Enhanced Risk Assessment in Asymptomatic Individuals With Exercise Testing and Framingham Risk Scores
http://www.100md.com
《循环学杂志》
the Center for Cardiovascular Disease Prevention, Divisions of Preventive Medicine and Cardiovascular Medicine (S.M.), Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Mass
Division of Cardiology (R.F.R.), Department of Medicine, University of California, San Francisco, San Francisco, Calif
Department of Epidemiology (A.R.S.), Johns Hopkins Bloomberg School of Public Health, Baltimore, Md
Ciccarone Preventive Cardiology Center (R.S.B.), Division of Cardiology, Department of Medicine, Johns Hopkins Medical Institutions, Baltimore, Md.
Abstract
Background— National Cholesterol Education Program Adult Treatment Panel III (ATP III) guidelines recommend the use of Framingham risk scores (FRS) for cardiovascular assessment of asymptomatic individuals. We hypothesized that risk prediction could be improved with 2 non-ECG exercise test measures, exercise capacity (metabolic equivalents, or METs) and heart rate recovery (HRR).
Methods and Results— An asymptomatic cohort with baseline treadmill tests (n=6126; 46% women, FRS <20%) was followed up prospectively for 20 years. Individuals with low (median or less) HRR or METs experienced 91% of all cardiovascular disease (CVD) deaths (225/246). After FRS adjustment, low HRR and METs individually were highly significant predictors of CVD death, but low HRR and METs together were associated with substantially higher risk (adjusted hazard ratio compared with high HRR/high METs for women 8.51, 95% CI 3.65 to 19.84; for men, 3.53, 95% CI 2.03 to 6.15; P<0.001 for both). At 10-year follow-up, FRS-adjusted CVD death risk associated with low HRR/low METs was less than at 20 years but remained significant (women 3.83, 95% CI 1.09 to 13.47, and men 2.70, 95% CI 1.11 to 6.55). The application of HRR/METs information to FRS assessment identified those at high risk (>0.5% annual CVD mortality) in half of women with FRS 6% to 9% and 10% to 19% and just under half of men with FRS 10% to 19%. Low HRR/low METs was also associated with an increased relative risk of CVD death in individuals with low-risk FRS (FRS <6% in women and <10% in men), but absolute CVD mortality rates were low in this subgroup.
Conclusions— Exercise testing may be a useful adjunct for clinical risk assessment in asymptomatic women with FRS 6% to 19% and men with FRS 10% to 19%.
Key Words: exercise prevention prognosis risk factors
Introduction
Cardiovascular disease (CVD) risk assessment is the cornerstone in tailoring the intensity of risk factor reduction to overall risk.1 Traditional risk factors that comprise the Framingham Risk Score (FRS) are the foundation for estimating risk in asymptomatic individuals.2 Although coronary heart disease (CHD) accounts for more than half a million deaths annually in the United States,3 most of the American adult population is at low or intermediate risk on the basis of the FRS.4 Recent National Health and Nutrition Examination Survey (NHANES) data found that 5% of men and <1% of women have 10-year FRS scores for hard CHD events >20%.5 Although many individuals, especially women and younger individuals, may have low or intermediate risk over a 10-year horizon, a substantial proportion of them are at significantly increased risk in the long term.4,6,7
Additional risk assessment with exercise testing has been proposed to identify individuals at intermediate FRS risk who may benefit from risk factor reduction, yet the high rate of false-positive ischemic ECG responses in asymptomatic populations has prevented its widespread recommendation.4,8–11 Recent studies suggest that risk prediction of exercise testing is improved when variables other than exercise-induced ischemia are used,12–19 yet no study has examined the predictive value of non-ECG exercise test parameters beyond FRS scoring for prediction of fatal CVD in low- or intermediate-risk asymptomatic populations. The purpose of the present study was to evaluate whether 2 measures obtained from exercise treadmill testing, exercise capacity (metabolic equivalents [METs]) and heart rate recovery (HRR), could provide incremental prognostic value for cardiovascular mortality in asymptomatic individuals at low or intermediate FRS risk.
Methods
Study Participants
Participants were derived from the Lipid Research Clinics Prevalence Study, a prospective cohort of North American men and women across 10 different geographic locations, from diverse socioeconomic and occupational groups.16,20,21 Individuals were screened at visit 1 for triglyceride and cholesterol levels. All those screened at visit 1 who had elevated lipids, plus a 15% random sample of visit 1 participants, were invited back for visit 2 for a baseline history, physical examination, fasting blood samples, and an exercise treadmill test. Most participants in this analysis (60%) came from the nonhyperlipidemic sample.
Altogether, 8652 of 13 852 individuals who were screened at visit 2 underwent exercise tests (1972–1976). Of these, 2526 individuals were excluded from the present analysis. Excluded individuals included those with exercise duration <1 minute (to reduce potential confounding by preexisting disease or severe comorbidities); those with known or suspected CVD (myocardial infarction, angina, stroke, left ventricular hypertrophy, or digitalis use); those with FRS 20%; those with known diabetes mellitus or fasting glucose 7.0 mmol/L (126 mg/dL), because such individuals are considered CHD equivalents1; and those with missing FRS or exercise test data. We also excluded baseline age <30 or >70 years, treadmill test other than Bruce protocol, and those lost to follow-up (n=6). Of the 6126 participants in the present study, 54% were not included in a previous report from the Lipid Research Clinics on exercise testing.12
Follow-up was until death or December 31, 1995. The primary outcome was CVD death, ascertained with death certificates, interviews with next of kin, medical records, the National Death Index, and the Epidemiology Research Index.16 Secondary outcomes were CHD death and non-CVD death, ascertained in a similar manner to CVD death. All participants gave written informed consent at the time of enrollment. The study was approved by the Johns Hopkins Institutional Review Board.
Exercise Test Protocol
Participants underwent standard Bruce treadmill protocol exercise tests.22 The test was terminated when a predetermined target heart rate of 90% of maximal predicted heart rate for age and physical activity level was attained.23 In addition, the test could be terminated early at the discretion of the supervising physician for significant symptoms, arrhythmias, or hemodynamic or ST-segment changes, or if the participant was unwilling or unable to continue.
Exercise Test Variables
Peak exercise capacity, expressed in METs, was estimated from treadmill time.22 HRR was defined as the heart rate at peak exercise minus the heart rate at 2 minutes after exercise, in beats per minute (bpm).24 Ischemic ECG changes were defined as exercise-induced horizontal or downsloping ST-segment depression of 1 mm from baseline that occurred in the last stage of exercise or recovery.
Framingham Risk Scores
Sex-specific Framingham equations for 10-year absolute risk of hard CHD events (myocardial infarction or CHD death) based on the third report of the National Cholesterol Education Program’s Adult Treatment Panel III (ATP III) were used to calculate FRS values at baseline.2 Risk categories were defined according to FRS values of <6%, 6% to 9%, and 10% to 19%.2,9 Serum lipid levels were determined from fasting samples from visit 2. Smoking was defined as cigarette smoking at visit 2. Hypertension was defined as baseline systolic blood pressure 140 mm Hg, diastolic pressure 90 mm Hg, or treatment for hypertension. Medication use was assessed from questionnaires or examination of medications by trained interviewers.
