Vitamin D3 for Life

Live Enlightened

Physiology

Infant Vitamin D, Bone, Minerals and Health Outcomes: Full Evidence Review

Previous related pageInfant Vitamin D, Bone, Minerals and Health Outcomes: Study SummaryNext related pageInfant Vitamin D, Bone, Minerals and Health Outcomes: Study Summary
Study at a glance

What was studied?

Researchers tested whether vitamin D strategies changed infant bone measurements, blood minerals, rickets (a disorder of growing bone), reported infections or body composition. The trials involved different populations, including infants with low starting vitamin D status and infants born early.

What did the studies find?

Several trials in infants born at the usual stage of pregnancy raised vitamin D blood levels without demonstrating an added bone benefit. A maternal strategy given during pregnancy and continued after birth reduced cases meeting a laboratory definition of rickets in Bangladesh. This was not the same as X-ray-confirmed disease. A Finnish trial found no reduction in reported infections overall. One trial found modestly more lean tissue at 12 months, without showing stronger muscles or less body fat.

What does this mean?

Blood measurements, bone scans and diagnosed illness answer different questions. A benefit for one outcome does not establish benefit for all outcomes. Lean tissue includes soft tissue other than fat, including muscle; measuring its amount does not measure its strength.

What are the limits?

The studies used different doses, timing and populations. Some outcomes were uncertain or measured only in a small sample. Most infants in the Finnish trial already had sufficient vitamin D status. The findings do not establish a universal dose or blood target.

Read the full article below ↓

An infant’s vitamin D blood result and a health outcome answer different questions. Bone scans, blood calcium and phosphate, rickets assessments, reported infections and body composition measure different aspects of development. This review compares five independent trial groups to examine when vitamin D interventions changed those outcomes, when they did not, and why the findings depend on the infants and intervention studied. For a shorter account, read Infant Vitamin D, Bone, Minerals and Health Outcomes: Study Summary.

Start with the outcome, not only the blood concentration

Most supplementation trials measure 25-hydroxyvitamin D, written as 25(OH)D. This is the circulating metabolite used to assess vitamin D status. It is different from the vitamin D3 given in a supplement and from the active hormone, 1,25-dihydroxyvitamin D. Higher 25(OH)D confirms a difference in measured status; it does not establish stronger bone or improved immunity. The marker distinction is introduced in Vitamin D Testing: 25-OH-D vs 1,25-OH₂-D, and the conversion between vitamin D forms in Vitamin D Activation.

The distinction is especially important in infancy, when studies may measure mineral accumulation rather than fractures or diagnosed disease. Bone mineral content is the amount of mineral measured in a scan region. Bone density relates mineral to the size or volume measured by a particular imaging technique. These outcomes are related, but they are not interchangeable with each other or with a clinical event. The wider explanation in Bone Mineralisation, Turnover and Fractures Measure Different Outcomes separates these endpoints without transferring animal or adult results to infants.

Blood calcium and phosphate provide another kind of evidence. An unchanged concentration does not demonstrate that intestinal absorption, kidney losses or deposition in the skeleton were unchanged. Nor does a change in parathyroid hormone, a hormone involved in mineral regulation, independently prove a skeletal benefit. The trials below need to be read according to the endpoints they actually measured. The relevant general pathways are introduced in Vitamin D and Calcium and Vitamin D and Phosphate Regulation.

All supplement amounts describe research interventions. IU means international units. Trials in healthy term infants, infants selected for low status, and preterm infants receiving specialist care should not be treated as interchangeable dosing guides.

For current UK guidance, see NHS Vitamins for children. Infants born early need advice from their neonatal team.

Healthy term infants: more vitamin D did not mean more measured bone benefit

In Montreal, Gallo and colleagues randomised 132 healthy breastfed term infants at about one month to 400, 800, 1,200 or 1,600 IU of vitamin D3 daily. The assigned doses changed patterns of 25(OH)D attainment, but the study did not detect a corresponding advantage in growth or mineral accrual. About 74% completed the 11-month intervention, so the findings must also be interpreted with the losses to follow-up and limited power for secondary outcomes in mind (Gallo et al., 2013).

The highest-dose arm was stopped after high immunoassay measurements, and those infants were switched to 400 IU. Subsequent mass-spectrometry results were lower, while additional safety investigations were normal. The stopping decision is relevant to monitoring and measurement uncertainty. It is not proof of toxic injury. Equally, the absence of demonstrated injury in this small monitored study cannot establish universal long-term safety.

