INTRODUCTION

Accurate assessments of cardiovascular health and fitness are important for predicting at-risk individuals and for developing interventional therapeutic strategies (1). Accumulating evidence has established that clinical assessments of cardiorespiratory fitness improve risk stratification for adverse health outcomes and are a powerful tool for patient management (2–4). The American Heart Association released a statement indicating that cardiorespiratory fitness should be considered a clinical vital sign and should be assessed regularly in the clinic along with other preventative assessments (4). However, direct measures of cardiorespiratory fitness rely on determining peak oxygen uptake (V̇o_2Peak) during a graded exercise test conducted in a laboratory setting. Because measuring V̇o_2Peak during a graded exercise test can be costly and inconvenient, it is often more practical to estimate cardiorespiratory fitness using simple, noninvasive measures that are easily collected during exercise. One such measure is heart rate recovery (HRRec) after exercise testing, which is an index of vagal reactivation and is a strong predictor of morbidity and mortality in older adults (5–9).

HRRec is used in some clinical settings as a measure of autonomic dysfunction to identify high-risk cardiovascular disease patients. However, it is not typically used as a marker of cardiorespiratory fitness. For example, HRRec is predictive of long-term outcomes and survival in patients with coronary artery disease and congestive heart failure who have known autonomic nervous system dysfunction (7,8). The HRRec is also an independent risk factor for development of metabolic diseases, suggesting it is an informative marker for at-risk individuals (10,11). Evidence suggests that HRRec is a valid method to assess cardiorespiratory fitness. Cross-sectional studies report that physically active individuals have improved HRRec compared with their sedentary counterparts (12–15). Likewise, both V̇o_2peak and HRRec improve after an exercise regimen in longitudinal studies (16–18). In addition, HRRec and V̇o_2peak are highly associated in studies that include older adults, athletes, and physically active individuals (19–23). Together authors of these studies suggest that HRRec is a valid marker of cardiorespiratory fitness. However, authors of one study examined the association between cardiorespiratory fitness and HRRec in young, healthy, sedentary females and found no association between V̇o_2peak and submaximal HRRec (24). Therefore, despite evidence in other populations, the use of HRRec as an accurate assessment of cardiorespiratory fitness in young and sedentary but otherwise healthy (nonsmoking, nonhypertensive, nondiabetic) adults is unclear.

The purpose of this study was to investigate the validity of using HRRec as an estimate of cardiorespiratory fitness in young, sedentary adults by determining if a significant association exists between HRRec and V̇o_2peak during a maximal graded exercise test (GXT_max). It is well known that sedentary behavior is associated with cardiovascular disease risk factors (25). In the US, the amount of time young adults spent in sedentary behaviors increased by 12% from 2007 to 2016 (26). Thus, young adults are an emerging at-risk population, and valid clinical measures of cardiorespiratory fitness in this group are needed.

METHODS

The study was reviewed and approved by the University of Kentucky Office of Research Integrity Medical Institutional Review Board (16-0789-F6A), and participants provided written informed consent before inclusion in the study. Participants between the ages of 18 and 45 years were recruited for a previously reported intervention study (27). Each participant completed a Physical Activity Readiness-Questionnaire and Health History Form and were excluded if he or she had existing contraindications to the GXT_max as specified in the American College of Sports Medicine Guidelines for Exercise Testing and Prescription (28).

Anthropometric and body composition measures, including standing height, body mass, circumference measurements, and a total-body dual-energy x-ray absorptiometry (DXA) scan were performed. Participants were measured in lightweight clothing containing no metal and without shoes. Standing height was determined to the nearest 0.1 cm from a wall-fixed meter stick (Pittsburgh, Pennsylvania) with the participants' hands positioned on the hips during a maximal inhalation. Body mass was determined to the nearest 0.1 kg using a calibrated electric scale (Escali Corp., Burnsville, Minnesota). Circumference measurements (waist, abdominal, and hip) were taken in triplicate using a fiberglass anthropometric tape (Creative Health Products BMS-8) in accordance with the guidelines established by the Airlie Conference Proceedings (29). The mean of the 3 measures was used for analysis. Body composition was measured using total body DXA scans performed using a GE Lunar iDXA bone densitometer (Lunar Inc., Madison, Wisconsin). All female participants had negative urine pregnancy tests, which were taken immediately before DXA scanning. A single trained investigator completed and analyzed all scans using the Lunar software, Version 14.10. Total body DXA absolute fat-free mass (kg), mineral-free lean mass (kg), and absolute (kg) and relative (% of body mass) fat masses were determined for each participant.