Statistical Analyses
STATA software (version 8.2) was used for statistical analyses.25 All analyses were stratified by sex. Mean±SD values were calculated for continuous variables, and statistical comparisons were made with Student t tests or ANOVA. Categorical variables were compared by contingency tables and 2 statistic. All significance tests were 2-tailed, with probability values <0.05 considered significant. FRS was analyzed as a continuous and categorical variable. Sex-specific medians for peak METs and HRR were used to define 3 HRR/METs groups: (1) high HRR/high METs: both METs and HRR above the median; (2) low HRR or METs: either METs or HRR at or below the median; and (3) low HRR/low METs: both METs and HRR at or below the median.
FRS-adjusted Cox proportional hazards models were used to estimate the independent association of the 3 HRR/METs groups with time to death. The proportional hazard assumption was tested and satisfied with Schoenfeld residuals. To test the hypothesis of an incremental prognostic value for the addition of exercise test variables to FRS, we used likelihood ratio tests to determine whether models that included FRS and HRR/METs provided a significantly better fit than models limited to FRS alone. Interaction terms were used in regression models to test for interaction, and 2 statistical tests for interaction were obtained from likelihood ratio testing of models with and without interaction terms. We further explored the extent to which HRR/METs information may be used to move subjects from a risk of CVD death of <5/1000 (0.5%) per year to a risk of >5/1000, a rate used in recent European Task Force guidelines.26 An annual CVD death rate of 0.5% corresponds approximately to an annual hard CHD rate of 1.0%, because incident hard CHD events are twice as frequent as CVD deaths.6
Results
Baseline Characteristics
The cohort was middle-aged and mostly white (Table 1). Mean LDL cholesterol and triglyceride values were higher than current national averages, although the prevalence of hypertension and obesity was lower. Mean±SD FRS values for men and women were 6.4±5.4% and 1.9±2.6%, respectively (P<0.0001), both of which are considered low risk by ATP III. Sex-specific medians for HRR were 56 and 55 bpm for women and men, respectively, and for METs, they were 7.5 and 10.7, respectively. Individuals with low HRR/low METs were substantially older, had higher FRS, and were less likely to participate in regular exercise (Table 2) than those with high HRR or high METs. They also had more exercise test abnormalities, such as exercise-induced ischemia and inability to reach 90% or more of target heart rate. Only 19 individuals (0.3%) were taking -blockers or medication for irregular heart rate.
Kaplan-Meier Curves by HRR/METs
During a mean follow-up period of 20.5±3.6 years, there were 749 all-cause deaths. Fatal cardiovascular events occurred in 145 men (4.4%) and 101 women (3.6%), with 91% of these deaths (225/246) occurring in individuals with low HRR or low METs. Over the first 10 years of follow-up, there were 50 CVD deaths in men and 30 in women. Kaplan-Meier curves for survival free of CVD death diverged early in follow-up and continued to diverge for the group with both low HRR and low METs compared with the other 2 groups (Figure 1). Survival curves for those with either low HRR or low METs separated late in follow-up in both sexes.
Relative Risk of CVD Death by HRR/METs and FRS
HRR and METs separately were highly significant predictors of fatal CVD. After adjustment for FRS categories (FRS <6%, 6% to 9%, and 10% to 19%), hazard ratios for low compared with high METs for women and men, respectively, were 3.09 (95% CI 1.90 to 5.05) and 2.59 (95% CI 1.74 to 3.87; P<0.001 for both sexes). Similarly, the FRS-adjusted hazard ratios for low HRR for women and men, respectively, were 3.25 (95% CI 2.02 to 5.24) and 1.99 (95% CI 1.37 to 2.89; P<0.001 for both sexes).
When both HRR and METs were low (Table 3), the FRS-adjusted hazard ratios for women and men were higher, especially at 20-year follow-up (adjusted hazard ratios, respectively, were 8.51, 95% CI 3.65 to 19.84, and 3.53, 95% CI 2.03 to 6.15, P<0.001 for both sexes). There was no statistically significant interaction between HRR and METs categories for CVD death. When the 10- and 20-year follow-up hazard ratios were compared, the relative effect of FRS appeared to lessen with time, whereas that of HRR/METs appeared to increase. However, there was no statistically significant interaction between HRR/METs and follow-up period for the outcome of CVD death (Pinteraction=0.83 and 0.47 for men and women, respectively). There was a borderline statistically significant interaction of sex and HRR/METs (Pinteraction=0.08). Likelihood ratio tests comparing Cox regression models that included HRR/METs and FRS to models limited to FRS alone showed a highly statistically significant improvement to risk prediction for both sexes with the addition of HRR/METs, although the magnitude of improvement was stronger with longer duration of follow-up (P<0.0001 for both sexes).
At both 10- and 20-year follow-up, HRR/METs additionally stratified the relative risk for CVD death across all FRS categories, even after further adjustment for the range of FRS values within each FRS category (Figure 2). For individuals with FRS <6%, the impact of having low HRR/low METs on the relative risk of CVD death was somewhat higher after 20 years than 10 years (adjusted hazard ratio compared with individuals with high HRR/high METS, at 20 years: 7.86, 95% CI 3.73 to 16.56, P<0.001; at 10 years: 3.29, 95% CI 1.05 to 19.36, P=0.04). In the group with FRS 6% to 19%, the relative risk of low HRR/low METs was similar and highly statistically significant (P<0.001) at both 10- and 20-year follow-up.
Absolute Risk of CVD Death by HRR/METs and FRS
Average annual CVD mortality rates (per 1000 person-years) stratified by HRR/METs and FRS categories are shown in Table 4. Men with FRS <10% and women with FRS <6% had low absolute CVD mortality rates (<5/1000 or 0.5% per year), regardless of their HRR/METs. For individuals with FRS 6% to 19%, exercise testing clearly identified individuals with high absolute CVD mortality rates (>5/1000 per year, rates that are considered high risk by European Task Force guidelines and that correspond to rates of hard CHD events >1% per year6,26).
In men with FRS 10% to 19%, those with low HRR/low METs (399/847 or 47%) were identified as high risk on the basis of annual CVD mortality rates >5/1000. Similarly, for women with FRS 6% to 9% and 10% to 19%, those with low HRR/low METs (70/134 or 52% of women with FRS 6% to 9% and 34/62 or 55% of women with FRS 10% to 19%) were identified as high risk on the basis of annual rates >5/1000. Women with FRS 10% to 19% and either low HRR or low METs had a trend toward higher annual CVD mortality rates, but the confidence interval included values <5/1000.
ST-Segment Depression
When exercise-induced ST-segment depression 1 mm was added to the models in Table 2 that included FRS and HRR/METs, there was increased risk associated with ST-segment depression in men but not in women (adjusted hazard ratio for ST depression at 20 years for men 2.26, 95% CI 1.28 to 4.01, P=0.005, and 1.47, 95% CI 0.73 to 2.93, P=0.29 for women). At 10-year follow-up, there was a trend for increased risk in men (P=0.06), and ST-segment depression remained nonsignificant for women (P=0.99).
