Worry and Risk Perceptions as Independent and Interacting Predictors of Health Protective Behaviors

Abstract

Worry and risk perceptions are generally found to be independently associated with health-promoting behaviors, although it is unknown whether they interact in ways that potentially dampen the effect of either construct on behavior. In this hypothesis-generating study, cancer-related worry and risk perception, and their interaction, were used to predict odds of meeting 5-a-day fruit and vegetable consumption guidelines and engaging in any exercise using data from a nationally representative sample (N = 10,230). Risk perception was not associated with either behavior; worry was associated only with exercise (OR = 1.77, 95% CI: 1.16, 2.70, p < .01). More important, their interaction was associated with these behaviors in a counterintuitive manner; among those higher in worry, higher levels of risk perception were associated with lower vegetable consumption (OR = 0.79, 95% CI: 0.62, 1.00, p < .05) and exercise (OR = 0.77, 95% CI: 0.63, 0.95, p = .01). These results suggest the hypothesis that, among people high in worry, attempts to increase risk perception could be counterproductive. These and related findings suggest the importance of distinguishing worry from risk perception, and future research is necessary to determine the causal nature of these associations.

Notes

¹Analyses using HINTS 2007 often encounter mode effects, where the results are slightly but significantly different between the RDD and mail samples, for reasons such as demand characteristics in a phone interview and inability to clarify questions in a mailed questionnaire. As such, it is recommended to limit the aggregated data to the random digit dialing sample when combining HINTS 2007 data with HINTS data from other years where only random digit dialing was used (Moser, Cantor, & Waldron, 2009).

²Body mass index and personal history of cancer were included as covariates in analyses, but were not significant and did not change the pattern of results for either fruit and vegetable consumption or diet and exercise, and were thus not included in the final analyses.

Note. Means, standard deviations, and percentages reflect weighted values; counts reflect raw data.

This article not subject to US copyright law.