Journal Logo


Correlates of Heart Rate Recovery over 20 Years in a Healthy Population Sample


Author Information
Medicine & Science in Sports & Exercise: February 2012 - Volume 44 - Issue 2 - p 273-279
doi: 10.1249/MSS.0b013e31822cb190
  • Free


HR recovery (HRR) after graded exercise treadmill testing (GXT) is an established measure of autonomic nervous system function. Specifically, the rapid decline in HR after cessation of GXT is attributable to parasympathetic reactivation (28). Impaired parasympathetic reactivation, as indicated by a smaller difference between maximum HR and HR 1 or 2 min after exercise (i.e., slow HRR), is associated with elevated all-cause mortality, independent of other exercise test parameters (10,11,25). Most of these findings are based on symptomatic populations of middle-aged and older adults referred for exercise testing because they had symptoms of CHD. Less is known about the development of slow HRR over time in healthy young adults.

We hypothesized that the development of slow HRR in middle age is correlated with CHD risk factors (e.g., hypertension, diabetes) that have previously demonstrated an association with autonomic nervous system dysfunction. Prior population studies report that impaired autonomic function is associated with lower socioeconomic status (29), negative affect (e.g., depressive symptoms), physical inactivity (7,26), and obesity. Consequently, the goal of the present study was to investigate the association of a comprehensive set of sociodemographic characteristics, health behaviors, and clinical characteristics with the development of slow HRR in healthy young adults who were followed during 20 yr.


Study Population and Design

The Coronary Artery Risk Development in Young Adults (CARDIA) study is a longitudinal cohort of the evolution of CHD risk factors in a population sample of 5115 adults aged 18–30 yr at baseline (1985–1986). Black and white men and women were recruited from Birmingham, AL; Chicago, IL; Minneapolis, MN; and Oakland, CA. Participants were reexamined 2, 5, 7, 10, 15, and 20 yr after baseline; retention rates across examinations were 91%, 86%, 81%, 79%, 74%, and 72%, respectively. A detailed description of study design, sampling, and response rates is published (17). The sample for the present report was drawn from participants in the CARDIA Fitness Study, an ancillary investigation conducted in conjunction with the year 20 examination in 2005–2006 that measured fitness using a symptom-limited maximal treadmill test. Our study sample is composed of the 2730 participants who completed treadmill testing at baseline and the 20-yr follow-up examination and who did not have slow HRR at baseline. Institutional review boards at each study site reviewed the protocol and procedures and approved the research. All participants provided written informed consent.


Standardized protocols for data collection were used across study centers and examinations. Participants were asked to fast for at least 12 h before examination and to avoid smoking or engaging in heavy physical activity for at least 2 h before the examination.


Participants who did not have a history of ischemic or congenital heart disease, who were not taking cardiovascular medications other than blood pressure medications, whose blood pressure was less than 160/90 mm Hg, and who were not currently experiencing a febrile illness were eligible to complete treadmill testing. Procedures were the same at baseline and 20-yr follow-up and have been previously described (30). In brief, participants underwent a symptom-limited GXT using a modified Balke protocol to measure aerobic fitness. The protocol included up to nine 2-min stages of progressively increasing difficulty, and participants were encouraged to exercise as long as possible to maximum exertion. At the end of each stage, HR and blood pressure were measured, and participants were asked to rate their level of exertion on the Borg scale (3). Pretest HR was the HR recorded as participants were standing and before the initiation of the treadmill test. Peak HR was determined as the highest value recorded across all stages. HRR was calculated as the difference between peak HR and HR 2 min after test cessation. HRR of <22 beats·min−1 was defined as “abnormal” (28). We chose to use HRR at 2 min to represent parasympathetic reactivation because our protocol involves an active walking cooldown period. By contrast, at 1 min after exercise, the treadmill is still slowing down and lowering, which can introduce variability in the actual speed and grade depending on how long participants exercised on the treadmill (24). The speed and grade are more uniform at 2 min after peak exercise (28). Change in HRR during follow-up was calculated as the difference between HRR at the year 20 examination and baseline. Relative change in HRR was determined by dividing the absolute change by the baseline HRR. Fitness is estimated by the duration of the treadmill test, which is directly correlated with V˙O2max (2). Age-predicted maximum HR was determined using the Tanaka formula (208 − 0.7 × age) (32); participants were determined to have exercised to their maximal capacity if they achieved a peak heart rate that is >85% of their age-predicted HR.

