Abstract
In the clinical setting, administration of high daily or bolus doses of vitamin D is often solely based on 25-hydroxyvitamin D [25(OH)D] testing. This review summarizes the evidence of the effect of vitamin D on cardiovascular disease (CVD). Meta-analyses of randomized controlled trials (RCTs) have demonstrated that CVD risk markers, such as lipid parameters, inflammation markers, blood pressure, and arterial stiffness, are largely unaffected by vitamin D supplementation. Similar results have been obtained regarding CVD events and mortality from (meta)-analyses of RCTs, even in subgroups with 25(OH)D concentrations <50 nmol/l. Likewise, Mendelian randomization studies have indicated that the genetic reduction of the 25(OH)D concentration does not increase CVD risk. Some studies do not exclude the possibility of adverse vitamin D effects, such as elevated plasma calcium concentration and an increased CVD risk at a 25(OH)D concentration >125 nmol/l. Based on a conservative benefit–risk management approach, vitamin D doses beyond the nutritionally recommended amounts of 600 to 800 IE daily currently cannot be advised for the prevention of CVD events.
- Vitamin D
- cardiovascular disease
- meta-analysis
- randomized controlled trial
- Mendelian randomization study
- review
With more than 70,000 hits available in pubmed by January 2019, vitamin D is the vitamin with the greatest scientific interest. Moreover, the number of high-quality articles on vitamin D, such as large prospective cohort studies or randomized controlled trials (RCTs), and meta-analyses of these types of studies steadily increased from 1993 to 2017 (Figure 1). Likewise, Mendelian randomization studies have shed light on the regulation of vitamin D metabolism and on health-related functions of vitamin D.
To become biologically active, vitamin D has to be metabolized into 1,25-dihydroxy-vitamin D (1,25[OH])2D) through two hydroxylation steps, which occur mainly in the liver (25-hydroxylase) and the kidney (1α-hydroxylase) (1). Whereas 25-hydroxyvitamin D (25[OH]D) is the most abundant vitamin D metabolite in the circulation, 1,25(OH)2D is a steroid hormone, whose receptors are found in almost all tissues of the human body (1). In addition, a number of tissues are also capable of producing 1,25(OH)2D (2), indicating the great importance of this substance in human health and disease.
The present article is an update of a narrative review on the potential role of vitamin D in cardiovascular disease (CVD) (3).
Search Strategy
For this review, we performed a systematic literature search in pubmed for relevant publications released before January 31, 2019. We searched for the following terms: ‘vitamin D’ or ‘cholecalciferol’ or ‘calcitriol’ or ‘25-hydroxyvitamin D’ or ‘1,25-dihydroxyvitamin D’ combined with ‘cardiovascular disease’ or ‘lipid parameters’ or ‘inflammation markers’ or ‘blood pressure’ or ‘arterial stiffness’ or ‘vascular calcification’ or ‘cardiovascular mortality’ or ‘all-cause mortality’ or ‘overall mortality’. Personal collections on this topic, as well as references from selected articles, were also used to extend the search. Some articles are not cited due to space limitations.
Assessment of Vitamin D Status and its Clinical Implications Recommendations on Vitamin D Status and Intake
Circulating 25(OH)D is generally accepted as the indicator of vitamin D status (4). Nevertheless, there is an ongoing debate on the threshold for inadequate, adequate, and potentially harmful concentrations. Several authoritative institutions such as North American and European nutrition societies have set the threshold for adequate 25(OH)D concentrations at 50 nmol/l (Table I) (4-7). Most of the nutrition societies suggest that the oral vitamin D intake required to achieve this concentration is 600 IU-800 IU daily, depending on whether low or no cutaneous vitamin D synthesis is considered. The scientific basis and rationale underlying these recommendations have been described elsewhere (4-6). Vitamin D recommendations by nutrition societies are based on bone health (4-7), whereas other organizations (8, 9) also take potentially non-classical vitamin D effects into consideration. These organizations argue that higher doses are necessary for the prevention of vitamin D-related diseases (8, 9). They have classified 25(OH)D concentrations up to 125 nmol/L as the lower threshold for adequate 25(OH)D concentrations, and recommend an oral vitamin D intake of up to 5,000 IU vitamin D daily or more (Table I). Similar to the large disparity between some organizations and institutions regarding recommended intakes, the classification of potentially harmful 25(OH)D concentrations and the corresponding upper tolerable intake levels vary between 125 and 250-400 nmol/l, and 4,000 and 10,000 IU daily, respectively (4-10).
