VITAMIN D AND THE VASCULAR SENSITIVITY TO ANGIOTENSIN II IN OBESE CAUCASIANS WITH HYPERTENSION

Abstract

Obesity and vitamin D deficiency have both been linked to augmented activity of the tissue renin-angiotensin system (RAS). We investigated whether obesity status influenced the relationship between 25-hydroxyvitamin D [25(OH)D] and vascular RAS activity. 25(OH)D levels were measured in hypertensive obese (n=39) and hypertensive non-obese (n=58) Caucasian individuals. RAS activity was assessed via plasma renin activity, and evaluation of the vascular sensitivity to angiotensin II (AngII) using the mean arterial pressure (MAP) response to an infusion of AngII. Among obese subjects, 25(OH)D was an independent positive predictor of the MAP response to AngII (β = 0.70, r = 0.41, p < 0.01); lower 25(OH)D concentrations were associated with a blunted MAP response to AngII. In contrast, 25(OH)D did not significantly predict the vascular sensitivity to AngII in non-obese subjects (β = 0.10, r = 0.07, p = 0.62). A multivariable adjusted interaction model confirmed that the positive relationship between 25(OH)D and the vascular sensitivity to AngII strengthened with obesity (p-interaction = 0.03). These findings demonstrate a positive association between 25(OH)D and the vascular sensitivity to AngII in obese hypertensives, and further suggest that vascular RAS activity may progressively increase when 25(OH)D deficiency occurs in obesity. Future studies to evaluate the effect of vitamin D supplementation on vascular RAS activity in obesity are needed.

INTRODUCTION

Obesity and 25-hydroxyvitamin D [25(OH)D] deficiency are increasingly common epidemiologic scenarios that have both been implicated with dysregulated control of the renin-angiotensin system (RAS) ^1–4.

Obesity is considered to be a state of high RAS activity; adipocytes have been shown to produce all of the components of the RAS locally ^5–11. Excess activity of the RAS results in increased salt sensitivity, volume expansion, hypertension, and insulin resistance – all unfavorable characteristics of the obese state ^{10, 12–14}. Thus, hyperactivity of the RAS in obesity may represent a crucial target for cardiovascular risk modification.

Animal studies have strongly suggested that vitamin D negatively regulates the RAS by inhibiting renin. Mice lacking the vitamin D receptor or 1α-hydroxylase enzyme develop a phenotype of increased RAS activity with resultant hypertension (HTN) and cardiac hypertrophy, which are improved with RAS antagonism and vitamin D agonist therapy ^15–19. A similar inverse association between 1,25(OH)₂D and plasma renin activity (PRA) was described by Resnick et al. and Tomaschitz et al. in humans ^{20, 21}. We recently observed an inverse association between 25(OH)D and the renal vascular sensitivity to angiotensin II (AngII) in humans²², linking 25(OH)D deficiency with increased activity of the renal vascular RAS. Prior investigations have used the renal vascular and blood pressure responses to exogenous AngII as an inversely proportional measure of vascular RAS activity ^23–29.

Though 25(OH)D deficiency has been associated with both obesity and increased tissue RAS activity ^30–33, the influence of obesity on the relationship between 25(OH)D and RAS regulation has not been described. Given the heightened tissue RAS activity in obesity and hypertension, we hypothesized that the relationship between 25(OH)D and tissue RAS activity would be notably pronounced in obese subjects with hypertension, when compared to non-obese subjects with hypertension.

METHODS

Participants

We performed a retrospective analysis of hypertensive subjects previously studied in the HyperPath cohort. The HyperPath Project was designed to characterize the physiology and genetic underpinnings of cardiovascular disease in an international cohort of patients. Participants were studied at four collaborating centers: Brigham and Women’s Hospital (Boston, MA), University of Utah Medical Center (Salt Lake City, UT), Vanderbilt University Hospital (Nashville, TN), and Hopital European Georges Pompidou (Paris, France). Study protocols were approved by the Human Subjects Committees at each site, and informed written consent was obtained from each subject.

