Association between serum 25‐hydroxyvitamin D and the effects of Angiotensin II receptor blocker on renal function among African Americans: A post hoc analysis of a randomized placebo‐controlled trial

Abstract

We tested the hypothesis that vitamin D status may modify the effect of Angiotensin II receptor blocker (ARB) on renal function among African Americans. Sixty‐four participants were included in this ancillary study from a randomized, double‐blind, placebo‐controlled, crossover trial among normotensive African Americans to test the effect of ARB on stress response of blood pressure and renal sodium handling. The participants were randomly assigned to receive either ARB or placebo for one week, washed out for one week and then cross‐overed to receive the other intervention for one week. On the final day of each intervention, the participant underwent a mental stress test. Baseline serum 25‐hydroxyvitamin D [25(OH)D] level was measured in this ancillary study. Sixty‐four participants were included, aged 26.5 ± 10.2 years and 47% were female. Among the participants with the serum 25(OH)D concentrations in the low tertile, ARB treatment was associated with 2.58 mg/dL higher blood urea nitrogen (BUN) (P < .001) and was not associated with serum creatinine (SCr) or estimated glomerular filtration rate (eGFR) (Ps > .05). Among the participants in the high 25(OH)D tertile, ARB was associated with 1.59 mg/dL lower BUN (P < .001), 0.08 mg/dL lower SCr (P = .001), and 8.59 mL/min/1.73 m² higher eGFR (P = .001). The interactions between vitamin D and ARB on renal function were more significant during stress and recovery than at rest. The effects of ARB treatment on renal function are modified by the vitamin D status among African Americans. ARB may improve renal function only among the ones with optimal vitamin D status.

1. INTRODUCTION

Angiotensin II (Ang II) is a vasoconstrictive hormone that plays a critical role in renal sodium handling and has been implicated as being one of the major contributors to salt‐dependent hypertension. ¹ Angiotensin II receptor blocker (ARB) is recommended as one of the major treatments of hypertension. ARBs inhibit the actions of Ang II through selective binding of type 1 (AT₁) receptors in vascular smooth muscle ² and are effective in lowering blood pressure (BP), preventing major cardiovascular outcomes. ³ However, ARB is shown to be less effective in lowering BP in African Americans than their white counterparts in multiple clinical trials and observational studies. ⁴ , ⁵ , ⁶ , ⁷ Therefore, it is recommended to avoid ARB as a first‐line therapy of hypertension for African Americans. ⁸ , ⁹ But it is still far from clear why there is such racial disparity regarding the ARB treatment.

ARBs bind to AT₁ receptors on kidneys and are supposed to have renoprotective effects, especially among patients with diabetic and nondiabetic chronic kidney disease (CKD). ¹⁰ , ¹¹ However, there are contradictory findings of the relationship between ARB and renal function. Some studies found that ARB improved renal function. A meta‐analysis showed that ARBs reduced the risks of end‐stage renal disease and reduced the doubling of the serum creatinine (SCr) level among patients with diabetes and albuminuria. ¹² Clinical trials found that ARB was effective in lowering SCr concentration ¹³ and protecting against the progression of nephropathy due to type 2 diabetes. ¹⁴ ARB treatment also reduces proteinuria and protects renal function even in advanced renal failure. ¹⁵ However, other studies found impairment of renal function resulting from ARB treatment in high‐risk groups. The Heart failure Endpoint evaluation of Angiotensin II Antagonist Losartan (HEAAL) study found that compared with 50, 150 mg losartan led to a greater reduction in estimated glomerular filtration rate (eGFR) and an increased risk of acute rise in SCr. ¹⁶ , ¹⁷ In the Valsartan in Heart Failure Trial (Val‐HeFT), eGFR decreased 6.7 mL/min/1.73 m² after 36‐month valsartan treatment compared with 2.9 mL/min/1.73 m² in the placebo group. ¹⁸ ARB is also an independent risk factor for developing contrast‐induced nephropathy. ¹⁹ Among the patients with advanced CKD, discontinuation of angiotensin‐converting enzyme inhibitor (ACEI)/ARB increased eGFR significantly and delayed the onset of renal replacement therapy. ²⁰

