Statin administration improves vascular function in heart failure with preserved ejection fraction

Abstract

Heart failure with preserved ejection fraction (HFpEF) is characterized by impaired vascular endothelial function that may be improved by hydroxy-methylglutaryl-CoA (HMG-CoA) reductase enzyme inhibition. Thus, using a parallel, double-blind, placebo-controlled design, this study evaluated the efficacy of 30-day atorvastatin administration (10 mg daily) on peripheral vascular function and biomarkers of inflammation and oxidative stress in 16 patients with HFpEF [Statin: n = 8, 74 ± 6 yr, ejection fraction (EF) 52–73%; Placebo: n = 8, 67 ± 9 yr, EF 56–72%]. Flow-mediated dilation (FMD) and sustained-stimulus FMD (SS-FMD) during handgrip (HG) exercise, reactive hyperemia (RH), and blood flow during HG exercise were evaluated to assess conduit vessel function, microvascular function, and exercising muscle blood flow, respectively. FMD improved following statin administration (pre, 3.33 ± 2.13%; post, 5.23 ± 1.35%; P < 0.01), but was unchanged in the placebo group. Likewise, SS-FMD, quantified using the slope of changes in brachial artery diameter in response to increases in shear rate, improved following statin administration (pre: 5.31e⁻⁵ ± 3.85e⁻⁵ mm/s⁻¹; post: 8.54e⁻⁵ ± 4.98e⁻⁵ mm/s⁻¹; P = 0.03), with no change in the placebo group. Reactive hyperemia and exercise hyperemia responses were unchanged in both statin and placebo groups. Statin administration decreased markers of lipid peroxidation (malondialdehyde, MDA) (pre, 0.652 ± 0.095; post, 0.501 ± 0.094; P = 0.04), whereas other inflammatory and oxidative stress biomarkers were unchanged. Together, these data provide new evidence for the efficacy of low-dose statin administration to improve brachial artery endothelium-dependent vasodilation, but not microvascular function or exercising limb blood flow, in patients with HFpEF, which may be due in part to reductions in oxidative stress.

NEW & NOTEWORTHY This is the first study to investigate the impact of statin administration on vascular function and exercise hyperemia in patients with heart failure with preserved ejection fraction (HFpEF). In support of our hypothesis, both conventional flow-mediated dilation (FMD) testing and brachial artery vasodilation in response to sustained elevations in shear rate during handgrip exercise increased significantly in patients with HFpEF following statin administration, beneficial effects that were accompanied by a decrease in biomarkers of oxidative damage. However, contrary to our hypothesis, reactive hyperemia and exercise hyperemia were unchanged in patients with HFpEF following statin therapy. These data provide new evidence for the efficacy of low-dose statin administration to improve brachial artery endothelium-dependent vasodilation, but not microvascular reactivity or exercising muscle blood flow in patients with HFpEF, which may be due in part to reductions in oxidative stress.

INTRODUCTION

Heart failure with preserved ejection fraction (HFpEF) is a complex clinical syndrome affecting millions of Americans and has become the predominant form of heart failure (HF) in the community (1–3). A growing body of evidence implicates impaired vascular endothelial function as a key feature of HFpEF pathophysiology (4–7) that is correlated with greater symptoms (5) and a poorer prognosis (8). Although effective pharmacotherapies have been developed to treat heart failure with reduced ejection fraction (HFrEF), no optimal treatment strategy has been identified for HFpEF (9), rendering it one of the greatest unmet needs in cardiovascular medicine.

HFpEF is characterized by the high prevalence of noncardiac comorbidities (i.e., hypertension, pulmonary hypertension, obesity, diabetes, and chronic kidney disease) associated with chronic inflammation. The increasingly recognized paradigm for HFpEF posits that these comorbidities drive microvascular endothelial inflammation through a systemic proinflammatory and pro-oxidant state (10). These derangements in inflammation and redox balance result in endothelial nitric oxide synthase (eNOS) uncoupling and reduction in nitric oxide (NO) bioavailability (11–13), likely contributing to impaired endothelial function in this population. Vascular dysfunction in patients with HFpEF may also impact cardiovascular control during exercise in this patient group. Indeed, using small muscle mass exercise modalities that evoke minimal cardiopulmonary stress, our group (7, 14–16) and others (17) have identified a 15–25% attenuation in exercising skeletal muscle blood flow in patients with HFpEF, implicating disease-related changes in the peripheral circulation as the limiting factor in the overall hyperemic response to exercise. Therefore, treatment strategies aimed at reducing inflammation and improving redox balance to increase NO bioavailability may improve peripheral vascular function and exercise responses in this patient population.

Recent studies from our group have demonstrated the efficacy of both acute antioxidant administration (18) and 10-day enteral l-citrulline (an NO substrate) consumption (19) to improve vascular and functional outcomes in patients with HFpEF. Although findings from previous work demonstrate the potential of therapies targeting the NO-biosynthesis pathway to favorably impact the peripheral circulation, the potential benefit of more conventional pharmacologic agents on peripheral vascular function has not been widely studied in this patient population. One such agent is hydroxy-methylglutaryl-CoA (HMG-CoA) reductase enzyme inhibitors (statins). This drug class has been shown to exert numerous systemic, pleiotropic effects on the peripheral vasculature independent of their potent cholesterol-lowering capabilities, with evidence for anti-inflammatory and antioxidant effects that improve NO bioavailability and endothelial function (20). Although statins have been shown to improve endothelial-dependent vasodilation in patients with HFrEF (21, 22), whether this drug class confers any benefit on vascular outcomes in patients with HFpEF has not been investigated. Therefore, this study sought to investigate the impact of 30-day statin administration on vascular function, and to determine the contribution of changes in inflammation and oxidative stress to these functional outcomes, in patients with HFpEF. We hypothesized that statin administration would improve conduit vessel and microvascular function, and exercising muscle blood flow in patients with HFpEF, and that these improvements would be associated with a reduction in biomarkers of inflammation and oxidative stress.

