Habituation attenuates the sex-specific associations between ischemic pain, blood pressure, and arterial stiffness in young adults

Abstract

Cardiovascular reactivity (CVR) during physical stress is prognostic for incident cardiovascular disease. CVR is influenced by perceived pain. However, there is limited data on the effect of sex differences and repeated exposures to painful stimuli on CVR. We measured blood pressure (BP) and carotid-femoral pulse wave velocity (cf-PWV; an index of arterial stiffness) at rest, during isometric handgrip (HG) exercise at 30% of maximum voluntary contraction, and during postexercise circulatory occlusion (PECO) during two identical trials in 39 adults (20M/19F; 18–39 yr). We assessed participants’ perceived pain using a visual analog scale after the first minute of each stimulus. We collected BP during minute 2 of each stimulus and cf-PWV during minute 3 of each stimulus. In male participants, we observed moderate associations (Ps ≤ 0.023) between perceived pain and changes in brachial diastolic (ρ = 0.620) and mean BP (ρ = 0.597); central diastolic, mean, and systolic BP (ρs = 0.519–0.654); and cf-PWV (ρ = 0.680) during PECO in trial 1, but not trial 2 (Ps ≥ 0.162). However, in female participants, there were no associations between pain and CVR indices during either trial (Ps ≥ 0.137). Irrespective of sex, reductions in perceived pain during trial 2 relative to trial 1 were weakly to moderately associated (Ps ≤ 0.038) with reductions in brachial diastolic (ρ = 0.346), mean (ρ = 0.379), and systolic BP (ρ = 0.333); central mean (ρ = 0.400) and systolic BP (ρ = 0.369); and cf-PWV (ρ = 0.526). These findings suggest that 1) there are sex differences in pain modulation of CVR in young adults and 2) habituation blunts pain and CVR during PECO, irrespective of sex.

NEW & NOTEWORTHY We demonstrate sex differences in the association between pain perception and cardiovascular reactivity (CVR) during ischemic pain. We also demonstrate habituation to pain and reduced CVR during repeated exposure in a sex-independent manner. Accounting for sex differences and habituation may improve the prognostic utility of CVR.

INTRODUCTION

The cardiovascular reactivity (CVR) hypothesis proposes that individuals exhibiting exaggerated heart rate (HR) and/or blood pressure (BP) responses to stress are at greater risk for future cardiovascular disease (1). Though several prospective studies support this hypothesis (2–4), some findings do not support the prognostic value of CVR (5, 6). Acute pain during physical stress evokes a sympathetic nervous system response leading to increases in HR and BP (7) that contribute to the magnitude of CVR. Biological sex is suspected to account for variability in pain perception (8), with females typically demonstrating greater pain sensitivity (9, 10). However, females also typically demonstrate less drastic increases in BP during physical stimuli (11, 12). Thus, given a paucity of data (9, 10), there is a compelling rationale to examine whether there are sex differences in the association between pain perception and CVR. Furthermore, repeated exposure to pain-inducing physical stimuli can increase (sensitization) or decrease (habituation) subsequent pain perception (13). However, it is unclear whether there are sex differences in the adaptive processes to pain that may influence the association between pain and CVR.

The purpose of this study was to determine whether perceived pain is associated with indices of CVR [HR, BP, and carotid-femoral pulse wave velocity (cf-PWV)] during physical stress [isometric hand grip (HG) exercise] and an ischemic stimulus [postexercise circulatory occlusion (PECO)] in young males and females. We hypothesized that changes in CVR during HG and PECO would be positively correlated with perceived pain for male, but not female, participants based on a previous report that evaluated HR and BP, but not arterial stiffness (9). In addition, we sought to determine whether there are sex differences in changes in pain and/or CVR with repeated exposure to HG and PECO. Our secondary hypothesis was that repeated exposure to pain-inducing physical stimuli would attenuate pain perception (i.e., habituation) and consequently CVR, irrespective of sex.

METHODS

Findings from the present study were collected as part of a registered clinical trial (NCT05132556) with results that will be reported elsewhere. All procedures involving human subjects were approved by the Institutional Review Board at Georgia Southern University. The study was performed in accordance with the ethical standards of the Declaration of Helsinki. All participants provided written, informed consent to participate.