Renal impairment is more of a concern in African American patients when using ARB. ²¹ The underlying mechanism is not known. A retrospective cohort study found that the black patients hospitalized for acute myocardial infarction had the highest rate of acute renal failure at 12 months (4.3%) whereas the white patients had the lowest incidence (2.6%). ²² In addition, 17.2% of the African American patients with reduced left ventricular ejection fraction reduced or discontinued ACEI/ARB, among whom 57.6% were due to concerns of elevated SCr levels. ²¹

African American populations are prone to lower vitamin D status, ²³ which is associated with poor cardiovascular outcomes, ²⁴ cancer, ²⁵ and overall mortality. ²⁶ Previous studies suggest that vitamin D status contributes to the racial disparity in hypertension, ²⁷ cancer mortality, ²⁸ and albuminuria. ²⁹ Moreover, vitamin D may inhibit renin‐angiotensin system. ³⁰ Therefore, we aimed to test the hypothesis that the vitamin D status would modify the effect of ARB on renal function among African Americans in this ancillary study. Stress is a well‐established risk factor of hypertension and stress‐induced impairment of renal sodium excretion, ³¹ , ³² and antihypertensive therapy leads to disparate renal hemodynamic effects during exposure to mental stress as opposed to resting conditions. ³³ In addition, we further determined whether and how vitamin D modifies the effect of ARB on renal function during a stress test.

2. MATERIALS AND METHODS

2.1. Participants

Participants were recruited between 2008 and 2013. ³² , ³⁴ Inclusion criteria included: (a) African American by self‐report; (b) 18‐50 years old; (c) healthy normotensive based one self‐report and BP screening < 140/90 mm Hg. Exclusion criteria included: (a) on medication that influences BP; (b) pregnant. Among a total of 213 volunteers screened, 132 participants met the inclusion/exclusion criteria and completed the study protocol. A total of 64 participants had serum samples available and were included in this ancillary study. All the participants were classified into tertiles by their baseline serum hydroxyvitamin D [25(OH)D] concentrations. Informed consent was obtained from all participants prior to any measurements and a brief physical examination was done by a physician prior to testing. All procedures were approved by the Institution Review Board of Augusta University in accordance with the institutional guidelines.

2.2. Study design

The parental study was a randomized, double‐blind, placebo‐controlled, crossover clinical trial (NCT#: NCT02386293) (Figure S1). ³² , ³⁴ After screening, the participants underwent a pre‐test protocol prior to the treatment randomization to ensure that they did not have an adverse reaction to the ARB. Subsequently, participants were randomly assigned to receive either placebo or irbesartan (an ARB, 150 mg PO) for one week during which participants were on a fixed sodium‐controlled diet to allow similar levels of sodium balance at the time of testing. A dietitian identified foods with their sodium and potassium levels and participants selected their own diet to equal 4000 ± 200 mg of sodium and 2600 ± 200 mg of potassium per day. Overnight urine samples were collected for each night to assess dietary compliance. This method has been shown to reduce the variability of sodium intake, as estimated by overnight sodium excretion in free‐living individuals. ³⁵ On the final day of treatment, the participant performed the stress test protocol described below. This was followed by a one to two‐week washout period. The participants then followed the same procedure while on the other treatment. Randomization was performed and maintained by the university pharmacy. Participants were asked to refrain from drinking caffeinated or alcoholic beverages on the day before the studies.

2.3. Stress testing protocol

The stress testing protocol (Figure S1) ³² was a modified version of the standard stress testing protocol we have used successfully up to this point in over 1000 participants as described previously in both adult ³⁶ and pediatric populations. ³⁷ , ³⁸ , ³⁹ , ⁴⁰ The stressor was a competitive video game task played against another individual for monetary reward. On the testing day, the participants were tested in a private room in a comfortable chair. The participants took their last dose of treatment immediately before the stress testing. The testing period included: 45 minutes of rest (rest), 45 minutes of mental stress (stress), in which the participants played a video game, followed by another 45 minutes of rest (recovery). The participants watched a movie during Rest and Recovery periods.

2.4. Anthropometric and laboratory measurements

The hemodynamic responses were monitored by an automated BP machine (Dinamap Model 1864 SX, Critikon Inc) for systolic blood pressure (SBP) and diastolic blood pressure (DBP). Blood samples were obtained before and after rest, stress and recovery periods.

