Obesity demonstrates a direct relation with cardiovascular risk and all-cause mortality, while cardiorespiratory fitness demonstrates an inverse relation. In clinical practice, several cardiometabolic (CM) risk factors are commonly measured to gauge cardiovascular risk, but the interaction between fitness and obesity with regard to CM risk has not been fully explored. In this study, 2,634 Brazilian adults referred for employer-sponsored heath exams were assessed. Obesity was defined as body mass index >30 kg/m 2 or waist circumference >102 cm in men or >88 cm in women when body mass index was 25 to 30 kg/m 2 . Fitness was quantified by stage achieved on an Ellestad treadmill stress test, with those completing stage 4 considered fit. Hepatic steatosis was determined by ultrasound. CM risk factors were compared after stratifying patients into 4 groups: fit and normal weight, fit and obese, unfit and normal weight, and unfit and obese. Approximately 22% of patients were obese; 12% were unfit. Fitness and obesity were moderately correlated (ρ = 0.38 to 0.50). The sample included 6.5% unfit and normal-weight subjects and 16% fit and obese subjects. In overweight and obese patients, fitness was negatively associated with CM risk (p <0.01 for all values). In fit patients, increasing body mass index was positively associated with CM risk (p <0.01 for all values). In instances of discordance between fitness and obesity, obesity was the stronger determinant of CM risk. In conclusion, fitness and obesity are independently associated with CM risk. The effects of fitness and obesity are additive, but obesity is more strongly associated with CM risk when fitness and obesity are discordant. These findings underscore the need for weight loss in obese patients and suggest an unmeasured benefit of fitness.
Obesity has become one of the most pressing worldwide health issues. Obesity increases all-cause and cardiovascular mortality. According to the World Health Organization, the prevalence of obesity doubled from 1980 to 2008, and now roughly half a billion individuals worldwide are obese. The World Health Organization has estimated that 2.8 million patients die each year as a result of being overweight or obese. Many studies have shown that the mortality associated with obesity can be attenuated or even reversed with improved cardiorespiratory fitness. These investigations have demonstrated that those who are fit yet obese have lower cardiovascular mortality than those who are lean but unfit. However, it is not clear if the benefits of improved fitness can be fully appreciated on routinely measured cardiometabolic (CM) risk panels. We sought to investigate the interaction of fitness and obesity with regard to routinely measured traditional and nontraditional CM risk factors to explore if the previously observed mortality benefit of fitness could be accounted for in routine CM risk assessment.
Methods
A total of 2,634 asymptomatic Brazilian men and women, who were free of known heart disease, were evaluated during employer-sponsored clinical health examinations at the Preventive Medicine Center of Hospital Israelita Albert Einstein in São Paulo, Brazil, from November 2008 to July 2010. The examination consisted of a medical history questionnaire, laboratory evaluation of cardiovascular risk factors, abdominal ultrasound for determination of liver fat, and a symptom-limited Ellestad treadmill test. A total of 2,576 subjects (98%) had complete information on CM risk factors except for high-sensitivity C-reactive protein (hsCRP), and these made up the primary study population. A total of 88% of subjects had measurements of hsCRP, and these were retained in hsCRP-specific analyses.
All participants had anthropomorphic measurements of obesity, including height, weight, and waist circumference. Obesity was defined using body mass index (BMI) and waist circumference to measure central adiposity, adapted from the National Cholesterol Education Program Adult Treatment Panel III ranges for metabolic syndrome. Obesity was considered present for BMI >30 kg/m 2 or waist circumference >102 cm in men or >88 cm in women when BMI was ≥25 kg/m 2 . Those with BMIs of 25 to 30 kg/m 2 who did not meet the waist circumference cutoff for obesity were considered overweight.
Fitness was quantified by peak stage achieved on a maximal symptom-limited Ellestad treadmill stress test, which ranged from 1.7 mph at a 10% grade (roughly 4.6 METs) to 8 mph at a 15% grade (roughly 21.5 METs). As such, fitness was considered an ordinal variable ranging from stage 1 through stage 7. Those who completed stage 4 (3 minutes at 5 mph at 10% grade or roughly 12.1 METs) were considered fit, while those who could not were considered unfit. Stage 4 was chosen as the fitness cutoff because it corresponds to the 125th percentile of predicted exercise capacity of our cohort given their average age. Because of the small sample sizes at the extremes, we combined stages 1 and 2 and stages 6 and 7 for subsequent analyses after the basic correlation analysis. METs were calculated according to the guidelines for indirect oxygen consumption according to the American College of Sports Medicine.