GXT_max tests were completed using an indirect calorimetry testing system (Vmax Encore, Vyaire Medical, Yorba Linda, California) with an integrated electrocardiogram (ECG; 60 Hz sampling rate; Cardiosoft v6.51, GE Healthcare, Chicago, Illinois) and a treadmill ergometer. Before the exercise test, baseline heart rate (HR) and blood pressure (BP) were measured while the participant stood on the treadmill. During the test and recovery period, continuous cardiovascular measurements (HR and ECG) were monitored. During the continuous, progressive (speed and grade) tests, oxygen consumption (V̇o₂) was measured breath by breath and averaged over 1-min intervals. The GXT_max tests were performed using an incremental treadmill protocol, with 2-min workload stages, developed for a previous study (30). Authors of prior studies have shown that similar V̇o_2peak and maximal HR are achieved regardless of treadmill protocol (ramp versus incremental) (31,32). The initial stage of the test began with a walking speed of 3.2 mph (5.1 kph) and 0% grade. The test progressed with a 0.4 mph (0.6 kph) increase in speed and 2% increase in grade with each subsequent stage. During the final minute of each stage, BP (by manual auscultation) and ratings of perceived exertion (RPE; using the original 6–20 Borg Scale (33)) were recorded. HR was recorded in the last 10 s of each stage. The test was terminated for all participants at volitional fatigue (3–4.5 s for treadmill to fully stop). Verbal encouragement was given throughout the test. After completing the GXT_max, 5 min of passive recovery data (HR and BP) were taken while the participants remained standing on the treadmill. Achievement of V̇o_2peak (mL·kg⁻¹·min⁻¹) was defined as a participant's ability to obtain a minimum of 2 of the following criteria: respiratory exchange ratio ≥1.1 (determined by 1-min averaging), RPE ≥ 17, and/or age-predicted maximal HR achieved or exceeded (34). The highest V̇o₂ value observed during GXT was used for analysis.

Our primary analysis included HRRec data at 1, 2, and 3 min postexercise. Absolute HRRec (b·min⁻¹) was defined as the HR at 1, 2, and 3 min postexercise subtracted from the maximal HR achieved during the graded exercise test. Relative HRRec was defined as absolute HRRec divided by maximal HR, multiplied by 100. Also, the difference between 1- and 2-min HRRec was calculated as an index of the slow component of the postexercise HRRec (35).

STATISTICAL ANALYSIS

Data were analyzed using SPSS (Version 26, IBM, Armonk, NY). Descriptive data are presented as means ± SD. Independent sample t tests were conducted to assess differences in measured variables between males and females. Pearson's correlation coefficients were used to assess associations between V̇o_2peak and relative and absolute HRRec as well as measures of body composition. Correlation analyses were stratified to assess potential sex differences. Since obesity could influence the analysis, a secondary sensitivity analysis was performed to determine if adiposity influenced the findings. Pearson's partial correlation coefficients were used to assess associations between V̇o_2peak and relative and absolute HRRec in participants while controlling for body fat percentage. A one-way repeated measures analysis of variance (ANOVA) was performed to determine if relative and absolute HRRec differed between 1, 2, and 3 min postexercise. Significance was ascribed at P < 0.05.

RESULTS

All participants reported good health, including no known cardiovascular disease or hypertension, were medication free (except contraceptives), and did not participate in a structured exercise regimen at the time of the study or participate in greater than 2 h moderate-vigorous physical activity each week. Eleven of the 72 participants were excluded because they did not achieve V̇o_2peak (details below). The remaining 61 participants included in the analytic dataset (40 females) had varying adiposities (body mass index (BMI): 16.6–37.0 kg·min⁻²; body fat percentage: 12.4–51.7%). Males had significantly greater height, body mass, and waist circumference than females (Table 1).

TABLE 1. Participant characteristics and anthropometric measures.

