Introduction

Cardiovascular disease is a prominent cause of mortality worldwide. It is particularly common in patients with chronic kidney disease (CKD), where it occurs at rates 10 to 20 times greater than the general population. CKD is also accompanied by both an increased risk of vitamin D insufficiency and a reduction in the activity of 1α-hydroxylase, the enzyme that coverts 25-hydroxyvitamin D (25[OH]D), the predominant circulating form of vitamin D, to 1,25-dihydroxyvitamin D. The latter is the active, hormonal form of vitamin D that acts on target tissues through the intracellular vitamin D receptor (VDR). The VDR binds to vitamin D response elements (VDREs) on DNA, where it can influence gene regulation and influence a host of biologic activities. Historically, the primary role of vitamin D has been believed to be the regulation of calcium homeostasis, particularly the intestinal transport of calcium. Lack of 1,25-dihydroxyvitamin D in CKD leads to hypocalcemia and, consequently, secondary hyperparathyroidism with its associated bone disease. Supplemental forms of activated vitamin D, initially calcitriol and later synthetic vitamin D receptor agonists (VDRAs), have been used to ameliorate this process in CKD.

Early retrospective cohort studies in patients on dialysis studying the effects of these analogs found survival differences depending on the specific analog used [1]; a follow-up study showed that use of any activated vitamin D was associated with a 26% mortality reduction relative to those who received no such treatment, with death from cardiovascular disease reduced by nearly 50% [2]. A number of studies have since replicated these findings. Several epidemiologic studies in the general population have similarly linked greater vitamin D levels and use to reduced risk of hypertension and left ventricular hypertrophy (LVH), although no prospective clinical trial has yet demonstrated a causative relationship. Nonetheless, these early studies, combined with increasing animal data suggesting potential mechanisms, have raised the question of a role beyond mineral metabolism for vitamin D.

Increasing evidence points to a direct role for vitamin D on the cardiovascular system, either through direct action on the heart, or indirectly through hormonal influence or effects on the vasculature. This article reviews the current state of knowledge about vitamin D action with respect to hypertension and LVH, including both potential biologic mechanisms and human studies.

Animal Studies and Biologic Mechanisms

Early animal studies provided clues for understanding links between vitamin D and the cardiovascular system. Weishaar et al. [3] demonstrated that rats fed a diet deficient in vitamin D and raised with limited exposure to ultraviolet light (to prevent endogenous vitamin D production) developed cardiomegaly. This cardiomegaly was not reversed by normalization of calcium levels through dietary supplementation, suggesting that changes in mineral metabolism did not drive the changes in cardiac morphology. The same group also demonstrated increases in systolic blood pressure and enhanced cardiac contractility with the same regimen [4]. Supplementation with vitamin D has also been shown to abrogate endothelin-induced hypertrophy and atrial natriuretic peptide (ANP) production in cultured rat myocytes [5].

One likely mechanism by which vitamin D could influence the cardiovascular system is via the renin-angiotensin-aldosterone axis. Li et al. [6] examined mice lacking the VDR and found they developed a nearly fourfold increase in renin expression and angiotensin II production compared to wild-type mice. Systolic blood pressure, diastolic blood pressure, and cardiac mass were all increased in VDR-null animals. Treatment with the angiotensin-converting enzyme (ACE) inhibitor, captopril, eliminated the blood pressure difference in the groups whereas treatment of wild-type mice with 1,25-dihydroxyvitamin D suppressed renin production.

A follow-up study in VDR knockout mice further characterized this phenotype, demonstrating increased atrial natriuretic peptide (ANP) and hypertrophy of left ventricular cardiomyocytes [7]. As with prior studies, these changes were prevented by use of captopril. Cardiac renin levels were also increased; some believe these levels may be more relevant to these physiologic changes than circulating renin levels.

