Role of Blood Pressure Responses to Exercise and Vascular Insulin Sensitivity with Nocturnal Blood Pressure Dipping in Metabolic Syndrome

The World Health Organization identifies elevated blood pressure as the leading risk factor for cardiovascular disease (CVD) [1]. The American College of Cardiology and American Heart Association recently established new guidelines to define hypertension (HTN; ≥130/80 mm Hg) that classified ∼45% of American adults to have HTN [2, 3]. Importantly, 24-h ambulatory blood pressure monitoring (AMBP) better predicts CVD risk and target organ damage than the traditional, single clinic blood pressure assessment [4]. While systolic blood pressure (SBP) typically dips approximately 10–20% in the transition from daytime to nighttime, blunted dips (e.g., <10%) have been linked to both elevated CVD risk [5] as well as all-cause mortality, [6, 7] independent of daytime or clinical HTN status based on a single measurement. Thus, identifying key clinical factors that account for blood pressure circadian rhythm is needed to optimize care.

Traditional risk factors for clinic HTN (e.g., high dietary sodium, lack of physical activity, arterial stiffness, and obesity) have been well-described [8, 9]. However, factors affecting 24-h blood pressure are less understood. Interestingly, insulin resistance has emerged as a factor promoting HTN [10]. Hyperinsulinemia generally increases sodium reabsorption in the kidneys, in part, due to insulin overactivating sympathetic drive, and this is purported to promote essential HTN in patients with obesity-related type 2 diabetes (T2D) [11, 12] as well as metabolic syndrome (MetS) [13]. However, it is also important to acknowledge that the correlation between hyperinsulinemia and HTN in individuals with MetS may be somewhat protective in natriuresis, thereby favoring blood pressure homeostasis [14, 15]. In either case, the potential for greater sympathetic drive is considered to promote systemic and renal blood vessel vasoconstriction that raises blood pressure [16, 17]. Similarly, a hypertensive response to exercise (>210 mm Hg for men; >190 mm Hg for women), which is suggestive of elevated sympathetic drive, may also relate to reduced endothelial function given that impaired NO synthesis contributes to a vasoconstrictive state [18-20]. However, a major limitation of prior studies examining the role of insulin sensitivity is that they assessed clinic blood pressure, and none of them used the “gold-standard” euglycemic clamp to directly test insulin action on metabolic and vascular tissue. Thus, there is a knowledge gap about the role of insulin sensitivity in relation to 24-h AMBP. Moreover, few studies [21, 22] have has assessed blood pressure dipping in adults with MetS. This is concerning given that MetS is a cluster of risk factors inclusive of HTN and patients with MetS have a higher risk of CVD compared to those without [23]. In fact, it was reported that MetS-Score, a global risk assessment scoring system for MetS based on multivariate analysis [22], was a strong predictor of non-dipping and elevated nighttime SBP (>110/65 mm Hg) [5, 21]. Whether insulin sensitivity, endothelial function, and/or sympathovagal balance associates with nighttime SBP in MetS adults remains unclear. Therefore, the purpose of this study was to examine factors that relate to SBP dipping and 24-h blood pressure in patients with MetS. We hypothesized that metabolic and vascular insulin sensitivity as well as sympathovagal balance would significantly associate with AMBP in patients with MetS. We further hypothesized that patients who experienced a ≥10% drop in their SBP at night would have more favorable metabolic and vascular outcomes.