CHD Death, Non-CVD Death, and All-Cause Death
In secondary analyses for CHD death with models similar to those in Table 2 for 20-year follow-up, the adjusted hazard ratios for HRR/METs categories for CHD death in men were similar to those for CVD death. The hazard ratios for HRR/METs categories and CHD death in women were greater than those for CVD death, with a statistically significant interaction between sex and HRR/METs (2=5.99, Pinteraction=0.01). Women with low HRR/low METs had nearly 30-fold increased risk of CHD death compared with women with high HRR/high METs, after adjustment for FRS categories (P=0.001).
When similar analyses were performed for non-CVD death and for all-cause death at 20 years, the adjusted hazard ratios for HRR/METs for men and women were highly significant, and both were similar in magnitude to results for CVD death in men, with no significant interactions with sex. There were no significant interactions between high LDL cholesterol and HRR/METs with respect to any of the mortality outcomes.
Discussion
In this large, prospective North American cohort of asymptomatic individuals followed up for more than 2 decades, we found that 2 easily obtained non-ECG measures from exercise testing (HRR and METs) were strongly predictive of CVD death, incremental to FRS, and valid across FRS risk categories <20%. Using annual CVD mortality rates, we estimated that half of women and just under half of men with FRS 10% to 19%, plus half of women with FRS 6% to 9%, would be reclassified as high risk on the basis of having low HRR/low METs, thus providing clinically important risk information.
The US Preventive Services Task Force recently concluded that exercise testing had poor to fair accuracy in predicting events in asymptomatic individuals based on ECGs.10 Although several studies have previously examined the prognostic relevance of exercise tests,12–18,20,27–31 only a few studies analyzed how non-ECG exercise variables add to FRS estimates.19, 32–33
The present study differs from previous studies by focusing on low-intermediate risk individuals and by examining fatal cardiovascular outcomes in a very large number of men and women with lengthy follow-up. In one previous study, exercise capacity in women predicted all-cause mortality after adjustment for FRS, but CVD data were unavailable.32 In a recent Cleveland Clinic study of asymptomatic individuals, low HRR and low METs were significant predictors of all-cause mortality after global risk adjustment, but they were clinically useful only in those with highest global risk.19 However, the study may have been underpowered to detect clinically meaningful differences in lower-risk individuals compared with the present study. In addition, their sample was self-referred, and there were no cardiovascular outcome data.19 The Framingham Study recently reported that among asymptomatic men, additional prognostic information in FRS-adjusted models was obtained by ST-segment depression, failure to reach target heart rate, and decreased exercise capacity.33 These exercise test variables did not yield incremental prognostic information in women or low-risk men, but there were few CHD events in women.33
One potential strategy to apply exercise testing in the general population would be to screen individuals with intermediate risk (FRS 10% to 19% or 6% to 19%).9 The present data suggest that the impact of adding an exercise test to risk prediction would be to additionally identify as "high risk" (based on an annual CVD mortality rate >0.5%) a substantial proportion of individuals with FRS scores between 6% and 19%. Even though there were only 74 women in the FRS category of 10% to 19%, having a low HRR and low METs was a highly significant predictor in both men and women across all FRS categories. HRR/METs data may be used to inform clinicians’ decisions about the optimal degree of primary prevention. Although exercise testing also risk-stratified those with FRS <6%, their absolute CVD mortality rates were low, and thus exercise testing in this low-risk group may not provide important clinical information. The present study offers further data in women and men to support the call for a clinical outcomes trial that addresses survival benefit using targeted risk factor reduction and improved fitness strategies based on exercise testing.34
There are several limitations to the present study. One limitation is the generalizability of our risk estimates to nonwhites, even though the cohort was recruited from different geographic sites with a range of socioeconomic and occupational status.20,35 Another limitation is that 40% of the participants were enrolled on the basis of lipid abnormalities. Despite the higher prevalence of dyslipidemia in this cohort, their mean FRS values and mortality rates were similar to those found in populations not enriched with lipid abnormalities,6 with no significant interaction between high LDL cholesterol and mortality. The exercise test protocol used in the present study was a submaximal Bruce protocol, with termination at 90% of maximal predicted heart rate, and it is unclear whether our results would apply to other types of stress protocols. Our findings potentially may not be as applicable to soft CHD end points, such as angina or nonfatal myocardial infarction. A recent study suggested that exercise capacity was more predictive of fatal than nonfatal CVD events, although the potential mechanisms are unclear.36 Even in this large cohort, there were very few deaths in the high HRR/high METs group (the reference group), which may have resulted in large confidence intervals for some hazard ratios, such as those for CHD death in women. Finally, there are currently no large, randomized trials that show the clinical utility of screening exercise tests.
In conclusion, non-ECG exercise test results (HRR and METs) added important and incremental prognostic information to FRS in predicting fatal CVD in low- and intermediate-risk individuals. The application of these 2 simple, noninvasive exercise measures to individuals with FRS 6% to 19% may identify a significant proportion of individuals who are at high risk but who would have been classified as at low or intermediate risk by FRS alone. Selective use of exercise testing may help tailor primary prevention strategies that would match the level of risk to the intensity of therapy, with the potential to help close the sizable gap between detection of and death due to CVD.
Acknowledgments
The authors thank the LRC investigators, staff, and participants for their valuable contributions.
Disclosure
Dr Blumenthal has received research funding and served on the speakers’ bureaus of or received honoraria from Merck, Pfizer, Wyeth, Novartis, Kos, and the Maryland Atheletic Club Charitable Foundation, Lutherville, Md.
Footnotes
Presented in part at the 75th Annual Scientific Sessions of the American Heart Association, Chicago, Ill, November 17–20, 2002, and published in abstract form [Circulation. 2002;106(suppl II):II-601].
References
Goldman L, Garber AM, Grover SA, Hlatky MA. 27th Bethesda Conference: matching the intensity of risk factor management with the hazard for coronary disease events: Task Force 6: cost effectiveness of assessment and management of risk factors. J Am Coll Cardiol. 1996; 27: 1020–1030.
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) final report. Circulation. 2002; 106: 3143–3421.
American Heart Association. Heart Disease and Stroke Statistics—2004 Update. Dallas, Tex: American Heart Association; 2003.
Pasternak RC, Abrams J, Greenland P, Smaha LA, Wilson PW, Houston-Miller N. 34th Bethesda Conference: Task force #1: identification of coronary heart disease risk: is there a detection gap J Am Coll Cardiol. 2003; 41: 1863–1874.
Ford ES, Giles WH, Mokdad AH. The distribution of 10-year risk for coronary heart disease among US adults: findings from the National Health and Nutrition Examination Survey III. J Am Coll Cardiol. 2004; 43: 1791–1796.
Lloyd-Jones DM, Larson MG, Beiser A, Levy D. Lifetime risk of developing coronary heart disease. Lancet. 1999; 353: 89–92.
Lloyd-Jones DM, Wilson PW, Larson MG, Beiser A, Leip EP, D’Agostino RB, Levy D. Framingham risk score and prediction of lifetime risk for coronary heart disease. Am J Cardiol. 2004; 94: 20–24.
