Table 1 Parameters and conditions that should be controlled for an optimal evaluation of the effects of vitamin D in future RCTs

Conditions allowing an optimisation of the statistical power of the study (common conditions for trials of drugs and nutrients)

Sample size (number of participants) and trial duration must be appropriately calculated according to the frequency of the studied event in the recruited population. These points depend on the basal clinical status of the patients (larger sample and/or longer duration if the studied disease is not very active in the recruited patients)

Adherence/observance must be optimized (for example, new technologies such as SMS that are sent the day just before the treatment must be taken, in case of intermittent dosage, allow an easy reminding for the patients)

Conditions specific to a vitamin D trial

Choose to administrate vitamin D3 instead of D2, specially in case of intermittent dosage

Ensure that dietary calcium intakes of the participants are sufficient

Treat with daily doses or, in case of intermittent dosage, choose doses that are not too high (<or = 100000 IU) and not too spaced out (ideally < or = 1 month)

Choosing the dose will depend on the disease to be studied (search in the literature) but must be above 800 IU/day (often more)

Possible vitamin D supplements that were taken by the patients before the study must be stopped (paradoxically, it was found in some studies with a poor observance that some patients in the placebo group received in fact more vitamin D during the trial than some patients in the vitamin D group)

It will be important to recruit patients with low 25OHD serum levels (or at least much lower than the 25OHD levels that are targeted in the study) so that a frank increase of the 25OHD concentration may be observed on the one hand, and so that the placebo group is really insufficient/deficient on the other hand.