Hypertension, a major modifiable risk factor for cardiovascular disease, is estimated to affect nearly one third of adults in the United States.1 Serum calcium plays an important pathophysiologic role in cardiovascular2 and kidney function.3 However, the association between serum calcium levels and hypertension is not clear. Previous studies that examined the association between serum calcium and hypertension have shown mixed results.4–11 While some studies reported a positive association between serum calcium and hypertension,6–9 others reported an inverse4 or no association.5,10,11 Further, the majority of the studies that examined the association between serum calcium and hypertension were either limited by small sample size9 or by inadequate adjustment for confounders, including serum albumin,5,6 glomerular filtration rate (GFR), and serum phosphorous.6,8,9 In this context, we examined the association between serum ionized calcium, total calcium, and hypertension in a nationally representative sample of adults in the United States after adjusting for important confounders.
J Clin Hypertens (Greenwich). 2011;13:716–721. ©2011 Wiley Periodicals, Inc.
Serum calcium levels have been shown to be associated with cardiovascular disease (CVD); however, it is not clear whether serum calcium levels are related to hypertension, a risk factor for CVD. The authors examined the association between serum calcium and hypertension in a representative sample of US adults. A cross-sectional study of 12,405 third National Health and Nutrition Examination Survey participants 20 years and older was conducted. Serum total and ionized calcium levels were analyzed as quartiles. The main outcome of interest was hypertension (n=3437), defined as self-reported use of antihypertensive medication and/or systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg. Elevated serum total calcium levels were positively associated with hypertension, independent of potential confounders including C-reactive protein, estimated glomerular filtration rate, serum albumin, 25(OH)D, and phosphorous. Compared with the lowest quartile of serum total calcium (referent category), the multivariable odds ratio (95% confidence interval) of hypertension was 1.49 (1.15–1.93) for the highest quartile (P=.005). This association persisted in subgroup analyses stratified by sex, age, and race-ethnicity. In contrast, serum ionized calcium levels were not associated with hypertension. Higher serum total calcium levels are positively associated with hypertension in a representative sample of US adults.
The data for the current study are derived from the third National Health and Nutrition Examination Survey (NHANES III). Detailed description of the complex survey design and methods have been published elsewhere and are available online.12 In brief, NHANES employed a stratified, multistage, probability sample representative of the civilian noninstitutionalized US population with oversampling of non-Hispanic blacks and Mexican Americans to ensure adequate representation of these groups. Eligible participants were required to sign an informed consent form and ethics approval was obtained from the Human Subjects Committee in the US Department of Health and Human Service.
Of the 15,951 participants 20 years and older in the current study, serum calcium levels were available in 13,910 participants. After excluding those with missing values of blood pressure (BP) (n=24), those with self-reported cardiovascular disease (CVD) (n=1148), and those with missing values for other covariates included in the multivariable analysis including education, body mass index (BMI), serum phosphorous, albumin, total and high-density lipoprotein (HDL) cholesterol, plasma glucose, and glycated hemoglobin (n=333), 12,405 were available for the current analysis.
Assessment of outcome
Three sets of BP measurements were taken using a mercury sphygmomanometer by a physician at the Medical Examination Center (MEC) according to the standardized BP measurement protocols as recommended by the American Heart Association. The average of the 3 measures was taken as the systolic and diastolic BP of the participant for the current analysis. Patients were considered to have hypertension if they reported using current BP-reducing medication and/or had systolic BP ≥140 mm Hg and/or diastolic BP ≥90 mm Hg.
