Abstract
Objective
To predict risk for non-fatal cardiovascular disease (CVD) and disease-specific mortality using CRF algorithms that do not involve exercise testing.
Background
Cardiorespiratory fitness (CRF) is not routinely measured, as it requires trained personnel and specialized equipment.
Methods
Participants were 43,356 adults (21% women) from the Aerobics Center Longitudinal Study followed between 1974 and 2003. Estimated CRF was based on sex, age, body mass index, waist circumference, resting heart rate, physical activity level and smoking status. Actual CRF was measured by a maximal treadmill test.
Results
During a median follow-up of 14.5 years, 1,934 deaths occurred, 627 due to CVD. In a sub-sample of 18,095 participants, 1,049 cases of non-fatal CVD events were ascertained. After adjusting for potential confounders, both measured CRF and estimated CRF were inversely associated with risk of all-cause mortality, CVD mortality and non-fatal CVD incidence in men, and with all-cause mortality and non-fatal CVD in women. The risk reduction per 1-metabolic equivalent (MET) increase ranged approximately from 10 to 20 %. Measured CRF had a slightly better discriminative ability (c-statistic) than estimated CRF, and the net reclassification improvement (NRI) of measured CRF vs. estimated CRF was 12.3% in men (p<0.05) and 19.8% in women (p<0.001).
Conclusions
These algorithms utilize information routinely collected to obtain an estimate of CRF that provides a valid indication of health status. In addition to identifying people at risk, this method can provide more appropriate exercise recommendations that reflect initial CRF levels.
Keywords: cardiorespiratory fitness, mortality, cardiovascular disease, algorithms
INTRODUCTION
Low cardiorespiratory fitness (CRF) is associated with increased risk of cardiovascular disease (CVD), type 2 diabetes, and all-cause and disease-specific mortality (1–4). The most valid measure of CRF is cardiopulmonary exercise testing (CPX) with maximal oxygen uptake (VO2max) measured by ventilatory expired gas analysis (5). The common approach, however, is to calculate VO2max from total test time, which has demonstrated to be highly valid and easier to use widely (6,7). In a healthy 40 year-old, 70 kg man, 1 metabolic equivalent (MET) unit is defined as 3.5 mL O2 · kg−1 · min−1 (8), and every 1-MET increase has been associated with a 13% and 15% risk reduction of all-cause mortality and CVD events, respectively (9).
Unlike other important risk factors, CRF is not routinely measured. It requires trained personnel to administer an exercise test using specialized equipment. With the publication of non-exercise algorithms, it is now feasible to estimate CRF with reasonable accuracy using health indicators typically available in field and healthcare settings. Over two decades ago, we published the first equations (10) and many others have been developed more recently (11–15). Although they provide accurate estimates at the population level, these models were developed with cross-sectional data. Furthermore, age was included as a linear term and recent longitudinal data demonstrated that CRF declines nonlinearly with aging (16). To address this issue, we recently developed new longitudinal algorithms that estimate CRF changes associated with aging (17). The error estimates ranged from 1.41 to 1.69 METs (17).
Whether estimated CRF can predict health risk is yet to be determined. To the best of our knowledge, only Stamatakis et al. (18) have explored the association between estimated CRF and mortality. After a mean follow-up of 9 years, a higher level of estimated CRF was associated with a lower risk of mortality from all causes and CVD (18). Interestingly, estimated CRF had a better discriminative ability (as judged by c-statistic) than any of its modifiable components (body mass index [BMI], self-reported physical activity, and resting heart rate [RHR]) (18).
Given the underutilization of fitness testing and the potential of estimated CRF as demonstrated by Stamatakis et al. (18), the purpose of the present analyses is to examine the capacity of our new longitudinal CRF algorithms to predict incident CVD and disease-specific mortality. This study will add to the previous one (18) by comparing risk predictive capacity with measured CRF and by adding non-fatal major CVD events, in a large established database with long follow-up.
METHODS
Study population
The Aerobics Center Longitudinal Study (ACLS) is a prospective observational study of adult men and women who underwent preventive medical evaluations at the Cooper Clinic (Dallas, TX) (1). Participants were unpaid volunteers, mostly non-Hispanic whites, well educated, and worked in executive or professional positions. All participants provided written informed consent and the study protocol was approved annually.
Inclusion criteria for the present analysis were: no existing CVD (myocardial infarction [MI] or stroke) or cancer at baseline; achieving 85% or more of the individual’s age-predicted maximal heart rate (220 – age) during the treadmill exercise testing; BMI ≥18.5 kg/m2; ≥1 year of follow-up; complete data on CRF, mortality outcomes, covariables and all parameters included in the CRF algorithms (17). These criteria resulted in 43,356 individuals (21% women) aged 20 to 84 years who underwent a baseline examination between 1974 and 2002.
Baseline examination
The clinical examinations were completed after an overnight fast. Height and weight were measured on a physician’s scale and stadiometer, and BMI was calculated. Waist circumference (WC) was measured level with the umbilicus. Resting systolic and diastolic blood pressures were measured with a mercury sphygmomanometer using standard auscultation methods (19). Blood chemistries were analyzed with automated bioassays in the Cooper Clinic laboratory. Concentrations of total cholesterol and fasting plasma glucose were measured in accordance with the standards of the CDC Lipid Standardization Program (20).
Physical activity
A formerly validated questionnaire was used to assess self-reported leisure-time physical activity (21). A five-level physical activity index was created (17): no regular activity (level 0); some regular activity such as bicycling, swimming, racquet sports, and other strenuous sports, but not walking or jogging (level 1); walking or jogging < 10 miles per week (level 2); walking or jogging 10–20 miles (level 3); and walking or jogging > 20 miles (level 4). Walking and jogging were used as the basis for physical activity because they were the most common activities in this population. A second physical activity index was defined as inactive (levels 0–2) or active (levels 3, 4) to match as closely as possible the consensus recommendation of 150 minutes per week of aerobic activity (22).
