|
|
|
| Analysis of whole blood element level in children with autism spectrum disorder and construction of nomogram |
| ZHANG Yushan1, YU Mengxue2, Asimuguli KELIMU1 |
| 1. College of Public Health, Xinjiang Medical University, Urumqi 830011, Xinjiang Uygur Autonomous Region, China；2. Maternal and Child Health Hospital of Xinjiang Uygur Autonomous Region, Urumqi 830011, Xinjiang Uygur Autonomous Region, China |
|
|
|
|
Abstract Objective To analyze the relationship between the whole blood elements levels and the risk of autism spectrum disorder(ASD) occurrence in autistic children, and establish and verify an individualized nomogram to predict the risk of ASD in children. Methods The data of 226 children volunteers, including 113 ASD children and 113 healthy children, recruited in outpatient clinic of a Grade Ⅲ Level A hospital in Urumqi in 2018, were retrospectively analyzed. All 226 children were randomly divided by R software into development group (n＝182) and verification group (n＝44). Univariate Logistic regression and multivariate Logistic regression analysis were used to screen independent risk factors of ASD. The corresponding nomogram prediction model was drawn according to the regression coefficient. The area under the receiver operating characteristic (ROC) curve was used to estimate the distinction and the calibration degree. Results Multivariate Logistic regression analysis showed that sex and blood copper, zinc, and magnesium levels were independent risk factors for ASD children. Area under the ROC curve of the development group and the verification group were 0.782 and 0.740，which indicated that the new prediction model had good discriminant ability. The correction results of the prediction model were evaluated by Hosmer-Lemeshow goodness-of-fit test. The P values of the two sets of calibration tests were 0.867 and 0.816. The prediction probability of the model was consistent with the actual probability, which indicated that the prediction model had strong consistency. Conclusion Individual predictive models of ASD children are helpful for screening and early identification of children with high risk of ASD.
|
|
|
|
|
|
|
|
|