Participants

All participants were free of any overt cardiovascular, metabolic, or neurological disorders. Inclusion criteria included age between 18 and 39 yr and body mass index <40 kg/m². Exclusion criteria included current or recent use of pain or BP-modifying medication(s) within the past 6 mo. Forty individuals completed the study, but one female was excluded due to an abnormal heart rhythm for a final sample size of 39 (self-identified: 20 males, 10 White and 10 non-Hispanic Black; 19 females, 10 White and 9 non-Hispanic Black). Compared with female participants, male participants were taller (177 ± 6 vs. 163 ± 9 cm; P < 0.001), had a higher body mass (75 ± 11 vs. 65 ± 12 kg; P = 0.008), were older (23 ± 4 vs. 20 ± 2 yr; P = 0.002), and exhibited greater maximal handgrip strength (44 ± 9 vs. 31 ± 3 kg; P < 0.001). No sex difference in body mass index was detected (males: 24 ± 3 kg/m², females: 24 ± 5 kg/m²; P = 0.708).

Study Design

This study consisted of three visits. The initial visit included written consent followed by a review of medical history and anthropometric measurements. Maximal HG strength (left hand) was determined using the highest force from three efforts (Jamar HG dynamometer, Sammons Preston, Bolingbrook, IL). For all visits, testing was performed in the supine position with the shoulder adducted and neutrally rotated, the elbow flexed to 90°.

Study visits 2 and 3 (i.e., experimental trials 1 and 2) consisted of identical procedures. All testing was performed in a temperature-controlled (21–23°C) and dimly lit room (14). Participants reported for testing between 05:00 and 10:00 (matched ± 1 h between visits) following an overnight fast and avoidance of strenuous physical activity, caffeine, tobacco, over-the-counter medication, alcohol for 24 h, and water for 1 h (15). Following 10 min of supine rest, we collected at least duplicate measures of HR, BP, and cf-PWV and report the average herein.

Next, participants completed 3 min of isometric HG exercise at 30% of their maximal voluntary contraction. Ten seconds before the cessation of exercise, a brachial cuff was rapidly inflated to a suprasystolic pressure (∼240 mmHg) and the cuff remained inflated for 3 min to isolate the muscle metaboreflex and evoke an ischemic pain stimulus (i.e., PECO). After the first minute of both HG and PECO, participants were asked to provide a pain rating using a visual analog scale (0–10; 0 = no pain, 10 = most pain). HR and BP were acquired during the second minute and cf-PWV was acquired during the final minute of each stimulus. Experimental trials were separated by 2 to 14 days (male average = 7 ± 3 days, female average = 6 ± 3 days).

Cardiovascular Measures

We applied a brachial cuff to the participant’s right arm to acquire HR, brachial BP, and central BP using a SphygmoCor XCEL device (AtCor Medical, Naperville, IL). Central arterial stiffness was next estimated via cf-PWV, as previously described (16). All cf-PWV measures were obtained by one of two trained investigators (B.L.C. or J.D.V.), with only high-quality recordings (defined as intradevice quality index ≥80%) included in the analysis.

Circulating 17β-Estradiol

We used enzyme-linked immunosorbent assays (Enzo Life Sciences, Inc., Farmingdale, NY) to quantify endogenous concentrations of 17β-estradiol (kit sensitivity: 14 pg/mL). We collected plasma samples during both experimental trials in acid-citrate dextrose vacutainers. Samples were assayed in triplicate using a 1:4 dilution. The intra-assay coefficient of variation was 5.0%.

Statistical Analyses

Statistical analyses were performed using GraphPad Prism 9.3.1 and jamovi version 2.3.0. Descriptive characteristics are presented as mean and standard deviation (SD). Normality was confirmed through visual inspection of quantile-quantile plots. Resting cardiovascular measures were compared between sexes and trials using a two-way ANOVA. Spearman’s correlations were used to evaluate associations between perceived pain and change (Δ; post – pre) in cardiovascular measures during HG and PECO before and after adjustment for HG absolute contraction load (11). Pain ratings and cardiovascular responses (Δ; post – pre) during HG and PECO were compared between sexes and trials using two-way ANOVA or mixed-effects models (in the case of missing values). In all instances, when an ANOVA or mixed-effects model revealed a significant F value, post hoc Sidak pairwise comparisons were made. Spearman’s correlations were used to examine relations between changes in pain and changes in CVR between trials. For plasma 17β-estradiol, equivalence testing was conducted according to recent recommendations using the confidence interval method (17). Significance was set at an α-level of 0.05 for all analyses.