Pearson TA. New tools for coronary risk assessment: what are their advantages and limitations Circulation. 2002; 105: 886–892.
Greenland P, Smith SC Jr, Grundy SM. Improving coronary heart disease risk assessment in asymptomatic people: role of traditional risk factors and noninvasive cardiovascular tests. Circulation. 2001; 104: 1863–1867.
Fowler-Brown A, Pignone M, Pletcher M, Tice JA, Sutton SF, Lohr KN. Exercise tolerance testing to screen for coronary heart disease: a systematic review for the technical support for the U.S. Preventive Services Task Force. Ann Intern Med. 2004; 140: W9–W24.
Gibbons RJ, Balady GJ, Bricker JT, Chaitman BR, Fletcher GF, Froelicher VF, Mark DB, McCallister BD, Mooss AN, O’Reilly MG, Winters WL Jr, Antman EM, Alpert JS, Faxon DP, Fuster V, Gregoratos G, Hiratzka LF, Jacobs AK, Russell RO, Smith SC Jr. ACC/AHA 2002 guideline update for exercise testing: summary article: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1997 Exercise Testing Guidelines). Circulation. 2002; 106: 1883–1892.
Mora S, Redberg RF, Cui Y, Whiteman MK, Flaws JA, Sharrett AR, Blumenthal RS. Ability of exercise testing to predict cardiovascular and all-cause death in asymptomatic women: a 20-year follow-up of the Lipid Research Clinics Prevalence Study. JAMA. 2003; 290: 1600–1607.
Blair SN, Kampert JB, Kohl HW III, Barlow CE, Macera CA, Paffenbarger RS Jr, Gibbons LW. Influences of cardiorespiratory fitness and other precursors on cardiovascular disease and all-cause mortality in men and women. JAMA. 1996; 276: 205–210.
Blair SN, Kohl HW III, Barlow CE, Paffenbarger RS Jr, Gibbons LW, Macera CA. Changes in physical fitness and all-cause mortality: a prospective study of healthy and unhealthy men. JAMA. 1995; 273: 1093–1098.
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: 1351–1357.
Ekelund LG, Haskell WL, Johnson JL, Whaley FS, Criqui MH, Sheps DS. Physical fitness as a predictor of cardiovascular mortality in asymptomatic North American men: the Lipid Research Clinics Mortality Follow-up Study. N Engl J Med. 1988; 319: 1379–1384.
Kwok JM, Miller TD, Hodge DO, Gibbons RJ. Prognostic value of the Duke treadmill score in the elderly. J Am Coll Cardiol. 2002; 39: 1475–1481.
Myers J, Prakash M, Froelicher V, Do D, Partington S, Atwood JE. Exercise capacity and mortality among men referred for exercise testing. N Engl J Med. 2002; 346: 793–801.
Aktas MK, Ozduran V, Pothier CE, Lang R, Lauer MS. Global risk scores and exercise testing for predicting all-cause mortality in a preventive medicine program. JAMA. 2004; 292: 1462–1468.
Gordon DJ, Ekelund LG, Karon JM, Probstfield JL, Rubenstein C, Sheffield LT, Weissfeld L. Predictive value of the exercise tolerance test for mortality in North American men: the Lipid Research Clinics Mortality Follow-up Study. Circulation. 1986; 74: 252–261.
Cui Y, Blumenthal RS, Flaws JA, Whiteman MK, Langenberg P, Bachorik PS, Bush TL. Non-high-density lipoprotein cholesterol level as a predictor of cardiovascular disease mortality. Arch Intern Med. 2001; 161: 1413–1419.
Bruce RA, Kusumi F, Hosmer D. Maximal oxygen intake and nomographic assessment of functional aerobic impairment in cardiovascular disease. Am Heart J. 1973; 85: 546–562.
Sheffield LT, Haskell W, Heiss G, Kioschos M, Leon A, Roitman D, Schrott H. Safety of exercise testing volunteer subjects: the Lipid Research Clinics’ Prevalence Study experience. J Cardiac Rehab. 1982; 2: 395–400.
Shetler K, Marcus R, Froelicher VF, Vora S, Kalisetti D, Prakash M, Do D, Myers J. Heart rate recovery: validation and methodologic issues. J Am Coll Cardiol. 2001; 38: 1980–1987.
STATA Reference Manual, Release 8.1. College Station, Tex: Stata Corporation; 2003.
De Backer G, Ambrosioni E, Borch-Johnsen K, Brotons C, Cifkova R, Dallongeville J, Ebrahim S, Faergeman O, Graham I, Mancia G, Manger Cats V, Orth-Gomer K, Perk J, Pyorala K, Rodicio JL, Sans S, Sansoy V, Sechtem U, Silber S, Thomsen T, Wood D. European guidelines on cardiovascular disease prevention in clinical practice: Third Joint Task Force of European and Other Societies on Cardiovascular Disease Prevention in Clinical Practice. Eur Heart J. 2003; 24: 1601–1610.
Allen WH, Aronow WS, Goodman P, Stinson P. Five-year follow-up of maximal treadmill stress test in asymptomatic men and women. Circulation. 1980; 62: 522–527.
Lauer MS, Francis GS, Okin PM, Pashkow FJ, Snader CE, Marwick TH. Impaired chronotropic response to exercise as a predictor of mortality. JAMA. 1999; 281: 524–529.
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: 552–555.
Jouven X, Zureik M, Desnos M, Courbon D, Ducimetiere P. Long-term outcome in asymptomatic men with exercise-induced premature ventricular depolarizations. N Engl J Med. 2000; 343: 826–833.
Cheng YJ, Lauer MS, Earnest CP, Church TS, Kampert JB, Gibbons LW, Blair SN. Heart rate recovery following maximal exercise testing as a predictor of cardiovascular disease and all-cause mortality in men with diabetes. Diabetes Care. 2003; 26: 2052–2057.
Gulati M, Pandey DK, Arnsdorf MF, Lauderdale DS, Thisted RA, Wicklund RH, Al-Hani AJ, Black HR. Exercise capacity and the risk of death in women: the St James Women Take Heart Project. Circulation. 2003; 108: 1554–1559.
Balady GJ, Larson MG, Vasan RS, Leip EP, O’Donnell CJ, Levy D. Usefulness of exercise testing in the prediction of coronary disease risk among asymptomatic persons as a function of the Framingham risk score. Circulation. 2004; 110: 1920–1925.
Mark DB, Lauer MS. Exercise capacity: the prognostic variable that doesn’t get enough respect. Circulation. 2003; 108: 1534–1536.
Bush TL, Cowan LD, Barrett-Connor E, Criqui MH, Karon JM, Wallace RB, Tyroler HA, Rifkind BM. Estrogen use and all-cause mortality: preliminary results from the Lipid Research Clinics Program Follow-Up Study. JAMA. 1983; 249: 903–906.
