25-Hydroxyvitamin D deficiency is independently associated with cardiovascular disease in the Third National Health and Nutrition Examination Survey

25-Hydroxyvitamin D deficiency is independently associated with cardiovascular disease in the Third National Health and Nutrition Examination Survey

Atherosclerosis 205 (2009) 255–260 Contents lists available at ScienceDirect Atherosclerosis journal homepage: www.elsevier.com/locate/atheroscleros...

264KB Sizes 0 Downloads 49 Views

Atherosclerosis 205 (2009) 255–260

Contents lists available at ScienceDirect

Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis

25-Hydroxyvitamin D deficiency is independently associated with cardiovascular disease in the Third National Health and Nutrition Examination Survey Jessica Kendrick a , Giovanni Targher b , Gerard Smits a , Michel Chonchol a,∗ a b

Division of Renal Diseases and Hypertension, University of Colorado Health Sciences Center, Denver, USA Section of Endocrinology, Department of Biomedical and Surgical Sciences, University of Verona, Verona, Italy

a r t i c l e

i n f o

Article history: Received 18 July 2008 Received in revised form 12 September 2008 Accepted 30 October 2008 Available online 11 November 2008 Keywords: 25-Hydroxyvitamin D Cardiovascular disease Epidemiology

a b s t r a c t Objective: Serum 25-hydroxyvitamin D [25(OH)D] levels are inversely associated with important cardiovascular disease (CVD) risk factors. However, the association between 25(OH)D levels and prevalent CVD has not been extensively examined in the general population. Methods: We performed a cross-sectional analysis of data from the Third National Health and Nutrition Examination Survey (1988–1994) and examined the association between serum 25(OH)D levels and prevalence of CVD in a representative population-based sample of 16,603 men and women aged 18 years or older. Prevalence of CVD was defined as a composite measure inclusive of self-reported angina, myocardial infarction or stroke. Results: In the whole population, there were 1308 (8%) subjects with self-reported CVD. Participants with CVD had a greater frequency of 25(OH)D deficiency [defined as serum 25(OH)D levels <20 ng/mL] than those without (29.3% vs. 21.4%; p < 0.0001). After adjustment for age, gender, race/ethnicity, season of measurement, physical activity, body mass index, smoking status, hypertension, diabetes, elevated lowdensity lipoprotein cholesterol, hypertriglyceridemia, low high-density lipoprotein cholesterol, chronic kidney disease and vitamin D use, participants with 25(OH)D deficiency had an increased risk of prevalent CVD (odds ratio 1.20 [95% confidence interval (CI) 1.01–1.36; p = 0.03]). Conclusions: These results indicate a strong and independent relationship of 25(OH)D deficiency with prevalent CVD in a large sample representative of the US adult population. © 2008 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Cardiovascular disease (CVD) is a major cause of morbidity and mortality in the United States, accounting for greater than 900,000 deaths per year [1]. Death rates from CVD are partly due to the high prevalence of traditional CVD risk factors such as hypertension, type 2 diabetes and dyslipidemia [1]. Recent reports from the National Health and Nutrition Examination Survey [2–6] and several other cohort studies [7–11] have found that serum levels of 25-hydroxyvitamin D [25(OH)D] are inversely associated with hypertension, diabetes, carotid atherosclerosis, myocardial infarction, congestive heart failure, stroke, microalbuminuria and decreased kidney function.

∗ Corresponding author at: University of Colorado Health Sciences Center, Division of Renal Diseases and Hypertension, Box C-281, Denver, CO 80262, USA. Tel.: +1 303 399 6997; fax: +1 303 399 3131. E-mail address: [email protected] (M. Chonchol). 0021-9150/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2008.10.033

Although, persons with lower serum 25(OH)D levels appear to be at increased risk for cardiovascular risk factors [12,13], it is unclear if 25(OH)D deficiency is related with prevalent CVD in the US population. Humans get vitamin D from exposure to sunlight, from diet, and from dietary supplements [13]. Only a few natural food sources (i.e., diets rich in oily fish) contain significant amounts of ergocalciferol (vitamin D2 ) and cholecalciferol (vitamin D3 ), but many foods are now fortified with vitamin D [13]. Nonetheless, vitamin D insufficiency or deficiency persists in most of the world including North America and Europe possibly due to nutritional deficits and perhaps to avoidance of sunlight and the use of sunscreens [13]. 25(OH)D is formed in the liver and it represents the principle storage form of vitamin D, which is used to clinically monitor overall vitamin D3 status [13]. The discovery that most tissues in the body possess vitamin D receptors (VDR) has provided new insights into the broad functions of vitamin D and the non-calcemic adverse effects of its deficiency [13,14]. Although 25(OH)D has a relatively low affinity for the VDR [15], many cells including vascular smooth