Other measures

Age, race, and sex were self-reported, and medication use was determined using standard questionnaires. Height and weight were measured with a vertical ruler and a calibrated scale, respectively. Body mass index (BMI) was calculated as weight (kg)/height (m2); overweight and obesity were determined at BMI of 25–30 and ≥30 kg·m−2, respectively. Waist circumference was measured laterally midway between the iliac crest and the lowest lateral portion of the rib cage and anteriorly midway between the xiphoid process of the sternum and the umbilicus. Cigarette smoking was assessed by a standardized questionnaire at each examination, and respondents were categorized as current, former, or never smokers. Physical activity was assessed with the CARDIA physical activity questionnaire, an interviewer-administered instrument that asks participants to report the frequency of participation in 13 different categories of recreational sports and exercise in the past 12 months. Scores were computed by multiplying the frequency of participation by estimated intensity of activity and reported as “exercise units.” The reliability and validity of the instrument are comparable to those of other activity questionnaires (21,22). Depressive symptoms were assessed at year 5 using the Centers for Epidemiologic Studies Depression Scale (CES-D) with a score range of 0 to 60; symptom scores >16 are correlated with diagnosed depression. Changes in continuously measured health behaviors were determined as the difference between baseline measurements (1985–1986, except depressive symptoms, which were measured in 1990–1991) and measurements at final follow-up. To evaluate smoking behavior during 20 yr, we began by categorizing participants who reported smoking at year 20 as “current smokers,” participants who reported smoking at any other examination but not at year 20 as “former smokers,” and those who reported not smoking at any examination as “never smokers.”

After participants rested in a quiet room for 5 min, blood pressure was measured from participants in the seated position three times; the last two readings were averaged. Hypertension was determined if any of the following criteria was met: systolic blood pressure ≥140 mm Hg or a diastolic blood pressure ≥90 mm Hg on any visit or reported use of antihypertensive medication. For our change analysis, incident hypertension was defined in participants who were free of hypertension at baseline and who met the criteria for hypertension (defined above) at any subsequent visit.

Blood samples for measurement of glucose (16) and lipids (15,33) were collected according to standardized CARDIA procedures (17) and processed at central laboratories. Glucose was assayed at a central laboratory using the hexokinase method. Diabetes was determined when participants met any of the following criteria: measured fasting glucose levels (≥7.0 mmol·L−1) at examination years 7, 10, 15, or 20; self-report of oral hypoglycemic medications or insulin (all examinations); or a 2-h postload glucose (≥11.1 mmol·L−1) at examination years 10 and 20. For our change analysis, the incidence of diabetes during 20 yr was determined among participants who did not have diabetes at baseline on the basis of fasting glucose levels.

Metabolic syndrome was determined according to the American Heart Association modification (19) of the National Cholesterol Education Program Adult Treatment Panel III definition (14) if any three of the following criteria were met: 1) fasting glucose ≥5.7 mmol·L−1, 2) waist circumference >88 cm (women) or >102 cm (men), 3) SBP ≥130 mm Hg or DBP ≥85 mm Hg, 4) triglycerides ≥1.7 mmol·L−1, and 5) HDL cholesterol <1.3 mmol·L−1 in women or <1.04 mmol·L−1 in men. Participants who reported using medications for diabetes or hypertension control were classified as having met the criterion for elevated glucose or blood pressure, respectively. For our change analysis, incident metabolic syndrome was identified in participants who did not have metabolic syndrome at baseline.