Measurement of Exposure Variables
Information on participant’s demographic characteristics, educational attainment, cigarette smoking, alcohol consumption, physical activity, history of diabetes, hypertension, and medication use was assessed using a standardized questionnaire.12 Detailed description about the blood collection, processing, and quality-control checks are provided in the Laboratory Procedures Manual.13 Laboratory tests relevant to this analysis included serum total and ionized calcium, phosphorous, albumin, 25(OH) D, total and HDL cholesterol, glucose, glycated hemoglobin, creatinine, and C-reactive protein, and measurement of laboratory variables has been published before.14 Serum total calcium, phosphorous, and albumin were measured using a Hitachi 737 Analyzer (Boehringer Mannheim Diagnostics, Indianapolis, IN). Serum ionized calcium was measured using a NOVA 7+7 Electrolyte Analyzer (Nova Biomedical; Waltham, MA) adjusting the measured value for pH. Serum 25(OH) D was measured using a radioimmunoassay kit (Diasorin, Stillwater, MN) at the National Center for Environmental Health, CDC, Atlanta, GA. Diabetes was defined as a serum glucose ≥126 mg/dL after fasting for a minimum of 8 hours, a serum glucose ≥200 mg/dL for patients who fasted <8 hours before their NHANES visit, or self-reported current use of oral hypoglycemic medication or insulin. Serum creatinine was measured using a modified kinetic Jaffe reaction and GFR was estimated from serum creatinine using the Modification of Diet in Renal Disease equation.15
Serum total calcium levels were categorized into quartiles (<2.25, 2.25–2.30, 2.31–2.37, >2.37 mmol/L). We compared selected baseline characteristics of the study population by quartiles of total calcium using chi-square or analysis of variance as appropriate for the variable. We examined the association between quartiles of total calcium and hypertension in 3 multivariable-adjusted logistic regression models: model 1 adjusted for age (years) and sex (women, men); model 2 additionally adjusted for race-ethnicity (non-Hispanic whites, non-Hispanic blacks, Mexican Americans, others), smoking (never, former, current smoker), current alcohol intake (absent, present), BMI (kg/m2), physical inactivity (absent, present), diabetes mellitus (absent, present), serum total cholesterol (mg/dL), and HDL cholesterol (mg/dL); model 3 adjusted for variables in model 2 plus C-reactive protein (CRP) (mg/dL), estimated GFR (mL/min/1.73 m2), serum albumin (g/L), serum phosphorous (mg/dL), and serum 25(OH)D (ng/mL). We performed tests for linear trend by modeling total calcium quartiles as an ordinal variable in the corresponding multivariable logistic regression models. We also analyzed total calcium as a continuous variable (per standard deviation [SD] change). We then repeated the above analyses using serum ionized calcium quartiles (<1.21, 1.21–1.23, 1.24–1.26, >1.26 mmol/L). Since there was no significant association between serum ionized calcium levels and hypertension, we focused our subsequent analyses on total calcium levels. To examine the consistency of the association between total calcium and hypertension, we performed subgroup analyses stratified by age, sex, and race-ethnicity. Interactions were formally evaluated by including cross-product interaction terms in the corresponding multivariable models. All analyses were conducted by including sampling weights13 to account for unequal probabilities of selection, oversampling, and non-response using SUDAAN (version 8.0; Research Triangle Institute, Research Triangle Park, NC) and SAS (version 9.2; SAS Institute, Cary, NC) software. SEs were estimated using the Taylor series linearization method.
Table I shows the baseline characteristics of the study population. Compared with those in the lowest quartile of total calcium, those in the highest quartile were more likely to be younger, had higher levels of systolic and diastolic BP, total and HDL cholesterol, serum phosphorous, albumin, and 25(OH)D levels; were less likely to be female and non-Hispanic white; and had lower levels of BMI and CRP.
|Characteristics||Quartile 1 (n=3167)||Quartile 2 (n=2816)||Quartile 3 (n=3306)||Quartile 4 (n=3116)||P Valuea|
|Unweighted sample size|
|Smoking categories, %||.1985|
|Current drinker, %||53.68±1.96||53.24±1.81||56.70±1.72||55.53±1.74||.2121|
|Diabetes mellitus, %||5.48±0.63||6.01±0.64||5.37±0.57||5.98±0.57||.7784|
|Physical inactivity, %||49.13±1.46||46.66±1.64||46.65±1.59||49.47±1.75||.3435|
|Body mass index, kg/m2||26.53±0.15||26.44±0.19||26.61±0.19||26.34±0.15||.0293|
|Systolic blood pressure, mm Hg||120.98±0.68||120.88±0.63||121.66±0.51||122.71±0.58||.0008|
|Diastolic blood pressure, mm Hg||73.18±0.31||73.95±0.29||74.62±0.30||75.07±0.34||<.0001|
|Total cholesterol, mg/dL||197.83±1.17||201.82±1.33||203.46±1.30||209.13±1.16||<.0001|
|HDL cholesterol, mg/dL||50.64±0.47||50.81±0.54||50.82±0.57||51.06±0.45||.0003|
|Serum phosphorous, mg/dL||3.39±0.01||3.41±0.02||3.46±0.02||3.51±0.02||<.0001|
|Serum albumin, g/L||4.02±0.03||4.12±0.02||4.25±0.02||4.34±0.02||<.0001|
|Serum C-reactive protein, mg/dL||0.46±0.02||0.40±0.01||0.37±0.01||0.36±0.02||<.0001|
|Estimated GFR, mL/min/1.73m2||97.38±0.92||93.80±0.77||93.56±0.63||94.73±0.82||.08|
|Serum 25(OH) D, ng/mL||28.32±0.38||29.38±0.51||30.20±0.51||30.17±0.45||<.0001|
Table II shows the association between serum total calcium, ionized calcium levels, and hypertension. Increasing categories of total calcium were positively associated with hypertension in all 3 regression models (P<.01). The positive association between total calcium and hypertension persisted when total calcium was analyzed as a continuous variable (per SD increase). In contrast, no statistically significant association was observed between serum ionized calcium and hypertension in any model.
