Cardiorespiratory fitness (CRF) is defined as the body’s maximal ability to transport and utilize oxygen at the tissue level and is dependent primarily on maximal cardiac output and maximal arterial–venous oxygen difference as well as efficient shunting of blood to working skeletal muscles. CRF is influenced by genetic factors (4) and by habitual physical activity (16) and can be objectively measured via maximal exercise testing (2). Over the past three decades, a low level of CRF has been shown to be a significant predictor of various adverse health outcomes in men including all-cause (3,19,22,27,36) cardiovascular (3,6,10,22,30,36) and cancer (11,12,23,32) mortality as well as incidence of type 2 diabetes (24,31), hypertension (7), metabolic syndrome (17,21), and cardiovascular morbidity (20). Furthermore, Gupta et al. (15) recently showed that a single measurement of CRF significantly improved classification of both short-term and long-term risk for cardiovascular disease (CVD) mortality when added to traditional risk factors. The Cooper Center Longitudinal Study (CCLS) has made significant contributions to the literature in this regard. In a landmark article using age-adjusted quintiles to classify CRF categories, Blair et al. (3) demonstrated that CRF was strongly and inversely related to all-cause mortality in both men and women. Because the largest decrement in all-cause mortality was seen when comparing the lowest CRF quintile with the next higher quintile, the lowest quintile has been designated as “unfit” or “low CRF” in many CCLS articles published since then.
The notion that low CRF is a strong predictor of morbidity and mortality is not limited to the CCLS. For example, the Lipid Research Clinics Mortality Follow-up Study (10), Seattle Heart Watch (6), Baltimore Longitudinal Study of Aging (35), the Palo Alto Veterans Study (26), and the U.S. Railroad Study (33) have all shown that low CRF is significantly associated with CVD. In a recent meta-analysis by Kodama et al. (19) using 33 observational cohort studies, those with low CRF (<7.9 METs) had relative risk values of 1.70 and 1.56 for all-cause mortality and CVD events, respectively, when compared with those with high CRF (≥10.9 METs). When compared with individuals with intermediate CRF (7.9–10.8 METs), those with low CRF had relative risk values of 1.40 and 1.57 for all-cause mortality and CVD events, respectively (P < 0.001 for all comparisons).
Although it is evident that low CRF is a strong and independent risk factor for morbidity and mortality, very little is known regarding the gradient of risk within the low CRF category. For example, it is not known whether a risk gradient exists, whether it is linear, or whether there is a threshold for significantly increased risk within the low CRF category. Physical activity pattern and other coronary risk factor comparisons between generally healthy individuals at the low, intermediate, and high spectrum of the low CRF category have not been examined. In addition, studies concerning hemodynamic responses to maximal exercise testing; for example, double product reserve, are sparse.
Therefore, the primary purpose of the current study was to examine the risk gradient for all-cause mortality in otherwise apparently healthy men with low CRF. We will also compare physical activity patterns, other coronary risk factors, and hemodynamic variables across the spectrum of the low CRF category in this cohort.
Study participants and measurements
Briefly, the aim of the CCLS is to examine prospectively the association of health behaviors and chronic disease biomarkers with long-term health in men and women. Participants in the present study were 6251 men between the ages of 40 and 69 yr with low CRF, who completed baseline examinations at the Cooper Clinic in Dallas, TX, between 1971 and 2006. All participants were U.S. residents, and the majority (approximately 95%) of men were white and from middle-to-upper socioeconomic strata. After receiving a written informed consent from each participant, a clinical evaluation was performed and included a physician examination, personal and family health history, fasting blood chemistry assessment, anthropometry, resting ECG, and a maximal graded treadmill exercise test. Resting HR and resting blood pressure were measured after 5 min of rest in the recumbent position. Height and weight were measured using a stadiometer and standard physician’s scale. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared. All procedures were administered by trained technicians who followed standardized measurement protocols. The Cooper Clinic laboratory meets the quality control standards of the Centers for Disease Control and Prevention Lipid Standardization Program. The CCLS undergoes annual approval by the Cooper institutional review board.
