Vitamin D, Calcium, and Atherosclerotic Risk: Evidence from serum levels and supplementation studies

Abstract

Vitamin D and calcium have been traditionally been viewed in relation to bone health. However, recent research has suggested relations between these nutrients and cardiovascular disease (CVD). Specifically, evidence from both observational studies and clinical trials suggests that vitamin D may be related to lower risk of CVD. The picture for calcium is more complex. Dietary intake of calcium may be associated with lower CVD risk, while calcium supplementation may elevate CVD risk. In this review we summarize evidence of these relations, and comment on the recent Institute of Medicine (IOM) recommendations regarding use of vitamin D and calcium supplements.

Introduction

Atherosclerotic cardiovascular disease (CVD) is the number one cause of morbidity and mortality in developed countries. Therefore, there remains a great interest in identifying novel risk factors to identify those at risk for CVD, particularly markers that are modifiable/treatable.

Vitamin D is a fat-soluble vitamin obtained through cutaneous synthesis resulting from sun exposure, and through oral intake from food and supplement sources. Insufficient vitamin D, as assessed by low circulating 25-hydroxyvitamin D [25(OH)D], has recently drawn attention as a potential CVD risk factor. This possibility is of immense public health and clinical interest, given the vast potential for intervention as globally it is estimated that more than 50% of the world population has inadequate vitamin D levels¹. Although intriguing, much remains unknown about the relation of vitamin D to CVD events.

Calcium is vital to the functionality of all cells, though more than 99% of the calcium in the body is used as a structural component of bones and teeth². The biology of calcium and vitamin D is intricately intertwined. As such, these nutrients are often prescribed in combination for the promotion of skeletal health, particularly among postmenopausal women. Relations of calcium supplementation to reduced fracture risk are, however, perhaps less pronounced than commonly assumed^{3, 4}. As such, it is important to be cognizant of other ways in which calcium supplementation may impact health. In particular, recent studies have highlighted concerns regarding increased cardiovascular risk associated with calcium supplementation.

Vitamin D Metabolism

The major source of vitamin D is endogenous production via the action of the sun’s ultraviolet b (UVB) light on 7-dehydrocholesterol precursors in the skin, converting them to vitamin D3 (cholecalciferol)^{2, 5}. Exogenous D3 can also be obtained from the diet from oily fish, fortified foods, or supplements. Vitamin D can also be found in the form of D2 (ergocalciferol) which is produced in in small amounts by plants after UVB radiation of ergosterol, and can also be consumed in the diet as a supplement or from fortified food. Fortified foods provide the majority of the vitamin D in the American diet⁶; however most people meet their vitamin D needs predominantly through sun exposure⁶.

Vitamin D undergoes 25-hydroxylation in the liver, to form 25(OH)D, the primary circulating form of vitamin D and the metabolite that best reflects stores of vitamin D (i.e. sufficiency or deficiency states). The active metabolite, 1,25(OH)₂D, (also called calcitriol), is formed after 1-hydroxylation in the kidneys, and both 1,25(OH)₂D2 and 1,25(OH)₂D3 bind to the vitamin D receptor. Although the main source of 1,25(OH)₂D in the circulation is from renal 1-hydroxylation, it has been recently appreciated that many other tissues, including breast, colon and vascular smooth muscle cells also express 1-hydroxylase. In these tissues, the locally formed 1,25(OH)₂D serves as a cell-differentiating factor. Extra-renal production of 1,25(OH)₂D is thought to have a paracrine/autocrine function including an important role in innate (macrophage) immunity. Levels of activated 1,25(OH)₂D levels do not correlate well with 25(OH)D levels, and given tight regulation by parathyroid hormone (PTH), 1,25(OH)₂D levels may remain in the normal range even in the face of 25(OH)D deficiency^5–7.

Vitamin D and CVD Risk Factors

Suboptimal vitamin D is thought to influence CVD risk predominantly by acting on established CVD risk factors, namely hypertension, diabetes, and inflammation.

Vitamin D is involved in regulation of the renin-angiotensin system, and low 25(OH)D levels are associated with hypertension and hypertrophy of the left ventricle and vascular smooth muscle cells^8–10. The majority of observational prospective cohort studies^{11, 12}, though not all¹³, have reported inverse associations between 25(OH)D and risk of incident hypertension. Experimentally, several small clinical trials have found that supplementing vitamin D (either by dietary supplements or with UVB radiation) successfully reduced blood pressure^14–18. However, a recent small clinical trial enrolling patients with chronic kidney disease and left ventricular hypertrophy did not find the activated vitamin D analog paricalcitol compared to placebo reduced left ventricular mass or measures of diastolic dysfunction after 48 weeks of treatment¹⁹.

