Spatial Analysis and Risk Assessment Model Research of Arthritis Based on Risk Factors: China, 2011, 2013 and 2015

Arthritis is a public health issue that is of global concern. Arthritis is one of the chronic diseases with a high incidence of middle-aged and older adults. The patients have paid a heavy price for this and caused a substantial economic burden on society. In this study, we used spatial autocorrelation, spatial cluster analysis, multiple logistic regression, and random forest models to analyze the spatial distribution and possible risk factors for arthritis in elderly Chinese and assess arthritis risk. Global spatial autocorrelation analysis and significance test results show that Moran's I of arthritis spatial autocorrelation in 2011, 2013, and 2015 are statistically significant, so there is significant spatial autocorrelation three years. The results of local spatial autocorrelation and spatial clustering analysis show that the aggregation areas of arthritis patients are mainly in the southwest, northwest, and central China. Multivariate logistic regression analysis showed that gender, age, education level, Body Mass Index (BMI), Center for Epidemiologic Studies Depression Scale score (CES-D), altitude, region, weather temperature, hypertension, lung, liver, heart, stroke, digestive, and kidney disease were all arthritis affects factors (P <; 0.05). Compared with the multi-factor Logistic regression model, the random forest model better assesses performance and higher fit. The fitting accuracy is 82.2% in the random forest model, which is better than the multi-factor Logistic regression model (66.6%). According to the assessment risk map generated by the random forest model, Northeast, Southwest, Northwest, South, and Central are high-risk areas for arthritis. These results provide benchmark data for the control and prevention of arthritis diseases.