Sebastian Isbanner, Julia Carins, Taylor Willmott
Women generally consume a healthier diet than men; however, both fall short of dietary guidelines. Health behaviour theories are often applied in a gender-neutral manner, implicitly assuming that explanatory models operate similarly across groups. This conceptual replication tested the Capability, Opportunity, Motivation and Behaviour (COM-B) model for dietary behaviour, originally evaluated in a female-dominant sample, in a male-dominant sample (N = 1,358; 67% male). Using the same measurement battery and a closely aligned analytic approach, structural equation modelling indicated acceptable overall fit, with the COM-B model explaining 22% of the variance in diet quality. As in the original study, capability and opportunity were associated with behaviour indirectly through motivation, which was the strongest predictor of diet quality. At the indicator level, the male-dominant model retained action control and habits (capability), descriptive norms and the home food environment (opportunity), and self-efficacy, intentions, affective attitudes and perceived behavioural control (motivation), whereas identity, decision-making barriers, instrumental attitudes and the shopping environment indicator were not retained during trimming. These differences, relative to the female-dominant model, suggest potential variation in how COM-B constructs are operationalised across gender-dominant samples and should be treated as hypothesis-generating. Discriminant validity concerns (habits-action control; self-efficacy-intention) further indicated overlap between conceptually distinct constructs in the male-dominant sample. Overall, findings support the stability of the higher-order COM-B pathway while highlighting potential group differences in indicator contributions, with implications for measurement and population-sensitive intervention design.