Suboptimal vitamin D is also believed to elevate diabetes risk. The mechanism of action is thought to be mediated through regulation of plasma calcium levels, which in turn regulate insulin synthesis and secretion, and also through a direct action on pancreatic beta-cell function^20–23. Most^24–28, but not all²⁹, prospective epidemiological studies have shown an association between low serum 25(OH)D and diabetes incidence. Intervention studies evaluating the relation between vitamin D supplementation and markers of glucose metabolism have yielded mixed results, with some finding vitamin D supplementation to be protective, while others failed to show an association^{16, 23, 30–32}. These interventions were, however, limited in that they were often short in duration, included very few subjects, and/or had varying supplement formulations. Clearly additional research is needed.

Proinflammatory effects for low 25(OH)D levels and increased PTH have also been posited, as they appear to stimulate cytokine release from vascular smooth muscle cells and lymphocytes^33–36. In several small clinical trials, most in populations with non-cardiac inflammatory diseases, vitamin D supplementation has been shown to lower inflammatory marker levels^{37, 38}.

Additional mechanisms through which low vitamin D may influence CVD risk include its association with elevated triglyceride levels^39–41, promotion of atherogenic foam cells⁴², and direct cardiac effects, such as cardiomyocyte hypertrophy, apoptosis, and fibrosis of the left ventricle and vascular medial smooth muscle tissue^{33, 43–45}.

Observational Studies of Vitamin D and CVD Outcomes

The vast majority of the literature on vitamin D and CVD is based upon observational epidemiologic data. Low 25(OH)D levels have been associated with greater risk of combined CVD endpoints^{46, 47}, and CVD mortality^{47–53,54,55}, as well as specific CVD outcomes.

Relations between low 25(OH)D levels and myocardial infarction (MI) have been observed in three prospective cohort studies^{47, 56, 57}, several small case-control studies^58–60, and cross-sectional data from the third National Health and Nutrition Examination Survey (NHANES)⁶¹. In a prospective nested case-control study of 18,225 men from the Health Professional’s Follow-up Study, the risk of MI was 50% higher among men with serum 25(OH)D levels <15 ng/mL as compared to men with levels >30 ng/mL⁵⁶.

In the Cardiovascular Health Study of older adults, low levels of 25(OH)D <20 ng/ml were also associated with an increased risk of sudden cardiac death,⁶² as well as a composite outcome that included hip fractures, MIs, cancer, and death⁶³.

A few studies have explored the associations of low vitamin D with risk of cerebrovascular disease. In the Mini-Finland Health Survey, which included 6,219 initially healthy individuals who were followed for over 27 years, participants in the highest quintile of 25(OH)D levels were at lower risk of incident stroke than were those in the lowest quintile [HR:0.48 (95% CI: 0.31–0.75)]⁴⁸. Inverse associations between 25(OH)D levels and fatal stroke were also observed among members of the Intermountain Healthcare system⁴⁷, among white but not black participants in an analysis of NHANES-III data⁶⁴, in a sample of patients referred for coronary angiography⁶⁵, and in two small case-control studies^{66, 67}. Low dietary vitamin D intake predicts 34-year stroke incidence among Japanese-American men⁶⁸.

Low vitamin D levels have also been associated with heart failure. Members of the Intermountain Healthcare system who had low 25(OH)D (<15 ng/mL) were at double the risk of incident heart failure relative to those with levels >30 ng/mL⁴⁷. In a population of Germans referred for coronary angiography low vitamin D levels were also associated with increased risk of heart failure mortality⁶⁹. Additionally, low 25(OH)D has also been linked to congestive heart failure in cross-sectional data from NHANES⁶¹ and in a number of small clinical studies^70–73. In the one study to date, no association was observed between serum vitamin D levels and risk of atrial fibrillation⁷⁴.

In summary, the epidemiological evidence suggests that lower 25(OH)D levels are associated with greater risk of CVD events. However, caution should be used in interpreting these findings as residual or uncontrolled confounding may have contributed to these associations. Of additional potential concern is reverse causality bias, as individuals who are less healthy may participate in fewer outdoor activities as compared to healthy adults, which may in turn lead to less sunlight exposure and a greater propensity for suboptimal vitamin D levels.

Clinical Trials of Vitamin D and CVD Outcomes

The key question regarding the role of vitamin D in CVD risk is whether a causal relationship exists. To date, no large-scale randomized trials have been completed with clinical CVD events as the primary prespecified outcome⁷⁵.

Two randomized trials which tested moderate to high doses of vitamin D supplements (100,000 IU/4 months⁷⁶ and 1000 IU/d⁷⁷) without co-administration of calcium found a slight but nonsignificant reduction in CVD risk among those randomized to vitamin D versus placebo [meta-analysis HR:0.90 (95% CI: 0.77–1.05)]⁷⁸. Two other trials tested combined vitamin D and calcium supplementation versus placebo [including the large Women’s Health Initiative (WHI)], and both trials showed no differences in CVD risk between the treatment groups^{79, 80}. These studies were, however, limited. It is widely believed that in WHI the vitamin D supplement dose (400 IU/d) was insufficient to alter rates between the treatment groups^{37, 79}. Every 100 IU vitamin D ingested daily increases the 25(OH)D level by about 1 ng/mL^81–83. The other study was very small (n = 192), and likely underpowered for CVD events⁸⁰.