HTN was defined as an untreated seated diastolic blood pressure (DBP) > 100 mmHg, a DBP > 90 mmHg with one or more antihypertensive medications, or the use of two or more antihypertensive medications. All blood pressure measurements reflect the average of three readings with standard manual mercury sphygmomanometer. Anthropometric measurements were obtained upon admission to each center’s Clinical Research Center. Exclusion criteria included chronic kidney disease, coronary heart disease, heart failure, suggested or known causes of secondary hypertension, and active malignancy. The full scope of inclusion and exclusion criteria of this population have been described previously ^{34, 35}.

At the time of this retrospective analysis, 1248 subjects had been studied in the HyperPath cohort. 810 of these subjects had HTN, and 345 of these hypertensive subjects had available frozen plasma for further analysis. To avert erroneous interpretations of the RAS and its biologic effect, only subjects known to have normal RAS physiology were considered for inclusion. Based on prior dynamic phenotyping of these 345 subjects, 164 were previously classified as hypertensives who demonstrated normal circulating RAS physiology in response to sodium restriction and exogenous AngII infusion defined as: PRA elevation > 2.4 ng/mL/hr with upright posture and an aldosterone increase of > 15 ng/dL with AngII administration. This approach of studying an isolated phenotype of hypertension limits misinterpretations attributable to the heterogeneity of RAS regulation among subjects, especially when studying hypothesized modifiers of RAS such as obesity, vitamin D, and HTN. A total of 97 of the 164 subjects completed the AngII infusion protocol (below) and had AngII-stimulated MAP information available for analysis.

All 97 subjects were of Caucasian race. Individuals with body mass index (BMI) ≥ 30 kg/m² were considered “obese,” whereas all individuals with BMI < 30 kg/m² were considered “non-obese.” This dichotomization of obesity status was pursued over further stratifications of BMI as only 16 of the study participants had a lean BMI of < 25 kg/m².

Study Protocol

To minimize interference with RAS assessment, participants taking angiotensin converting enzyme inhibitors, angiotensin receptor blockers, or mineralocorticoid receptor antagonists were transitioned to amlodipine and/or hydrochlorothiazide 3 months prior to study initiation. Two weeks prior to study initiation, all anti-hypertensive medications were withdrawn. Subjects were maintained on a fixed potassium (80 mmol/day), calcium (1000mg/day), and high sodium (200 mEq/24h) diet for 4–7 days leading to study initiation. This dietary sodium content approximates the daily salt consumption in Western countries ^{36, 37}. External sodium balance and diet compliance was confirmed with a 24-hr urine sodium excretion of ≥ 150 mmol.

Subjects were admitted to the Clinical Research Center and maintained in supine position for 1 night and 1 day. Baseline mean arterial pressure (MAP) was determined while supine between the hours of 8:00 AM and 10:00 AM, following 10 hours of overnight rest using the average of five readings from a Dinamap automated device (Critikon, Tampa, FL). All subjects underwent infusion with AngII with 3 ng/kg/min for 55 minutes, with MAP readings recorded every 2 minutes. The AngII-stimulated MAP was defined as the average of 5 readings in the last 10 minutes of the infusion, and used as a measure for intrinsic systemic vascular RAS activity. Baseline blood sampling was also obtained in the morning following overnight rest, collected on ice and kept frozen until assayed.

Biochemical Assessments

Blood samples collected during the study protocol were used to measure electrolytes and PRA. PRA was measured as previously described ³⁸. 25(OH)D was measured from the baseline frozen blood samples obtained on the original day of study using the DiaSorin, Inc. radioimmunoassay (Stillwater, MN, USA). This assay has a sensitivity of 4 ng/mL with a coefficient of variation ranging from 4.4–8.4%. Since 25(OH)D was measured from stored frozen samples, we thawed, aliquoted, and measured 25(OH)D levels from 19 participants who also had 25(OH)D levels measured on fresh samples from the original day of study. The correlation coefficient comparing levels from fresh and frozen samples was 0.97.