Measured CRF
Measured CRF was quantified as the duration of a symptom-limited maximal treadmill exercise test using a modified Balke protocol (1,23). Patients were encouraged to give maximal effort, and the test endpoint was volitional exhaustion or termination by the physician for medical reasons. We calculated METs from the final treadmill speed and grade (24). Exercise treadmill duration on this protocol is highly correlated (r ≥ 0.92) with measured peak oxygen uptake (6,7). Participants were classified into lower, middle and upper groups based on age- (20–39, 40–49, 50–59, ≥ 60 yr) and sex-specific thirds of METs distribution.
Estimated CRF
Four different sex-specific algorithms were created to estimate CRF, based on age, BMI (or percent body fat), WC, RHR, physical activity index (in two or five levels) and smoking status, as previously described and validated (17). For a higher applicability, the present analyses focus on those algorithms that include BMI (rather than body fat) and physical activity in two levels (rather than five):
-
Women
Estimated CRF (METs) = 14.7873 + (age × 0.1159) – (age2 × 0.0017) – (BMI × 0.1534) – (WC × 0.0085) – (RHR × 0.0364) + (active × 0.5987) – (smoker × 0.2994)
-
Men
Estimated CRF (METs) = 21.2870 + (age × 0.1654) – (age2 × 0.0023) – (BMI × 0.2318) – (WC × 0.0337) – (RHR × 0.0390) + (active × 0.6351) – (smoker × 0.4263)
where active = 1 if the participant was classified as physically active, 0 if inactive; and smoker = 1 if current smoker, 0 if not. Once the algorithms were implemented, participants were classified into lower, middle and upper groups based on age- (20–39, 40–49, 50–59, ≥ 60 yr) and sex-specific thirds of METs distribution.
Assessment of outcomes
Participants were followed from baseline examination until the date of death or 31 December 2003. Mortality surveillance was based on the national death index (NDI). The underlying cause of death was determined from the NDI report or by a nosologist’s review of official death certificates. CVD mortality was defined by International Classification of Diseases, Ninth Revision (ICD-9) codes 390 to 449.9 before 1999 and Tenth Revision (ICD-10) codes I00 to I78 during 1999–2003 (25).
Incidence of non-fatal CVD events was ascertained in a subsample of 18,095 individuals (20% women) from responses to mail-back health surveys in 1982, 1999, and 2004. The aggregate survey response rate across all survey periods in the ACLS is 65% to 75% (26). Baseline health histories and clinical measures were similar between responders and non-responders and between early and late responders (27).
Non-fatal CVD endpoints were defined as diagnosis by a physician of MI, stroke, or a coronary revascularization procedure. In participants reporting multiple events, the first event was used for analysis. In a random sample of these endpoints, we applied a standard definition for defining and adjudicating MI, revascularization and stroke (28). The percentage of agreement between reported events and participants’ medical records was 88, 100, and 89% for MI, revascularization, and stroke, respectively (26).
Statistical Analysis
Participants’ baseline characteristics were summarized based on sex and estimated CRF level, using analysis of the variance (ANOVA) and chi-square tests. We used Cox proportional hazards regression to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) according to CRF levels (both measured and estimated). Multivariable analyses included these covariables: age (years), examination year, alcohol intake [heavy drinker or not: >14 and >7 drinks / wk for males and females, respectively (29)], presence or absence of hypercholesterolemia (total cholesterol ≥240 mg/dL or previous physician diagnosis), hypertension (resting blood pressure ≥ 140/90 mm Hg or previous physician diagnosis), diabetes mellitus (fasting blood glucose ≥126 mg/dL, previous physician diagnosis or use of insulin), abnormal resting or exercise ECG, and parental history of CVD. Sensitivity analyses compared the performance of the algorithms with five and two levels of physical activity, and also excluded intermediate covariables that could be on the causal pathway (hypercholesterolemia, hypertension, diabetes). Cumulative hazard plots grouped by exposure categories suggested no appreciable violations of the proportional hazards assumption.
To compare the risk predictive capacity of measured CRF and estimated CRF, we constructed ROC curves with corresponding areas under the curve (AUC). The AUC (also known as c-statistic) is a function of both the sensitivity and specificity of the model across all of its values, and it represents the ability of the score to discriminate future cases from non-cases (30). Risk predictive capacity of estimated CRF was also compared with its modifiable constituent components (BMI, WC, RHR and physical activity). We used existing methods (18) to develop a continuous clustered score. After z-score conversion of each variable [z = (value - mean) / standard deviation (SD)], the four z-scores were summed and the sum was divided by 4 to compile a score with units of SD. The continuous variable for physical activity (originally active or inactive) was calculated as MET-minutes per week (22). As it is protective, the z-score from physical activity was multiplied by -1. Smoking status was not included in this analysis as it is defined as a dichotomous variable (current smoker or not).
Finally, we calculated the net reclassification improvement (NRI) for all-cause mortality between estimated CRF and measured CRF. Based on subsequent observed cases, this index integrates proportions of appropriate and inappropriate reclassifications between two risk prediction models (30). NRI was calculated as (18,31):
NRI = [P(up|case) – P(down|case) + P(down|non-case) – P(up|non-case)] × 100
where P is the proportion of participants moving up or down in terms of predicted risk category. The NRI was statistically examined by an asymptotic test (31).
Data analyses were performed using SPSS version 20.0 (IBM Inc.) and all p values are 2-sided with an alpha level of 0.05.
RESULTS
Descriptive characteristics of the study population are presented in Table 1. Men and women with higher levels of estimated CRF had lower BMI, WC, RHR, total cholesterol and glucose concentrations (except in women), and systolic and diastolic blood pressures. Participants with higher levels of estimated CRF were more likely to be physically active, and less likely to be smokers, heavy drinkers (except in women), to have hypercholesterolemia, diabetes mellitus, hypertension or abnormal ECG.
