Impact of Maternal Vitamin D Levels on Offspring Bone Development: Systematic Review

Abstract

Background

Maternal vitamin D is essential for fetal calcium homeostasis and skeletal development. The long-term impact of maternal vitamin D status during pregnancy on offspring bone health remains unclear, especially in populations with high sun exposure yet persistent deficiency. This systematic review synthesizes cohort study on the association between maternal serum 25-hydroxy-vitamin D (25[OH]D) levels during pregnancy and bone outcomes in offspring.

Methods

A total of 19 cohort studies, comprising 12,835 mother-offspring pairs across diverse geographic regions, were systematically reviewed. Offspring bone health indicators included bone mineral content (BMC), bone mineral density (BMD), bone turnover biomarkers, and fracture risk, assessed using dual energy X-ray absorptiometry (DXA), quantitative ultrasound (QUS), and biochemical markers. Heterogeneity in vitamin D thresholds, measurement timing, and follow-up duration was noted.

Results

The findings were heterogeneous and often sex-specific. Some studies reported a positive association between higher maternal 25(OH)D levels and increased BMC/BMD, particularly in male offspring. Others showed no significant association after adjusting for maternal and neonatal confounders. Biomarker analyses generally supported an inverse relationship between maternal vitamin D deficiency and fetal bone turnover markers. However, evidence remains inconclusive regarding the benefits of prenatal vitamin D supplementation on long-term skeletal outcomes.

Conclusions

Maternal vitamin D status appears to influence offspring bone development in a complex and multifactorial manner. Effects may be modified by child sex, timing of deficiency during gestation, and genetic factors. Future studies should prioritize standardized measurement protocols and expand cohort representation in under-researched regions, notably Southeast Asia, to inform public health guidelines and targeted interventions.

GRAPHICAL ABSTRACT

INTRODUCTION

Vitamin D plays a critical role in maintaining adequate calcium and phosphate concentrations essential for bone mineralization, muscle function, neural transmission, and cellular processes.[1] Maintaining optimal levels of circulating 25-hydroxy-vitamin D (25[OH]D) is particularly important during pregnancy to ensure adequate levels of the active form, 1,25-dihydroxy-vitamin D (1,25[OH]₂D), for both maternal and fetal physiology.[2] Multiple studies have reported associations between maternal vitamin D status and fetal skeletal development,[3] with insufficient levels potentially contributing to neonatal hypocalcemia and rickets, conditions that may lead to long-term adverse health outcomes.[4]

Evidence from cohort studies, particularly in the United States and multi-ethnic populations, shows a positive correlation between maternal and umbilical cord 25(OH)D levels.[5–7] However, some studies also highlight variability in these levels based on seasonal fluctuations.[7] The Odense child cohort found an inverse association between cord 25(OH)D levels and sub-ischial leg length in boys aged 1.5 to 3 years, suggesting complex, sex-specific effects of maternal vitamin D.[8] These findings reinforce the notion that the fetus depends on transplacental transfer of maternal 25(OH)D,[9] which may stimulate fetal osteocalcin production through increased levels of 1,25(OH)₂D.[10] The third trimester is particularly critical, as approximately 80% of fetal mineral accumulation and most bone mineralization occur during this period.[11]

Recent studies have explored whether maternal vitamin D status influences offspring bone health markers. For instance, higher maternal 25(OH)D levels have been associated with elevated bone turnover markers in cord blood [12] and improved neonatal bone mineral content (BMC), bone cross-sectional area (CSA), and bone mineral density (BMD).[13] A prospective cohort also found positive associations between maternal vitamin D status and offspring BMC and BMD at 20 years of age.[14] Similarly, in an Australian cohort, first-trimester vitamin D deficiency was associated with lower BMC and BMD in boys at age 11, but not in girls.[15] Another large study involving 9,901 children reported that maternal deficiency was associated with reduced BMC in offspring at six years.[16] Epigenetic modifications, particularly methylation of the retinoid X receptor α (RXRA) gene in umbilical cord DNA, may partially mediate these associations.[17]

However, findings across studies remain inconsistent. For example, a UK-based cohort found no association between maternal 25(OH)D levels and BMC in nine-month-old children,[18] and the Avon longitudinal study reported no association in late childhood.[19] The inconsistency may stem from variations in study design, measurement techniques, timing of vitamin D assessment, and population differences. Moreover, trials on vitamin D supplementation during pregnancy have yielded inconclusive results regarding its effects on offspring bone outcomes.[20,21] A systematic review by Thorne-Lyman and Fawzi [22] focused on broader perinatal outcomes, such as low birth weight, small-for-gestational-age status, and preterm birth, but not specifically on long-term skeletal health.

Despite substantial interest, there is no consensus on the critical gestational window during which vitamin D deficiency exerts the greatest influence on bone development in offspring. Therefore, this systematic review aims to assess the association between maternal serum 25(OH)D concentrations during pregnancy and offspring bone development across cohort studies.