Laukkanen JA, Kurl S, Salonen R, Rauramaa R, Salonen JT. The predictive value of cardiorespiratory fitness for cardiovascular events in men with various risk profiles: a prospective population-based cohort study. Eur Heart J. 2004; 25: 1428–1437(Samia Mora, MD, MHS; Rita)
Division of Cardiology (R.F.R.), Department of Medicine, University of California, San Francisco, San Francisco, Calif
Department of Epidemiology (A.R.S.), Johns Hopkins Bloomberg School of Public Health, Baltimore, Md
Ciccarone Preventive Cardiology Center (R.S.B.), Division of Cardiology, Department of Medicine, Johns Hopkins Medical Institutions, Baltimore, Md.
Abstract
Background— National Cholesterol Education Program Adult Treatment Panel III (ATP III) guidelines recommend the use of Framingham risk scores (FRS) for cardiovascular assessment of asymptomatic individuals. We hypothesized that risk prediction could be improved with 2 non-ECG exercise test measures, exercise capacity (metabolic equivalents, or METs) and heart rate recovery (HRR).
Methods and Results— An asymptomatic cohort with baseline treadmill tests (n=6126; 46% women, FRS <20%) was followed up prospectively for 20 years. Individuals with low (median or less) HRR or METs experienced 91% of all cardiovascular disease (CVD) deaths (225/246). After FRS adjustment, low HRR and METs individually were highly significant predictors of CVD death, but low HRR and METs together were associated with substantially higher risk (adjusted hazard ratio compared with high HRR/high METs for women 8.51, 95% CI 3.65 to 19.84; for men, 3.53, 95% CI 2.03 to 6.15; P<0.001 for both). At 10-year follow-up, FRS-adjusted CVD death risk associated with low HRR/low METs was less than at 20 years but remained significant (women 3.83, 95% CI 1.09 to 13.47, and men 2.70, 95% CI 1.11 to 6.55). The application of HRR/METs information to FRS assessment identified those at high risk (>0.5% annual CVD mortality) in half of women with FRS 6% to 9% and 10% to 19% and just under half of men with FRS 10% to 19%. Low HRR/low METs was also associated with an increased relative risk of CVD death in individuals with low-risk FRS (FRS <6% in women and <10% in men), but absolute CVD mortality rates were low in this subgroup.
Conclusions— Exercise testing may be a useful adjunct for clinical risk assessment in asymptomatic women with FRS 6% to 19% and men with FRS 10% to 19%.
Key Words: exercise prevention prognosis risk factors
Introduction
Cardiovascular disease (CVD) risk assessment is the cornerstone in tailoring the intensity of risk factor reduction to overall risk.1 Traditional risk factors that comprise the Framingham Risk Score (FRS) are the foundation for estimating risk in asymptomatic individuals.2 Although coronary heart disease (CHD) accounts for more than half a million deaths annually in the United States,3 most of the American adult population is at low or intermediate risk on the basis of the FRS.4 Recent National Health and Nutrition Examination Survey (NHANES) data found that 5% of men and <1% of women have 10-year FRS scores for hard CHD events >20%.5 Although many individuals, especially women and younger individuals, may have low or intermediate risk over a 10-year horizon, a substantial proportion of them are at significantly increased risk in the long term.4,6,7
Additional risk assessment with exercise testing has been proposed to identify individuals at intermediate FRS risk who may benefit from risk factor reduction, yet the high rate of false-positive ischemic ECG responses in asymptomatic populations has prevented its widespread recommendation.4,8–11 Recent studies suggest that risk prediction of exercise testing is improved when variables other than exercise-induced ischemia are used,12–19 yet no study has examined the predictive value of non-ECG exercise test parameters beyond FRS scoring for prediction of fatal CVD in low- or intermediate-risk asymptomatic populations. The purpose of the present study was to evaluate whether 2 measures obtained from exercise treadmill testing, exercise capacity (metabolic equivalents [METs]) and heart rate recovery (HRR), could provide incremental prognostic value for cardiovascular mortality in asymptomatic individuals at low or intermediate FRS risk.
Methods
Study Participants
Participants were derived from the Lipid Research Clinics Prevalence Study, a prospective cohort of North American men and women across 10 different geographic locations, from diverse socioeconomic and occupational groups.16,20,21 Individuals were screened at visit 1 for triglyceride and cholesterol levels. All those screened at visit 1 who had elevated lipids, plus a 15% random sample of visit 1 participants, were invited back for visit 2 for a baseline history, physical examination, fasting blood samples, and an exercise treadmill test. Most participants in this analysis (60%) came from the nonhyperlipidemic sample.
Altogether, 8652 of 13 852 individuals who were screened at visit 2 underwent exercise tests (1972–1976). Of these, 2526 individuals were excluded from the present analysis. Excluded individuals included those with exercise duration <1 minute (to reduce potential confounding by preexisting disease or severe comorbidities); those with known or suspected CVD (myocardial infarction, angina, stroke, left ventricular hypertrophy, or digitalis use); those with FRS 20%; those with known diabetes mellitus or fasting glucose 7.0 mmol/L (126 mg/dL), because such individuals are considered CHD equivalents1; and those with missing FRS or exercise test data. We also excluded baseline age <30 or >70 years, treadmill test other than Bruce protocol, and those lost to follow-up (n=6). Of the 6126 participants in the present study, 54% were not included in a previous report from the Lipid Research Clinics on exercise testing.12
Follow-up was until death or December 31, 1995. The primary outcome was CVD death, ascertained with death certificates, interviews with next of kin, medical records, the National Death Index, and the Epidemiology Research Index.16 Secondary outcomes were CHD death and non-CVD death, ascertained in a similar manner to CVD death. All participants gave written informed consent at the time of enrollment. The study was approved by the Johns Hopkins Institutional Review Board.
Exercise Test Protocol
Participants underwent standard Bruce treadmill protocol exercise tests.22 The test was terminated when a predetermined target heart rate of 90% of maximal predicted heart rate for age and physical activity level was attained.23 In addition, the test could be terminated early at the discretion of the supervising physician for significant symptoms, arrhythmias, or hemodynamic or ST-segment changes, or if the participant was unwilling or unable to continue.
Exercise Test Variables
Peak exercise capacity, expressed in METs, was estimated from treadmill time.22 HRR was defined as the heart rate at peak exercise minus the heart rate at 2 minutes after exercise, in beats per minute (bpm).24 Ischemic ECG changes were defined as exercise-induced horizontal or downsloping ST-segment depression of 1 mm from baseline that occurred in the last stage of exercise or recovery.
Framingham Risk Scores
Sex-specific Framingham equations for 10-year absolute risk of hard CHD events (myocardial infarction or CHD death) based on the third report of the National Cholesterol Education Program’s Adult Treatment Panel III (ATP III) were used to calculate FRS values at baseline.2 Risk categories were defined according to FRS values of <6%, 6% to 9%, and 10% to 19%.2,9 Serum lipid levels were determined from fasting samples from visit 2. Smoking was defined as cigarette smoking at visit 2. Hypertension was defined as baseline systolic blood pressure 140 mm Hg, diastolic pressure 90 mm Hg, or treatment for hypertension. Medication use was assessed from questionnaires or examination of medications by trained interviewers.
