|Home | About | Journals | Submit | Contact Us | Français|
Maternal 25-hydroxy-vitamin D [25(OH)D] status in pregnancy has been associated with offspring bone development and adiposity. Vitamin D has also been implicated in postnatal muscle function but little is known about a role for antenatal 25(OH)D exposure in programming muscle development.
We investigated the associations between maternal plasma 25(OH)D status at 34 weeks gestation and offspring lean mass and muscle strength at 4 years of age.
A prospective UK population-based mother-offspring cohort: the Southampton Women’s Survey (SWS).
12583 non-pregnant women were initially recruited into SWS, of which 3159 had singleton pregnancies. 678 mother-child pairs were included in this analysis.
At 4 years of age, offspring assessments included hand grip strength (Jamar Dynamometer) and whole body DXA (Hologic Discovery) yielding lean mass and %lean mass. Physical activity was assessed by 7-day accelerometry (Actiheart) in a subset of children (n=326).
Maternal serum 25(OH)D concentration in pregnancy was positively associated with offspring height-adjusted hand grip strength (β=0.10 SD/SD, p=0.013), which persisted after adjustment for maternal confounding factors, duration of breastfeeding and child’s physical activity at 4 years (β=0.13 SD/SD, p=0.014). Maternal 25(OH)D was also positively associated with offspring %lean mass (β=0.11 SD/SD, p=0.006), but not total lean mass (β=0.06, p=0.15). This however did not persist after adjustment for confounding factors (β=0.09 SD/SD, p=0.11).
This observational study suggests that intrauterine exposure to 25(OH)D during late pregnancy might influence offspring muscle development through an effect primarily on muscle strength rather than muscle mass.
It is well established that vitamin D is important for muscle function in postnatal life. Firstly, the vitamin D receptor (VDR) has been isolated in skeletal muscle (1), and polymorphisms in the VDR are related to differences in muscle strength (2). Secondly, severe vitamin D deficiency can present with a proximal myopathy, which improves with vitamin D supplementation (3, 4). Thirdly, subclinical vitamin D insufficiency has been associated with reduced physical performance and muscle function in adolescent girls and older adults (5, 6), although trials of vitamin D supplementation have had inconsistent results with regard to improvements in muscle strength (7, 8).
Additionally, evidence is accruing that maternal serum 25-hydroxy-vitamin D [25(OH)D] concentrations during pregnancy might influence offspring body composition in childhood (9-12). Thus, in observational studies, maternal antenatal serum 25(OH)D concentrations have been associated positively with bone mass (11-13), and negatively with fat mass (9) in the offspring. Although there are scant data relating postnatal muscle development to intrauterine 25(OH)D exposure, birth weight, a marker of prenatal nutrition, has been associated with muscle mass and grip strength throughout the lifecourse from childhood to older age (14-21), consistent with a potential role for early life influences in long term muscle development. We therefore aimed, using a population-based mother-offspring cohort study (Southampton Women’s Survey), to test the hypothesis that maternal serum 25(OH)D concentrations during pregnancy are positively associated with markers of muscle size and strength in the offspring at 4 years of age.
The Southampton Women’s Survey (SWS) is a study of 12583, initially non-pregnant, women aged 20-34 years, resident in the city of Southampton, UK (22). Assessments of lifestyle, diet and anthropometry were performed at study entry (April 1998 – December 2002), and, for women who became pregnant, again at 11 and 34 weeks gestation.
The SWS was conducted according to the guidelines laid down in the Declaration of Helsinki, and the Southampton and South West Hampshire Research Ethics Committee approved all procedures (06/Q1702/104). Written informed consent was obtained from all participating women and by a parent or guardian with parental responsibility on behalf of their children.
At the pre-pregnancy interview details of maternal parity, highest educational attainment and social class were obtained and height and weight were measured. At 34 weeks gestation the women were reweighed and triceps skinfold thickness measured. Details of dietary supplements, smoking status and walking speed were ascertained by direct interview.
