MR1451 Vitamin D and Longevity (VIDAL) Trial: randomised feasibility study

Vitamin D and Longevity (VIDAL) Trial is a randomised feasibility study started by the London School of Hygiene & Tropical Medicine (LSHTM). Instigated with support from interested collaborators in Queen Mary University of London and University College London. This is a feasibility study for a larger, main trial which currently has not received funding to be conducted. The question the main trial will address is whether vitamin D supplementation reduces morbidity and increases lifespan in men and women aged 65 to 84 in response to a growing literature base suggesting that such trials are required. The International Agency for Research on Cancer (IARC) Working Group on vitamin D and cancer, of which the chief investigator was a member, reviewed the epidemiological evidence on vitamin D and cancer and concluded that the evidence was strong for colorectal cancer but inconclusive for other individual cancers (IARC Working Group on Vitamin D, 2008). Their report concluded: “The only way to further address the cause-effect issue is to organise new randomized trials to evaluate the impact of vitamin D on all-cause mortality and on the incidence and mortality from common conditions including cancer. These trials should make sure that key parameters of vitamin D status (e.g., serum 25-hydroxyvitamin D levels before and in trial) can be assessed.”

Many people in the UK do not receive enough sun exposure to achieve a satisfactory vitamin D blood level. Epidemiological evidence suggests that people with low vitamin D levels are at increased risk for several diseases, including heart disease, various infections and some types of cancer. However, similar epidemiological evidence of benefit for various other vitamins was disproved when randomized trials were conducted. LSHTM is therefore planning a large (n=20,000) trial to evaluate the effect of vitamin D supplementation on mortality and morbidity among older members of the general population. For the primary endpoint of overall mortality in the main trial the necessity of placebo control is not clear.

The primary aim of the feasibility study is to establish the procedures required to conduct the main trial. We shall determine the recruitment rate and the compliance rate (the proportion adhering to allocated treatment over 2 years) in practices with placebo control, and in practices with open control. We shall also determine contamination rates in those

allocated to placebo and those allocated to open control (ie self-administration of >400 IU vitamin D per day or equivalent). IMP adherence and use of additional vitamin D supplements will be studied both by self-report and from blood levels of 25(OH) vitamin D at entry and at 2 years.

This feasibility study, which is funded by the Government’s Health Technology Assessment agency, will compare a placebo controlled trial in 10 GP practices against an open randomised trial in 10 practices, randomising 800 people aged 65-84 double blind to monthly placebo versus monthly vitamin D for 2 years, and 800 to no treatment versus 2 years of monthly vitamin D. Participants were selected from the GP area within the age ranges based on their availability for study processes conducted via the GP surgery as well as the consideration of multiple exclusion criteria based in medical history (for example particular diagnoses or vitamin levels in participants) and lifestyle choices (plans to move/emigrate or the taking of dietary supplements).

LSHTM currently holds identifiable, record level data against the total cohort of participants (1615) for the following datasets:

HES – Admitted Patient Care (2012/13 – 2018/19)

HES – Critical Care (2012/13 – 2018/19)

MRIS – Cause of Death Report

MRIS – Cohort Event Notification Report

MRIS – Flagging Current Status Report

One of the research objectives is to establish the feasibility of flagging for lifelong follow-up through national registers for death and cancer registration, and tracing through for hospital admissions and diagnoses. Comparing hospital admissions, cancer registrations and deaths between participants who have and have not taken vitamin D will answer the question of whether high-dose vitamin D supplementation might help to prolong life and reduce instances of various diseases.

The other primary objectives of the trial are to determine the recruitment rate and the compliance rate (the proportion adhering to allocated treatment over 2 years) in practices with placebo control, and in practices with open control. The institution shall also determine contamination rates in those allocated to placebo and those allocated to open control (ie self-administration of >400 IU vitamin D per day or equivalent). IMP adherence and use of additional vitamin D supplements were studied both by self-report and from blood levels of 25(OH) vitamin D at entry and at 2 years.

The outcomes of this feasibility study will assist with a decision as to the study design of the main trial by providing unbiased data on adverse events in each treatment arm. This will be compared to the (potentially) biased real-time collection of adverse events during the trial to see whether an open allocation trial (where people know which treatment they’ve been allocated to) is more biased in this regard than a blinded design (where patients either receive treatment or placebo). This in turn will help us to present a reasoned argument for the study design that should be adopted in the main trial.

