Five-year cardiometabolic multimorbidity risk model for Chinese middle-aged and older adults
In 5,388 initially CMM-free Chinese middle-aged and older adults, a logistic regression model using nine predictors estimated five-year CMM risk with a validation AUC of 0.732. [1]
Automatic summary · not editorially reviewed
The key findings
The analysis included 5,388 initially CMM-free middle-aged and older adults from the 2015–2020 CHARLS surveys, split 70/30 into training and validation sets. [1]
Nine predictors were selected from 31 candidates, and the logistic regression model outperformed the XGBoost model on the validation set. [1]
The logistic regression model reached a validation AUC of 0.732 (95% CI: 0.703–0.761) for five-year CMM risk. [1]
The source at a glance
- Study type / format
- Prediction model development and validation using cohort survey data (abstract)
- Population / context
- 5,388 Chinese middle-aged and older adults initially free of cardiometabolic multimorbidity
- Summary based on
- Study abstract
- Original publication
- March 12, 2026
CHARLS cohort followed for five-year CMM onset
The study used data from the 2015 to 2020 China Health and Retirement Longitudinal Study (CHARLS) surveys. It included 5,388 middle-aged and older adults who were initially free of cardiometabolic multimorbidity (CMM). The outcome was the five-year risk of CMM onset. The dataset was randomly split into a training set (70%) and a validation set (30%).
Supporting source [1]
Nine predictors from 31 candidates; logistic model best
From 31 potential variables, LASSO regression with 10-fold cross-validation and correlation analysis identified nine key predictors: systolic blood pressure, BMI, fasting blood glucose, total cholesterol, triglycerides, uric acid, age, comorbidities and pain. Both an Extreme Gradient Boosting model and a logistic regression model were built. The logistic regression model performed better and more stably on the validation set, with an area under the curve of 0.732 (95% CI: 0.703–0.761). Calibration curves indicated reliable accuracy, and decision curve analysis showed clinical net benefit across a range of risk thresholds.
Supporting source [1]
Abstract-only scope of the CMM prediction model
The authors describe a reliable, clinically applicable logistic regression model with a nomogram for five-year CMM risk in Chinese middle-aged and older adults, intended to help identify high-risk individuals for early intervention. These are the authors' characterizations. Only the abstract was reviewed; the full text was not reviewed, so details such as predictor definitions, handling of missing data and the exact CMM definition could not be checked. The model is a risk-prediction tool, not evidence that intervention lowers CMM risk.
Supporting source [1]
How to interpret this
Abstract only; the full text was not reviewed. Predictor definitions, missing-data handling and the exact CMM definition are not reported in the abstract. The model predicts risk; it does not show that early intervention reduces CMM.
Based on the study abstract; the full paper, tables and supplementary data were not reviewed.
Source & supporting passages
Li H, Ma X, Cui T, He W, Zhu L, Zhang H. · europepmc.org · March 12, 2026