RESULTS

Resting Cardiovascular Measures

Table 1 provides resting cardiovascular measures during both trials. Male participants exhibited greater cf-PWV than females (P = 0.045), but this difference was no longer observed after controlling for age (P = 0.493). There was also a significant effect of trial indicating lower brachial (P = 0.011) and central (P = 0.027) diastolic BP and brachial mean BP (P = 0.027) during trial 2.

Table 1.

Resting cardiovascular measures

	Females			Males			P Values
	n	Trial 1	Trial 2	n	Trial 1	Trial 2	Trial	Sex	Interaction
Heart rate, beats/min	19	65 (7)	66 (11)	20	61 (8)	62 (8)	0.085	0.798	0.871
cf-PWV, m/s	19	5.7 (0.8)	5.6 (1.0)	19	6.1 (0.7)	6.1 (0.6)	0.615	0.045	0.638
Brachial BP, mmHg
Systolic	19	117 (12)	115 (12)	20	121 (7)	120 (8)	0.208	0.129	0.136
Diastolic	19	70 (9)	68 (7)	20	69 (7)	68 (6)	0.011	0.857	0.218
Mean	19	86 (10)	83 (8)	20	86 (7)	85 (6)	0.027	0.603	0.188
Central BP, mmHg
Systolic	19	103 (12)	101 (12)	20	104 (6)	105 (8)	0.298	0.409	0.095
Diastolic	19	71 (9)	68 (8)	20	70 (7)	69 (6)	0.027	0.873	0.175
Mean	19	84 (11)	81 (9)	20	83 (7)	83 (7)	0.054	0.997	0.122

Values are means (SD); n, number of participants. BP, blood pressure; cf-PWV, carotid femoral-pulse wave velocity. Boldface indicates significance.

Cardiovascular Responses during Pain-Inducing Physical Stress and Ischemic Pain

In trial 1 during HG, females and males exhibited increases in all measured cardiovascular variables (P < 0.01). During PECO, female participants exhibited significant increases (P < 0.05) in cf-PWV and BP, but not in HR (P = 0.081). In trial 2 during HG, female participants exhibited increased BP (P < 0.05), a trend for cf-PWV (P = 0.052) but not for HR (P = 0.666). During PECO, male participants exhibited increased cf-PWV and BP during both trials (P < 0.01), but HR did not increase in trial 1 (P = 0.091) or trial 2 (P = 0.379).

Associations between Perceived Pain and Cardiovascular Reactivity

In female participants, there were no associations between perceived pain and CVR before or after adjusting for HG absolute contraction load in trial 1 (unadjusted: Ps ≥ 0.360; adjusted: Ps ≥ 0.296), or trial 2 (unadjusted: Ps ≥ 0.291; adjusted: Ps > 0.259). Similarly, in male participants, there were no associations between perceived pain and CVR during trial 1 (unadjusted: Ps ≥ 0.079; adjusted: Ps ≥ 0.087) or trial 2 (unadjusted: Ps ≥ 0.282; adjusted: Ps ≥ 0.336).

Associations between perceived pain and CVR during PECO are displayed in Table 2. In female participants, there were no associations during trial 1 (unadjusted: Ps ≥ 0.137; adjusted: Ps ≥ 0.289) or trial 2 (unadjusted: Ps ≥ 0.502; adjusted Ps ≥ 0.390). However, in male participants, during trial 1 there were associations between perceived pain and changes in brachial diastolic (P = 0.005) and mean BP (P = 0.007); central diastolic (P = 0.002), mean BP (P = 0.008), and systolic BP (P = 0.023); and cf-PWV (P = 0.002). All of these associations remained (Ps ≤ 0.026) after adjusting for absolute HG contraction load, although associations between perceived pain and changes in cf-PWV were weakened after adjusting for changes in mean BP (ρ = 0.475, P = 0.054). In contrast, there were no associations between perceived pain and CVR during trial 2 in male participants (unadjusted: Ps ≥ 0.162; adjusted: Ps ≥ 0.149).

Table 2.