Serum 25(OH)D concentrations were measured using enzyme immunoassay (Immunodiagnostic Systems). The intra‐ and inter‐assay coefficients of variation (CV) were 5.6% and 6.6%, respectively. Our laboratory is certified by the vitamin D external quality assessment scheme (DEQAS), an international program monitoring accuracy of 25(OH)D measurements. Baseline 25(OH)D concentrations were measured in the serum samples drawn before randomization in this ancillary study.

Freshly collected blood samples were sent to Pathology Lab at Augusta University for measurement of comprehensive metabolic panel analysis including blood urea nitrogen (BUN) and serum creatinine (SCr). The inter‐ and intra‐assay CVs for creatinine are 11% and 6%. The inter‐ and intra‐assay CVs for BUN are 12% and 6%. eGFR was estimated by CKD‐EPI creatinine equation ⁴¹ . The equation is

$$\text{eGFR}=141\times min{\left(\frac{S_{\text{Cr}}}{\mathrm{\kappa }},1\right)}^{\mathrm{\alpha }}\times max{\left(\frac{S_{\text{Cr}}}{\mathrm{\kappa }},1\right)}^{‐1.209}\times 0.{993}^{\text{Age}}\times \mathrm{\beta }\times \mathrm{\gamma },$$

where S _Cr is the standardized serum creatinine, κ equals to 0.7 in females or 0.9 in males, α equals to −0.329 in females or −0.411 in males, β equals to 1.018 in females or 1 in males, and γ equals to 1.159 in Black or 1 in White.

2.5. Statistical analysis

The general characteristics of the participants are presented as mean ± standard deviation (SD) for continuous variables and N (%) for categorical variables. Two‐tailed t test was conducted to examine the differences of variables between ARB and placebo periods. Effects of ARB on BUN, eGFR, and SCr were investigated using mixed‐effects linear regression while incorporating repeated measured data and controlling for age, sex, and BMI as confounding variables. Mixed‐effects linear regression was also used to assess the interaction of ARB and 25(OH)D on BUN, SCr, and eGFR by including their interaction. In addition, the effects of ARB on BUN, eGFR, and SCr was estimated in each 25(OH)D tertile separately. All the analyses were performed for the markers measured at rest, stress, and recovery separately to fulfill the stress testing protocol. In addition, we also constructed the mixed‐effect models, which comprised all observations during the three stress testing periods with additional adjustment of the stress testing period (denoted as “overall”). A P‐value < .05 was considered significant. All analyses were performed using Stata version 12.0 (StataCorp.).

3. RESULTS

3.1. General characteristics

A total of 64 participants were included, aged 26.5 ± 10.2 years and 47% were female. At baseline, BMI was 30.2 ± 8.2 kg/m² and 25(OH)D concentration was 44.9 ± 14.8 nmol/L. A total of 19 participants (30%) were vitamin D sufficient [25(OH)D ≥ 50 nmol/L]. The cutoff points of 25(OH)D tertiles were 36.5 and 49.2 nmol/L, respectively. Among them, 53 were from ARB group and 54 were from placebo group (Table 1). ARB treatment reduced overall SBP (P = .026) and DBP (P < .001). During the stress test, DBPs were significantly lower when on ARB than on placebo during every period (Ps < .05). There was no significant association between ARB treatment and the markers of renal function (BUN, SCr, or eGFR, Ps > .05).

Table 1.

General characteristics

Characteristics	Overall (N = 64)
Baseline
Age (y)	26.5 ± 10.2
Female (%)	30 (47)
BMI (kg/m2)	30.2 ± 8.2
25(OH)D (nmol/L)	44.9 ± 14.8

	ARB (N = 53)	Placebo (N = 54)	P
BUN (mg/dL)
Overall a	11.4 ± 3.7	10.8 ± 3.5	.199
Rest	11.7 ± 3.9	10.9 ± 3.5	.285
Stress	11.2 ± 3.7	10.9 ± 3.5	.720
Recovery	11.1 ± 3.6	10.5 ± 3.6	.475
SCr (mg/dL)
Overall a	0.87 ± 0.19	0.86 ± 0.18	.672
Rest	0.90 ± 0.21	0.87 ± 0.18	.588
Stress	0.85 ± 0.16	0.86 ± 0.17	.789
Recovery	0.84 ± 0.18	0.83 ± 0.17	.752
eGFR (mL/min/1.73 m2)
Overall a	126.5 ± 23.7	125.9 ± 19.9	.790
Rest	123.1 ± 23.4	124.1 ± 21.9	.834
Stress	128.1 ± 20.6	125.8 ± 19.7	.591
Recovery	129.6 ± 21.6	128.2 ± 17.8	.758
SBP (mm Hg)
Overall a	115.1 ± 12.5	118.3 ± 10.6	.026
Rest	114.4 ± 11.4	115.8 ± 10.0	.511
Stress	116.4 ± 13.2	120.4 ± 10.2	.109
Recovery	114.4 ± 13.2	119.1 ± 11.6	.118
DBP (mm Hg)
Overall a	61.8 ± 7.8	66.5 ± 7.9	<.001
Rest	60.4 ± 7.2	64.0 ± 7.4	.014
Stress	63.3 ± 8.1	67.8 ± 7.4	.007
Recovery	61.9 ± 8.2	68.2 ± 8.4	.002