Vitamin D Testing and Dosing
In the clinical setting, testing of vitamin D status and recommendations on vitamin D supplementation are often largely based on the aforementioned measurement of circulating 25(OH)D concentrations (11, 12). Vitamin D supplements can be used to guarantee an adequate vitamin D status, i.e. 25(OH)D concentrations > 50 nmol/l. However, in some European countries, such as Germany, vitamin D preparations are legally regarded as dietary supplements only if the daily dosage is within the official dietary reference intake of 800 IU (13). Nonetheless, in cases of a 25(OH)D concentration below 50 nmol/l, even higher doses (see before) or bolus doses (e.g. 50,000 IU vitamin D) are also frequently administered in the clinical setting. These high-dose preparations are used for restoring, correcting, or influencing physiological functions, and/or are intended for the purpose of healing, alleviating or preventing diseases. Therefore, they formally have to be considered as drugs (13). Actually, they would require approval as a drug if vitamin D was a novel substance and not a 100-year-old known vitamin, and if diseases other than vitamin D-dependent rickets or osteomalacia should be prevented or treated.
Critique regarding vitamin D blood testing. Measurement of 25(OH)D as the exclusive parameter for the assessment of vitamin D status is acceptable in large studies in the apparently healthy general population. In the clinical setting, however, caution is necessary when using this parameter as the sole criterion for assessing vitamin D status, especially if the measurement is used to recommend oral vitamin D doses much higher than 600 to 800 IU daily. Plasma 25(OH)D concentrations account for only 10% of the body's vitamin D content (14). This limits the use of this parameter in estimating vitamin D supply to its target tissues correctly. Notably, the increment in circulating 25(OH)D decreases sharply at vitamin D doses beyond 1200 IU daily (15) and circulating 25(OH)D concentrations above 100 nmol/l (16). Moreover, circulating 25(OH)D is influenced by common genetic variants in the 25-hydroxylase (CYP2R1) gene and the vitamin D binding protein gene (17). Inter-assay variability of 25(OH)D measurement is another issue, pointing to the importance of assay standardization (18), not only for the comparison of different studies with each other, but also for the correct measurement of individual 25(OH)D concentrations. Although it has been discussed whether measurement of bioactive 25(OH)D may be superior to measuring the total 25(OH)D for the vitamin D status assessment, there is little evidence to date for this assumption (19-21), with a potential exception in hormonal contraceptive users (22).
We should also bear in mind that the concentration of 1,25(OH)2D, the active vitamin D hormone, in the circulation accounts for only 1/1000 of the concentration of its substrate 25(OH)D. Because of parathyroid hormone (PTH) induces stimulation of renal 1α-hydroxylase, even in cases of low circulating 25(OH)D concentrations, circulating 1,25(OH)2D concentrations remain relatively constant and become only significantly substrate-dependent at extremely low 25(OH)D concentrations (4). Although vitamin D deficiency-induced secondary hyperparathyroidism can result in bone loss (23), the effect on other organ systems, such as the cardiovascular system, is less clear. Low circulating 1,25(OH)2D concentrations in some diseases, such as end-stage heart failure or end-stage kidney disease, should not be a priori considered to be caused by inadequate 25(OH)D availability, but may be the result of the activation of mechanisms to protect the human body from calcium and phosphate intoxication (24). Likewise, in patients with advanced heart failure and preserved kidney function, secondary hyperparathyroidism is not necessarily caused by vitamin D deficiency, but may be a disease-related compensatory effect, because PTH may also exert beneficial cardiac effects, such as increased heart rate, myocardial blood flow, and cardiac output (25, 26). The assessment of vitamin D status by biochemical blood parameters is further complicated by the fact that local regulation of 1,25(OH)2D in extra-renal tissues can be independent of substrate availability, i.e. 25(OH)D concentrations (27, 28). Finally, vitamin D receptor Cdx2, Fok1, Bsm1, Apa1, Bgl1, Taq1, and Poly (A) gene polymorphisms may influence receptor-mediated cellular vitamin D effects (29). Altogether, the various factors that influence vitamin D metabolism make health-related recommendations difficult if they are based solely on the measurement of circulating vitamin D metabolite concentrations or on associations of vitamin D metabolites with clinical outcomes. This limitation should also be considered when interpreting results of observational studies regarding CVD. Therefore, RCTs are necessary, not only to avoid the problem of unexplained confounding, but also to determine the required vitamin D dose and to specify the diseases in which vitamin D supplementation is useful. Ideally, RCTs should be performed in patients with vitamin D deficiency and well-described genetic backgrounds. In addition, they should be designed for the prevention or treatment of well-described diseases. Furthermore, they should be adequately powered to assess whether or not vitamin D is able to influence clinically relevant endpoints. Both efficacy and safety should be adequately considered.