Table 1.
Baseline Characteristics of Study Participants by Sex and Estimated CRF Level, Aerobics Center Longitudinal Study 1974–2002
| Men (n = 34,211) | Women (n = 9,145) | |||||
|---|---|---|---|---|---|---|
| Estimated CRF | Estimated CRF | |||||
| Characteristic | Lower (n = 11,402) |
Middle (n = 11,405) |
Upper (n = 11,404) |
Lower (n = 3,048) |
Middle (n = 3,049) |
Upper (n = 3,048) |
| Age (years) | 44.9 (10.0) | 44.5 (9.7) | 44.0 (9.5)† | 44.9 (11.0) | 44.5 (10.3) | 44.0 (10.7)* |
| Body mass index (kg/m2) | 30.2 (3.6) | 25.9 (1.7) | 23.5 (1.7)† | 27.1 (4.5) | 22.5 (2.0) | 20.8 (1.5)† |
| Waist circumference (cm) | 103.8 (9.4) | 92.4 (5.7) | 84.9 (6.3)† | 82.3 (11.2) | 71.4 (7.0) | 67.5 (5.3)† |
| Resting heart rate (bpm) | 66.6 (10.5) | 60.4 (8.8) | 53.6 (8.4)† | 71.0 (10.3) | 64.6 (7.8) | 56.7 (7.5)† |
| Treadmill time (min) | 15.1 (3.9) | 18.3 (3.9) | 22.0 (4.3)† | 11.2 (3.6) | 13.7 (3.8) | 16.5 (4.4)† |
| Measured CRF (METs) | 10.3 (1.8) | 11.8 (1.9) | 13.6 (2.3)† | 8.5 (1.7) | 9.7 (1.8) | 10.9 (2.1)† |
| Estimated CRF (METs) | 10.7 (1.2) | 12.4 (0.7) | 13.8 (0.8)† | 9.0 (0.9) | 10.1 (0.5) | 10.9 (0.6)† |
| Total cholesterol (mg/dL) | 215.6 (41.3) | 210.0 (43.2) | 199.4 (44.5)† | 205.7 (39.5) | 198.1 (45.5) | 192.9 (34.7)† |
| Glucose (mg/dL) | 104.2 (22.1) | 99.4 (15.2) | 97.7 (86.9)† | 96.5 (15.4) | 96.3 (167.5) | 91.9 (11.0) |
| Blood pressure (mmHg) | ||||||
| Systolic | 125.2 (13.4) | 120.3 (12.7) | 118.0 (12.9)† | 117.7 (14.9) | 112.1 (14.1) | 109.6 (13.8)† |
| Diastolic | 84.7 (9.6) | 80.8 (9.0) | 78.2 (8.8)† | 79.2 (9.6) | 75.7 (9.1) | 73.9 (9.1)† |
| Physically Active | 645 (5.7) | 1,505 (13.2) | 5,157 (45.2)† | 143 (4.7) | 314 (10.3) | 1,224 (40.2)† |
| Current smokers | 2,766 (24.3) | 1,983 (17.4) | 915 (8.0)† | 338 (11.1) | 301 (9.9) | 148 (4.9)† |
| Heavy drinkers | 986 (8.6) | 981 (8.6) | 797 (7.0)† | 297 (9.7) | 334 (11.0) | 361 (11.8)* |
| Hypercholesterolemia | 4,044 (35.5) | 3,238 (28.4) | 2,281 (20.0)† | 888 (29.1) | 616 (20.2) | 468 (15.4)† |
| Diabetes mellitus | 866 (7.6) | 404 (3.5) | 249 (2.2)† | 169 (5.5) | 101 (3.3) | 85 (2.8)† |
| Hypertension | 5,168 (45.3) | 3,120 (27.4) | 2,105 (18.5)† | 799 (26.2) | 440 (14.4) | 342 (11.2)† |
| Abnormal ECG | 957 (8.4) | 804 (7.0) | 714 (6.3)† | 270 (8.9) | 216 (7.1) | 158 (5.2)† |
| Parental history of CVD | 3,152 (27.6) | 3,144 (27.6) | 3,133 (27.5) | 824 (27.0) | 842 (27.6) | 800 (26.2) |
Values are means (SD).
CRF, cardiorespiratory fitness; MET, metabolic equivalent.
Analysis of the variance (ANOVA) was used to test differences between groups in each sex: * p<0.01;
p<0.001.
The median (25–75th percentiles) follow-up period for mortality was 14.5 (5.7–20.1) years. A total of 1,934 participants died, 627 from CVD. For non-fatal CVD incidence, follow-up period was 7.7 (2.9–16.5) years and 1,049 cases were registered. Tables 2 and 3 show HRs and 95% CIs by CRF levels. In men, both measured CRF and estimated CRF were inversely associated with risk of all-cause mortality (HR per 1-MET increase, measured CRF: 0.87, 0.85–0.89; estimated CRF: 0.85, 0.82–0.88), CVD mortality (measured CRF: 0.83, 0.79–0.86; estimated CRF: 0.81, 0.77–0.86), and non-fatal CVD incidence (measured CRF: 0.90, 0.87–0.93; estimated CRF: 0.89, 0.85–0.93), after adjustments for potential confounders (table 2). In women, both measured CRF and estimated CRF were inversely associated with risk of all-cause mortality (measured CRF: 0.91, 0.85–0.99; estimated CRF: 0.87, 0.75–0.99) and non-fatal CVD incidence (measured CRF: 0.77, 0.67–0.90; estimated CRF: 0.76, 0.58–0.99) (table 3). Excluding hypercholesterolemia, hypertension and diabetes as confounders slightly strengthened some of these results (data not shown).
Table 2.