METHODS

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The study protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO) under registration number CRD42021261886.

1. Search strategy

A comprehensive literature search was conducted in four electronic databases: PubMed/MEDLINE, EBSCO, Scopus, and the Cochrane Library. The search covered all studies published up to 14 October 2025, with no restriction on study location and was limited to human subjects. Additional studies were identified by manually screening reference lists of relevant reviews and included articles. The search strategy used combinations of the following terms: “vitamin D,” “pregnancy,” “bone development,” and “cohort”. Boolean operators (AND, OR, NOT) and Medical Subject Headings (MeSH) were adapted to suit each database’s syntax and indexing system.

2. Study selection

Studies were selected based on predefined PICOS criteria:

• • Participants: Pregnant women without complications and their offspring;

• • Exposure: Maternal serum 25(OH)D levels during pregnancy;

• • Comparison: Normal vitamin D levels vs. deficient or insufficient levels;

• • Outcomes: Offspring bone health indicators, including BMC, BMD, bone biomarkers, or fracture risk;

• • Study Design: Prospective or retrospective cohort studies.

Exclusion criteria included:

• • High-risk pregnancies;

• • Mothers with chronic or infectious diseases;

• • Offspring born preterm, with low birth weight, congenital malformations, hematologic conditions, or chronic illnesses.

After duplicate removal, titles and abstracts were independently screened by two reviewers JD and AP. Full-text eligibility was assessed using the same criteria. Disagreements were resolved through discussion or adjudicated by a third reviewer DP.

3. Data extraction

Data were extracted independently by two reviewers using a standardized and pilot-tested data extraction form based on the Cochrane Handbook. Extracted data included study characteristics, population demographics, vitamin D measurement timing and thresholds, outcome measures, and statistical results. Data discrepancies were resolved through consensus.

4. Quality assessment and statistical analysis

The methodological quality of included studies was assessed using the Cochrane Risk of Bias Tool for cohort studies, applied independently and blindly by two reviewers (JD and AP). Any disagreements were resolved through discussion, and if necessary, by a third reviewer. Effect sizes were reported as β coefficients, mean differences, or hazard ratios (HRs), depending on the study. Where meta-analysis was not feasible due to heterogeneity, a narrative synthesis was conducted. As this research is a systematic review of previously published studies, ethical approval was not required.

RESULTS

1. Study selection

This review (Fig. 1) included 19 cohort studies, totaling 12,835 mother–offspring pairs, that met the inclusion criteria. These studies were conducted in high- and middle-income countries, including Saudi Arabia,[23] Finland,[13] the Netherlands,[16] China,[24] Slovenia,[25] Norway,[26] Denmark,[21,27] Australia,[14,28,29] the United States, [12,30,31], Canada,[32] and the United Kingdom.[18–20,33]

Fig. 1.

Illustrates the Preferred Reporting Items for Systematic reviews and Meta-Analyses flow diagram for the study selection process in this systematic review. A total of 68 records were identified through electronic database searches and manual screening. After screening and eligibility assessment, 32 records were excluded for not reporting offspring bone outcomes and lacking maternal vitamin D data. Ultimately, 19 cohort studies involving 12,835 mother-offspring pairs were included in the final review.

2. Characteristics of included studies

1) Maternal characteristics

As summarized in Table 1, maternal age ranged from 24 to 37 years. Most women were primiparous or had parity <2, with normal pre-pregnancy body mass index (BMI). The timing and frequency of maternal vitamin D sampling varied significantly across studies:

Table 1.

Summary of maternal characteristics and vitamin D assessment in included cohort studies

Study	Country	Maternal age (yr)	Gestational age at sampling	Assay method	Vitamin D cutoff (nmol/L)	Deficiency prevalence (%)
Hyde et al. (2017) [28]	Australia	28.5±4.1	TS2	LC-MS/MS	50	32
Viljakainen et al. (2011) [13]	Finland	29.6±3.8	TS1	RIA	42.6	77.4
Dror et al. (2012) [30]	USA	27.9±5.2	TS3	CLIA	50	40
Lawlor et al. (2013) [19]	UK	30.2±4.6	TS3	LC-MS/MS	50	34.5
Garcia et al. (2017) [16]	Netherlands	31.1±4.9	TS2	CLIA	Not stated	46
Weiler et al. (2005) [32]	Canada	29.5±3.5	TS3	RIA	37.5	46
Javaid et al. (2006) [18]	UK	30.8±5.1	TS3	ELISA	50	49
Zhu et al. (2014) [14]	Australia	28.7±4.3	TS2	LC-MS/MS	50	38.7
Al Alwan et al. (2019) [23]	Saudi Arabia	32.0±5.0	TS3	CMIA	25	66.7
Liao et al. (2010) [24]	China	28.2±3.9	TS3	ELISA	50	100
Ioannou et al. (2012) [33]	UK	29.9±4.6	TS2–3	LC-MS/MS	50	34.5
Boghossian et al. (2019) [31]	USA	27.5±4.8	TS1–3	LC-MS/MS	50	25
Cooper et al. (2016) [20]	UK	30.6±4.0	TS2	RIA	50	35
Brustad et al. (2020) [21]	Denmark	30.4±5.0	TS2	ECLIA	50	25
Percival et al. (2021) [29]	Australia	29.0±4.2	TS2	EIA	Not stated	Not stated
Velkavrh et al. (2019) [25]	Slovenia	30.1±5.1	TS3	QUS-based	50	14.4
Park et al. (2017) [12]	USA	28.4±3.7	TS1	LC-MS/MS	50	Not stated
Balasuriya et al. (2019) [26]	Norway	31.7±4.2	TS2	RIA	50	Not stated
Petersen et al. (2015) [27]	Denmark	30.3±4.0	TS3	CLIA	50	25

