A large prospective cohort was assembled in 2021 and followed through 2025. The investigators enrolled 65,347 participants between January and December 2021 and conducted follow-up surveys between January and December 2025. The cohort contributed a mean follow-up time of 4.08 ± 0.27 years. By the end of follow-up, 4,029 individuals—15.8% of the cohort—had been diagnosed with thyroid nodules. These case counts form the basis for the logistic regression models and downstream analyses reported by the authors.
The exposure of interest was the triglyceride glucose-waist height ratio (TyG-WHtR), a composite metric combining the triglyceride–glucose index with an adiposity measure (waist-to-height ratio). The study team evaluated TyG-WHtR at baseline and as a cumulative metric over time. They categorized participants into quartiles for categorical comparisons and also modeled the metric per 1–standard-deviation (1-SD) increment to quantify continuous associations. The abstract did not provide formulaic details for the TyG-WHtR calculation nor the exact method for deriving the cumulative measure; those methodological specifics were not reported in the source abstract.
Investigators applied multivariable binary logistic regression to estimate the association between TyG-WHtR and incident thyroid nodules. In fully adjusted models, participants in the highest quartile of baseline TyG-WHtR had an odds ratio (OR) of 1.57 (95% CI: 1.22–2.02) compared with those in the lowest quartile. The cumulative TyG-WHtR showed a nearly identical effect size for highest versus lowest quartiles, with an OR of 1.57 (95% CI: 1.25–1.96). Modeling TyG-WHtR continuously produced smaller but statistically measurable effects: each 1-SD increment in baseline TyG-WHtR corresponded to an OR of 1.15 (95% CI: 1.05–1.26), and each 1-SD increment in cumulative TyG-WHtR corresponded to an OR of 1.17 (95% CI: 1.07–1.27). The authors therefore reported that both baseline and time-averaged TyG-WHtR were independently associated with higher odds of developing thyroid nodules.
Restricted cubic spline (RCS) analyses were used to assess potential nonlinear relationships between TyG-WHtR and thyroid nodule risk. The RCS models indicated linear dose–response relationships for baseline and cumulative TyG-WHtR; the P values for nonlinearity were greater than 0.05, which the authors interpreted as no evidence of departure from linearity. For discrimination, receiver operating characteristic (ROC) curves compared four metabolic metrics: baseline TyG, waist-to-height ratio (WHtR), baseline TyG-WHtR, and cumulative TyG-WHtR. The cumulative TyG-WHtR achieved the highest area under the curve (AUC = 0.539) for predicting thyroid nodule status in this dataset. The reported AUC indicates modest discrimination in this population, according to the numeric value provided in the abstract.
The study team performed subgroup and sensitivity analyses and reported that these secondary analyses collectively confirmed the robustness of the observed associations. The abstract did not list which subgroups were examined, nor did it detail the sensitivity analysis strategies or results by subgroup. Therefore, while the authors noted consistency across these checks, the exact strata and covariate adjustments included in those analyses were not specified in the abstract text.
The authors concluded that higher TyG-WHtR was independently associated with increased risk of incident thyroid nodules, and that cumulative TyG-WHtR showed a stronger association than baseline measurement. The abstract also advised that maintaining favorable TyG and waist circumference levels over time may reduce thyroid nodule risk in this Chinese population. The source abstract, however, did not report several methodological details that clinicians and researchers often seek: it did not specify the full list of covariates included in the multivariable models, the diagnostic criteria or imaging protocols used to define thyroid nodules, the exact formula for calculating TyG-WHtR, nor the approach used to derive the cumulative TyG-WHtR metric. The abstract likewise did not present subgroup definitions or the number of participants contributing to each subgroup. These omissions limit the ability to assess residual confounding, detection bias from variable imaging, or potential effect modification by age, sex, or comorbidities—issues that would need to be examined in the full manuscript.
Unanswered clinical questions remain. For example, the prognostic significance of TyG-WHtR-related nodules (benign versus malignant pathology) was not reported in the abstract; the authors did not provide data on nodule size, sonographic risk features, or subsequent interventions. Those outcome-level details were not included in the source abstract. Future readers should consult the full article for granular methods and subgroup results before translating these associations into clinical screening or risk-reduction strategies.