256

J. Kendrick et al. / Atherosclerosis 205 (2009) 255–260

muscle and endothelial cells also express 1-␣-hydroxylase that converts 25(OH)D to the active form 1,25(OH)2 D locally [16]. Thus, extra-renally produced 1,25(OH)2 D primarily serves autocrine or paracrine cell-specific functions, instead of endocrine functions. Because the implications of vitamin D insufficiency or deficiency for overall health could be substantial, we examined the association between prevalent CVD and 25(OH)D levels among the US adult population. To examine the relation of serum 25(OH)D deficiency with clinical CVD defined as a composite measure inclusive of angina, myocardial infarction and stroke, we analyzed data collected from a large representative sample of the US adult population enrolled in the Third National Health and Nutrition Examination Survey (NHANES III). 2. Methods 2.1. Study population and sample Data for this analysis were obtained from the NHANES III, a national cross-sectional survey conducted by the National Center for Health Statistics between 1988 and 1994 [17]. The survey was designed to obtain information representative to the non-institutionalized US population on health and nutritional status between 1988 and 1994. A stratified, multistage sampling design was used, with over-sampling of non-Hispanic Blacks, Mexican–Americans and persons over the age of 60 years. Data collection consisted of a standardized questionnaire, which was administered during a home interview followed by a detailed physical examination and included collection of blood specimens at a mobile examination center or at the participant’s home [17]. This analysis was restricted to 20,050 adult participants of 18 years or older. Responders, who had a missing serum 25(OH)D measurements (n = 261) or had incomplete data (n = 3186) for the calculation of estimated glomerular filtration rate (GFR) by the abbreviated Modification of Diet in Renal Disease formula (MDRD) [18], were excluded (n = 3447). Thus, the final sample used in this study included 16,603 adult participants. 2.2. Primary predictor and outcome The primary predictor or independent variable was 25(OH)D level which was obtained from the laboratory results data file. Serum 25(OH)D levels were measured using an INCSTAR 25(OH)D two step assay procedure with a coefficient of variation of less than 10%. The first step in the procedure involves the rapid extraction of 25(OH)D from the serum using acetonitrile. Following extraction, the treated sample is assayed by using an equilibrium radioimmunoassay procedure. This method is based on an antibody with specificity to 25(OH)D [17]. The sample, antibody, and tracer are incubated at 20–25 ◦ C for 90 min. A second antibody-precipitating complex is used to achieve phase separation. The radioimmunoassay method tends to overestimate the level of 25(OH)D because the antibody recognizes all forms of dihydroxyvitamin D and D steroids. However, in humans these are usually present in picomolar concentrations [17]. The outcome or dependent variable of interest was the presence of clinically manifest CVD, defined as a composite measure inclusive of self-reported angina pectoris, myocardial infarction or stroke. These self-reported diseases were extracted from the interview portion. The NHANES III survey included a questionnaire where participants were specifically asked whether they had a history of angina, myocardial infarction, or stroke [17]. Participants answered questions regarding angina based on the World Health Organization (WHO) Rose Angina questionnaire [19]. This ques-

