Covid-19 & Vitamin D: UK unhelpful clinical trial

London – Greenwich and Canary Wharf

It has been established very well during the Covid-19 pandemic that a low blood levels of Vitamin D recorded in advance of illness (Santander, Heidelberg, Israel ) is the most significant but potentially reversible risk indicator of critical disease and death. Vitamin D deficiency is also the common factor within the major at-risk groups, the obese, the elderly, and those of South Asian and Black African ethnicity who live in the UK and other northern European countries. 

It has been established that Vitamin D in its "raw" form when given to the critically ill in Intensive Care Units (ICU) is not effective (Brazil), the simple reason being that as we know, it takes up to two weeks for the Vitamin D to be converted in the liver into its hydroxylated pre-active form, 25(OH)D, also known as Calcifediol. When Calcifediol is given by mouth it takes just two hours to achieve good blood levels. It is Calcifediol, 25(OH)D that we measure in the blood as Vitamin D.

Randomised controlled trials (Cordoba and Barcelona) have shown that when Calcifediol is given to patients admitted to hospital on account of severe Covid-19, it is remarkably successful in reducing the risk of escalation to ventilatory care and death. Why it has not been used widely in the UK and elsewhere is a mystery.

Correction of vitamin D deficiency

It would be a sensible public health initiative to correct Vitamin D deficiency within the population, especially in those at particular risk. A "safe" level of Vitamin D in the blood should be achieved, ideally before admission to hospital. Vitamin D is cheap, readily available, and safe when given in physiological doses checked by blood levels. 

It is clear that a low blood level of vitamin D (=Vitamin D deficiency) can be a serious disadvantage, and restoration to an ideal blood level of 40ng/ml (100nmol/L) would be sensible. But how to achieve this is the challenge. 

The physiological way to obtain Vitamin D is by exposure of the skin to the sun, at times when at close to sea level the sun is more than 45 degrees above the horizon, judged by the length of the shadow being less that the height. This could be recommended but an excessive single exposure leading to burn is to be avoided.

It has been demonstrated that not only people with ideal blood levels of Vitamin D have a reduced risk of critical or fatal Covid-19, but also people who are taking vitamin D supplements at the time of hospital admission on account of Covid-19 (Tameside, UK). 

The biology of the pivotal role of Vitamin D, in its fully activated form 1,25(OH)D, in the escalation of natural defensive immunity has been established during the past forty years.

With the onset of a serious pandemic due to a new respiratory virus, it would seem to be sensible to act on a knowledge of medical science supported by early reports of the dangers of low blood levels of Vitamin D. Correcting Vitamin D deficiency would have been of great potential benefit, with no danger. I have drawn attention to the report of just a single case of vitamin D toxicity occurring during the pandemic.

Randomised controlled trial – RCT – from UK

Official advice concerning the use of Vitamin D during the pandemic was very negative, ridiculously so given the circumstances and 200,000 Covid-19 deaths in the UK. The pandemic emergency resulted in vaccines being issued on the basis of Emergency Use Authorisation (EUA), meaning that only short term and rudimentary RCTs in healthy people had been completed. 

When it came to vaccinations in pregnancy there were no RCTs, which can be regarded as negligent. Safety was to be judged on the basis of "post–marketing surveillance". And this is the important thing: the main purpose of RCTs is to establish safety, following the thalidomide tragedy of the 1960s. When taking medicines, effectiveness is less importance than safety. A medicine might not do good, but it must not do harm – primum non nocere.

Whereas the "vaccines" against Covid-19 are a new RNA technology with safety being far from clear, Vitamin D is a natural substance that has been in use for a century. The standard dose range is known, with appropriate blood levels. The exceptionally rare side-effect of Vitamin D intoxication (hypervitaminosis D) is always the result of major dose error and is both easily recognised and completely reversible. The need for RCTs of Vitamin D was much less pressing than for vaccines.

The challenge of prospective randomised trials of vitamin D, or anything else, is that if low-risk individuals are to be studied, very large numbers must be recruited if the endpoint of critical illness or death is to be reached in significant numbers. Administration costs are inevitably very high, and follow-up times will be long. It imight be sensible to concentrate on a smaller number of individuals who are at high risk of serious or fatal Covid-19. Hence the success of the RCTs from Spain, small numbers showing a big effect with statistical significance.

I have also shown in previous Blog posts the wisdom of Sir Austin Bradford Hill, whose criteria of pragmatic "proof" (it is never absolute) identifies an RCT as being only one component. For ethical and operational reasons an RCT might not be possible, or be of only limited value. There is also the wisdom of Blaise Pascal, who realised that decision-making is also pragmatic, a trade-off between likely benefit and possible adverse outcome.

United Kingdom - the COVIDENCE UK study

Early in the pandemic in the UK, a large community-based clinical trial of Vitamin D in the prevention of Covid-19 was proposed by Professor Adrian Martineau. Funding of the study became an immediate problem as a large administrative staff would be essential and there would be no pharmaceutical company support. However the study was undertaken and the result became available in early 2022.

6200 adults were incorporated into the trial, with a number of exclusion criteria including current Vitamin D supplementation. 2958 of these were randomised to receive Vitamin D.

2690 (86.8%) of those to receive VItamin D had blood Vitamin D levels less than 30ng/ml, 75nmol/ml. This shows the extent of clinically significant Vitamin D deficiency. They had suboptimal blood levels as identified in the observational studies described in Heidelberg, Israel, and other places. 