At 34 weeks gestation, a venous blood sample was obtained and an aliquot of maternal serum was frozen at −80°C. Serum 25(OH)D concentrations were analysed by radioimmunoassay (Diasorin, Minnesota, USA). This assay measures both 25(OH)D2 and 25(OH)D3. The assay met the requirements of the UK National Vitamin D External Quality Assurance Scheme, and intra- and interassay CVs were <10%.
There were 3159 singleton live births. The children were followed up at birth and during infancy. Duration of breastfeeding was determined from feeding histories obtained at 6 and 12 months of age. Consecutive subsets of children have been assessed postnatally; 900 children underwent DXA measurements at the Osteoporosis Centre at Southampton General Hospital at 4 years of age.
At this visit, height was measured using a Leicester height measurer (Seca Ltd., Birmingham, UK) and weight (in underpants only) measured using calibrated digital scales (Seca Ltd., Birmingham, UK). A whole body Dual-energy Xray Absorptiometry (DXA) scan was obtained using a Hologic Discovery instrument (Hologic Inc. Bedford, MA, USA) in paediatric scan mode (software Apex 3.1), yielding fat mass, lean mass and bone mineral content (BMC). As children with greater adiposity also tend to have higher absolute lean mass (23), percent fat mass and percent lean mass were subsequently derived using a three-compartment model (fat mass, lean mass and bone mineral content) to provide an indication of a more favourable body composition. Furthermore, the variable lean mass adjusted for fat mass was generated to remove any effect of lean mass increasing with fat mass. The coefficient of variation for body composition analysis for the DXA instrument was 1.4-1.9%. The reliability of DXA in small subjects has been demonstrated previously (24, 25).
Grip strength was measured using a Jamar handgrip dynamometer (Promedics, Blackburn, UK) with a standardised approach (26). The dynamometer was adjusted to fit the hand size of each individual and three measurements of each hand were taken with the maximum from all six measurements being used in the analysis. Due to the learning and tiring effect in grip strength assessment, which can lead to some variability across measurements, and to encourage children to get as high a score as possible (26), we opted to use the maximum of six measurements as our main outcome. Test-retest reliability has previously been demonstrated in this age group (27), and the coefficient of variation of the six measurements was 11%, which is similar to other studies (28). Additional sensitivity analyses were undertaken using average grip strength. Grip strength was adjusted for the child’s height.
In a subset of children (n=326), habitual physical activity was assessed using an Actiheart combined accelerometer and heart rate monitor (Cambridge Neurotechnology Ltd, Cambridge, UK) worn continuously for 7 days except during bathing and swimming. The detailed methodology has been described previously (29). Moderate, vigorous and very vigorous activity levels were grouped to give the primary exposure measure (MVPA).
Differences in demographic characteristics and body composition of the children by gender were explored using t-tests and Mann-Whitney U tests for normally and non-normally distributed variables, respectively. Owing to differences between boys and girls the body composition variables were adjusted for the sex of the child. In order to allow for subsequent comparison of effect sizes in univariate and multivariable linear regression models, the exposures and outcomes (offspring body composition, physical activity, grip strength, and maternal late pregnancy 25(OH)D concentration) were standardised using a Fisher-Yates transformation to a normally distributed variable with a mean of 0 and a standard deviation of 1. These analyses thus yielded standardised regression coefficients (SD per SD). In the first multivariable model (Model 1) we included a number of child (sex, age, height, milk intake at 4 years, duration of breastfeeding) and maternal factors (parity, late pregnancy walking speed, late pregnancy smoking status, triceps skinfold thickness at 34 weeks gestation, age at delivery and social class). Additionally we explored whether using maternal BMI (either measured pre-pregnancy or at 6 months post-delivery) as a measure of adiposity instead of late pregnancy triceps skinfold thickness in model 1 changed the associations. In further analyses, offspring time in MVPA was subsequently added (Model 2) and we additionally determined whether inclusion of either season of maternal 25(OH)D measurement, birth or 4 year assessment in the models would change the associations. Seasons were defined as winter (December to February), Spring (March-May), Summer (June-August) and Autumn (September-November). All analysis was performed using Stata v12.0 (Statacorp, College Station, Texas, USA). A p value of <0.05 was accepted as statistically significant, and given the observational nature of the study, together with the substantial collinearity amongst both predictors and outcomes, testing for multiple comparisons was felt to be inappropriate (30).
