Abstract

Background:

Populations with low vitamin D status, such as blacks living in the US or UK, have increased blood pressure (BP) compared with whites. We analyzed the association between serum 25-hydroxyvitamin D (25OHD) and BP to determine whether low 25OHD explains any of the increased BP in blacks.

Methods:

The Third US National Health and Nutrition Examination Survey (NHANES III) is a cross-sectional survey representative of the US civilian population during 1988 to 1994. Analyses were restricted to 12,644 people aged ≥20 years with measurements of BP and 25OHD, after excluding those on hypertensive medication.

Results:

Adjusted mean serum 25OHD was lowest in non-Hispanic blacks (49 nmol/L), intermediate in Mexican Americans (68 nmol/L), and highest in non-Hispanic whites (79 nmol/L). When participants were divided into 25OHD quintiles, mean (standard error) systolic BP was 3.0 (0.7) mm Hg lower (P = .0004) and diastolic BP was 1.6 (0.6) mm Hg lower (P = .011) for participants in the highest quintile (25OHD ≥85.7 nmol/L) compared with the lowest (25OHD ≤40.4 nmol/L), adjusting for age, sex, ethnicity, and physical activity. Further adjustment for body mass index (BMI) weakened the inverse association between 25OHD and BP, which remained significant for systolic BP (P < .05). The inverse association between 25OHD and systolic BP was stronger in participants aged ≥50 years than younger (P = .021). Ethnic differences in 25OHD explained about half of the increased hypertension prevalence in non-Hispanic blacks compared with whites.

Conclusions:

Vitamin D status, which is amenable to intervention by safely increasing sun exposure or vitamin D supplementation, was associated inversely with BP in a large sample representative of the US population. Am J Hypertens 2007;20: 713–719 © 2007 American Journal of Hypertension, Ltd.

Alterations in calcium metabolism are known to influence blood pressure (BP) regulation.1,2 Calciotropic hormones, including vitamin D, may have a role in this regulation.3,4 A receptor to 1,25-dihydroxyvitamin D has been described in smooth muscle tissue, supporting a potential role for vitamin D in the regulation of smooth muscle contraction, and therefore BP.5 A positive association between serum 25-hydroxyvitamin D (25OHD), which is increased in vitamin D deficiency, and BP has been reported in a US study.6

Ultraviolet B radiation, the main source of vitamin D, has been shown to lower BP in Poles7 and in Germans with mild untreated hypertension,8 although serum levels of sun-induced 25-hydroxyvitamin D3 were similar in newly detected hypertensive cases compared with matched controls in New Zealand.9

The possibility that vitamin D status is inversely related to BP may explain some of the well-known regional and ethnic variations in hypertension.10,11 Blood pressure is higher in American and British blacks compared with whites.12,13 These ethnic differences in BP are consistent with low levels of vitamin D in blacks,14 because of decreased skin synthesis secondary to increased skin pigmentation,15 compared with whites.

The recent National Health and Nutrition Examination Survey (NHANES III) measured serum vitamin D status and BP in a sample representative of the US population. A previous short report, limited to fasting participants attending the morning examination of NHANES III (n = 8421), failed to detect a significant association between serum 25OHD and adjusted odds ratios of hypertension, analyzed as one component of the metabolic syndrome.16 It is also unclear from this report whether participants with treated hypertension were excluded from statistical analyses. The current article analyzes a larger sample of NHANES III participants (both fasting and nonfasting) to increase statistical power, after excluding those on treatment for hypertension to avoid possible bias from treatment effects. The specific aims are to examine: (1) whether vitamin D is inversely related to both systolic and diastolic BP; and, if so, (2) whether ethnic differences in vitamin D status explain any ethnic difference in BP.

Methods

The NHANES III is a cross-sectional survey representative of the US civilian noninstitutionalized population carried out during 1988 to 1994 by the National Center for Health Statistics of the Centers for Disease Control and Prevention. A stratified, multistage sampling design was used to recruit participants from household clusters, with oversampling of non-Hispanic blacks and Mexican Americans. After an initial interview at home, participants visited mobile centers, where they had an extensive physical examination. Full details of all survey methods, including sampling, interview, examination, laboratory measurement of blood samples, ethical approval, and informed consent, have been published.17

Study Population

A total of 23,258 adults, aged ≥20 years, were invited to take part in the survey. Of these 18,825 were interviewed at home, 16,573 of whom attended mobile examination centers. In the home interview, information was collected on a wide range of variables including: age, sex, ethnicity (self-assigned as either non-Hispanic white, non-Hispanic black, Mexican American, other), past history of ever being told by a physician or other health professional of having hypertension, and the number of times a range of common physical activities was undertaken in leisure time during the previous month.17 Metabolic equivalents (MET) were assigned for each physical activity, and participants aged ≥60 years were classified as doing moderate or vigorous activities if the MET for any activity was ≥3.0 or ≥6.0, respectively, whereas those aged 20 to 59 years were similarly classified if the MET for any activity was ≥3.5 or ≥7.0, respectively.18

At the mobile examination centers, participants were dressed in underpants, disposable light clothing, and slippers while being weighed on electronic scales in kilograms, to two decimal places. Height was measured with a fixed stadiometer to the nearest millimeter.17 Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters. The date of the examination, by calendar month, was used to account for seasonal variation in sun exposure.