Statistical Analyses

Analyses were conducted to evaluate the association between 25(OH)D concentrations and the MAP response to an infusion of AngII in the obese and non-obese subgroups separately, and thereafter to evaluate whether these relationships were significantly different.

Demographic data are presented as mean values ± standard error of means. Student’s t-tests were used to compare means between the two independent populations with normal distributions, and the non-parametric Wilcoxon Ranks test was used for non-normal distributed data (PRA measurements). Chi-square and Fisher exact testing were used to detect statistical differences between categorical group frequencies. Pearson correlation and linear regression models were used to test for relationships between continuous variables. PRA was log transformed when used as a continuous variable for modeling. Continuous interaction analysis was performed to evaluate whether increasing BMI was an effect modifier of the relationship between 25(OH)D and MAP response to AngII. This interaction model included 25(OH)D, BMI, age, gender, baseline MAP, and an interaction term of BMI and 25(OH)D (each as a continuous variable). The level for significance for all tests conducted was set at α=0.05, with all reported p-values as two-tailed. Data analyses were performed using SAS statistical software, v9.1 (Cary, NC, USA).

RESULTS

Study Population

Characteristics of the two study populations are shown in Table 1. Obese individuals tended to have lower 25(OH)D levels and PRA, although this disparity was not statistically significant. There were no differences in baseline and AngII-stimulated MAP between obese and non-obese subjects. The majority of participants (89%) had 25(OH)D levels < 30 ng/mL.

Table 1.

Baseline characteristics of the study population by obesity status.

Characteristics	BMI < 30 kg/m2	BMI ≥ 30 kg/m2	p-value
n	58	39	-
Mean Age (y)	46.8 ± 1.2	46.1 ± 1.5	0.69
Gender (% female)	47	44	0.83
Mean BMI (kg/m2)	26.2 ± 0.4	32.1 ± 0.3	<0.0001
Mean 25(OH)D (ng/mL)	23.7 ± 1.1	21.2 ± 1.2	0.12
Plasma Renin Activity (ng/mL/hr)	0.50 (0.29, 0.90)*	0.31 (0.20, 0.80)*	0.13
24-hr urine sodium (mmol)	223.3 ± 8.9	228.4 ± 8.9	0.70
Baseline MAP (mmHg)	105.0 ± 1.7	102.2 ± 1.8	0.27
AngII Stimulated MAP (mmHg)	115.2 ± 1.7	115.1 ± 2.0	0.98

Results reported as means ± standard error of means, and as

^*median with interquartile ranges.

25(OH)D and RAS activity in Obesity

Among obese individuals, there was a modest but non-significant inverse association between 25(OH)D and circulating PRA (r = −0.17, β = −0.02, p = 0.32). However, a significant positive association was observed between 25(OH)D levels and the MAP response to AngII (r=0.41, β=0.70, p<0.01); low 25(OH)D concentrations predicted a blunted vascular sensitivity to AngII, while higher 25(OH)D concentrations were associated with improved vascular sensitivity to AngII. In a multivariable linear regression model including age, gender, BMI, and baseline MAP, 25(OH)D remained a significant independent predictor of the MAP response to AngII (Table 2). This multivariable model explained a large portion of the variability in the MAP response to AngII (model R²=0.67), with 25(OH)D explaining 11% of the variability in MAP response to AngII.

Table 2.

Multivariable adjusted relationships between 25(OH)D and other relevant variables with the MAP response to AngII in obese (top) and non-obese (bottom) subjects. Variables ordered by standardized effect estimates within each obesity status sub-group.