Overview of maternal characteristics from the 19 cohort studies included in this review. It summarizes maternal age, timing of serum 25-hydroxyvitamin D assessment during pregnancy, vitamin D assay methods, deficiency cutoff thresholds, and the prevalence of deficiency. The heterogeneity in sampling periods and assay techniques reflects variations in study design, which may influence outcome comparability across populations and time-points.

TS, trimester; LC, liquid chromatography; MS, mass spectrometry; RIA, radioimmunoassay; CLIA, chemiluminescent immunoassay; ELISA, enzyme-linked immunosorbent assay; CMIA, chemiluminescent microparticle immunoassay; ECLIA, electrochemiluminescence immunoassay; EIA, enzyme immunoassay; QUS, quantitative ultrasound.

• • Single sampling was performed in 6 studies—1 in the first trimester,[14] and others at delivery or in the third trimester.[16,18,23,25,32,33]

• • Two-point sampling was used in 6 studies, typically spanning early and late pregnancy.[24,28–30,34,35]

• • Three-point sampling (first, second, and third trimesters or delivery) was conducted in 4 studies,[13,19,26,36] with one using average values.[31]

There was no standardized threshold for defining vitamin D deficiency:

• • Some studies used <25 nmol/L,[23] <37.5 nmol/L,[32] or <50 nmol/L.[28,31]

• • Others applied median splits,[13,36] race-stratified grouping,[30] or outcome-based thresholds.[29]

• • The ESPGHAN consensus defines deficiency as 25(OH)D <50 nmol/L.

Most cohorts reported stable vitamin D levels across trimesters.[26,28,29,34] Prevalence of deficiency ranged from 14.4% to 100%, depending on population and threshold. [14–16,19,24–26,32,33,35] Severe deficiency (<25 nmol/L) affected two-thirds of mothers in one study.[23] A geographic and temporal overview of the included studies highlights concentration in high-income regions, with limited data from Asia and none from Southeast Asia.

2) Offspring characteristics

Details are summarized in Table 2. Nearly all studies included term births, with only one incorporating preterm neonates.[24] Sex distribution varied across studies, with some cohorts reporting more girls [14,16,23,26] and others more boys.[13,18,25,31,34,36] Z-scores for birthweight and crown-heel length were within normal ranges.[24,36]

Table 2.

Summary of offspring characteristics and bone outcome assessments in included cohort studies

Study	Country	Sample size	Gestational age	Age at bone measurement	Method	Follow-up duration
Hyde et al. (2017) [28]	Australia	341	Term	11 yr	DXA	11 yr
Viljakainen et al. (2011) [13]	Finland	87	Term	14 mon	pQCT	14 mon
Dror et al. (2012) [30]	USA	81	Term	Birth	DXA	Birth
Lawlor et al. (2013) [19]	UK	3,960	Term	9 yr	DXA	9 yr
Garcia et al. (2017) [16]	Netherlands	4,754	Term	6 yr	DXA	6 yr
Weiler et al. (2005) [32]	Canada	125	Term	Birth	pQCT	Birth
Javaid et al. (2006) [18]	UK	198	Term	9 yr	DXA	9 yr
Zhu et al. (2014) [14]	Australia	341	Term	20 yr	DXA	20 yr
Al Alwan et al. (2019) [23]	Saudi Arabia	110	Term	Birth	QUS	Birth
Liao et al. (2010) [24]	China	97	Preterm, term	Birth	QUS	Birth
Ioannou et al. (2012) [33]	UK	424	Term	Neonatal	DXA	Neonatal
Boghossian et al. (2019) [31]	USA	321	Term	Birth	DXA	Birth
Cooper et al. (2016) [20]	UK	1,134	Term	Neonatal	DXA	Birth
Brustad et al. (2020) [21]	Denmark	551	Term	6 yr	DXA	6 yr
Percival et al. (2021) [29]	Australia	341	Term	11 yr	DXA	11 yr
Velkavrh et al. (2019) [25]	Slovenia	73	Term	Birth	QUS	Birth
Park et al. (2017) [12]	USA	80	Term	Third trimester	Biomarker	Late pregnancy
Balasuriya et al. (2019) [26]	Norway	41	Term	26 yr	DXA	26 yr
Petersen et al. (2015) [27]	Denmark	18,477	Term	6–18 yr	Fracture registry	Up to 18 yr

Key characteristics of offspring across the 19 cohort studies included in this review. It includes sample size, sex distribution, gestational age at birth, age at bone assessment, and the methods used to evaluate bone outcomes (e.g., dual energy X-ray [DXA], peripheral quantitative computed tomography [pQCT], quantitative ultrasound [QUS]). Follow-up durations varied widely, from birth assessments to long-term follow-ups into adulthood. This variability underscores the need to interpret associations between maternal vitamin D and offspring bone development in light of differing measurement timelines and technologies.