View Fullscreen

During the GXT, all participants in the analytic dataset, except 3 females, were above an absolute HRRec of 18 b·min⁻¹, a cutoff value for abnormal 1-min HRRec observed in a previous study (Supplemental Table 1; https://dx.doi.org/10.6084/m9.figshare.14691099) (36). V̇o_2peak was not associated with absolute HRRec at 1, 2, or 3 min when males and females were examined separately or when the entire cohort of participants was combined (Table 2; Figure 1). Absolute HRRec was also not significantly associated with V̇o_2peak at 1, 2, or 3 min when controlling for body fat percentage. Increased absolute HRRec at 3 min was associated with an increased waist circumference for the total study group only. However, absolute HRRec at 1, 2, and 3 min were not significantly associated with age, BMI, or other anthropometric measures for males, females, or the total study group (Table 2). Males had a significantly greater V̇o_2peak than females (Table 3). The difference between absolute HRRec at 1 and 2 min was also not associated with V̇o_2peak in males, females, or the total study group.

TABLE 2. Pearson correlation coefficients among absolute HRRec, V̇o_2peak, and anthropometric and body composition measures.

View Fullscreen

TABLE 3. V̇o_2peak, HR, and BP responses to maximal graded exercise test.

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

Age (r = −0.28; P = 0.03), BMI (r = −0.47; P < 0.01), and waist (r = −0.27; P = 0.04), abdominal (r = −0.43; P < 0.01), and hip circumferences (r = −0.46; P < 0.01) were each negatively associated with V̇o_2peak for the total study group. Males, compared with females, had higher HR at peak exercise as well as higher systolic and diastolic BP at baseline, peak exercise, and during recovery (Table 3).

Body composition varied in the cohort (body fat percentage range: 12.4–51.7%). Increased absolute HRRec at 1 min was associated with increased fat-free mass in the total study group only. However, fat mass, body fat percentage, and mineral-free lean mass were not associated with absolute HRRec in males, females, or the total study group (Table 2). Increased body fat percentage and fat mass were associated with a reduced V̇o_2peak in males and females separately as well as the total study group. Also, increased mineral-free lean mass and fat-free mass were associated with an increased V̇o_2peak for the total study group only (Figure 2). Males had greater fat-free mass and mineral-free lean mass than females (Table 1). Additionally, females had greater body fat percentage than males (Table 1).

View Figure

• Download as Powerpoint
• Download Figure

Since peak HR varied within the cohort (range: 157–210 b·min⁻¹), we also examined the relationship between V̇o_2peak and relative HRRec, which accounts for variability in peak HR. Relative HRRec at 1, 2, or 3 min were not significantly associated with measures of V̇o_2peak when males and females were examined separately or when the entire cohort of participants was combined (Figure 1, Table 4). Relative HRRec was also not significantly associated with V̇o_2peak at 1, 2, or 3 min when controlling for body fat percentage. Greater relative HRRec at 3 min was associated with increasing age and increasing fat mass for the total study group only. However, relative HRRec measures were not associated with BMI or any other anthropometric and body composition measure for males, females, or the total study group (Table 4).

TABLE 4. Pearson correlation coefficients among relative HRRec, V̇o_2peak, and anthropometric and body composition measures.

View Fullscreen

DISCUSSION

HRRec is recognized as a prognostic measure and predictor of mortality in older adults (5,6,9). HRRec and V̇o_2peak are used to inform clinical practices in older adults because they are associated with health and longevity (2,3,9). Furthermore, authors of numerous studies have shown that HRRec and V̇o_2peak are greater in physically active than sedentary participants and after completing various exercise regimens, suggesting an important physiological relationship between HRRec and cardiorespiratory fitness (14,16,17,37). However, the utility of HRRec as an indicator of cardiorespiratory fitness may vary by population and has not been well studied in young, sedentary adults.

Consistent with this study, Tonello et al. (24) also examined the association between cardiorespiratory fitness and HRRec in young (mean age: 34.5 years) adults not participating in structured exercise and found no association between V̇o_2peak and 1-, 2-, 3-, and 5-min HRRec. While we and Tonello et al. (24) both studied young sedentary adults, the 2 studies used somewhat different methods. Our study included both sexes and determined V̇o_2peak and HRRec from treadmill exercise testing, while Tonello et al. (24) studied only females and used a cycle ergometer. Maximal oxygen uptake is lower when measured by cycle ergometer than a treadmill (38,39). Tonello et al. (24) used a submaximal test to measure HRRec (that induced a HR at 86% age-predicted max), while our participants performed a GXT_max to volitional fatigue for determination of HRRec. Thus, despite distinct methodological differences, the fact that our 2 studies had similar results is strong evidence that HRRec may not be a valid indicator of cardiorespiratory fitness in young, sedentary adults.

