Objective
To objective of the present study was to investigate the initial evidence for construct validity of the Brazilian Mindful Eating Scale (MES-B), developed to assess mindfulness in the context of eating behavior.
Method
A total of 236 overweight or obese Brazilian women, aged between 18 and 59 years, answered the Mindful Eating Scale (MES) and questionnaires related to eating behavior, mindfulness, anxiety, depression, self-compassion, and eating disorders. Data were analyzed using confirmatory factor analysis. Models were adjusted according to modification indices, if theoretically and semantically appropriate, respecting a > 10 threshold. McDonald’s ω coefficients were estimated as a reliability measure. Measurement invariance was tested by Multiple Indicators, Multiple Causes (MIMIC) models.
Results
The number of MES items had to be reduced from 28 to 24 for better scale reliability, due to low factor loadings and high cross-loadings on more than 1 item. In this reduced scale, McDonald’s omega coefficients (ω) ranged from 0.47 to 0.83, with only 2 factors (acceptance and awareness) showing an ω greater than 0.70. Unhealthy eating habits, measured by the Dutch Eating Behavior Questionnaire (DEBQ) and Binge Eating Scale (BES), were negatively associated with mindful eating.
Conclusion
The Brazilian Mindful Eating Scale (MES-B), consisting of 24 items and six domains, was initially and successfully validated. This proposed version may be promising but should be evaluated in all genders and not only in the overweight female population.
Preregistration
ClinicalTrials.gov Identifier: NCT02893150 – MindEat.