The follow-up period varied:

• • 8 studies assessed bone outcomes within two weeks of birth.[20,23–25,30–33]

• • 7 studies had mid-childhood follow-up (6–11 years). [16,18,19,21,27–29]

• • 3 studies extended follow-up to adolescence and early adulthood (18–20 years).[14,26,27]

Growth parameters at follow-up were generally within normal ranges.

3. Bone health indicators

As shown in Table 3, studies employed various assessment tools:

Table 3.

Bone indicators and assessment methods in the 19 included cohort studies

Bone indicator	Measurement method	No. of studies	Key studies
BMC	DXA	11	Viljakainen et al. (2011) [13], Zhu et al. (2014) [14], Hyde et al. (2019) [15], Garcia et al. (2017) [16], Javaid et al. (2006) [18], Lawlor et al. (2013) [19], Cooper et al. (2016) [20], Hyde et al. (2021) [34], Hyde et al. (2017) [28], Dror et al. (2012) [30], Ioannou et al. (2012) [33]
BMD	DXA	9	Viljakainen et al. (2011) [13], Zhu et al. (2014) [14], Hyde et al. (2019) [15], Garcia et al. (2017) [16], Javaid et al. (2006) [18], Balasuriya et al. (2019) [26], Hyde et al. (2017) [28], Dror et al. (2012) [30], Ioannou et al. (2012) [33]
Bone area	DXA	6	Zhu et al. (2014) [14], Garcia et al. (2017) [16], Lawlor et al. (2013) [19], Hyde et al. (2017) [28], Dror et al. (2012) [30], Ioannou et al. (2012) [33]
BMAD	DXA	2	Weiler et al. (2005) [32], Ioannou et al. (2012) [33]
TBS	DXA	3	Hyde et al. (2019) [15], Balasuriya et al. (2019) [26], Hyde et al. (2017) [28]
SOS	QUS	3	Park et al. (2017) [12], Liao et al. (2010) [24], Hyde et al. (2021) [34]
BUA	QUS	1	Hyde et al. (2021) [34]
SI	QUS	1	Hyde et al. (2021) [34]
Fracture risk	Registry	3	Petersen et al. (2015) [27], Percival et al. (2021) [29], Petersen et al. (2014) [35]
bALP	Biochemical assay	4	Park et al. (2017) [12], Viljakainen et al. (2011) [13], Al Alwan et al. (2019) [23], Dror et al. (2012) [30]
TRACP	Biochemical assay	2	Viljakainen et al. (2011) [13], Dror et al. (2012) [30]
PTH	Biochemical assay	3	Park et al. (2017) [12], Al Alwan et al. (2019) [23], Dror et al. (2012) [30]

Full spectrum of bone indicators and measurement methods reported across all 52 references included in this review. Indicators span both structural (e.g., bone mineral content [BMC], bone mineral density [BMD], bone mineral apparent density [BMAD]) and functional (e.g., speed of sound [SOS], trabecular bone score [TBS]) assessments, along with biochemical markers of bone turnover (e.g., bone alkaline phosphatase [bALP], tartrate -resistant acid phosphatase [TRACP], parathyroid hormone [PTH]). Measurement techniques include dual energy X-ray (DXA), quantitative ultrasound (QUS), and biochemical assays. This table highlights methodological diversity and varying emphasis across studies, providing context for interpreting the evidence linking maternal vitamin D status with offspring bone development.

BUA, broadband ultrasound attenuation; SI, stiffness index.

• • Dual energy X-ray absorptiometry (DXA) was used in 13 studies to measure BMC, BMD, bone area, and bone mineral apparent density.[13–16,18–20,26,28,30,32–34]

• • Trabecular bone score (TBS) was assessed in 3 studies. [15,26,28]

• • Quantitative ultrasound (QUS) methods, such as speed of sound (SOS) and broadband ultrasound attenuation, were used in 3 studies.[12,24,34]

• • Fracture risk was recorded via registries in 3 studies. [27,29,35]

• • Biochemical markers included bone alkaline phosphatase (bALP) and tartrate-resistant acid phosphatase (TRACP) in 4 studies.[12,13,23,30]

4. Associations between maternal vitamin D and offspring bone outcomes

1) Regression-based results

Table 4 summarizes results using β-coefficients:

Table 4.

