggests that a lower fat mass at birth is not necessarily associated with lower adiposity in later childhood.

By the age of 4 years, the pattern of association between maternal vitamin D status and fat mass in the SWS children had reversed, such that lower status in late pregnancy was predictive of greater adiposity. At this age the association was not robust to adjustment for the effects of confounding influences. However the same pattern of association between lower vitamin D status in late pregnancy and greater offspring adiposity was evident at 6 years, and this remained after taking account of a wide range of confounders. There are two studies that are relevant to our findings. Firstly, in a recent study of Indian children aged 9.5 years⁸, percentage fat determined using bioimpedance analysis was greater in offspring of mothers with a lower vitamin D status assessed at 28-32 weeks gestation although this was of borderline significance after adjustment for the effects of maternal covariates. Unlike our study, in the Indian children positive associations were found between maternal vitamin D status and a measure of fat-free mass (arm-muscle area). There are methodological differences between the studies that may be important, as we assessed body composition directly using DXA. However it may also be difficult to compare the findings as so many maternal characteristics differ, including body composition and vitamin D status. For example 67% of Indian women studied had low 25(OH)D (<50 nmol/l), compared with 35% in the present study. A study that allows a more direct comparison with our current data measured 178 children aged 9 years from a previous Southampton birth cohort, whose body composition was also assessed using DXA¹³. In contrast with the present findings, maternal vitamin D status in late pregnancy was not related to offspring fat mass in the earlier study. It is not clear why these findings differ. Apart from differences in the size of the cohorts there are other differences between the cohorts that may be relevant. Amongst these, the children in the previous study were approaching puberty and the greater use of vitamin D supplements (22% compared with 7%) and higher maternal vitamin D status in late pregnancy (median 25(OH)D 62 vs 50 nmol/l) in the SWS may be important. In our continued follow-up of the SWS children we will be able to reassess the relationship between maternal vitamin D status and body composition in later childhood, but the differences between the findings of the studies are currently not fully explained.

Further investigation of the association between vitamin D status and 6 year fat mass in the SWS children demonstrated a possible threshold effect at a maternal 25(OH)D around 64 nmol/l; the negative association between vitamin D and fat mass was confined to the 48% of women who had lower values; there was no association above this threshold. When we examined the predictors of maternal vitamin D status we found that season of measurement was a key influence, such that women who reached 34 weeks gestation in the summer months had much higher vitamin D concentrations than those measured in the winter (Figure 2). Such seasonal effects on vitamin D status are expected. However, vitamin D status was also increased by use of supplements. In the UK an additional 10μg/day has been recommended since 1991. Although NICE guidance in 2008²² endorsed the original recommendation, use of supplements is not routine in the UK. This is borne out in this sample of SWS women; 8.5% of women were taking the recommended amount of vitamin D in late pregnancy. It is important to note that among these women 94% had vitamin D concentrations above 64 nmol/l, irrespective of the season when they were studied.

Strengths and weaknesses

The SWS provides data from a contemporary cohort of women and their offspring from a wide range of sociodemographic backgrounds. Vitamin D status was measured in a large number of participants and outcomes were assessed at three different time points. We used DXA to provide measurements of fat mass and fat-free mass, and by including length or height measures in the models we were able to describe effects independent of infant or childhood stature. Because we have detailed data on this cohort we were able to control for a large number of other potential confounding factors, including important predictors of childhood fat mass such as maternal education, pre-pregnancy BMI, smoking in pregnancy and duration of breastfeeding. For example, whilst we have previously shown that excessive weight gain in pregnancy is associated with offspring adiposity¹⁶, we were able to demonstrate that the association between vitamin D status and fat mass at 6 years was independent of this effect (Table 2). A further consideration is that of reverse causality, as increased storage of vitamin D in women who have greater fat stores results in lower 25(OH)D^23-25. An apparent association between lower maternal vitamin D status and adiposity in their offspring could therefore be due to the effects of shared diet and lifestyle, rather than causally related to variations in status. Whilst we cannot exclude this possibility, the association between vitamin D status and 6 year fat mass was independent of maternal pre-pregnancy BMI in our study (Table 2); we therefore do not think this would explain the pattern of associations we observed.

Although the use of DXA is well-validated in adults, there are potential problems with neonates and young children due to their low absolute BMC and tendency to move. However, we used specific paediatric software and movement artifact was modest; the few individuals with excessive movement at each time point were excluded from our analyses. The accuracy of DXA for the assessment of body composition in small animals has been demonstrated in piglets^26;27.

DXA measurements and vitamin D status were available on a subset of 977 children and their mothers, who, when compared with a similar group without measurements, were better educated, less likely to smoke in pregnancy, older, taller and more likely to be from a white ethnic group. However, unless the associations between vitamin D status and offspring body composition are different in the remainder of the cohort, it is unlikely that selection bias could explain the relationships observed.

Interpretation and implications

We will be able to address some of the implications of the findings of the present study in our continued follow-up of these children. Whilst the findings require extension and replication, the mechanisms that link vitamin D insufficiency to a lower fat mass in the baby at birth but greater adiposity in later childhood also need to be elucidated. The role of vitamin D in adipocyte metabolism is complex^4;5 and the mechanisms that link vitamin D status to adiposity are currently unknown⁴. One interpretation of our data is that there are programmed effects on the fetus arising from maternal vitamin D insufficiency that remain with the individual and that may predispose them to gain excess body fat in later childhood. We speculate that there may be different routes to childhood obesity arising in prenatal life since both insufficient (associated with maternal vitamin D insufficiency or inadequate pregnancy weight gain¹⁶), as well as excess (associated with maternal adiposity and excessive pregnancy weight gain¹⁶) maternal nutrition have been shown to be associated with greater adiposity in childhood.

Our study shows that low vitamin D status in pregnancy, among a population of healthy pregnant women living in the south of England, is common. These data contribute to a new evidence base that links maternal vitamin D insufficiency to adverse outcomes in the offspring. Whilst further studies are needed, the findings add weight to current concerns about the prevalence of low vitamin D status among women of reproductive age in the UK⁶. Analyses of an ongoing vitamin D supplementation trial amongst pregnant women will provide a further opportunity to assess the associations between vitamin D insufficient and offspring body composition in an experimental study²⁸.

What is already known on this topic

• Low vitamin D status is common in women of reproductive age in the UK.

• Low vitamin D status has been linked to higher BMI and to greater adiposity in children and adults, but the consequences of low maternal status in pregnancy for offspring body composition is not known.

What this study adds

• In a large general population sample of women, lower vitamin D status in late pregnancy was associated with greater adiposity in their offspring at the age of 6 years.

• These data suggest that maternal vitamin D insufficiency may influence fetal development and thereby increase the risk of childhood adiposity in the offspring.

Abbreviations

SWS

Southampton Women’s Survey

IOM

Institute of Medicine

DXA

Dual X-ray Absorptiometry