Hazard Ratios for Disease-specific Mortality and Non-fatal CVD According to CRF Levels in Men
| Measured CRF | Estimated CRF | |||||
|---|---|---|---|---|---|---|
| cases / n | Hazard Ratios (95% CI) | cases / n | Hazard Ratios (95% CI) | |||
| Model 1* | Model 2† | Model 1* | Model 2† | |||
| All-cause mortality | ||||||
| Lower | 793/11329 | 1 (ref) | 1 (ref) | 642/11402 | 1 (ref) | 1 (ref) |
| Middle | 524/12164 | 0.64 (0.57–0.71) | 0.68 (0.61–0.76) | 557/11405 | 0.74 (0.66–0.82) | 0.79 (0.70–0.88) |
| Upper | 401/10718 | 0.49 (0.44–0.55) | 0.56 (0.49–0.63) | 519/11404 | 0.59 (0.53–0.67) | 0.67 (0.59–0.75) |
| P for linear trend | < 0.001 | < 0.001 | < 0.001 | < 0.001 | ||
| Per 1-MET | 0.84 (0.83–0.86) | 0.87 (0.85–0.89) | 0.82 (0.79–0.84) | 0.85 (0.82–0.88) | ||
| CVD mortality | ||||||
| Lower | 304/11329 | 1 (ref) | 1 (ref) | 249/11402 | 1 (ref) | 1 (ref) |
| Middle | 164/12164 | 0.52 (0.43–0.63) | 0.60 (0.49–0.72) | 170/11405 | 0.59 (0.48–0.71) | 0.66 (0.54–0.80) |
| Upper | 109/10718 | 0.34 (0.27–0.43) | 0.45 (0.36–0.57) | 158/11404 | 0.47 (0.38–0.57) | 0.59 (0.48–0.73) |
| P for linear trend | < 0.001 | < 0.001 | < 0.001 | < 0.001 | ||
| Per 1-MET | 0.78 (0.75–0.81) | 0.83 (0.79–0.86) | 0.76 (0.72–0.80) | 0.81 (0.77–0.86) | ||
| Non-fatal CVD | ||||||
| Lower | 353/4335 | 1 (ref) | 1 (ref) | 309/4048 | 1 (ref) | 1 (ref) |
| Middle | 346/5140 | 0.72 (0.62–0.83) | 0.80 (0.69–0.93) | 327/4902 | 0.76 (0.65–0.89) | 0.84 (0.71–0.98) |
| Upper | 278/4985 | 0.47 (0.40–0.55) | 0.57 (0.48–0.67) | 341/5510 | 0.56 (0.48–0.65) | 0.66 (0.56–0.78) |
| P for linear trend | < 0.001 | < 0.001 | < 0.001 | < 0.001 | ||
| Per 1-MET | 0.87 (0.84–0.89) | 0.90 (0.87–0.93) | 0.84 (0.80–0.88) | 0.89 (0.85–0.93) | ||
CI, confidence interval; CRF, cardiorespiratory fitness; CVD, cardiovascular disease; MET, metabolic equivalent.
Adjusted for age and examination year.
Adjusted for age, examination year, alcohol intake (heavy drinker or not), hypercholesterolemia, hypertension, diabetes, abnormal resting or exercise ECG, and parental history of CVD (present or not for each).
Table 3.
Hazard Ratios for Disease-specific Mortality and Non-fatal CVD According to CRF Levels in Women
| Measured CRF | Estimated CRF | |||||
|---|---|---|---|---|---|---|
| cases / n | Hazard Ratios (95% CI) | cases / n | Hazard Ratios (95% CI) | |||
| Model 1* | Model 2† | Model 1* | Model 2† | |||
| All-cause mortality | ||||||
| Lower | 109/3143 | 1 (ref) | 1 (ref) | 74/3048 | 1 (ref) | 1 (ref) |
| Middle | 66/3076 | 0.92 (0.68–1.25) | 0.91 (0.67–1.24) | 77/3049 | 0.83 (0.60–1.14) | 0.82 (0.59–1.13) |
| Upper | 41/2926 | 0.75 (0.52–1.07) | 0.74 (0.52–1.07) | 65/3048 | 0.68 (0.49–0.95) | 0.67 (0.48–0.94) |
| P for linear trend | 0.281 | 0.275 | 0.079 | 0.068 | ||
| Per 1-MET | 0.92 (0.85–0.99) | 0.91 (0.85–0.99) | 0.87 (0.75–0.99) | 0.87 (0.75–0.99) | ||
| CVD mortality | ||||||
| Lower | 25/3143 | 1 (ref) | 1 (ref) | 21/3048 | 1 (ref) | 1 (ref) |
| Middle | 16/3076 | 1.05 (0.56–1.97) | 1.08 (0.57–2.04) | 15/3049 | 0.63 (0.32–1.22) | 0.70 (0.35–1.37) |
| Upper | 9/2926 | 0.80 (0.37–1.72) | 0.89 (0.41–1.94) | 14/3048 | 0.58 (0.29–1.17) | 0.65 (0.32–1.31) |
| P for linear trend | 0.793 | 0.905 | 0.230 | 0.410 | ||
| Per 1-MET | 0.92 (0.78–1.09) | 0.95 (0.81–1.13) | 0.83 (0.63–1.08) | 0.84 (0.64–1.12) | ||
| Non-fatal CVD | ||||||
| Lower | 42/1206 | 1 (ref) | 1 (ref) | 27/1024 | 1 (ref) | 1 (ref) |
| Middle | 15/1175 | 0.35 (0.19–0.63) | 0.36 (0.20–0.66) | 27/1239 | 0.64 (0.38–1.11) | 0.73 (0.42–1.26) |
| Upper | 15/1254 | 0.35 (0.19–0.63) | 0.38 (0.21–0.69) | 18/1372 | 0.34 (0.19–0.63) | 0.38 (0.20–0.70) |
| P for linear trend | < 0.001 | < 0.001 | 0.003 | 0.008 | ||
| Per 1-MET | 0.75 (0.65–0.87) | 0.77 (0.67–0.90) | 0.69 (0.54–0.89) | 0.76 (0.58–0.99) | ||
CI, confidence interval; CRF, cardiorespiratory fitness; CVD, cardiovascular disease; MET, metabolic equivalent.
