Adaptação transcultural e avaliação psicométrica da Teruel Orthorexia Scale para uma amostra de praticantes de exercício físico
Data
2020-11-12
Autores
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Proposições: (i) realizar adaptação transcultural da Teruel Orthorexia Scale (TOS) para a língua portuguesa; (ii) avaliar as propriedades psicométricas da escala para uma amostra de praticantes de exercício físico; (iii) verificar a prevalência de aspectos característicos de Ortorexia Nervosa e Saudável entre os participantes. Material e Métodos: O conteúdo da versão original da TOS (espanhol) foi adaptado para a língua portuguesa seguindo protocolo padronizado. A versão em português da TOS foi testada em estudo piloto com 30 adultos brasileiros. Neste, verificou-se que a maioria dos participantes (>75%) compreendeu bem o conteúdo da versão em português e, portanto, o estudo definitivo foi realizado. A amostra foi composta por indivíduos que frequentavam academias da cidade de Araraquara/SP para realizar exercícios físicos. Os participantes preencheram a versão em português da TOS e um questionário para caracterização da amostra. As propriedades psicométricas do modelo bifatorial da TOS (original) foram avaliadas para a amostra. Na validade fatorial, utilizou-se os índices razão de qui-quadrado pelos graus de liberdade (χ 2 /gl), Root Mean Square Error of Approximation (RMSEA), Comparative Fit Index (CFI) e Tucker-Lewis Index (TLI). A variância extraída média (VEM) e o coeficiente de determinação (r2 ) foram calculados para investigar as validades convergente e discriminante, respectivamente. Além disso, foi verificada a confiabilidade dos fatores por meio dos coeficientes alfa ordinal (α) e ômega (ω). A invariância também foi avaliada e utilizou-se análise multigrupos (Δχ 2 ). Ainda, os escores médios de cada fator da TOS foram calculados e as prevalências foram estimadas. Resultados: Participaram do estudo 226 sujeitos (homens=63,7%) com média de idade de 27,8 (desviopadrão=5,1) anos. O modelo bifatorial da TOS apresentou adequada validade fatorial (χ 2 /gl=2,74; RMSEA=0,09; CFI=0,94; TLI=0,93), convergente (VEM=0,58/0,53) e discriminante (r2=0,19), bem como boa confiabilidade (α=0,91/0,90; ω=0,87/0,78) para a amostra de estudo. Além disso, o modelo bifatorial da TOS foi invariante quando testado em subamostras independentes e entre homens e mulheres (Δχ 2 : p=0,074-0,748). Quanto aos escores médios dos fatores da TOS, observou-se que apenas a Ortorexia Nervosa foi diferente entre homens e mulheres sendo que o público feminino apresentou valores mais elevados (p=0,001). Na amostra total, a prevalência de aspectos típicos de Ortorexia Nervosa foi de 5,3% e de Ortorexia Saudável foi de 41,1%. Conclusão: A versão em português da TOS apresentada neste estudo foi bem compreendida pelos participantes e poderá ser utilizada em contexto brasileiro. Para a amostra estudada, o modelo bifatorial da TOS foi válido e confiável. Ainda, foram encontradas prevalências importantes de aspectos disfuncionais e “normais” quanto à ingestão de alimentos saudáveis. Portanto, o uso da TOS em futuras investigações pode ser útil para rastrear indivíduos com comportamentos alimentares de risco visando desenvolver estratégias de ação mais direcionadas e efetivas. Palavras chave: Ortorexia; Escala; Validade; Confiabilidade; Tradução; Exercício Físico.