CRF was quantified as the duration of a maximal treadmill exercise test using a modified Balke protocol (2). The treadmill test began at a speed of 88 m·min−1 and 0% elevation. At the end of the first minute, elevation was increased to 2%, then by 1% each minute thereafter. At 25 min, elevation remained at 25% while speed was increased 5.4 m·min−1 each minute until volitional fatigue. Exercise duration from this protocol has been shown to correlate highly (r = 0.92) with directly measured maximal oxygen uptake in men (28). Patients were given verbal encouragement to achieve maximal effort during the test, and those that did not achieve at least 85% of age-predicted HRmax (n = 1515) were excluded from the analyses to prevent potential misclassification of CRF. To standardize exercise test performance, we computed maximal MET (1 MET = 3.5 mL O2·kg−1·min−1) levels of CRF on the basis of the final treadmill speed and grade (1). Men were divided into age-specific quintiles of CRF as described previously (3). In addition, in accordance with previous CCLS studies (3), individuals scoring in the first quintile of CRF in their age group were categorized as low fit. Self-reported physical activity was ascertained using the Physical Activity Index (PAI), which has been described elsewhere (25). HR reserve was calculated by subtracting resting HR from the HRmax achieved during the treadmill test. Double product was calculated by multiplying systolic blood pressure and HR, whereas double product reserve was calculated by subtracting resting double product from maximal exercise double product.
Smoking history was obtained from a standardized questionnaire and grouped categorically for analysis (never, past, or current smoker). Men with previously diagnosed CVD, cancer, or diabetes at baseline (n = 3696) as well as men with BMI <18.5 kg·m−2 (n = 80), those with less than 1 yr of follow-up (n = 276), and those with an abnormal resting or exercise ECG (n = 4117) were excluded from the analyses. After these exclusions, the resultant sample size of the cohort was 6251.
Vital status was ascertained primarily using the National Death Index (NDI). NDI was established by the National Center for Health Statistics to serve as a centralized database of death record information contained within state vital statistics offices. The NDI has a sensitivity of 96% and a specificity of 100% for determining deaths in the general population (34). Once we identified possible decedents, Departments of Vital Statistics in the appropriate states were contacted and official copies of death certificates were requested. We compared information on the death certificates with clinical records to confirm that the death certificate matched the individual. Deaths were identified using the International Classification of Diseases, 9th revision, for deaths occurring before 1999 and 10th revision for deaths during 1999–2008.
Study participants were followed for mortality from the date of their examination to the date of death for decedents or to December 31, 2008, for survivors. We computed descriptive characteristics according to vital status and computed man-years of exposure as the sum of follow-up time among decedents and survivors. After their exercise test, men who scored in the first quintile were categorized as having low CRF and were then further divided into tertiles (low/low, LL; mid/low, ML; high/low, HL) using the age categories of 40–49, 50–59, and 60–69 yr. We next compared descriptive characteristics of the cohort across tertiles of low CRF within these same age categories. After adjustment for examination year and smoking status, all-cause mortality rates across low CRF tertiles in each age group were computed. Separate survival curves were generated for smokers and nonsmokers to evaluate the survival experience of each of the tertiles of low CRF. Cox proportional hazard regression analysis was used to estimate hazard ratios and 95% confidence intervals of all-cause mortality according to exposure categories. All P values are two-sided, and P < 0.05 was considered statistically significant. We used SAS version 9.2, (2008; SAS Institute, Cary, NC) for all statistical analyses.