Eighteen adults (n = 18; 13 F/5 M; 53.2 ± 6.5 years; body mass index [BMI] 35.8 ± 4.5 kg/m²) were recruited via social media and/or newspaper advertisements from the Charlottesville, VA community. Participants were nonsmoking, sedentary (exercise <60 min week), weight stable (≤2 kg over the previous 3 months) and screened for MetS based on the National Cholesterol Education Panel Adult Treatment Panel (ATP) III criteria. Participants were excluded if they had chronic disease (e.g., cardiovascular, renal, hepatic, pulmonary, etc.) or were taking antidiabetic (e.g., sulfonylureas, biguanides, α-glucosidase inhibitors, etc.), weight loss (e.g., orlistat, lorcaserin, etc.), lipid-lowering medications (e.g., statins), and/or medications affecting blood pressure as well as heart rate (HR)/rhythm (e.g., beta-blockers, ACE-inhibitors, etc.) based on self-reported medical history. Blood analyses were performed prior to physical exam (e.g., electrolytes, estimated glomerular filtration rate [eGFR], etc.) that also included a resting electrocardiogram and peak exercise stress test to rule out heart dysfunction. All participants provided written and verbal informed consent approved by the University of Virginia Institutional Review Board (IRB#19364, NCT03355469).

Body mass (kilograms) and height (meters) were measured using a digital scale and stadiometer, respectively, to assess BMI. Total body fat and lean body mass were measured using dual-energy X-ray absorptiometry (Lunar Prodigy; GE Healthcare). Participants completed a continuous incremental peak oxygen consumption (VO₂peak) test on a treadmill with indirect calorimetry (CareFusion, V_max CART, Yorba Linda, CA, USA) using routine criteria (e.g., plateau in VO₂, >1.1 RER, or within 10% of age-predicted HRmax [220 – age]). A self-selected speed was chosen by the participant and incline rose by 2.5% every 2 min until participants reached VO₂peak. Blood pressure was obtained during the last 30 s of rest and stages 1 (baseline), 3, 4, and peak. HR was continuously monitored using a 12-lead electrocardiogram during exercise up to peak (HRpeak) to estimate sympathetic activity. HR was also tested 1- and 2-min post-exercise to assess HR recovery (HRR; peak – post-exercise), while participants walked at 1.5 mph at 2.5% grade to estimate parasympathetic withdrawal. In addition, rate pressure product (RPP = HR × SBP) was calculated for each stage of exercise.

Participants were provided a standardized American Heart Association diet that consisted of 55% carbohydrate, 30% fat, and 15% protein for 24-h prior to measurements. Caloric needs were determined using indirect calorimetry and multiplying resting metabolic rate by an activity factor of 1.2. Dietary information was analyzed using The Food Nutrition Processor Analysis Software (version 11.1; ESHA Research, Salem, OR, USA). Participants were also instructed to refrain from vigorous exercise 72-h prior to measurements as well as alcohol and caffeine 24-h prior to the clamp procedure and receiving the AMBP monitor.

All participants wore the ABPM (Welch Allyn ABPM-6100, Skaneateles Falls, NY, USA) cuff on their upper left arm following the manufacturer’s instructions for a period of 24 h, including rest/sleep (bedtime and wake times were self-selected). Cuffs were set to record daytime blood pressure every 30 min and nighttime blood pressure every hour at night. Unsuccessful recordings, defined as an incomplete inflation, were excluded from the data to ensure measurements were accurate. Participants were then categorized as “dippers” (n = 7; 4 F) if they experienced a ≥10% decrease in mean daytime to mean nighttime SBP or “non-dippers” (n = 11; 9 F) if they experienced <10% decrease or rise in SBP from daytime to nighttime [5, 24].

Participants reported to the Clinical Research Unit after an ∼10-h overnight fast. Participants sat semi-supine in a quiet, dimly lit room for approximately 5 min before clinical blood pressure was recorded using an automated platform (DINAMAP Procare 400; GE Medical Systems). Two intravenous catheters were then placed in the right arm for blood draws and insulin/glucose infusion, respectively. Insulin was infused at a primed (250 mU/m²/min) constant rate (40 mU/m²/min) for 120 min with a variable infusion of dextrose 20% to achieve a blood glucose concentration of 90 mg/dL. Blood glucose was analyzed every 5 min during the procedure. Insulin concentrations were assessed at 0, 90, 105, and 120 min and averaged during the last 30 min to reflect steady-state. Metabolic insulin sensitivity was defined as glucose infusion rate (GIR) during the final 30 min of the clamp. Baseline and 120-min brachial SBP and diastolic blood pressure (DBP) were assessed and standardized to the left arm, and data were averaged across three consecutive measurements at both time points, with a 1-min break between each recording. HR was also recorded at baseline and 120 min.