The Finnish VIDI trial was substantially larger. It compared 400 with 1,200 IU daily from two weeks to 24 months in 975 infants included in the analysis, of whom 823 completed follow-up. At birth, 95.7% were classified as vitamin D sufficient. Investigators used peripheral quantitative computed tomography, an imaging technique that examines bone density and structure, to study the tibia, or shin bone (Rosendahl et al., 2018).

The higher dose raised 25(OH)D but did not add a detected tibial bone benefit. Movement and missing scans affected the imaging data, and the largely sufficient Northern European population limits transfer to markedly deficient settings. Nevertheless, this randomised comparison is useful evidence against assuming that additional supplementation above the lower strategy must improve bone measurements in already sufficient infants.

These two trials did not test withholding vitamin D from severely deficient infants. Their null higher-versus-lower-dose findings therefore do not contradict a role for vitamin D in preventing deficiency-related disease. They address the incremental benefit of the specific dose comparisons, not every possible comparison between deficiency and adequacy.

Low birth status: a bone null with meaningful uncertainty

A later Montreal trial addressed infants whose 25(OH)D was below 50 nmol/L at birth. Ninety-eight term infants were randomised at about one month to 400 or 1,000 IU daily. Forty-one infants with sufficient birth status formed a separate nonrandomised reference group. The bone report was a prespecified secondary analysis of this trial (Gharibeh et al., 2023).

The higher dose produced higher 25(OH)D without a demonstrated mineral advantage. At 12 months, the reported whole-body bone-mineral-content contrast was 10.78 g, with a 95% confidence interval from −1.79 to 23.34 g. Because that interval includes zero, the result does not establish a benefit. Because it also includes positive differences, it is not proof that the two regimens are equivalent. The study’s modest sample and attrition matter when interpreting this uncertainty.

Selection for low status at birth does not mean all infants had the same status when supplementation began, or that severe deficiency was equally represented. The nonrandomised reference group also cannot establish the causal effect of being born sufficient. The strongest comparison remains the randomised difference between the two assigned supplement strategies.

This trial helps connect the sufficient-population findings to a different starting population, but it does not justify a universal statement that changing vitamin D status never affects mineralisation. Its finding is specific to the doses, timing, population and measurements used. A conflict between narrative and table wording for a vitamin D metabolite ratio also prevents drawing a reliable mechanistic conclusion from that ratio here.

Biochemical rickets in a deficient setting: a different result and a different intervention

The Maternal Vitamin D for Infant Growth trial, or MDIG, took place in Bangladesh and randomised 1,300 pregnant women to prenatal supplementation strategies, including one higher-dose strategy continued for six months after birth. The combined strategy provided mothers with 28,000 IU of vitamin D3 each week, starting at 17 to 24 weeks of pregnancy and continuing until six months after birth. These were research doses given to mothers, not an infant dosing schedule. All maternal groups also received 500 mg calcium daily. A later analysis screened 790 infants at six to 12 months for biochemical signs of rickets (Lautatzis et al., 2024).

There were 13 biochemical cases among 165 screened infants in the placebo group, compared with two among 154 in the higher-dose prenatal-plus-postpartum group. The main results table gives a risk ratio of 0.16 with a 95% confidence interval of 0.04 to 0.72. This is evidence of fewer cases meeting the study’s biochemical definition under that combined strategy. Prenatal-only comparisons were not statistically significant.

The definition and diagnostic process are crucial. A biochemical case was identified using an investigator-consensus definition. Alkaline phosphatase, an enzyme measured in blood, had to be at least 450 U/L, or at least 350 U/L together with one or more of the following: calcium at or below 2.2 mmol/L, phosphate at or below 1.6 mmol/L, or intact parathyroid hormone at or above 6.9 pmol/L. These study-specific laboratory criteria were not an X-ray diagnosis. Across the analysis, 39 infants met that definition. Only 16 underwent radiography, and four of those had radiographic rickets. It would therefore be incorrect to describe all 39 biochemical cases as X-ray-confirmed rickets or to treat incomplete radiography as confirmation that the other cases had no skeletal disease.

This was a secondary analysis in a setting with maternal deficiency, not a direct infant 400-versus-higher-dose experiment in largely sufficient term infants. Screening covered a subset of the original trial, at varying ages, and all mothers received calcium. The result supports the tested combined maternal strategy’s effect on the measured biochemical endpoint. It does not isolate a universal infant dose, prove the same reduction in clinical fractures, or establish that vitamin D alone explained every component of the intervention context. The different question of treating established disease is covered in Calcium Availability and Vitamin D Response in Nutritional Rickets: Full Study. Its calcium-treatment comparisons should not be mistaken for this maternal prevention strategy.

