Objective: To analyze the characteristics of fine particulate matter (PM2.5) pollution in Urumqi City, Xinjiang Uygur Autonomous Region from 2016 to 2023 and establish a prediction model, so as to provide the reference for air pollution prevention and control. Methods: PM2.5 monitoring data of Urumqi City from 2016 to 2023 were collected through the website of Ministry of Ecology and Environment of China. The changing trend of PM2.5 concentration was analyzed using temporal chart and seasonal index. PM2.5 monthly average concentrations from 2016 to 2023 were used to establish an autoregressive integrated moving average (ARIMA) model, and the data in 2023 was fitted and compared with the actual values, using mean absolute percentage error (MAPE) to evaluate the effectiveness of the model, and PM2.5 monthly average concentration from 2024 to 2025 was predicted. Results: PM2.5 daily average concentration in Urumqi City showed a decreasing trend from 2016 to 2023 (rs=-0.239, P<0.001), with high seasonal indexes in January, February and December, indicating certain seasonal characteristics. The optional model was ARIMA (1, 0, 0) (1, 1, 0)12, with the value of Akaike information criterion, corrected Akaike information criterion, and Bayesian information criterion being 727.38, 727.88 and 737.10, respectively. PM2.5 monthly average concentration in 2023 was fitted and compared with the actual values, with an absolute error range of 0.31-7.45 μg/m3, a relative error range of 0.01-0.53, and MAPE of 14.42%. PM2.5 monthly average concentration in Urumqi City from 2024 to 2025 was predicted to be consistent with the trend from 2016 to 2023. Conclusions PM2.5 concentration in Urumqi City showed a tendency towards a decline from 2016 to 2023, and was relatively high in winter. ARIMA (1, 0, 0) (1, 1, 0)12 can be used for short-term prediction of PM2.5 pollution in Urumqi City.

Objective: To construct a prediction model of non-alcoholic fatty liver disease (NAFLD) in middle-aged and elderly patients with type 2 diabetes mellitus (T2DM), so as to provide basis for early screening and prevention of T2DM complicated with NAFLD. Methods: Patients aged 45 years and above and diagnosed with T2DM in Karamay Hospital of People's Hospital of Xinjiang Uygur Autonomous Region in 2021 were collected as the study subjects. The data of general demographic characteristics and biochemical test results were collected. The patients were randomly divided into training group (n=3 241) and validation group (n=1 389) according to the ratio of 7∶3. LASSO regression and multivariable logistic regression model were used to select predictive factors. The nomograph model for prediction of NAFLD risk in T2DM patients was established. The predictive value of the model was evaluated using the receiver operating characteristic (ROC), adjusted curve and decision clinical analysis. Results: Totally 4 630 T2DM cases were included, including 1 279 cases (27.62%) complicated with NAFLD. LASSO regression and multivariable logistic regression analysis identified gender, age, diastolic blood pressure, body mass index, alanine transaminase, triglycerides, low density lipoprotein cholesterol and platelet count as risk prediction factors for NAFLD in T2DM patients. The area under the ROC curve was 0.823 (95%CI: 0.814-0.832) for the training group and 0.809 (95%CI: 0.799-0.818) for the validation group, and Hosmer-Lemeshow test showed a good fitting effect (P>0.05). Decision curve analysis showed higher net clinical benefit of using the predictive model to predict NAFLD risk when the risk threshold probability was 0.27 to 0.85. Conclusion The nomogram model established has a good predictive value for the risk of NAFLD in T2DM patients aged 45 years and above.