In total, these results strongly suggested that, at least in animal models, vitamin D is an important regulator of the renin-angiotensin-aldosterone axis. Zhou et al. [8] confirmed these findings in the 1α-hydroxylase knockout mouse. These mice lack the enzyme responsible for converting the inactive 25(OH) D to the active 1,25-dihydroxyvitamin D [8]. Within 4 weeks, animals developed systolic hypertension, cardiac hypertrophy (with thickening of the intraventricular septum as well as individual cardiomyocytes), and impaired cardiac function (with reduced fractional shortening and ejection fraction). The renin-angiotensin system was upregulated in both the kidney and in cardiac tissues. Treatment with 1,25-dihydroxyvitamin D, which bypassed the missing enzyme, restored normal cardiac physiology. Abnormalities in calcium and phosphorous could be corrected with either dietary supplementation or 1,25-dihydroxyvitamin D, but the former did not correct the abnormalities in cardiac structure and function, confirming that the cardiovascular effects of vitamin D are independent of its effects on mineral metabolism. As with prior studies, treatment with inhibitors of the renin-angiotensin system (including captopril and losartan) prevented the observed changes with the exception of increased renin production, presumably because these agents acted downstream of renin in this pathway.

These results suggest that vitamin D or its analogs might be used therapeutically in high-risk individuals for the prevention or amelioration of cardiovascular disease. One potential model in which to study these effects is the Dahl salt-sensitive rat model. These animals, when supplied with a high-salt diet, develop LVH and congestive heart failure. They also become profoundly vitamin D deficient, a phenomenon thought to occur because of urinary losses. Bodyak et al. [9] used this model to assess the effects of paricalcitol, an active vitamin D analog. Paricalcitol-treated animals demonstrated improved cardiac contractility by echocardiography, suggesting a direct benefit on cardiac function. Cardiac catheterization demonstrated lower end-diastolic pressures when compared with untreated animals, and heart size was reduced in animals receiving paricalcitol. Paricalcitol treatment was also associated with reduced cardiac mRNA expression for atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP), and renin. Interestingly, systemic blood pressures were not significantly affected by paricalcitol therapy, suggesting that the observed cardiac changes were due to direct effects on the myocardium. Indeed, ventricular gene expression profiles revealed over 1,000 genes that were affected by paricalcitol exposure. This study included a retrospective analysis of human subjects that found improved diastolic function and septal and posterior wall thicknesses in individuals on hemodialysis who received paricalcitol (compared with their untreated counterparts), suggesting the observed effects on cardiac physiology may translate to humans.

Summary

Several animal studies suggest that vitamin D has the potential to suppress renin production and, thus, interrupt a pathway central to hypertension and LVH. There is increasing evidence that vitamin D is associated not only with endocrine changes, but also changes in cardiac structure and function. However, the majority of these studies involve extreme phenotypes (eg, VDR knockouts or profoundly deficient animals), and may not be directly applicable to physiologic effects in humans.

Epidemiologic Studies

A number of cross-sectional studies have evaluated the association between vitamin D levels and blood pressure (Table 1). Woodhouse et al. [10] investigated the seasonal variations in blood pressure and its relationship to ambient temperature and observed that the incidence of hypertension increased with higher latitude, and higher recordings are noted in winter than in summer months. Data from the INTERSALT study has suggested that a rise in blood pressure is proportional to distance from the equator and speculated that the ultraviolet light may contribute to the geographic variation in blood pressure differences [11]. As early as the 1980s, studies reported higher blood pressure in patients with low 25(OH) D levels [12, 13].

Table 1 Summary of recent epidemiologic studies evaluating associations between vitamin D, hypertension, and left ventricular hypertrophy
Full size table