tionnaire has been used widely and is validated for the diagnosis of angina [20–22]. Participants were defined as having angina if they reported any of the following: ever having pain or discomfort in the chest, chest pain occurring with walking uphill or in a hurry, chest pain causing them to stop and rest, chest pain relieved by rest or within 10 min of resting, substernal chest pain or pain radiating to the left arm. We considered participants to have had a myocardial infarction if they answered yes to the question “Has a doctor ever told you that you had a heart attack?” Participants who answered yes to the question “Has a doctor ever told you that you had a stroke?” were analyzed as having a past history of a stroke. These measures have also been used widely to study the prevalence and the natural history of coronary heart disease and stroke [23,24]. 2.3. Other measurements We chose covariates as potential confounding factors based on prior studies or based on their biological plausibility. The following covariates were included in analysis: age, gender, race/ethnicity, season of measurement, physical activity, body mass index (BMI), smoking status, hypertension, diabetes, elevated low-density lipoprotein cholesterol (LDL-C) and triglycerides levels, decreased high-density lipoprotein cholesterol (HDL-C) levels, chronic kidney disease (CKD), and use of vitamin D supplements. Age was stratified into groups: 18–29, 30–39, 40–49, 50–59, 60–69, and ≥70 years. Race/ethnicity was broken into four categories: non-Hispanic White, non-Hispanic Black, Mexican–American, and other. BMI was calculated as weight in kilograms divided by the square of height in meters and classified according to World Health Organization guidelines. Smoking status was classified as never, former, or current. Information was also collected on the use of vitamin D supplements in the previous months and the number of time a range of nine common physical activities were undertaken during leisure time in the previous month. A physical exercise score was formed by summing the number of activities reported. Due to the small number of subjects reporting more than 4 activities, the scale was restricted to values of 0 through 4. Participants were defined as having diabetes when a physician had ever told them that they had diabetes, were taking hypoglycemic medications or had a fasting glucose concentration ≥126 mg/dL. Hypertension was diagnosed if the participants reported being told by a physician that they have high blood pressure, were taking antihypertensive medications, or the average of three blood pressure readings was ≥140/90 mmHg. Fasting levels of HDL-C and triglycerides were measured enzymatically with a Hitachi 704 Analyzer (Boehringer Mannheim Diagnostics, Indianapolis, IN). LDL-C concentration was calculated using the Friedewald’s equation (i.e., LDL-C = total cholesterol-HDLC - triglyceride/5), except for those with triglycerides exceeding 400 mg/dL [25]. Participants with LDL-C ≥ 160 mg/dL, triglycerides ≥ 150 mg/dL or HDL-C ≤ 35 mg/dL were considered to have high LDL-C and triglycerides or low HDL-C levels, respectively. Glomerular filtration rate was estimated from the abbreviated MDRD Study formula [18] as follows: estimated GFR = 186.3 × (serum creatinine mg/dL)−1.154 × age−0.203 × (0.742 if female) × (1.21 if Black). Serum creatinine measurements were recalibrated to the Cleveland Clinic “standard” assay used in the development of the MDRD-GFR prediction equation to ensure validity of the results [26]. In these analyses, CKD was defined as estimated GFR less than 60 mL/min/1.73 m2 . 2.4. Analytical methods Unadjusted cross tabulations were first run for each of the covariates and 25(OH)D with prevalent CVD to determine risk rates

J. Kendrick et al. / Atherosclerosis 205 (2009) 255–260

257

for the two CVD groups. Chi-square statistics were calculated for each table. Observations were weighted to reflect the general US population as of early 1990s, using weights calculated for that purpose by the National Health Statistics [27]. These analyses were followed by univariate logistic regression models to determine odds ratios (OR) for the unadjusted covariates and the main predictor (i.e., CVD). Backward stepping logistic regression modeling was performed to determine the ability of 25(OH)D, categorized into <20 ng/mL vs. higher in keeping with the definition for 25(OH)D deficiency [13], to predict prevalent CVD in the presence of a broad range of potential confounding variables (see above). The analysis was repeated categorizing serum 25(OH)D levels into <30 ng/mL vs. higher as a level of 30 ng/mL or greater can be considered to indicate sufficient vitamin D [13]. Of note, season of measurement was also included in the fully adjusted regression model to account for seasonal variations in NHANES III. Terms in the model with pvalues <0.05 were retained. Differences in the effect of 25(OH)D on prevalent CVD by race was explored in subset analyses and by tests of interaction. Analyses were conducted using SAS-callable SUDAAN statistical software (Research Triangle Institute, Research Triangle Park, NC), which takes into account the complex design of the survey [28].

Table 1 Age/gender-adjusted, weighted proportions of NHANES III participants with cardiovascular risk factors grouped according to presence/absence of clinically manifest cardiovascular disease (n = 16,603).

3. Results 3.1. Prevalence of traditional cardiovascular risk factors in participants with CVD Among 16,603 adult participants for whom self-reported CVD history was available, there were 1308 (8%) subjects with CVD, 681 (4.1%) of whom with a history of angina, 537 (3.2%) with a history of myocardial infarction and 309 (1.9%) with a history of stroke. Of note, some participants had more than one identifying CVD event. Table 1 shows the age- and gender-adjusted, weighted prevalence of cardiovascular risk factors among participants with and without clinical CVD. The prevalence of traditional risk factors was higher in those with CVD, including older age, obesity, smoking history, hypertension, diabetes, dyslipidemia (i.e., elevated LDL-C, hypertriglyceridemia or low HDL-C), and CKD. Non-Hispanic Whites had a higher proportion of CVD whereas Mexican–Americans had a higher proportion of subjects without CVD. In addition, participants with prevalent CVD tended to report a decreased physical activity when compared to persons without CVD.