|Cases, No.||Age- and Sex-Adjusted OR (95% CI)||Multivariate Model 1 OR (95% CI)a||Multivariate Model 2 OR (95% CI)b|
|Ionized calcium, mmol/L|
|Quartile 1 (<1.21 mmol/L)||2845 (836)||1.00 (Referent)||1.00 (Referent)||1.00 (Referent)|
|Quartile 2 (1.21–1.23 mmol/L)||3170 (856)||0.87 (0.69–1.08)||0.87 (0.69–1.09)||0.88 (0.70–1.10)|
|Quartile 3 (1.24–1.26 mmol/L)||3360 (855)||0.96 (0.79–1.17)||0.92 (0.76–1.13)||0.93 (0.75–1.14)|
|Quartile 4 (>1.26 mmol/L)||3030 (890)||1.02 (0.80–1.32)||0.98 (0.75–1.28)||0.97 (0.74–1.27)|
|Per SD increase in ionized calcium||12,405 (3437)||1.02 (0.93–1.10)||1.00 (0.92–1.09)||0.99 (0.91–1.08)|
|Total calcium, mmol/L|
|Quartile 1 (<2.25 mmol/L)||3167 (866)||1.00 (Referent)||1.00 (Referent)||1.00 (Referent)|
|Quartile 2 (2.25–2.30 mmol/L)||2816 (766)||1.20 (0.95–1.51)||1.13 (0.90–1.42)||1.13 (0.90–1.41)|
|Quartile 3 (2.31–2.37 mmol/L)||3306 (861)||1.30 (1.06–1.59)||1.17 (0.97–1.42)||1.15 (0.95–1.40)|
|Quartile 4 (>2.37 mmol/L)||3116 (944)||1.69 (1.33–2.13)||1.52 (1.19–1.94)||1.49 (1.15–1.93)|
|Per SD increase in total calcium||12,405 (3437)||1.20 (1.10–1.32)||1.16 (1.07–1.27)||1.15 (1.05–1.26)|
In subgroup analyses, the positive association between total calcium and hypertension was consistently present in men and women (Table III), those 60 years and older, and those younger than 60 years (Table IV). When stratified by race-ethnicity (Table V), the association was stronger in non-Hispanic whites and non-Hispanic blacks but not significant among other race-ethnicities. There was no interaction by sex (P=.55), age (P=.43), or race-ethnicity (P=0.26) in the association between total calcium and hypertension.
|Total Calcium, mmol/L||Cases, No.||Age- and Sex-Adjusted OR (95% CI)||Multivariate Model 1 OR (95% CI)a||Multivariate Model 2 OR (95% CI)b|
|Quartile 1 (<2.25 mmol/L)||1334 (442)||1.00 (Referent)||1.00 (Referent)||1.00 (Referent)|
|Quartile 2 (2.25–2.30 mmol/L)||1235 (383)||1.21 (0.82–1.79)||1.12 (0.75–1.68)||1.11 (0.75–1.65)|
|Quartile 3 (2.31–2.37 mmol/L)||1582 (414)||1.22 (0.85–1.76)||1.08 (0.75–1.55)||1.06 (0.74–1.52)|
|Quartile 4 (>2.37 mmol/L)||1563 (388)||1.53 (1.08–2.18)||1.39 (0.95–2.04)||1.36 (0.93–2.00)|
|Quartile 1 (<2.25 mmol/L)||1833 (424)||1.00 (Referent)||1.00 (Referent)||1.00 (Referent)|
|Quartile 2 (2.25–2.30 mmol/L)||1581 (383)||1.15 (0.85–1.56)||1.11 (0.80–1.52)||1.13 (0.83–1.54)|
|Quartile 3 (2.31–2.37 mmol/L)||1724 (447)||1.26 (0.92–1.75)||1.18 (0.87–1.61)||1.21 (0.90–1.62)|
|Quartile 4 (>2.37 mmol/L)||1553 (556)||1.64 (1.23–2.18)||1.48 (1.10–1.98)||1.50 (1.10–2.04)|
|Total Calcium, mmol/L||Cases, No.||Age- and Sex-Adjusted OR (95% CI)||Multivariate Model 1 OR (95% CI)a||Multivariate Model 2 OR (95% CI)b|
|Age ≥60 y|