METHODS

Patients

New York Heart Association (NYHA) Class I–III HFpEF patients were recruited from the HF clinics at the University of Utah Health Sciences Center and the Salt Lake City Veterans Affairs Medical Center. All patients were stable on optimized, guideline-directed pharmacotherapy for ≥3 mo before study enrolment, and no medications were withheld prior to study visits. Patients with HFpEF were screened in clinic and included upon criteria consistent with the TOPCAT trial (23), which is as follows: 1) HF defined by the presence of ≥1 symptom at the time of screening (paroxysmal nocturnal dyspnea, orthopnea, dyspnea on exertion) and one sign (edema, elevation in jugular venous distention) in the previous 12 mo; 2) left ventricular ejection fraction (LVEF) ≥ 45%; 3) controlled systolic blood pressure; and 4) either ≥1 hospitalization in the previous 12 mo for which HF was a major component of hospitalization, or B-type natriuretic peptide (BNP) in the previous 60 days ≥100 pg/mL. Exclusion criteria for patients with HFpEF included a history of statin intolerance, current statin use, elevated liver function tests [aspartate aminotransferase (AST) and alanine aminotransferase (ALT)], atrial flutter, uncorrected primary valvular disease, smokers, orthopedic limitations, hormonal replacement therapy, dementia, severe cardiopulmonary obstructive disease, end-stage renal disease, type I diabetes, class III obesity (body mass index ≥ 40 kg/m²), uncontrolled hypertension, severe renal insufficiency, or end-stage malignancy. All experimental procedures and protocols were approved by the University of Utah and the Salt Lake City Veterans Affairs Institutional Review Board (IRB_138675), and written informed consent was obtained from all subjects prior to participation in the study.

Study Protocol

Using a parallel, double-blind, placebo-controlled design, participants were randomized to either 30 days of atorvastatin or placebo administration. All patients reported to the laboratory for testing before (pre) and after (post) 30-day intervention. The study visits were similar in duration and time of day. All patients reported to experimental visits after abstaining from food for >4 h, alcohol and caffeine for >8 h, and strenuous physical activity exercise for >24 h. After baseline testing, patients were provided a 30-day supply of either atorvastatin (10 mg daily) or placebo pills of similar taste, color, and appearance. Patients were asked to consume their last intervention medication >12 h prior to postintervention testing. To minimize the risk of known side effects of statins (muscle myalgia and elevated liver enzymes), and therefore subject attrition (24), we chose atorvastatin 10 mg, once daily, for statin administration as it is typically the lowest dose prescribed clinically. All data collection took place in a thermoneutral environment. A single investigator performed all analyses, blinded to the randomization of the study visits.

Data Acquisition, Techniques, and Measurements

Biomarkers of inflammation and oxidative stress.

Blood was sampled from the antecubital vein, and plasma and serum samples were stored at −80°C for subsequent analyses. Protein carbonyl, a marker of oxidant damage, was measured with a protein carbonyl ELISA kit (NWK-PCK01; Northwest Life Science Specialties, Vancouver, WA). Lipid peroxidation, a marker of oxidant damage, was assessed by plasma malondialdehyde (MDA) levels (LPO-586; Bioxytech, Foster City, CA). Serum proinflammatory biomarkers C-reactive protein (CRP), interleukin-6 (IL-6), and tumor necrosis factor α (TNF-α) were also evaluated. Serum CRP was measured using a liquid-phase, double-antibody radioimmunoassay (R&D Systems: Minneapolis, MN). High sensitivity IL-6 and TNF-α concentrations were measured using solid-phase sandwich ELISA kits (HS600B and DTA00C, respectively, R&D Systems). Total antioxidant capacity was evaluated by determining the ferric-reducing ability of plasma (FRAP). The antioxidant potential of the sample was determined using a ferrous iron standard curve. The reagents utilized are as follows: 300 mM acetate buffer, pH 3.6, 10 mM Fe²⁺ (FeSO₄·7H₂O) in acetate buffer, 10 mM TPTZ (2,4,6 tripyridyl-s-triazine) in 40 mM HCl, and 20 mM FeCl₃·6H₂O in 40 mM HCl. Endogenous antioxidant capacity, assessed by catalase and superoxide dismutase activity, was also assayed in the plasma (Cayman Chemical, Ann Arbor, MI).

Cardiovascular measurements.

Heart rate was monitored continuously from a standard three-lead ECG (ECG100C, Biopac, Goleta, CA) and beat-to beat arterial blood pressure was recorded continuously using finger photoplethysmography (Finometer, Finapres Medical Systems BV, Amsterdam, The Netherlands) on the nondominant limb. An automated sphygmomanometer (Omron) was placed on the dominant arm, and was utilized to calibrate blood pressure values obtained by Finapres. Brachial artery blood velocity and vessel diameter were assessed using a Logiq E9 ultrasound Doppler system (GE Medical Systems, Milwaukee, WI). A linear array transducer operating at 14 MHz was utilized for vessel wall imaging, and was collected, simultaneously, with the same transducer at a Doppler frequency of 5 MHz in high-pulsed repetition frequency mode (2–25 kHz). Sample volume was optimized in relation to vessel diameter and centered within the vessel. An angle of insonation of 60° was used for all measurements. Commercially available software (Logiq E9) was used to calculate angle-corrected, time-averaged, and intensity-weighted mean blood velocity. Brachial artery vasodilation was determined offline from end-diastolic, ECG R-wave triggered images collected from the Logiq E9 using automated edge-detection software (Medical Imaging Applications, Coraville, IA).