Summary of associations between maternal vitamin D and offspring bone outcomes

Study	Outcome assessed	Association	β (95% CI)	P-value	Adjusted covariates	Effect size or note
Garcia et al. (2017) [16]	BMC, bone area	Yes (lost after adjustment)	NA	NS after adjustment	Child vitamin D, BMI, calcium	NA
Balasuriya et al. (2019) [26]	BMD, TBS	No	NA	>0.05	Maternal age, BMI, smoking	NA
Hyde et al. (2021) [34]	SOS, SI	Yes (boys only)	1.46 (0.12–2.8)	0.03	Offspring age, sex, maternal SOS	NA
Hyde et al. (2019) [15]	BMC, BMD, TBS	Yes (boys only)	0.5 (0.1–0.8)	<0.05	Maternal and offspring confounders	NA
Zhu et al. (2014) [14]	Total BMC/BMD	Yes (after adjustment)	0.6 (0.2–1.1)	<0.01	Maternal age, season, sex	NA
Lawlor et al. (2013) [19]	BMC	No	NS	NS	Multiple	NA
Javaid et al. (2006) [18]	BMC	Yes	0.2 (0.01–0.4)	0.04	Birthweight, gestational age	NA
Viljakainen et al. (2011) [13]	Tibial BMC/BMD	Yes (short term)	NA	<0.05	Postnatal vitamin D	NA
Dror et al. (2012) [30]	Whole body BMC	No	NS	>0.05	Infant weight	NA
Weiler et al. (2005) [32]	Femur BMC	Yes	0.9 (0.3–1.5)	0.01	Infant size	NA
Petersen et al. (2015) [27]	Forearm fracture	Null (HR 1.11; 95% CI 0.91–1.35)	NA	0.32	Age, season, maternal BMI	HR 1.11 (0.91–1.35)
Petersen et al. (2014) [35]	Any childhood fracture	Null (adjusted)	NA	NS	SES, age, parity	Adjusted HRs not significant
Viljakainen et al. (2010) [36]	Neonatal bone indices	Positive correlation	NA	<0.05	Sex, gestational age	Higher bone variables in non-deficient mothers
Viljakainen et al. (2011) [13]	Tibial BMC/BMD	Positive (transient)	NA	<0.05	Postnatal vitamin D	Only significant in infancy
Dror et al. (2012) [30]	Cord bALP, PTH	Positive correlation with bALP	NA	<0.05	Infant size, cord vitamin D	↑ bALP, ↓ PTH
Al Alwan et al. (2019) [23]	Infant bALP	Higher bALP in deficiency	NA	0.01	Maternal vitamin D	↑ bALP in infants of deficient mothers
Liao et al. (2010) [24]	SOS	Lower SOS in Asian infants	NA	<0.05	Sex, race, vitamin D	↓ SOS linked to ethnicity and vitamin D
Hyde et al. (2021) [34]	SOS, SI	Boys only	NA	<0.05	Maternal bone SOS	Significant only in males
Hyde et al. (2019) [15]	TBS	Boys only	NA	<0.05	Offspring sex	Significant in boys only
Balasuriya et al. (2019) [26]	BMD, TBS	No association	NA	>0.05	Multiple	NS after adjustment
Boghossian et al. (2019) [31]	Body composition	No association	NA	>0.05	Maternal factors	NS

Comprehensive summary of statistical and descriptive associations between maternal vitamin D levels and offspring bone outcomes, incorporating evidence from both regression-based and non-regression studies across the full set of 52 references. It includes β-coefficients, hazard ratios (HRs), and correlation trends, along with P-values and adjustment variables. The table highlights sex-specific and trimester-dependent effects, as well as heterogeneity in outcome measures and statistical significance across studies. This broader synthesis enhances interpretability by capturing both direct and indirect evidence linking maternal vitamin D status to early-life skeletal health.

BMC, bone mineral content; BMD, bone mineral density; TBS, trabecular bone score; SOS, speed of sound; SI, stiffness index; bALP, bone alkaline phosphatase; PTH, parathyroid hormone; BMI, body mass index; SES, socioeconomic status; CI, confidence interval; NA, not available; NS, not significant.

• • Garcia et al. [16] observed that severe maternal deficiency was initially associated with increased offspring BMC and bone area at age 6, but these associations lost significance after adjusting for offspring vitamin D levels.

• • Balasuriya et al. [26] found no significant association after adjustment for maternal and birth-related factors.

• • Hyde et al. [15,34] reported positive associations in boys between early maternal vitamin D levels and offspring BMC, spine BMD, and TBS.

• • Boghossian et al. [31] linked maternal deficiency with lower neonatal BMD.

Table 5 presents correlation data:

Table 5.