1334 were given a lower dose of vitamin D supplement, 800 units per day, and 1356 were given a higher dose, 3200 units per day.

The controls did not have blood testing and they were informed that they were not to receive a Vitamin D supplement. This is important as, with the Heidelberg and Israel studies, the identification of Vitamin D deficiency before the study would have created the ethical dilemma if they were not to have the deficiency corrected.

There was no placebo given to the control group. This weakened the trial and in the knowledge that they were not to receive Vitamin D that was being tested, the controls would be likely to take a Vitamin D suplement not given to them. In fact 49.9% of them did so on at least one occasion.

The trial commenced in December 2020, with six month follow-up of individuals. The vaccination process was under way at that time and 89.1% of the subjects received one or more doses of a Covid-19 vaccination during the study period. This would obviously complicate the interpretation of the results.

The baseline Vitamin D levels can be seen in Figure 1. Both units in current use are shown, nmol/L and ng/ml. We can see no significance difference between the two treatment groups at the baseline. We see again the high prevalence of low sub-optimal blood levels of Vitamin D.

Figure 1. Baseline blood levels of Vitamin D

In Figure 2 we can see the blood levels of Vitamin D at the end of the study. The ideal blood levels can be judged to be greater than 75nmol/L, 30ng/ml, and preferable to be 100nmol/L, 40ng/ml. We can see that the former level is achieved by taking 800units per day, and the higher level by taking 3,200units per day.

In previous posts I have generally advised a supplement of 3,000units per day (the dose that I take, but as 20,000units once each week) but I have also emphasised the importance of checking blood levels so as to monitor the appropriate dose of the supplement. I have drawn parallels with monitoring the treatment of other hormone deficiencies, namely thyroxine and TSH levels in hypothyroidism, and glucose in diabetes.

Figure 2. End of study blood levels of Vitamin D

It is interesting to note that mean blood level of Vitamin D in the "No offer" group at the end of the study was 66.6nmol/L, 26.6ng/ml. As mentioned above the baseline blood Vitamin D levels in the "No offer" group, the intended controls, were not measured. There is no reason why they should not have been the same as in the treatment groups, which is about 41nmol/L, 16.5ng/ml. The fact that at the end of the study the mean blood level was 66.6nmol/L, 26.6ng/ml, indicates that many of those in the "No offer" were taking Vitamin D supplement of their own initiative, as some admitted. The increase would be 66.5%.

Clinical outcome was of course reported in the results of the study. We can see in Figure 3 that there were no significant differences in the cases of Covid-19 in the three groups, 2.97% in the high dose group, 3.63% in the low dose group, and 2.64% in the "No offer" group. There was similarly no significant difference in hospital admission rate. Ventilation was necessary in only one in each group, and there were no deaths.

Figure 3. Covid-19 infections related to Vitamin D supplement

Figure 4 shows us that there were very few acute respiratory infections, with no significant differences between the groups. The infection rate was highest in the Vitamin D high dose group.

Figure 4. Acute respiratory infection related to Vitamin D supplement

Conclusions

The main conclusion of the study as mentioned in the paper is that a public health policy of Test and Treat Vitamin D Deficiency is a practical proposition. To quote from the paper:

"Ultimately, however, this trial was designed to investigate the effectiveness of a pragmatic 'test–and–treat' approach to boosting population vitamin D status, rather than  biological efficacy of vitamin D to prevent ARIs [acute respiratory infections], and our findings should be interpreted accordingly."

It is not possible to form a conclusion concerning the effectiveness or otherwise of Vitamin D supplementation, but it appears that in prevention among asymptomatic people, a dose of 3,200units is perhaps of no advantage over a lower dose of 800units each day. However the absence of Covid-19 and respiratory deaths in all groups indicates that the study was underpowered (too small) to assess effect on deaths. To do so it would have been necessary recruit at least ten times the number of subjects.

This was a field study and the supplement dose was not adjusted to meet the needs of individual people. It was "one size fits all", but there might be an advantage of personalised approach with intermediate blood testing.

Of greatest importance in assessing Vitamin D efficacy is that the study was compromised by the "control" group being uncontrolled. It would have been scientifically more robust if they had been given a placebo, but that was not the case. This might have been for ethical rather than operational reasons, and similarly no base-line Vitamin D testing for the "control" group. However for a "test-and-treat" evaluation, the management of controls was not important and indeed the presence of controls was not really necessary.

In practice, those randomised to be controls were informed that they were not to receive a Vitamin D supplement, termed "no offer". It would appear that they did not want to miss out on what would be a very safe and potentially great advantage in the face of the pandemic, and many accepted Pascal's Wager, that is they took Vitamin D in unknown amounts, but sufficient to increase the mean blood level of Vitamin D by 66.5%. 

The other confounding factor was the distribution of vaccines at the same time of the study. This was among all groups but it could confuse and diminish any assessment of clinical outcome.

Of 6,000 people taking Vitamin D supplement (including a significant but unknown number of "controls"), only three required ventilatory care for Covid-19 and there were no deaths. This in itself tells us the value of taking a Vitamin D supplement to correct deficiency.

Tracing for and testing of vitamin D deficiency was achieved in this study. 

Assessment of correction of Vitamin D deficiency was inconclusive.

There are of course several other studies, and a study from Mexico will be the subject of the next Blog post.