Adjusted for age and examination year.
Adjusted for age, examination year, alcohol intake (heavy drinker or not), hypercholesterolemia, hypertension, diabetes, abnormal resting or exercise ECG, and parental history of CVD (present or not for each).
Table 4 presents the discrimination statistics of measured CRF and estimated CRF. Cstatistic values (AUC) were slightly higher for measured CRF than estimated CRF. In both cases, the discriminative ability was always higher for CVD mortality than for any other outcome. In general, c-statistic values were higher in women than men. The lowest c-statistic value was 0.61 (estimated CRF discriminating non-fatal CVD incidence in men) and the highest was 0.74 (measured CRF discriminating CVD mortality in women). As it can be observed in Table 5, the discriminative ability of estimated CRF was higher to that from any of its modifiable components, separately or together, in men and women, and for all outcomes.
Table 4.
Discrimination Statistics of Measured CRF and Estimated CRF for Disease-specific Mortality and Non-fatal CVD
| Measured CRF | Estimated CRF | |||||
|---|---|---|---|---|---|---|
| All-cause mortality |
CVD mortality |
Non-fatal CVD |
All-cause mortality |
CVD mortality |
Non-fatal CVD |
|
| Men (n = 34,211) | ||||||
| Cases | 1,718 | 577 | 977* | 1,718 | 577 | 977* |
| Area under the curve | 0.67 | 0.72 | 0.62 | 0.63 | 0.68 | 0.61 |
| (95% CI)† | (0.66–0.69) | (0.70–0.74) | (0.60–0.64) | (0.61–0.64) | (0.66–0.70) | (0.59–0.62) |
| Sensitivity‡ | 0.46 | 0.53 | 0.36 | 0.37 | 0.43 | 0.32 |
| Specificity§ | 0.68 | 0.67 | 0.70 | 0.67 | 0.67 | 0.72 |
| Women (n = 9,145) | ||||||
| Cases | 216 | 50 | 72* | 216 | 50 | 72* |
| Area under the curve | 0.70 | 0.74 | 0.71 | 0.64 | 0.73 | 0.68 |
| (95% CI)† | (0.66–0.73) | (0.68–0.80) | (0.64–0.77) | (0.60–0.68) | (0.66–0.81) | (0.62–0.74) |
| Sensitivity‡ | 0.50 | 0.50 | 0.58 | 0.34 | 0.42 | 0.38 |
| Specificity§ | 0.66 | 0.66 | 0.67 | 0.67 | 0.67 | 0.72 |
Subsample of 14,460 men and 3,635 women.
Calculated from inverted CRF (as it is protective).
The proportion of cases captured by lower CRF group (highest risk).
The proportion of non-cases captured by combined middle and upper CRF groups.
Table 5.
Area under the curve (AUC) for the modifiable constituent components of the non-exercise CRF algorithms (separately and clustered) and for estimated CRF
| AUC (95% CI) | ||||||
|---|---|---|---|---|---|---|
| Men (n = 34,211) | Women (n = 9,145) | |||||
| All-cause mortality |
CVD mortality | Non-fatal CVD* |
All-cause mortality |
CVD mortality | Non-fatal CVD* |
|
| Body mass index | 0.49 (0.48–0.51) | 0.53 (0.51–0.55) | 0.53 (0.51–0.55) | 0.50 (0.47–0.54) | 0.51 (0.43–0.58) | 0.56 (0.50–0.63) |
| Waist circumference | 0.53 (0.52–0.55) | 0.58 (0.55–0.60) | 0.56 (0.55–0.58) | 0.53 (0.49–0.57) | 0.56 (0.49–0.63) | 0.59 (0.52–0.65) |
| Resting heart rate | 0.55 (0.53–0.56) | 0.56 (0.54–0.59) | 0.52 (0.50–0.54) | 0.52 (0.48–0.56) | 0.62 (0.55–0.69) | 0.56 (0.49–0.62) |
| Physical activity† | 0.58 (0.57–0.59) | 0.57 (0.55–0.60) | 0.54 (0.52–0.55) | 0.61 (0.58–0.65) | 0.61 (0.54–0.69) | 0.59 (0.53–0.65) |
| Clustered score | 0.55 (0.54–0.57) | 0.58 (0.56–0.61) | 0.56 (0.54–0.57) | 0.56 (0.53–0.60) | 0.61 (0.54–0.69) | 0.62 (0.56–0.68) |
| Estimated CRF† | 0.63 (0.61–0.64) | 0.68 (0.66–0.70) | 0.61 (0.59–0.62) | 0.64 (0.60–0.68) | 0.73 (0.66–0.81) | 0.68 (0.62–0.74) |
Subsample of 14,460 men and 3,635 women.
Calculated from inverted physical activity and CRF (as they are protective).
Finally, Table 6 shows the reclassification statistics for all-cause mortality between both CRF methods. Compared to estimated CRF, measured CRF reclassified correctly 12.7% of men and 20.8% of women who died (i.e. they were reclassified to a higher risk category). The overall NRI was 12.3% for men (p<0.05) and 19.8% for women (p<0.001).
Table 6.