Statistical Analyses

We described baseline characteristics (1985–1986) stratified by status of slow HRR at year 20 in 2005–2006 and compared values using t-tests for continuous variables, chi-square tests for most categorical variables, and Fisher exact tests for categorical variables with a low prevalence (e.g., hypertension, diabetes, and metabolic syndrome). Next, we compare distributions of the relative change in HRR during 20 yr by sex, race, and race–sex. Before calculating the odds of developing slow HRR during 20 yr, we tested whether selected characteristics at baseline, namely, demographic characteristics (i.e., age category, sex, and race), baseline HRR below the median (among those ≥22 beats·min−1 at baseline), hypertension, and diabetes, modified the association of baseline risk factors with incident slow HRR using multiplicative interaction terms. The only factor that exceeded our threshold of P < 0.05 for meaningful interaction was baseline glucose. However, we report stratified findings for glucose in the text. For all other covariates, we present odds ratios (OR) and 95% confidence intervals (CI) of having slow HRR at year 20 per SD from the mean (i.e., standardized changes) or according to a referent (categorical variables) and adjusted for baseline HRR. Multivariable models included all independent baseline characteristics in a single model. Estimates for waist circumference and diastolic blood pressure were generated in separate multivariable models that did not include BMI and SBP, respectively. We calculated the odds of having slow HRR at follow-up by changes in clinical measurements and health behaviors during 20 yr, adjusted for baseline demographic characteristics and HRR. ORs and 95% CIs for changes in continuous measures were calculated per SD increase from the mean over time. Odds for incident hypertension, diabetes, and metabolic syndrome were generated in the subsets of participants who did not have those conditions at baseline (n = 2679, 2632, and 2681, respectively). Estimates with a P value < 0.05 were considered statistically significant. All analyses were conducted using SAS version 9.2 (SAS Institute, Cary, NC).


From among the 2788 participants who had valid HRR measures at baseline and the 20-yr follow-up, 58 (2.1%) had slow HRR at baseline. After excluding those participants, we had 2730 in our study sample, of whom 56% were female and 44% were black, and the mean age was 25.1 yr (Table 1). The distribution of health behaviors and clinical characteristics was similar to that of the full CARDIA cohort (17). After 20 yr, 135 participants (5%) had slow HRR (HRR < 22 beats·min−1). Participants who had slow HRR in 2005–2006 were significantly more likely to be female and black, and a lower proportion had less than a high school education. As compared with participants who had normal HRR, participants who had slow HRR had higher baseline BMI and a larger waist and had a higher prevalence of hypertension and diabetes. In addition, participants with slow HRR in 2005–2006 were more likely than their counterparts with normal HRR to smoke at baseline, to report more depressive symptoms and less physical activity, and to exercise for less time on the treadmill.

Baseline characteristicsa of all participants with baseline HRR ≥22 beats·min−1, stratified by incident slow HRR in 2005–2006.

Between baseline and year 20, HRR declined from a mean ± SD of 44 ± 11 beats·min−1 at baseline to 40 ± 12 beats·min−1 (an average of 3.6 beats·min−1) in the total sample. When changes in HRR were calculated relative to the baseline levels, men and black participants experienced greater declines. The relative change in men as compared with women was −7.4% ± 27.8% versus −3.4% ± 31.6%, respectively (P < 0.05). Black participants experienced significantly (P < 0.01) larger relative declines in HRR (−9.0% ± 31.1%) than white participants (−2.1% ± 28.7%), and across race–sex groups, black men (−11.7% ± 27.2%), black women (−7.2% ± 33.4%), and white men (−4.5% ± 27.8%) experienced relative declines from their baseline value, whereas white women did not change relative to baseline (0.0% ± 29.4%) (P < 0.01). Findings were similar when we restricted analyses to the cohort of participants who exercised to at least 85% of their age-predicted HR (data not shown). Despite differences in the rate of change in HRR by race and sex, testing for effect measure modification in the relationship of baseline clinical factors and health behaviors by race and sex yielded no statistically significant differences. All results are presented pooled by race and sex group.