Brachial artery flow-mediated dilation and reactive hyperemia.

Following 20 min of supine rest, baseline measurements of brachial artery blood velocity and vessel diameter were taken for 1 min. Immediately following baseline measurements, a blood pressure cuff, placed on the right arm distal to the Doppler probe measurement site, was inflated to a suprasystolic pressure (250 mmHg) for 5 min. The cuff was then rapidly deflated, and brachial artery vessel diameter and blood velocity were obtained continuously for 2 min. Reactive hyperemia (RH) was quantified as the cumulative brachial artery blood flow [i.e., area-under-the-curve (AUC)] for the 2-min period post cuff release.

Handgrip exercise.

Participants rested in the supine position for ∼20 min before the start of data collection with the right arm abducted at 90°. The elbow joint was extended at heart level to allow individuals to perform handgrip (HG) exercise. First, maximal voluntary contraction (MVC) was established by taking the highest value of three maximal contractions using a handgrip dynamometer (Biopac Systems, Goleta, CA). Intermittent isometric HG exercise was then performed at four workloads (3 and 5 kg, and 30 and 45% of MVC, 1 Hz, 3 min per exercise stage), with force output displayed to provide visual feedback. If relative workload values were within 0.5 kg of an absolute workload, those exercise trials were averaged and combined. Brachial artery (BA) blood velocity, diameter, heart rate, and arterial blood pressure were obtained simultaneously during the last minute of each exercise stage. For the slope of BA diameter to shear rate, shear rate was calculated based on the formula: shear rate (s⁻¹) = 8 V_mean/arterial diameter. Due to physical limitations, two participants were unable to complete HG exercise.

Data analyses.

Brachial blood flow was calculated based on the formula: brachial blood flow (mL/min) = [V_mean × π (vessel diameter/2)² × 60]. Forearm vascular conductance (FVC) was calculated as follows: FVC (mL/min/mmHg) = forearm blood flow (FBF) (mL/min)/mean arterial pressure (MAP) (mmHg). FMD was quantified using the greatest increase in brachial artery vessel diameter during the 2-min period following cuff release. SS-FMD was calculated based on the formula: Slope (m) = brachial artery vessel diameter (y)/shear rate (x), across all HG exercise intensities. Shear rate was calculated as: shear rate (s⁻¹) = 8 V_mean/arterial diameter. Cumulative area-under-the-curve (AUC) values for blood flow and shear rate were integrated via the trapezoid rule and calculated as: (Σ{yi[x(i + 1) − xi] + (1/2)[y(i + 1) − yi][x(i + 1) − xi]}).

Statistical analyses.

Statistical analysis was performed with commercially available software (GraphPad Prism version 8.0.0 for Windows, GraphPad Software, San Diego, CA). A Student’s unpaired t test was used to determine mean difference for subject characteristics, resting MAP, heart rate, and FMD and RH AUC at baseline. Changes from baseline measures for FMD, RH, and circulating biomarkers were compared across the intervention using two-way (group × time), repeated-measures ANOVA. Changes from baseline measures for MAP, FVC, and FBF during progressive intensity HG exercise were compared across the intervention using three-way (group × time × workload), repeated-measures ANOVA. Šídák's post hoc testing was used when a significant main effect is detected. Pearson correlation coefficient testing was performed to examine the relationship between plasma biomarkers. Significance for the statistical analyses was accepted at P < 0.05. Values are presented as means ± SD.

RESULTS

Subject characteristics at baseline are presented in Table 1. Statin and placebo groups were similar for sex, age, body mass index, and handgrip strength (MVC). Baseline diastolic blood pressure was lower in the Statin group (P = 0.04); however, systolic and mean arterial pressure and resting heart rate were not different between groups (P > 0.05; Table 1). There were also no significant differences in New York Heart Association functional class (NYHA FC), comorbidities, medication types, or medication types taken between groups (P > 0.05 for all comparisons). Echocardiographic characteristics and clinical biomarkers are listed in Table 2. Groups were well-matched for ejection fraction (EF) at baseline (Statin, 65 ± 7%; Placebo, 64 ± 6%; P > 0.05).

Table 1.

Patient characteristics

	Statin	Placebo
Patient characteristics
n, M/F	1/7	3/5
Age, yr	74 ± 6	67 ± 9
Mass, kg	88 ± 24	95 ± 25
BMI, kg/m2	33 ± 8	34 ± 7
MVC, kg	14.8 ± 8.8	17.8 ± 5.3
Heart rate, beats/min	64 ± 12	70 ± 14
Systolic blood pressure, mmHg	121 ± 19	119 ± 7
Diastolic blood pressure, mmHg	69 ± 13*	80 ± 5
Mean arterial blood pressure, mmHg	86 ± 13	93 ± 4
NYHA class II, n	6	4
NYHA class III, n	1	3
Comorbidities
Diabetes	1	2
COPD	0	0
CAD	2	0
Hypertension	6	6
Pulmonary hypertension	5	7
OSA	4	6
Medications
β-Blockers	4	1
ACEi	1	0
ARB	0	2
Loop diuretics	7	6
Aldosterone antagonists	6	6
Rx types taken	3.4 ± 2.2	3.3 ± 1.6

Values for Rx types taken are means ± SD. ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; BMI, body mass index; CAD, coronary heart disease; COPD, chronic obstructive pulmonary disease; NYHA class, New York Heart Association functional classification; OSA, obstructive sleep apnea; Rx, prescription.

^*Student's unpaired t test; significant difference, P < 0.05.

Table 2.