Associations between perceived pain during postexercise circulatory occlusion and cardiovascular responses (absolute Δ) before and after adjusting or handgrip absolute contraction load

	Females						Males
	Trial 1			Trial 2			Trial 1			Trial 2
	n	ρ	P	n	ρ	P	n	ρ	P	n	ρ	P
	Unadjusted
Heart rate, beats/min	18	0.079	0.756	19	0.107	0.662	19	−0.059	0.809	19	−0.051	0.836
cf-PWV, m/s	16	0.389	0.137	19	−0.082	0.739	18	0.680	0.002	17	0.355	0.162
Brachial BP, mmHg
Systolic	18	0.249	0.319	19	0.127	0.604	19	0.450	0.053	19	0.212	0.384
Diastolic	18	0.176	0.486	19	−0.122	0.618	19	0.620	0.005	19	0.234	0.335
Mean	18	0.138	0.572	19	0.001	0.997	19	0.597	0.007	19	0.252	0.298
Central BP, mmHg
Systolic	18	0.220	0.381	19	0.037	0.881	19	0.519	0.023	19	0.211	0.386
Diastolic	18	0.125	0.620	19	−0.164	0.502	19	0.654	0.002	19	0.290	0.229
Mean	18	0.189	0.453	19	−0.027	0.913	19	0.591	0.008	19	0.288	0.232
	Adjusted
Heart rate, beats/min	18	0.012	0.963	19	0.097	0.703	19	−0.054	0.831	19	−0.150	0.553
cf-PWV, m/s	16	0.293	0.289	19	−0.156	0.536	18	0.678	0.003	17	0.378	0.149
Brachial BP, mmHg
Systolic	18	0.047	0.858	19	0.117	0.645	19	0.449	0.062	19	0.124	0.624
Diastolic	18	0.008	0.976	19	−0.171	0.496	19	0.636	0.005	19	0.186	0.461
Mean	18	0.048	0.851	19	−0.036	0.886	19	0.609	0.007	19	0.255	0.306
Central BP, mmHg
Systolic	18	−0.001	0.997	19	−0.001	0.997	19	0.521	0.026	19	0.144	0.568
Diastolic	18	−0.069	0.794	19	−0.216	0.390	19	0.678	0.002	19	0.243	0.330
Mean	18	−0.003	0.992	19	−0.064	0.801	19	0.607	0.008	19	0.293	0.238

n, number of participants. BP, blood pressure; cf-PWV, carotid femoral-pulse wave velocity. Boldface indicates significance.

Effects of Repeated Exposure on Perceived Pain and Cardiovascular Reactivity

Pain ratings and CVR during HG are displayed in Fig. 1. For pain (Fig. 1A), a trial effect was observed (P = 0.017) whereby ratings were lower during trial 2, but no sex differences were observed. Regarding cardiovascular measures, adjusted models assessing diastolic and mean BP with baseline values as covariates did not influence statistical outcomes, thus unadjusted values are presented. There were attenuated increases in heart rate (Fig. 1B, P = 0.049); brachial systolic BP (Fig. 1D, P = 0.016), mean BP (Fig. 1F, P = 0.028); and central systolic BP (Fig. 1G, P = 0.011) and mean BP (Fig. 1I, P = 0.031) during trial 2 compared with trial 1. No other effects of trial (Ps ≥ 0.063), biological sex (Ps ≥ 0.520), or interactions (Ps ≥ 0.478) were observed. In addition, controlling for perceived pain did not influence statistical outcomes (P > 0.05). Owing to a lack of sex differences in pain perception or CVR responses, correlations between changes in pain and changes in CVR between trials were analyzed in the sample as a whole. No associations between trial changes in pain and changes in CVR were observed during HG (Ps ≥ 0.421).

Figure 1.

Perceived pain (A) and changes (Δ; post – pre) in heart rate (B), carotid-femoral pulse wave velocity (cf-PWV; C), brachial blood pressure (BP; D–F), and central BP (G–I) during isometric handgrip exercise in trial 1 (white bars) and trial 2 (gray bars).