Endothelial function was assessed using flow-mediated dilation (FMD) at the brachial artery of the left arm using a 12–3 MHz range linear transducer approximately 5 cm proximal to the antecubital crease using B-mode ultrasound (Epiq 7C Ultrasound Machine; Philips Medical Systems, Andover, MA, USA). A blood pressure cuff was placed around the forearm and baseline images of the brachial artery were taken prior to inflation. The cuff was manually inflated to at least 50 mm Hg over the participant’s resting SBP for 5 min and then deflated. Thereafter, images were captured every 10 s post-deflation for 2 min to determine postischemic peak diameter and the time to reach peak diameter (time to peak). Measurements were collected at 0 min (baseline) and approximately 120 min (steady-state) of the clamp procedure. FMD (unscaled) was defined as the percent change in peak diameter compared to the baseline diameter as previously performed by our group [25, 26]. Vascular insulin sensitivity was defined as FMD at 120 min to mimic GIR. In addition, FMD was allometrically scaled to account for potential baseline diameter influence. All images were stored in Digital Imaging and Communication in Medicine format for analysis and images were assessed using commercially available software (Brachial Analyzer for Research v.6, Medical Imaging Applications LLC., Coralville, IA, USA).

Plasma glucose was measured immediately after collection using a glucose oxidase assay (YSI 2700, Yellow Springs, OH, USA). Sodium, potassium, chloride, calcium, BUN, and creatine were analyzed via the hospital medical laboratory. eGFR was estimated using chronic kidney disease (CKD)-EPI calculations accordingly for renal function from these analytes. Insulin was centrifuged at 4°C for 10 min at 3,000 rpm and stored at −80°C until analysis via radioimmunoassay.

Data were analyzed with SPSS Software (version 27, 2020) and graphs were produced via GraphPad Prism (version 9.2.0, 2021). Normality was assessed by Shapiro-Wilk tests. Outliers were assessed using the interquartile range method and sensitivity analyses were used to confirm removal. Independent two sample two-tailed t tests were performed to assess group differences when appropriate. A Fisher’s exact test was used to assess differences in sex and racial/ethnic distribution between groups. A two-way (group × time) ANOVA with repeated measures was used to assess group differences (dipper vs. non-dipper) during the clamp procedure (0-min/baseline vs. 120-min) and graded exercise test (stages 1-peak) when appropriate. Pearson correlations were used to examine the strength of associations among variables. Given that dippers had lower concentrations of high-density lipoprotein and low-density lipoprotein cholesterol, correlations were performed as partial correlations controlling for high-density lipoprotein and low-density lipoprotein, when appropriate. Effect sizes were calculated using Cohen D (baseline) and partial η² (group × time) when appropriate [27, 28]. We interpret Cohen D with 0.2, 0.5, or 0.8, and partial η² with 0.01, 0.06, and 0.14 representing small, medium, and large effect sizes, respectively. Data are reported as mean ± SD and significance was accepted as p < 0.05.

There were no significant differences in ATP III criteria, age, VO₂peak, and body composition between dippers and non-dippers (shown in Table 1), although VO₂peak effect sizes were large in dippers versus non-dippers. Similarly, there were no significant differences in blood sodium, potassium, chloride, creatine, calcium, BUN, or eGFR (shown in Table 1). There were no statistically significant differences in sex distribution (p = 0.33) and racial/ethnic distribution (p = 0.31) between groups.

