A Robust Feature Selection Analysis of Early Childhood Development Predictors in Thailand Using Stability Selection Early Childhood Development Predictor Selection Using Stability Selection
Article Sidebar
Main Article Content
Abstract
Early identification of developmental delays is critical for timely intervention. However, the specialized training and time required for frequent professional evaluations create a high clinical demand. This highlights the need for proxy predictors, features that are strong predictors of developmental delays that are also easily observable by caregivers, which can act as early triggers for professional referral. This study aimed to identify robust clinical and environmental predictors of suspected developmental delay across five DSPM domains and overall development among children in Chonburi, Thailand. Using a cross-sectional dataset (N=300) with 243 engineered features, a two-stage statistical pipeline was employed: (1) Stability Selection (100 iterations, 50% subsampling) to identify variables with a selection frequency ≥ 0.50; and (2) Non-parametric Bootstrapping (1,000 iterations) with multivariate logistic regression to calculate Median Odds Ratios (OR) and 95% Confidence Intervals (CI). Dental plaque emerged as a primary predictor, significantly increasing the risk for overall developmental delay (OR = 5.36) and linguistic delays. Other stable risk factors included starting electronic media between ages 2.0–2.9 years (OR = 6.95), a history of childhood pneumonia (OR = 5.96), and receiving the DSPM manual only during vaccination visits (OR = 5.47). Conversely, avoiding electronics (OR = 0.29) and proactive DSPM utilization (OR = 0.30) were significant protective factors for expressive language. No stable predictors reached the threshold for motor or social domains. Findings emphasize the sentinel role of caretakers’ health literacy, oral health, and digital hygiene in developmental screening. Poor oral health, early digital device usage, and lack of health literacy through DSPM are proxy predictors for developmental delay, while caretakers’ health literacy through utilization of DSPM and avoiding usage of electronic devices lowers the likelihood of developmental delay. Pediatric nursing interventions should prioritize caretakers’ health literacy and shift from reactive to proactive DSPM utilization to improve developmental outcomes.
Article Details
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
The authors retain copyright and permit the journal the copyright of first publication
Articles, once having passed the review process and accepted for publication in the CDMH Journal, are copyrighted under the CDMH Journal, Department of Mental Health, Ministry of Public Health. Please be aware distribution of CDMH Journal content for commercial purposes without permission is expressly prohibited. However, distribution with intent to educate, advocate, or spread awareness within the general public and research communities is permitted and encouraged with the understanding that the CDMH Journal Editorial Board do not hold jurisdiction or liability for any accompanying comments, text, or information from third parties, either in favor for or against the original article’s assertions, conclusions, methodology, or content.
References
Abramovitz, I., Zini, A., Atzmoni, M., Kedem, R., Zur, D., Protter, N. E., & Almoznino, G. (2021). Cognitive performance and its associations with dental caries: Results from the dental, oral, medical epidemiological (DOME) records-based nationwide study. Biology, 10(3), 178.
Bodrumlu, E. H., Demiriz, L., & Toprak, S. (2018). Relationship between severe early childhood caries and dental development. European journal of paediatric dentistry, 19, 156-160.
Bornstein, M. H., Hahn, C. S., & Suwalsky, J. T. (2013A). Physically developed and exploratory young infants contribute to their own long-term academic achievement. Psychological Science, 24(10), 1906–1917. https://doi.org/10.1177/0956797613482721
Bornstein, M. H., Hahn, C. S., & Wolke, D. (2013B). Systems and cascades in cognitive development and academic achievement. Child Development, 84(1), 154–162. https://doi.org/10.1111/j.1467-8624.2012.01849.x
Das, A. C., Dai, R., Lokshin, A., Salam, S., & Smith, L. (2026). Identifying Predictive Combinations of Biomarkers for Early Cancer Detection with Stability Selection in Combination with Ensemble Learning. Statistics in Biosciences, 1-29.
Ghareeb, Z. F. (2023). A comparative study between shrinkage methods (ridge-lasso) using simulation. Periodicals of Engineering and Natural Sciences, 11(2), 36-47.
Hyde, R., O’Grady, L., & Green, M. (2022). Stability selection for mixed effect models with large numbers of predictor variables: A simulation study. Preventive Veterinary Medicine, 206, 105714. https://doi.org/10.1016/j.prevetmed.2022.105714
Jantanavivat, W., Nopmaneejumruslers, K., & Seree, P. (2022). Effects of the language development program based on early childhood developmental surveillance and promotion. International Journal of Child Development & Mental Health, 10(2).
Maneekunwong, K., Trakooltorwong, P., Kodyee, S., Wutthipan, N., Sopajaree, C., & Moonpanane, K. (2025). Electronic device usage during COVID-19 and suspected developmental delay in rural Thai children: A case-control study. Asia-Pacific Social Science Review, 25(3), 4.
Meinshausen, N., & Bühlmann, P. (2010). Stability selection. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 72(4), 417–473. https://doi.org/10.1111/j.1467-9868.2010.00740.x
Ministry of Public Health. (2020). Developmental surveillance and promotion manual (DSPM). Department of Health.
Räikkönen, K., Forsen, T., Henriksson, M., Kajantie, E., Heinonen, K., Pesonen, A. K., ... & Eriksson, J. G. (2009). Growth trajectories and intellectual abilities in young adulthood: The Helsinki Birth Cohort study. American Journal of Epidemiology, 170(4), 447–455. https://doi.org/10.1093/aje/kwp147
Rithipukdee, N., & Kusol, K. (2022). Factors associated with the suspected delay in the language development of early childhood in southern Thailand. Children, 9(5), 662. https://doi.org/10.3390/children9050662
Russell, S. J., & Norvig, P. (2021). Artificial Intelligence: A Modern Approach (4th ed.). Pearson.
Saputro, D. R. S., Wahyu, N. L., & Widyaningsih, Y. (2025). Performance of ridge regression, least absolute shrinkage and selection operator, and elastic net in overcoming multicollinearity. Journal of Multidisciplinary Applied Natural Science, 5(2), 370-382.
Schneider, W., Niklas, F., & Schmiedeler, S. (2014). Intellectual development from early childhood to early adulthood: The impact of early IQ differences on stability and change over time. Learning and Individual Differences, 32, 156–162. https://doi.org/10.1016/j.lindif.2014.02.001
Thaineua, V., Sirithongthaworn, S., Kanshana, S., Isaranurak, S., Karnkawinpong, O., Benjaponpitak, A., ... & Tangviriyapaiboon, D. (2024). A 9‐year retrospective cohort study of the monitoring and screening of childhood developmental delay in Thailand. Child: Care, Health and Development, 50(2), e13233. https://doi.org/10.1111/cch.13233
Wehrens, R., Franceschi, P., Vrhovsek, U., & Mattivi, F. (2011). Stability-based biomarker selection. Analytica Chimica Acta, 705(1-2), 15–23. https://doi.org/10.1016/j.aca.2011.02.025
Zhou, J., Sun, J., Liu, Y., Hu, J., & Ye, J. (2013). Patient risk prediction model via top-k stability selection. In Proceedings of the 2013 SIAM International Conference on Data Mining (pp. 55–63). Society for Industrial and Applied Mathematics. https://doi.org/10.1137/1.9781611973310.7
Zhu, Z., Chang, S., Cheng, Y., Qi, Q., Li, S., Elhoumed, M., ... & Sudfeld, C. R. (2019). Early life cognitive development trajectories and intelligence quotient in middle childhood and early adolescence in rural western China. Scientific Reports, 9(1), 18315. https://doi.org/10.1038/s41598-019-54848-x