Baseline characteristics for the cohort are presented in Table 1. Because the Cooper Clinic did not begin measuring HDL-cholesterol (HDL-C) and LDL-cholesterol (LDL-C) until 1978, there is a substantial amount of missing data for these variables. There were 1259 deaths identified during an average of 19.1 ± 10.4 yr of follow-up and 119,394 man-years of exposure. Overall, the baseline characteristics of survivors were more favorable than those of decedents. In Table 2, we compare clinical characteristics of the LL, ML, and HL groups. With the exception of LDL-C concentration and family history of CVD, all of the variables were significantly (P < 0.01) associated with categories of CRF; more favorable values were seen across incremental CRF categories. Self-reported physical activity differed significantly across LL, ML, and HL groups. The proportions of men who reported no physical activity were 67.3%, 59.5%, and 51.1% across LL, ML, and HL groups, respectively (P trend < 0.0001). For men reporting some physical activity, a higher percentage of men was active across LL, ML, and HL groups (P < 0.0001).
In Table 3, all-cause mortality rates across LL, ML, and HL groups are shown for each age group. For men ages 40–49 yr, a significant inverse trend for adjusted all-cause mortality was seen across LL, ML, and HL groups (57.0, 31.1, and 34.4 deaths per 10,000 man-years, respectively; P trend = 0.007). A similar pattern was observed in the 50- to 59-yr age group (102.9, 79.2, and 64.6 deaths per 10,000 man-years for LL, ML, and HL groups, respectively; P trend = 0.02). Although a similar pattern was also observed in the 60- to 69-yr age group (209.7, 125.3, and 118.3 deaths per 10,000 man-years for LL, ML, and HL groups, respectively), the trend approached but did not reach statistical significance (P = 0.09). Further adjustment for BMI did not materially change these results.
When using CRF as a continuous rather than a categorical variable, we found that for every 1-min increment in treadmill time, there were corresponding 9%, 11%, and 15% decreased risk of mortality for 40- to 49-yr-old, 50- to 59-yr-old, and 60- to 69-yr-old groups, respectively. Because clinicians may be more familiar with METs than the modified-Balke treadmill protocol, we also report that for every 1-MET increment in treadmill performance, there were corresponding 16%, 20%, and 23% decreased risk of mortality for these same age groups.
In Table 4, we compare resting and maximal exercise test hemodynamic variables across LL, ML, and HL groups. HR reserve, resting and maximal double product, as well as double product reserve values were more favorable across LL, ML, and HL groups (P trend < 0.0001). In addition, the percentage of predicted of HRmax attained during the treadmill test was higher across the three groups (P < 0.001).
In Figure 1A (smokers) and B (nonsmokers), we present survival curves for LL, ML, and HL men. In both smokers and nonsmokers, men within the LL group experienced significantly greater all-cause mortality risk than ML or HL men, with greater decrement of mortality between the LL and ML groups than that between the ML and HL groups.
Although it is well-known that a nonlinear gradient in all-cause mortality exists between the least-fit and next–least-fit quintiles of apparently healthy men, little is known regarding the risk gradient within the least-fit quintile (low CRF). Thus, the primary purpose of the current study was to examine whether a risk gradient exists for all-cause mortality in otherwise apparently healthy men with low CRF. We also examined clinical characteristics as well as physical activity levels and hemodynamic variables across this group of men.
Our group (13) recently reported on cardiovascular risk factor profiles across quintiles of CRF in a cohort of 59,820 male and 22,192 female Cooper Clinic patients. There was significant inverse relation between CRF and triglycerides, non–HDL-C, triglyceride:HDL-C ratio, and resting blood pressure as well as decreased prevalence of obesity, metabolic syndrome, diabetes, and cigarette smoking across CRF quintiles. Our current observation shows that this relation remains consistent within the lowest quintile of CRF.
Although reports on the mortality risk gradient among men with low CRF (quintile 1) are quite sparse, there is one report comparing mortality within the lowest two CRF quintiles. In a population of 1535 apparently healthy Veterans Administration hospital patients (approximately 95% male), Mandic et al. (26) sought to identify reasons for the disproportionate decrement in mortality between quintiles 1 and 2. Clinical characteristics, exercise test variables, and physical activity patterns were compared in the two groups, who were followed for 8.7 ± 5.3 yr. After adjustment for age and other potentially confounding variables, the approximately 60% difference in all-cause mortality between the first and second quintiles of CRF was found to be due more to differences in physical activity patterns between the two quintiles as opposed to differences in clinical characteristics; that is, health status.