Data were available for 678 mother-offspring pairs who had maternal serum 25(OH)D status in late pregnancy and offspring body composition by DXA and grip strength measurement at 4 years. The characteristics of the mothers and children are presented in tables 1 and and2,2, respectively.
The mothers included in this study were of similar age (mean±sd) at delivery (30.7±3.8 years vs 30.6±3.9 years, p=0.69) and parity (51.3% vs 51.0% nulliparous, p=0.88) but had achieved a higher educational level (25% vs 21% had a higher degree, p<0.001) compared with mothers in the SWS cohort whose children did not participate in this study. Additionally fewer mothers included in this study smoked in late pregnancy (9.9% vs 16.9%, p=0.001).
The boys and girls were of similar age, height and weight, but the girls had lower total and percent lean mass (both p<0.0001) (Table 2). Although absolute grip strength was greater in the boys than in the girls (8.5±1.7kg vs 8.2±1.7kg, p=0.023), after adjustment for child’s height, this difference was attenuated and became statistically non-significant (8.5±1.5kg vs 8.3±1.6kg, p=0.072). Physical activity indices were similar in boys and girls (Table 2).
A significant positive correlation was identified between maternal serum 25(OH)D concentration in late pregnancy and offspring height-adjusted grip strength at 4 years (β=0.10 SD/SD, p=0.013; Figure 1), such that for every standard deviation increase in maternal serum 25(OH)D, offspring height-adjusted grip strength increased by 0.15kg (95% CI 0.03, 0.27kg). This association persisted after adjustment for confounding factors (Model 1: β=0.08 SD/SD, p=0.040). Furthermore, in the 326 children who had physical activity monitoring, the child’s mean daily time in MVPA was positively associated with height-adjusted hand grip strength (β=0.13 SD/SD, p=0.011), and inclusion of time in MVPA strengthened the association between maternal 25(OH)D and offspring grip strength (Model 2: β=0.13 SD/SD, p=0.014; Table 3). Although the associations between maternal 25(OH)D and offspring grip strength appeared somewhat more robust in the girls than the boys (Table 3), the test for a statistical interaction between maternal serum 25(OH)D concentration and grip strength by child sex did not achieve statistical significance (p=0.30).
Maternal serum 25(OH)D concentration in late pregnancy was positively associated with offspring percent lean mass (β=0.11 SD/SD, p=0.006) (figure 1) and lean mass adjusted for fat mass (β=0.08 SD/SD, p=0.035), but not total lean mass (β=0.06 SD/SD, p=0.15). The associations with percent lean mass and lean mass adjusted for fat mass however were attenuated after the addition of potential confounding maternal and child factors (Model 1), and just failed to achieve statistical significance (β=0.07 SD/SD, p=0.062 and β=0.05 SD/SD, p=0.051, respectively).
There was a high correlation between the maximum and the mean of six measurements in our cohort (r=0.94, p<0.0001): In further sensitivity analyses, we repeated the analysis using average grip strength instead of maximum grip strength, and the relationship with maternal 25(OH)D in later pregnancy was almost identical (β=0.10 SD/SD, p=0.012; Model 1: β=0.08 SD/SD, p=0.042; Model 2: β=0.12, p=0.016). The associations did not differ when maternal BMI was included in the multivariate models as a measure of adiposity instead of maternal triceps thickness in late pregnancy. Finally, the inclusion of either season of 25(OH)D measurement, season of birth or season of 4 year assessment did not alter the relationships.