Echocardiography

	Statin	Placebo
Echocardiogram
Ejection fraction, %	65 ± 7	64 ± 6
LV IVSD, cm	0.97 ± 0.11	0.98 ± 0.27
LV PWD, cm	0.98 ± 0.09	0.85 ± 0.16
LV ID diastole, cm	4.43 ± 0.49	4.77 ± 0.60
Peak E wave, cm/s	1.02 ± 0.34	0.88 ± 0.25
Peak A wave, cm/s	0.98 ± 0.23	0.89 ± 0.19
E/A ratio	1.12 ± 0.59	1.00 ± 0.23
E/E′ lateral wall ratio	10.8 ± 2.9	9.81 ± 5.26
Mitral E-wave deceleration time, ms	234 ± 60	321 ± 222

Values are means ± SD. A wave, peak velocity of late transmitral flow; BA, brachial artery; BNP, brain natriuretic peptide; E wave, peak velocity of early diastolic transmitral flow; E′, peak velocity of early diastolic mitral annular motion; ID, internal dimension; IVSD, interventricular septum thickness at end-diastole; LV, left ventricle; NT-proBNP, NH₂-terminal pro-B-type natriuretic peptide; PWD, posterior wall thickness; ST2, suppression of tumorigenicity.

FMD and RH

Prior to the intervention, no differences in baseline brachial artery diameter, blood flow, peak brachial artery diameter, cumulative shear rate at peak diameter, or RH were observed between groups (P > 0.05; Table 3). Similarly, brachial artery FMD, assessed by both peak percentage change (Statin, 3.33 ± 2.13%; Placebo, 2.57 ± 2.03%; P > 0.05; Fig. 1A) and absolute change in brachial artery diameter (Statin, 1.3 ± 0.8 mm; Placebo, 01.1 ± 0.7 mm; P > 0.05; Fig. 1B) were not different between groups at baseline. At day 30, no differences in baseline brachial artery diameter, blood flow, or cumulative shear rate at peak diameter were evident (P > 0.05; Table 3). However, statin administration increased brachial artery FMD, assessed by both peak percentage change (pre, 3.33 ± 2.13%; post, 5.23 ± 1.35%; P < 0.001; Fig. 1A) and absolute change in brachial artery diameter (pre, 01.3 ± 0.8 mm; post, 02.0 ± 0.5 mm; P < 0.01; Fig. 1B), whereas the placebo group remained unchanged versus preintervention (P > 0.05; Fig. 1). Furthermore, these increases in brachial artery FMD persisted even after normalization for shear rate (P < 0.01; Fig. 2). RH was unchanged (pre, 344 ± 228 mL; post, 435 ± 335 mL; P = 0.21; Table 3) following the statin intervention.

Table 3.

FMD and RH before and after intervention

	Statin		Placebo		P Value	Effect Size (Cohen’s d)
	Pre	Post	Pre	Post
Baseline BA diameter, mm	4.08 ± 1.05	4.02 ± 1.08	4.60 ± 1.35	4.44 ± 1.27	0.60	0.23
Baseline blood flow, mL/min	71.4 ± 50.6	80.0 ± 49.6	55.7 ± 10.6	69.1 ± 40.8	0.74	0.12
Peak BA diameter, mm	4.21 ± 1.03	4.22 ± 1.08	4.71 ± 1.35	4.54 ± 1.27	0.28	0.38
Cumulative shear rate at peak dilation, s−1	37,681 ± 12,535	36,148 ± 12,015	38,002 ± 19,347	33,880 ± 12,437	0.85	0.60
Time to peak dilation, s	65 ± 23	53 ± 16	66 ± 24	61 ± 21	0.64	0.33

Values are means ± SD. BA, brachial artery; FMD, flow-mediated dilation; Post, after intervention; Pre, before intervention; RH, reactive hyperemia. Two-way (group × time) ANOVA with RM, P < 0.05.

Figure 1.

Brachial artery flow-mediated dilation (FMD), expressed as percent change (A) and absolute change (B) from baseline before (Pre) and after (Post) placebo and statin administration. [Statin: n = 8, 1M/7F; Placebo: n = 8, 3 M/5F; two-way (group × time) ANOVA with RM].

Figure 2.

Brachial artery flow-mediated dilation (FMD) normalized for shear rate area-under-the-curve (SR AUC), expressed as percent change (A) and absolute change (B) from baseline before (Pre) and after (Post) placebo and statin administration. [Statin: n = 8, 1M/7F; Placebo: n = 8, 3M/5F; two-way (group × time) ANOVA with RM].

SS FMD during Progressive Intensity Dynamic Handgrip Exercise

Due to physical limitations, two participants (Statin, 1 F; Placebo, 1 F) were unable to complete HG exercise. Brachial artery shear rate increased with increasing workload in both groups. There were no group differences for brachial artery diameter or shear rate at baseline, 3-kg work rates, or 5-kg work rates, pre- or postintervention (P > 0.05 for all comparisons; Table 4). The slope of brachial artery diameter to shear rate was not different between groups preintervention (P > 0.05; Fig. 3). Following statin administration, the slope of brachial artery diameter to shear rate significantly increased (pre: 5.31e⁻⁵ ± 3.85e⁻⁵ mm/s⁻¹; post: 8.54e⁻⁵ ± 4.98e⁻⁵ mm/s⁻¹; P = 0.03), whereas the placebo group remained unchanged (pre: 5.97e⁻⁵ ± 3.87e⁻⁴ mm/s⁻¹; post: 5.57e⁻⁵ ± 4.86e⁻⁴ mm/s⁻¹; P = 0.88).

Table 4.

