Intention to change risky health behaviours: is the Theory of Planned Behaviour a valid model, and can we improve upon it?
Consequences of risky health behaviours (smoking, drinking excess alcohol, illicit drug use and unsafe sex) are a leading cause of morbidity and mortality. Healthcare struggles to shoulder some of the burden of these behaviours and focus is shifting to prevention through encouraging behavioural change. The Theory of Planned Behaviour (TPB) is perhaps the most widely used behavioural change model. TPB seeks to explain the influence of attitude, perceived behavioural control and normative beliefs on intention to change. It is becoming clear, however, TPB alone does not fully explain intention to change. We aim to discover if this model is valid in a ‘risky behaviour’ population, and if certain moderating factors increase the model’s predictive power.
Data from The Healthy Foundations Lifestages Segmentation Survey was the basis for this secondary analysis project. 2427 eligible participants were identified, a TPB model built from the dataset, and logistic regression used to assess model significance and fit. Social deprivation index, health literacy level, frequency of social interaction and number of significant life event variables were created and used to moderate the model using logistic regression to explore effect on intention to change. Data was controlled for age, gender, presence of comorbid long-term illness and GHQ-12 score (a validated screening tool for mental health problems).
The TPB model was partially supported; significant associations with attitude (OR: 1.640, p = 0.000, 95% CI: 1.336-2.013) and perceived behavioural control (OR: 1.847, p = 0.001, 95% CI: 1.305-2.615) were observed. Normative beliefs have little influence in our population. Further analysis revealed positive findings with higher levels of health literacy, frequent interaction with social groups and low social deprivation index scores. Patients that were more health literate had greater perceived control of their ability to change (OR 2.996, p = 0.015, 95% CI 1.243-7.224), as did those who frequently attended social groups (OR 2.908, p = 0.001, 95% CI 1.557-5.430) and those who lived in less deprived areas (OR 2.707, p = 0.000, 95% CI 1.561-4.695). Attitude towards ability to change behaviour was found to be significantly higher in the less deprived (OR 1.932, p = 0.001, 95% CI 1.435-2.601) and more socially active groups (OR 1.770, p = 0.001, 95% CI 1.249-2.509).
Attitude and perceived control are the most important predictors of intention to change in a risky health behaviour population. Improving a patient’s health literacy and encouraging regular participation in social groups may further increase the number responding positively to attempts to change their risky behavioural habits. Further research into health literacy is particularly pertinent to Primary Care. Whilst social deprivation has been highlighted as another area for improvement, the healthcare system may be less well placed than other public services to influence this.