Baseline characteristics that were associated with the presence of slow HRR at year 20, taking into account baseline HRR, are reported in Table 2. Male sex, education, physical activity levels, and treadmill test duration were inversely associated with the odds of having slow HRR in 2005–2006, whereas black race, waist circumference, BMI, overweight and obesity (relative to normal weight), fasting glucose, hypertension, metabolic syndrome, high depressive symptoms, and current smoking were associated with higher odds of having slow HRR. There was evidence of significant effect measure modification for the association of baseline glucose by baseline HRR. At year 20, 99 participants who had HRR below the median at baseline had developed slow HRR as compared with only 36 from among those with HRR above the median at baseline. Baseline glucose was significantly positively associated with incident slow HRR in participants whose HRR was at or above the median at baseline (OR = 1.50, 95% CI = 1.15–1.96 per 0.58 mmol·L−1 (SD) higher), whereas there was no association in participants whose HRR was below the baseline median (OR = 1.04, 95% CI = 0.87–1.24 per SD). In multivariable models, the baseline characteristics that remained statistically significantly associated with higher odds of having slow HRR were BMI, baseline fasting glucose, and current smoking. In a separate multivariable model that included waist circumference instead of BMI, waist circumference also remained positively associated with HRR after adjustment for other factors. In a separate multivariable model that did not include SBP, the association of DBP with development of slow HRR was null (OR = 1.14, 95% CI = 0.94–1.40). Years of education remained inversely associated with the odds of having slow HRR in multivariable models.

Baseline characteristicsa associated with slow HRR at year 20 examination (OR, 95% CI).

The odds of having slow HRR at year 20 were significantly higher per SD increase in BMI and waist circumference (OR = 1.33 and 1.55, respectively) and with the incidence of diabetes, hypertension, and the metabolic syndrome and among current smokers (OR = 2.04, 2.06, 2.78, and 2.27, respectively) after adjustment for age, race, sex, education, and baseline HRR levels (Table 3). By contrast, the odds were lower as percent change in treadmill test duration and physical activity change increased during follow-up (OR = 1.35 and 3.52, respectively). Although baseline depressive symptom scores were positively associated with slow HRR at year 20, increasing depressive symptoms during 15 yr were not associated with higher odds of having slow HRR.

Adjusted associationa of changes in risk factors during 20 yr and HRR in 2005–2006.


During 20 yr, HRR declines in healthy young adults; however, adults who have adverse CHD risk factor profiles (e.g., higher fasting glucose, higher depressive symptoms, overweight or obesity, cigarette smoking) in young adulthood or who develop CHD risk factors over time are more likely to have slow HRR (<22 beats·min−1) by early middle age. Although a notably higher proportion of women and black participants had slow HRR at follow-up, those demographic differences were eliminated once health behaviors, clinical characteristics, and education were taken into account. Findings from our observational longitudinal study indicate that risk factors for slow HRR and clinical CHD are shared.

Several parameters from a standard GXT are used clinically to identify patients with functional cardiovascular abnormalities that place them at high risk for morbidity and mortality. After cessation of the test, the HR falls from its maximal output to pretest resting levels. The rate of HRR to pretest levels is determined in part by cardiorespiratory fitness but primarily by the reactivation of the parasympathetic division of the autonomic nervous system. Contributions of the parasympathetic nervous system to recovery have been determined in pharmacologic blockade studies that block parasympathetic functioning and note that HR does not fall as fast or as far with blockade as compared with the control condition of no blockade (28). Prior research demonstrates that slow HRR, independent of other exercise test parameters, is associated with two to three times higher mortality (depending on the cut points selected) (10,11,25). Most studies have reported mortality estimates related to cut points for HRR (2 min at <22 beats·min−1 and 1 min at <42 beats·min−1); however, Cole et al. (10) reported that each SD lower in 1-min HRR (9 beats·min−1) was associated with a doubling in risk of mortality (HR = 2.1, 95% CI = 1.8–2.5). Although the absolute change in HRR cannot be directly compared with our findings, Cole et al. present a figure demonstrating a graded increase in mortality with slower HRR. It is possible that participants experiencing greater declines during 20 yr, particularly those that exceed the average of 3.6 beats·min−1, may experience higher mortality.

Most mortality studies have been carried out in convenience samples of patients referred for exercise testing on the basis of their symptoms of heart disease or existing high-risk cardiovascular profile. Although cardiorespiratory fitness may be strongly associated with HRR, prior studies that adjust for treadmill duration as an estimate of fitness may be incomplete; statistical adjustment for direct measures of fitness (e.g., V˙O2max determined by gas exchange) may be needed to fully establish the independence of parasympathetic reactivation versus fitness on HRR.