Blood pressure was measured at the mobile examination centers by physicians with mercury sphygmomanometers using a standard protocol.17 Up to three measurements were collected from each participant while in the sitting position, and if more than two measurements were collected, the last two were averaged. Systolic BP was defined as the point at which the first Korotkoff sound was heard; the diastolic BP was the level of mercury 2 mm below where the last sound was heard. Hypertension was defined as systolic >140 mm Hg or diastolic >90 mm Hg.12 Pulse pressure was calculated as the difference between systolic and diastolic pressures.19

Blood samples collected during the examination were centrifuged, aliquoted, and frozen to −70°C on site, and shipped on dry ice to central laboratories where they were stored at −70°C until analysis.17 Serum 25OHD was measured by a radioimmunoassay kit after extraction with acetonitrile (DiaSorin, Stillwater, MN) by the National Center for Environmental Health, CDC, Atlanta, GA. Serum 25OHD concentrations ranged from 8.7 to 243.6 nmol/L after excluding one person with a 25OHD value of 400.1 nmol/L.

Data in this report are restricted to non-Hispanic white, non-Hispanic black, and Mexican-American adults ≥20 years who attended the mobile examination centers (n = 12,644), after excluding those who were on current treatment for hypertension (n = 2649), had no serum 25OHD measurement (n = 664), had no BP measurement (n = 28), had no BMI measurement (n = 25), or were of “other” nationalities (n = 563).

Statistical analyses were carried out with SUDAAN (version 9.0.0; Research Triangle Park, NC), using the sampling weights for the mobile examination centers to adjust for oversampling of non-Hispanic blacks and Mexican Americans, and to correct standard errors for any design effect arising from clustered sampling. PROC REGRESS was used to calculate adjusted means and regression coefficients, whereas PROC CROSSTAB was used to calculate adjusted relative risks.

Results

Adjusted mean serum 25OHD concentrations varied between categories of demographic and lifestyle variables (Table 1). Vitamin D was higher in men than in women and declined with increasing age. With regard to ethnicity, vitamin D was lowest in non-Hispanic blacks, intermediate in Mexican Americans, and highest in non-Hispanic whites. Mean level of 25OHD decreased with increasing BMI quintile. Vitamin D concentrations were lowest in participants who did no leisure-time physical activity during the previous month, compared with those who were physically active, and increased in a stepwise fashion with increasing frequency of activity. The expected seasonal variation in 25OHD was also present, with levels being lowest in January to April and highest in July to October, after adjusting for covariates. Mean serum 25OHD was similar for participants who had ever been told by a physician they had hypertension, but were not on current antihypertensive medication, and those who had never been told they have hypertension, indicating that this variable was not related to vitamin D status. Hence, it was not adjusted for in further analyses.

Table 1

Mean (SE) serum 25-hydroxyvitamin D concentration (nmol/L), adjusted for all other variables

