The TyG index and PLR are independently and jointly associated with mental health indicators in Chinese adults, and their combination identified by LASSO regression (AUC = 0.744) may enhance the ability to identify individuals at elevated risk for mental health problems.
The TyG index was positively associated with multiple SCL-90 mental health dimensions in Chinese adults.
The platelet-to-lymphocyte ratio (PLR) showed positive associations with mental health dimensions similar to the TyG index, but also extended to additional domains.
Individuals in the high-TyG/high-PLR combined group had significantly higher scores across nearly all SCL-90 dimensions compared to other groups.
Among six machine learning classifiers evaluated, LASSO regression demonstrated the best overall predictive performance for mental health problems.
PLR and the TyG index were identified as the strongest positive predictors of mental health problems in the LASSO regression model.
Traditional logistic regression and linear regression were used to evaluate independent associations of TyG and PLR with overall and domain-specific mental health.
This research suggests that two blood-based markers — the triglyceride-glucose (TyG) index, which reflects insulin resistance, and the platelet-to-lymphocyte ratio (PLR), which reflects inflammation — are each independently linked to poorer mental health in Chinese adults. The study analyzed data from 725 adults and used a widely recognized mental health questionnaire (SCL-90) to assess symptoms across multiple dimensions including depression, anxiety, hostility, and others. People with high levels of both markers simultaneously had significantly worse scores across nearly all mental health dimensions measured. The researchers also tested whether combining these two biomarkers could help predict who is at elevated risk for mental health problems, using both traditional statistical methods and six machine learning approaches. A method called LASSO regression performed best, achieving an AUC (a measure of predictive accuracy) of 0.744, and identified PLR and the TyG index as the strongest predictors. Machine learning models generally outperformed traditional logistic regression in distinguishing individuals with mental health problems from those without. This research suggests that routinely collected blood test results — specifically markers of metabolic and inflammatory status — may carry information relevant to mental health risk. The combination of the TyG index and PLR could potentially serve as a low-cost screening tool to help identify adults who may benefit from further mental health assessment, though the findings come from a single hospital in China and would need to be validated in broader and more diverse populations before clinical application.
Zhou J, Yin S, Xue S, Sun C, Zhao Y. (2026). Combined predictive value of TyG index and PLR for mental health in Chinese adults: A machine learning approach.. PloS one. https://doi.org/10.1371/journal.pone.0355937