The number of successful measurements recorded during the day (p = 0.84), night (p = 0.43), and 24-h period (p = 0.99) were similar between groups (shown in Table 2). As expected by design, dippers had a significantly greater drop in SBP from daytime to nighttime than non-dippers (16.63 ± 5.2 vs. 1.83 ± 5.6%, p < 0.01). Daytime SBP (153.4 ± 21.5 vs. 145.3 ± 21.9 mm Hg, p = 0.45) and nighttime SBP (127.4 ± 16.3 vs. 142.2 ± 19.5 mm Hg, p = 0.76) were not statistically different between dippers and non-dippers. Additionally, there were no significant differences in fasting SBP (134.1 ± 8.6 vs. 133.1 ± 7.4 mm Hg, p = 0.81) or DBP (81.1 ± 5.6 vs. 78.4 ± 5.9 mm Hg, p = 0.37) measured the morning of the clamp.

There were no group differences in fasting glucose or insulin concentrations as well as metabolic insulin sensitivity as reflected by GIRs (2.2 ± 0.6 vs. 2.8 ± 1.3 mg/kg/min, p = 0.22). While there were no group differences in glucose concentrations at 120 min of the clamp between dippers and non-dippers (2.2 ± 0.6 vs. 2.8 ± 1.3 mg/dL, p = 0.22), dippers had greater post-clamp insulin concentrations (n = 14 due to technical assay issues, 119.1 ± 14.2 vs. 75.5 ± 17.8 µU/mL, p < 0.01). There were no group differences in pre-occlusion brachial artery diameter at 0 min (4.2 ± 0.9 vs. 3.8 ± 0.8 mm, p = 0.35) or 120 min of the clamp (4.1 ± 0.8 vs. 3.8 ± 0.8 mm, p = 0.21). Similarly, there were no group differences in post-occlusion brachial artery diameter at 0 (4.3 ± 1.0 vs. 3.8 ± 0.8 mm, p = 0.21) or 120 min of the clamp (4.2 ± 0.9 vs. 3.7 ± 0.7 mm, p = 0.25, shown in Table 3). Time to peak was similar between dippers and non-dippers at both 0 min (n = 17, 52.4 ± 32.9 vs. 31.9 ± 22.5, p = 0.15) and 120 min (n = 15, 71.2 ± 33.2 vs. 61.4 ± 43.0, p = 0.63) of the clamp. There were no statistical differences in unscaled FMD at baseline (6.2 ± 2.1 vs. 6.0 ± 1.8%, p > 0.99) or 120 min of the clamp between dippers and non-dippers (7.0 ± 1.1 vs. 5.5 ± 1.6%, p = 0.15; shown in Fig. 1). Similarly, there were no differences in scaled FMD at baseline (6.72 ± 1.05 vs. 6.87 ± 0.81%, p > 0.99) or 120 min of the clamp between groups (7.00 ± 0.81 vs. 6.67 ± 0.61%, p = 0.83). However, it is worth noting that there were medium effect sizes for insulin to promote increases in FMD in dippers versus non-dippers (shown in Fig. 1). There were no differences in baseline SBP (129.9 ± 13.5 vs. 133.1 ± 7.4, p = 0.53) or DBP (77.2 ± 11.4 vs. 78.4 ± 5.9, p = 0.78) between dippers and non-dippers. Similarly, there were no differences in SBP (126.4 ± 7.2 vs. 134.5 ± 11.2, p = 0.13) or DBP (77.1 ± 8.3 vs. 82.3 ± 11.0, p = 0.33) at 120 min of the clamp between dippers and non-dippers. While baseline HR was similar between dippers and non-dippers during the clamp (62.0 ± 8.6 vs. 66.8 ± 9.9, p = 0.32), dippers had a significantly lower HR at 120 min (60.1 ± 7.0 vs. 72.0 ± 9.9, p = 0.02). Greater insulin-stimulated FMD correlated with increased SBP dipping correlated (r = 0.52, p = 0.03; shown in Fig. 1), but the relationship was no longer significant when controlling for HR slope (r = 0.42, p = 0.09). Additionally, the relationship was no longer significant when allometrically scaling FMD (r = 0.26, p = 0.30; shown in Fig. 1).