After an intriguing report by Shishehbor et al. (29) that lower socioeconomic status was associated with slower HRR in a patient sample, we investigated whether those findings were also present in a healthy sample or whether the behaviors and clinical profiles of persons with lower socioeconomic status explained such an association. Education is a marker of socioeconomic status that is inversely associated with morbidity and mortality from cardiovascular disease (12). Adverse health risk behaviors cluster among adults with lower education, and in turn, rates of obesity, hypertension, and diabetes are higher (20). In CARDIA, healthy adults were sampled from three US communities and one large health maintenance organization, and the distribution of participants was roughly balanced by educational attainment (17). Our findings concur with prior reports; education was inversely associated with the odds of having slow HRR during 20 yr. A residual association between education and HRR remained even once health behaviors (e.g., physical activity and smoking status), other demographic characteristics (e.g., age, race, sex), and metabolic factors (e.g., BMI, waist circumference, blood pressure, and fasting glucose) were taken into account. Education is a marker for multiple unmeasured parameters that are linked to socioeconomic status such as diet, financial insecurity, discrimination, stress, access to health care, and neighborhood cohesiveness that may not be captured by one simple marker at the individual level (12).

The biological plausibility of our observations arises from the considerable literature demonstrating that metabolic disorders, namely, diabetes and hypertension, are associated with impaired autonomic nervous system function. Autonomic neuropathy is an established microvascular complication of diabetes. Chronic hyperglycemia may degrade the peripheral vasculature leading to neuropathy (4). Alternatively, autonomic nervous system dysfunction may occur earlier in the course of disease than previously thought—possibly secondary to obesity and insulin resistance. Early dysfunction may interfere with hepatic glucose production, circulating glucose uptake, and muscle insulin sensitivity, thus prompting the development of diabetes (5,6,9). Our findings demonstrate that the association of fasting glucose with incident slow HRR is even stronger among participants whose HRR is already low at baseline. Similarly, the association between autonomic dysfunction and hypertension seems bidirectional. Prior reports indicate not only that HR variability, an estimate of autonomic nervous system function, is impaired with increasing blood pressure but also that lower HR variability is positively associated with the development of hypertension (27). By contrast, longitudinal investigations suggest that slower HRR does not precede the development of metabolic syndrome; rather, the presence of one or more components of metabolic syndrome is associated with greater declines in HRR over time (23). Consequently, our findings of a positive association of baseline fasting glucose and the incidence of diabetes, hypertension, and metabolic syndrome with slow HRR during follow-up are consistent with the existing literature.

Health behaviors, namely, smoking and physical inactivity, were each associated with a higher likelihood of having slow HRR at follow-up. Activity is associated with a more favorable autonomic profile in other population studies that used different estimates of autonomic functioning (26). In a prior CARDIA investigation, we demonstrated that participants who were more physically active had faster HRR (7). Most prior studies report less favorable resting autonomic nervous system function and autonomic response to provocative maneuvers in smokers as compared with nonsmokers (1,8). Our longitudinal findings extend beyond prior cross-sectional reports to suggest that these adverse health behaviors precede the onset of autonomic impairment.

Unlike many prior studies of correlates of HRR that were conducted in patient-referred samples, our study was conducted in a sample of young adults who participated in a longitudinal population study on cardiovascular health. Findings from our study are generalizable to the large proportion of the population without symptoms necessitating referral to exercise testing. Using repeated assessment of HRR in our longitudinal cohort study, we could determine factors associated with the development of slow HRR during follow-up. By contrast, most prior studies included a single measure of HRR.

Our findings must be interpreted in light of some limitations. HRR is not a direct measure of autonomic nervous system dysfunction but rather is an estimate of parasympathetic responsiveness to a specific physiologic maneuver (i.e., exercise). However, prior studies that have applied pharmacologic blockade have validated the contributions of the parasympathetic nervous system to HRR (18,31). In addition, HRR is strongly positively correlated with components of HR variability thought to represent parasympathetic modulation of HR (13). Further studies with measures of autonomic nervous system function that represent different components of function (e.g., sympathetic input) are warranted to confirm our observations.