More recently, Martins et al. [14] performed a secondary analysis of data from the Third National Health and Nutrition Examination Survey. This analysis included 7,186 male and 7,902 female adult participants from the United States. The mean 25(OH) D level in the overall population was 30 ng/mL. Investigators observed lower 25(OH) D levels in women, elderly patients (≥60 years), and in patients with hypertension, obesity, and diabetes mellitus. The adjusted prevalence of hypertension was significantly higher in the first than in the fourth quartile of serum 25(OH) D levels (odds ratio 1.30; 95% CI, 1.13–1.49]; P = 0.001). Similar inverse relationships between serum 25(OH)D levels and blood pressure have been confirmed in other analyses of The Third National Health and Nutrition Examination Survey [15, 16], in large population cohorts from other countries [17, 18], and have also been demonstrated when the blood pressure was measured by 24-hour measurements [19•]. However, these observations regarding 25(OH) D levels and blood pressure have not been consistent in cross-sectional studies [20–22]. Reis et al. [20] reported findings on participants from the Rancho Bernardo Study (age 44–96 years, 410 men and 660 women). They observed no association of 25(OH) D with components of metabolic syndrome, including hypertension. The authors acknowledged that the non-significant findings in this cohort may have been due to reduced variability in 25(OH) D levels in this population due to their residence in a southern California community with a sunny and temperate year-round climate. Rueda et al. [21] performed a cross-sectional analysis of 298 severely obese patients, and although in unadjusted analyses patients in the highest quartile of 25(OH)D were less likely to present high blood pressure (odds ratio 0.35; 95% CI, 0.16–0.77), this association lost its significance after adjustment for age, season, and body mass index. Among 1,205 elderly participants from the Longitudinal Aging Study Amsterdam, Snijder et al. [22] analyzed associations between serum 25(OH) D and blood pressure. In this analysis, serum 25(OH) D levels were not significantly associated with blood pressure. The low prevalence of vitamin D deficiency (approximately 10%) may have contributed to this non-significant association.

Cross-sectional studies (reporting significant as well as non-significant associations) lack establishment of temporal association because the outcome and exposure are measured simultaneously. Prospective cohort studies eliminate this bias. Association between vitamin D and blood pressure has been examined in a number of cohorts (Table 1) [23••, 24–27]. As recently reported by Wei and Giovannucci [23••], these cohorts include the Nurses’ Health Study I (n = 121,700 female nurses aged 37–64 years at baseline in 1984), Nurses’ Health Study II (n = 116,671 female nurses aged 27–44 years at baseline in 1991), and Health Professionals Follow-up Study (n = 51,529 male health professionals aged 40–75 years at baseline in 1986). In the Health Professionals Follow-up Study and Nurses’ Health Study, prospective analyses of 25(OH) D and vitamin D intake and incident hypertension were conducted by Forman et al. [24]. During 4 years of follow-up, in a multivariate analysis those with 25(OH) D levels <15 ng/mL had a relative risk for incident hypertension of 2.67 (95% CI, 1.05–6.79) compared with those whose levels were >30 ng/mL [24]. A similar inverse association between 25(OH) D and incident hypertension of similar magnitude was reported in a nested case–control study in the Nurses’ Health Study II [25]. Women in the lowest quartile of 25(OH) D had an increased odds of incident hypertension (odds ratio 2.21; 95% CI, 1.57–3.12; P for trend = 0.001). In a multivariate adjustment, this relationship was attenuated but remained significant (odds ratio 1.66; 95% CI, 1.11–2.48; P for trend = 0.01). However, when associations between the intake of vitamin D and the risk of incident hypertension were studied in Nurses’ Health Study I, Nurses’ Health Study II, and in the Health Professionals’ Follow-up Study, higher intake of vitamin D was not associated with a lower risk of incident hypertension in a multivariate analysis [26]. Although it is possible that the participants in these three cohorts had vitamin D intake that was likely below the necessary intake to reduce the incidence of hypertension, when the investigators compared participants who consumed ≥1,600 to < 400 IU per day and those who consumed ≥1,000 to < 200 IU per day, no significant associations were found.

In the recently published Tromso study, investigators confirmed the cross-sectional association between serum 25(OH)D and blood pressure, but failed to show serum 25(OH)D as a significant predictor of incident hypertension [28••]. Also, there was no association between change in serum 25(OH) D and change in blood pressure. Contrary to this, in a population-based longitudinal Michigan Bone Health and Metabolism Study, a single-time baseline measure of serum 25(OH)D levels among adult women was associated with a significantly greater incidence of systolic hypertension a decade later [29•].

