Emerging studies have investigated the association between different anthropometric indices with diabetes risk but the results were inconsistent. The aims of the study were to examine the associations of different anthropometric indices with incident diabetes risk and whether novel anthropometric indices improve diabetes prediction beyond traditional indices among elderly Chinese. Nine thousand nine hundred sixty-two elderly individuals (age ≥ 60 years old) derived from the prospective Dongfeng-Tongji cohort were included. Hazard ratio (HR) and corresponding 95% confidence interval (CI) were evaluated by Cox proportional hazard model to examine the associations between traditional anthropometric indices (body mass index [BMI], waist circumference [WC], waist-to-height ratio [WHtR]), novel anthropometric indices (visceral adiposity index [VAI], a body shape index [ABSI], body roundness index [BRI]) and diabetes risk. Receiver operating characteristic (ROC) curve and area under curve (AUC) were applied to compare the novel anthropometric indices with the traditional indices in diabetes prediction. During mean 4.6 years of follow-up, 614 incident cases of type 2 diabetes (T2D) were identified. Significant positive associations were detected between BMI, WC, WHtR, VAI and BRI and incident T2D risk. For ABSI, no significant association was observed in either men or women. BMI was the strongest predictor in diabetes in men (AUC = 0.655) comparable with the other anthropometric indices (P < 0.05). Similar as men, BMI was the strongest predictor (AUC = 0.635) in women. Except for WC, the AUC of BMI was larger than WHtR, VAI, and BRI. In contrast, ABSI was not a good predictor in either men (AUC = 0.507) or women (AUC = 0.503). In elderly Chinese, BMI, WC, WHtR, VAI and BRI were positively associated with incident T2D risk. Among them, BMI was the strongest predictor in both men and women.

Limited evidence exists about the relationship between serum carotenoid and mortality in metabolic syndrome (MetS) patients, and the effects of medication use on this association remains unclear. The study encompassed 2,521 MetS patients from the National Health and Nutrition Examination Survey (NHANES) 2001–2006 and 2017–2018. A total of 7 serum carotenoids were evaluated. Death data were sourced from the National Death Index, with causes assessed using ICD-10 codes. Bayesian kernel machine regression (BKMR) and random survival forest (RSF) were utilized to investigate serum carotenoid mixture on mortality and identify key carotenoids. “Qgcompint” R package was used to explore the modifying effects of medication use. The serum carotenoid levels at baseline ranged from 0.04 to 1.37 µmol/L. During a follow-up of 15.1 years, there were 696 deaths (27.61%), with 247 (35.49%) by cardiovascular disease (CVD), 148 (21.26%) by cancer, and 301 (43.25%) by other diseases. Individual and combined serum carotenoids were negatively associated with all-cause mortality (HR range:0.70–0.88, 95%CI range:0.56–0.99, all P < 0.05). α-carotene (VIMP = 0.223 in RSF) and lutein/zeaxanthin (PIP = 1.000 in BKMR) emerged as the greatest contributors to all-cause mortality. Lipid-lowering drugs attenuate the negative effect of serum carotenoids on MetS patients’ mortality (Pint = 0.014). The present study identified a protective effect of serum carotenoid on mortality in MetS patients, which was probably weakened by lipid-lowering drugs. Early dietary interventions for MetS patients taking lipid-lowering drugs, particularly those rich in carotenoids like α-carotene and lutein/zeaxanthin, could help reduce mortality.

This study aimed to examine the longitudinal associations of estimated pulse wave velocity (ePWV) and its longitudinal change with incident frailty in older adults. A total of 2,016 older adults (≥60 years, baseline non-frail) were included in baseline analysis, selected from an initial cohort of 4,129. Of these, 1,137 had repeated ePWV data for change analysis. Arterial stiffness was measured using ePWV, and its longitudinal change from baseline to follow-up was defined as ΔePWV. Frailty was assessed using a frailty index (FI), analyzed as both continuous and binary outcome (FI ≥ 0.25). Multivariable regression models evaluated associations of baseline ePWV and ΔePWV with incident frailty. After a mean follow-up of 7 years, 394 of 2,016 participants (19.5%) developed frailty. Each 1 m/s increase in baseline ePWV was associated with higher frailty risk (OR = 1.21, 95% CI: 1.10–1.33); the highest quartile had more than twice the risk of the lowest (OR = 2.11, 95% CI: 1.36–3.28). Among 1,137 participants with repeated measures, each 1 m/s increase in ΔePWV was associated with 21.4% higher frailty odds (OR = 1.21, 95% CI: 1.03–1.43), and rapid increase group had higher risk than decline/stable group (OR = 1.91, 95% CI: 1.13–3.23). Results were consistent in continuous FI and most sensitivity analyses. However, after excluding hypertension and cardiovascular disease from the FI, the associations were attenuated and no longer statistically significant. These findings indicate that higher baseline ePWV and increases in ePWV over follow-up were associated with an increased risk of incident frailty, although part of this association may be related to the cardiovascular components within the FI. ePWV may serve as a useful indicator for identifying individuals at elevated risk of frailty.

The association between organochlorine pesticides (OCPs)/synthetic pyrethroids (SPs) and thyroid cancer (TC) remains poorly understood, with metabolic mechanisms unexplored. We conducted a 1:1 age- and sex-matched case-control study (n = 668). Serum levels of 27 target analytes (19 OCPs and 8 SPs) were quantified; subsequent analyses were restricted to 13 compounds (10 OCPs and 3 SPs) with detection frequencies ≥85%. Eight machine learning (ML) algorithms with Shapley Additive Explanations (SHAP) were used to identify key pollutants in the 334 case-control pairs. Untargeted metabolomics was performed in a subset of 50 age- and sex-matched case-control pairs. Mixture effects were assessed by Bayesian kernel machine regression (BKMR) and weighted quantile sum (WQS) regression. Furthermore, the Latent Unknown Clustering Integrating Multi-Omics Data (LUCID) model was employed to integrate exposure and metabolic data, enabling the identification of TC patient subgroups and the exploration of underlying metabolic mechanisms. Participants (mean age 45.2 years, 82.3% female) had serum OCPs at 0.007-0.333 ng/mL and SPs at 0.046-0.095 ng/mL. ML algorithms identified fenpropathrin, β-BHC, cyhalothrin, α-BHC, and p,p'-DDD as the top five contributors to TC. Elevated OCPs/SPs exposure was significantly associated with increased TC risk (WQS: adjusted OR = 1.45, 95%CI = 1.34-2.24, P = 0.019; LUCID: OR = 9.33). Fenpropathrin was the primary contributor (BKMR posterior inclusion probability = 1.00; WQS weight = 67.6%). A total of 45 significant differential metabolites (DMs) were identified (VIP ≥1, P < 0.05, and qualitative level 1). LUCID revealed a distinct TC cluster characterized by upregulated S-sulfo-L-cysteine/adenosine and downregulated 2-hydroxycaprylic acid. OCPs/SPs mixtures, driven by fenpropathrin, disrupt amino acid/nucleotide metabolism while suppressing organic acid metabolism, representing a potential TC-associated metabolic signature.

Thyroid cancer (TC) is the most common endocrine malignancy (Dai et al. 2020; Xie et al. 2021). The incidence of TC has increased substantially over the past few decades (Pellegriti et al. 2013). I...