Fatores de risco perinatais para o diagnóstico precoce do transtorno do espectro autista em goiás: um estudo caso controle com linhage de bases de dados públicas

Abstract

INTRODUCTION: Autism Spectrum Disorder (ASD) is a neurodevelopmental condition with multifactorial etiology, whose prevalence has increased globally, although data on perinatal risk factors and regional particularities remain scarce in Brazil. OBJECTIVE: To identify pre- and perinatal risk factors associated with ASD diagnosis in children up to five years of age in the state of Goiás, Brazil. METHODS: This was a case-control study using deterministic linkage of health databases, utilizing the Live Birth Information System (SINASC) and the Outpatient Health Actions Registry – Psychosocial Care (RAAS-PSI). The population included live births in Goiás during 2019, with follow-up until 2024 to identify ASD diagnoses (ICD-10 F84). Controls were selected through simple random sampling from births without the outcome. Multiple imputation by chained equations (MICE) was applied for missing data, and hierarchical logistic regression with LASSO variable selection was performed. RESULTS: A total of 96,112 births and 1,033 ASD cases were identified, yielding a cumulative prevalence of 1.07% (95%CI 1.01-1.14). In the multivariate analysis with LASSO selection, male sex (OR=3.35; 95%CI 2.86-3.92; p<0.001) and the presence of congenital malformations (OR=2.48; 95%CI 1.26-4.88; p=0.008) remained significantly associated with ASD diagnosis. Classic variables from the literature such as advanced parental age, prematurity, and low birth weight showed no statistical significance in the adjusted models. CONCLUSION: Male sex and congenital malformations constitute the main risk factors for early ASD diagnosis in the studied population. Public database linkage proved to be a feasible tool for producing national evidence, although the socioeconomic homogeneity of the sample may have masked other associations. The findings reinforce the need for neurodevelopmental surveillance in higher vulnerability groups and for diagnostic strategies sensitive to gender particularities.