Variable

2 p-value

No event (N = 15295)

Event (CVD) (N = 1308)

Gender (%) Females Males

51.9 48.1

52.2 47.8

0.8795

Age in years (%) 18–29 30–39 40–49 50–59 60–69 ≥70

27.7 23.8 18.6 11.7 9.9 8.3

8.8 11.2 14.2 14.5 20.7 30.6

<0.0001

Ethnicity (%) Non-Hispanic White Non-Hispanic Black Mexican–American Other

76.1 10.6 5.4 7.9

77.8 11.6 3.8 6.8

<0.0309

BMI (kg/m2 ) (%) <25 25–29 ≥30

46.8 32.0 21.2

37.2 33.2 29.6

Smoking status (%) Current Former Never

48.3 23.6 28.1

36.2 34.5 29.3

<0.0001

Physical activity score (%) 0 1 2 3 4 Hypertension (%) Diabetes (%) High LDL cholesterol (%) Low HDL cholesterol (%) High triglycerides (%) Chronic kidney disease (%) Vitamin D supplement (%) 25(OH)D deficiency (<20 ng/mL)

20.2 27.4 21.6 14.1 16.7 20.3 4.2 15.4 35.6 29.1 2.9 0.9 21.4

33.8 32.5 21.1 6.7 5.9 48.9 15.9 24.8 48.2 48.6 14.1 1.2 29.3

<0.0001

<0.0001

<0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.2667 <0.0001

Abbreviations: BMI = body mass index; CVD = cardiovascular disease; HDL = highdensity lipoprotein; LDL = low-density lipoprotein; 25(OH)D = 25-hydroxyvitamin D.

3.2. Cross-sectional relationship between 25(OH)D deficiency and prevalent cardiovascular disease The weighted mean serum 25(OH)D concentration was 29.6 ± 11.9 ng/mL (range: 4.0–160 ng/mL) for the whole population. The overall prevalence of 25(OH)D deficiency (<20 ng/mL) was 22%. Participants with CVD had a higher frequency of 25(OH)D deficiency (<20 ng/mL) than those without (29.3 vs. 21.4%, p < 0.0001); similarly, the mean serum 25(OH)D concentrations were significantly lower (p < 0.0001) in those with CVD (27.1 ± 9.7 ng/mL) than in those without (29.8 ± 12.2 ng/mL). In addition, Fig. 1 shows the age-adjusted prevalence of CVD across different serum levels of 25(OH)D. Notably, the increase in the prevalence of CVD with decreasing levels of 25(OH)D was linear (F-value = 8.14; p < 0.0001 for trend), suggesting a dose–effect relationship. At a 25(OH)D level greater or equal to 30 ng/mL, the age-adjusted prevalence of CVD was 6%, increasing to 10.5% for those participants with a 25(OH) D level <20 ng/mL. The age-adjusted prevalence of serum levels of 25(OH)D less than 20 ng/mL in participants with

Fig. 1. Age-adjusted prevalence of clinical cardiovascular disease by levels of 25hydroxyvitamin D.

258

J. Kendrick et al. / Atherosclerosis 205 (2009) 255–260

was categorized as <30 ng/mL vs. higher. Also in this case, there was an independent association between 25(OH)D deficiency (i.e., <30 ng/mL) and prevalent CVD [multiple-adjusted OR: 1.15 95% CI 1.01–1.31; p = 0.04]. 4. Discussion

Fig. 2. Age-adjusted prevalence of 25-hydroxyvitamin D deficiency (<20 ng/mL) in participants with clinical cardiovascular disease by gender.

Fig. 3. Age-adjusted prevalence of 25-hydroxyvitamin D deficiency (<20 ng/mL) in participants with clinical cardiovascular disease across race/ethnicity groups.

