Controversy in the link between vitamin D supplementation and hypertension

Abstract

Vitamin D affects a range of pathophysiological processes pertinent to the control of blood pressure, including endothelial function, inflammation and renin–angiotensin system activity. Observational data show a clear relationship between 25-hydroxyvitamin D levels and both current blood pressure and incident hypertension. However, recent trial data have shown no significant effect of vitamin D supplementation on blood pressure, even at high doses, low vitamin D levels and in patients with high baseline blood pressure. Vitamin D might still benefit cardiovascular health through mechanisms other than blood pressure reduction, but data from large trials are required to show this. In the meantime, vitamin D has no place in controlling blood pressure either at the individual or the population level.

Keywords:

• cardiovascular
• hypertension
• meta-analysis
• trials
• vitamin D

There are few interventions in healthcare touted as a cure for all ills in the way that vitamin D has been over the past 20 years. Cancer, metabolic diseases, autoimmune disease, infection and cardiovascular disease have all been linked to low circulating 25-hydroxyvitamin D (25OHD) levels (the definition of which continues to be a focus of controversy), and the more enthusiastic proponents of vitamin D therapy continue to push for universal supplementation. Despite the enthusiasm, trial data have been less than encouraging in many disease areas. The relationship between blood pressure and vitamin D reflects this wider controversy, with initially encouraging observational data, but less positive intervention results. What progress has been made in this area recently and is the evidence now any clearer?

Biological plausibility

There are a number of mechanisms through which vitamin D may act on blood pressure and influence vascular health. Vitamin D receptors have been identified on many tissues, including vascular endothelium, vascular smooth muscle and cardiac muscle cells [1]. Animal and human studies have demonstrated that vitamin D downregulates the renin–angiotensin–aldosterone system (RAAS) [2,3], a key player in the regulation of vascular tone. Vitamin D has also been shown to have a beneficial effect on endothelial function [4,5] with a number of mechanisms including cytokine modulation [6] and effects on VEGF production [7]. Vitamin D analogs may have beneficial effects on renal artery endothelial function via reduction in oxidative stress and downregulation of AT1 receptors [8]. Furthermore, vitamin D suppresses parathyroid hormone (PTH) levels, and elevated PTH itself has vasculotoxic effects, including driving left ventricular hypertrophy [9]. Hence, it is certainly biologically plausible that vitamin D could influence blood pressure and vascular health.

Observational data

Observational data have demonstrated relationships between vitamin D levels and a number of cardiovascular risk factors including blood pressure. Cross-sectional studies have evidenced an inverse association between serum 25OHD levels and blood pressure [10]. Longitudinal observational studies examining vitamin D and cardiovascular outcomes have shown lower 25OHD levels to be associated with a high risk of incident hypertension. Meta-analysis of three cohort studies in subjects with self-reported hypertension revealed a 1.8-fold higher risk of new hypertension diagnosis in subjects with the lowest vitamin D levels [11], and similar results were seen in a more recent meta-analysis of eight cohorts comprising over one-quarter of a million participants [12]. In contrast, the large Women’s Health Initiative trial [13] did not demonstrate an association between baseline 25OHD levels and blood pressure over their 7-year follow-up period. Observational data are inherently limited by both confounding factors and the complex interplay between a number of cardiovascular risk factors, include age, obesity, smoking, low levels of physical activity, genetics and inflammation. To try and circumvent some of these issues, Mendelian randomization studies have been attempted [14], relating allelic variations in genes that influence 25OHD levels with blood pressure levels and incident hypertension. This analysis demonstrated that for every 10% increase in predicted 25OHD levels, blood pressure was 0.4/0.3 mmHg lower, and that the risk of incident hypertension was 8.1% lower.