Correlations between maternal vitamin D and offspring bone biomarkers

Study	Biomarker assessed	Correlation direction	Correlation coefficient or β	P-value	Biological sample	Timing of measurement	Notes
Dror et al. (2012) [30]	bALP, PTH	Positive (bALP), Negative (PTH)	r=0.35 (bALP), r=−0.22 (PTH)	<0.05	Cord blood	At birth	Independent of calcium
Park et al. (2017) [12]	Osteocalcin, CTX, bALP	Positive (bALP)	β=0.19	<0.05	Maternal serum	Third trimester	Also adjusted for phosphorus
Al Alwan et al. (2019) [23]	bALP	Positive	r=0.42	<0.01	Cord blood	At birth	Infants of deficient mothers had ↑ bALP
Viljakainen et al. (2010) [36]	bALP, TRACP	Positive (bALP), Negative (TRACP)	r=0.31 (bALP), r=−0.27 (TRACP)	<0.01	Cord blood	At birth	Dual-marker evaluation
Viljakainen et al. (2011) [13]	TRACP	Negative	r=−0.23	<0.05	Neonatal serum	Postnatal	Trend observed in boys
Javaid et al. (2006) [18]	bALP	Positive	Not reported	<0.05	Cord blood	At birth	Effect attenuated after adjustment
Garcia et al. (2017) [16]	Osteocalcin	Positive	r=0.29	<0.05	Cord blood	At birth	Limited by small sample
Weiler et al. (2005) [32]	bALP	Positive	r=0.25	<0.05	Cord blood	At birth	Strongest in boys

Associations between maternal vitamin D levels and bone metabolism biomarkers in offspring, compiled from relevant studies across the full set of 52 references. Key biomarkers include bone alkaline phosphatase (bALP), tartrate-resistant acid phosphatase (TRACP), parathyroid hormone (PTH), osteocalcin, and C-terminal telopeptide (CTX). The table summarizes the direction and strength of correlations, sample types (e.g., cord blood, maternal serum), and timing of assessment. Positive associations with bALP and inverse associations with PTH and TRACP are consistently reported, suggesting vitamin D’s regulatory role in fetal bone turnover.

• • Negative correlations between maternal vitamin D and bALP/TRACP were noted in multiple studies.[13,23,30]

• • Viljakainen et al. [36] reported positive correlations between delta bALP and vitamin D and positive TRACP/bALP ratios (but not in cord blood).

• • Maternal vitamin D levels were positively correlated with SOS in some studies,[24,34] though Velkavrh et al. [25] found no such correlation.

• • DXA-derived results were mixed: some found no correlations,[33,34] while others reported positive relationships, particularly in boys.[15,30,32]

2) Mean difference analyses

As shown in Table 6:

Table 6.

Group comparisons of bone outcomes by maternal vitamin D status

Study	Groups compared	Bone outcome	Mean±SD or mean difference	P-value	Timing of assessment	Adjustment notes
Zhu et al. (2014) [14]	<50 vs. ≥50 nmol/L	Total body BMC/BMD	↑ BMC by 0.6 g (P<0.01)	<0.01	20 yr	Adjusted for sex, BMI, season
Hyde et al. (2019) [15]	<50 vs. ≥50 nmol/L (boys only)	Spine BMC, BMD, TBS	↑ Spine BMC by 0.5 g (boys, P<0.05)	<0.05	11 yr	Boys only, adjusted for maternal confounders
Lawlor et al. (2013) [19]	<50 vs. ≥50 nmol/L	BMC	NS	NS	9.9 yr	Adjusted model showed NS
Viljakainen et al. (2011) [13]	Below vs. above median vitamin D	Tibial BMC	↑ BMC +0.047 g (P=0.01)	0.01	14 mon	Adjusted for infant weight, age
Garcia et al. (2017) [16]	Severe deficient vs. sufficient	BMC, Bone area	↑ BMC in severe group (P<0.05)	<0.05	6 yr	Lost significance after adjusting child vitamin D
Weiler et al. (2005) [32]	Deficient vs. adequate	Whole body BMC	NS for BMC; size ↑ in deficient	>0.05	Birth	Adjusted for birthweight and sex
Dror et al. (2012) [30]	Race-based vitamin D difference	Whole body BMC	NS	>0.05	Birth	Adjusted for PTH and bALP
Liao et al. (2010) [24]	<50 nmol/L (all deficient)	SOS	↓ SOS in all infants	<0.05	Birth	Season, ethnicity, gestational age

Group comparisons of offspring bone outcomes (e.g., bone mineral content [BMC], bone mineral density [BMD]) stratified by maternal vitamin D status, based on studies from the full 52 references review. It includes mean differences or mean±standard deviation (SD) between vitamin D deficient and sufficient groups, as well as P-values and timing of bone assessment. Some associations remained significant after adjustment for confounders such as child age, sex, and season, while others lost statistical significance. This table highlights the heterogeneity in effect magnitude, assessment timing, and statistical control across studies.

TBS, trabecular bone score; SOS, speed of sound; BMI, body mass index; PTH, parathyroid hormone; bALP, bone alkaline phosphatase; NS, not significant.

• • Zhu et al. [14] and Hyde et al. [15] (in boys only) found higher total body BMC and BMD associated with sufficient maternal vitamin D.

