Revisiting the BMI-based obesity paradox in mortality risk across the cardiovascular-kidney-metabolic syndrome continuum

The obesity paradox, generally defined by body mass index (BMI), refers to the observation that individuals with overweight or obesity have lower mortality risk in certain conditions. However, inherent observational confounding and reverse causation may bias these findings, and whether genetic susceptibility shows the same pattern remains unclear. To provide an epidemiological and genetic context for the obesity paradox in cardiovascular-kidney-metabolic (CKM) syndrome, two-sample Mendelian randomization (MR) analyses were conducted to investigate the potential causal effects of higher BMI on susceptibility to six CKM conditions, and Global Burden of Disease (GBD) 2021 data were used to quantify the proportions of deaths from these conditions attributable to high BMI. Genome-wide association studies (GWAS) further characterized the genetic associations with CKM phenotypes. To investigate the obesity paradox among CKM stages, associations of measured BMI with all-cause and cardiovascular disease (CVD) mortality were evaluated in UK Biobank (n = 488,656) and NHANES (n = 13,114) using Cox regression, restricted cubic splines (RCS), and two-piecewise Cox models. In UK Biobank, a BMI genetic risk score (BMI-GRS) was conducted to examine whether genetic susceptibility to higher BMI exhibited mortality association patterns comparable to those observed for measured BMI. Exploratory proteomic mediation analyses were subsequently conducted among participants with advanced CKM stages. Genetically predicted higher BMI was associated with greater susceptibility to all six conditions, and GBD analyses indicated a corresponding high-BMI-attributable mortality burden. CKM GWAS showed enrichment of lipid and lipoprotein metabolism pathways. In both cohorts, measured BMI showed broadly U-shaped associations with mortality, with the fitted curve nadir generally shifting toward higher BMI as CKM stage advanced. For all-cause mortality, the estimated breakpoint increased from 25.88 to 27.06 kg/m2 across stages 2–4 in UK Biobank and from 22.10 kg/m2 in stage 0 to approximately 30 kg/m2 in stages 2–4 in NHANES, with similar trends for CVD mortality. Conversely, BMI-GRS showed approximately linear adverse associations with all-cause mortality (HR per 1-SD increment 1.05, 95% CI 1.04–1.06) and CVD mortality (HR 1.10, 95% CI 1.07–1.12), and mortality risk increased progressively with advancing CKM stage across BMI-GRS strata (P for trend < 0.001). Exploratory proteomic mediation analyses highlighted FABP4, ADM, FSTL3, and LEP as candidate protein signals related to the BMI–mortality association in advanced CKM stages. The apparent obesity paradox in observational studies may partly reflect the limitations of BMI as a proxy for adiposity, rather than a ture protective effect of excess adiposity. What is currently known about this topic? Higher BMI raises cardiovascular, kidney and metabolic disease risk. Yet in advanced disease, overweight or obesity often predicts lower mortality—the obesity paradox. The CKM framework links adiposity, metabolic and kidney disease to CVD across a continuum. What is the key research question? Across the CKM continuum, does genetic susceptibility to higher BMI exhibit the same mortality pattern as that observed for measured BMI? What is new? Measured BMI shows broadly U-shaped mortality associations, with the nadir of the fitted curve increasing as CKM stage advances, whereas BMI-GRS shows approximately linear adverse mortality associations, with no survival advantage. How might this study influence clinical practice? Findings support using adiposity measures beyond BMI to assess prognostic risk.