Vitamin D3 for Life

Live Enlightened

Article

Calcium Availability and Vitamin D Response in Nutritional Rickets: Full Study

Previous related pageVitamin D in Nutrient NetworksNext related pageCalcium Availability and Vitamin D Response in Nutritional Rickets: Study Summary
Study at a glance

Did adding vitamin D help children receiving calcium for rickets?

The trial involved 72 Nigerian children with nutritional rickets, a condition in which growing bones do not harden properly. Their calcium intake was very low. All received calcium treatment; chance determined who also received vitamin D2. The trial tested the added effect of vitamin D during calcium treatment, not vitamin D alone or prevention in healthy children.

Did adding vitamin D improve recovery from rickets?

At 24 weeks, 67% receiving both nutrients and 44% receiving calcium alone met the recovery definition, which combined an X-ray score with a blood marker. The direct comparison narrowly missed the usual statistical threshold. An analysis accounting for starting differences favoured combined treatment, but its estimate had a wide range of uncertainty. Recovery was more frequent with combined treatment at 12 weeks, suggesting faster improvement. However, one forearm mineral measurement favoured calcium alone.

What did the other rickets and calcium studies add?

A small absorption experiment found that active vitamin D rose without a measurable increase in the percentage of calcium absorbed. Another found a lower absorption percentage but estimated that higher intake would still supply more absorbed calcium. A calcium-dose trial found different responses for X-ray healing and bone density. A larger prevention trial in Afghan infants found no significant overall reduction in rickets or improvement in growth with intermittent vitamin D.

What are the limits of these rickets treatment findings?

The main trial had small, unequal groups and was registered after completion. It did not measure the processes that might explain recovery. The studies used different populations, treatments and outcomes, so they cannot be treated as one experiment. Achieved blood vitamin D levels do not establish a personal target. The treatments tested in research are not instructions for treating a child at home or adding other supplements.

Read the full article below ↓

Nutritional rickets is a disorder in which a child’s growing bones and growth plates fail to mineralise properly because the supply or handling of essential nutrients is inadequate. This review examines a trial of vitamin D added to calcium treatment, then compares absorption and prevention studies. The question is how calcium availability changes the interpretation of vitamin D response, from a blood measurement to skeletal recovery.

The question is about a nutrient relationship

A response to vitamin D can mean several different things: a higher blood concentration, altered intestinal absorption, faster improvement on an X-ray, or a lasting improvement in health. These outcomes are connected, but one cannot automatically stand in for another. An experiment needs to specify which response it measures before its findings can be interpreted.

The calcium question makes that distinction concrete. Even efficient absorption can deliver little calcium when little is eaten. Conversely, increasing the calcium supply can increase the amount absorbed without increasing the percentage absorbed. Nutrient availability therefore belongs in the explanation of a response, alongside the intervention itself. This is the relationship explored more broadly in Vitamin D in Nutrient Networks.

The trial considered here addresses treatment in children with established disease. Its comparison asks about the additional contribution of vitamin D during calcium treatment. It cannot, by that design, establish what vitamin D would accomplish without calcium treatment, or what either intervention would accomplish in a healthy child.

The treatment trial and its findings

Thacher and colleagues’ 2014 trial enrolled 72 Nigerian children with radiographically active rickets who attended a clinic with leg deformities. All were older than one year; median ages were 55 months in the combined-treatment group and 42 months in the calcium group. They had very low calcium intakes, approximately 170 mg/day, and substantial short stature. Most had regular sunlight exposure. This was a specific nutritional and disease phenotype, not a sample of healthy children identified through a low vitamin D test.

The mean starting concentration of 25-hydroxyvitamin D, abbreviated 25(OH)D, was 30.2 nmol/L. This circulating metabolite is commonly used to assess vitamin D status; it is different from the active hormone, 1,25-dihydroxyvitamin D. The baseline measurement establishes the trial’s starting context without proving that one nutrient alone caused every child’s rickets.

Coin-toss allocation produced 44 children receiving calcium plus vitamin D and 28 receiving calcium with B-complex comparator tablets. Both groups were prescribed about 938 mg elemental calcium twice daily as tested limestone mixed with food. The combined group also received 50,000 IU vitamin D₂ every four weeks for 24 weeks. The vitamin D doses were directly observed; calcium adherence was assessed by weighing unused limestone at return visits.

Sixty-eight children completed treatment. Estimated median calcium adherence was 96% with combined treatment and 99% with calcium alone. Returned-container weights measure disappearance of the preparation, not proof that every child finished every meal containing it. This matters because the amount prescribed and the amount actually swallowed may differ.