Pubmed-listed vitamin D publications during the last 25 years. RCT: Randomized controlled trial.
Recommended and safe circulating 25-hydroxyvitamin D concentrations, as well as daily intake values of vitamin D in adults
Biochemical and Clinical CVD Risk Markers
Dyslipoproteinemia, high concentrations of pro-inflammatory cytokines, high blood pressure, and high values of parameters of arterial stiffness are considered to be risk factors for CVD. A large recent meta-analysis (30) summarized data from RCTs regarding the effect of vitamin D supplementation on lipid parameters and the inflammation marker high-sensitive C-reactive protein (hs-CRP). The average vitamin D dose was ~3,000 IU/day; ~two-thirds of studies had mean baseline 25(OH)D levels <50nmol/l, and the increase in circulating 25(OH)D was 48±23 nmol/l. Data indicate a significant reduction in total cholesterol, LDL-cholesterol, and triglycerides of −0.15 (95%CI=−0.25 - −0.04) mmol/l, −0.10 (95%CI=−0.20 - −0.003) mmol/l, and −0.12 (95%CI=−0.23 - −0.003) mmol/l, respectively; an increase in HDL-cholesterol of 0.09 (95%CI=0.00-0.17) mmol/l and a reduction in hs-CRP of −0.20 (95%CI=−0.34 - −0.06) mg/dl by vitamin D supplementation. There was substantial heterogeneity among studies, and subgroup analysis indicates that the vitamin D effect on triglycerides and HDL-cholesterol was higher if study participants were supplemented for ≥6 months. However, there was no significant vitamin D effect according to baseline 25(OH)D concentration, daily vitamin D dose, or calcium co-administration on lipid parameters. Regarding hs-CRP, concentrations were marginally lower at doses ≥4000 IU vitamin D/day than at doses <4,000 IU daily. Another meta-analysis in patients with HF (31) reported a significant suppression of the pro-inflammatory cytokine tumor necrosis factor-α, whereas concentrations of interleukin-6 and CRP remained unaffected. It was concluded that vitamin D supplementation may have specific, but modest effects on inflammatory markers in this group of patients. The aforementioned large meta-analysis (30) also presented data on blood pressure, and parameters of arterial stiffness, such as peak wave velocity and Augmentation Index. Data of 39 included RCTs indicate a small, but significant reduction in systolic blood pressure of −0.102±0.04 mmHg (95%CI=−0.20 - −0.03) and diastolic blood pressure of −0.07±0.03 mmHg (95%CI=−0.14 - −0.006) by vitamin D supplementation. The effects were more pronounced if in-study 25(OH)D ≥86 nmol/l were achieved, the daily vitamin D dose was ≥4,000 IU, and the duration of intervention was ≥6 months, but was unaffected by baseline 25(OH)D concentration. Generally, results cover the 95%CI of an earlier meta-analysis incorporating individual patient data from 27 RCTs, concluding that vitamin D supplementation is ineffective as an agent for lowering blood pressure (32). In that meta-analysis, the mean difference in systolic and diastolic blood pressure between individuals assigned to vitamin D or placebo was -0.5 (95%CI=−1.2-0.4) mmHg and 0.2 (95%CI=−0.3-0.7) mmHg. Both meta-analyses are also in general agreement with a large Mendelian randomization study (33), indicating a non-significant reduction of -0.37 mmHg (−0.73-0.003; p=0.052) in systolic blood pressure and a significant reduction of −0.29 mmHg (−0.52 - −0.07; p=0.01) in diastolic blood pressure with each 10% increase in genetically determined 25(OH)D concentration. With regard to arterial stiffness, the large meta-analysis by Mirhosseini et al. (30) could not show any significant effects of vitamin D supplementation, neither on pulse wave velocity, nor on Augmentation Index. Altogether, some significant but small effects of vitamin D supplementation on biochemical CVD risk markers cannot be ruled out. However, the clinical relevance of these effects is questionable, since clinical surrogate parameters of CVD risk such as blood pressure and arterial stiffness seem to be largely unaffected by vitamin D supplementation.