|Quartile 1 (<2.25 mmol/L)||1000 (557)||1.00 (Referent)||1.00 (Referent)||1.00 (Referent)|
|Quartile 2 (2.25–2.30 mmol/L)||793 (470)||1.21 (0.89–1.64)||1.14 (0.82–1.58)||1.13 (0.82–1.56)|
|Quartile 3 (2.31–2.37 mmol/L)||809 (489)||1.23 (0.95–1.61)||1.18 (0.91–1.54)||1.15 (0.89–1.48)|
|Quartile 4 (>2.37 mmol/L)||812 (555)||1.88 (1.49–2.38)||1.73 (1.34–2.22)||1.65 (1.27–2.14)|
|Age <60 y|
|Quartile 1 (<2.25 mmol/L)||2167 (309)||1.00 (Referent)||1.00 (Referent)||1.00 (Referent)|
|Quartile 2 (2.25–2.30 mmol/L)||2023 (296)||1.16 (0.82–1.62)||1.11 (0.80–1.55)||1.10 (0.79–1.52)|
|Quartile 3 (2.31–2.37 mmol/L)||2497 (372)||1.24 (0.94–1.64)||1.09 (0.83–1.43)||1.09 (0.84–1.41)|
|Quartile 4 (>2.37 mmol/L)||2304 (389)||1.49 (1.04–2.12)||1.36 (0.95–1.95)||1.34 (0.93–1.94)|
|Total Calcium, mmol/L||Cases, No.||Age- and Sex- Adjusted OR (95% CI)||Multivariate Model 1 OR (95% CI)a||Multivariate Model 2 OR (95% CI)b|
|Quartile 1 (<2.25 mmol/L)||1323 (420)||1.00 (Referent)||1.00 (Referent)||1.00 (Referent)|
|Quartile 2 (2.25–2.30 mmol/L)||1188 (376)||1.28 (0.94–1.75)||1.20 (0.89–1.62)||1.19 (0.88–1.60)|
|Quartile 3 (2.31–2.37 mmol/L)||1370 (423)||1.43 (1.11–1.84)||1.27 (1.00–1.59)||1.23 (0.97–1.55)|
|Quartile 4 (>2.37 mmol/L)||1163 (418)||1.79 (1.31–2.45)||1.59 (1.16–2.18)||1.51 (1.09–2.09)|
|Quartile 1 (<2.25 mmol/L)||743 (228)||1.00 (Referent)||1.00 (Referent)||1.00 (Referent)|
|Quartile 2 (2.25–2.30 mmol/L)||713 (216)||1.24 (0.95–1.63)||1.22 (0.92–1.63)||1.21 (0.91–1.60)|
|Quartile 3 (2.31–2.37 mmol/L)||910 (264)||1.07 (0.80–1.43)||1.03 (0.75–1.41)||1.00 (0.76–1.32)|
|Quartile 4 (>2.37 mmol/L)||1071 (342)||1.46 (1.15–1.87)||1.37 (1.07–1.75)||1.32 (1.03–1.70)|
|Quartile 1 (<2.25 mmol/L)||1101 (218)||1.00 (Referent)||1.00 (Referent)||1.00 (Referent)|
|Quartile 2 (2.25–2.30 mmol/L)||915 (174)||0.66 (0.39–1.12)||0.60 (0.35–1.02)||0.57 (0.33–0.98)|
|Quartile 3 (2.31–2.37 mmol/L)||1026 (174)||0.73 (0.43–1.24)||0.70 (0.41–1.20)||0.63 (0.36–1.09)|
|Quartile 4 (>2.37 mmol/L)||882 (184)||1.04 (0.60–1.81)||1.00 (0.57–1.76)||0.85 (0.47–1.54)|
Higher serum total calcium levels were found to be positively associated with hypertension in a representative sample of US adults without CVD. This association was independent of age, sex, race-ethnicity, smoking, alcohol intake, BMI, physical activity, diabetes mellitus, total cholesterol, HDL cholesterol, CRP, estimated GFR, serum albumin, serum vitamin D, and serum phosphorous. Further, the positive association between serum total calcium and hypertension was consistently present in subgroups of sex, age, and race-ethnicity and was persistent when serum calcium was analyzed as a continuous variable. In contrast, serum ionized calcium was not significantly associated with hypertension.