Intervention data

Early meta-analyses examining the effect of vitamin D supplementation on blood pressure suggested a possible benefit – one meta-analysis found a marginally significant 4/3 mmHg reduction in blood pressure across eight trials, with benefit confined to those trials with a mean baseline systolic blood pressure above 140 mmHg [15]. Another meta-analysis conducted around the same time did not find any benefit to supplementation [16]. At that time, few studies had been performed with blood pressure as their primary outcome measure; studies were mostly small, of short duration, and were criticized for using doses of vitamin D that were too low to produce sufficiently large changes in 25OHD levels.

The past 5 years have seen publication of a large number of trials examining the effect of vitamin D and related analogs on measures of blood pressure and vascular health, such as arterial stiffness, endothelial function and left ventricular mass. A recent meta-analysis [17] combined data from 46 randomized controlled trials (n = 4541) of vitamin D supplementation and found no statistically significant treatment effect on either systolic (0.0 mmHg; 95% CI -0.8–0.8) or diastolic blood pressure (-0.1 mmHg; 95% CI -0.6–0.5) compared with placebo intervention. The use of meta-regression and analysis of individual patient data from these trials failed to identify any subgroup that benefited from supplementation – those with high baseline blood pressure, low baseline 25OHD levels, high baseline PTH levels or those not taking RAAS system blockers still failed to benefit. Even high doses of vitamin D (above 2000 units of vitamin D3 per day) failed to show any effect.

Where next?

It seems clear that vitamin D supplementation is not going to provide dramatic blood pressure reductions for individual patients, and the current evidence does not suggest that population-level supplementation would reduce blood pressure. Simply continuing to increase the dose in trials is unlikely to change these conclusions, and although the largest trials to date have enrolled only a few hundred patients, the effect size that a large hypertension trial would need to change the current meta-analysis result would be greatly at variance with all results to date. Some caveats are worth spelling out, however. First, few patients of African or African-American descent have been enrolled into vitamin D trials. It is possible that this group might derive greater benefit, and one recent trial does suggest a potential benefit in this group [18]. Second, most trials are of relatively short duration (12 months or less), and although no relationship between trial length and degree of blood pressure reduction was noted in the recent meta-analysis, it is still possible that several years of supplementation might produce beneficial effects, for example, by prevention of vascular remodeling or vascular calcification. Finally, a lack of effect on blood pressure does not necessarily obviate an effect on cardiovascular events via other mechanisms. Several very large vitamin D supplementation trials [19] (ranging from 2000 to 20,000 patients) are currently recruiting or are in follow-up, most are recording cardiovascular events as a part of their suite of outcomes, and data from these trials will provide an answer as to whether vitamin D supplementation can reduce vascular events, notwithstanding the lack of effect seen on blood pressure.

The relationship between low 25OHD levels, risk factors and disease states is complex. Low 25OHD levels may be a consequence of other cardiovascular risk factors such as obesity, smoking, aging and low physical activity, leading to less sunlight exposure and less efficient production in the skin. The relationship may be further confounded by the fact that the inflammatory response, found in many disease states and itself implicated in hypertension and cardiovascular disease, also appears to cause a fall in 25OHD levels [20]. Mendelian randomization studies provide a partial solution, as well as potentially showing the effect of integrated 25OHD levels over a lifetime, but they bring their own problems – not all alleles affecting 25OHD levels have a consistent effect [14], and there is doubt as to whether some genes affecting 25OHD levels do so without effects on non-vitamin D systems that might affect blood pressure. Thus, even the best observational data struggle to disentangle cause from effect, and it is critical that we maintain the proper degree of skepticism surrounding the disconnect between observational data and trial results. Blaming inadequate trials is unhelpful if there is actually no causal association between low 25OHD levels and high blood pressure. As ever in medicine, a cautious approach to treatment appears wise – current evidence does not support the use of vitamin D supplementation for treatment of hypertension at a population or individual level, and we need to wait for the results of ongoing large trials before deciding whether vitamin D supplementation has a place in the wider armamentarium of therapy for cardiovascular disease.

Financial & competing interests disclosure

MD Witham has received grant funding for research on vitamin D from the Chief Scientist Office (Scottish Government), Diabetes UK, Heart Research UK, ME Research UK and Tenovus Scotland. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.