The parent trial did not find improved infant length at one year (Roth et al., 2018, corrected 2021). Its growth report and the later rickets analysis describe the same underlying trial. A biochemical benefit and a growth null can coexist because they are different endpoints. Counting them as independent confirmations, or using either to erase the other, would misrepresent the evidence.

Preterm infants: mineral markers are not a bone-outcome trial

A Chiang Mai trial randomised 38 preterm or very-low-birthweight infants, 19 per group, to supplemental vitamin D3 at 400 or 800 IU daily. The final assessment occurred at six weeks of chronological age, after approximately four to five weeks of exposure. The higher-dose group had higher mean 25(OH)D, 47.3 versus 32.0 ng/mL, but no detected advantage in calcium, phosphate or alkaline phosphatase, an enzyme measured as part of the biochemical assessment (Rueang-Amnat et al., 2025).

Both groups received additional nutrition and mineral care, so the assigned supplement was not the whole intake. The study doses replaced the routine hospital vitamin D supplement. They were not added to a further routine 600 IU. This detail matters when interpreting exposure and prevents accidentally turning the study into a larger-dose comparison than it actually was.

The study did not measure improvement in bone imaging, fractures or clinical rickets. Unchanged mineral markers therefore cannot be presented as proof that skeletal mineral deposition was unchanged, just as higher 25(OH)D cannot be presented as proof that it improved. The short follow-up, single centre, unblinded clinical staff, feeding differences and marked sex imbalance between groups restrict broader conclusions. Short-term monitoring cannot establish safety for all infants or longer durations.

Infection: a clinical question with a null overall comparison

VIDI also collected parent-reported infection information. The overall infection incidence-rate ratio for the higher: versus lower-dose groups was 1.00, with a 95% confidence interval of 0.93 to 1.06. The point estimate indicates no difference in the observed overall rate; the interval describes the uncertainty around that estimate. It should not be converted into a claim that the study proved precisely identical infection risk (Rosendahl et al., 2018).

The outcome came from parent diaries rather than laboratory confirmation of every infection. It is therefore evidence about reported illness under the tested regimen, not a direct measurement of every immune pathway or every pathogen-specific response. Post hoc seasonal findings do not replace the main overall comparison. Nor does the null show that vitamin D signalling has no immune function. It shows that the higher dose did not reduce the measured overall infection rate in this predominantly sufficient infant population. Mechanistic evidence is considered separately in Vitamin D and Innate Responses: Recognition, Peptides and Function: Full Evidence Review. Changes in immune signals or laboratory defence cannot replace an infant trial of reported illness.

Body composition: retain the positive finding without enlarging it

Razaghi and colleagues reported body composition from the same low-birth-status Montreal trial later used for the bone analysis. They found modestly higher absolute lean mass at 12 months with 1,000 rather than 400 IU daily, but not at three months. There was no corresponding advantage in percentage lean mass, the length-adjusted lean-mass index or fat outcomes (Razaghi et al., 2022).

Absolute lean mass is a body-composition measurement, not a test of muscle strength or developmental ability. The combination of findings does not establish better physical function, prevention of obesity or a lasting developmental effect. The unavailable supplement and difficulties interpreting the adjusted table restrict this review to the cautious qualitative main-report finding, without precise adjusted effects or subgroup claims.

This positive signal matters because a review of the bone findings alone could leave the false impression that every measured outcome was null. It deserves inclusion alongside its limitations. It is one result from one trial, not independent confirmation added to the Gharibeh report. Whether it reflects a clinically meaningful benefit requires evidence beyond the endpoint reported here.

How to interpret the whole pattern

The strongest consistent distinction is between raising circulating status and demonstrating added benefit in a specific outcome. Term-infant dose comparisons did not show added bone benefit, including in a trial selected for low birth status. The deficient-setting maternal trial found fewer biochemical rickets cases with a combined prenatal and postpartum strategy. The same trial did not show improved length. These results concern different populations, interventions and endpoints, so they do not require a single all-positive or all-null explanation.

Blood calcium, phosphate, hormone measurements and imaging contribute different information. None should be used as a substitute for a clinical outcome that the study did not assess. Similarly, the infection null and the limited lean-mass signal cannot be assembled into a claim that vitamin D either improves or has no effect on “all systems”. Each conclusion needs its own population, comparison and time point.