SS-FMD before and after intervention

	Statin		Placebo		P Value
	Pre	Post	Pre	Post
Baseline BA diameter, mm	4.00 ± 0.88	3.84 ± 0.78	4.51 ± 1.30	4.38 ± 1.10	0.96
3 kg BA diameter, mm	4.14 ± 0.86	4.04 ± 0.71	4.64 ± 1.49	4.47 ± 1.20	0.99
5 kg BA diameter, mm	4.21 ± 0.87	4.19 ± 0.77	4.59 ± 1.37	4.65 ± 1.30	0.89
30% MVC BA diameter, mm	4.16 ± 0.85	4.08 ± 0.73	4.72 ± 1.26	4.48 ± 1.12	0.70
45% MVC BA diameter, mm	4.25 ± 0.86	4.31 ± 0.76	4.96 ± 1.29	4.71 ± 1.16	0.32
Baseline shear rate, s−1	183 ± 118	189 ± 68	143 ± 75	158 ± 71	0.87
3 kg shear rate, s−1	489 ± 178	496 ± 109	445 ± 195	490 ± 248	0.58
5 kg shear rate, s−1	597 ± 188	722 ± 268	562 ± 240	623 ± 294	0.35
30% MVC shear rate, s−1	554 ± 165	565 ± 148	557 ± 202	488 ± 229	0.36
45% MVC shear rate, s−1	619 ± 160	702 ± 133	605 ± 230	640 ± 263	0.38
Slope of brachial artery diameter to shear, mm/s−1	5.31e−5 ± 3.85e−5	8.54e−5 ± 4.98e−5*	5.97e−5 ± 3.87e−5	5.57e−5 ± 4.86e−5	0.03

Values are means ± SD. BA, brachial artery; MVC, maximal voluntary contraction; Post, after intervention; Pre, before intervention; SS-FMD, sustained stimulus flow-mediated dilation. *Two-way (group × time) ANOVA with RM, P < 0.05.

Figure 3.

Brachial artery sustained-stimulus flow-mediated dilation (SS-FMD), quantified as the slope of brachial artery diameter in response to increases in shear rate during hand grip exercise, before (Pre) and after (Post) placebo and statin administration. [Statin: n = 7, 1M/6F; Placebo: n = 7, 3 M/4F; two-way (group × time) ANOVA with RM].

Exercise Hyperemia

There were no group differences in force of maximal voluntary contraction (MVC) (P > 0.05; Table 1). Heart rate progressively increased with increasing workload similarly in both groups (P > 0.05 for all comparisons; Table 1). As expected, FBF, MAP, and FVC progressively increased with the increasing work rate during handgrip exercise for both groups (Fig. 4, A–C). Prior to the intervention, no group differences in FBF, MAP, or FVC were found at baseline, 3-kg work rate, or 5-kg work rates (P > 0.05 for all comparisons; Fig. 4, A–C). Following the intervention, statin administration had no effect on FBF, MAP, or FVC at baseline or during handgrip exercise at either 3-kg or 5-kg work rate (P > 0.05 for all comparisons; Fig. 4, A–C).

Figure 4.

Forearm blood flow (A), mean arterial pressure (B), forearm vascular conductance (C), and heart rate (D) during handgrip exercise before (Pre) and after (Post) placebo and statin administration. [Statin: n = 7, 1M/6F; Placebo: n = 7, 3M/4F; three-way (group × time × workload) ANOVA with RM].

Circulating Biomarkers and Vascular Function

Circulating serum cholesterol and biomarkers of inflammation, oxidative stress, and antioxidant capacity are presented in Table 5. As expected, statin administration reduced serum low-density lipoprotein cholesterol (LDL-C) (P = 0.01), whereas triglycerides and high-density lipoprotein cholesterol (HDL-C) remained unaffected (P > 0.05). Malondialdehyde (MDA) was significantly reduced (≈23%) following statin administration (P = 0.04, Cohen’s d = 1.46; Fig. 5), whereas other markers of oxidative stress, antioxidant capacity, and inflammation were unchanged (P > 0.05 for all comparisons). No statistically significant relationship was observed between changes in FMD and circulating LDL-C (r = 0.659; P = 0.83) or MDA (r = 0.286; P = 0.49).

Table 5.

Circulating biomarkers before and after intervention

	Statin		Placebo		P Value	Effect Size (Cohen’s d)	Coefficient of Variation, %
	Pre	Post	Pre	Post
Serum cholesterol
LDL-C, mg/dL	100 ± 33	60 ± 17*	114 ± 38	122 ± 24	0.01	4.36
HDL-C, mg/dL	67 ± 15	66 ± 13	60 ± 21	60 ± 23	0.84	0.47
Triglycerides, mg/dL	107 ± 48	91 ± 42	113 ± 62	113 ± 31	0.48	0.94
Antioxidants
SOD, U/mL	1.65 ± 0.83	2.52 ± 1.66	1.79 ± 2.06	1.34 ± 0.69	0.13	0.71	4.6
Catalase, nmol	40.0 ± 17.2	30.0 ± 13.6	34.5 ± 14.0	32.3 ± 23.9	0.47	0.17	9.4
Oxidative stress
Protein carbonyl, nmol/mg	0.16 ± 0.06	0.17 ± 0.07	0.12 ± 0.05	0.12 ± 0.06	0.69	0.68	14
Antioxidant capacity
FRAP, mM	1.05 ± 0.21	1.03 ± 0.24	0.86 ± 0.26	0.87 ± 0.27	0.56	0.65	3.4
Inflammation
IL-6, pg/mL	4.63 ± 2.12	5.12 ± 2.66	3.45 ± 1.40	3.40 ± 1.57	0.44	0.64	2.9
CRP, ng/mL	3,364 ± 1,790	2,707 ± 1,497*	3,027 ± 2,202	3,578 ± 1,678	0.04	0.58	6.2
TNF-α, pg/mL	0.97 ± 0.31	1.07 ± 0.32	1.06 ± 0.53	1.20 ± 0.60	0.57	0.43	4.1

Values are means ± SD. CRP, c-reactive protein; FRAP, ferric reducing ability of plasma; HDL-C, high-density lipoprotein cholesterol; IL-6, interleukin-6; LDL-C, low-density lipoprotein cholesterol; MDA, malondialdehyde; Post, after intervention; Pre, before intervention; SOD, superoxide dismutase; TNF-α, tumor necrosis factor α. *Two-way (group × time) ANOVA with RM, P < 0.05.