BMI and 25(OH)D concentration were measured only once for each participant on the first visit. Age, sex, BMI, and 25(OH)D concentration were summarized among the total participants (N = 64).

^aOverall means the average of the measurements throughout the mental stress test.

3.2. Effects of ARB on BUN, SCr, and eGFR

Figures 1A, 2A, and 3A show the levels of BUN, SCr, and eGFR at each period of the stress test. BUN and SCr were increased, and eGFR was decreased continuously from rest to stress, and to recovery period. The average levels of the BUN were higher when on ARB than on placebo, but the differences were not significant (Ps > .05). The levels of SCr and eGFR on ARB and on placebo were close to each other (Ps > .05). Table 2 presents the effects of ARB on BUN, SCr, and eGFR during each period of the stress test regardless of vitamin D status. There was no significant effect of ARB on BUN, SCr, or eGFR overall, at rest, stress, or recovery (Ps > .05), except that ARB was associated with higher eGFR in overall stress testing period (P = .044).

Figure 1.

Mean BUN levels during stress test among all participants (A) and among the participants in each of the 25(OH)D tertile (B low tertile, C middle tertile, and D high tertile). Dashed line: ARB (drug); solid line: placebo (placebo)

Figure 2.

Mean creatinine levels during stress test among all participants (A) and among the participants in each of the 25(OH)D tertile (B low tertile, C middle tertile, and D high tertile). Dashed line: ARB (drug); solid line: placebo (placebo)

Figure 3.

Mean eGFR levels during stress test among all participants (A) and among the participants in each of the 25(OH)D tertile (B low tertile, C middle tertile, and D high tertile). Dashed line: ARB (drug); solid line: placebo (placebo)

Table 2.

Effects of ARB on BUN, eGFR, and SCr during stress test

	β (95% CI)	P
BUN
Overall	0.33 (−0.20, 0.87)	.218
Rest	0.67 (−0.40, 1.74)	.219
Stress	0.16 (−1.11, 1.41)	.808
Recovery	0.38 (−0.97, 1.74)	.579
SCr
Overall	−0.02 (−0.04, 0.01)	.161
Rest	0.01 (−0.05, 0.06)	.840
Stress	−0.03 (−0.08, 0.02)	.236
Recovery	−0.02 (−0.07, 0.03)	.468
eGFR
Overall	3.05 (0.09, 6.01)	.044
Rest	0.58 (−5.78, 6.94)	.858
Stress	3.81 (−2.05, 9.66)	.203
Recovery	2.48 (−4.57, 9.53)	.491

Mixed‐effects models were adjusted for age, sex, and BMI. In the overall model, stress testing periods were also adjusted. Bold font represents significant association.

Abbreviations: BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; SCr, serum creatinine.

3.3. Interactions between the effects of ARB and 25(OH)D on BUN, eGFR, and SCr

Table 3 presents the results of mixed‐effects models investigating whether the serum levels of 25(OH)D at baseline modified the effects of ARB on BUN, SCr, and eGFR while adjusting for age, sex, and BMI. To better understand the results, marginal effects were plotted based on the mixed‐effects models (Figures S2‐S4).

Table 3.