Obesity Status	Variable	β	Standardized β	Partial R2	p-value
OBESE (BMI ≥ 30 kg/m2)	Baseline MAP (mmHg)	0.85	0.72	0.540	<0.0001
	25(OH)D (ng/mL)	0.37	0.22	0.112	0.049
	BMI (kg/m2)	0.76	0.11	0.034	0.290
	Gender (female)	2.61	0.09	0.022	0.399
	Age (y)	0.06	0.04	0.005	0.694
NON-OBESE (BMI < 30 kg/m2)	Baseline MAP (mmHg)	0.66	0.67	0.448	<0.0001
	Age (y)	0.22	0.16	0.042	0.148
	Gender (female)	3.20	0.13	0.028	0.240
	25(OH)D (ng/mL)	−0.09	−0.06	0.006	0.590
	BMI (kg/m2)	0.20	0.04	0.003	0.700

25(OH)D and RAS activity in the Non-Obese

To determine whether the positive influence of 25(OH)D on the vascular sensitivity to AngII was specific to obesity, we evaluated whether 25(OH)D predicted the MAP response to AngII in non-obese subjects. Akin to obesity, there was a modest but non-significant inverse association between 25(OH)D and circulating PRA in non-obese individuals (r = −0.12, β = −0.01, p = 0.40). However, in contrast to obesity, the association between 25(OH)D and the MAP response to AngII was weak and non-significant in both univariate (r = 0.07, β = 0.10, p = 0.62) and multivariate modeling (model R²=0.48) (Table 2).

Obesity as an Effect Modifier

In accordance with our findings, suggesting an interaction between obesity status and 25(OH)D in predicting the vascular sensitivity to AngII, we used an adjusted, continuous interaction, model to evaluate whether increasing BMI was an effect modifier of the relationship between 25(OH)D and AngII-stimulated MAP. This multivariable model (adjusted for age, gender, BMI, and baseline MAP) confirmed that the adjusted effect estimates for the obese and non-obese groups were significantly different (p-interaction = 0.03, R² = 0.55).

DISCUSSION

Obesity and 25(OH)D deficiency are common pathologies in humans that have been increasingly observed to aggregate in tandem^30–32. Despite the high prevalence of HTN in obesity, very little is understood regarding its mechanisms. Efforts to further elucidate underlying mechanisms of HTN in obesity, and/or simple and effective methods to regulate RAS in obesity, may have tremendous implications in preventative health. The epidemiologic concurrence of obesity and 25(OH)D deficiency is relevant because both obesity^{5–8, 11} and 25(OH)D deficiency^{15–19, 21, 22, 39–41} have been linked to augmented RAS activity, which is a known contributor to cardiovascular disease^{10, 12–14}.

Our investigation included a population of Caucasian hypertensive men and women, who were largely 25(OH)D insufficient (< 30 ng/mL) according to current convention³. We employed the MAP response to an infusion of AngII to evaluate vascular RAS activity; where vascular RAS activity is inversely proportional to the MAP response to AngII. When dichotomized by obesity status, both obese and non-obese subgroups had similar 25(OH)D concentrations and similar MAP responses to AngII. In contrast, when the MAP response to AngII was evaluated as a function of 25(OH)D concentrations we observed that 25(OH)D levels positively predicted the vascular sensitivity to AngII in obesity, but not in the non-obese. These findings may suggest that 25(OH)D deficiency in obesity may result in appreciably augmented vascular RAS activity, whereas with higher 25(OH)D levels vascular RAS activity may be decreased. In contrast, among the non-obese subjects in our study, variations in 25(OH)D levels may not result in detectable changes in vascular RAS activity.

Our findings are consistent with and extend those from prior reports. We detected an inverse association between circulating PRA and 25(OH)D levels that was not significant; suggesting that the relationship between 25(OH)D status and the RAS may not be adequately quantified by measuring traditional circulating RAS components when in high sodium balance ^22–27. While animal studies have provided convincing support for 1,25(OH)₂D as an antagonist of renin expression^15–19, there is a paucity of studies exploring the RAS regulatory role of vitamin D in humans^20–22, with almost no attention to 25(OH)D deficiency in obesity. In concert with our current report, we recently demonstrated that 25(OH)D deficiency was associated with blunted renal vascular sensitivity to AngII in normotensive humans in high sodium balance²². Herein, we extend those findings by demonstrating that the influence of vitamin D on the vascular sensitivity to AngII may predominate in obesity. Furthermore, we demonstrate a similar inverse association between 25(OH)D and AngII sensitivity in the systemic vasculature. Taken together, our findings suggest the regulation of the vascular tissue RAS may be significantly influenced by the interaction of 25(OH)D and adiposity status.