Our current report examines differences in mortality between tertiles of men within the lowest CRF quintile. We observed lower BMI, resting double product, glucose, total cholesterol, and triglycerides, lower smoking prevalence, as well as higher HDL-C and HRmax values across LL, ML, and HL groups. Accordingly, the prevalence of metabolic syndrome (14) was decreased across groups (65%, 56.5%, and 49.2% for LL, ML, and HL groups, respectively; P < 0.001). We also observed higher prevalence of smoking in the LL group (32.2%) than that in the ML and HL groups (26.2% and 25%, respectively). Although the differences in these clinical characteristics were statistically significant, they were not likely large enough to fully explain the steep inverse mortality risk gradient within our low CRF cohort. The differences in many of these clinical characteristics are likely due to the amount of physical activity performed. We note that although LDL-C values were not significantly different across groups, the triglyceride:HDL-C ratio (a proxy for LDL-C particle size and insulin resistance) was significantly lower across groups (6.54, 5.27, and 4.80 for LL, ML, and HL groups, respectively; P < 0.008). This may be an important factor in explaining some of the mortality differences between the three groups. Of interest, the trend across groups for the triglyceride:HDL-C ratio paralleled that for waist circumference. Also of interest is the fact that family history of myocardial infarction/stroke was nearly identical across groups. Because genetics is thought to explain approximately 30%–50% of the population variance in CRF (5) and is a powerful risk factor for many chronic diseases, we must also consider genetics as a factor that may have influenced our results.
The survival curves for nonsmoking men were shifted to the right as compared with the survival curves for smoking men. Although this finding was expected, the curves were very similar in smokers and nonsmokers in that there was a marked difference in survival between LL and ML men, with much less of a difference in survival between ML and HL men. Because this pattern was observed in both smokers and nonsmokers, this would indicate that the greatly increased mortality in LL men was not due solely to cigarette smoking.
Hemodynamic variables both at rest and during maximal exercise were more favorable across our LL, ML, and HL groups. In a study of 15,247 older CCLS men (age 40–59 yr) with varying degrees of CRF, Cheng et al. (8) found that HR reserve was inversely related to all-cause mortality. Among 1759 male veterans who underwent maximal exercise testing, Sadrzadeh Rafie et al. (29) found that age-adjusted double product reserve was the strongest hemodynamic variable in terms of cardiovascular prognosis. Furthermore, the double product reserve was shown to have greater prognostic power than the MET level achieved on the treadmill. Thus, the improved survival of our men across LL, ML, and HL groups may be due partly to more favorable hemodynamic variables (e.g., higher HR reserve and higher double product reserve) across these three groups.
To qualify for the ML or HL CRF category, a 60- to 69-yr-old man would need to achieve maximal MET levels of 7.4–8.1 or 8.2–8.5, respectively. This is equivalent to covering approximately 1 or 1.1 miles, respectively, in the Cooper 12-min Run–Walk Test (9) or achieving a treadmill time of approximately 4.5 or 5.5 min, respectively, on a standard Bruce Treadmill Exercise Test. A much greater reduction in all-cause mortality risk can be achieved for LL CRF men who are able to move to the next highest CRF quintile (quintile 2). These levels of CRF can be achieved by many, perhaps even most, apparently healthy men with LL CRF through modest amounts and intensities of regular physical activity such as brisk walking for 30 min on most days. In a study comparing the effects of weight loss and aerobic exercise training on CAD risk factors in sedentary, obese middle-age and older men, Katzel et al. (18) showed a 17% mean increase (P < 0.001) in CRF in 49 subjects who completed 9 months of moderate aerobic training. The significant increase in V˙O2max in the aerobic training group occurred in the absence of weight loss, showing that the increase in CRF was due to a true training effect and not simply due to weight loss.