Serum total calcium is the total sum of 3 forms, ionized or free, protein-bound and soluble form complexed with anions such as bicarbonate and phosphate. Around 50% of total serum calcium is in the ionized form, 40% in the bound form mainly to albumin, and 10% bound to anions.16 Ionized calcium, the physiologically active form in the blood, is an accurate indicator of calcium homeostasis.17 However, ionized calcium measurement is more expensive and is affected by a number of factors including method of collection, choice of anticoagulant, pH changes, and variability of reference range used by different laboratories using different analyzers.18 Total calcium measurement is least affected by these changes, correlates well with ionized calcium measurement, and is routinely used in clinical practice to assess calcium status in health and disease. Our study was unique in that we examined the association between both serum total and ionized calcium and hypertension separately.
In the current study, we found that serum total calcium levels were positively associated with hypertension, whereas ionized calcium levels were not. The magnitude of association between serum total calcium and hypertension, its persistence after multivariate adjustment for potential confounders, and the consistency of the association in subgroups of sex, age, and race-ethnicity render support to the validity of our study findings. The finding of a null association between serum ionized calcium and hypertension was also similarly robust.
An association between serum total calcium and hypertension is plausible, including a direct effect on vasculature by enhanced vascular resistance,19 alteration in extracellular binding of calcium,20 interaction between serum calcium, and other cations such as sodium, potassium and magnesium,21 renal vasoconstriction causing kidney dysfunction,22 and hyperactivity of renin-angiotensin system from hyperparathyroidism.23 Animal studies have demonstrated elevated total calcium concentrations but normal ionized calcium levels in spontaneously hypertensive rats.24 Treatment of hypertensive patients with calcium channel blockers lower BP by inhibiting transmembrane transport of calcium through membrane channels.25
Our finding of a positive association between elevated serum total calcium levels and hypertension is consistent with several previous studies.1,6–34 Jorde and colleagues, in a large cross-sectional study of 12,865 men and 14,293 women in Norway, reported that serum total calcium was positively associated with hypertension in both men and women after adjusting for age, BMI, heart rate, and cholesterol.6 Kesteloot and associates, in a cross-sectional study of 4167 men and 3891 women in Belgium, documented that higher serum total calcium was positively associated with hypertension in both men and women after adjusting for serum creatinine and other confounders.7 Philips and colleagues,8 in a cross-sectional study involving 7735 healthy middle-aged British men, reported that higher serum total calcium was associated with hypertension even after adjusting for serum albumin. Rinner and coworkers,9 in a small sample of 182 Dutch adults, reported that higher serum total calcium was positively associated with hypertension after adjusting for age, BMI, and albumin and in the subgroup analysis stratified by sex, the association was found to be stronger in women than in men. In our study, consistent with the majority of literature in this field,6,7,26 we found that higher serum total calcium levels were positively associated with hypertension in both men and women. In contrast to serum calcium, dietary calcium has been shown to be inversely associated with hypertension.27–30
In the current study, we found no association between serum ionized calcium and hypertension. This is in keeping with 3 previous reports that suggested a similar null association between serum ionized calcium levels and hypertension, including the study by Buckley and colleagues31 in 325 male industrial workers; the study by Andersen and associates in 70 men and women from Denmark32; and the study by Resnik and colleagues33 in a clinical sample of 200 normotensive and hypertensive patients. However, in contrast to our findings, some previous studies have reported an inverse association between serum ionized calcium and hypertension. Folsom and coworkers4 reported an inverse association between serum ionized calcium and hypertension in a small sample of 56 patients in Minnesota. Vargas and coworkers,11 in a previous study using NHANES III data, reported an inverse association between serum ionized calcium and hypertension in the subgroup of younger Mexican-American men only in the dataset. Finally, there is some evidence to suggest that the association between ionized calcium and hypertension may vary by plasma renin levels. Hunt and colleagues,34 in a study involving 875 normotensive patients observed a positive association between ionized calcium and BP among the high renin group and an inverse association between ionized calcium and BP among the low renin group. This differential relationship of ionized calcium with BP across categories of plasma renin activity suggests that the association between ionized calcium and hypertension is modified by plasma renin and that ionized calcium in isolation may not be independently related to BP.30,34 In the current study, we did not have measurements of plasma renin levels to validate or disprove this hypothesis.
The major advantages of our study include the large sample size, vigorous methodology, and availability of both ionized and total calcium measurements. The cross-sectional nature of the study limits making causal inferences. Further, bias due to unmeasured confounding, for example, plasma renin activity, could not be excluded.
Elevated total calcium levels were found to be positively associated with hypertension in a nationally representative sample of US adults.
Acknowledgments and disclosures: This study was partially funded by an American Heart Association National Clinical Research Program grant (AS) and NIH/NIEHS grant 1R03ES018888-02(AS). There are no conflicts of interest related to this manuscript.