Interaction between the effects of ARB and 25(OH)D on BUN, eGFR and SCr

	ARB		25(OH)D		Interaction
	β (95% CI)	P	β (95% CI)	P	β (95% CI)	P
BUN
Overall	18.95 (12.82, 25.07)	<.001	2.26 (−0.25, 4.77)	.077	−4.93 (−6.54, −3.31)	<.001
Rest	13.88 (1.30, 26.46)	.031	1.60 (−1.46, 4.66)	.307	−3.50 (−6.82, −0.18)	.039
Stress	20.78 (6.54, 35.02)	0.004	2.76 (−0.52, 6.04)	.099	−5.47 (−9.24, −1.71)	.004
Recovery	27.84 (13.09, 42.59)	<.001	3.12 (−0.37, 6.60)	0.080	−7.32 (−11.24, −3.41)	<.001
SCr
Overall	−0.34 (0.03, 0.65)	.032	0.01 (−0.11, 0.12)	.880	−0.10 (−0.18, −0.01)	.024
Rest	0.35 (−0.32, 1.02)	.309	−0.01 (−0.16, 0.14)	.936	−0.09 (−0.27, 0.09)	.314
Stress	0.41 (−0.16, 0.98)	.156	0.03 (−0.11, 0.17)	.664	−0.12 (−0.27, 0.03)	.127
Recovery	0.67 (0.03, 1.30)	.039	−0.02 (−0.17, 0.13)	.773	−0.18 (−0.35, −0.02)	.032
eGFR
Overall	−28.62 (−65.09, 7.86)	.124	1.50 (−13014, 16.13)	.841	8.39 (−1.23, 18.02)	.088
Rest	−31.79 (−109.77, 46.20)	.424	2.78 (−15.03, 20.58)	.760	8.59 (−11.98, 29.17)	.413
Stress	−53.58 (−124.54, 17.37)	.139	−3.61 (−21.33, 14.11)	.690	15.24 (−3.52, 33.99)	.111
Recovery	−81.83 (−166.03, 2.36)	.057	2.82 (−16.31, 21.95)	.773	22.50 (0.14, 44.85)	.049

Models were adjusted for age, sex, and BMI. In the overall model, stress testing periods were also adjusted. 25(OH)D was log transformed. Bold font represents significant association.

Abbreviations: BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; SCr, serum creatinine.

The serum concentration of 25(OH)D significantly modified the effects of ARB on BUN for every period during the stress test (Ps < .05), while 25(OH)D concentration itself was not associated with BUN (Ps > .05). Figure S2 shows that ARB increased BUN among the participants whose serum 25(OH)D concentrations were lower than 46.79 nmol/L, while decreased BUN among the other participants whose serum 25(OH)D concentrations were higher than 46.79 nmol/L. Similar patterns were also observed during each period of the stress test with cutoff points of 25(OH)D at 52.83, 44.51, and 44.71 nmol/L for rest, stress, and recovery, respectively.

The interactions between the effects of ARB and 25(OH)D on SCr were significant only at recovery (P = .032) and overall (P = .024). Similarly, 25(OH)D concentration itself was not associated with SCr (Ps > .05). Figure S3 shows that ARB was associated with higher SCr among the participants whose serum 25(OH)D concentrations were lower than 31.11 nmol/L. The association reversed after the cutoff point of 31.11 nmol/L. The cutoff point of 25(OH)D during the recovery period was 38.01 nmol/L.

The interactions between the effects of ARB and 25(OH)D on eGFR were significant only at recovery (P = .049) and borderline significant overall (P = .088). Circulating concentration of 25(OH)D itself was not associated with eGFR (Ps > .05). Figure S4 shows that, at recovery, ARB was associated with lower eGFR among the participants whose serum 25(OH)D concentrations were lower than 38.01 nmol/L. the association reversed after 38.01 nmol/L.

3.4. Effects of ARB on BUN stratified by 25(OH)D tertiles

Among the participants in the low tertile of serum 25(OH)D concentrations, ARB was associated with 2.58 mg/dL higher BUN (P < .001) compared with placebo overall. The changes in BUN from placebo to ARB during each stress testing period were similar. Among the participants in the high tertile of serum 25(OH)D concentrations, ARB was associated with 1.59 (P < .001), 0.30 (P = .708), 2.21 (P = .001), and 2.71 mg/dL (P < .001) lower BUN overall and during rest, stress, and recovery period, respectively. But the difference at rest was not statistically significant. Among the participants in the middle 25(OH)D tertile group, there was no significant association between ARB and BUN (Ps > .05) (Table 4 and Figure 1).

Table 4.

