Abstract
Background: Understanding how aspects of behaviour change theories effect the success of interventions is critical for improving the overall efficacy of public health interventions. The present study aims to investigate whether there is a mediating relationship between the Transtheoretical Model (TTM) constructs of decisional balance (DB) and self-efficacy (SE) and fruit and vegetable intakes using pre-existing data from The 5+YourWay® Study. The 5+YourWay® Study was a randomised control trial conducted in a nationally representative sample of 25-60 year old New Zealanders that tested the efficacy of TTM-based, computer-tailored print communications administered over 3 months, with follow-ups 6 and 12 months from baseline.
Methods: Prior to testing for mediation items from pre-existing, validated, combined fruit and vegetable DB and SE scales were used to explore and confirm separate fruit scales and vegetable scales for each construct using principal components analysis (PCA) and confirmatory factor analysis (CFA). The PCA and CFA were conducted using split samples from the baseline cross-sectional data provided by 2,132 respondents to The 5+YourWay® Study screening survey. The reliability and construct and criterion-related validity of the scales were tested using the entire baseline sample. To ascertain whether it would be possible to test for mediation, multivariate analysis of variance was used to determine whether fruit intake and vegetable intake had changed across the 12-month study in the tailored intervention (n = 385), generic intervention (n = 384) and control (n = 386) groups. Structural equation models were constructed to test whether DB and SE mediated the effect the intervention had on fruit intake and vegetable intake between baseline and the 3-month, 6-month and 12-month follow-ups. Evidence of mediation was established using the Baron and Kenny causal steps method and the significance of mediated effects was tested using the Sobel method.
Results: The PCA and CFA identified three DB subscales for fruit DB and vegetable DB. Overall the DB subscales were reliable and had marginal validity. The single fruit SE scale and two vegetable SE subscales had good validity and reliability. Fruit intake increased in the tailored group between baseline and 3 months by 0.59 servings per day [(95%CI: 0.47, 0.71) p < 0.0001] and between baseline and 6 months by 0.70 servings per day [(95%CI: 0.56, 0.84) p < 0.0001] and remained stable for the rest of the study. In the tailored group vegetable intake increased by 0.47 serving per day [(95%CI: 0.29, 0.66) p< 0.0001] between baseline and 3 months, which was maintained for the duration of the study. None of the DB subscales mediated the effect the intervention had on fruit intake or vegetable intake. However, fruit SE accounted for 14.9% of the effect of the intervention on fruit intake between baseline and 3 months and 30.9% between baseline and 6 months. ‘Main meal’ vegetable SE accounted for 12.6% of the effect of the intervention on vegetable intake at 3 months, whilst ‘other occasions’ vegetable SE accounted for 24.0% of the effect of the intervention at six months and 45% of the effect of the intervention at 12 months.
Conclusions: The TTM-based, tailored communications used in The 5+YourWay® Study were effective at producing significant changes in fruit intake and vegetable intake in a nationally representative sample of New Zealanders aged 25-60 years. SE mediated the increases in fruit intake and vegetable intake, but DB had no mediating effect. Future interventions aiming to increases fruit and vegetable intake should focus on increasing an individual’s self-efficacy to perform these behaviours.