Variable Category Number 25-hydroxyvitamin D mean (SE) nmol/L P 
Sex Male 6097 78 (0.8) * 
 Female 6547 73 (0.8) <.0001 
Age (y) 20–29 3127 81 (1.1) * 
 30–39 2901 78 (1.2) .026 
 40–49 2128 73 (1.0) <.0001 
 50–59 1295 72 (0.7) <.0001 
 60–69 1434 70 (0.9) <.0001 
 ≥70 1759 67 (0.9) <.0001 
Race/Ethnicity non-Hispanic black 3479 49 (0.7) <.0001 
 Mexican American 3866 68 (0.9) <.0001 
 non-Hispanic white 5299 79 (0.7) * 
BMI quintile ≤22.1 2499 80 (0.7) * 
 22.2–24.6 2511 79 (1.2) .11 
 24.7–27.1 2528 75 (0.8) <.0001 
 27.2–30.6 2583 73 (1.0) <.0001 
 ≥30.7 2523 67 (0.8) <.0001 
Leisure-time physical activity (times in last month) None 2652 69 (0.9) * 
 Moderate <12 3514 73 (0.7) .004 
 Moderate ≥12 4334 78 (0.9) <.0001 
 Vigorous <12 1404 76 (1.4) <.0001 
 Vigorous ≥12 740 81 (1.6) <.0001 
Month of year Jan–Feb 2106 68 (2.1) <.0001 
 Mar–Apr 2481 68 (1.3) <.0001 
 May–Jun 2174 73 (1.7) <.0001 
 Jul–Aug 2045 82 (1.6) * 
 Sep–Oct 1930 81 (1.2) .87 
 Nov–Dec 1908 72 (1.3) <.0001 
Self-reported hypertension Yes 1731 76 (1.3) .95 
 No 10,913 75 (0.6) * 
 Total 12,644   
Variable Category Number 25-hydroxyvitamin D mean (SE) nmol/L P 
Sex Male 6097 78 (0.8) * 
 Female 6547 73 (0.8) <.0001 
Age (y) 20–29 3127 81 (1.1) * 
 30–39 2901 78 (1.2) .026 
 40–49 2128 73 (1.0) <.0001 
 50–59 1295 72 (0.7) <.0001 
 60–69 1434 70 (0.9) <.0001 
 ≥70 1759 67 (0.9) <.0001 
Race/Ethnicity non-Hispanic black 3479 49 (0.7) <.0001 
 Mexican American 3866 68 (0.9) <.0001 
 non-Hispanic white 5299 79 (0.7) * 
BMI quintile ≤22.1 2499 80 (0.7) * 
 22.2–24.6 2511 79 (1.2) .11 
 24.7–27.1 2528 75 (0.8) <.0001 
 27.2–30.6 2583 73 (1.0) <.0001 
 ≥30.7 2523 67 (0.8) <.0001 
Leisure-time physical activity (times in last month) None 2652 69 (0.9) * 
 Moderate <12 3514 73 (0.7) .004 
 Moderate ≥12 4334 78 (0.9) <.0001 
 Vigorous <12 1404 76 (1.4) <.0001 
 Vigorous ≥12 740 81 (1.6) <.0001 
Month of year Jan–Feb 2106 68 (2.1) <.0001 
 Mar–Apr 2481 68 (1.3) <.0001 
 May–Jun 2174 73 (1.7) <.0001 
 Jul–Aug 2045 82 (1.6) * 
 Sep–Oct 1930 81 (1.2) .87 
 Nov–Dec 1908 72 (1.3) <.0001 
Self-reported hypertension Yes 1731 76 (1.3) .95 
 No 10,913 75 (0.6) * 
 Total 12,644   
*

Reference category for P value.

Mean BP varied inversely with vitamin D status, with systolic, diastolic, and pulse pressure each being significantly (P < .05) lower in the highest quintile of serum 25OHD (≥85.7 nmol/L) compared with the lowest quintile (≤40.4 nmol/L) after adjusting for age, sex, ethnicity, and leisure-time physical activity (Table 2). However, further adjustment for BMI attenuated BP differences; therefore, only systolic BP and pulse pressure varied significantly (P < .05) between vitamin D quintiles (Table 2). In contrast, adding serum calcium to the model had little effect on the BP differences between vitamin D quintiles, indicating that the inverse association between serum 25OHD and BP is independent of serum calcium. Mean serum 25OHD levels did not vary (P = .09) between hypertensive cases (72.9 nmol/L) and controls (75.6 nmol/L), adjusting for age, sex, and ethnicity, indicating that cases and controls had overlapping 25OHD distributions.

Table 2

Adjusted mean (SE) blood pressure (mm Hg) by quintile of serum 25-hydroxyvitamin D

  Blood pressure (mm Hg): adjusted for sex, age, ethnicity, and leisure-time physical activity 
Vitamin D quintile (nmol/L) Mean (SE) Mean difference (SE) Mean difference (SE) also adjusted for BMI 
Systolic     
 ≤40.4 2545 122.2 (0.6) 
 40.5–53.9 2533 121.2 (0.4) −0.9 (0.6) −0.8 (0.6) 
 54.0–68.1 2516 120.0 (0.4) −2.2 (0.7) −1.8 (0.6) 
 68.2–85.6 2520 119.6 (0.4) −2.5 (0.6) −1.8 (0.7)* 
 ≥85.7 2530 119.1 (0.4) −3.0 (0.7) −1.8 (0.7)* 
P value (Wald F)   .0004 .045 
Diastolic     
 ≤40.4 2545 74.5 (0.5) 
 40.5–53.9 2533 74.1 (0.3) −0.4 (0.5) −0.3 (0.5) 
 54.0–68.1 2516 73.4 (0.3) −1.1 (0.6) −0.8 (0.6) 
 68. –85.6 2520 73.6 (0.3) −0.9 (0.6) −0.3 (0.6) 
 ≥85.7 2530 72.9 (0.3) −1.6 (0.6) −0.7 (0.5) 
P value (Wald F)   .011 .34 
Pulse pressure     
 ≤40.4 2545 47.7 (0.5) 
 40.5–53.9 2533 47.1 (0.4) −0.6 (0.5) −0.5 (0.5) 
 54.0–68.1 2516 46.6 (0.3) −1.1 (0.5)* −1.0 (0.5) 
 68.2–85.6 2520 46.0 (0.4) −1.7 (0.5) −1.5 (0.5) 
 ≥85.7 2530 46.3 (0.4) −1.4 (0.6)* −1.1 (0.6) 
P value (Wald F)   .012 .055 
  Blood pressure (mm Hg): adjusted for sex, age, ethnicity, and leisure-time physical activity 
Vitamin D quintile (nmol/L) Mean (SE) Mean difference (SE) Mean difference (SE) also adjusted for BMI 
Systolic     
 ≤40.4 2545 122.2 (0.6) 
 40.5–53.9 2533 121.2 (0.4) −0.9 (0.6) −0.8 (0.6) 
 54.0–68.1 2516 120.0 (0.4) −2.2 (0.7) −1.8 (0.6) 
 68.2–85.6 2520 119.6 (0.4) −2.5 (0.6) −1.8 (0.7)* 
 ≥85.7 2530 119.1 (0.4) −3.0 (0.7) −1.8 (0.7)* 
P value (Wald F)   .0004 .045 
Diastolic     
 ≤40.4 2545 74.5 (0.5) 
 40.5–53.9 2533 74.1 (0.3) −0.4 (0.5) −0.3 (0.5) 
 54.0–68.1 2516 73.4 (0.3) −1.1 (0.6) −0.8 (0.6) 
 68. –85.6 2520 73.6 (0.3) −0.9 (0.6) −0.3 (0.6) 
 ≥85.7 2530 72.9 (0.3) −1.6 (0.6) −0.7 (0.5) 
P value (Wald F)   .011 .34 
Pulse pressure     
 ≤40.4 2545 47.7 (0.5) 
 40.5–53.9 2533 47.1 (0.4) −0.6 (0.5) −0.5 (0.5) 
 54.0–68.1 2516 46.6 (0.3) −1.1 (0.5)* −1.0 (0.5) 
 68.2–85.6 2520 46.0 (0.4) −1.7 (0.5) −1.5 (0.5) 
 ≥85.7 2530 46.3 (0.4) −1.4 (0.6)* −1.1 (0.6) 
P value (Wald F)   .012 .055 
*