CVD across gender and racial/ethnic groups is demonstrated in Figs. 2 and 3. The prevalence of 25(OH)D deficiency was higher in females than in males, and racial/ethnic minorities, including non-Hispanic Black and Mexican–Americans. Of note, the average serum 25(OH)D levels among non-Hispanics Whites, nonHispanic Blacks, Mexican–Americans and those of other race were 31.8 ± 16.1, 19.4 ± 5.4, 25.2 ± 4.1 and 24.8 ± 12.9 ng/mL, respectively. Table 2 shows that 25(OH)D deficiency was associated with a 57% higher risk for prevalent CVD. After adjustment for age, gender, race/ethnicity, season of measurement, physical activity, BMI, smoking status, hypertension, diabetes, dyslipidemia, CKD and vitamin D use, participants with 25(OH)D deficiency had a 20% increased odds of CVD [multiple-adjusted OR:1.20 95% CI: 1.01–1.36; p = 0.03]. Testing for a race by 25(OH)D interaction revealed none (p = 0.4482). In the fully adjusted regression model, other independent predictors of prevalent CVD were age, male gender, hypertension, diabetes, smoking, low HDL-C, high triglycerides and CKD. Similar results were observed when 25(OH)D Table 2 Cross-sectional association between 25(OH)D deficiency and clinical cardiovascular disease (CVD) in the whole population (n = 16,603). Serum 25(OH)D Levels

Unadjusted OR (95% CI) Demographic-adjusted ORa (95% CI) Fully adjusted OR* (95% CI)

≥20 ng/mL (n = 12,947)

<20 ng/mL (n = 3656)

1.0 (REF) 1.0 (REF) 1.0 (REF)

1.57 (1.45, 1.68) 1.45(1.34, 1.58) 1.20 (1.01, 1.36)

OR, odds ratio; CI, confidence interval. a Adjusted for age, gender, race/ethnicity. * Adjusted for age, gender, race/ethnicity, season of measurement, physical activity score, body mass index, smoking status, hypertension, diabetes, dyslipidemia (i.e., high low-density lipoprotein cholesterol and triglycerides, low high-density lipoprotein cholesterol), chronic kidney disease and vitamin D use.

In the present study, we have shown that serum levels of 25(OH)D were below the recommended levels [13] for a large portion of the general adult population [i.e., 22% of participants had a serum 25(OH)D level <20 ng/mL], and that 25(OH)D deficiency was associated with a composite of self-reported angina, myocardial infarction and stroke, independently of several established risk factors, in a nationally representative sample of the US adult population. These findings converge with a growing body of evidence in suggesting that serum 25(OH)D levels may have clinical relevance in identifying patients with CVD [11,12,29,30]. Wang and colleagues have recently shown that low 25(OH)D levels (i.e., <15 ng/mL) are independently associated with incident CVD in 1739 Framingham Offspring Study participants followed for 5.4 years [11]. Similarly, other European studies in type 2 diabetic patients have shown that 25(OH)D levels are inversely associated with prevalent CVD [9] and carotid intima-medial thickening [10]. Despite the differences in the patient populations and definition of 25(OH)D deficiency in those studies compared to ours, we found similar results regarding the relationship of 25(OH)D deficiency and CVD. Furthermore, the Framingham Offspring study only included participants of White race whereas we included participants from different ethnic backgrounds. Thus, our findings extend previous observations in a representative sample of the US population. They further suggest that the pathophysiology that predisposes 25(OH)D deficiency to CVD in epidemiological studies became operative, even at 25(OH)D levels below 30 ng/mL. More recently, 25(OH)D deficiency has also been found to be associated with all-cause and cardiovascular mortality [29,30]. A recent study of White European patients undergoing coronary angiography found that participants in the lowest quartile of 25(OH)D levels had an increased risk of all-cause mortality and cardiovascular mortality after adjustment for traditional cardiovascular risk factors (hazard ratio 2.08, 95% CI 1.60–2.70 and hazard ratio 2.22, 95% CI 1.57–3.13 respectively) [29]. A similar relationship has been found in the general US population [30]. Melamed and colleagues evaluating NHANES III participants, found that 25(OH)D deficiency (<17.8 ng/mL) was associated with a 26% increased risk of all-cause mortality (hazard ratio 1.26, 95% CI 1.08–1.46) [30]. VDRs are ubiquitously distributed in the brain, skeletal muscle, pancreas, heart, blood vessels and immune cells, such as T and B lymphocytes and monocytes. The ability to generate the 1,25(OH)2 D in these cells is dependent upon the availability of the precursor molecule 25(OH)D derived from the circulating plasma and the availability of the converting enzyme 1-␣-hydroxylase. Locally synthesized 1,25(OH)2 D in these tissues binds to the intranuclear VDRs in an autocrine/paracrine manner [13,14]. The binding of 1,25(OH)2 D to the VDR has been found to (a) suppress the renin–angiotensin system [31,32], (b) promote or prevent apoptosis as required for tissue development [13], and (c) be required for cell proliferation and differentiation [13]. The ubiquitous availability of 1-␣-hydroxylase and VDRs, in addition to the wide scope of the 25(OH)D effects in extra-renal tissues, suggests that the autocrine/paracrine vitamin D system is primarily designed to fulfill the immediate local needs by coordinated local regulation [13,14], thus circumventing the dependence on the circulating 1,25(OH)2 D pool, which is regulated entirely by systemic calcium homeostatic factors.

