Propriedades psicométricas dos instrumentos de avaliação do comportamento alimentar no transtorno do espectro autista

Abstract

The assessment of eating behavior in individuals with Autism Spectrum Disorder (ASD) represents a relevant challenge, particularly in view of the scarcity of psychometrically robust instruments, especially those designed for adolescents and adults. Eating behavior, in addition to its impact on nutritional and clinical status, also directly influences educational processes, including learning, school participation, pedagogical organization, and educational inclusion. In this context, this doctoral thesis aimed to critically analyze the instruments used to assess eating behavior in individuals with ASD, with an emphasis on tools applied to adolescents and adults, examining their structural characteristics, psychometric properties, and applicability in clinical, educational, and research settings. This thesis is structured as a collection of articles, comprising three complementary studies. The first article mapped the instruments used in the scientific literature to assess eating behavior in individuals with ASD, identifying an intense concentration on tools developed for childhood and significant gaps for more advanced age groups. The second article critically analyzed the psychometric properties of the identified instruments, revealing high structural heterogeneity and recurrent weaknesses in the reported metric evidence. The third article focused on a critical synthesis of the available evidence regarding the Swedish Eating Assessment for Autism Spectrum Disorders (SWEAA), the instrument most frequently used with adolescents and adults, systematically applying the COSMINGRADE criteria to assess the quality of the evidence. The results indicated that, although there is widespread use of instruments to assess eating behavior in individuals with ASD, the available psychometric evidence is fragmented, predominantly based on isolated indices of internal consistency, and insufficient to support individualized clinical and educational decision-making, particularly for adolescents and adults. The findings further emphasize that the use of assessment instruments requires a critical understanding of their theoretical assumptions, metric limitations, and contexts of application, constituting an essential formative competence in the education and training of health and education professionals. It is concluded that the main contribution of this thesis lies in elucidating the methodological limitations of the field and in advocating for consistent psychometric advances that can qualify evaluative, pedagogical, and educational practices at the interface between education, health, and inclusion.