P < .05;

P < .01;

P < .001 compared with vitamin D quintile <40.4 nmol/L.

Possible effect modification of the relationship between BP and serum 25OHD was examined with multiple regression analyses using BP as a continuous variable. Inverse associations between BP and 25OHD existed in all ethnic groups. Coefficients (SE) from regressing BP (mm Hg) as the dependent variable against serum 25OHD (nmol/L), adjusting for age, sex, and leisure-time physical activity were: for systolic BP, −0.022 (P = .0017) in non-Hispanic whites, −0.024 (P = .09) in non-Hispanic blacks, and −0.031 (P = .0024) in Mexican Americans; and for diastolic BP, −0.015 (P = .0060) in non-Hispanic whites, −0.018 (P = .09) in non-Hispanic blacks, and −0.030 (P = .0004) in Mexican Americans.

However, the association between BP and serum 25OHD varied with age after adjusting for sex, ethnicity, and leisure-time physical activity (Table 3). Regression models were run with serum 25OHD and a product term of 25OHD times a dummy variable for age (≥50 years = 1, <50 years = 0). In these models, serum 25OHD was inversely associated with both systolic and diastolic BP but not with pulse pressure (P = .56). In addition, there were significant (P < .05) negative age interactions for systolic BP and pulse pressure, indicating that these two measures of BP decreased more with increasing 25OHD in people ≥50 years of age compared with younger participants. Calculations based on these coefficients indicate that increasing serum 25OHD from 20 to 100 nmol/L predicts, in people aged <50 and ≥50 years, respectively, a decrease of 1.8 mm Hg and 4.6 mm Hg in systolic BP, a decrease of 2.1 mm Hg and 0.8 mm Hg in diastolic BP, and an increase of 0.3 mm Hg and a decrease of 3.8 mm Hg in pulse pressure.

Table 3

Regression coefficients (β) for blood pressure (mm Hg) regressed against serum 25-hydroxyvitamin D (nmol/L) and interaction product term of age × 25-hydroxyvitamin D, adjusted for sex, ethnicity, and leisure-time physical activity

 25OHD* 25OHD x age 50 y 
Blood pressure (mm Hg) B (SE) P β (SE) P 
Systolic −0.023 (0.005) .0001 −0.034 (0.014) .021 
Diastolic −0.027 (0.005) <.0001 0.017 (0.009) .07 
Pulse pressure 0.004 (0.006) .56 −0.051 (0.013) .0003 
 25OHD* 25OHD x age 50 y 
Blood pressure (mm Hg) B (SE) P β (SE) P 
Systolic −0.023 (0.005) .0001 −0.034 (0.014) .021 
Diastolic −0.027 (0.005) <.0001 0.017 (0.009) .07 
Pulse pressure 0.004 (0.006) .56 −0.051 (0.013) .0003 
*

Serum 25-hydroxyvitamin D;

If age ≥50 years, then age 50 = 1; if age <50 years, then age 50 = 0.

Ethnic-specific mean BPs are shown in Table 4. Within each gender, non-Hispanic blacks had higher age-adjusted systolic and diastolic BPs than Mexican Americans and non-Hispanic whites, whereas Mexican Americans had BPs similar to non-Hispanic whites, except for a higher systolic BP in women. Mean pulse pressure was also higher in non-Hispanic blacks and Mexican Americans compared with non-Hispanic whites.