In this study, we examined HRRec at 1, 2, and 3 min, as these are the measures that have been associated with cardiorespiratory fitness in other populations (19,23,40). HRRec after exercise is orchestrated by both the parasympathetic and sympathetic branches of the autonomic nervous system (41). Parasympathetic reactivation is predominately responsible for the decrease in HR immediately after exercise, while sympathetic withdrawal occurs more gradually (41,42). For this reason, authors of previous studies have considered HRRec at 1 and 2 min an indicator of vagal reactivation (5,6). In our study of young sedentary adults, HRRec at 1, 2, and 3 min was not significantly associated with cardiorespiratory fitness in our total analytic dataset or when stratified by sex. Thus, vagal reactivation may not be a reliable indicator of cardiorespiratory fitness in this population.

Our study cohort spanned a large range of adiposity, and 47% of participants were overweight or obese (BMI ≥ 25 kg·min⁻²). Previous studies found that increased obesity is associated with reduced 1-min HRRec (43,44). In contrast, our data revealed that an increased waist circumference was associated with a greater absolute 3-min HRRec, and increased fat mass was associated with a greater relative 3-min HRRec. However, this unexpected finding may be due to the difference in physiological significance of the 3-min HRRec measure compared with the 1 min (i.e., sympathetic withdrawal versus parasympathetic reactivation). Also, our previous data showed that HRRec after a GXT_max was similar in healthy-weight and obese children, indicating that young individuals with poor body composition can have normal vagal reactivation after exercise (45).

Physical activity status may be another factor influencing the relationship between V̇o_2peak and HRRec. Authors of studies in young adults, which included both physically active and sedentary participants, reported a significant association between V̇o_2peak and HRRec (22,23,40). Authors of studies have also found that subjectively and objectively measured physical activity were associated with HRRec (23,24). Although our subjects did not participate in structured exercise, incidental activity may have influenced HRRec, as shown by Tonello et al. (24). In fact, fat-free mass, which is affected by sedentary behavior (46), was associated with 1-min HRRec in our cohort. Age may also be an important factor since a significant association between V̇o_2peak and HRRec after a maximal treadmill test was observed in older adults with congestive heart failure (19). Thus, sedentary behavior and young age appear to be important contributing factors when determining if HRRec is a valid indicator of cardiorespiratory fitness.

We also found that V̇o_2peak was associated with body fat and fat-free body composition measures. Although V̇o_2peak is expressed relative to body mass, the composition of body mass varies. In agreement with our findings, previous researchers have reported that greater body fat percentage was associated with reduced V̇o_2peak and greater fat-free mass was associated with increased V̇o_2peak (47,48).

Some limitations of our study existed. First, our HRRec measures were collected during passive recovery while participants remained standing. Authors of other studies collected either active recovery measures or passive recovery measures while participants were seated or lying down (6,23,36). However, immediately moving participants to a seated or supine position after a maximal exercise test can be difficult in practice. Since our goal was to inform clinical utility, we collected measures while participants remained standing. Second, our sample size was small with varying adiposities. However, since we designed this study to inform on clinical utility of HRRec in the general population, our inclusion criteria included young, relatively healthy, and sedentary individuals, and no exclusion criteria regarding obesity status were implemented. Third, we did not control for dietary supplements that may have been consumed during the study. Fourth, we did not control for the phase of the menstrual cycle when the GXT_max was performed for female participants. It is possible that phase of the menstrual cycle may influence maximal oxygen uptake (49).

Since the clinical utility of HRRec was first introduced in the late 1990s, many authors have investigated HRRec as a measure of cardiovascular health and fitness in a variety of populations (5,20,23,24). HRRec has been shown to be a useful diagnostic and prognostic tool for coronary artery disease, heart failure, and mortality in older adults, including healthy participants and heart failure patients (5–7). However, few studies have been performed in young, sedentary adults, which is a rapidly expanding and at-risk population. Valid measures of cardiorespiratory fitness are needed to identify at-risk individuals at a young age and develop interventional therapeutic strategies.

CONCLUSION

The prevalence of cardiovascular disease among US adults is a staggering 49% of the population (50). We found that HRRec measures were not significantly associated with V̇o_2peak in a sample of young, sedentary adults. While HRRec measures have been used as a clinical indicator of health and morbidity in other populations, they are not a reliable indicator of cardiorespiratory fitness in sedentary young adults.