Vitamin D3: Optimal Levels, Dosing and Testing

What Vitamin D Actually Is

Vitamin D is not really a vitamin. Ultraviolet B radiation converts 7-dehydrocholesterol in skin into cholecalciferol, which the liver hydroxylates to 25-hydroxyvitamin D, the circulating storage form that laboratories measure, and which the kidney and many other tissues then convert to 1,25-dihydroxyvitamin D, the active hormone. That hormone binds a nuclear receptor and alters gene transcription across immune cells, bone, muscle, gut and vascular tissue.

Two features explain most of the confusion in the field. First, the storage form has a half-life of two to three weeks, so concentrations change slowly and single measurements are reasonably stable. Second, local tissue conversion means that immune cells can generate active hormone for their own use, which is why immune outcomes track with the circulating storage form even though the active hormone is tightly regulated.

Latitude, skin pigmentation, age, body composition and sun behaviour all affect production. Above roughly 37 degrees of latitude, winter ultraviolet B is insufficient for meaningful synthesis regardless of time spent outdoors. Adipose tissue sequesters vitamin D, so a higher body mass generally requires a higher intake for the same serum concentration. Skin synthesis capacity declines substantially with age.

What the Large Trials Showed

Observational data linking low vitamin D concentrations to almost every adverse outcome created enormous expectations. The large randomised trials that followed were, on the whole, sobering, and understanding why is more useful than either the enthusiasm or the backlash.

A large United States trial of 2,000 IU per day over roughly five years found no reduction in cancer incidence or cardiovascular events as primary endpoints, with signals in secondary analyses for cancer mortality and for autoimmune conditions. A large trial in Australia using monthly high-dose supplementation found no mortality benefit. Trials of very high intermittent dosing in older adults have in some cases reported more falls, not fewer.

The critical detail is baseline status. Participants in these trials largely started with adequate concentrations, because population fortification and widespread supplementation have shifted the baseline. Trials that enrol people with genuinely low concentrations continue to show benefit on bone, on respiratory infection frequency in meta-analysis, and on markers of immune regulation. The trials did not show that vitamin D does not matter. They showed that adding more to someone who has enough does not help, which is exactly what a nutrient rather than a drug should be expected to do.

Target Concentrations and Where Guidance Diverges

Guidance bodies disagree, and the disagreement is about purpose rather than about data.

• Below 20 ng/mL (50 nmol/L). Widely agreed as inadequate. Associated with impaired calcium absorption and bone remodelling.
• 20 to 30 ng/mL (50 to 75 nmol/L). Sufficient for bone health in most population-level guidance, and the point at which parathyroid hormone typically stabilises.
• 30 to 50 ng/mL (75 to 125 nmol/L). The range most optimisation-oriented practice targets, on the basis of immune and metabolic association data rather than randomised outcome data.
• Above 60 ng/mL (150 nmol/L). No demonstrated additional benefit. Above roughly 100 ng/mL, calcium handling concerns become real.

The honest framing is that 30 to 40 ng/mL is defensible and 30 ng/mL is where the strongest evidence sits. Chasing 70 or 80 ng/mL is not supported by outcome data and carries a real, if small, calcium load consideration.

Dosing, Titration and Cofactors

Response to a given dose varies severalfold between individuals, driven by body composition, absorption, genetics of the binding protein and baseline status. This is why fixed-dose advice is weak and measurement is strong.

As a rough starting point, 1,000 IU per day raises serum 25-hydroxyvitamin D by roughly 5 to 10 ng/mL in an average adult, with diminishing returns at higher intakes. Common maintenance intakes fall between 1,000 and 4,000 IU per day, the latter being the tolerable upper intake level in most guidance. Higher doses are used for repletion under clinical supervision.

Three practical points. Vitamin D3 is more effective than D2 at raising and maintaining serum concentrations. Absorption is fat-dependent, so take it with the largest meal of the day, which can change serum response substantially compared with a fasted dose. And daily dosing is preferable to large intermittent boluses, since several trials using monthly or annual high doses produced null or unfavourable results, possibly because a large bolus disrupts the enzymatic regulation of the active hormone.

Cofactors that actually matter

Vitamin D increases calcium absorption. Where that calcium goes depends partly on vitamin K2, which activates matrix Gla protein in vascular tissue and osteocalcin in bone. This is the mechanistic basis for pairing the two, though controlled human evidence for the combination remains limited. Magnesium is required as a cofactor for the enzymes that hydroxylate vitamin D, so magnesium status can influence how well a given dose translates into circulating 25-hydroxyvitamin D.

Testing: What, When and How Often

The correct test is serum 25-hydroxyvitamin D, reported in ng/mL or nmol/L. It is widely available, inexpensive and stable enough that a single value is informative. Measuring 1,25-dihydroxyvitamin D is not useful for status assessment, because it is tightly regulated and can appear normal or high when stores are low.

Practical timing: measure before starting, then again after eight to twelve weeks on a stable dose, since serum concentrations take that long to plateau. After that, twice a year is sufficient for most people, ideally once in late winter, which is the annual low point, and once in late summer, the annual high. If a single measurement per year is all that is available, late winter is more informative.

Season, latitude, holidays in sunny places and changes in body composition all shift the result, which is a reason to interpret a value in the context of when it was taken rather than as a fixed personal number.

Safety, Upper Limits and Who Needs Caution

Vitamin D toxicity is rare and requires sustained very high intake, typically above 10,000 IU per day for months, or a dosing error. It presents through hypercalcaemia: nausea, excessive thirst and urination, confusion, and in severe cases kidney involvement. The margin between common supplemental doses and toxicity is wide, but it is not infinite, and fat solubility means it accumulates.

Specific caution applies to several groups. Anyone with granulomatous conditions such as sarcoidosis, certain lymphomas, or primary hyperparathyroidism can convert vitamin D to the active hormone in an unregulated way and become hypercalcaemic at ordinary doses, so supplementation in these situations belongs under clinical supervision. Thiazide diuretics reduce calcium excretion and can compound the effect. Anyone taking high-dose vitamin D alongside calcium supplementation should be measuring rather than assuming.

The Temporal View

Vitamin D is the clearest example of a compound whose correct dose is a moving target. It changes with season, with latitude if someone travels or relocates, with body composition, with age, and with sun behaviour. A dose established in February is frequently wrong by July.

A sensible structure is: measure, titrate to a target in the 30 to 40 ng/mL range, hold steady, re-measure in the opposite season, and adjust. That is two measurements a year and one or two dose changes, which is a modest amount of effort for the compound where individual variation in response is largest.

The failure mode here is not overdosing. It is taking a fixed 5,000 IU per day for a decade because someone once said it was the right number, without ever checking whether it produced a sensible concentration in this particular body.