LSHTM will be acting as the sole Data Controller and Processor. The Data Controller has assessed the purpose for processing this data and identified the legal basis for processing NHS Digital data under General Data Protection Regulation as Article 6(1)(e) ‘processing is necessary for the performance of a task carried out in the public interest…’ & Article 9(2)(j) ‘processing is necessary for archiving purposes in the public interest…’. LSHTM have considered the evidence that several epidemiological studies report a correlation between low circulating level of vitamin D and the risk of developing various cancers accompanied by evidence that 80% of men and 87% of women aged over 65 years living independently in the UK have low levels of vitamin D in line with recent studies. LSHTM propose that the requirement for processing NHS Digital data to aid in the completion of this study and the overall project is in line with the public interest.

Data from all 1615 patients randomised to the trial will be included for the statistical analysis. The following steps have been completed in order to trace participant data from NHS Digital.

• LSHTM provided each participant's unique Study ID, NHS number, date of birth and post code to NHS Digital. Data containing personal identifiers will be transferred to and from NHS Digital using the secure data transfer portal.

• The participant's identifiers (Study ID, NHS number, date of birth and postcode) were linked to medical records held by NHS Digital on cancer registrations (provided by NHS Digital on behalf of Public Health England), deaths (sourced from civil registration data and provided by NHS Digital on behalf of the Office for National Statistics) and the Hospital Episodes Statistics database on hospital admissions.

• This linked data flow to LSHTM. This data will be stored on the secure server at LSHTM which can be accessed only by the LSHTM study team using a unique network password. No-one else outside of the LSHTM study team have access to any of the NHS Digital data from this application.

Participant medical data has been linked to study data, including blood samples depicting vitamin D levels, obtained from participants over the course of their two visits with participating practises. The study team have conducted analysis of NHS Digital data linked to Study data to assess level of vitamin D against overall health. The main analyses will be the matched pair comparison of recruitment and overall compliance rates between placebo control and open control designs. Details of additional vitamin D supplement use will be recorded at entry, and retrospectively at the 2-year final visit. Failure to report taking at least 19 doses of allocated IMP or to attend the 2-year final visit are included in the definition of noncompliance, so there will be no missing data for the main analyses. Change in blood level of 25(OH) vitamin D from entry to exit will be analysed in relation to self-reported compliance. The effect of treatment on serum 25(OH)D both overall by allocated treatment, and in compliant participants, will also be analysed in relation to the pre-treatment 25(OH)D level. Other principal outcome data from this trial will include deaths, cancer cases and hospital admissions from major causes during and after the trial. This is the reason for requesting the mortality, cancer registration and HES data from NHS Digital. The data from a year preceding the beginning of the trial are required to provide baseline rates for deaths, cancer cases and hospital admissions. Deaths, cancer registration rates and hospital admission rates for major disease categories will be estimated by treatment allocation group (vitamin D versus no treatment) and baseline serum vitamin D level (blood samples were taken at entry to the trial).

The data will only be viewed, accessed and analysed by substantive employees of the London School of Hygiene & Tropical Medicine (LSHTM). There are two collaborators in this study – University College London and Queen Mary University of London – but they do not have access to the data.

A sole study-specific paper has been produced as a result of this feasibility study:

Cluster randomised comparison of open versus double-blind individual randomisation in people aged 65–84 years: the VIDAL (Vitamin D And Longevity) feasibility RCT - Rake C, Gilham C, Bukasa L, Ostler R, Newton M, Wild JP, et al.

This paper has recently been finalised and accepted for publication by the peer reviewed NIHR Health Technology Assessment journal.

The results of this feasibility study will be reported in peer reviewed scientific journals with accompanying conference presentations and submissions to regulatory authorities. The institution will also target the scientific medical community by submitting a report to the peer-reviewed British Medical Journal. Due to the unavailability of cancer registration data from April 2017, LSHTM have not received cancer data over the course of the original version of the agreement. As a result, the applicant have confirmed they have been unable to produce particular outputs including submitting these reports.