Table 4

Age-adjusted mean (SE) blood pressure (mm Hg) by sex and ethnicity

 Systolic pressure (mm Hg) Diastolic pressure (mm Hg) Pulse pressure (mm Hg) 
Demographic group Mean (SE) Mean difference (SE) Mean (SE) Mean difference (SE) Mean (SE) Mean difference (SE) 
Men       
• Non-Hispanic black (n = 1622) 125.9 (0.3) 3.2 (0.5) 77.4 (0.5) 1.5 (0.6)* 48.5 (0.3) 1.7 (0.5) 
• Mexican American (n = 1956) 123.5 (0.5) 0.8 (0.5) 75.8 (0.5) −0.1 (0.5) 47.7 (0.4) 0.9 (0.5) 
• Non-Hispanic white (n = 2519) 122.7 (0.4) 75.9 (0.3) 46.8 (0.4) 
Women       
• Non-Hispanic black (n = 1857) 120.5 (0.4) 3.9 (0.5) 72.7 (0.4) 1.8 (0.4) 47.8 (0.3) 2.1 (0.4) 
• Mexican American (n = 1910) 118.4 (0.3) 1.8 (0.4) 70.5 (0.4) −0.3 (0.4) 47.9 (0.3) 2.2 (0.4) 
• Non-Hispanic white (n = 2780) 116.6 (0.3) 70.9 (0.3) 45.7 (0.3) 
Both sexes       
• Non-Hispanic black (n = 3479) 123.1 (0.2) 3.5 (0.4) 75.0 (0.3) 1.6 (0.3) 48.1 (0.3 1.8 (0.3) 
• Mexican American (n = 3866) 121.0 (0.3) 1.4 (0.4) 73.1 (0.4) −0.2 (0.4) 47.8 (0.3) 1.6 (0.4) 
• Non-Hispanic white (n = 5299) 119.6 (0.3) 73.3 (0.2) 46.2 (0.3) 
 Systolic pressure (mm Hg) Diastolic pressure (mm Hg) Pulse pressure (mm Hg) 
Demographic group Mean (SE) Mean difference (SE) Mean (SE) Mean difference (SE) Mean (SE) Mean difference (SE) 
Men       
• Non-Hispanic black (n = 1622) 125.9 (0.3) 3.2 (0.5) 77.4 (0.5) 1.5 (0.6)* 48.5 (0.3) 1.7 (0.5) 
• Mexican American (n = 1956) 123.5 (0.5) 0.8 (0.5) 75.8 (0.5) −0.1 (0.5) 47.7 (0.4) 0.9 (0.5) 
• Non-Hispanic white (n = 2519) 122.7 (0.4) 75.9 (0.3) 46.8 (0.4) 
Women       
• Non-Hispanic black (n = 1857) 120.5 (0.4) 3.9 (0.5) 72.7 (0.4) 1.8 (0.4) 47.8 (0.3) 2.1 (0.4) 
• Mexican American (n = 1910) 118.4 (0.3) 1.8 (0.4) 70.5 (0.4) −0.3 (0.4) 47.9 (0.3) 2.2 (0.4) 
• Non-Hispanic white (n = 2780) 116.6 (0.3) 70.9 (0.3) 45.7 (0.3) 
Both sexes       
• Non-Hispanic black (n = 3479) 123.1 (0.2) 3.5 (0.4) 75.0 (0.3) 1.6 (0.3) 48.1 (0.3 1.8 (0.3) 
• Mexican American (n = 3866) 121.0 (0.3) 1.4 (0.4) 73.1 (0.4) −0.2 (0.4) 47.8 (0.3) 1.6 (0.4) 
• Non-Hispanic white (n = 5299) 119.6 (0.3) 73.3 (0.2) 46.2 (0.3) 

Both sexes: blood pressure also adjusted for sex.

*

P < .05;

P < .01;

P < .001 compared with non-Hispanic white.

The contribution of ethnic differences in vitamin D status to the ethnic variations in systolic BP is shown in Figs. 1 and 2. The mean difference in systolic BP was 3.5 mm Hg among non-Hispanic blacks and 1.4 mm Hg among Mexican Americans compared with non-Hispanic whites, after adjusting for age and sex. When quintile of vitamin D or BMI was added to the model, the mean difference in systolic BP among non-Hispanic blacks decreased more for 25OHD (to 2.1 mm Hg) than for BMI (to 2.7 mm Hg), indicating that ethnic variations in vitamin D explained more of the increased systolic BP in non-Hispanic blacks (about one-third) than ethnic differences in BMI, which explained about one-fifth. When both vitamin D and BMI were adjusted for, the mean difference (1.9 mm Hg) was similar to adjusting for vitamin D alone, indicating that their effects were not additive. The same pattern was seen for diastolic BP and pulse pressure (data not shown). In contrast, the addition of vitamin D and BMI to the model had similar effects on the difference in BP between Mexican Americans and non-Hispanic whites, and their effects were additive, as there was only a very small mean ethnic difference (0.1 mm Hg) after adjusting for both variables (Fig. 1).