Aims: (i) to perform a cross-cultural adaptation of the Teruel Orthorexia Scale (TOS) for the Portuguese language; (ii) to evaluate the psychometric properties of the scale in a sample of Brazilian gym users; (iii) to verify the prevalence of characteristic aspects of Orthorexia Nervosa and Healthy Orthorexia among the participants. Material and Methods: The content of the original version of TOS (Spanish) was adapted to the Portuguese language following standardized protocol. The Portuguese version of TOS was tested in a pilot study with 30 Brazilian adults. In this study, it was found that the majority of the participants (>75%) understood the content of the Portuguese version; therefore, the definitive study was carried out. The sample consisted of individuals who attended gyms in the city of Araraquara (SP) to perform physical exercises. The volunteers completed Portuguese version of TOS and a questionnaire on sample’s characteristics as paper and pencil survey. The psychometric properties of the TOS bifactorial model (original) were assessed for study sample. Factorial validity was estimated using chi-square ratio to degrees of freedom (χ 2 /df), Root Mean Square Error of Approximation (RMSEA), Comparative Fit Index (CFI) e Tucker-Lewis Index (TLI). The average variance extracted (AVE) and the determination coefficient (r2 ) were calculated to investigate convergent and discriminant validities, respectively. Furthermore, the reliability of the factors was evaluated using alpha ordinal (α) and omega (ω) coefficients. The invariance was assessed by multigroup analysis (Δχ 2 ). The mean scores of each factor of TOS were calculated and the prevalence were estimated. Results: A total of 226 subjects (men=63.7%) with a mean age of 27.8 (standard deviation=5.1) years participated. TOS bifactorial model showed adequate factorial validity (χ 2 /df=2.74; RMSEA=.09; CFI=.94; TLI=.93), convergent (AVE=.58/.53) and discriminant (r2=.19), as well as good reliability (α=.91/.90; ω=.87/.78) for sample. In addition, the bifactorial model of TOS was invariant across independent subsamples and among men and women samples (Δχ 2 : p=.074-.748). Regarding mean scores, only Orthorexia Nervosa factor was different among men and women, with female sample showing higher values (p=.001). In the total sample, the prevalence of typical aspects of Nervous Orthorexia was 5.3% and Healthy Orthorexia was 41.1%. Conclusion: The Portuguese version of the TOS presented in this study was well understood by the participants and may be used in Brazilian context. For the sample studied, the TOS bifactorial model was valid and reliable. Still, important prevalence of dysfunctional and normal aspects were found regarding the intake of healthy foods. Therefore, the use of TOS in future investigations can be useful to screen individuals with risky eating behaviors in order to develop more targeted and effective action strategies.
Aims: (i) to perform a cross-cultural adaptation of the Teruel Orthorexia Scale (TOS) for the Portuguese language; (ii) to evaluate the psychometric properties of the scale in a sample of Brazilian gym users; (iii) to verify the prevalence of characteristic aspects of Orthorexia Nervosa and Healthy Orthorexia among the participants. Material and Methods: The content of the original version of TOS (Spanish) was adapted to the Portuguese language following standardized protocol. The Portuguese version of TOS was tested in a pilot study with 30 Brazilian adults. In this study, it was found that the majority of the participants (>75%) understood the content of the Portuguese version; therefore, the definitive study was carried out. The sample consisted of individuals who attended gyms in the city of Araraquara (SP) to perform physical exercises. The volunteers completed Portuguese version of TOS and a questionnaire on sample’s characteristics as paper and pencil survey. The psychometric properties of the TOS bifactorial model (original) were assessed for study sample. Factorial validity was estimated using chi-square ratio to degrees of freedom (χ 2 /df), Root Mean Square Error of Approximation (RMSEA), Comparative Fit Index (CFI) e Tucker-Lewis Index (TLI). The average variance extracted (AVE) and the determination coefficient (r2 ) were calculated to investigate convergent and discriminant validities, respectively. Furthermore, the reliability of the factors was evaluated using alpha ordinal (α) and omega (ω) coefficients. The invariance was assessed by multigroup analysis (Δχ 2 ). The mean scores of each factor of TOS were calculated and the prevalence were estimated. Results: A total of 226 subjects (men=63.7%) with a mean age of 27.8 (standard deviation=5.1) years participated. TOS bifactorial model showed adequate factorial validity (χ 2 /df=2.74; RMSEA=.09; CFI=.94; TLI=.93), convergent (AVE=.58/.53) and discriminant (r2=.19), as well as good reliability (α=.91/.90; ω=.87/.78) for sample. In addition, the bifactorial model of TOS was invariant across independent subsamples and among men and women samples (Δχ 2 : p=.074-.748). Regarding mean scores, only Orthorexia Nervosa factor was different among men and women, with female sample showing higher values (p=.001). In the total sample, the prevalence of typical aspects of Nervous Orthorexia was 5.3% and Healthy Orthorexia was 41.1%. Conclusion: The Portuguese version of the TOS presented in this study was well understood by the participants and may be used in Brazilian context. For the sample studied, the TOS bifactorial model was valid and reliable. Still, important prevalence of dysfunctional and normal aspects were found regarding the intake of healthy foods. Therefore, the use of TOS in future investigations can be useful to screen individuals with risky eating behaviors in order to develop more targeted and effective action strategies.
Descrição
Palavras-chave
Ortorexia, Escala, Validade, Confiabilidade, Tradução, Exercício Físico, Physical Exercise, Orthorexia, Reliability