Clinical Endpoints
Some recent meta-analyses have summarized data from RCTs regarding the effect of vitamin D supplementation on CVD outcomes (34-36). Data indicate that vitamin D supplementation influences neither non-fatal CVD events, such as the risk of myocardial infarction, stroke or ischemic heart disease, nor the risk of CVD deaths. However, these meta-analyses were primarily based on studies in which CVD outcomes were only secondary endpoints. Therefore, the results of two recent very large vitamin D supplementation studies in the elderly general population are important: In a New Zealand study (37), 5,110 community-resident adults aged 50 to 84 years were assigned to a monthly vitamin D bolus of 100,000 IU vitamin D or placebo for a mean duration of 3.34 years. The primary endpoint was incident CVD and death, including a pre-specified subgroup analysis in participants with baseline 25(OH)D concentrations <50 nmol/l. In a nationwide US study (38), 25,871 men 50 years of age or older and women 55 years of age or older received 2,000 IU vitamin D daily, marine n-3 fatty acids or placebo by a two-by-two factorial design for a median duration of 5.3 years. The primary endpoint was a composite of myocardial infarction, stroke, or death from cardiovascular causes. In the New Zealand study (37), the incidence of the primary endpoint was reached by 11.8% in the vitamin D group and 11.5% in the placebo group, yielding an adjusted hazard ratio of 1.02 (95%CI=0.87-1.20). Similar results were seen for participants with baseline vitamin D levels <50 nmol/l and for secondary endpoints such as myocardial infarction, angina, heart failure, hypertension, arrhythmias, arteriosclerosis, stroke, and venous thrombosis. In the US study (38), the hazard ratio of the primary endpoint was for the vitamin D versus the placebo group 0.97 (95%CI=0.85-1.12). Likewise, the hazard ratio did not differ significantly between the study groups for an expanded composite of major cardiovascular events plus coronary revascularization and the individual components of major cardiovascular events. Moreover, subgroup analysis did not indicate an interaction of age, sex, race, body mass index or baseline 25(OH)D level (<50 nmol/l and ≥50 nmol/l) with the study group with regard to major CVD events. In both aforementioned RCTs, the prevalence of 25(OH)D levels classified by the IOM as deficient (<30 nmol/l) (4), was not explicitly specified, but was apparently very low. Results published by a European consortium of eight prospective studies using individual patient data and standardized 25(OH)D values indicate the highest CVD mortality at 25(OH)D levels less than 30 nmol/l (39). Likewise, a large meta-analysis of cohort studies (40) reported a sharp non-linear increase in the risk of CVD events and CVD mortality at 25(OH)D concentrations below 37 nmol/l. A beneficial vitamin D effect on CVD risk can thus at best be expected in individuals with baseline 25(OH)D concentrations within the deficiency range, i.e. <30 nmol/l. In line with this assumption, vitamin D supplementation was not associated with a reduction in CVD events in a very recent meta-analysis of RCTs in individuals not selected for initial 25(OH)D levels below 30 nmol/l (41).
In line with the results of the RCTs, large Mendelian randomization studies indicate that genetically lowered 25(OH)D levels are not associated with increased risk of coronary artery disease (CAD) or myocardial infarction (42, 43). Genetically enhanced 25(OH)D levels of 20 nmol/l were more likely to be associated with a marginally higher risk of CVD mortality (odds ratio=1.30; 95%CI=0.93-1.82) (44). However, results were obtained in the clinical setting and may reflect the association between 25(OH)D and disease progression, rather than disease occurrence. With respect to the associations between the polymorphisms of vitamin D receptor and CAD, two meta-analyses (45, 46) provided inconsistent results. Whereas one meta-analysis (45) concluded that the Apa 1, Fok 1, Taq 1, and Bsm 1 polymorphisms of the VDR gene may not be associated with genetic susceptibility to CAD, the other meta-analysis (46) concluded that the Fok 1 polymorphism may play a protective role in CAD and the Taq 1 polymorphism is associated with a significant increase in CAD risk. The analysis by Alizadeh et al. (45) was based on nine studies involving a total of 5,259 cases and 1,981 controls, and the analysis by Lu et al. (46) was based on seven studies involving 2,306 CAD patients and 4,151 controls.