Statistical Analyses
STATA software (version 8.2) was used for statistical analyses.25 All analyses were stratified by sex. Mean±SD values were calculated for continuous variables, and statistical comparisons were made with Student t tests or ANOVA. Categorical variables were compared by contingency tables and 2 statistic. All significance tests were 2-tailed, with probability values <0.05 considered significant. FRS was analyzed as a continuous and categorical variable. Sex-specific medians for peak METs and HRR were used to define 3 HRR/METs groups: (1) high HRR/high METs: both METs and HRR above the median; (2) low HRR or METs: either METs or HRR at or below the median; and (3) low HRR/low METs: both METs and HRR at or below the median.
FRS-adjusted Cox proportional hazards models were used to estimate the independent association of the 3 HRR/METs groups with time to death. The proportional hazard assumption was tested and satisfied with Schoenfeld residuals. To test the hypothesis of an incremental prognostic value for the addition of exercise test variables to FRS, we used likelihood ratio tests to determine whether models that included FRS and HRR/METs provided a significantly better fit than models limited to FRS alone. Interaction terms were used in regression models to test for interaction, and 2 statistical tests for interaction were obtained from likelihood ratio testing of models with and without interaction terms. We further explored the extent to which HRR/METs information may be used to move subjects from a risk of CVD death of <5/1000 (0.5%) per year to a risk of >5/1000, a rate used in recent European Task Force guidelines.26 An annual CVD death rate of 0.5% corresponds approximately to an annual hard CHD rate of 1.0%, because incident hard CHD events are twice as frequent as CVD deaths.6
Results
Baseline Characteristics
The cohort was middle-aged and mostly white (Table 1). Mean LDL cholesterol and triglyceride values were higher than current national averages, although the prevalence of hypertension and obesity was lower. Mean±SD FRS values for men and women were 6.4±5.4% and 1.9±2.6%, respectively (P<0.0001), both of which are considered low risk by ATP III. Sex-specific medians for HRR were 56 and 55 bpm for women and men, respectively, and for METs, they were 7.5 and 10.7, respectively. Individuals with low HRR/low METs were substantially older, had higher FRS, and were less likely to participate in regular exercise (Table 2) than those with high HRR or high METs. They also had more exercise test abnormalities, such as exercise-induced ischemia and inability to reach 90% or more of target heart rate. Only 19 individuals (0.3%) were taking -blockers or medication for irregular heart rate.
Kaplan-Meier Curves by HRR/METs
During a mean follow-up period of 20.5±3.6 years, there were 749 all-cause deaths. Fatal cardiovascular events occurred in 145 men (4.4%) and 101 women (3.6%), with 91% of these deaths (225/246) occurring in individuals with low HRR or low METs. Over the first 10 years of follow-up, there were 50 CVD deaths in men and 30 in women. Kaplan-Meier curves for survival free of CVD death diverged early in follow-up and continued to diverge for the group with both low HRR and low METs compared with the other 2 groups (Figure 1). Survival curves for those with either low HRR or low METs separated late in follow-up in both sexes.
Relative Risk of CVD Death by HRR/METs and FRS
HRR and METs separately were highly significant predictors of fatal CVD. After adjustment for FRS categories (FRS <6%, 6% to 9%, and 10% to 19%), hazard ratios for low compared with high METs for women and men, respectively, were 3.09 (95% CI 1.90 to 5.05) and 2.59 (95% CI 1.74 to 3.87; P<0.001 for both sexes). Similarly, the FRS-adjusted hazard ratios for low HRR for women and men, respectively, were 3.25 (95% CI 2.02 to 5.24) and 1.99 (95% CI 1.37 to 2.89; P<0.001 for both sexes).
When both HRR and METs were low (Table 3), the FRS-adjusted hazard ratios for women and men were higher, especially at 20-year follow-up (adjusted hazard ratios, respectively, were 8.51, 95% CI 3.65 to 19.84, and 3.53, 95% CI 2.03 to 6.15, P<0.001 for both sexes). There was no statistically significant interaction between HRR and METs categories for CVD death. When the 10- and 20-year follow-up hazard ratios were compared, the relative effect of FRS appeared to lessen with time, whereas that of HRR/METs appeared to increase. However, there was no statistically significant interaction between HRR/METs and follow-up period for the outcome of CVD death (Pinteraction=0.83 and 0.47 for men and women, respectively). There was a borderline statistically significant interaction of sex and HRR/METs (Pinteraction=0.08). Likelihood ratio tests comparing Cox regression models that included HRR/METs and FRS to models limited to FRS alone showed a highly statistically significant improvement to risk prediction for both sexes with the addition of HRR/METs, although the magnitude of improvement was stronger with longer duration of follow-up (P<0.0001 for both sexes).
At both 10- and 20-year follow-up, HRR/METs additionally stratified the relative risk for CVD death across all FRS categories, even after further adjustment for the range of FRS values within each FRS category (Figure 2). For individuals with FRS <6%, the impact of having low HRR/low METs on the relative risk of CVD death was somewhat higher after 20 years than 10 years (adjusted hazard ratio compared with individuals with high HRR/high METS, at 20 years: 7.86, 95% CI 3.73 to 16.56, P<0.001; at 10 years: 3.29, 95% CI 1.05 to 19.36, P=0.04). In the group with FRS 6% to 19%, the relative risk of low HRR/low METs was similar and highly statistically significant (P<0.001) at both 10- and 20-year follow-up.
Absolute Risk of CVD Death by HRR/METs and FRS
Average annual CVD mortality rates (per 1000 person-years) stratified by HRR/METs and FRS categories are shown in Table 4. Men with FRS <10% and women with FRS <6% had low absolute CVD mortality rates (<5/1000 or 0.5% per year), regardless of their HRR/METs. For individuals with FRS 6% to 19%, exercise testing clearly identified individuals with high absolute CVD mortality rates (>5/1000 per year, rates that are considered high risk by European Task Force guidelines and that correspond to rates of hard CHD events >1% per year6,26).
In men with FRS 10% to 19%, those with low HRR/low METs (399/847 or 47%) were identified as high risk on the basis of annual CVD mortality rates >5/1000. Similarly, for women with FRS 6% to 9% and 10% to 19%, those with low HRR/low METs (70/134 or 52% of women with FRS 6% to 9% and 34/62 or 55% of women with FRS 10% to 19%) were identified as high risk on the basis of annual rates >5/1000. Women with FRS 10% to 19% and either low HRR or low METs had a trend toward higher annual CVD mortality rates, but the confidence interval included values <5/1000.
ST-Segment Depression
When exercise-induced ST-segment depression 1 mm was added to the models in Table 2 that included FRS and HRR/METs, there was increased risk associated with ST-segment depression in men but not in women (adjusted hazard ratio for ST depression at 20 years for men 2.26, 95% CI 1.28 to 4.01, P=0.005, and 1.47, 95% CI 0.73 to 2.93, P=0.29 for women). At 10-year follow-up, there was a trend for increased risk in men (P=0.06), and ST-segment depression remained nonsignificant for women (P=0.99).