J. Kendrick et al. / Atherosclerosis 205 (2009) 255–260

The biological mechanisms by which 25(OH)D might protect against the development and progression of CVD have not been fully delineated [31,32]. Previous studies have clearly demonstrated an inverse relation between serum 25(OH)D levels and important CVD risk factors, including hypertension, diabetes, obesity and hypertriglyceridemia [2,33,34]. Moreover, it has been reported a positive correlation of 25(OH)D concentration with insulin sensitivity and a negative effect of hypovitaminosis D (<20 ng/mL) on pancreatic beta cell function [8]. Not only is 25(OH)D deficiency associated with important CVD risk factors, but it has also been suggested to be a contributing factor in the pathogenesis of chronic heart failure, as lower 25(OH)D levels have been demonstrated in these patients [12,35]. In addition, several clinical and experimental studies demonstrated that 1,25(OH)2 D is one of the most potent negative endocrine regulators of the renin–angiotensin system, and suggested that the cardioprotective effects of 25(OH)D could be partly mediated by the renin–angiotensin system [7,31,32]. Indeed, the renin–angiotensin system plays a central role in the regulation of blood pressure, volume homeostasis, endothelial function, vascular remodeling and fibrinogenesis [36]. Finally, some studies documented that vitamin D3 can inhibit various aspects of inflammation [13,14]. Long-term vitamin D3 supplementation in vitamin D-deficient individuals markedly reduced plasma levels of C-reactive protein, tissue matrix-metalloproteinase and its inhibitors [35,37], and had beneficial effects on the elastic properties of the common carotid artery in postmenopausal women [38]. This study has some important limitations. First, because this study is cross-sectional, the present analysis is limited in its ability to establish causal or temporal relationships between 25(OH)D and clinical CVD. Second, CVD outcome was based on self-report, whereby participants were notified by their treating physician of the diagnosis, with potential for misclassification bias. Third, the secondary prevention after an initial CVD event might be successful in reducing traditional risk factors, thus attenuating the observed associations of risk factors with self-reported CVD. Several studies have addressed the reliability of self-reported data [39–42] and the results vary by patient population and study design. Researchers do agree that higher validity is noted when structured interviews or questionnaires are used and when shorter time frames are used for patients to recall data. NHANES III used structured methods for data collection which increases the validity of our results. Unfortunately, the data set does not provide information about the date of onset of CVD or any subsequent life style changes, both of which could have occurred long before the patient’s participation in the study. Finally, we were unable to evaluate any possible effects of parathyroid hormone and 1,25(OH)2 D, because these measurements were not performed in this study. Notwithstanding these limitations, this analysis has several important strengths. First, it is the largest and most comprehensive national survey to estimate the association between clinical CVD and 25(OH)D deficiency among race groups of the US adult population. Second, NHANES III used uniform methods to collect data on serum 25(OH)D levels and CVD. Third, the availability of extensive and complete data on a wide range of CVD risk factors allowed us to ensure giving an unbiased estimate between 25(OH)D deficiency and CVD. Finally, the design of NHANES III allows the results to be extrapolated to the entire US civilian non-institutionalized adult population. In summary, our study has shown that 25(OH)D deficiency is associated with prevalent CVD in a representative sample of the US adult population independently of numerous confounding factors. This association suggests that 25(OH)D is an important and underestimated risk factor for CVD, and that vitamin D replacement in persons with insufficient or deficient 25(OH)D levels might help

