Predictive Model using AI may detect Hyperuricemia Risk Early among those on low dose Aspirin
Hyperuricemia is primarily attributed to increased serum uric acid (SUA) concentration, considered the primary gout factor. In the US, approximately 38 million adults, up to 16.9% of the population, are affected by hyperuricemia.
According to an original research article published in Frontiers in Pharmacology, researchers concluded that a predictive model established by XGBoost algorithms could help clinicians detect hyperuricemia risk early in people taking low-dose aspirin.
Hyperuricemia is a severe condition related to gout and cardiovascular diseases. Low-dose aspirin was reported to inhibit uric acid excretion, leading to hyperuricemia. To decrease hyperuricemia-related CVD, this study identified the risk of hyperuricemia in people taking aspirin.
The data for this study were collected from the NHANES between 2011 and 2018. The analysis included participants who answered the “Preventive Aspirin Use” questionnaire with positive answers. The study used six machine learning algorithms, and the eXtreme Gradient Boosting (XGBoost) model was selected to predict the risk of hyperuricemia.
Key findings from the study are:
- Out of 805 participants enrolled, 190 participants had hyperuricemia.
- The participants were divided into a training set and a testing set (ratio of 8:2).
- The area under the curve for the training and testing set was 0.864 and 0.811, respectively.
- The SHapley Additive exPlanations (SHAP) method evaluated the modelling performance.
- The feature ranking interpretation presented that eGFR, BMI, and waist circumference were the three most essential features for hyperuricemia in those taking aspirin.
- The factors correlated with the development of hyperuricemia are triglyceride, hypertension, total cholesterol, HDL, LDL, age, race, and smoking.
Study limitations include a cross-sectional study, small sample size, and the duration of people taking low-dose aspirin was not known.
They said we leveraged an ML model trained on NHANES data to develop a hyperuricemia model for those taking aspirin. XGBoost model can p help clinicians detect hyperuricemia risk early.
Reference
Zhu, B. et al. Prediction of hyperuricemia in people taking low-dose aspirin using a machine learning algorithm: a cross-sectional study of the National Health and Nutrition Examination Survey. Frontiers in Pharmacology, 14. https://doi.org/10.3389/fphar.2023.1276149
Next Story