CHD Death, Non-CVD Death, and All-Cause Death
In secondary analyses for CHD death with models similar to those in Table 2 for 20-year follow-up, the adjusted hazard ratios for HRR/METs categories for CHD death in men were similar to those for CVD death. The hazard ratios for HRR/METs categories and CHD death in women were greater than those for CVD death, with a statistically significant interaction between sex and HRR/METs (2=5.99, Pinteraction=0.01). Women with low HRR/low METs had nearly 30-fold increased risk of CHD death compared with women with high HRR/high METs, after adjustment for FRS categories (P=0.001).
When similar analyses were performed for non-CVD death and for all-cause death at 20 years, the adjusted hazard ratios for HRR/METs for men and women were highly significant, and both were similar in magnitude to results for CVD death in men, with no significant interactions with sex. There were no significant interactions between high LDL cholesterol and HRR/METs with respect to any of the mortality outcomes.
Discussion
In this large, prospective North American cohort of asymptomatic individuals followed up for more than 2 decades, we found that 2 easily obtained non-ECG measures from exercise testing (HRR and METs) were strongly predictive of CVD death, incremental to FRS, and valid across FRS risk categories <20%. Using annual CVD mortality rates, we estimated that half of women and just under half of men with FRS 10% to 19%, plus half of women with FRS 6% to 9%, would be reclassified as high risk on the basis of having low HRR/low METs, thus providing clinically important risk information.
The US Preventive Services Task Force recently concluded that exercise testing had poor to fair accuracy in predicting events in asymptomatic individuals based on ECGs.10 Although several studies have previously examined the prognostic relevance of exercise tests,12–18,20,27–31 only a few studies analyzed how non-ECG exercise variables add to FRS estimates.19, 32–33
The present study differs from previous studies by focusing on low-intermediate risk individuals and by examining fatal cardiovascular outcomes in a very large number of men and women with lengthy follow-up. In one previous study, exercise capacity in women predicted all-cause mortality after adjustment for FRS, but CVD data were unavailable.32 In a recent Cleveland Clinic study of asymptomatic individuals, low HRR and low METs were significant predictors of all-cause mortality after global risk adjustment, but they were clinically useful only in those with highest global risk.19 However, the study may have been underpowered to detect clinically meaningful differences in lower-risk individuals compared with the present study. In addition, their sample was self-referred, and there were no cardiovascular outcome data.19 The Framingham Study recently reported that among asymptomatic men, additional prognostic information in FRS-adjusted models was obtained by ST-segment depression, failure to reach target heart rate, and decreased exercise capacity.33 These exercise test variables did not yield incremental prognostic information in women or low-risk men, but there were few CHD events in women.33
One potential strategy to apply exercise testing in the general population would be to screen individuals with intermediate risk (FRS 10% to 19% or 6% to 19%).9 The present data suggest that the impact of adding an exercise test to risk prediction would be to additionally identify as "high risk" (based on an annual CVD mortality rate >0.5%) a substantial proportion of individuals with FRS scores between 6% and 19%. Even though there were only 74 women in the FRS category of 10% to 19%, having a low HRR and low METs was a highly significant predictor in both men and women across all FRS categories. HRR/METs data may be used to inform clinicians’ decisions about the optimal degree of primary prevention. Although exercise testing also risk-stratified those with FRS <6%, their absolute CVD mortality rates were low, and thus exercise testing in this low-risk group may not provide important clinical information. The present study offers further data in women and men to support the call for a clinical outcomes trial that addresses survival benefit using targeted risk factor reduction and improved fitness strategies based on exercise testing.34
There are several limitations to the present study. One limitation is the generalizability of our risk estimates to nonwhites, even though the cohort was recruited from different geographic sites with a range of socioeconomic and occupational status.20,35 Another limitation is that 40% of the participants were enrolled on the basis of lipid abnormalities. Despite the higher prevalence of dyslipidemia in this cohort, their mean FRS values and mortality rates were similar to those found in populations not enriched with lipid abnormalities,6 with no significant interaction between high LDL cholesterol and mortality. The exercise test protocol used in the present study was a submaximal Bruce protocol, with termination at 90% of maximal predicted heart rate, and it is unclear whether our results would apply to other types of stress protocols. Our findings potentially may not be as applicable to soft CHD end points, such as angina or nonfatal myocardial infarction. A recent study suggested that exercise capacity was more predictive of fatal than nonfatal CVD events, although the potential mechanisms are unclear.36 Even in this large cohort, there were very few deaths in the high HRR/high METs group (the reference group), which may have resulted in large confidence intervals for some hazard ratios, such as those for CHD death in women. Finally, there are currently no large, randomized trials that show the clinical utility of screening exercise tests.
In conclusion, non-ECG exercise test results (HRR and METs) added important and incremental prognostic information to FRS in predicting fatal CVD in low- and intermediate-risk individuals. The application of these 2 simple, noninvasive exercise measures to individuals with FRS 6% to 19% may identify a significant proportion of individuals who are at high risk but who would have been classified as at low or intermediate risk by FRS alone. Selective use of exercise testing may help tailor primary prevention strategies that would match the level of risk to the intensity of therapy, with the potential to help close the sizable gap between detection of and death due to CVD.
Acknowledgments
The authors thank the LRC investigators, staff, and participants for their valuable contributions.
Disclosure
Dr Blumenthal has received research funding and served on the speakers’ bureaus of or received honoraria from Merck, Pfizer, Wyeth, Novartis, Kos, and the Maryland Atheletic Club Charitable Foundation, Lutherville, Md.
Footnotes
Presented in part at the 75th Annual Scientific Sessions of the American Heart Association, Chicago, Ill, November 17–20, 2002, and published in abstract form [Circulation. 2002;106(suppl II):II-601].
References
Goldman L, Garber AM, Grover SA, Hlatky MA. 27th Bethesda Conference: matching the intensity of risk factor management with the hazard for coronary disease events: Task Force 6: cost effectiveness of assessment and management of risk factors. J Am Coll Cardiol. 1996; 27: 1020–1030.
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) final report. Circulation. 2002; 106: 3143–3421.
American Heart Association. Heart Disease and Stroke Statistics—2004 Update. Dallas, Tex: American Heart Association; 2003.
Pasternak RC, Abrams J, Greenland P, Smaha LA, Wilson PW, Houston-Miller N. 34th Bethesda Conference: Task force #1: identification of coronary heart disease risk: is there a detection gap J Am Coll Cardiol. 2003; 41: 1863–1874.
Ford ES, Giles WH, Mokdad AH. The distribution of 10-year risk for coronary heart disease among US adults: findings from the National Health and Nutrition Examination Survey III. J Am Coll Cardiol. 2004; 43: 1791–1796.
Lloyd-Jones DM, Larson MG, Beiser A, Levy D. Lifetime risk of developing coronary heart disease. Lancet. 1999; 353: 89–92.
Lloyd-Jones DM, Wilson PW, Larson MG, Beiser A, Leip EP, D’Agostino RB, Levy D. Framingham risk score and prediction of lifetime risk for coronary heart disease. Am J Cardiol. 2004; 94: 20–24.