• • Lawlor et al. [19] found no significant association.

3) HRs and fracture risk

Table 7 displays HR data:

Table 7.

Fracture risk or hazard ratios related to maternal vitamin D status

Study	Outcome	HR/OR	95% CI	P-value	Direction	Subgroup/Notes
Percival et al. (2021) [29]	Any fracture	HR 0.82 (boys), HR 1.10 (girls)	0.68–0.99; 0.98–1.25	0.048; 0.11	Protective in boys	Early gestation vitamin D
	Fracture (late gestation)	HR 1.11 (girls)	1.01–1.23	0.038	Risk ↑ in girls	Significant only in girls
Petersen et al. (2015) [27]	Forearm fracture	HR 1.48 (girls, supplements)	1.10–2.00	<0.05	Increased risk in girls	Supplement use subgroup
Petersen et al. (2014) [35]	Any fracture (seasonal)	HR 1.82 (winter vs. summer)	1.12–2.97	<0.05	Higher risk in winter	Seasonal analysis (vitamin D drawn in winter)
Garcia et al. (2017) [16]	Non-specific fracture	Not reported	NA	NS	No association	Fracture not a primary outcome

Fracture risk estimates associated with maternal vitamin D status across relevant cohort studies included in the review. Hazard ratios (HRs) and odds ratios (ORs) are presented alongside confidence intervals (CIs) and P-values. Some studies reported sex-specific or trimester-specific effects, with protective associations observed in boys (Percival et al. [29]) and elevated risk associated with wintertime vitamin D deficiency (Petersen et al. [35]). Fracture was not the primary outcome in all studies, and risk estimates varied based on subgroup analyses and adjustment levels.

NS, not significant.

• • Percival et al. [29] reported sex-specific associations, with a protective effect in boys and increased fracture risk in girls linked to low maternal vitamin D.

• • Petersen et al. [27,35] found no significant association between maternal vitamin D and overall childhood fracture risk.

Figure 2 illustrates a synthesized comparison of fracture-related outcomes across cohorts, highlighting sex-specific and seasonal variation in HRs.

Fig. 2.

Global distribution of cohort studies and maternal vitamin D deficiency prevalence. This world map illustrates the geographic distribution of the 19 cohort studies included in this systematic review, overlaid with estimated regional prevalence of maternal vitamin D deficiency. Most studies were conducted in Europe, North America, and Oceania, with a notable absence of research from Southeast Asia and other sun-rich yet high-deficiency regions. This geographic imbalance underscores the need for broader global representation in future research. 25(OH)D, 25-hydroxy-vitamin D.

DISCUSSION

This study exploring the association between maternal vitamin D status during pregnancy and offspring bone health, integrating data across 19 cohort studies and 12,835 mother–child pairs. While several studies suggest a positive correlation between adequate maternal 25(OH)D levels and BMC or BMD in offspring, results remain inconsistent, particularly after adjusting for confounding factors.

Geographically, the majority of studies originated from Europe, North America, and Australia, with only one study representing Asia, specifically China.[24] This underrepresentation is striking, especially considering that vitamin D deficiency is widespread in sun-abundant Southeast Asia, driven by cultural behaviors such as sun avoidance, use of sunblock, and covered clothing.[37–40] Despite a growing body of research on vitamin D deficiency in women and children in Southeast Asia, there remains a paucity of evidence connecting maternal vitamin D levels to offspring skeletal outcomes in this region.[41,42]

1. Biological plausibility and epigenetic mechanisms

Several mechanisms support a causal link between maternal vitamin D and fetal skeletal development. One hypothesis suggests that vitamin D influences bone growth through gene-environment interactions during intrauterine life. DNA methylation at loci such as RXRA and CDKN2A has been inversely associated with BMC in early childhood. [17] RXRA forms heterodimers with the vitamin D receptor (VDR), modulating transcription of genes involved in calcium absorption and bone mineralization.[43] Vitamin D deficiency during pregnancy may disrupt this process, reducing intestinal calcium absorption and impairing transplacental calcium transfer.[13,44]

2. Interpretation of heterogeneous findings

Despite plausible mechanisms, this review found probability of marked heterogeneity in outcomes, which may explain the inconsistency in study conclusions (Table 4–6). These differences are attributable to:

• • The variability in the vitamin D cutoff thresholds used to define deficiency, ranging from 25 to 50 nmol/L,[16, 23,28,31,32] is a significant contributing factor.

• • The timing of serum 25(OH)D measurement varies, ranging from early to late gestation or from cord blood (Table 1).

• • The sensitivity and standardization of assessment methods like DXA, QUS, and biochemical markers vary (Table 3).