Mean (95% CI) difference in systolic blood pressure for non-Hispanic blacks compared with non-Hispanic whites, adjusted for age, sex, serum 25-hydroxyvitamin D (25OHD), and body mass index (BMI).

Mean (95% CI) difference in systolic blood pressure for Mexican Americans compared with non-Hispanic whites, adjusted for age, sex, serum 25-hydroxyvitamin D (25OHD), and body mass index (BMI).

When the relative risk of hypertension compared with non-Hispanic whites was calculated, for non-Hispanic blacks it decreased from 1.93 (95% CI 1.55, 2.41), adjusting for age and sex, to 1.40 (95% CI 1.04, 1.89), adjusting also for vitamin D, and 1.66 (95% CI 1.34, 2.07), adjusting also for BMI. For Mexican Americans relative to non-Hispanic whites, the relative risk of hypertension decreased from 1.11 (95% CI 0.84, 1.46), adjusting for age and sex, to 0.98 (95% CI 0.72, 1.33), adjusting for vitamin D, and 0.97 (95% CI 0.73, 1.15), adjusting for BMI. This indicates that ethnic differences in vitamin D explain about half of the increased prevalence of hypertension in non-Hispanic blacks compared with non-Hispanic whites.

Discussion

These results from a nationally representative US sample show that systolic BP and pulse pressure are inversely associated with serum 25OHD. Although the differences we observed in BP between vitamin D quintiles are small, the random measurement error arising from measuring BP at a single interview, and from a single measurement of vitamin D status using a blood sample collected at the same interview, is likely to have resulted in attenuation of the observed association between these two variables.20 Moreover, the vitamin D–related BP differences reported in this article have public health significance, as a 2 to 3 mm Hg decrease in systolic BP would produce an approximate 10% to 15% decline in cardiovascular mortality.20

The inverse association between serum 25OHD and systolic BP has clinical significance, because the latter variable is a better predictor of coronary heart disease risk than diastolic BP, particularly in older people.21 The inverse association between serum 25OHD and pulse pressure suggests that vitamin D may lower systolic BP by increasing arterial compliance.19 This may occur through serum 25OHD–dependent autocrine production of 1,25-dihydroxyvitamin D in vascular smooth muscle cells, which inhibits smooth muscle cell growth,22 or through the influence of parathyroid hormone (which is increased by vitamin D deficiency),11 or by direct suppression of the renin-angiotensin system.3,23

The interpretation of the significance of our 25OHD findings depends on whether it is appropriate to include BMI as a confounder, or whether the latter is an intermediary in the causal pathway linking vitamin D and BP and therefore should not be adjusted when evaluating the full effect of vitamin D. The reduced serum 25OHD level for participants in the high BMI quintiles (Table 1) is probably due to sequestering of vitamin D within the increased fat mass of obese people.24 However, there is also evidence that low vitamin D status, by causing parathyroid hormone (PTH) excess and calcium influx into adipocytes, may promote weight gain.25

In addition to the possibility of random measurement error weakening the observed inverse association between serum 25OHD and BP, other limitations of this study include its cross-sectional design, which cannot separate cause and effect, and the possibility of other lifestyle variables (in addition to physical activity and obesity) associated with vitamin D status, confounding the association between vitamin D and BP.

Although our findings are consistent with previous studies of ultraviolet B radiation and BP,7,8 they contrast with some previous observational studies of dietary vitamin D, which provides only part of the supply to humans. The oral vitamin D studies have shown inconsistent results, with an inverse association being observed between systolic BP and oral intake of vitamin D in Iowa womens26 but not in a recent analysis of three US cohorts.27 Supplementation with a vitamin D analog (α-calcidol) has been shown to lower BP in patients with impaired glucose tolerance in Sweden,28 but supplementation with vitamin D3 did not lower BP in elderly men and women in England.29 The likely explanation for the failure of the latter study to lower BP may be insufficient change of vitamin D in the treated group. For example, the English study only increased 25OHD levels by 52% from 34 to 52 nmol/L.29 In contrast, 25OHD levels more than doubled from 25.7 to 64.8 nmol/L among the treated group in a German study, which observed a significant reduction in both systolic BP and pulse rate.30

Age- and gender-adjusted mean BPs were higher in non-Hispanic blacks than in Mexican Americans and non-Hispanic whites, consistent with results from the first phase (1988 to 1991) of NHANES III.13 The finding that ethnic differences in vitamin D status explained about half of the increased prevalence of hypertension in non-Hispanic blacks, compared with non-Hispanic whites, supports the previous suggestion that low vitamin D levels in non-Hispanic blacks may be a factor in their increased hypertension prevalence.10,11 Ethnic differences in BMI also made a small contribution to the increased risk of hypertension in both non-Hispanic blacks and Mexican Americans compared with non-Hispanic whites. Other lifestyle factors not analyzed in this article, such as low intake of potassium and high intake of sodium, may contribute to the increased BP levels in non-Hispanic blacks,31 whereas the contribution of genetic factors to ethnic differences is probably modest compared with the role of lifestyle.32

In summary, we have found increased systolic BP in people with low serum 25OHD levels in a representative US sample. This finding may have public health significance, as vitamin D levels can easily, and cheaply, be increased by a modest increase in sun exposure or vitamin D supplementation. However, first it needs to be confirmed by large, well-designed intervention studies.