Reclassification of the Predicted Risk of All-cause Mortality in Men and Women on the basis of Estimated CRF vs. Measured CRF
| Estimated CRF | Measured CRF | Reclassified as higher risk |
Reclassified as lower risk |
Net
correctly reclassified (%) |
||
|---|---|---|---|---|---|---|
| Lower (highest risk) |
Middle | Upper | ||||
| Men | ||||||
| Cases (n=1,718) | ||||||
| Lower (highest risk) | 487 | 132 | 23 | 468 | 250 | 12.7 |
| Middle | 232 | 230 | 95 | |||
| Upper | 74 | 162 | 283 | |||
| Non-cases (n=32,493) | ||||||
| Lower (highest risk) | 6,758 | 3,406 | 596 | 6,886 | 6,760 | −0.4 |
| Middle | 2,964 | 5,126 | 2,758 | |||
| Upper | 814 | 3,108 | 6,963 | |||
| Net Reclassification improvement (NRI) | 12.3 % (p<0.05) | |||||
| Women | ||||||
| Cases (n=216) | ||||||
| Lower (highest risk) | 54 | 16 | 4 | 81 | 36 | 20.8 |
| Middle | 37 | 24 | 16 | |||
| Upper | 18 | 26 | 21 | |||
| Non-cases (n=8,929) | ||||||
| Lower (highest risk) | 1,794 | 874 | 306 | 2,099 | 2,007 | −1.0 |
| Middle | 868 | 1277 | 827 | |||
| Upper | 372 | 859 | 1,752 | |||
| Net Reclassification improvement (NRI) | 19.8 % (p<0.001) | |||||
CRF, cardiorespiratory fitness.
All the analyses (tables 2–6) were repeated using the algorithms with the five-level physical activity variable. The results were virtually the same, with a risk predictive capacity and discriminative ability very similar to that provided by the two-level physical activity algorithms (data not shown).
DISCUSSION
The purpose of the present study was to investigate the association of estimated CRF, based on longitudinal algorithms, with disease-specific mortality and non-fatal CVD events in middle-aged men and women. Previous studies had shown that these (17) and other non-exercise equations (10–13,15) estimate CRF with reasonable accuracy at the population level. Now we show that estimated CRF, calculated from typically available health indicators, significantly predicted future risk of non-fatal CVD as well as all-cause and CVD mortality, after adjustment for standard risk factors. However, these algorithms can still be refined for a better risk prediction performance, as measured CRF presented a better discriminative capacity (c-statistic) and reclassified correctly a significant proportion of cases (NRI).
Clinicians have long been aware that patients capable of high levels of physical exertion have a better prognosis than those with limited exercise capacity. Data from the ACLS and other epidemiologic studies indicate that individuals with low CRF are much more likely to develop hypertension (32), diabetes (32,33), and metabolic syndrome (32,34) and to have higher rates of death due to CVD (3,35), cancer (36), and all causes (1,3,37). Many experts have recommended CRF testing in asymptomatic and symptomatic men and women of all ages (38), and the American Heart Association recently highlighted the need for a national CRF registry (39). The present algorithms are a practical alternative for an estimate of CRF and a useful tool for identifying persons at risk. The potential clinical implications are substantial, as this method could be applied to electronic medical record systems and easily determined on patients in clinical practice or health maintenance plans.
The risk reduction per 1-MET increase observed in our study is consistent with previous findings using measured CRF (9), and also with the only study so far investigating estimated CRF and mortality (18). Similar to our results, these studies have suggested a 10–20% risk reduction per 1-MET increase, with the effect being slightly higher for CVD events than for allcause mortality (9,18). Between genders, our findings indicate a similar protective trend in men and women, although in females there was no association with CVD mortality and the linear trend for all-cause mortality did not reach statistical significance. The smaller number of cases in women likely decreased the statistical power as suggested in previous ACLS studies (26). In the Lipid Research Clinics study (40), CRF predicted CVD mortality risk in women and men, whereas in the Framingham Heart Study, CRF was significantly associated with coronary heart disease events in men but not in women (41). The discriminative capacity (c-statistic) of our models for CVD mortality (from 0.68 to 0.74) was close to that observed for the Framingham risk score and other similar prediction models (0.75–0.80), based on the combination of multiple independent risk markers (30). The c-statistic values reported by Stamatakis et al. (18) for CVD deaths were also comparable (0.73–75).
The potential mechanisms for the protective role of estimated CRF could be the same as those attributed to measured CRF, achieved in most cases through healthy lifestyle habits: lower levels of adiposity, blood pressure and chronic inflammation, higher insulin sensitivity and glycemic control, more favorable lipid profile, enhanced endothelial function, improved cardiac autonomic regulation, and preserved functional capacity and cognitive ability during aging, among others (22). In our study, those participants with higher levels of estimated CRF had lower BMI, WC, RHR, cholesterol and glucose concentrations, systolic and diastolic blood pressures, and were more likely to be physically active and less likely to be smokers, to have hypercholesterolemia, diabetes mellitus or hypertension.
Notable features make this a robust set of findings: the possibility to compare risk predictive capacity between measured CRF and estimated CRF; a large set of major outcomes studied (all-cause mortality, CVD mortality and non-fatal CVD incidence); the large sample size and long follow-up; and the broad range of potential confounders taken into account. Some limitations must also be mentioned and addressed in future research. Measured CRF was not directly assessed by gas analysis, but indirectly calculated from treadmill speed and grade. Although CPX is considered the gold standard, both assessed and calculated METs from exercise tests have shown to be among the strongest predictors of adverse events in a prognostic model (9). The lack of information on diet and medication use/adherence may have introduced some residual confounding. The most heavily weighted variable (physical activity) was self-reported, although any objective alternative would be less practical. And finally, participants were mostly non-Hispanic whites, well-educated and with professional positions, so we do not know how well our algorithms would predict health risk in other populations. Cross-validation studies are needed to investigate the generalizability of these results by testing the algorithms’ predictive capacity in other cohorts.
In conclusion, our longitudinal algorithms utilize information routinely collected to obtain an estimate of CRF that provides a valid indication of health status. Although the method can still be refined, estimated CRF significantly predicted risk of non-fatal CVD and all-cause and disease-specific mortality. In addition to identifying people at risk, this method can be potentially utilized to provide more appropriate exercise recommendations that reflect initial CRF levels.