Overdosing of Vitamin D and CVD
In the two aforementioned large population-based RCTs (37, 38), the incidence of hypercalcemia, which is the hallmark of vitamin D intoxication, did not differ between participants assigned to vitamin D or placebo, and there was no evidence for adverse vitamin D effects on the cardiovascular system. In the New Zealand and US studies, mean in-study 25(OH)D concentrations in the participants with available 25(OH)D data were 132 nmol/l and 104 nmol/l, respectively. Moreover, a large meta-analysis of cohort studies by Zhang et al. (40) did not provide evidence regarding the adverse effects of 25(OH)D levels between 100 and 137 nmol/l on CVD events in the general population. However, data should not be generalized to all groups of individuals. Some large cohort studies from the clinical setting indicate an increase in CVD events at 25(OH)D levels above 100 nmol/l (47-49). In an RCT in patients with advanced heart failure (50), a daily vitamin D3 supplement of 4,000 IU for 3 years resulted in a greater need for mechanical circulatory support implants. In end-stage heart failure patients, the devices are implanted as last option to prevent death. The need of device implantation was highest in the subgroup of patients achieving in-study 25(OH)D levels above 100 nmol/l. Compared to placebo, vitamin D also resulted in significantly higher plasma calcium levels and a non-significant higher incidence of hypercalcemia (6.2% vs. 3.1%). Vitamin D-induced hypercalcemia has been associated with vascular calcification (51,52) and mild hypercalcemia had been reported in 28% (n=213) and 2% (n=12) of infants at 6 and 12 months, respectively, during daily vitamin D supplementation of 400 or 1,200 IU (53). The infants with mild hypercalcemia at 12 months of age had an average 25(OH)D concentration of 110 nmol/l. A recent meta-analysis regarding long-term (≥1 year), high dose (median of the calculated daily dose: 4,000 IU) vitamin D supplementation and hypercalcemia events reported a risk ratio of 1.93 (95%CI=1.00-3.73; p=0.05) in the vitamin D group vs. the placebo group (54). The clinical relevance of the slight increase in plasma calcium levels at physiological daily vitamin D doses is unknown at present. However, higher plasma calcium levels are non-linearly associated with an increased incidence of heart failure (55). Since the evidence for beneficial vitamin D effects on CVD risk is lacking, caution regarding vitamin D supplementation is necessary. In nutritional science, a conservative benefit-risk management approach is proposed for food components: only convincing evidence should be taken into account for beneficial effects, but also probable and possible evidence for adverse effects (56). Therefore, circulating 25(OH)D levels of 100 to 125 nmol/l should not be exceeded. In some adults, 4,000 IU vitamin D daily may already be too high to avoid 25(OH)D levels >100 nmol/l (50). Thus, the approach to legally regard vitamin D preparations as dietary supplements only if the daily dosage is within the official dietary reference intake of 800 IU (13), seems to make sense.
Conclusion
Experimental studies have shown that animals lacking vitamin D action, induced either by deletion of the vitamin D receptor or by vitamin D-depleted diets, develop vascular calcification and atherosclerosis (57, 58). However, available data from large prospective cohort studies and RCTs indicate that no beneficial vitamin D effects on the cardiovascular system can be expected in individuals with 25(OH)D levels above 30 nmol/l. In the adult US and European populations, the prevalence of circulating 25(OH)D concentrations below 30 nmol/l is 8% and 13%, respectively (59, 60). Potential future studies regarding vitamin D and CVD risk should focus on this group of individuals. However, the time window for these studies is closing since food fortification with vitamin D has already been introduced in some countries and is recommended for other countries as a safe and cost-effective strategy for preventing deficient 25(OH)D levels (61).
Until food fortification with vitamin D is implemented in the general population, a nutritional supplement of 600 to 800 IU vitamin D daily can be used to prevent a decrease in circulating 25(OH)D concentrations below 30 nmol/l if skin synthesis of vitamin D is low or absent. However, convincing evidence for reducing the risk of CVD by vitamin D supplementation is lacking.
Footnotes
Authors' Contributions
AZ and SP both contributed to the concept and design of the work. AZ drafted the manuscript. SP critically reviewed it for its intellectual content. Both Authors approved the final version of the article.
Conflicts of Interest
The Authors have no conflicts of interest to declare regarding this study.
- Received June 18, 2019.
- Revision received July 2, 2019.
- Accepted July 3, 2019.
- Copyright© 2019, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.