We thank the Centers for Disease Control and Prevention, Hyattsville, MD, for making the data available for analysis. R.S. was responsible for drafting the text. R.S. and C.B. were responsible for data analysis. M.S. advised on the data analysis. All authors edited and critically reviewed the manuscript.

References

1.
Miller
GD
,
DiRienzo
DD
,
Reusser
ME
,
McCarron
DA
:
Benefits of dairy product consumption on blood pressure in humans: a summary of the biomedical literature
.
J Am Coll Nutr
 
2000
;
19
(
Suppl 2
):
147S
164S
.
2.
Brown
EM
:
The extracellular Ca++-sensing receptor: central mediator of systemic calcium homeostatis
.
Ann Rev Nutr
 
2000
;
20
:
507
533
.
3.
Resnick
LM
,
Muller
FB
,
Laragh
JH
:
Calcium-regulating hormones in essential hypertension
.
Ann Intern Med
 
1986
;
105
:
649
654
.
4.
Brickman
AS
,
Nyby
MD
,
von Hungen
K
,
Eggena
P
,
Tuck
ML
:
Calciotropic hormones, platelet calcium, and blood pressure in essential hypertension
.
Hypertension
 
1990
;
16
:
515
522
.
5.
Kawashima
H
:
Receptor for 1,25-dihydroxyvitamin D in a vascular smooth muscle cell line derived from rat aorta
.
Biochem Biophys Res Commun
 
1987
;
146
:
1
6
.
6.
Sowers
MR
,
Wallace
RB
,
Hollis
BW
,
Lemke
JH
:
Relationship between 1,25-dihydroxyvitamin D and blood pressure in a geographically defined population
.
Am J Clin Nutr
 
1988
;
48
:
1053
1056
.
7.
Kokot
F
,
Schmidt-Gayk
H
,
Wiecek
A
,
Mleczko
Z
,
Bracel
B
:
Influence of ultraviolet light on plasma vitamin D and calcitonin levels in humans
.
Kidney Int
 
1989
;
36
(
Suppl 27
):
S143
S146
.
8.
Krause
R
,
Buhring
M
,
Hopfenmuller
W
,
Holick
MF
,
Sharma
AM
:
Ultraviolet B and blood pressure
.
Lancet
 
1998
;
352
:
709
710
.
9.
Scragg
R
,
Holdaway
I
,
Singh
V
,
Metcalf
P
,
Baker
J
,
Dryson
E
:
Serum 25-hydroxyvitamin D3 levels in newly detected hypertension
.
Am J Hypertens
 
1995
;
8
:
429
432
.
10.
Scragg
R
:
Sunlight, vitamin D and cardiovascular disease
. In:
Crass
M.F.
,
Avioli
L.V.
(Eds).
Calcium-Regulating Hormones and Cardiovascular Function
 .
CRC Press
:
Boca Raton, FL
,
1994
.
213
237
.
11.
Rostand
SG
:
Ultraviolet light may contribute to geographic and racial blood pressure differences
.
Hypertension
 
1997
;
30
:
150
156
.
12.
Burt
VL
,
Whelton
P
,
Roccella
EJ
,
Brown
C
,
Cutler
JA
,
Higgins
M
,
Horan
MJ
,
Labarthe
D
:
Prevalence of hypertension in the adult US population: results from the third National Health and Nutrition Examination Survey, 1988–1991
.
Hypertension
 
1995
;
25
:
305
313
.
13.
Chaturvedi
N
,
McKeigue
PM
,
Marmot
MG
:
Resting and ambulatory blood pressure differences in Afro-Caribbeans and Europeans
.
Hypertension
 
1993
;
22
:
90
96
.
14.
Looker
AC
,
Dawson-Hughes
B
,
Calvo
MS
,
Gunter
EW
,
Sahyoun
NR
:
Serum 25-hydroxyvitamin D status of adolescents and adults in two seasonal subpopulations from NHANES III
.
Bone
 
2002
;
30
:
771
777
.
15.
Clemens
TL
,
Adams
JS
,
Henderson
SL
,
Holick
MF
:
Increased skin pigment reduces the capacity of skin to synthesize vitamin D3
.
Lancet
 