An important issue that arises when examining mortality risk across tertiles of the lowest quintile of CRF is perspective. In a subsequent analysis (Fig. 2), we compare trends in mortality rates across tertiles of the lowest CRF quintile and the next highest quintile. The trend across these four groups was significant for the age groups of 40–49 and 50–59 yr (P < 0.003) and approached significance for the 60- to 69-yr-old group (P = 0.09). This suggests that the high mortality risk in quintile 1 is not being driven primarily by men in the LL group and helps confirm the strong inverse association between CRF and mortality across the spectrum of lower levels of CRF. From a public health perspective, although all low-fit men are at substantially higher mortality risk than men with higher levels of CRF, it is important to emphasize that even a very modest improvement in CRF (e.g., 1 MET) of low-fit men will result in approximately 20% lower mortality risk. Such an improvement can be achieved through modest increases in physical activity and weight loss (if necessary). Approximately 60% of men in the lowest CRF quintile reported no physical activity. As is often stated in the exercise science field, a little bit of physical activity is better than none and more is better within reasonable limits.
Among the strengths of the current study are a large and well-characterized cohort of men, an extensive follow-up with a relatively large number of deaths for analysis, and the use of an objective measure for CRF. We adjusted mortality rates for several confounders including age (where applicable), examination year, smoking, and BMI. To decrease the likelihood that preexisting subclinical disease was present at baseline, we excluded men with a BMI <18.5 kg·m−2 and those with less than 1 yr of follow-up. Limiting our analyses to men with three or more years of follow-up (n = 5928) did not materially change our findings (results not shown), which increases our confidence level that our results were not due to the presence of a preexisting subclinical disease. We examined CRF as both a categorical and continuous variable. When using the latter, we found our results to be consistent with those of Mandic et al. (26), who demonstrated a 21% reduction in all-cause mortality for every 1-MET increase in CRF at the low end of the CRF spectrum (quintiles 1 and 2). To our knowledge, ours is the first study to examine the gradient of all-cause mortality risk among men with very low CRF.
This study also has limitations. The cohort is primarily White and from middle-to-upper socioeconomic strata; therefore, our findings must be cautiously interpreted when generalized to women or other populations of men. However, the homogeneity of sociodemographic factors in our population sample strengthens the internal validity of our findings by reducing potential confounding by these issues. We did not have sufficient data for medication use to include in this analysis, which is a potential confounder. We also did not have more extensive information on smoking habits, such as number of pack-years. Because the Cooper Clinic did not begin to measure HDL-C and LDL-C until 1978, we have some missing data for these variables, which in turn limited our ability to report on prevalence of the metabolic syndrome in the cohort. Finally, the relatively small sample size (n = 374) and resultant number of deaths (n = 138) in the 60- to 69-yr-old age group likely contributed to the nonsignificant mortality trend in this age group (P < 0.09). However, we point out that the absolute differences in mortality risk across tertiles in this age group are robust, with an approximately 40% decrease in risk between the LL tertile and the remaining two tertiles.
Previous work with this and other cohorts has shown that men with low CRF (quintile 1) are at disproportionately greater risk for all-cause mortality relative to men with moderate CRF (quintiles 2–3). However, it is important to recognize that within the low CRF group, men with the lowest CRF levels (LL) are at substantially greater risk than men with low CRF who are in the ML and HL groups. Thus, it seems imperative to first identify LL CRF men and then target this very high-risk group for physical activity intervention. These results are consistent with 2008 physical activity guidelines (16), which state that those with the lowest level of physical activity are at the greatest risk and that these individuals seem to benefit from small increases in physical activity. At the same time, we must reiterate that men in the ML and HL groups still have a higher mortality rate than men in the second quintile of CRF. Thus, it is extremely important to identify and target ML and HL men for physical activity intervention as well.
To summarize, we observed significantly lower adjusted all-cause mortality across LL, ML, and HL groups. These differences in mortality are likely partially explained by a combination of variables including coronary risk factor and hemodynamic response differences as well as physical activity differences among the three groups. Health and fitness professionals should recognize very low CRF as a major and modifiable risk factor for all-cause mortality. Additional investigations regarding the mechanisms responsible for increased mortality among men with very low levels of CRF are warranted.
