Work

WHtR identifies higher ultrasound-detected fatty liver prevalence below the BMI overweight threshold but adds little beyond BMI and TyG

Year: 2026
Type: article
Study design: Observational Study
Abstract: Waist-to-height ratio (WHtR) may identify a central-adiposity phenotype with a high prevalence of fatty liver among adults whose body mass index (BMI) is below the Chinese overweight threshold (< 24 kg/m²). Whether this information improves individual-level discrimination beyond metabolic measures remains uncertain. We evaluated the phenotyping value of WHtR, its incremental contribution beyond BMI plus the triglyceride–glucose (TyG) index, and the performance of TyG-derived anthropometric composite indices. We retrospectively analysed 3469 adults who attended routine health examinations at Fusui County People’s Hospital. We first compared the prevalence of ultrasound-detected fatty liver across four BMI–WHtR phenotypes and then examined the joint BMI–WHtR association using a continuous generalized additive model. For the secondary classification analysis, unpenalized logistic models were evaluated with 20 repeats of outcome-stratified 10-fold cross-validation. All models included age, sex, systolic blood pressure and diastolic blood pressure; additional terms comprised BMI, waist circumference, WHtR, TyG, prespecified combinations or TyG-derived composite indices. Ultrasound-detected fatty liver was present in 839 participants (24.2%). Among the 1933 participants with BMI < 24 kg/m², those with WHtR ≥ 0.5 represented 11.0% of the BMI stratum but accounted for 31.4% of its cases of ultrasound-detected fatty liver. The prevalence of ultrasound-detected fatty liver was 23.0% in this group, compared with 6.2% among participants with both low BMI and low WHtR. After adjustment for age, sex and TyG, the prevalence ratio for the low-BMI/high-WHtR phenotype was 2.63 (95% CI, 1.90–3.63). Prevalence rose gradually across the joint BMI–WHtR distribution, with no abrupt change at BMI 24 kg/m² or WHtR 0.5. Adding TyG to BMI increased the cross-validated ROC AUC from 0.837 to 0.865. Adding WHtR to BMI plus TyG increased the ROC AUC to 0.867, a paired difference of 0.00163 (95% CI, − 0.00009 to 0.00337). The corresponding differences were − 0.00008 (95% CI, − 0.00424 to 0.00416) for precision–recall AUC and − 0.00053 (95% CI, − 0.00132 to 0.00023) for the Brier score. ​WHtR identified a high-prevalence phenotype of ultrasound-detected fatty liver among adults below the Chinese BMI overweight threshold. Its addition to BMI plus TyG, however, yielded little further discrimination, with no clear improvement in precision–recall performance or overall prediction error. TyG-derived composite indices showed no clear overall advantage over the simpler BMI-plus-TyG model. Waist and height measurements may support low-cost phenotypic stratification in routine county hospital health examinations, but these single-centre cross-sectional data do not support using WHtR alone as a screening threshold to guide liver ultrasound referrals.
Language: English
Subfield: Hepatology
Field: Medicine
Open Access status: gold
Estimated APC paid: $2,690
OpenAlex ID: W7229602856
FWCI:
Cites: 14
Cited by:
Related to: 10