A machine learning-based diagnostic model for children with autism spectrum disorders complicated with intellectual disability.

Background: Early detection of children with autism spectrum disorder (ASD) and comorbid intellectual disability (ID) can help in individualized intervention. Appropriate assessment and diagnostic tools are lacking in primary care. This study aims to explore the applicability of machine learning (ML) methods in diagnosing ASD comorbid ID compared with traditional regression models. Method: From January 2017 to December 2021, 241 children with ASD, with an average age of 6.41 ± 1.96, diagnosed in the Developmental Behavior Department of the Children's Hospital Affiliated with the Medical College of Zhejiang University were included in the analysis. This study trained the traditional diagnostic models of Logistic regression (LR), Support Vector Machine (SVM), and two ensemble learning algorithms [Random Forest (RF) and XGBoost]. Socio-demographic and behavioral observation data were used to distinguish whether autistic children had combined ID. The hyperparameters adjustment uses grid search and 10-fold validation. The Boruta method is used to select variables. The model's performance was evaluated using discrimination, calibration, and decision curve analysis (DCA). Result: Among 241 autistic children, 98 (40.66%) were ASD comorbid ID. The four diagnostic models can better distinguish whether autistic children are complicated with ID, and the accuracy of SVM is the highest (0.836); SVM and XGBoost have better accuracy (0.800, 0.838); LR has the best sensitivity (0.939), followed by SVM (0.952). Regarding specificity, SVM, RF, and XGBoost performed significantly higher than LR (0.355). The AUC of ML (SVM, 0.835 [95% CI: 0.747-0.944]; RF, 0.829 [95% CI: 0.738-0.920]; XGBoost, 0.845 [95% CI: 0.734-0.937]) is not different from traditional LR (0.858 [95% CI: 0.770-0.944]). Only SVM observed a good calibration degree. Regarding DCA, LR, and SVM have higher benefits in a wider threshold range. Conclusion: Compared to the traditional regression model, ML model based on socio-demographic and behavioral observation data, especially SVM, has a better ability to distinguish whether autistic children are combined with ID.

Intelligence profiles and adaptive behaviors of high-functioning autism spectrum disorder and developmental speech and language disorders.

Objective: The present study was aimed at investigating the intelligence profiles and adaptive behaviors of children with high-functioning autism spectrum disorder (HFASD) and developmental speech and language disorders (DSLDs). We compared the similarities and differences of cognitive capabilities and adaptive functions and explored their correlations in the HFASD and DSLDs groups. Methods: 128 patients with HFASD, 111 patients with DSLDs and 114 typically developing (TD) children were enrolled into our study. Wechsler Intelligence Scale for Children-IV (WISC-IV) and Adaptive Behavior Assessment System-II (ABAS-II) were respectively applied to evaluate intelligence profiles and adaptive behaviors. Intelligence quotient (IQ) scores and adaptive functioning scores among the HFASD, DSLDs and TD groups were compared through one-way ANOVA. Pearson correlation coefficient was applied to examine the relationships between WISC indices and ABAS domains. Results: Outcomes showed significantly poorer intelligence profiles and adaptive behaviors in HFASD and DSLDs groups. Both children with HFASD and DSLDs demonstrated impairments in verbal comprehension and executive functions. Processing speed and working memory were the predominant defects of children with HFASD and DSLDs in the field of executive functions, respectively. Whereas perceptual reasoning was a relative strength for them. Children with DSLDs had balanced scores of all the domains in ABAS-II; nevertheless, HFASD individuals demonstrated striking impairments in Social domain. Correlation analysis showed IQs of children with HFASD were positively correlated with all the domains and General Adaptive Composite (GAC) of ABAS-II. Additionally, IQs were positively correlated with Conceptual domain and GAC for children with DSLDs. Compared with DSLDs group, intelligence displayed stronger correlations with adaptive behaviors in HFASD group. Conclusion: Our study expanded insights regarding intelligence profiles and adaptive behaviors of children with HFASD and DSLDs. Moreover, this study made breakthroughs in discovering positive correlations between IQs and adaptive functions in the two neurodevelopmental disorders.

The Application of the First Year Inventory for ASD Screening in China.

PURPOSE: The First Year Inventory (FYI) is a parent-report instrument, and is developed to assess behaviors of 12-month-old infants that could suggest risk for an eventual diagnosis of autism. This study was designed to examine the application of the FYI in the Chinese community. DESIGN AND METHODS: FYIs were completed at a community health center by 541 families during the child's physical examination at 12 months of age from 2013 to 2015. The weighted risk scores used in this study were based on US norms, and compared the FYI differences between China and the U.S. RESULTS: The total risk scores ranged from 5 to 42 points; the 95th percentile cutoff was 27.00(9.8 points higher than the 95th percentile cutoff in the US), the 98th percentile cutoff was 29.66(7.04 points higher than the 98th percentile cutoff in the US), and the 99th percentile cutoff was 31.83. Higher risk scores were found for boys than girls. Mothers with a junior college education reported significantly higher FYI risk scores than other three groups including high school, college graduates and post-graduates. CONCLUSIONS: There were no significant effects of birth parity, investigator, or investigation year on risk scores. Large-scale longitudinal research is encouraged in the future to develop an early detection model of autism in China.

THE RELATIONSHIP BETWEEN SERUM 25-HYDROXYVITAMIN D AND GLUCOSE HOMEOSTASIS IN OBESE CHILDREN AND ADOLESCENTS IN ZHEJIANG, CHINA.

OBJECTIVE: Evidence of the association between vitamin D, insulin resistance, and oral disposition index (oDI) in obese children and adolescents is limited. To fill this research gap, we measured serum 25-hydroxyvitamin D (25[OH]D) levels in obese children and analyzed the relationship between serum 25(OH)D levels and glucose homeostasis. METHODS: Altogether, 348 obese and 445 nonobese children and adolescents (age, 6 to 16 years) were enrolled in this study. Obese children were divided into 4 subgroups: normal glucose tolerance (NGT), impaired fasting glucose (IFG), impaired glucose tolerance (IGT), and combined IFG and IGT (IFG+IGT) according to oral glucose tolerance test results. We measured serum 25(OH)D levels and calculated the homeostasis model assessment (HOMA) of insulin resistance (IR), the whole-body insulin sensitivity index (WBISI), and the disposition index. RESULTS: The levels of 25(OH)D in the obese group were significantly lower than in the nonobese group; serum 25(OH)D level in the NGT subgroup was higher than those of the other 3 subgroups, and it was significantly inversely correlated with logHOMA-IR (r = -0.090; P = .045) and positively correlated with logWBISI and logHOMA-oDI (r = 0.091, P = .049; and r = 0.108, P = .046, respectively). Obese patients with vitamin D deficiency thus have a significantly higher risk of disturbances in glucose metabolism. CONCLUSION: 25(OH)D deficiency or insufficiency is quite common in obese children and adolescents in Zhejiang, China. Obese patients with 25(OH)D deficiency (<30 nmol/L) are shown to be at higher risk for abnormal glucose metabolism.

A new diterpene from the stems of Manihot esculenta.

A new diterpene, named yucalexin P-23 (1), together with three known compounds, yucalexin P-15 (2), protocatechuic acid (3), and catalpinic acid (4), was isolated from the stems of Manihot esculenta. Their structures were determined by means of physicochemical evidence and spectral analysis. Compounds 3 and 4 exhibited antimicrobial activity.