Blood specimens were collected after an overnight fast. Laboratory analysis included a standard lipid panel, fasting glucose, aspartate aminotransferase, alanine aminotransferase, and γ-glutamyl transpetidase, all of which were analyzed using a Vitros platform automated laboratory system (Johnson & Johnson Clinical Diagnostics, New Brunswick, New Jersey). Hs-CRP levels were determined by immunonephelometry (Dade-Behring GmbH, Mannheim, Germany). Hepatic steatosis was assessed by ultrasonography after ≥6-hour fasting using an Acuson XP-10 (Siemens Medical Solutions USA, Inc., Mountain View, California). The diagnosis was made by 2 board-certified radiologists, blinded to the results of the laboratory tests, as a pattern of bright liver with contrast between the hepatic and renal parenchyma. This method is widely used and has been previously validated. Metabolic syndrome was defined according to the International Diabetes Federation.
Age, gender, CM risk factors, and 10-year Framingham risk score for the overall study population were determined by computing frequencies for categorical variables, means with SDs for normally distributed continuous variables, and medians with interquartile ranges for skewed continuous variables. Pearson’s correlation coefficients were calculated to assess the relations of stage achieved on the Ellestad treadmill stress test with BMI, waist circumference, and obesity.
In the subset of participants who were either overweight or obese (all participants with BMIs >25 kg/m 2 ), the effect of increasing fitness level was determined by comparing CM risk factors across Ellestad stages using Pearson’s chi-square tests for categorical variables, 1-way analysis of variance for continuous variables with normal distributions, and Kruskal-Wallis tests for continuous variables with skewed distributions. In the subset of participants who were defined as fit (Ellestad stage ≥4), the effect of increasing BMI was determined by comparing CM risk factors across BMI quartiles using Pearson’s chi-square tests, 1-way analysis of variance, and Kruskal-Wallis tests.
To assess the effect of fitness on CM risk in the absence and presence of obesity, the total study population was divided into 4 groups: group 1, fit and not obese; group 2, unfit and not obese; group 3, fit and obese; and group 4, unfit and obese. CM risk factors were compared across these 4 study groups using Pearson’s chi-square tests, 1-way analysis of variance, and Kruskal-Wallis tests. CM risk in the unfit and not obese group was compared to that in the fit and obese group using the same tests. All analyses were performed using Stata version 12 (StataCorp LP, College Station, Texas).
Results
The baseline characteristics of the study population are listed in Table 1 . The study population had a mean age of 42.8 ± 8.7 years, with average BMIs of 23.5 ± 3.7 kg/m 2 for women and 26.9 ± 3.5 kg/m 2 for men. Obesity affected 11.6% of the women and 24.3% of the men. The average waist circumferences were 78.7 ± 9.5 cm for women and 94.8 ± 9.8 cm for men. The mean calculated 10-year Framingham risk score was 3.8 ± 4.3%.
| Characteristic | Value |
|---|---|
| Age (yrs) | 42.8 ± 8.7 |
| Women | 21.0% |
| BMI (kg/m 2 ) | 26.2 ± 3.8 |
| Obesity ∗ | 21.7% |
| Systolic blood pressure (mm Hg) | 116.8 ± 11.9 |
| Diastolic blood pressure (mm Hg) | 76.2 ± 7.8 |
| Hypertension | 12.4% |
| Fasting glucose (mg/dl) | 89.8 ± 10.2 |
| Diabetes mellitus | 1.2% |
| Low-density lipoprotein | |
| mg/dl | 131.1 ± 33.6 |
| mmol/L | 3.39 ± 0.87 |
| HDL | |
| mg/dl | 50.0 ± 13.5 |
| mmol/L | 1.29 ± 0.35 |
| Triglycerides | |
| mg/dl | 116 (83–165) |
| mmol/L | 1.31 (0.93–2.15) |
| Lipid-lowering drugs | 8.1% |
| Current smoker | 8.3% |
| Alcohol use † | 3 (2–6) |
| hsCRP (mg/L) (n = 2,576) | 1.2 (0.6–2.4) |
| γ-glutamyl transpeptidase (U/L) | 38.3 ± 31.5 |
| Hepatic steatosis | 35.1% |
| Metabolic syndrome | 17.3% |
| 10-year Framingham risk (%) | 3.8 ± 4.3 |
∗ BMI >30 kg/m 2 or waist circumference >102 cm in men or >88 cm in women when BMI was 25 to 30 kg/m 2 .