In this prospective mother-offspring study, we have identified a number of key associations between maternal serum 25-hydroxy-vitamin D concentration in late pregnancy and offspring muscle development. Thus maternal 25(OH)D status was positively associated with offspring grip strength at 4 years. This finding persisted after adjustment for a number of potential confounding factors relating to maternal/ childhood lifestyle, body build and physical activity. The weaker relationships between offspring muscle mass and maternal 25(OH)D status are consistent with the notion that the association between maternal vitamin D and offspring grip strength might be mediated via an effect on muscle function partly independently of an increase in muscle mass.
The strength of this study is in the detailed phenotyping of the mother-offspring pairs and its prospective design. Although the children included in this study were born to mothers who were slightly older and tended to be better educated than mothers of children not included, they do represent a wide range of maternal age and family backgrounds and all comparisons were internal. However, there are a number of limitations to this study. Firstly, although DXA is well validated in adults there are some problems in children because of their smaller size and tendency to move. Body composition assessment by DXA of small subjects has been previously validated using biochemical assessment of carcass nitrogen content and lipid extraction to determine lean and fat mass, respectively, in piglets which were sacrificed immediately after DXA scanning (24); we used specific paediatric software and movement artefact was minimal. The few scans with excess movement artefact were excluded from the analysis. Secondly, measurement of grip strength in children is less straightforward than in adults, but the children were able to cooperate, and the validity and reproducibility of grip strength measurement in this age group has been demonstrated previously (29;30). Indeed, the coefficient of variation across the six measurements was 11%, which is similar to that reported in other studies (28). Despite the greater precision of DXA, stronger relationships with 25(OH)D were identified with grip strength than with lean mass, suggesting that any noise introduced by random variation in the grip measurements did not prevent the detection of meaningful associations. Thirdly, children may remove physical activity monitors and we did not systematically record this. However, we accounted for non-wear time in the analysis of the accelerometer output. Finally, it is not possible in this observational study to determine whether the observed associations are causal.
There are few previous data relating maternal 25(OH)D concentrations during pregnancy to offspring muscle development. Findings from the Mysore-Parthenon Study, a prospective mother-offspring birth cohort in India, demonstrated greater arm muscle area at 5 and 9 years in children born to vitamin D replete [serum 25(OH)D>50nmol/l] compared with deplete [25(OH)D<50nmol/l] mothers(10). Consistent with these results, in the Avon Longitudinal Study of Parents and Children (ALSPAC) a positive association between maternal estimated ultraviolet B exposure in the third trimester (a proxy for maternal 25(OH)D concentrations) and offspring lean mass determined by DXA at 9 years was observed (12). In contrast to our findings, in the Mysore study, no difference in grip strength was identified at 9 years of age between children born to mothers defined as vitamin D deficient and replete at 28-32 weeks gestation (10). However, there are marked differences across these three populations in terms of exposure definition (maternal 25(OH)D concentration or sunlight exposure), 25(OH)D assay technique, confounding factors considered, and the socioeconomic status, childhood body composition, 25(OH)D distribution and age of the children studied, making direct comparison difficult. Furthermore, racial differences in vitamin D metabolism have been demonstrated (31) and increases in sex-hormones at the inception of puberty, together with greater exposure to manual work, may have obscured any association between maternal pregnancy serum 25(OH)D concentration and offspring grip strength at 9 years in the Mysore study.
Taken together, these previous studies are consistent with a positive association between maternal 25(OH)D concentration during pregnancy and offspring muscle development. However, our findings suggest that this relationship might be mediated partly via muscle function, rather than purely by muscle size. Such a disparity between the influence of muscle size and strength has been observed in relation to outcomes such as disability and mortality in adult cohorts (32) and there is good evidence that vitamin D might influence muscle strength in postnatal life: The vitamin D receptor has been isolated in skeletal muscle (1), and myopathy is a prominent feature of vitamin D deficiency in both infants and adults; histological studies have demonstrated atrophy of the type II muscle fibres in vitamin D deficient subjects (33, 34). These fibres are necessary for rapid bursts of speed and power, and are therefore likely to be involved in the action required for grip strength assessment. Vitamin D supplementation may improve muscle strength, although the results of randomised controlled trials are inconsistent (7, 8). Importantly, intramuscular fat accumulation appears to be inversely associated with 25(OH)D concentration independent of BMI and muscle area (35) and muscle adiposity is negatively related to muscle strength (36, 37). Anthropometric measures such as arm muscle area cannot distinguish lean mass from intramuscular fat infiltration and hence might explain some of the observed discrepancy in associations between maternal 25(OH)D and offspring muscle strength compared with muscle mass. Given that fibre number is largely set in utero, and muscle size increases by hypertrophy postnatally (38, 39), another possibility is that maternal 25(OH)D concentrations might in some way influence fibre number, or motor unit size, more than overall mass. Such effects, in the context of maternal undernutrition, have been demonstrated in animal models (40).