Pain ratings and CVR during PECO are displayed in Fig. 2. Pain ratings were significantly lower during trial 2 (Fig. 2A, P = 0.001), but no sex differences were observed. Adjusted models assessing changes in diastolic and mean BP with baseline values as covariates did not influence study findings, thus unadjusted values are presented. There were attenuated increases in brachial (Fig. 2D, P = 0.012) and central (Fig. 2G, P = 0.022) systolic BP during trial 2 compared with trial 1. No other effects of trial (Ps ≥ 0.075), biological sex (Ps ≥ 0.063), or interactions (Ps ≥ 0.345) were observed, and controlling for perceived pain did not influence statistical outcomes (P > 0.05). In the sample as a whole, associations were observed between repeated exposure-related changes in pain and changes in brachial diastolic (ρ = 0.346, P = 0.031), mean (ρ = 0.379, P = 0.017), and systolic BP (ρ = 0.333, P = 0.038); central mean (ρ = 0.400, P = 0.012) and systolic BP (ρ = 0.369, P = 0.021); and cf-PWV (ρ = 0.526, P < 0.001); data not shown.

Figure 2.

Perceived pain (A) and changes (Δ; post – pre) in heart rate (B), carotid-femoral pulse wave velocity (cf-PWV; C), brachial blood pressure (BP; D–F), and central BP (G–I) during postexercise cuff occlusion in trial 1 (white bars) and trial 2 (gray bars).

17β-Estradiol and Perceived Pain

Female participants’ circulating 17β-estradiol was not different between trials (243 ± 145 vs. 227 ± 122 pg/mL; P = 0.702). Equivalence testing of 17β-estradiol measures between trials using the two one-sided test (TOST) revealed P = 0.115 for Δ_upper and P = 0.004 for Δ_lower (90% confidence interval = −17.3 to 52.4; equivalence zone = −42.5 to 42.5). During trial 1 17β-estradiol was associated with perceived pain during HG (ρ = −0.524, P = 0.021) but not PECO (ρ = −0.420, P = 0.073). During trial 2, estrogen levels were not related to pain during HG (ρ = −0.211, P = 0.400) or PECO (ρ = −0.066, P = 0.795).

DISCUSSION

The purpose of this study was to determine whether biological sex influences the association between perceived pain and CVR during physical stress and an ischemic pain stimulus. We also sought to determine whether repeated exposure to these stimuli would influence pain perception and/or CVR. The main findings from this investigation were that perceived pain during an initial testing session was associated with BP and arterial stiffness responses during ischemic pain in young males, but not in females. However, a dissociation of the association between perceived pain and CVR in males was observed during a second testing session. Finally, habituation of pain perception and CVR was observed with repeated exposure to these stimuli, independent of sex.

Findings from the present investigation, and others (9), demonstrate that there are sex differences in the association between pain and CVR. Though the underlying reason for this sex difference is uncertain, more variable responses to pain in females compared with males may contribute (18). Sex hormones have been identified as a possible source of pain-related variability (8), and several recent studies have found that females are more sensitive to pain during periods of low estrogen (19). Consistent with this notion, we observed an inverse association between plasma estrogen concentrations and perceived pain in female participants during trial 1. However, the dissociation of this association during trial 2 highlights the nuanced nature of this association, which is supported by rodent studies showing that while ovariectomy induces mechanical and thermal hyperalgesia (20), inhibiting estrogen receptors decreases pain threshold (21). Regardless, we observed no difference in plasma estrogen concentration in female participants during trial 1 and trial 2, suggesting that intracycle variations in estrogen likely did not influence study findings.

In contrast to the lack of sex difference in pain perception and CVR reported herein, previous literature has made a case for greater pain sensitivity (9, 10) and smaller increases in BP (11, 12) to physical stress in female versus male participants. However, a recent systematic review of research on biological sex and pain perception concluded that females and males have comparable ischemic pain thresholds (22). Moreover, CVR during physical stress (i.e., cold pressor test) has been demonstrated to be influenced by the appraised gender relevance of the stressor (23), with some studies showing a lack of sex difference in CVR and others demonstrating females exhibit greater CVR than males (24). Collectively, these mixed findings highlight the difficulty of untangling the role of biological sex from the complex interplay of biopsychosocial factors that may influence physiological responses to pain, and the need for more research in this area.