This is a selected evidence review, not a systematic review or a treatment guideline. It supports careful interpretation of infant trials while preserving uncertainty, meaningful positive findings and credible nulls. It does not establish an optimum dose, a universal blood target or a personal supplementation plan. Its central finding is that the biological importance of vitamin D and the benefit of increasing a particular dose are separate scientific questions.

Sources and access

Full principal reports and claim-relevant material were assessed. Unavailable MDIG subgroup supplements and Razaghi supplementary detail are not used to claim subgroup effects or comprehensive robustness. The corrected MDIG parent report informs the growth interpretation. Gharibeh and Razaghi are one trial, and the two MDIG reports are one trial. The study-title links below identify the publications rather than endorse every implication of their titles.

Effect of Different Dosages of Oral Vitamin D Supplementation on Vitamin D Status in Healthy, Breastfed Infants: A Randomized Trial. Gallo S et al. JAMA. 2013;309:1785 to 1792. DOI: 10.1001/jama.2013.3404.

Effect of Higher vs Standard Dosage of Vitamin D3 Supplementation on Bone Strength and Infection in Healthy Infants: A Randomized Clinical Trial. Rosendahl J et al. JAMA Pediatrics. 2018;172:646 to 654. DOI: 10.1001/jamapediatrics.2018.0602.

Effect of Vitamin D Supplementation on Bone Mass in Infants With 25-Hydroxyvitamin D Concentrations Less Than 50 nmol/L: A Prespecified Secondary Analysis of a Randomized Clinical Trial. Gharibeh N et al. JAMA Pediatrics. 2023;177:353 to 362. DOI: 10.1001/jamapediatrics.2022.5837.

Maternal Vitamin D Supplementation and Infantile Rickets: Secondary Analysis of a Randomized Trial. Lautatzis ME et al. Pediatrics. 2024;153:e2023063263.

Vitamin D Supplementation in Pregnancy and Lactation and Infant Growth. Roth DE et al. New England Journal of Medicine. 2018;379:535 to 546. Same MDIG trial; read with the 2021 correction.

Vitamin D Supplementation in Pregnancy and Lactation and Infant Growth. Correction. New England Journal of Medicine. 2021;385:1728. Parity coding and infant sex records corrected; not a separate trial.

Improving Vitamin D Status in Preterm Newborns: A Randomized Trial of 800 vs. 400 IU/Day. Rueang-Amnat N et al. Nutrients. 2025;17:1888. DOI: 10.3390/nu17111888.

Correction of neonatal vitamin D status using 1000 IU vitamin D/d increased lean body mass by 12 months of age compared with 400 IU/d: a randomized controlled trial. Razaghi M et al. American Journal of Clinical Nutrition. 2022;115:1612 to 1625. Same trial as Gharibeh; the title refers to correcting status, not an erratum.

Related reading

Original sources

Source

Vitamins for children

For current UK guidance, see NHS Vitamins for children.

Publication details

For current UK guidance, see NHS Vitamins for children. Infants born early need advice from their neonatal team.

Read more →
Source

Effect of Vitamin D Supplementation on Bone Mass in Infants With 25-Hydroxyvitamin D Concentrations Less Than 50 nmol/L: A Prespecified Secondary Analysis of a Randomized Clinical Trial

Published in Gharibeh N et al. JAMA Pediatrics (2023).

Publication details

Effect of Vitamin D Supplementation on Bone Mass in Infants With 25-Hydroxyvitamin D Concentrations Less Than 50 nmol/L: A Prespecified Secondary Analysis of a Randomized Clinical Trial. Gharibeh N et al. JAMA Pediatrics. 2023;177:353 to 362. DOI: 10.1001/jamapediatrics.2022.5837.

Read more →
Source

Correction of neonatal vitamin D status using 1000 IU vitamin D/d increased lean body mass by 12 months of age compared with 400 IU/d: a randomized controlled trial

Published in Razaghi M et al. American Journal of Clinical Nutrition (2022).

Publication details

Correction of neonatal vitamin D status using 1000 IU vitamin D/d increased lean body mass by 12 months of age compared with 400 IU/d: a randomized controlled trial. Razaghi M et al. American Journal of Clinical Nutrition. 2022;115:1612 to 1625. Same trial as Gharibeh; the title refers to correcting status, not an erratum.

Read more →