We wish to thank Dr. Kenneth H. Cooper for establishing the CCLS, as well as the Cooper Clinic physicians, patients, and technicians.
There is no funding to report for this work. Each author declares no conflict of interest.
The results of this study do not constitute endorsement by the American College of Sports Medicine.
1. American College of Sports Medicine. Guidelines for exercise testing and prescription
. Philadelphia (PA): Lippincott, Williams & Wilkins; 2009.
2. Balke B, Ware RW. An experimental study of physical fitness of Air Force personnel. U S Armed Forces Med J
. 1959; 10 (6): 675–88.
3. Blair SN, Kohl HW III, Paffenbarger RS Jr, Clark DG, Cooper KH, Gibbons LW. Physical fitness and all-cause mortality. A prospective study of healthy men and women. JAMA
. 1989; 262 (17): 2395–401.
4. Bouchard C, Malina R, Perusse L. Genetics of cardiorespiratory fitness phenotypes. In: Genetics of Fitness and Physical Performance
. Champaign (IL): Human Kinetics, 1997: 243–6.
5. Bouchard C, An P, Rice T, et al. Familial aggregation of VO2max
response to exercise training: results from the HERITAGE Family Study. J Appl Physiol (1985)
. 1999; 87 (3): 1003–8.
6. Bruce RA, Hossack KF, DeRouen TA, Hofer V. Enhanced risk assessment for primary coronary heart disease events by maximal exercise testing: 10 years’ experience of Seattle Heart Watch. J Am Coll Cardiol
. 1983; 2 (3): 565–73.
7. Chase NL, Sui X, Lee DC, Blair SN. The association of cardiorespiratory fitness and physical activity with incidence of hypertension in men. Am J Hypertens
. 2009; 22 (4): 417–24.
8. Cheng YJ, Macera CA, Church TS, Blair SN. Heart rate reserve as a predictor of cardiovascular and all-cause mortality in men. Med Sci Sports Exerc
. 2002; 34 (12): 1873–8.
9. Cooper KH. A means of assessing maximal oxygen intake. Correlation between field and treadmill testing. JAMA
. 1968; 203 (3): 201–4.
10. 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 (21): 1379–84.
11. Evenson KR, Stevens J, Cai J, Thomas R, Thomas O. The effect of cardiorespiratory fitness and obesity on cancer mortality in women and men. Med Sci Sports Exerc
. 2003; 35 (2): 270–7.
12. Farrell SW, Cortese GM, Lamonte MJ, Blair SN. Cardiorespiratory fitness, different measures of adiposity, and cancer mortality in men. Obesity (Silver Spring)
. 2007; 15 (12): 3140–9.
13. Grundy SM, Barlow CE, Farrell SW, Vega GL, Haskell WL. Cardiorespiratory fitness and metabolic risk. Am J Cardiol
. 2012; 109: 988–93.
14. Grundy SM, Cleeman JI, Daniels SR, et al. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation
. 2005; 112 (17): 2735–52.
15. Gupta S, Rohatgi A, Ayers CR, et al. Cardiorespiratory fitness and classification of risk of cardiovascular disease mortality. Circulation
. 2011; 123 (13): 1377–83.
16. Haskell WL, Lee IM, Pate RR, et al. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Circulation
. 2007; 116 (9): 1081–93.
17. Jurca R, LaMonte MJ, Church TS, et al. Associations of muscle strength and aerobic fitness with metabolic syndrome in men. Med Sci Sports Exerc
. 2004; 36 (8): 1301–7.
18. Katzel LI, Bleecker ER, Colman EG, Rogus EM, Sorkin JD, Goldberg AP. Effects of weight loss vs aerobic exercise training on risk factors for coronary disease in healthy, obese, middle-aged and older men. A randomized controlled trial. JAMA
. 1995; 274 (24): 1915–21.