In summary, HRR declines in healthy adults with aging. However, whether that decline reaches levels that have been previously determined to pose a higher risk for mortality depends on the presence and development of CHD risk factors. With the exception of educational attainment, the characteristics most strongly associated with the odds of having slow HRR in middle age are modifiable. Future intervention studies should measure changes in HRR before and after an intervention to determine whether improvements in parasympathetic reactivity are a plausible mediating pathway between the health behavior of physical activity and cardiovascular and metabolic outcomes.

Work on this article was supported by grant 5 R01 HL078972 from the National Heart, Lung, and Blood Institute and was partially supported by contracts N01-HC-48047, N01-HC-48048, N01-HC-48049, N01-HC-48050, and N01-HC-95095 from the National Heart, Lung, and Blood Institute/National Institutes of Health.

The authors thank the CARDIA study participants and staff for their valuable contributions.

This work was previously presented in part at the 54th Annual Scientific Sessions of the American College of Sports Medicine in New Orleans, LA, in 2007.

The results of the present study do not constitute endorsement by the American College of Sports Medicine.


1. Barutcu I, Esen AM, Kaya D, et al.. Cigarette smoking and heart rate variability: dynamic influence of parasympathetic and sympathetic maneuvers. Ann Noninvasive Electrocardiol. 2005; 10 (3): 324–9.
2. Blair SN, Kampert JB, Kohl HW 3rd, et al.. Influences of cardiorespiratory fitness and other precursors on cardiovascular disease and all-cause mortality in men and women. JAMA. 1996; 276 (3): 205–10.
3. Borg GA. Perceived exertion. Exerc Sport Sci Rev. 1974; 2: 131–53.
4. Boyko EJ. The epidemiology of diabetic neuropathy. In: Veves A, editor. Clinical Management of Diabetic Neuropathy. Totowa (NJ): Humana Press, Inc.; 1998. p. 1–12.
5. Carnethon MR, Golden SH, Folsom AR, Haskell W, Liao D. Prospective investigation of autonomic nervous system function and the development of type 2 diabetes: the Atherosclerosis Risk in Communities study, 1987–1998. Circulation. 2003; 107 (17): 2190–5.
6. Carnethon MR, Jacobs DR Jr, Sidney S, Liu K. Influence of autonomic nervous system dysfunction on the development of type 2 diabetes: the CARDIA study. Diabetes Care. 2003; 26 (11): 3035–41.
7. Carnethon MR, Jacobs DR Jr, Sidney S, et al.. A longitudinal study of physical activity and heart rate recovery: CARDIA, 1987–1993. Med Sci Sports Exerc. 2005; 37 (4): 606–12.
8. Carnethon MR, Liao D, Evans GW, Cascio WE, Chambless LE, Heiss G. Correlates of the shift in heart rate variability with an active postural change in a healthy population sample: the Atherosclerosis Risk in Communities study. Am Heart J. 2002; 143 (5): 808–13.
9. Carnethon MR, Prineas RJ, Temprosa M, Zhang ZM, Uwaifo G, Molitch ME. The association among autonomic nervous system function, incident diabetes, and intervention arm in the Diabetes Prevention Program. Diabetes Care. 2006; 29 (4): 914–9.
10. Cole CR, Blackstone EH, Pashkow FJ, Snader CE, Lauer MS. Heart-rate recovery immediately after exercise as a predictor of mortality. N Engl J Med. 1999; 341: 1351–7.
11. Cole CR, Foody JM, Blackstone EH, Lauer MS. Heart rate recovery after submaximal exercise testing as a predictor of mortality in a cardiovascularly healthy cohort. Ann Intern Med. 2000; 132 (7): 552–5.
12. Craft LL, Perna FM. The benefits of exercise for the clinically depressed. Prim Care Companion J Clin Psychiatry. 2004; 6 (3): 104–11.
13. Evrengul H, Tanriverdi H, Kose S, et al.. The relationship between heart rate recovery and heart rate variability in coronary artery disease. Ann Noninvasive Electrocardiol. 2006; 11 (2): 154–62.
14. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001; 285 (19): 2486–97.
15. Folsom AR, Burke GL, Ballew C, et al.. Relation of body fatness and its distribution to cardiovascular risk factors in young blacks and whites. The role of insulin. Am J Epidemiol. 1989; 130 (5): 911–24.
16. Folsom AR, Jacobs DR Jr, Wagenknecht LE, et al.. Increase in fasting insulin and glucose over seven years with increasing weight and inactivity of young adults. The CARDIA Study. Coronary Artery Risk Development in Young Adults. Am J Epidemiol. 1996; 144 (3): 235–46.
17. Friedman GD, Cutter GR, Donahue RP, et al.. CARDIA: study design, recruitment, and some characteristics of the examined subjects. J Clin Epidemiol. 1988; 41 (11): 1105–16.
18. Goldberger JJ, Challapalli S, Tung R, Parker MA, Kadish AH. Relationship of heart rate variability to parasympathetic effect. Circulation. 2001; 103 (15): 1977–83.
19. Grundy SM, Brewer HB Jr, Cleeman JI, Smith SC Jr, Lenfant C, and for the Conference Participants. Definition of metabolic syndrome: report of the National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues related to definition. Circulation. 2004; 109 (3): 433–8.
20. Hsiung HM, Smiley DL, Zhang XY, et al.. Potent peptide agonists for human melanocortin 3 and 4 receptors derived from enzymatic cleavages of human beta-MSH(5–22) by dipeptidyl peptidase I and dipeptidyl peptidase IV. Peptides. 2005; 26 (10): 1988–96.
21. Jacobs DR, Hahn L, Haskell WL, Pirie P, Sidney S. Validity and reliability of short physical activity history: CARDIA Study and the Minnesota Heart Health Program. J Cardiopulm Rehabil. 1989; 9: 448–59.
22. Jacobs DR Jr, Ainsworth BE, Hartman TJ, Leon AS. A simultaneous evaluation of 10 commonly used physical activity questionnaires. Med Sci Sports Exerc. 1993; 25 (1): 81–91.
23. Kizilbash MA, Carnethon MR, Chan C, Jacobs DR, Sidney S, Liu K. The temporal relationship between heart rate recovery immediately after exercise and the metabolic syndrome: the CARDIA study. Eur Heart J. 2006; 27 (13): 1592–6.
24. Morise AP. Heart rate recovery: predictor of risk today and target of therapy tomorrow? Circulation. 2004; 110 (18): 2778–80.
25. Nishime EO, Cole CR, Blackstone EH, Pashkow FJ, Lauer MS. Heart rate recovery and treadmill exercise score as predictors of mortality in patients referred for exercise ECG. JAMA. 2000; 284 (11): 1392–8.
26. Rennie KL, Hemingway H, Kumari M, Brunner E, Malik M, Marmot M. Effects of moderate and vigorous physical activity on heart rate variability in a British study of civil servants. Am J Epidemiol. 2003; 158 (2): 135–43.
27. Schroeder EB, Liao D, Chambless LE, Prineas RJ, Evans GW, Heiss G. Hypertension, blood pressure, and heart rate variability: the Atherosclerosis Risk in Communities (ARIC) study. Hypertension. 2003; 42 (6): 1106–11.
28. Shetler K, Marcus R, Froelicher VF, et al.. Heart rate recovery: validation and methodologic issues. J Am Coll Cardiol. 2001; 38 (7): 1980–7.
29. Shishehbor MH, Litaker D, Pothier CE, Lauer MS. Association of socioeconomic status with functional capacity, heart rate recovery, and all-cause mortality. JAMA. 2006; 295 (7): 784–92.
30. Sidney S, Haskell WL, Crow R, et al.. Symptom-limited graded treadmill exercise testing in young adults in the CARDIA study. Med Sci Sports Exerc. 1992; 24 (2): 177–83.
31. Sundaram S, Carnethon M, Polito K, Kadish AH, Goldberger JJ. Autonomic effects on QT–RR interval dynamics after exercise. Am J Physiol Heart Circ Physiol. 2008; 294 (1): H490–7.
32. Tanaka H, Monahan KD, Seals DR. Age-predicted maximal heart rate revisited. J Am Coll Cardiol. 2001; 37 (1): 153–6.
33. Warnick GR, Benderson J, Albers JJ. Dextran sulfate–Mg2+ precipitation procedure for quantitation of high-density–lipoprotein cholesterol. Clin Chem. 1982; 28 (6): 1379–88.


©2012The American College of Sports Medicine