Recovery required both an improved X-ray score and a serum alkaline phosphatase concentration at or below the study’s threshold. Alkaline phosphatase is an enzyme associated with bone formation that is often elevated in active rickets. Its inclusion added a biochemical measure of recovery to the structural assessment. These were trial endpoint definitions, not universal diagnostic cutoffs for every child.

MeasureReported finding
Main endpointRadiographic score ≤1.5 plus alkaline phosphatase ≤350 U/L.
24-week result29/43 (67%) with calcium plus vitamin D versus 11/25 (44%) with calcium; unadjusted p=0.06.
Adjusted analysisOdds ratio 5.4; 95% confidence interval 1.3 to 27; p=0.02. Model adjusted for baseline alkaline phosphatase, radiographic score, 25(OH)D and height for age.
Earlier recoveryAt 12 weeks, 26% versus 4.2% met the combined endpoint; p=0.03.
Individual endpointsAdjusted radiographic scores differed significantly at 12 weeks (p=0.01), but not at 24 weeks (p=0.12). Alkaline phosphatase differences were not significant.
Vitamin D measurementsFinal mean 25(OH)D: 55.4 versus 37.9 nmol/L. Baseline status did not detectably modify treatment response.
Contrasting skeletal findingAdjusted distal forearm mineral gains favoured calcium alone: 0.093 versus 0.053 g; p=0.02. Estimates adjusted for baseline mineral content and height; proximal gains did not differ significantly.

How to read the evidence without overstating it

The associated registry, NCT00949832, was first posted in July 2009, after its recorded study completion in April 2007. It describes 109 participants and a broader set of investigations, so it is not a one-to-one record of the published comparison. Its retrospective timing limits independent confirmation that the published outcome definitions and analyses were specified before results were known. The registry identifies participant masking; it does not establish blinded outcome assessment for this comparison.

The direct 24-week comparison remained uncertain at the conventional 0.05 significance threshold, while the adjusted model favoured combined treatment. The model accounted for initial biochemical severity, radiographic severity, vitamin D status and height for age. These are clinically relevant starting differences. Adjustment nevertheless introduces model assumptions, and the wide confidence interval, from 1.3 to 27, shows how imprecisely the treatment association was estimated.

The trial originally planned for 40 children per group and fell short, particularly in the calcium-only arm. Randomisation supports a causal comparison, but unequal small groups can retain important chance imbalances. Reporting both the observed counts and the adjusted result is more informative than selecting whichever analysis gives the stronger headline.

Similarly, failure to detect an interaction with starting vitamin D status does not establish identical benefits at every starting concentration. An interaction analysis can miss a real difference when the information available is limited.

The odds ratio is not a multiplication of the probability of recovery. Recovery was already common, so saying that vitamin D made recovery “5.4 times as likely” would misrepresent the model. The observed proportions, 67% and 44%, are the clearer starting point; the adjusted odds ratio supplies a separate analysis of the same limited dataset.

Requiring both endpoint criteria means a child could improve on an X-ray without yet being classified as recovered. The significant 12-week radiographic difference, followed by a nonsignificant 24-week difference, makes recovery timing part of the result. The evidence is consistent with earlier improvement, rather than a demonstrated improvement in every final skeletal measurement.

The different skeletal measurements also deserve independent attention. A radiographic rickets score assesses features of disease at growing bone ends. A mineral-content measurement quantifies mineral in the scanned region. Neither is a complete description of bone health. A favourable result for one does not erase a contrary result for the other, and neither alone establishes a reduction in future fractures.

Taken together, the trial supports a possible acceleration of recovery with combined treatment, while leaving uncertainty about its final magnitude and the consistency of skeletal benefit. That is a narrower conclusion than a claim that vitamin D improves every aspect of calcium handling.

Why a higher vitamin D measurement need not mean greater absorption

A separate experiment by Thacher and colleagues in 2009 examined 17 Nigerian children with rickets using oral and intravenous stable calcium isotopes. Children received a single 50,000 IU dose of either vitamin D₂ or D₃; absorption was assessed before dosing and three days afterwards. Mean fractional calcium absorption was essentially unchanged, from 52.6% to 53.2%, despite an increase in circulating 1,25-dihydroxyvitamin D. The two vitamin forms did not produce detectably different absorption responses.

This was a small, short experiment without an untreated comparison group. It cannot establish long-term equivalence between D₂ and D₃, exclude effects in severe deficiency, or explain an entire course of skeletal healing. Its useful contribution is more specific: a vitamin D metabolite can rise without a measurable increase in calcium absorption under the tested conditions.