Two meta-analyses of randomized vitamin D supplementation trials on any health condition found a reduced mortality from any cause for those randomized to supplements as compared to placebo. The first meta-analysis showed a statistically significant reduction in mortality, albeit modest⁸⁴. The more recent one from the Cochrane Collaboration was a meta-analysis of 50 randomized vitamin D supplementation trials, with a total of nearly 100,000 participants, which found a trend in reduced mortality from any cause for those randomized to vitamin D supplements as compared to placebo of borderline statistical significance [HR: 0.97 (95% CI: 0.94–1.00). Notably, most trials included only elderly women (older than 70 years)⁸⁵. Thus, it is uncertain how generalizable this finding is to men and younger adults. Also, the meta-analyses are limited by combining heterogenous studies involving different types of vitamin D, differing dosing, and differing patient populations. Despite being unclear which conditions contributed to the lower mortality, this finding is intriguing.

Calcium Metabolism

Dairy products, such as milk, yogurt, and cheese are the most common sources of dietary calcium. As noted in the introduction, calcium is involved in the function of all cells. In addition to its major roles in forming and maintaining bones, calcium is also involved in cell metabolism, muscle contraction, blood coagulation, and transmitting nerve impulses.

The physiology of calcium and vitamin D are intricately intertwined. Calcitriol, the active hormone form of vitamin D (1,25(OH)₂D), enhances intestinal absorption of calcium². Low concentrations of 1,25(OH)₂D are associated with impaired calcium absorption, a negative calcium balance, and a compensatory rise in PTH. Elevated PTH stimulates the osteoclast activity, thereby resulting in increased release of calcium from the bones and higher serum calcium levels. Levels of serum calcium are highly regulated in the body, thus bone mass will be lost in order to maintain serum calcium levels. Given the complex interrelations between calcium and vitamin D, these nutrients are often prescribed in combination for the promotion of skeletal health, particularly among postmenopausal women who are at high risk of osteoporosis and fracture.

Observational Studies of Calcium and CVD Risk

Dietary calcium has been reported to be associated with lower risk of obesity, hypertension, and the metabolic syndrome,⁸⁶ with generally favorable relations reported between calcium markers of glucose/insulin homeostasis or diabetes^87–89, and lipid levels⁹⁰.

As has recently been reviewed elsewhere⁹¹, results of studies of dietary calcium intake and CVD risk have been mixed, with a few studies showing an inverse (protective) association, and many others yielding null results. Interestingly, a recent epidemiologic paper from the EPIC-Heidelberg cohort (n=23, 980 participants aged 35–64 years free of CVD at baseline) found a benefit for moderate intake of total dietary calcium with the third quartile (mean=820 mg/day) compared to lowest quartile having about a 30% lower risk of MI (HR 0.69, 95% CI 0.50–0.94)⁹¹. However, users of calcium supplements compared to non-users had increased risk of MI (HR 2.39, 95% CI 1.12–5.12). This data suggests that large boluses of calcium taken once or twice daily by supplements do not confer the same metabolic effect as calcium ingested throughout the day in small doses via a balanced diet.

Clinical Trials of Calcium Supplements and Cardiovascular Risk

Calcium supplements modestly increase bone mineral density and reduce fracture risk^{3, 4}. A relatively strong body of evidence suggests that calcium supplementation is associated with lower blood pressure, compared to placebo. In a 2006 meta-analysis of 40 randomized clinical trials reported that calcium supplementation (mean daily dose = 1200 mg) reduced systolic BP by 1.86 (95% CI: 0.81 to 2.91) mmHg and diastolic BP by 0.99 (95% CI: 0.37 to 1.61) mmHg⁹². However recent studies have suggested possible CVD harm associated with calcium supplementation, although results are inconclusive.

A clinical trial of calcium supplements (without vitamin D) vs. placebo conducted in New Zealand in 2008 suggested an increased risk of MI associated with calcium supplements which was not statistically significant when additional adjudicated events were included⁹³. This was followed by a 2010 meta-analysis from the same group including 15 clinical trials (5 with patient level data) which again found an increased risk of MI for those assigned to calcium therapy (without vitamin D) [HR:1.31 (95% CI 1.02–1.67)] but no statistically increased risk for mortality or the composite CVD outcome⁹⁴. This was in contrast to another 2010 meta-analysis of clinical trials which found no statistically significant increased CVD risk for calcium supplementation alone or in combination with vitamin D⁷⁸.