There was no significant difference in resting HR between dippers and non-dippers during the graded exercise test (71.6 ± 3.2 vs. 73.1 ± 7.2 bpm, p = 0.67). However, non-dippers had an elevated HR at both stage 3 (104.7 ± 5.6 vs. 119.9 ± 13.8 bpm, p = 0.03) and stage 4 (113.6 ± 5.3 vs. 132.8 ± 13.3 bpm, p < 0.01) of exercise (shown in Fig. 2). Non-dippers also had an elevated HR slope during exercise (6.5 ± 0.9 vs. 8.2 ± 1.7 bpm/min, p = 0.03). Further, SBP_slope was higher in non-dippers compared to dippers (4.7 ± 2.3 vs. 7.2 ± 2.5 mm Hg/min, p = 0.05; shown in Fig. 2). Non-dippers had an elevated RPP stage 4 (17,828.0 ± 3,769.9 vs. 23,129.2 ± 4,495.4 mm Hg/bpm, p = 0.05) and peak (29,254.7 ± 5,744.4 vs. 31,610.2 ± 12,241.4 mm Hg/bpm, p = 0.03; shown in Fig. 2) of exercise as well as a steeper rise in RPP slope (1,577.1 ± 298.5 vs. 2,277.7 ± 519.0 mm Hg/bpm/min, p < 0.01). HRR did not differ between dippers and non-dippers at 1 min (17.9 ± 7.1 vs. 20.0 ± 3.7 bpm, p = 0.43) nor 2 min after exercise (33.3 ± 8.0 vs. 38.2 ± 4.9 bpm, p = 0.13, shown in Fig. 2). VO₂ rose comparably across stage 1–4, although we observed a significant interaction seemingly from stage 4 to peak (p = 0.01; shown in Fig. 2). Interestingly, SBP dipping did not correlate with SBP responses during exercise (r = −0.14, p = 0.59) or HR during exercise (r = −0.14, p = 0.60).

The major finding from this study is that SBP dipping is paralleled by attenuated elevations in blood pressure and HR compared with non-dippers. Moreover, there were no statistical group differences in metabolic or vascular insulin sensitivity in patients with MetS despite medium effect sizes for both outcomes. These findings suggest insulin action on endothelial function and/or glucose metabolism is unlikely a primary factor in explaining blood pressure dipping, although additional work confirming these results are warranted based on effect sizes. Indeed, prior work in adolescents with obesity reported that decreased insulin sensitivity as measured via homeostatic model assessment of insulin resistance was associated with blunted dipping when adjusting for daytime SBP, age, and BMI [29]. Interestingly, when examining the role of pre/post-clamp insulin concentrations in the present study though in relation to SBP dipping (n = 14), higher post-clamp insulin levels were observed in dippers (shown in Table 3). Moreover, we detected a strong effect size difference between groups in circulating insulin among dippers. Given that insulin levels were higher in dippers during the clamp (due to possibly alterations in insulin clearance and/or synthesis), it would suggest that per unit of insulin our non-dippers were to our surprise more insulin sensitive than dippers. These findings are in contrast to previous work in adults with HTN and T2D that have shown insulin to increase daytime SBP through increased sympathetic stimulation and sodium retention [30, 31]. We do not have a readily available explanation for this observation in our study. As such, our collective work highlights, with no clinical differences in sodium, creatine, eGRF, or potassium that the elevated metabolic effects of insulin observed in dippers likely have little clinical consequence in this cohort on nocturnal blood pressure as previously noted [11, 15, 30].

