Impact of cardiorespiratory fitness on the obesity paradox in patients with heart failure

Abstract

Objective: To determine the impact of cardiorespiratory fitness (FIT) on survival in relation to the obesity paradox in patients with systolic heart failure (HF).

Patients and methods: We studied 2066 patients with systolic HF (body mass index [BMI] ≥18.5 kg/m(2)) between April 1, 1993 and May 11, 2011 (with 1784 [86%] tested after January 31, 2000) from a multicenter cardiopulmonary exercise testing database who were followed for up to 5 years (mean ± SD, 25.0±17.5 months) to determine the impact of FIT (peak oxygen consumption <14 vs ≥14 mL O2 ∙ kg(-1) ∙ min(-1)) on the obesity paradox.

Results: There were 212 deaths during follow-up (annual mortality, 4.5%). In patients with low FIT, annual mortality was 8.2% compared with 2.8% in those with high FIT (P<.001). After adjusting for age and sex, BMI was a significant predictor of survival in the low FIT subgroup when expressed as a continuous (P=.03) and dichotomous (<25.0 vs ≥25.0 kg/m(2)) (P=.01) variable. Continuous and dichotomous BMI expressions were not significant predictors of survival in the overall and high FIT groups after adjusting for age and sex. In patients with low FIT, progressively worse survival was noted with BMI of 30.0 or greater, 25.0 to 29.9, and 18.5 to 24.9 (log-rank, 11.7; P=.003), whereas there was no obesity paradox noted in those with high FIT (log-rank, 1.72; P=.42).

Conclusion: These results indicate that FIT modifies the relationship between BMI and survival. Thus, assessing the obesity paradox in systolic HF may be misleading unless FIT is considered.

FIGURE 1.

Kaplan-Meier analysis according to body mass index (BMI) in the overall group. Log-rank, 4.8; P=.09. Line A shows patients with a BMI of 30.0 or greater. Of 728 patients, 546, 166, and 72 were still alive and tracked at 1, 3, and 5 years, respectively; 63 died. Line B shows patients with a BMI of 25.0 to 29.9. Of 768 patients, 577, 181, and 89 were still alive and tracked at 1, 3, and 5 years, respectively; 81 died. Line C shows patients with a BMI of 18.5 to 24.9. Of 570 patients, 395, 128, and 51 were still alive and tracked at 1, 3, and 5 years, respectively; 68 died.

FIGURE 2.

Kaplan-Meier analysis according to body mass index (BMI) in the low cardiorespiratory fitness group (oxygen consumption <14 mL O₂ • kg⁻¹ • min⁻¹). Log-rank, 11.7; P=.003. Line A shows patients with a BMI of 30.0 or greater. Of 334 patients, 223, 64, and 17 were still alive and tracked at 1, 3, and 5 years, respectively; 41 died. Line B shows patients with a BMI of 25.0 to 29.9. Of 275 patients, 195, 63, and 27 were still alive and tracked at 1, 3, and 5 years, respectively; 45 died. Line C shows patients with a BMI of 18.5 to 24.9. Of 192 patients, 106, 34, and 7 were still alive and tracked at 1, 3, and 5 years, respectively; 42 died.

FIGURE 3.

Kaplan-Meier analysis according to body mass index (BMI) in the high cardiorespiratory fitness group (oxygen consumption ≥14 mL O₂ • kg⁻¹ • min⁻¹). Log-rank, 1.72; P=.42. Line A shows patients with a BMI of 30.0 or greater. Of 394 patients, 322, 101, and 54 were still alive and tracked at 1, 3, and 5 years, respectively; 22 died. Line B shows patients with a BMI of 25.0 to 29.9. Of 493 patients, 381, 117, and 60 were still alive and tracked at 1, 3, and 5 years, respectively; 36 died. Line C shows patients with a BMI of 18.5 to 24.9. Of 378 patients, 288, 93, and 43 were still alive and tracked at 1, 3, and 5 years, respectively; 26 died.