† Measured by the Alcohol Use Disorders Identification Test.
Stage achieved on the Ellestad treadmill stress test was inversely correlated with BMI, waist size, and obesity ( Table 2 ). There was a greater correlation between fitness stage and obesity in men than women. Of the 3 anthropometric measures, waist size was most strongly correlated with fitness stage in men and women.
| Correlation Coefficient | Men | Women |
|---|---|---|
| Exercise stage and BMI | −0.423 | −0.390 |
| Exercise stage and waist size | −0.503 | −0.401 |
| Exercise stage and obesity | −0.376 | −0.307 |
In a cohort of overweight and obese participants (all participants with BMIs >25 kg/m 2 ), increasing fitness levels demonstrated a progressively lower burden of CM risk factors ( Table 3 ). All measured risk factors showed progressive decreases across increasing fitness levels (p <0.001 for all). In fit subjects, ascending BMI quartiles demonstrated increasing levels of CM risk factors ( Table 4 ). All measured risk factors increased as BMI quartile increased (p <0.001 for all).
| Ellestad Stage Achieved † | Median hsCRP (mg/L) † | Median Triglyceride/HDL Ratio | Glucose (mg/dl) | Systolic Blood Pressure (mm Hg) | Metabolic Syndrome | Hepatic Steatosis | γ-Glutamyl Transpeptidase (IU/L) |
|---|---|---|---|---|---|---|---|
| 1 or 2 (n = 17) | 4.0 | 3.3 | 92.9 | 129.2 | 58.8% | 58.8% | 24.2 |
| 3 (n = 190) | 2.3 | 3.3 | 93.8 | 124.1 | 37.9% | 61.1% | 41.9 |
| 4 (n = 918) | 1.5 | 3.2 | 92.5 | 120.5 | 31.4% | 57.1% | 38.7 |
| 5 (n = 329) | 1.2 | 2.6 | 90.8 | 118.2 | 15.2% | 34.0% | 25.3 |
| 6 or 7 (n = 84) | 0.9 | 2.1 | 88.7 | 116.7 | 7.1% | 19.1% | 15.2 |
| p value | <0.001 | <0.001 | 0.001 | <0.001 | <0.001 | <0.001 | <0.001 |
∗ Includes all participants with BMIs >25 kg/m 2 regardless of waist circumference.
† Estimated METs (kcal/kg h) for stages 1 to 7, respectively, 4.6, 7.4, 9.6, 12.1, 16.4, 19.0, and 21.5.
| BMI Quartile (kg/m 2 ) | Median hsCRP (mg/L) ∗ | Median Triglyceride/HDL Ratio | Glucose (mg/dl) | Systolic Blood Pressure (mm Hg) | Metabolic Syndrome | Hepatic Steatosis | γ-Glutamyl Transpeptidase (IU/L) |
|---|---|---|---|---|---|---|---|
| First (15.8–23.6) (n = 586) | 0.8 | 1.6 | 85.2 | 109.4 | 1.2% | 6.5% | 26.5 |
| Second (23.7–25.8) (n = 610) | 0.9 | 2.2 | 89.6 | 115.9 | 6.4% | 24.4% | 34.9 |
| Third (25.9–28.3) (n = 571) | 1.3 | 2.8 | 90.8 | 118.1 | 18.6% | 39.4% | 43.5 |
| Fourth (28.4–44.6) (n = 505) | 1.8 | 3.4 | 93.8 | 123.0 | 41.2% | 70.5% | 49.7 |
| p value | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
