Obesity paradox in Japanese elderly patients with atrial fibrillation: A pooled analysis of two large-scale registries

Background Several studies on cardiovascular diseases have demonstrated the "obesity paradox", in which overweight and obese patients tend to have a better prognosis than underweight patients. However, only a few large-scale prognostic studies have investigated this paradox in patients with atrial fibrillation (AF). Purpose This study aims to investigate the presence of obesity paradox in patients with AF and to explore potential differences between elderly and non-elderly patients. Methods The association between body mass index (BMI) and prognosis was evaluated among Japanese AF patients (n = 7156) using two large-scale registries. Patients were categorized into four groups based on BMI: underweight (BMI < 18.5; n = 369), healthy (18.5 ≤ BMI < 25; n = 4370), overweight (25 ≤ BMI < 30; n = 1985), and obese (30 ≤ BMI; n = 410) groups. The primary endpoints were all-cause mortality and cardiovascular death. Results The mean age of the participants was 72 ± 10 years, with a mean follow-up of 1227 ± 440 days. The underweight group was older, had a higher proportion of females and showed a lower prevalence of hypertension, diabetes mellitus, dyslipidemia, and current smoking (all p < 0.05). Additionally, this group had lower levels of hemoglobin (Hb), creatinine clearance, triglycerides, and HbA1c but higher levels of HDL cholesterol and brain natriuretic peptide (BNP) (all p < 0.05). The obesity group had a higher CHADS2 score and lower HAS-BLED score (all p < 0.05). The use of anticoagulants did not differ among the groups. For all participants, all-cause mortality and cardiovascular death increased progressively from the obese to underweight groups, supporting the presence of the obesity paradox. Poor prognosis was associated with male sex, lower BMI, Hb and creatinine clearance levels, advanced age, and elevated BNP levels. The elderly patients (age ≥ 75 years; n = 3055) showed a similar trend in all-cause mortality and cardiovascular death, although no significant associations between BMI and the prognoses were observed in the non-elderly patients (age < 75 years: n = 4101) as shown in Figure 1. Multivariate Cox regression analysis revealed that lower BMI remained a significant risk factor for all-cause mortality and cardiovascular death among elderly patients (p < 0.001, both) but not among non-elderly patients. The association between BMI and the hazard ratio (crude) for all-cause mortality showed a decrease in the right shoulder drop from the underweight to the obese group in the elderly patients but a U-shaped in the non-elderly patients (Figure 2). Similar trends were observed in the hazard ratio for cardiovascular death. Conclusions This study confirms the presence of the obesity paradox in Japanese patients with AF, particularly among the elderly. These findings underscore the clinical importance of assessing underweight in the elderly and highlight the need for targeted interventions to improve outcomes in this population.