19. Kodama S, Saito K, Tanaka S, et al. Cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women: a meta-analysis. JAMA
. 2009; 301 (19): 2024–35.
20. Lakka TA, Venalainen JM, Rauramaa R, Salonen R, Tuomilehto J, Salonen JT. Relation of leisure-time physical activity and cardiorespiratory fitness to the risk of acute myocardial infarction. N Engl J Med
. 1994; 330 (22): 1549–54.
21. Lamonte MJ, Barlow CE, Jurca R, Kampert JB, Church TS, Blair SN. Cardiorespiratory fitness is inversely associated with the incidence of metabolic syndrome: a prospective study of men and women. Circulation
. 2005; 112 (4): 505–12.
22. Laukkanen JA, Lakka TA, Rauramaa R, et al. Cardiovascular fitness as a predictor of mortality in men. Arch Intern Med
. 2001; 161 (6): 825–31.
23. Laukkanen JA, Pukkala E, Rauramaa R, Makikallio TH, Toriola AT, Kurl S. Cardiorespiratory fitness, lifestyle factors and cancer risk and mortality in Finnish men. Eur J Cancer
. 2010; 46 (2): 355–63.
24. Lee DC, Sui X, Church TS, Lee IM, Blair SN. Associations of cardiorespiratory fitness and obesity with risks of impaired fasting glucose and type 2 diabetes in men. Diabetes Care
. 2009; 32 (2): 257–62.
25. Macera CA, Jackson KL, Hagenmaier GW, Kronenfeld JJ, Kohl HW, Blair SN. Age, physical activity, physical fitness, body composition, and incidence of orthopedic problems. Res Q Exerc Sport
. 1989; 60 (3): 225–33.
26. Mandic S, Myers JN, Oliveira RB, Abella JP, Froelicher VF. Characterizing differences in mortality at the low end of the fitness spectrum. Med Sci Sports Exerc
. 2009; 41 (8): 1573–9.
27. Myers J, Kaykha A, George S, et al. Fitness versus physical activity patterns in predicting mortality in men. Am J Med
. 2004; 117 (12): 912–8.
28. Pollock ML, Bohannon RL, Cooper KH, et al. A comparative analysis of four protocols for maximal treadmill stress testing. Am Heart J
. 1976; 92 (1): 39–46.
29. Sadrzadeh Rafie AH, Sungar GW, Dewey FE, Hadley D, Myers J, Froelicher VF. Prognostic value of double product reserve. Eur J Cardiovasc Prev Rehabil
. 2008; 15 (5): 541–7.
30. Sandvik L, Erikssen J, Thaulow E, Erikssen G, Mundal R, Rodahl K. Physical fitness as a predictor of mortality among healthy, middle-aged Norwegian men. N Engl J Med
. 1993; 328 (8): 533–7.
31. Sawada SS, Lee IM, Naito H, et al. Long-term trends in cardiorespiratory fitness and the incidence of type 2 diabetes. Diabetes Care
. 2010; 33 (6): 1353–7.
32. Sawada SS, Muto T, Tanaka H, et al. Cardiorespiratory fitness and cancer mortality in Japanese men: a prospective study. Med Sci Sports Exerc
. 2003; 35 (9): 1546–50.
33. Slattery ML, Jacobs DR Jr. Physical fitness and cardiovascular disease mortality. The US Railroad Study. Am J Epidemiol
. 1988; 127 (3): 571–80.
34. Stampfer MJ, Willett WC, Speizer FE, et al. Test of the National Death Index. Am J Epidemiol
. 1984; 119 (5): 837–9.
35. Talbot LA, Morrell CH, Metter EJ, Fleg JL. Comparison of cardiorespiratory fitness versus leisure time physical activity as predictors of coronary events in men aged < or = 65 years and > 65 years. Am J Cardiol
. 2002; 89 (10): 1187–92.
36. Wei M, Kampert JB, Barlow CE, et al. Relationship between low cardiorespiratory fitness and mortality in normal-weight, overweight, and obese men. JAMA
. 1999; 282 (16): 1547–53.