A body shape index outperforms body mass index in predicting all-cause mortality but not non-fatal cardiovascular outcomes
M Domislovic, J Kos, A Jelakovic, I Pecin, J Josipovic, V Domislovic, L Gellineo, I Brzic, M Abramovic Baric, V Matijevic, A Belancic, M Bilobrk, L Bilajac, K Capak, B JelakovicAbstract
Background/Introduction
Body mass index (BMI) is widely used to assess obesity-related risk but does not capture body fat distribution. A body shape index (ABSI), reflecting central adiposity, may better predict adverse outcomes.
Purpose
We compared the predictive value of BMI and ABSI for all-cause mortality and non-fatal outcomes.
Methods
This prospective cohort study, with a median follow-up period of 9.7 years (interquartile range 7.3–11.8), included 820 participants from the ENAH project. Participants were stratified into tertiles according to their ABSI z-scores. Cox regression models assessed associations of BMI z-score and ABSI z-score with all-cause mortality and a composite non-fatal cardiovascular, cerebrovascular, and renal outcome, adjusted for age, gender, smoking, hypertension, diabetes, and eGFR. Discrimination was evaluated using receiver operating characteristic (ROC) analysis. Non-fatal outcomes were cardiovascular (myocardial infarction, hearth failure, atrial fibrillation), cerebrovascular (CVI, TIA), renal (haemodialysis).
Results
Across increasing ABSI tertiles, participants were older (47 vs 63 years), more often male (20% vs 42%), had higher waist circumference (88 vs 102 cm), lower eGFR (86.9 vs 79.3 mL/min/1.73 m²), and higher prevalence of hypertension (52% vs 74%) and diabetes (9.6% vs 16.5%) (all p<0.05), while BMI differed only modestly (28.1 vs 27.3 kg/m², p=0.015). Outcome rates increased markedly across ABSI tertiles (31.0%, 35.4%, 50.9%; p<0.001). In adjusted Cox models, ABSI z-score was independently associated with all-cause mortality (HR 1.15 per SD, 95% CI 1.01–1.32; p=0.034), whereas BMI z-score was not (HR 0.95, 95% CI 0.79–1.14; p=0.595). In the combined model, ABSI remained significant (HR 1.15; p=0.041), while BMI did not (p=0.911). Neither ABSI nor BMI predicted non-fatal outcomes. Kaplan–Meier analysis showed a significant mortality gradient across ABSI tertiles (Log-rank p<0.001), but not across BMI tertiles (p=0.487). ROC analysis demonstrated superior distinction of all-cause mortality for ABSI compared with BMI (AUC 0.65 vs 0.51; ΔAUC 0.14; p<0.001), with no difference for non-fatal outcomes.
Conclusions
ABSI strongly outperforms BMI in predicting all-cause mortality, while neither index predicts non-fatal cardiovascular, cerebrovascular, or renal outcomes. Central adiposity appears to be a more relevant determinant of long-term survival than overall body mass.Picture