Clinically, these findings could have long-term health benefits. There is evidence for the tracking of muscle function and mass (41-46); Gabel et al demonstrated significant tracking of muscle function over a 15 month period in pre-school children (41), and others have shown tracking of muscle strength through childhood and into early adulthood (43). Indeed our own studies of fetal and postnatal growth suggest that the majority of children have settled onto a sustained growth trajectory by the age of four years (47). Muscle strength peaks in young adulthood before declining, and low grip strength in adulthood has been associated with poor health outcomes including diabetes, falls, fractures and all-cause mortality (48, 49). Accordingly it would be expected that the greater muscle strength identified at 4 years of age in children born to mothers with higher vitamin D levels would track into adulthood, and this method of increasing peak muscle mass might be one approach to addressing the increasing burden of sarcopenia. Previous research has demonstrated that for every standard deviation reduction in grip strength in older women, the risk of incident falls and fractures during follow-up over the subsequent 3 to 9 years was increased by 33% and 25%, respectively (50). We observed a 0.25SD difference in height-adjusted grip strength between the children in the lowest and highest quartiles of maternal vitamin D status, thus if this difference were maintained into adulthood, it might translate into an 8% reduction in falls risk and a 6% decrease in fracture risk.
In summary, in this observational study, maternal serum 25(OH)D concentration in late pregnancy was associated positively with offspring hand grip strength at 4 years independent of child’s height, and a range of other maternal and childhood confounding factors. In contrast, a weaker positive non-significant association was observed with offspring percent lean mass, observations which would be consistent with maternal 25(OH)D status influencing muscle function more than muscle mass. These results suggest that vitamin D supplementation in pregnancy might lead to improved muscle development in the offspring. However formal testing of this hypothesis in an interventional setting (51, 52) should be undertaken prior to the development of any clinical recommendations.
NCH and RM are joint first author. We thank the mothers who gave us their time; and a team of dedicated research nurses and ancillary staff for their assistance. This work was supported by grants from the Medical Research Council, British Heart Foundation, Arthritis Research UK, National Osteoporosis Society, International Osteoporosis Foundation, Cohen Trust, NIHR Southampton Biomedical Research Centre, University of Southampton and University Hospital Southampton NHS Foundation Trust, and NIHR Musculoskeletal Biomedical Research Unit, University of Oxford. Participants were drawn from a cohort study funded by the Medical Research Council and the Dunhill Medical Trust. The research leading to these results has also received funding from the European Union’s Seventh Framework Programme (FP7/2007-2013), project EarlyNutrition under grant agreement n°289346. We thank Mrs G Strange and Mrs R Fifield for helping prepare the manuscript.
Study registration: www.birthcohorts.net (20/09/2011)
Disclosure Statement: NCH, RJM, AAS, GN, JHD, MKJ, SMR, HMI and CC have nothing to declare. KMG has acted as a consultant to Abbott Nutrition and Nestle Nutrition, and has received reimbursement for speaking at an Abbott Nutrition Conference on Pregnancy Nutrition and Later Health Outcomes, at a Nestle Nutrition Institute Workshop and at a workshop funded by the International Life Sciences Institute (ILSI Europe). He is part of an academic consortium that has received research funding from Abbott Nutrition, Nestec and Danone.