A recent meta-analysis summarized data on 78,028 participants from 14 cross-sectional and four prospective studies. This analysis concluded that 25(OH) D levels are inversely associated with hypertension [30•]. In this analysis, the pooled odds ratio of hypertension was 0.73 (95% CI, 0.63–0.84) for the highest versus the lowest category of blood 25(OH) D concentration. The authors also investigated a dose–response relationship between 25(OH) D levels and hypertension and reported the odds ratio for a 16 ng/mL increment in blood 25(OH) D concentration of 0.84 (95% CI, 0.78–0.90). The majority of studies in this meta-analysis reported estimates adjusted for season, physical activity, age, and body weight [30•].

Observational studies have also reported an association between vitamin D and LVH. Patients with congestive heart failure frequently have mean serum 25(OH) D concentrations in the insufficiency or even deficiency range [31–33]. Pilz et al. [34] measured 25(OH) D levels in 3,299 white patients who were referred for coronary angiography. They reported negative correlation between 25(OH) D levels and impaired left ventricular function (P for trend < 0.001), and this association remained significant after multivariable adjustments. In this cross-sectional study, the adjusted hazard ratios for death due to heart failure and for sudden cardiac deaths were 2.84 (95% CI, 1.20–6.74) and 5.05 (95% CI, 2.13–11.97), respectively, when comparing patients with severe vitamin D deficiency with persons in the optimal range.

Although the majority of cross-sectional and prospective studies confirm an inverse association between the serum 25(OH) D level and blood pressure, a causal relationship can not be established due to inability to adjust for possible confounding. Randomized controlled trials are needed to establish the causal association.

Summary

Many, but not all, cross-sectional and prospective studies suggest an inverse association between serum 25(OH) D level and blood pressure; this relationship may only be evident in the setting of severe vitamin D insufficiency. Fewer studies have examined the relationship with LVH and left ventricular function, although some data suggest an association. Although these results are supportive of biologic mechanisms suggested by animal studies, the potential for confounding, particularly with respect to advanced heart failure, is sizable.

Randomized Controlled Trials

In one of the very first randomized controlled trials on this topic, Lind et al. [35] evaluated the effects of alphacalcidol (1 μg) in 29 patients with mild intermittent hypercalcemia. Investigators observed an inverse relationship between serum calcium levels and diastolic blood pressure prior to intervention. Treatment with alphacalcidol raised the serum calcium by 0.28 mg/dL during a 6-month trial period and caused a significant reduction of diastolic blood pressure by 9.2 mm Hg compared with placebo (P < 0.01). The same investigators then demonstrated both systolic and diastolic blood pressure lowering by alphacalcidol in older male patients (61–65 years of age) with impaired glucose tolerance in a double-blind, placebo-controlled study over 12 weeks [36]. Since then, a number of randomized controlled trials have been performed on this topic (Table 2) [37, 38, 39•, 40–46, 47•, 48, 49]; however as reported in two recent meta-analyses, the majority of these studies have been small and of variable methodologic quality [50, 51••]. In meta-analysis of 10 randomized trials, Pittas et al. [51••] reported that supplementation with vitamin D non-significantly reduced systolic blood pressure (weighted mean difference, -1.9 mm Hg [95% CI, -4.2 to 0.4 mm Hg]) and did not affect diastolic blood pressure (weighted mean difference, -0.1 mm Hg [95% CI, -0.7 to 0.5 mm Hg]). Similar non-significant reduction in systolic blood pressure was observed in a meta-analysis by Witham et al. [50]; although, this meta-analysis reported a significant reduction in diastolic blood pressure (−3.1 mmHg [95% CI, -5.5 to −0.6]). The dose of vitamin D preparations varied considerably among these trials (from 400 to 8,500 units/day).

