Perceived risk, self-efficacy and outcome beliefs as predictors of physical activity and BMI: 5-year follow-up

Background: Perceived risk, self-efficacy and outcome beliefs are central concepts in several health behaviour theories, but their long-term effects on lifestyle changes remain unclear. We examined, in a five-year follow-up, whether perceived risks of diabetes and cardiovascular disease (CVD) predicted changes in physical activity (PA) or Body Mass Index (BMI kg/m²), or whether low PA and high BMI rather predicted higher perceived risks. Further, we examined whether perceived risks, self-efficacy, and outcome beliefs together predicted changes in PA or BMI. Methods: Participants were high diabetes risk participants (N=432) and randomly selected low to average risk participants (N=477) from FINRISK 2002 study, followed-up in 2007. Perceived absolute lifetime risks of diabetes and CVD, health action self-efficacy, outcome beliefs, and PA were self-reported; weight and height were assessed in health examinations. Structural equation models were adjusted for age and gender. Findings: In cross-lagged autoregressive models, we found no associations between perceived risks and PA, but higher baseline BMI predicted higher perceived risk of diabetes (β=.15, P<.001) and CVD (β=.10, P=.009) in the low/average risk group. In structural regression models, self-efficacy predicted increased PA (low/average risk group: β=.13, P=.011; high risk group: β=.17, P=.012), but not BMI change, whereas perceived risks or outcome beliefs did not predict changes in PA or BMI. Discussion: Perceived risk may rather reflect actual risk factors, such as high BMI and low PA, than predict lifestyle changes. Interventions aiming for long-term health behaviour change and maintenance should target risk perceptions together with other socio-cognitive factors, such as self-efficacy.