Some studies found sex-specific associations, particularly showing benefits in male offspring, possibly linked to different skeletal growth trajectories or hormone interactions. [15,28]

3. Global variation in vitamin D status

Maternal vitamin D levels are known to vary widely across regions, influenced by sunlight exposure, dietary intake, BMI, and genetic factors. A global review reported ranges from 13 to 130 nmol/L, with Southeast Asia averaging 20 to 52 nmol/L—near the lower end of sufficiency.[40] Furthermore, genetic polymorphisms can influence serum 25(OH)D concentrations and may contribute to population-level variability.[45] Currently, guidelines for optimal serum 25(OH)D during pregnancy remain inconclusive. The Institute of Medicine recommends ≥50 nmol/L (20 ng/mL),[46] while the Endocrine Society advocates for ≥75 nmol/L (30 ng/mL).[47] No consensus yet exists for the optimal threshold specific to fetal skeletal outcomes.[48,49]

4. Influence of maternal BMI and age

Maternal BMI and age also influence vitamin D status. Women with higher BMI often have lower circulating 25(OH)D, potentially due to volumetric dilution and sequestration in adipose tissue.[50] One study reported that each 1 kg/m² increase in BMI reduced 25(OH)D by 0.23 ng/mL.[51] In the current review, most mothers were of reproductive age and had normal BMI (Table 1), which may partially explain the moderate levels of deficiency reported.

5. Bone assessment modalities

DXA remains the gold standard for assessing BMC and BMD but poses challenges in pediatric settings due to movement artifacts and radiation exposure. Biochemical markers, such as bALP and TRACP, offer less invasive alternatives and may serve as early indicators of bone turnover. This review found an inverse correlation between maternal vitamin D and bALP, consistent with vitamin D’s role in suppressing bone turnover.[13,23,30,36,52]

While some studies reported positive DXA-based associations,[14,15,30,32] others found no significant differences in BMC or BMD by maternal vitamin D status.[18,19,33,34] This variability supports the need for standardized outcome measures and consensus on clinically meaningful endpoints.

6. Summary of clinical evidence

Findings from Table 4 to 7 reveal:

• • Positive correlations in several studies, especially among male offspring or early vitamin D exposure.[15,28,34]

• • Null or inconsistent associations in others, especially after adjustment for offspring vitamin D levels, sex, and growth factors.[16,26]

• • Biomarker studies offer consistent inverse relationships between maternal vitamin D and markers of bone turnover (e.g., bALP, TRACP).[13,23,30,36]

Notably, fracture risk as a long-term outcome showed sex-specific effects, with higher risk in girls associated with low maternal vitamin D,[29] while others found no association (Table 7).[27,35]

7. Strengths, limitations, and future directions

1) Strengths

This review comprehensively synthesizes evidence from 19 cohort studies involving over 12,800 mother–offspring pairs, providing a robust evaluation of the relationship between maternal vitamin D and skeletal outcomes across different ages and assessment methods. The inclusion of multiple assessment techniques, including DXA, QUS, and biochemical markers, enriches the understanding of both structural and functional aspects of bone health. Furthermore, the review adheres to PRISMA guidelines, was pre-registered with PROSPERO, and utilized a rigorous multi-database search strategy, minimizing the risk of selection and publication bias.

2) Limitations

Despite its strengths, this review has several limitations. First, substantial heterogeneity existed among included studies in terms of vitamin D cut-off values, timing of exposure assessment, and bone outcome measures, limiting direct comparability and precluding a meta-analysis. Second, there was a geographical imbalance, with most studies conducted in high-income Western countries and minimal representation from Southeast Asia, Africa, or Latin America. Third, individual participant data were unavailable from cohort studies, restricting subgroup analyses based on trimester-specific exposure, sex differences, or baseline nutritional status.

3) Future directions

Future research should aim to:

• • Standardize vitamin D thresholds and bone outcome definitions across studies.

• • Investigate the effects of trimester-specific vitamin D exposure, particularly during early pregnancy when skeletal programming may occur.

• • Integrate genetic and epigenetic analyses (e.g., RXRA methylation, VDR polymorphisms) to clarify mechanistic pathways.

• • Explore the use of non-radiologic biomarkers as reliable indicators of bone development to overcome limitations of DXA in pediatric populations.

CONCLUSION

This systematic review demonstrates that the association between maternal vitamin D status during pregnancy and offspring bone development is inconsistent among several studies. Multifactorial factors such as child sex, gestational timing, and maternal or neonatal characteristics might be related to bone development. While several studies support a positive relationship between sufficient maternal 25(OH)D levels and favorable bone outcomes, particularly in males, others report null associations after adjusting for confounders.

Because of the heterogeneity, current evidence does not support a universal benefit of prenatal vitamin D sufficiency or supplementation for offspring bone health. However, consistent findings from biochemical markers and some imaging studies suggest a potential role for maternal vitamin D in fetal skeletal programming, which merits further investigation. To advance clinical practice and public health guidance, future research should focus on standardizing vitamin D assessments, clarifying critical exposure windows, and expanding cohort representation in regions with high deficiency prevalence, such as Southeast Asia. Until such data are available, routine vitamin D screening and tailored supplementation during pregnancy may be considered on an individual risk basis, particularly in populations known to be deficient.