In addition, further funding will be sought for a larger “main” trial (for which this is a feasibility study), involving up to 20,000 participants across 200+ GP practices in the United Kingdom. The main trial is subject to funding from the National Institute for Health Research (NIHR). NIHR will assess the results of this feasibility in terms of whether substantial results are collected and replicated in the larger “main” trial.

NIHR have indicated that they would also like to see the results of several similar large-scale international vitamin D trials, due in the next few years, before they will consider funding a similar large scale study in the UK. These include the D-Health trial of 25,000 subjects in Australia and the TIPS-3 trial of 5713 participants across 9 countries, including Canada and India.

There is a growing consensus that most people in the UK do not receive enough sun exposure to achieve a satisfactory vitamin D blood level, and that current recommendations on the optimal daily dose of vitamin D should be increased. Eighty percent of men and 87% of women aged over 65 years living independently in the UK have blood levels of 25hydroxyvitamin D (25(OH)D the measure of vitamin D status) below the recommended concentration for optimal health (75 nmol/l or 30 ng/ml). Epidemiological evidence suggests that people with low blood levels may be at increased risk for several diseases, including heart disease, various infections and some cancers. However, similar epidemiological evidence of overall benefit for various other vitamins was disproved when randomized trials were conducted, and no adequately powered trial has tested vitamin D in doses that are high enough to achieve blood 25 (OH)D concentration > 75 nmol/l. Although the majority of observational studies report associations between vitamin D deficiency and susceptibility to a range of pathologies, some studies are null, and a few report opposite associations. A large randomized trial is therefore required to assess whether vitamin D supplementation can decrease the risk of various diseases and increase longevity. The importance of a clinical trial of oral supplementation of vitamin D is emphasized by the finding that in practice it is difficult for the majority of UK residents to obtain optimal vitamin D from sunlight or diet. An effective measure of a positive impact for the main trial would be a change in opinion regarding the current recommendations for vitamin D supplement. Amended recommendations would ideally outline that adults and children over the age of one should consider taking a daily supplement containing 10mcg of vitamin D, particularly during autumn and winter, while people who have a higher risk of vitamin D deficiency are being advised to take a supplement all year round. Higher risk groups include people whose skin has little or no exposure to the sun, like those in care homes, or people who cover their skin when they are outside, in addition to people with dark skin, from African, African-Caribbean and South Asian backgrounds, who may also not get enough vitamin D from sunlight in the summer.

If self-administered vitamin D supplementation were shown to confer substantial health benefits it would be routinely recommended and widely adopted. This would also provide a rationale for a national policy of vitamin D supplementation for the general population, a review of the relative risks and benefits of sun exposure, and a revision of existing policy on vitamin D fortification of foods. If no benefit or unforeseen disadvantages are shown this will also be a valuable contribution to knowledge. The target audience for such findings would be the Department of Health and National Institute for Health Research.

An important public health priority is therefore to demonstrate the feasibility of a large randomized trial of prolonged vitamin D supplementation in older people, and to show that this will increase serum 25(OH)D to >=75 nmol/l in the majority of subjects. The main trial for which this is the feasibility study will be a randomized trial with 20,000 participants followed for 10 years. A trial of that size is needed to detect the 7% reduction that vitamin D supplementation might plausibly achieve in total mortality in healthy adults aged over 65. As well as demonstrating an expected increase in circulating 25hydroxyvitamin D levels, the feasibility study will provide estimates of cost and establish the study design and procedures required for the main trial. This feasibility study will be used to trial & inform the methodology of the main study, including study design (“open” versus “placebo control” acceptability, costs, patient compliance and contamination) as well as overall costs, efficiency, data management, contacting and setting up centres, refining questionnaires and databases, patient acceptability, GP practice availability etc. This will highlight the most effective method of carrying out this study and replicate conditions applied during the study practises.

Due to the unavailability of cancer registration data from April 2017, LSHTM have not received cancer data over the course of the original version of the agreement. As a result, LSHTM confirmed they have been restricted in producing outputs outlined above.

LSHTM projects that supplementing the omitted cancer data for use within the trial will enable the expected outputs to be produced and the appropriate reports be directed back to stakeholders assessing the feasibility of the main trial. Once this process is complete, LSHTM will be in a position to identify benefits realised as a result of the processing of NHS Digital data and the wider project.