That distinction helps frame Vitamin D Testing: 25-OH-D vs 1,25-OH₂-D. The measurements refer to different metabolites. Neither number, by itself, reports how many milligrams of calcium have crossed the intestine or how well a growth plate is recovering.

There is no logical contradiction between an unchanged short-term absorption measurement and a possible improvement in recovery over months. They concern different outcomes and observation periods. Establishing the connection would require measurements of the proposed intermediate process alongside the clinical outcome, preferably within the same study. Without that evidence, a mechanism remains an explanation to test rather than an observed result.

Absorption efficiency and calcium supply must be separated

Oramasionwu and colleagues’ 2008 study followed calcium absorption during treatment rather than comparing vitamin D preparations. Ten children provided usable paired measurements after two weeks of calcium supplementation. Median fractional absorption fell from 72% to 57%. The investigators nevertheless estimated that the increased calcium intake would deliver substantially more absorbed calcium. They did not directly measure absorption from every component of the children’s daily intake.

This small before-and-after study had no untreated control, and only eight children remained for the three-month assessment. Its test-meal measurements should not be treated as exact measurements of whole-day absorption. It nevertheless illustrates why a lower absorption percentage need not represent a worse nutritional result.

Consider an illustrative calculation, separate from the study data. Absorbing 60% of 200 mg supplies 120 mg; absorbing 40% of 600 mg supplies 240 mg. The percentage has fallen while the amount has doubled. This arithmetic does not predict an individual child’s absorption, but it shows why the denominator matters.

Nor does the absorbed amount equal the amount retained in bone. Retention also depends on losses and on how the body uses the absorbed mineral. Intake, absorption and retention are separate quantities. Vitamin D and Calcium provides the wider physiological setting for those relationships.

The distinction also changes how a plausible explanation should be tested. If the proposal is that a treatment increases intestinal efficiency, the relevant measurement is fractional absorption. If the proposal is that it supplies more calcium for recovery, total availability and retention become relevant. A study of one quantity cannot automatically verify the other proposal.

What a calcium-dose trial adds

In a trial published in 2016, Thacher and colleagues randomised 65 children to 500, 1,000 or 2,000 mg calcium daily for 24 weeks. Radiographic improvement was faster with the two higher doses than with 500 mg, but 2,000 mg did not demonstrate faster radiographic healing than 1,000 mg. The proportions reaching the radiographic threshold at the final visit did not differ significantly. Serum 25(OH)D was not measured; participants were unblinded, although radiograph readers were masked.

The highest dose did, however, produce a faster increase in forearm diaphyseal areal bone mineral density than either lower dose. The diaphysis is the shaft of a long bone; areal density describes mineral relative to the scanned area. This is a different outcome from recovery of the growth-plate abnormalities scored on radiographs. Omitting it would make the dose-response evidence appear more uniformly flat than it was.

The balanced interpretation is endpoint-specific: increasing calcium beyond the middle dose did not accelerate radiographic healing, but it did change another skeletal measurement. Neither outcome establishes a universally optimal dose, and unmeasured vitamin D status prevents a clean assessment of how that status modified the response.

A dose comparison and an add-on comparison have different evidential roles. One asks which of several calcium regimens performs better. The other asks whether vitamin D contributes when calcium is already provided. Combining their insights is useful; treating their treatment groups as interchangeable arms of one larger randomised experiment would not be justified.

A larger prevention trial gives a different perspective

Crowe and colleagues’ Afghan trial randomised 3,046 infants to vitamin D₃ or placebo every three months for 18 months. Rickets was assessed in 631 children with usable radiographs. The reported prevalence was 5.3% with vitamin D and 5.5% with placebo, with an odds ratio of 0.96 and a 95% confidence interval of 0.48 to 1.92. Overall growth did not improve significantly.

A borderline height finding appeared in children with higher dietary calcium intake. However, the formal test of whether the treatment effect differed by calcium intake was not significant, with p=0.355. Calcium intake was estimated rather than randomised. That subgroup result therefore does not establish that adequate calcium caused a better vitamin D response.

This was a prevention question in younger children, using intermittent dosing, rather than treatment of established rickets with calcium supplied. Those differences limit direct comparison with the treatment trial. They also explain why the larger study cannot simply be treated as a decisive refutation or confirmation of the smaller one.

The broader issue is explored in Why Vitamin D Studies Disagree. Before combining apparently conflicting findings, identify the population, background nutrition, treatment, comparator, outcome and follow-up period. Agreement on the phrase “vitamin D trial” does not make the underlying questions identical.