Effects of ARB on BUN, eGFR, and SCr during mental stress stratified by 25(OH)D tertiles

	Low 25(OH)D tertile		Middle 25(OH)D tertile		High 25(OH)D tertile
	β (95% CI)	P	β (95% CI)	P	β (95% CI)	P
BUN
Overall	2.58 (1.64, 3.51)	<.001	0.00 (−0.86, 0.87)	.997	−1.59 (−2.29, −0.90)	<.001
Rest	2.36 (0.29, 4.43)	.025	0.14 (−1.46, 1.73)	.866	−0.30 (−1.88, 1.27)	.708
Stress	2.85 (0.59, 5.12)	.014	−0.83 (−2.88, 1.23)	0.431	−2.21 (−3.46, −0.97)	.001
Recovery	2.97 (1.32, 4.62)	<.001	−0.24 (−2.69, 2.20)	.845	−2.71 (−4.09, −1.34)	<.001
SCr
Overall	−0.02 (−0.06, 0.02)	.285	0.04 (−0.01, 0.09)	.162	−0.08 (−0.12, −0.03)	.001
Rest	0.01 (−0.07, 0.10)	.804	0.06 (−0.06, 0.17)	.315	−0.05 (−0.13, 0.04)	.271
Stress	−0.03 (−0.09, 0.04)	.426	0.00 (−0.05, 0.05)	.902	−0.10 (−0.20, 0.01)	.077
Recovery	‐−0.01 (−0.07, 0.05)	.633	0.02 (−0.08, 0.12)	.732	−0.14 (−0.21, −0.07)	<.001
eGFR
Overall	2.82 (−2.17, 7.80)	.268	−1.89 (−7.38, 3.60)	.499	8.59 (3.64, 13.54)	.001
Rest	−0.72 (−12.39, 10.94)	.903	−4.32 (−15.91, 7.27)	.465	5.92 (−3.89, 15.73)	.237
Stress	1.51 (−8.02, 11.04)	.756	1.12 (−6.60, 8.83)	.777	12.01 (0.13, 23.88)	.048
Recovery	1.43 (−7.90, 10.76)	.764	−1.27 (−15.05, 12.51)	.857	14.92 (6.17, 23.68)	.001

Models were adjusted for age, sex, BMI, and stress testing periods. In the overall model, stress testing periods were also adjusted. Bold font represents significant association.

Abbreviations: BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; SCr, serum creatinine.

Angiotensin II receptor blocker was not associated with SCr among the participants in the low or middle tertile of 25(OH)D. Among the participants in the high tertile, ARB was associated with 0.08 (P = .001), 0.05 (P = .271), 0.10 (P = .077), and 0.14 mg/dL (P < .001) lower SCr overall and during rest, stress and recovery period, respectively. But the associations during rest and stress periods were not statistically significant (Table 4 and Figure 2).

Angiotensin II receptor blocker was not associated with eGFR among the participants in the low or middle tertile of 25(OH)D. Among the participants in the high tertile, ARB was associated with 8.59 (P = .001), 5.92 (P = .237), 12.01 (P = .048), and 14.92 mL/min/1.73 m² (P = .001) higher eGFR overall and during rest, stress, and recovery period, respectively. But the association during rest was not statistically significant (Table 4 and Figure 3). Since circulating 25(OH)D concentrations are highly correlated with obesity, additional subgroup analyses were performed stratified by obesity status (Table S1).

4. DISCUSSION

To the best of our knowledge, this is the first study to explore the role of vitamin D in the relationship between ARB and renal function among African Americans. Overall, baseline vitamin D status modifies the effects of ARB treatment on renal function. Among the participants in the low tertile of baseline 25(OH)D, ARB increases BUN; while among the participants in the high tertile of 25(OH)D, ARB decreases BUN and SCr, and increases eGFR. The interactions between vitamin D and ARB on renal function are more significant during stress and recovery than at rest.