Figure 5.

Plasma concentration of malondialdehyde (MDA), a biomarker of oxidative stress, before (Pre) and after (Post) placebo and statin administration. [Statin: n = 8, 1M/7F; Placebo: n = 8, 3M/5F; two-way (group × time) ANOVA with RM].

DISCUSSION

To our knowledge, this is the first study to investigate the impact of statin administration on vascular function and exercise hyperemia in patients with HFpEF. In support of our hypothesis, endothelial-dependent vasodilation, as determined by both conventional FMD testing and brachial artery vasodilation in response to sustained elevations in shear rate during handgrip exercise, increased significantly following statin administration. Interestingly, reactive hyperemia was unchanged after the 30-day statin regimen, suggesting that the efficacy of this drug class to confer a beneficial effect in conduit vessel function may not extend to the peripheral microvasculature. Exercising skeletal muscle blood flow was also unaffected following statin therapy, and thus it appears that this intervention may not directly impact the pathways responsible for impaired exercise hyperemia in patients with HFpEF. In addition, statin administration resulted in a marked reduction in MDA, a biomarker of lipid peroxidation, providing new evidence for the efficacy of this drug class to mitigate oxidative damage in this patient group. Taken together, these findings provide new evidence for the efficacy of statins to improve vascular endothelial function in patients with HFpEF, likely mediated through reductions in oxidative stress and increased NO bioavailability.

Conduit Vessel Function and Statins in HFpEF

Brachial artery FMD testing provides a noninvasive index of endothelium-dependent conduit artery vasodilation that has been utilized in both healthy and diseased populations, with a number of investigations identifying an attenuated FMD response in patients with HFpEF compared with both patients with hypertension (25, 26) and healthy, age-matched controls (27). A more recent investigation from our group extended these findings, having found that FMD was diminished in patients with HFpEF versus healthy age-matched controls, although, notably, these differences were abolished after FMD was normalized for the shear stimulus (4). Although the extent of endothelial dysfunction in patients with HFpEF is unclear, most evidence to date suggest some manner of endothelial function in this population (28). In addition to the traditional assessment of FMD to a transient shear stress stimulus provoked by distal cuff occlusion, endothelium-dependent vasodilation can also be assessed using the so-called “sustained shear stimulus” approach provoked by exercise, vasodilator drug infusion, or limb heating (29). Importantly, our group has previously demonstrated that brachial artery vasodilation in response to step-wise increases in sustained shear during handgrip exercise is markedly reduced in young, healthy subjects following eNOS blockade [intra-arterial infusion of N-monomethyl-l-arginine (l-NMMA)] (30), supporting the validity of this experimental paradigm for determination of NO-dependent vasodilation.

In the present study, 30-day atorvastatin administration significantly improved brachial artery FMD, expressed as both percent change (Fig. 1) and percent change normalized for the shear stimulus (Fig. 2) in patients with HFpEF. Furthermore, the sustained-stimulus FMD response, quantified by determining the slope of changes in brachial artery diameter across the range of shear stimuli during handgrip exercise, was also improved after statin administration (Fig. 3). Together, these data provide new evidence for the efficacy of low-dose HMG-CoA reductase inhibition to improve brachial artery endothelium-dependent vasodilation in patients with HFpEF. Although statin therapy has been shown to improve endothelial function in other populations, including HFrEF (21, 22), to our knowledge this study is the first to report similar beneficial effects on the vasculature to patients with HFpEF.

This evidence of “plasticity” in the peripheral circulation adds to growing number of studies examining the potential benefit of drugs targeting the l-Arginine-NO-cyclic guanosine-3′,5′-monophosphate (l-Arg-NO-cGMP) pathway in patients with HFpEF. Indeed, over the past decade, several randomized clinical trials have evaluated the efficacy of inorganic nitrates/nitrites, soluble guanylate cyclase stimulators and activators, phosphodiesterase-5 inhibitors, and angiotensin receptor neprilysin inhibition on patient-centered outcomes such as quality of life (QOL) (31). Although many of these studies have produced negative or neutral results, smaller studies focused on vascular end points have proven more promising. Previous investigations from our group have provided preliminary evidence for the ability of both antioxidant (18) and l-citrulline (19) administration to improve vascular outcomes in patients with HFpEF, findings that support continued study to evaluate whether more conventional HF pharmacotherapies such as soluble guanylate cyclase stimulators and activators can improve endothelial health in this patient population. Beyond the statin-induced improvement in FMD observed in the present study, the additional effects of this drug class exert that may positively impact cardiovascular health, including reduction of serum cholesterol, inhibition of vascular smooth muscle cell proliferation and migration (32, 33), and attenuation of sympathetic nerve activity (34). Considering the heterogeneous and polymorbid nature of the HFpEF clinical syndrome (35), the nonvascular, pleiotropic effects that statin therapy may confer could be particularly useful in treating this patient population, and should be further explored.