Table 2 Summary of recent randomized controlled trials addressing effects of vitamin D supplementation on hypertension and left ventricular hypertrophy
Full size table

The longest and largest randomized trial to investigate this has been the Women’s Health Initiative [45]. In this trial, combined low-dose vitamin D3 (400 IU/d) and calcium carbonate supplementation (1,000 mg/d) had no effect on self-reported incident hypertension after 7 years of follow-up. Also, there was no difference in blood pressure measurements at 7 years between the intervention and control arms. However, this study was not specifically designed to investigate effects of vitamin D on blood pressure (it was originally designed to examine whether vitamin D and calcium supplementation would reduce fracture and cancer risks), and it is possible that the dose of vitamin D administered was sub-optimal to have any effects on blood pressure. In the recently published Vitamin D Receptor Activator for Albumin Lowering (VITAL), investigators aimed to assess whether selective activation of vitamin D by paricalcitol could be used to further reduce albuminuria in patients with diabetic nephropathy who are receiving either angiotensin-converting enzyme inhibitors or angiotensin receptor blockers [52]. In this multinational, placebo-controlled, double-blind trial, patients treated with 2 μg of paricalcitol demonstrated a sustained reduction in albuminuria. In addition, administration of 2 μg of paricalcitol was associated with lowering of systolic blood pressure (range −3 to −9 mm Hg; P = 0.033 vs placebo) during the treatment phase, and blood pressure values returned to baseline at 30 and 60 days after the completion of treatment. However, this study was also not specifically designed to examine the effects of paricalcitol on blood pressure.

As noted in previous paragraphs, although there are well-established biological mechanisms that indicate a role of vitamin D in control of blood pressure, there remains a discrepancy between the observational data and the currently available randomized controlled data regarding vitamin D and hypertension. Confounding in observational cohorts, as well as limitations of current randomized controlled trials in terms of their sample size, poor methodologic quality, and administration of possible sub-optimal intervention dose, are possible explanations for this discrepancy and emphasize the need for adequately powered, high-quality, randomized controlled trials designed specifically to investigate effects of vitamin D supplementation on blood pressure.

Although a few randomized controlled trials report improvement in cardiovascular events associated with vitamin D administration, the data have not been consistent [53–55]. Furthermore, at present there are no data available from randomized controlled trials specifically designed to assess the impact of vitamin D supplementation on LVH. Previous clinical trials on this topic have been non-controlled, have had small sample size, and have been conducted predominantly in patients with end-stage renal disease [56, 57]. Park et al. [56] performed echocardiography assessments over a 15-week period in 15 hemodialysis patients with secondary hyperparathyroidism before and after calcitriol treatment and 10 hemodialysis control patients with secondary hyperparathyroidism not receiving calcitriol therapy. Their study found a significant decrease in the interventricular wall thickness (from 13.9 ± 3.6 to 12.8 ± 3.1 mm; P = 0.01), left ventricular posterior wall thickness (from 12.5 ± 2.4 to 11.3 ± 1.8 mm; P < 0.05), and left ventricular mass index (from 178 ± 73 to 155 ± 61 g/m2; P < 0.01) before and after treatment. These changes were not seen in the control patients. Similar reductions in left ventricular mass index have been reported in another non-controlled trial by Kim et al. [57].

Baseline characteristics of study subjects from Paricalcitol Capsules Benefits in Renal Failure - Induced Cardiac Morbidity (PRIMO), a multinational, randomized, double-blinded trial of oral paricalcitol in subjects with chronic kidney disease (stages 3–4), mild-to-moderate LVH, and a left ventricular ejection fraction <50% have recently been reported and results from this trial are awaited at this time [58••]. The primary endpoint in this trial is change in the left ventricular mass index and the secondary endpoint is improvement in diastolic function after 48 weeks of treatment.

Summary

Controlled trials of vitamin D in hypertension have yielded mixed results. Early studies with active vitamin D analogs suggested a blood pressure–lowering effect, although many subsequent studies have either failed to find a significant effect of treatment using nutritional forms of vitamin D or have found modest effects. Many studies have been hampered by limited study quality or lack of blood pressure as a primary outcome measure. Data on LVH are encouraging but limited; definitive trials in high-risk populations are still ongoing.

Conclusions

Animal studies have demonstrated potential to suppress rennin production by vitamin D and interrupt pathways involved in the pathogenesis of HTN and LVH. Observational studies in humans support these animal studies however are limited by multiple possible confounders. Data from randomized controlled trials on HTN are inconsistent and those on LVH are limited. Further randomized controlled trials are needed to assess the effects of vitamin D preparations (both natural and activated) on these outcomes.