ACKNOWLEDGMENTS
The authors thank the Cooper Clinic physicians and technicians for collecting the data, and staff at the Cooper Institute for data entry and data management.
This work was supported by National Institutes of Health grants (AG06945, HL62508, R21DK088195) and in part by unrestricted research grants from The Coca-Cola Company and Spanish Ministry of Education (EX-2010-1008).
ABBREVIATIONS LIST
- ACLS
Aerobics Center Longitudinal Study.
- BMI
body mass index.
- CRF
cardiorespiratory fitness.
- CVD
cardiovascular disease.
- MET
metabolic equivalent.
- NRI
net reclassification improvement.
- RHR
resting heart rate.
- VO2max
maximal oxygen uptake.
- WC
waist circumference.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
The authors declare no relationship with industry that might pose a conflict of interest.
REFERENCES
- 1.Blair SN, Kohl HW, 3rd, 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;262:2395–2401. doi: 10.1001/jama.262.17.2395. [DOI] [PubMed] [Google Scholar]
- 2.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. European heart journal. 2004;25:1428–1437. doi: 10.1016/j.ehj.2004.06.013. [DOI] [PubMed] [Google Scholar]
- 3.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. doi: 10.1056/NEJMoa011858. [DOI] [PubMed] [Google Scholar]
- 4.Gulati M, Pandey DK, Arnsdorf MF, et al. Exercise capacity and the risk of death in women: the St James Women Take Heart Project. Circulation. 2003;108:1554–1559. doi: 10.1161/01.CIR.0000091080.57509.E9. [DOI] [PubMed] [Google Scholar]
- 5.Balady GJ, Arena R, Sietsema K, et al. Clinician's Guide to cardiopulmonary exercise testing in adults: a scientific statement from the American Heart Association. Circulation. 2010;122:191–225. doi: 10.1161/CIR.0b013e3181e52e69. [DOI] [PubMed] [Google Scholar]
- 6.Pollock ML, Bohannon RL, Cooper KH, et al. A comparative analysis of four protocols for maximal treadmill stress testing. Am Heart J. 1976;92:39–46. doi: 10.1016/s0002-8703(76)80401-2. [DOI] [PubMed] [Google Scholar]
- 7.Pollock ML, Foster C, Schmidt D, Hellman C, Linnerud AC, Ward A. Comparative analysis of physiologic responses to three different maximal graded exercise test protocols in healthy women. Am Heart J. 1982;103:363–373. doi: 10.1016/0002-8703(82)90275-7. [DOI] [PubMed] [Google Scholar]
- 8.Byrne NM, Hills AP, Hunter GR, Weinsier RL, Schutz Y. Metabolic equivalent: one size does not fit all. J Appl Physiol (1985) 2005;99:1112–1119. doi: 10.1152/japplphysiol.00023.2004. [DOI] [PubMed] [Google Scholar]
- 9.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;301:2024–2035. doi: 10.1001/jama.2009.681. [DOI] [PubMed] [Google Scholar]
- 10.Jackson AS, Blair SN, Mahar MT, Wier LT, Ross RM, Stuteville JE. Prediction of functional aerobic capacity without exercise testing. Medicine and science in sports and exercise. 1990;22:863–870. doi: 10.1249/00005768-199012000-00021. [DOI] [PubMed] [Google Scholar]
- 11.Heil DP, Freedson PS, Ahlquist LE, Price J, Rippe JM. Nonexercise regression models to estimate peak oxygen consumption. Medicine and science in sports and exercise. 1995;27:599–606. [PubMed] [Google Scholar]
- 12.Matthews CE, Heil DP, Freedson PS, Pastides H. Classification of cardiorespiratory fitness without exercise testing. Medicine and science in sports and exercise. 1999;31:486–493. doi: 10.1097/00005768-199903000-00019. [DOI] [PubMed] [Google Scholar]
- 13.Jurca R, Jackson AS, LaMonte MJ, et al. Assessing cardiorespiratory fitness without performing exercise testing. American journal of preventive medicine. 2005;29:185–193. doi: 10.1016/j.amepre.2005.06.004. [DOI] [PubMed] [Google Scholar]
- 14.Mailey EL, White SM, Wojcicki TR, Szabo AN, Kramer AF, McAuley E. Construct validation of a non-exercise measure of cardiorespiratory fitness in older adults. BMC Public Health. 2010;10:59. doi: 10.1186/1471-2458-10-59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Nes BM, Janszky I, Vatten LJ, Nilsen TI, Aspenes ST, Wisloff U. Estimating V.O 2peak from a nonexercise prediction model: the HUNT Study, Norway. Medicine and science in sports and exercise. 2011;43:2024–2030. doi: 10.1249/MSS.0b013e31821d3f6f. [DOI] [PubMed] [Google Scholar]
- 16.Jackson AS, Sui X, Hebert JR, Church TS, Blair SN. Role of lifestyle and aging on the longitudinal change in cardiorespiratory fitness. Arch Intern Med. 2009;169:1781–1787. doi: 10.1001/archinternmed.2009.312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Jackson AS, Sui X, O'Connor DP, et al. Longitudinal cardiorespiratory fitness algorithms for clinical settings. American journal of preventive medicine. 2012;43:512–519. doi: 10.1016/j.amepre.2012.06.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Stamatakis E, Hamer M, O'Donovan G, Batty GD, Kivimaki M. A non-exercise testing method for estimating cardiorespiratory fitness: associations with all-cause and cardiovascular mortality in a pooled analysis of eight population-based cohorts. European heart journal. 2013;34:750–758. doi: 10.1093/eurheartj/ehs097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Maslow AL, Sui X, Colabianchi N, Hussey J, Blair SN. Muscular strength and incident hypertension in normotensive and prehypertensive men. Medicine and science in sports and exercise. 2010;42:288–295. doi: 10.1249/MSS.0b013e3181b2f0a4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Myers GL, Cooper GR, Winn CL, Smith SJ. The Centers for Disease Control-National Heart, Lung and Blood Institute Lipid Standardization Program. An approach to accurate and precise lipid measurements. Clin Lab Med. 1989;9:105–135. [PubMed] [Google Scholar]