Microvascular Dysfunction in HFpEF

In contrast to the ambivalence in the literature regarding the degree of macrovascular dysfunction in patients with HFpEF, the preponderance of evidence suggests a disease-related decrement in microvascular dysfunction in both the coronary and peripheral vascular beds (36). Although several experimental techniques are available for determining microvascular reactivity in vivo, the degree to which blood flow increases following a period of distal cuff occlusion, known as reactive hyperemia (RH), is a well-established approach that is assessed during the FMD testing procedure. Using this technique, we have previously identified a ∼30% reduction in RH in patients with HFpEF compared with healthy, age-matched controls (4), and subsequently identified the ability of enteral l-citrulline (a substrate for NO biosynthesis) administration to improve RH in patients with HFpEF (19), supporting the efficacy of therapies targeting the NO biopathway to improve microvascular reactivity in this population. Considering that statin therapy has also been shown to improve NO bioavailability through anti-inflammatory and antioxidant properties (37), we hypothesized that 30-day statin therapy would improve microvascular reactivity in patients with HFpEF. Surprisingly, RH was unchanged following statin administration (Table 3), suggesting that HMG-CoA reductase inhibition is unable to alter the pathways responsible for the RH response. Although there are likely several factors that may explain this lack of efficacy, uncertainty regarding the role of NO in governing the RH response is an important consideration. Although some investigations demonstrate the contribution of NO in post-ischemic hyperemia (38, 39), others have challenged these findings, suggesting NO plays a minimal role in this response (40, 41). Indeed, although NO has been shown to be a key mediator of the FMD response, Crecelius et al. (42) elegantly demonstrated that activation of inwardly rectifying potassium channels and Na⁺/K+-ATPase contributes to almost all of the RH response, with a minimal contribution from NO. This ambiguity regarding the role of NO in the RH response may explain the present findings demonstrating that statin administration may improve FMD, but not RH in patients with HFpEF. However, as noted earlier, there is initial evidence for the effectiveness of increased NO substrate to augment the RH response in patients with HFpEF (19). The disparity between these former and current findings highlights the need for additional studies to establish the mechanisms responsible for microvascular dysfunction in patients with HFpEF in parallel with experiments focused on evaluating the efficacy of interventions for improving microvascular health in this patient group.

Exercise Hyperemia in HFpEF

An emerging body of evidence suggests that exercising muscle blood flow is reduced in patients with HFpEF (7, 14–17). Specific to the handgrip exercise paradigm used in the present study, we have previously demonstrated marked reductions in brachial artery blood flow and forearm vascular conductance (FVC) during handgrip exercise in patients with HFpEF compared with healthy (14) and hypertensive (15) controls that could not be attributed to changes in central hemodynamics, suggesting decrements in exercise capacity may be attributed to disease-related impairments in exercise hyperemia. Furthermore, there is new evidence for the effectiveness of aerobic exercise training to improve exercising limb blood flow during submaximal exercise in patients with HFpEF (17), suggesting that impairments in exercise hyperemia may be remediable in this population.

Contrary to our hypothesis, statin administration did not alter exercising skeletal muscle blood flow or FVC during rhythmic handgrip exercise (Fig. 4). This lack of efficacy is in contrast to earlier work in young, healthy individuals identifying a small but significant improvement in limb blood flow during submaximal knee-extensor exercise following 6 mo of high dose (80 mg daily) atorvastatin administration (43), and may suggest that a longer treatment period and/or higher dosing is needed to alter the exercise hyperemic response. The work rates utilized in the present study are an additional consideration, as we have previously demonstrated that the contribution of NO to vasodilation during handgrip exercise is most evident at very high work rates (20–24 kg), with a minimal effect at rest or at lower exercise intensities in young, healthy subjects (30). Thus, it is possible that in the present study, the work rates of handgrip exercise were too low to reveal differences in the hyperemic response during handgrip exercise following statin administration. Thus, although the work rates selected for the present study were tailored to maximize the probability that patients could complete all stages of handgrip exercise, it is possible that statin-induced improvements in exercise hyperemia are only evident during higher-intensity exercise that relies more heavily on NO signaling.

Impact of Statins on Biomarkers of Inflammation and Oxidative Stress in HFpEF

The comorbidity-inflammation paradigm for HFpEF posits that noncardiac comorbidities drive the development and severity of HFpEF through a cascade of events resulting from microvascular endothelial inflammation, redox imbalance, and reduced NO bioavailability that may diminish vascular function and contribute to severe exercise intolerance, a hallmark of the disease (10, 44–46). A body of evidence demonstrates the pleiotropic, beneficial effects of statins that include anti-inflammatory and antioxidant properties (20–22, 47–49) that may mitigate key aspects of HFpEF pathophysiology. Thus, in addition to the assessment of functional vascular outcomes, we also sought to examine the impact of statin administration on plasma biomarkers of inflammation and oxidative stress to explore the potential mechanisms for statin-induced improvements in vascular function.

The present investigation revealed that 30-day statin therapy significantly reduced MDA, a product of lipid peroxidation and marker for oxidative damage, in patients with HFpEF (Fig. 5). Although reductions in plasma concentrations of MDA following statin therapy could, in part, be attributed to the observed decrease in LDL-C (Table 3), lipid peroxidation may also be impacted by the generation of free radicals and antioxidant activity (50). Indeed, Wilson et al. (51) demonstrated that simvastatin attenuates increases in MDA and preserves vascular endothelial function in experimental hypercholesterolemia in the absence of cholesterol lowering, suggesting that statins also exert antioxidant properties independent of their lipid-lowering capabilities. Regardless, considering the impact of oxidative stress on the NO biosynthesis pathway, the observed reduction in this marker of oxidative stress likely, in part, contributes to the significant improvement in vascular endothelial function following statin administration in patients with HFpEF. However, linear regression analyses failed to identify a significant relationship between these biomarkers and results from FMD testing, suggesting that statin-induced changes in redox balance may not play a deterministic role in the improvement of endothelial-dependent vasodilation following treatment. Although C-reactive protein (CRP) directionally improved in most subjects (-≈20%, Table 3) following statin administration, this did not reach statistical significance. This finding was surprising considering previous evidence demonstrating that statin therapy reduces CRP (52–55), and may indicate that a higher statin dose is needed to provoke a detectable improvement in this proinflammatory biomarker.