Alternatively, it is reasonable to anticipate that endothelial function may relate to SBP dipping. FMD is a measure of shear rate-induced NO production independent of insulin-activated NO release through Akt/eNOS-related pathways [32, 33]. Although the findings of this study suggest that neither fasting nor insulin-stimulated FMD are statistically different between dippers and non-dippers, dippers had medium effect size results favoring time to peak during fasting as well as vascular insulin sensitivity. However, despite observing a significant correlation between insulin-stimulated FMD at 120 min of the clamp with SBP dipping, these results were no longer significant after controlling for HR slope. Our work appears partially at odds with recent work in patients with CKD that reported higher baseline brachial artery FMD in dippers compared to non-dippers, and that baseline/fasting FMD correlated with lower nighttime SBP [18]. Similarly, prior work in patients with HTN reported that carotid-femoral pulse wave velocity, the “gold standard” for measuring arterial stiffness, was associated with blunted SBP dipping [34]. Furthermore, another study examining brachial artery FMD and carotid-femoral pulse wave velocity in normotensive participants with MetS reported decreased endothelial function and increased arterial stiffness [35]. An important consideration between prior work and data presented here is that our MetS cohort had relatively high baseline FMD (∼6%) compared with others (∼3%–4%) [16]. This suggests that our patients had relatively normal fasting endothelial function, and shear rate-induced NO was not likely a primary factor contributing to SBP dipping. Our observation is consistent with recent work examining the role of microvascular function in relation to SBP dipping, whereby adults with stage-1 HTN exhibited normal fasting microvascular endothelial function [36]. While prior work reported elevated endothelin-1 expression, indicative of vasoconstrictive pathways, in patients with T2D compared to lean healthy controls under hyperinsulinemic conditions [37], no study to date has assessed insulin-stimulated FMD responses in relation to dipping status. Given medium effect size in dippers, it is speculative that insulin promoted Akt-related eNOS activation more than non-dippers [38]. Thus, our work expands on current literature as this is the first study to examine insulin-stimulated FMD in SBP dipping.

Another possible explanation is that SBP dipping relates to low sympathetic and high parasympathetic activity. In fact, Jeong et al. [18] reported greater muscle sympathetic nerve activity as determined by microneurography in non-dippers compared to dippers in patients with CKD. Similarly, a recent meta-analysis examining autonomic function and dipping patterns found that sustained sympathetic overdrive has been linked to essential HTN [39]. In the present study, we demonstrate that the rise in SBP and HR was elevated in non-dippers compared to dippers during exercise. Further, the higher RPP between dippers and non-dippers during the submaximal-graded exercise test is also of interest since VO₂ appeared similar, albeit peak fitness tended to be higher in dippers than non-dippers. While our work is consistent with views that autonomic function may be an important factor regulating ambulatory blood pressure in patients with MetS, future research should employ more direct measurements of sympathetic activity (i.e., MSNA or plasma catecholamines) to elucidate the relationship between blood pressure responses to exercise and sympathetic activity. The clinical consequence of greater submaximal SBP is that non-dippers may experience greater cardiac workload during exercise compared to dippers. However, it is important to acknowledge that previous work has suggested that an RPP above 30,000 is associated with a heightened risk for CVD [40] and both groups have an average peak RPP value at or above 30,000, which may be more reflective of their MetS status. In either case, we observed no statistical difference in HRR between groups, suggesting that parasympathetic activity was comparable between dippers and non-dippers. Collectively, we interpret the greater rises in HR and RPP during the graded exercise test in non-dippers to reflect heightened sympathetic activity [38, 41, 42]. Importantly, however, it should be noted that a graded exercise test is not a direct measure of sympathetic activity and additional work is warranted to directly measure sympathetic activation before and during insulin stimulation to confirm this hypothesis. This is clinically relevant because sympathetic overactivation has been linked to sudden cardiac arrest in a variety of conditions [39, 43, 44].