A particularly novel aspect of the present study was that we associated pain and consequent changes in arterial stiffness. The association between pain and changes in cf-PWV in our study is consistent with a previous investigation that demonstrated that pain elicited from thermal heat was associated with pulse transit time and pulse wave amplitude, a surrogate for arterial stiffness, in a cohort of young male and female adults (25). The exact mechanisms linking acute pain to elevated arterial stiffness are not entirely clear. However, the leading speculation is that acute pain leads to increases in sympathetic nerve activity (SNA) resulting in elevated arterial stiffness via increases in BP and subsequently distending pressure as well as adrenergic-mediated increases in smooth muscle tone (26–28). Although the current investigation did not entail direct recordings of SNA, it is well appreciated that acute pain (29) and ischemia (11) increase SNA. Thus, we speculate that in the present study central command and metaboreflex raised SNA and consequently BP and cf-PWV. Notably, associations between perceived pain and changes in BP and cf-PWV were exclusively observed during trial 1. Theoretically, the disassociation between pain and CVR observed during trial 2 may reflect heterogeneity in dynamic adaptive processes to pain, though this is hypothetical and warrants confirmation.

Consistent with our hypothesis, we observed habituation of pain during HG and PECO that was independent of sex. This observation is consistent with a previous study demonstrating that when thermal pain is applied to the same skin site habituation is the most likely adaptive outcome (13). Importantly, coinciding with reductions in pain, we report declines in BP and cf-PWV responses to PECO that were independent of sex but related to the degree of pain habituation. This habituation of CVR is consistent with work demonstrating reduced BP and HR responses to repeated mental stress in young males and females (30, 31). However, to the best of our knowledge, we are the first to demonstrate an association between the habituation of pain and CVR in response to pain-inducing physical stress. This finding extends upon previous work demonstrating single time-point associations between pain and CVR by demonstrating bidirectional changes in these measures. Thus, our cross-sectional findings highlight the importance of pain in modulating CVR during physical stress.

Temporal trends of cardiovascular health in young adults reveal a widening of sex differences characterized by an elevated and growing prevalence of high BP in young males (32). In the present study, though BP values were numerically higher in males than females, these differences were not statistically significant. Moreover, sex differences in arterial stiffness were also detected but appeared to be driven by age. This lack of clarity regarding sex differences in our cardiovascular outcomes may be explained by our relatively modest sample size. When evaluating our findings, we were somewhat surprised to see a sex-independent reduction in diastolic and mean BP during trial 2. To control for the possible influence of this confound we adjusted models comparing changes in these measures with baseline values but no differences relative to the unadjusted models were found. Nonetheless, future studies incorporating a similar battery of cardiovascular measures are encouraged to consider a baseline familiarization session to enhance measurement reliability.

Another potential limitation of our study is that we solely focused on associations between pain intensity and CVR. Pain induced by physical stimuli, such as exercise and ischemia, can be quantified in a variety of ways including intensity, threshold, and tolerance. However, pain intensity seems to be particularly influential when considering associations with CVR (9). Although this protocol did not consider interindividual differences in arm dominance, adjusting associations between pain and CVR by HG absolute contraction load did not influence study findings. Finally, we did not control for the menstrual phase by separating trials across cycles. Although the menstrual cycle may alter pain perception (33), a recent study found no influence of the menstrual cycle on ischemic pain perception (34). In studies reporting variability in pain perception across the menstrual cycle, an influence of gonadal hormone fluctuations on nociceptive processing is often identified as a principal contributor to this phenomenon. Importantly, our participants did not exhibit a difference in estrogen concentrations between trials, though complementary measures of progesterone would have provided greater certainty as to the lack of gonadal hormonal influence on pain perception between trials. Notable strengths of this investigation include the evaluation of central BP and cf-PWV responses, in addition to HR and brachial BP, during stress using valid and highly reliable equipment (35). Furthermore, we characterized pain and cardiovascular responses during exercise and ischemia.

We demonstrate sex differences in the association between pain perception and CVR during PECO, consistent with the previous research in the field (9). Extending upon these findings, we report associations between pain intensity and changes in central arterial stiffness. Interestingly, repeated exposure to these stimuli yielded a dissociation between pain and CVR, as well as habituation of pain perception and CVR that was independent of sex. Taken together, these findings indicate that 1) sex differences in the association between pain perception and CVR to pain-inducing physical stimuli and 2) habituation to pain and CVR observed with repeated exposure should be considered when evaluating CVR to physical stressors. Accounting for these factors may help to improve the prognostic utility of CVR as a means to forecast the development of disease.

GRANTS

This project was supported by a Georgia Southern University Internal Seed Grant (to G. J. Grosicki) and National Heart, Lung, and Blood Institute Grants K01HL160772 (to J. C. Watso) and K01HL147998 (to A. T. Robinson).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

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