Impact of Statins on Serum Cholesterol and Vascular Function in HFpEF

As expected, 30-day statin administration reduced LDL-C, whereas triglycerides and HDL-C were unaffected (Table 5). An important consideration is whether the observed improvements in vascular function were related to the pleiotropic effects of statin therapy, or simply its cholesterol-lowering effect. Our analysis revealed that changes in FMD were not correlated with changes in LDL-C, suggesting improvements in vascular function were not related to the cholesterol-lowering effect of statin administration. Although statins indeed provide cardiovascular benefits related to reductions in serum cholesterol, an abundance of evidence supports that statins also exert numerous pleiotropic effects independent of reductions in circulating cholesterol concentration. Indeed, Laufs et al. (56) have demonstrated that treatment with atorvastatin 80 mg improves endothelial function in healthy, normocholesterolemic young men within 24 h, which is weeks before any significant changes in cholesterol would likely be observed. Furthermore, Gounari et al. (22) used a double-blind, placebo-controlled crossover trial with rosuvastatin and ezetimibe (a selective cholesterol-absorption inhibitor) to compare the effects of these hypercholesterolemic drugs on endothelial function in patients with HFrEF. In that study, rosuvastatin, but not ezetimibe, significantly improved brachial artery FMD in these patients despite similar reductions in plasma lipid levels, suggesting statins improve vascular endothelial function by mechanisms independent of lipid-lowering.

Perspectives

HFpEF is characterized by a chronic proinflammatory state that drives disease progression and peripheral vascular dysfunction (10). Considering its prevalence coupled with lack of effective pharmacotherapies to improve outcomes, HFpEF has been considered one of the greatest unmet clinical needs in cardiovascular medicine. Therefore, it is imperative that novel pharmacologic approaches continue to be explored to optimize the treatment of this condition. Findings from the present investigation suggest that statin administration improves redox balance and improves endothelial-dependent vasodilation in this patient population. The present study supports previous findings from our group suggesting therapies targeting the NO biosynthesis pathway can improve peripheral vascular dysfunction in patients with HFpEF (18, 19), as well as new evidence for the novel use of this drug class to improve vascular-endothelial function in this patient population.

Experimental Considerations and Limitations

We administered atorvastatin at 10 mg, the lowest clinically prescribed dose. Although this approach served to minimize the risk of known side effects (muscle myalgia and elevated liver enzymes), and therefore subject attrition (24), we cannot exclude the possibility that higher dosing and/or longer duration of treatments could have a greater effect size or additional benefits that were not revealed with a lower dose. By design, no prescribed medications were withheld on study days, a standard approach by our group (4, 7, 14–16, 57) that enables the opportunity to study these patients in a “real-world” setting. However, given the variety of pharmacologic agents utilized to optimize the clinical care of patients with HFpEF (Table 1), we cannot exclude the possibility that this approach may have influenced the observed vascular and biomarker responses. Considering previous evidence demonstrating the efficacy of high-intensity lipophilic statins to reduce muscle sympathetic nerve activity in patients with HFpEF (34), it is possible that sympathetic nerve activity may have been modulated following statin administration in the present investigation. Another consideration is that we did not assess quality of life scores across the intervention. We did, however, assess 6 minute walk test (6MWT) distance, which has been shown to be associated with health-related quality of life (58). Nevertheless, statins did not impact 6MWT distance in the present study, and therefore it is unlikely that QOL was favorably impacted. Although the lack of post-treatment echocardiography precludes assessment of potential statin-induced changes in cardiac function in the present study, previous work failed to demonstrate an improvement in diastolic function after 24 wk of statin therapy in patients with hypercholesterolemia and diastolic dysfunction (59). We also acknowledge that the calorimetric assessment of MDA utilized in the present study does not distinguish between bound and free fractions of MDA in plasma, and may therefore provide an incomplete assessment of systemic oxidative stress (60). Finally, we acknowledge the possibility that the relatively small sample size in the present study may have limited the statistical power of some comparisons, particularly correlational analyses.

Conclusions

To the best of our knowledge, this study is the first to investigate the impact of statin administration on vascular function and exercise hyperemia in patients with HFpEF. In support of our hypothesis, both conventional FMD testing and brachial artery vasodilation in response to sustained elevations in shear rate during handgrip exercise increased significantly in patients with HFpEF following statin administration, beneficial effects that were accompanied by a decrease in biomarkers of oxidative damage. In contrast, statin administration did not alter RH or exercising muscle blood flow, suggesting that this intervention may not directly impact the pathways responsible for disease-related decrements in microvascular function and exercise hyperemia in this patient group. Collectively, this study provides new evidence for the efficacy of statin therapy to improve endothelial-dependent vasodilation in patients with HFpEF, likely mediated through reductions in oxidative stress and improved NO bioavailability.

DATA AVAILABILITY

Data will be made available upon reasonable request.

GRANTS

This work was funded in part by the NIH Grants HL162856 and HL139451 and the U.S. Department of Veterans Affairs Grants CX002152 and IK2RX003670.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.J.I. and D.W.W. conceived and designed research; J.J.I., J.K.A., S.M.R., J.C.C., J.M.S., and K.B. performed experiments; J.J.I. analyzed data; J.J.I., J.K.A., S.M.R., J.C.C., J.M.S., J.Z., V.R., K.B., C.L.M., J.J.R., and D.W.W. interpreted results of experiments; J.J.I. prepared figures; J.J.I. drafted manuscript; J.J.I., J.K.A., S.M.R., J.C.C., J.M.S., J.Z., V.R., K.B., C.L.M. J.J.R., and D.W.W. edited and revised manuscript; J.J.I., J.K.A., S.M.R., J.C.C., J.M.S., J.Z., V.R., K.B., C.L.M., J.J.R., and D.W.W. approved final version of manuscript.