259

to decrease their risk of CVD. These findings may have important public health implications given the high prevalence of 25(OH)D deficiency in the general population and the safety and low cost of treating vitamin D deficiency. Clinicians should consider routinely testing for 25(OH)D as a part of preventive health care, especially in those at risk of CVD. Future interventional and experimental studies are needed to determine whether the correction of vitamin D deficiency could contribute to CVD prevention, and to identify potential mechanisms of any preventive effect of vitamin D replacement on the development of CVD. Disclosures None of the authors have any conflicts of interest. Acknowledgments The principal investigator had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Financial support: Amgen Fellowship Grant. References [1] Cooper R, Cutler J, Desvigne-Nickens P, et al. Trends and disparities in coronary heart disease, stroke, and other cardiovascular diseases in the United States: findings of the national conference on cardiovascular disease prevention. Circulation 2000;102:3137–47. [2] Martins D, Wolf M, Pan D, et al. Prevalence of cardiovascular risk factors and the serum levels of 25-hydroxyvitamin D in the United States: data from the Third National Health and Nutrition Examination Survey. Arch Intern Med 2007;167:1159–65. [3] Scragg R, Sowers M, Bell C. Serum 25-hydroxyvitamin D, ethnicity, and blood pressure in the Third National Health and Nutrition Examination Survey. Am J Hypertens 2007;20:713–9. [4] Scragg R, Sowers M, Bell C. Serum 25-hydroxyvitamin D, diabetes, and ethnicity in the Third National Health and Nutrition Examination Survey. Diabetes Care 2004;27:2813–8. [5] de Boer IH, Ioannou GN, Kestenbaum B, Brunzell JD, Weiss NS. 25Hydroxyvitamin D levels and albuminuria in the Third National Health and Nutrition Examination Survey (NHANES III). Am J Kidney Dis 2007;50:69–77. [6] Chonchol M, Scragg R. 25-Hydroxyvitamin D, insulin resistance, and kidney function in the Third National Health and Nutrition Examination Survey. Kidney Int 2007;71:134–9. [7] Krause R, Buhring M, Hopfenmuller W, Holick MF, Sharma AM. Ultraviolet B and blood pressure. Lancet 1998;352:709–10. [8] Chiu KC, Chu A, Go VL, Saad MF. Hypovitaminosis D is associated with insulin resistance and beta cell dysfunction. Am J Clin Nutr 2004;79:820–5. [9] Cigolini M, Iagulli MP, Miconi V, et al. Serum 25-hydroxyvitamin D3 concentrations and prevalence of cardiovascular disease among type 2 diabetic patients. Diabetes Care 2006;29:722–4. [10] Targher G, Bertolini L, Padovani R, et al. Serum 25-hydroxyvitamin D3 concentrations and carotid artery intima-media thickness among type 2 diabetic patients. Clin Endocrinol 2006;65:593–7. [11] Wang TJ, Pencina MJ, Booth SL, et al. Vitamin D deficiency and risk of cardiovascular disease. Circulation 2008;117:503–11. [12] Zittermann A, Schleithoff SS, Tenderich G, et al. Low vitamin D status: a contributing factor in the pathogenesis of congestive heart failure? J Am Coll Cardiol 2003;41:105–12. [13] Holick MF. Vitamin D deficiency. N Engl J Med 2007;357:266–81. [14] Dusso AS, Brown AJ, Slatopolsky E. Vitamin D. Am J Physiol Renal Physiol 2005;289:F8–28. [15] Ritter CS, Armbrecht HJ, Slatopolsky E, Brown AJ. 25-Hydroxyvitamin D(3) suppresses PTH synthesis and secretion by bovine parathyroid cells. Kidney Int 2006;70:654–9. [16] Stumpf WE, Sar M, Reid FA, Tanaka Y, DeLuca HF. Target cells for 1,25dihydroxyvitamin D3 in intestinal tract, stomach, kidney, skin, pituitary, and parathyroid. Science 1979;206:1188–90. [17] The Third National Health and Nutrition Examination Survey (NHANES III 198894) reference manuals and reports [book on CD-ROM]. Bethesda: National Center for Health Statistics; 2002. [18] National Kidney Foundation: K/DOQI Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification, and Stratification. Am J Kidney Dis. 2002; 39(Suppl. 1):S1–S266. [19] Rose GA. The diagnosis of ischemic heart pain and intermittent claudication in field surveys. Bull World Health Organ 1962;27:645–58.