Acute renal failure is one of the most prevalent severe adverse effects after initiating ACEI/ARB treatment. ²² Therefore, identifying factors that would attenuate the adverse renal effect is of great value for therapeutic treatment of hypertension. Other than age, sex, race, and the presence of CKD, which were associated with the risk of acute renal failure when using ARB, ²² , ⁴² our study found that vitamin D status, is also a modifiable factor that can be easily targeted in the clinical practice. In addition, vitamin D is a well‐tolerated pleiotropic hormone. A recent result from the Vitamin D and Omega‐3 Trial (VITAL) found no significant differences in adverse events between the vitamin D group and placebo group with respect to incident diagnoses of hypercalcemia, kidney stones, or gastrointestinal symptom. ⁴³

This study is the first that investigated the association between vitamin D and the renal effects of ARB in humans. Our data showed that ARB improved renal function only among the participants in the high tertile of 25(OH)D. However, among the participants in the low tertile of 25(OH)D, ARB increased BUN, which is a sign of renal function impairment. We estimated kidney function with multiple indicators (BUN, SCr, and eGFR) and found consistent and robust effects of the interactions between ARB and vitamin D status on each outcome variable. The vitamin D status has shown similar moderation effect on the treatment of ACEI/ARB for atrial fibrillation (AF). A retrospective study found that the incidence of AF was lower among the patients who initiated ACEI/ARB compared to the ones without ACEI/ARB when stratified by their 25(OH)D concentrations. ⁴⁴ Moreover, the benefit of ACEI/ARB use was attenuated in 25(OH)D deficiency, which suggests that 25(OH)D may act as a cofactor in the mitigation of AF by RAAS inhibition. ⁴⁴

The underlying biological mechanisms of the interaction between vitamin D status and ARB treatment on the renal function are unknown. The major side effect of the ARB treatment is the compensatory increases of plasma renin activity and circulating Ang II levels. ⁴⁵ Several animal models show that vitamin D suppresses renin expression. A study showed that ARB treatment caused a drastic stimulation of renin expression in both wild type and vitamin D receptor (VDR)‐null mice, but renin expression was much higher in the VDR‐null mice. ⁴⁶ Yuan et al ⁴⁷ suggested that vitamin D could suppress renin gene transcription by blocking the activity of the cyclic adenosine monophosphate (AMP) response element in the renin gene promoter in vivo. In mice, the blockade of the compensatory renin increase by vitamin D significantly enhances the efficacy of ARB and produces synergistic therapeutic effects in combination therapy. ⁴⁸ In diabetic mice, ARB and vitamin D each alone moderately ameliorated albuminuria and glomerular damage, but their combined use showed a dramatic therapeutic synergism, and these effects were accompanied by blockade of the compensatory increase of renin production and Ang I/II accumulation in kidney. ⁴⁹

African Americans show exaggerated BP responses to stress, which produce vascular injury underlying the premature development of hypertension and associated target organ damage. ³¹ , ⁵⁰ We previously found that stress increased renal sympathetic nerve activity and induced sodium retention through activation of Ang II. ³² This study showed that the interactions between vitamin D and ARB on renal function were more significant during stress and recovery than at rest. Similarly, a previous study found that ACEI did not change the GFR at rest, but increased GFR with mental stress. ³³ Stress increases the burden of kidney, in the meanwhile, stress also makes kidney more sensitive to stimulations such as ARB treatment.

There are several limitations of this study. First, the statistical power was compromised by the modest sample size, short term intervention (one week), and lowest recommended dosage of ARB for safety purposes. Second, the study participants were all African American normotensives, and the findings may not be generalizable to patients with hypertension or to whites. In addition, vitamin D status was measured by serum 25(OH)D concentration at baseline. Therefore, no causal relationship can be inferred about the effect of vitamin D in this scenario. A randomized controlled trial with 2‐by‐2 design of ARB and vitamin D treatment is warranted to verify our findings.

5. CONCLUSION

The baseline vitamin D status modified the effects of ARB treatment on renal function among African Americans. ARB improves renal function only among the participants with higher levels of vitamin D.

CONFLICT OF INTEREST

None.

AUTHOR CONTRIBUTIONS

Li Chen involved in conceptualization, formal analysis, and original draft preparation. Haidong Zhu involved in investigation, review, and editing. Gregory A. Harshfield involved in investigation and funding acquisition. Ying Huang involved in investigation. Yanbin Dong involved in investigation, supervision, review and editing.

Supporting information

Supplementary Material

Chen L, Zhu H, Harshfield GA, Huang Y, Dong Y. Association between serum 25‐hydroxyvitamin D and the effects of Angiotensin II receptor blocker on renal function among African Americans: A post hoc analysis of a randomized placebo‐controlled trial. J Clin Hypertens. 2020;22:1871–1880. 10.1111/jch.13997