Pearson TA. New tools for coronary risk assessment: what are their advantages and limitations Circulation. 2002; 105: 886–892.
Greenland P, Smith SC Jr, Grundy SM. Improving coronary heart disease risk assessment in asymptomatic people: role of traditional risk factors and noninvasive cardiovascular tests. Circulation. 2001; 104: 1863–1867.
Fowler-Brown A, Pignone M, Pletcher M, Tice JA, Sutton SF, Lohr KN. Exercise tolerance testing to screen for coronary heart disease: a systematic review for the technical support for the U.S. Preventive Services Task Force. Ann Intern Med. 2004; 140: W9–W24.
Gibbons RJ, Balady GJ, Bricker JT, Chaitman BR, Fletcher GF, Froelicher VF, Mark DB, McCallister BD, Mooss AN, O’Reilly MG, Winters WL Jr, Antman EM, Alpert JS, Faxon DP, Fuster V, Gregoratos G, Hiratzka LF, Jacobs AK, Russell RO, Smith SC Jr. ACC/AHA 2002 guideline update for exercise testing: summary article: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1997 Exercise Testing Guidelines). Circulation. 2002; 106: 1883–1892.
Mora S, Redberg RF, Cui Y, Whiteman MK, Flaws JA, Sharrett AR, Blumenthal RS. Ability of exercise testing to predict cardiovascular and all-cause death in asymptomatic women: a 20-year follow-up of the Lipid Research Clinics Prevalence Study. JAMA. 2003; 290: 1600–1607.
Blair SN, Kampert JB, Kohl HW III, Barlow CE, Macera CA, Paffenbarger RS Jr, Gibbons LW. Influences of cardiorespiratory fitness and other precursors on cardiovascular disease and all-cause mortality in men and women. JAMA. 1996; 276: 205–210.
Blair SN, Kohl HW III, Barlow CE, Paffenbarger RS Jr, Gibbons LW, Macera CA. Changes in physical fitness and all-cause mortality: a prospective study of healthy and unhealthy men. JAMA. 1995; 273: 1093–1098.
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: 1351–1357.
Ekelund LG, Haskell WL, Johnson JL, Whaley FS, Criqui MH, Sheps DS. Physical fitness as a predictor of cardiovascular mortality in asymptomatic North American men: the Lipid Research Clinics Mortality Follow-up Study. N Engl J Med. 1988; 319: 1379–1384.
Kwok JM, Miller TD, Hodge DO, Gibbons RJ. Prognostic value of the Duke treadmill score in the elderly. J Am Coll Cardiol. 2002; 39: 1475–1481.
Myers J, Prakash M, Froelicher V, Do D, Partington S, Atwood JE. Exercise capacity and mortality among men referred for exercise testing. N Engl J Med. 2002; 346: 793–801.
Aktas MK, Ozduran V, Pothier CE, Lang R, Lauer MS. Global risk scores and exercise testing for predicting all-cause mortality in a preventive medicine program. JAMA. 2004; 292: 1462–1468.
Gordon DJ, Ekelund LG, Karon JM, Probstfield JL, Rubenstein C, Sheffield LT, Weissfeld L. Predictive value of the exercise tolerance test for mortality in North American men: the Lipid Research Clinics Mortality Follow-up Study. Circulation. 1986; 74: 252–261.
Cui Y, Blumenthal RS, Flaws JA, Whiteman MK, Langenberg P, Bachorik PS, Bush TL. Non-high-density lipoprotein cholesterol level as a predictor of cardiovascular disease mortality. Arch Intern Med. 2001; 161: 1413–1419.
Bruce RA, Kusumi F, Hosmer D. Maximal oxygen intake and nomographic assessment of functional aerobic impairment in cardiovascular disease. Am Heart J. 1973; 85: 546–562.
Sheffield LT, Haskell W, Heiss G, Kioschos M, Leon A, Roitman D, Schrott H. Safety of exercise testing volunteer subjects: the Lipid Research Clinics’ Prevalence Study experience. J Cardiac Rehab. 1982; 2: 395–400.
Shetler K, Marcus R, Froelicher VF, Vora S, Kalisetti D, Prakash M, Do D, Myers J. Heart rate recovery: validation and methodologic issues. J Am Coll Cardiol. 2001; 38: 1980–1987.
STATA Reference Manual, Release 8.1. College Station, Tex: Stata Corporation; 2003.
De Backer G, Ambrosioni E, Borch-Johnsen K, Brotons C, Cifkova R, Dallongeville J, Ebrahim S, Faergeman O, Graham I, Mancia G, Manger Cats V, Orth-Gomer K, Perk J, Pyorala K, Rodicio JL, Sans S, Sansoy V, Sechtem U, Silber S, Thomsen T, Wood D. European guidelines on cardiovascular disease prevention in clinical practice: Third Joint Task Force of European and Other Societies on Cardiovascular Disease Prevention in Clinical Practice. Eur Heart J. 2003; 24: 1601–1610.
Allen WH, Aronow WS, Goodman P, Stinson P. Five-year follow-up of maximal treadmill stress test in asymptomatic men and women. Circulation. 1980; 62: 522–527.
Lauer MS, Francis GS, Okin PM, Pashkow FJ, Snader CE, Marwick TH. Impaired chronotropic response to exercise as a predictor of mortality. JAMA. 1999; 281: 524–529.
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: 552–555.
Jouven X, Zureik M, Desnos M, Courbon D, Ducimetiere P. Long-term outcome in asymptomatic men with exercise-induced premature ventricular depolarizations. N Engl J Med. 2000; 343: 826–833.
Cheng YJ, Lauer MS, Earnest CP, Church TS, Kampert JB, Gibbons LW, Blair SN. Heart rate recovery following maximal exercise testing as a predictor of cardiovascular disease and all-cause mortality in men with diabetes. Diabetes Care. 2003; 26: 2052–2057.
Gulati M, Pandey DK, Arnsdorf MF, Lauderdale DS, Thisted RA, Wicklund RH, Al-Hani AJ, Black HR. Exercise capacity and the risk of death in women: the St James Women Take Heart Project. Circulation. 2003; 108: 1554–1559.
Balady GJ, Larson MG, Vasan RS, Leip EP, O’Donnell CJ, Levy D. Usefulness of exercise testing in the prediction of coronary disease risk among asymptomatic persons as a function of the Framingham risk score. Circulation. 2004; 110: 1920–1925.
Mark DB, Lauer MS. Exercise capacity: the prognostic variable that doesn’t get enough respect. Circulation. 2003; 108: 1534–1536.
Bush TL, Cowan LD, Barrett-Connor E, Criqui MH, Karon JM, Wallace RB, Tyroler HA, Rifkind BM. Estrogen use and all-cause mortality: preliminary results from the Lipid Research Clinics Program Follow-Up Study. JAMA. 1983; 249: 903–906.
Laukkanen JA, Kurl S, Salonen R, Rauramaa R, Salonen JT. The predictive value of cardiorespiratory fitness for cardiovascular events in men with various risk profiles: a prospective population-based cohort study. Eur Heart J. 2004; 25: 1428–1437(Samia Mora, MD, MHS; Rita)