260

J. Kendrick et al. / Atherosclerosis 205 (2009) 255–260

[20] Ford ES, Giles WH. Serum vitamins, carotenoids, and angina pectoris: findings from the National Health and Nutrition Examination Survey III. Ann Epidemiol 2000;10:106–16. [21] Sorlie PD, Cooper L, Schreiner PJ, et al. Repeatability and validity of the Rose questionnaire for angina pectoris in the Atherosclerosis Risk in Communities Study. J Clin Epidemiol 1996;49:719–25. [22] Graff-Iverson S, Selmer R, Lochen ML. Rose angina predicts 23-year coronary heart disease mortality in women and men aged 40–49 years. Heart 2008;94:482–6. [23] Bravata DM, Wells CK, Gulanski B, et al. Racial disparities in stroke risk factors: the impact of socioeconomic status. Stroke 2005;36:1507–11. [24] Simon JA, Hudes ES. Serum ascorbic acid and cardiovascular disease prevalence in U.S. adults: the Third National Health and Nutrition Examination Survey (NHANES III). Ann Epidemiol 1999;9:358–65. [25] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of lowdensity lipoprotein cholesterol in plasma without the use of the preparative ultracentrifuge. Clin Chem 1972;18:499–501. [26] Coresh J, Astor BC, Greene T, Eknoyan G, Levey AS. Prevalence of chronic kidney disease and decreased kidney function in the adult US population: Third National Health and Nutrition Examination Survey. Am J Kidney Dis 2003;41:1–12. [27] National Center for Health Statistics. Plan and operation of the Third National Health and Nutrition Examination Survey 1988–1994. Hyattsville, MD: National Center for Health Statistics; 1995. [28] Shah BV, Barnvell BG, Bieler GS. SUDAAN user’s manual version 7.5. Research Triangle Park, NC: Research Triangle Institute; 1997. [29] Dobnig H, Pilz S, Scharnagl H, et al. Independent association of low serum 25hydroxyvitamin D and 1,25-dihydroxyvitamin D levels with all-cause mortality and cardiovascular mortality. Arch Intern Med 2008;168:1340–9. [30] Melamed ML, Michos ED, Post W, Astor B. 25-hydroxyvitamin D levels and the risk of mortality in the general population. Arch Intern Med 2008;168:1629–37.

[31] Li YC. Vitamin D regulation of the renin–angiotensin system. J Cell Biochem 2003;88:327–31. [32] Li YC, Kong J, Wei M, et al. 1,25-Dihydroxyvitamin D3 is a negative endocrine regulator of the renin–angiotensin system. J Clin Invest 2002;110:229–38. [33] Zittermann A, Schleithoff SS, Koerfer R. Putting cardiovascular disease and vitamin D insufficiency into perspective. Br J Nutr 2005;94:483–92. [34] Rostand SG. Ultraviolet light may contribute to geographic and racial blood pressure differences. Hypertension 1997;30:150–6. [35] Schleithoff SS, Zittermann A, Tenderich G, et al. Vitamin D supplementation improves cytokine profiles in patients with congestive heart failure: a doubleblind, randomized, placebo-controlled trial. Am J Clin Nutr 2006;83:754–9. [36] Jacoby DS, Rader DJ. Renin–angiotensin system and atherothrombotic disease: from genes to treatment. Arch Intern Med 2003;163:1155–64. [37] Timms PM, Mannan N, Hitman GA, et al. Circulating MMP9, vitamin D and variation in the TIMP-1 response with VDR genotype: mechanisms for inflammatory damage in chronic disorders? QJM 2002;95:787–96. [38] Braam LA, Hoeks AP, Brouns F, et al. Beneficial effects of vitamin D and K on the elastic properties of the vessel wall in postmenopausal women: a follow-up study. Thromb Haemost 2004;91:373–80. [39] Bergmann MM, Byers T, Freedman DS, Mokdad A. Validity of self-reported diagnoses leading to hospitalization: a comparison of self-reports with hospital records in a prospective study of American adults. Am J Epidemiol 1998;147:969–77. [40] Walker MK, Whincup PH, Shaper AG, et al. Validation of patient recall of doctordiagnosed heart attack and stroke: a postal questionnaire and record review comparison. Am J Epidemiol 1998;148:355–61. [41] Lampe FC, Walker M, Lennon LT, et al. Validity of a self-reported history of doctor-diagnosed angina. J Clin Epidemiol 1999;52:73–81. [42] St. Sauver JL, Hagen PT, Cha SS, et al. Agreement between patient reports of cardiovascular disease and patient medical records. Mayo Clin Proc 2005;80:203–10.