1982
;
1
:
74
76
.
16.
Ford
ES
,
Ajani
UA
,
McGuire
LC
,
Liu
S
:
Concentrations of serum vitamin D and the metabolic syndrome among US adults
.
Diabetes Care
 
2005
;
28
:
1228
1230
.
17.
National Center for Health Statistics
Third National Health and Nutrition Examination Survey, 1988–1994, reference manuals and reports (CD-ROM)
 .
Centers for Disease Control and Prevention
:
Hyattsville, MD
,
1996
.
18.
Crespo
CJ
,
Keteyian
SJ
,
Heath
GW
,
Sempos
CT
:
Leisure-time physical activity among US adults: results from the Third National Health and Nutrition Examination Survey
.
Arch Intern Med
 
1996
;
156
:
93
98
.
19.
Smulyan
H
,
Safar
ME
:
The diastolic blood pressure in systolic hypertension
.
Ann Intern Med
 
2000
;
132
:
233
237
.
20.
Prospective Studies Collaboration
Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies
.
Lancet
 
2002
;
360
:
1903
1913
.
21.
Chobanian
AV
,
Bakris
GL
,
Black
HR
,
Cushman
WC
,
Green
LA
,
Izzo
JL
,
Jones
DW
,
Materson
BJ
,
Oparil
S
,
Wright
JT
,
Roccella
EJ
:
The seventh report of the Joint National Committee on Prevention, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report
.
JAMA
 
2003
;
289
:
2560
2572
.
22.
Somjen
D
,
Weisman
Y
,
Kohen
F
,
Gayer
B
,
Limor
R
,
Sharon
O
,
Jaccard
N
,
Knoll
E
,
Stern
N
:
25-hydroxyvitamin D3-1α-hydroxylase is expressed in human vascular smooth muscle cells and is upregulated by parathyroid hormone and estrogenic compounds
.
Circulation
 
2005
;
111
:
1666
1671
.
23.
Li
YC
,
Kong
J
,
Wei
M
,
Chen
ZF
,
Liu
SQ
,
Cao
LP
:
1,25-dihydroxyvitamin D3 is a negative endocrine regulator of the renin-angiotensin system
.
J Clin Invest
 
2002
;
110
:
229
238
.
24.
Swordsman
J
,
Matsuoka
LY
,
Chen
TC
,
Lu
Z
,
Holick
MF
:
Decreased bioavailability of vitamin D in obesity
.
Am J Clin Nutr
 
2000
;
72
:
690
693
.
25.
McCarty
MF
,
Thomas
CA
:
PTH excess may promote weight gain by impeding catecholamine-induced lipolysis: implications for the impact of calcium, vitamin D, and alcohol on body weight
.
Med Hypotheses
 
2003
;
61
:
535
542
.
26.
Sowers
MF
,
Wallace
RB
,
Lemke
JH
:
The association of intakes of vitamin D and calcium with blood pressure among women
.
Am J Clin Nutr
 
1985
;
42
:
135
142
.
27.
Forman
JP
,
Bischoff-Ferrari
HA
,
Willett
WC
,
Stampfer
MJ
,
Curhan
GC
:
Vitamin D intake and risk of incident hypertension: results from three large prospective studies
.
Hypertension
 
2005
;
46
:
676
682
.
28.
Lind
L
,
Lithell
H
,
Skarfors
E
,
Wide
L
,
Ljunghall
S
:
Reduction of blood pressure by treatment with alphacalcidol
.
Acta Med Scand
 
1988
;
223
:
211
217
.
29.
Scragg
R
,
Khaw
K-T
,
Murphy
S
:
Effect of winter oral vitamin D3 supplementation on cardiovascular risk factors in elderly adults
.
Eur J Clin Nutr
 
1995
;
49
:
640
646
.
30.
Pfeifer
M
,
Begerow
B
,
Minne
HW
,
Nachtigall
D
,
Hansen
C
:
Effects of a short-term vitamin D3 and calcium supplementation on blood pressure and parathyroid hormone levels in elderly women
.
J Clin Endocrinol Metab
 
2001
;
86
:
1633
1637
.
31.
Douglas
JG
,
Bakris
GL
,
Epstein
M
,
Ferdinand
KC
,
Ferrario
C
,
Flack
JM
,
Jamerson
KA
,
Jones
WE
,
Haywood
J
,
Maxey
R
,
Ofili
EO
,
Saunders
E
,
Schiffrin
EL
,
Sica
DA
,
Sowers
JR
,
Vidt
DG
:
Management of high blood pressure in African Americans: consensus statement of the Hypertension in African Americans Working Group of the International Society on Hypertension in Blacks
.
Arch Intern Med
 
2003
;
163
:
525
541
.
32.
Tang
H
,
Jorgenson
E
,
Gadde
M
,
Kardia
SL
,
Rao
DC
,
Zhu
X
,
Schork
NJ
,
Hanis
CL
,
Risch
N
:
Racial admixture and its impact on BMI and blood pressure in African and Mexican Americans
.
Hum Genet
 
2006
;
119
:
624
633
.