Background: The predictive value of the psychosocial constructs of Theory of Planned Behaviour (TPB) on subsequent dietary habits has not been previously investigated in a multivariate approach that includes demographic factors and past dietary behaviour among adults. The aim of this study was to investigate to what extent TPB constructs, including intention, attitudes, subjective norms, perceived behavioural control, and perceived social norms, measured at age 25 predicted four eating behaviours (intake of fruits and vegetables, whole grains, total fat and added sugar) eight years later.
Methods: Two hundred and forty men and 279 women that participated in the Oslo Youth Study were followed from 1991 to 1999 (mean age 25 and 33 years, respectively). Questionnaires at baseline (1991) included the constructs of the TPB and dietary habits, and at follow-up (1999) questionnaires included demographic factors and diet. For the assessment of diet, a food frequency questionnaire (FFQ) with a few food items was used at baseline while an extensive semi-quantitative FFQ was used at follow-up.
Results: Among men, attitudes, subjective norms and previous eating behaviour were significant predictors of fruit and vegetable intake, while education and past eating behaviour were predictive of whole grain intake in multivariate analyses predicting dietary intake at follow-up. For women, perceived behavioural control, perceived social norms and past behaviour were predictive of fruit and vegetable intake, while subjective norms, education and past eating behaviour were predictive of whole grain intake. For total fat intake, intention was predictive for men and perceived behavioural control for women. Household income and past consumption of sugar-rich foods were significant predictors of added sugar intake among men, while past intake of sugar-rich foods was a significant predictor of added sugar intake among women.
Conclusion: After adjusting for potential confounding factors, all psychosocial factors assessed among young adults appeared predictive of one or more eating behaviours reported eight years later. Results point to the influence of psychosocial factors on future eating behaviours and the potential for interventions targeting such factors.