Next, there was a recent re-analysis of the data from WHI. While the overall WHI results showed no evidence for CVD benefit or harm with assignment to calcium plus vitamin D therapy vs. placebo⁷⁹, Bolland et al found an interaction that suggested women who were not taking personal calcium supplements before randomization had a higher risk for CVD events when assigned to calcium/vitamin D therapy during the trial compared to placebo but no alteration in CVD risk for women already taking supplements before entry into the study⁹⁵. However, post-hoc subgroup analyses should always be viewed with caution. Furthermore, there was also no dose response seen which leads to the question if calcium is so harmful, why did the group taking the most not fare the worst? It is possible that a threshold exists, whereby calcium supplementation beyond a certain level does not result in further increased risk. An updated 2011 meta-analysis of placebo controlled trials by these authors found that calcium supplements with or without vitamin D modestly increases the risk of MI [RR:1.24 (95% CI: 1.07–1.45)].⁹⁵

The Calcium and Heart Disease Quandary

Counter to studies of dietary calcium intake and CVD risk, most^96–98, though not all⁹⁹, studies of serum calcium levels have shown a positive relation with CVD risk. These higher serum calcium levels are thought to promote CVD and atherogenesis through vascular calcifications and increased coagulability^{3, 4}. It has been suggested that the potential harm of calcium supplements is that large boluses can transiently increase serum calcium levels, even above the normal range, for several hours after ingestion. This may potentially cause similar vascular effects in these normocalcemic patients, as described in patients with higher serum calcium levels. However the current data is insufficient to confirm or refute this concern. More studies are clearly needed to clarify this potential risk.

Food sources, which have not been linked to increased CVD risk, may be the best source of calcium, perhaps because calcium ingested from food sources is absorbed more slowly in smaller amounts throughout the day, thus not substantially changing serum calcium levels. Pending future studies, given the only modest reduction in fractures and the mildly increased CVD risk, there may be no net benefit for calcium supplementation in post-menopausal women and possibly even potential harm.

Current IOM Recommendations Regarding Supplement Use

As of the 2010 Institute of Medicine (IOM) report, the recommended daily allowance of vitamin D for adults is 600 IU/day and 800 IU/day for those aged<70 and ≥70 years, respectively.¹⁰⁰ Notably, the IOM report did not recommend vitamin D supplementation for CVD prevention and suggested serum levels >20 ng/ml may be sufficient for health (rather than the higher targets>30 or >40 ng/ml recommended by many experts in the field). The calcium recommended daily allowances from those same guidelines range from 1000–1200 mg/day for adults.

The IOM authors emphasized cautious interpretation of observational epidemiological studies, as correlation does not equal causation, and stressed the need for large-scale clinical trials. Historically, several micronutrients have appeared to be associated with lower CVD risk in observational epidemiologic studies, but have failed to show an effect when tested in supplementation trials.

Several randomized clinical trials are currently underway, which will hopefully provide insight as to whether suboptimal vitamin D is causally related to CVD risk. The largest is the VITamin D and OmegA-3 TriaL (VITAL). It is a 2×2 factorial, double-blind, placebo-controlled clinical trial which will randomize 20,000 participants to 2000 IU/day of vitamin D (without calcium) and omega-3 dietary supplements¹⁰¹. VITAL began recruitment in 2010 and will follow participants for 5 years for cancer and CVD outcomes. Hopefully trials such as VITAL will provide important insight into whether associations between vitamin D and CVD are causal. Concerns about VITAL trial do, however exist. For instance, 25(OH)D deficiency was not a criteria for enrollment. Likely vitamin D replacement would most benefit those with 25(OH)D deficiency and may have little benefit in those who already have adequate 25(OH)D levels. Additionally, if vitamin D impacts CVD risk through diabetes, hypertension, and inflammation, given the age of participants (i.e. men ≥50 years, women ≥55 years), it is possible that the intervention is taking place too late in the natural history of the disease.

Conclusions

In summary, vitamin D status remains a marker of interest given the encouraging findings from observational data and the limited clinical trials to date. Although biologically plausible mechanisms have been proposed and the epidemiologic evidence is intriguing, only clinical trials are capable of establishing causality. Thus, pending the results of ongoing clinical trials, we agree with the IOM that presently there is insufficient evidence to recommend vitamin D supplements for the sole purpose of CVD prevention. At this time, the evidence for calcium supplements and increased CVD risk is also inconclusive but raises some concern. Moderate intake of calcium from dietary foods may be the best way to achieve target daily calcium goals, although calcium supplements may be needed for individuals not meeting the recommended daily allowances as outlined by IOM. Given remaining uncertainty about how to best advise patients, large designated randomized clinical trials with CVD events as the primary clinical endpoint are warranted for both calcium and vitamin D supplements.