What these studies leave unresolved about nutrient networks

The evidence directly examined here concerns calcium supply, vitamin D exposure, absorption and skeletal outcomes. It does not establish a universal combination of nutrients that maximises vitamin D signalling. To make that claim, an experiment would need to measure the relevant nutrients and distinguish their individual contributions from their interactions.

Phosphate belongs in the physiological explanation of mineralisation and growth-plate recovery. The Afghan manuscript discusses how disturbed calcium balance can be accompanied by hormonal changes that increase renal phosphate loss. That proposed pathway should be distinguished from a direct demonstration that it explains a particular treatment response. Vitamin D and Phosphate Regulation develops the underlying regulatory relationships.

Magnesium raises a separate question. Whether magnesium availability limits an enzyme-dependent process cannot be established from a comparison that does not manipulate magnesium or adequately characterise its status. The relevant background is covered in Vitamin D and Magnesium Function. A calcium treatment result provides no basis for automatically adding magnesium, vitamin K or other nutrients to the tested regimen.

The focal trial did not measure calcium absorption, parathyroid hormone or fibroblast growth factor 23, a hormone involved in phosphate regulation. It therefore cannot identify which intermediate pathway caused the treatment difference. The complementary evidence is brought together in Vitamin D in Nutrient Networks: Availability, Metabolism and Biological Response, which separates mineral supply from metabolism, receptor activity and clinical outcomes.

Why these results do not establish a universal blood-test target

The trial itself supplies the reason to discuss achieved blood concentrations. Children who met the recovery endpoint had a mean final 25(OH)D of 56.4 nmol/L, compared with 37.7 nmol/L among those who did not, with p<0.001. That is an observed association within this experiment. It is not evidence that 56.4 nmol/L is a treatment target: it is a group average after treatment, not a concentration to which children were randomly assigned.

The concentration reached can reflect treatment delivery, absorption, starting status and other characteristics related to recovery. Grouping children by that final concentration no longer preserves the original randomised comparison. A trial assigning different concentration targets would be needed to test whether deliberately maintaining a particular target improves outcomes.

This is an application of Correlation vs Causation in Vitamin D. An association between a final concentration and recovery can be informative without showing that deliberately raising everyone to that concentration would reproduce the association. A causal claim about a target requires a design capable of testing the target.

The distinction explained in Vitamin D Thresholds vs Targets is therefore practical here: an endpoint definition, a diagnostic cutoff and an average concentration among responders answer different questions. None can simply be substituted for the others.

The reader should therefore keep three questions separate: what was measured, what intervention was assigned, and what outcome changed. Vitamin D Status vs Vitamin D Effect addresses that distinction. Describing a blood concentration as “better” is incomplete unless the intended biological or clinical benefit is specified.

Where the findings can and cannot be applied

The strongest application is to the specific treatment question: children with established nutritional rickets and very low calcium intake. Extending the findings to healthy adults, athletes, cardiovascular prevention or routine supplement combinations would require additional evidence. A shared nutrient does not make those populations or outcomes equivalent.

Childhood also introduces a developmental context. A growing skeleton presents questions about growth plates and mineral accrual that are not captured by an adult blood-test comparison. Vitamin D in Childhood Development provides that wider context. Findings about radiographic healing should not be silently converted into claims about final adult height or lifelong skeletal strength.

Treatment doses reported in research describe the experimental protocols; they are not general intake recommendations. In particular, the limestone preparation used in the focal trial was tested for its calcium content and contaminants. That detail cannot be transferred to an untested material. Nutritional rickets requires clinical assessment, and these studies do not provide a basis for selecting or escalating a child’s treatment at home.

The conclusion supported by the evidence

Calcium availability is part of the context in which vitamin D responses must be interpreted. The treatment evidence suggests that adding vitamin D during calcium treatment may speed recovery from nutritional rickets, but the statistical uncertainty and differing skeletal outcomes limit the strength of that conclusion. Absorption experiments show why a metabolite change cannot substitute for a functional measurement, while prevention evidence limits generalisation beyond the treatment setting. The defensible lesson is to interpret the nutrient supply, intervention and measured outcome together, without converting their relationship into a universal supplement formula or blood-test target.

For the shorter account of the trial’s question, findings and practical limits, read Calcium Availability and Vitamin D Response in Nutritional Rickets: Study Summary.

Primary source and publication details

Related reading

Original sources

Trial registry

Vitamin D and Genetics in Nutritional Rickets

Mayo Clinic. Vitamin D and Genetics in Nutritional Rickets.

Publication details

Mayo Clinic. Vitamin D and Genetics in Nutritional Rickets. ClinicalTrials.gov identifier: NCT00949832. Registry record; not an additional independent trial.

Read more →