To evaluate the role of body fat-content in the complex relationship between adiposity, 25(OH)D status, and RAS regulation, we explored whether BMI modified the relationship between 25(OH)D and vascular sensitivity to AngII. Continuous multivariable interaction analysis confirmed that increasing BMI significantly modified the relationship between 25(OH)D and the vascular sensitivity to AngII. These findings may suggest that with increasing adiposity, where tissue RAS activity is suspected to increase, the RAS-inhibitory influence of vitamin D is detectable. In contrast, in non-obese subjects where baseline tissue RAS activity is considered to be lower, the RAS-inhibitory effect of vitamin D may be too low to detect, or negligible in high sodium balance.

Contrary to our findings in humans, previous studies have demonstrated the RAS inhibitory effects of vitamin D analogs by showing an inverse relationship between circulating PRA and biologic vitamin D action^{15–17, 20, 21, 41}. This may be attributed to the fact that our study subjects were equilibrated on a PRA-suppressive high sodium diet. This diet-induced dampening of circulating PRA could reduce levels to the point where small differences are not readily distinguished within our sample size. Alternatively, this observation may suggest that vitamin D regulation of the RAS in humans is determined at the individual tissue level by “local renin-angiotensin systems,” or target organ AngII sensitivity, rather than the circulating RAS. The role of local renin angiotensin systems in vascular and organ structure pathology has gained significant credence, especially in adipose tissue, where we observed the most pronounced effect ^{9, 42}.

Our findings must be interpreted in the context of our study design. Our analysis was cross-sectional, and therefore cannot demonstrate causality or directionality. The analyzed sample size was small, thus resultant positive and negative findings will require validation in a larger cohort. However, to our knowledge this is the first study focused on the role of 25(OH)D in regulating RAS in obesity, and was optimized to control for many potential confounders of the RAS (i.e. - dietary salt, pharmacologic agents, and hypertension phenotype). Detecting differences in RAS components is challenging in high sodium balance where the RAS is suppressed; however, our study diet approximated the average Western dietary sodium intake^{36, 37}. The study population was comprised exclusively of Caucasian hypertensive subjects of homogenous HTN phenotype, thus limiting the generalizability of our findings to other ethnicities, races, other phenotypes of hypertension, and to normotensives, where RAS physiology is known to be different ^43–46. Most prior investigations in this area have reported active vitamin D levels (calcitriol: 1,25[OH]₂D); however, we focused on measuring 25(OH)D levels because of their stability, assay reliability, and direct relevance to clinical medicine. Even though the time of the year and seasonality are known to influence 25(OH)D levels³, our analysis was focused on evaluating the physiologic effect of 25(OH)D concentrations on the vascular sensitivity to AngII at the time of study; therefore we did not adjust for these factors. Parathyroid hormone, circulating calcium, and intra-cellular calcium, have long been associated in a complex interaction with dietary sodium, dietary calcium, RAS activity and hypertension^{47, 48}. Though our study design controlled dietary sodium and calcium intake, because calcium and parathyroid hormone were not directly measured, we cannot comment on whether associations between 25(OH)D and RAS activation were independent of these factors.

In conclusion, vitamin D may have significant RAS regulatory effects in humans, especially in obese individuals with hypertension. 25(OH)D deficiency, in the setting of obesity, may represent a state of heightened vascular RAS activity. Demonstrating that vitamin D favorably modulates the vascular sensitivity to AngII in obesity could provide substantial credence to the hypothesis that vitamin D deficiency in obesity heightens vascular tissue RAS activity, while its supplementation subdues it. Future prospective studies evaluating the effect of vitamin D supplementation in obesity could shed further insight on whether vitamin D may play a causal role in regulating tissue RAS activity.

ABBREVIATIONS

25(OH)D

25-hydroxyvitamin D

AngII

angiotensin II

BMI

body mass index

HTN

hypertension

MAP

mean arterial pressure

PRA

plasma renin activity

RAS

renin-angiotensin system