- 21.Blair SN, Kannel WB, Kohl HW, Goodyear N, Wilson PW. Surrogate measures of physical activity and physical fitness. Evidence for sedentary traits of resting tachycardia, obesity, and low vital capacity. American journal of epidemiology. 1989;129:1145–1156. doi: 10.1093/oxfordjournals.aje.a115236. [DOI] [PubMed] [Google Scholar]
- 22.Physical Activity Guidelines Advisory Committee Report. Washington, DC: U.S. Department of Health and Human Services; 2008. Physical Activity Guidelines Advisory Committee. [DOI] [PubMed] [Google Scholar]
- 23.Balke B, Ware RW. An experimental study of physical fitness of Air Force personnel. U S Armed Forces Med J. 1959;10:675–688. [PubMed] [Google Scholar]
- 24.ACSM's Guidelines for Exercise Testing and Prescription. 8th ed. Philadelphia: Lippincott Williams and Wilkins; 2009. American College of Sports Medicine. [DOI] [PubMed] [Google Scholar]
- 25.Anderson RN, Minino AM, Hoyert DL, Rosenberg HM. Comparability of cause of death between ICD-9 and ICD-10: preliminary estimates. Natl Vital Stat Rep. 2001;49:1–32. [PubMed] [Google Scholar]
- 26.Sui X, LaMonte MJ, Blair SN. Cardiorespiratory fitness as a predictor of nonfatal cardiovascular events in asymptomatic women and men. American journal of epidemiology. 2007;165:1413–1423. doi: 10.1093/aje/kwm031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Macera CA, Jackson KL, Davis DR, Kronenfeld JJ, Blair SN. Patterns of non-response to a mail survey. J Clin Epidemiol. 1990;43:1427–1430. doi: 10.1016/0895-4356(90)90112-3. [DOI] [PubMed] [Google Scholar]
- 28.Luepker RV, Apple FS, Christenson RH, et al. Case definitions for acute coronary heart disease in epidemiology and clinical research studies: a statement from the AHA Council on Epidemiology and Prevention; AHA Statistics Committee; World Heart Federation Council on Epidemiology and Prevention; the European Society of Cardiology Working Group on Epidemiology and Prevention; Centers for Disease Control and Prevention; and the National Heart, Lung, and Blood Institute. Circulation. 2003;108:2543–2549. doi: 10.1161/01.CIR.0000100560.46946.EA. [DOI] [PubMed] [Google Scholar]
- 29.National Institute on Alcohol Abuse and Alcoholism (NIAAA) Moderate & Binge Drinking. [Accessed October 29, 2013]; http://www.niaaa.nih.gov/alcohol-health/overview-alcohol-consumption/moderate-bingedrinking.
- 30.Lloyd-Jones DM. Cardiovascular risk prediction: basic concepts, current status, and future directions. Circulation. 2010;121:1768–1777. doi: 10.1161/CIRCULATIONAHA.109.849166. [DOI] [PubMed] [Google Scholar]
- 31.Pencina MJ, D'Agostino RB, Sr, D'Agostino RB, Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med. 2008;27:157–172. doi: 10.1002/sim.2929. discussion 207–12. [DOI] [PubMed] [Google Scholar]
- 32.Carnethon MR, Gidding SS, Nehgme R, Sidney S, Jacobs DR, Jr, Liu K. Cardiorespiratory fitness in young adulthood and the development of cardiovascular disease risk factors. JAMA. 2003;290:3092–3100. doi: 10.1001/jama.290.23.3092. [DOI] [PubMed] [Google Scholar]
- 33.Sui X, Hooker SP, Lee IM, et al. A prospective study of cardiorespiratory fitness and risk of type 2 diabetes in women. Diabetes care. 2008;31:550–555. doi: 10.2337/dc07-1870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.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;112:505–512. doi: 10.1161/CIRCULATIONAHA.104.503805. [DOI] [PubMed] [Google Scholar]
- 35.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;276:205–210. [PubMed] [Google Scholar]
- 36.Evenson KR, Stevens J, Cai J, Thomas R, Thomas O. The effect of cardiorespiratory fitness and obesity on cancer mortality in women and men. Medicine and science in sports and exercise. 2003;35:270–277. doi: 10.1249/01.MSS.0000053511.02356.72. [DOI] [PubMed] [Google Scholar]
- 37.Gulati M, Black HR, Shaw LJ, et al. The prognostic value of a nomogram for exercise capacity in women. N Engl J Med. 2005;353:468–475. doi: 10.1056/NEJMoa044154. [DOI] [PubMed] [Google Scholar]
- 38.Mark DB, Lauer MS. Exercise capacity: the prognostic variable that doesn't get enough respect. Circulation. 2003;108:1534–1536. doi: 10.1161/01.CIR.0000094408.38603.7E. [DOI] [PubMed] [Google Scholar]
- 39.Kaminsky LA, Arena R, Beckie TM, et al. The importance of cardiorespiratory fitness in the United States: the need for a national registry: a policy statement from the american heart association. Circulation. 2013;127:652–662. doi: 10.1161/CIR.0b013e31827ee100. [DOI] [PubMed] [Google Scholar]
- 40.Mora S, Redberg RF, Cui Y, et al. 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. doi: 10.1001/jama.290.12.1600. [DOI] [PubMed] [Google Scholar]
- 41.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. doi: 10.1161/01.CIR.0000143226.40607.71. [DOI] [PubMed] [Google Scholar]