There are limitations to this study that warrant discussion. The sample population was relatively small and homogenous (∼94% White and ∼72% female [all postmenopausal]). This limits the generalizability of our results to the population as a whole. While previous work by Hyman et al. [45] reported similar dipping patterns across White, Black, and Hispanic populations [45], more racial/ethnic disaggregated data on AMBP is needed. It is also worth noting that muscle sympathetic nerve activity is a more direct measurement of sympathetic function than using a graded exercise test with HR and SBP measurements [46, 47], despite changes in HR being linked to insulin resistance and interpreted to reflect neural activity [48]. Notably, guidelines for recording AMBP suggest that a minimum of 8 measures should be recorded during sleep and 15 during wake time [49]. However, given that sleep measurements were taken once every hour and 10 participants (4 dippers, 6 non-dippers) reported getting less than 8 h of sleep, these individuals fell below this threshold. Additionally, FMD was not corrected for shear rate due to technical issues with obtainment of blood flow velocity. Nonetheless, FMD unscaled is an independent factor predicting CVD and provides clinical insight [50]. Additionally, since we did not continuously capture arterial diameter post-deflation, it is likely that we captured peak and not maximal postischemic diameter. Finally, participants in this study were dichotomized as dippers and non-dippers. We acknowledge that there are additional dipping subcategories known as extreme dippers (>20% drop in SBP) and risers (<0% drop in SBP) [5]. Risers, also known as reverse dippers, are most at risk for CVD and stroke compared to all other dipping profiles in older participants [51]. Extreme dippers also have an increased risk of CVD and are more susceptible to brain and cardiac hypoperfusion than other categories [52]. In the present study, three (2 F/1 M) non-dippers exhibited a rising pattern, and one (M) dipper exhibited extreme dipping patterns. However, removal of these participants did not affect the overall pattern of responses (data not shown). Regardless, future work should consider further characterizing these SBP dipping/non-dipping phenotypes to improve precision medical care.

In conclusion, our findings demonstrate that patients with MetS who experience a ≥10% dip in nocturnal SBP have greater SBP, HR, and RPP responses during exercise. While there were no statistical differences in fasting or insulin-stimulated FMD between groups, it is worth noting dippers had modest effects to raise insulin-stimulated endothelial function versus non-dippers. When taken with comparable HRR responses post-exercise, these overall findings suggest that sympathetic activity, and to a lesser extent vascular insulin sensitivity, may be an important factor modulating shifts from daytime to nighttime blood pressure in patients with MetS. Future work should consider direct assessments of sympathetic function, with and without insulin, to understand mechanisms explaining blood pressure dipping in MetS. Further work awaits investigation as well in examining how exercise, with or without pharmaceutical or dietary intervention, impacts nocturnal blood pressure to minimize CVD risk.

We thank the dedicated research assistants of the Applied Metabolism & Physiology Lab and participants for their effort. We also thank Dr. Eugene J. Barrett, M.D., Ph.D. as well as Ms. Linda Jahn, M.Ed., R.N. for training and assistance with ultrasound measures as well as the nursing staff of the Clinical Research Unit for technical assistance, and Ms. Lisa Farr, M.Ed. from the Exercise Physiology Core Lab for fitness testing.

This study protocol was reviewed and approved by the University of Virginia Institutional Review Board (IRB#19364, NCT03355469) and conducted according to the principles of the World Medical Association Declaration of Helsinki. All participants provided written and verbal consent to participate in the study.

The authors have no conflicts of interest to declare.

This study was supported by the National Institutes of Health R01-HL130296 (S.K.M.).

All authors reviewed and consented to publication. The paper was written by N.R.S. and S.K.M. with expert contributions by E.M.H., A.M.S., and P.K. All data collection, were contributed by E.M.H., S.L.M., A.C.B., U.S.C., and S.K.M. with FMD measurements and analysis specifically conducted by E.M.H. and S.L.M. All authors have agreed to submission and publication of the paper.

Availability of data and material is subject to permission by the contributing principal investigators of the study.