Associations between risk attributes, "control" and "dread", with perceptions of importance of cancer risk factors

Abstract

Background: Previous studies indicate lay perceptions of the importance of lifestyle factors in cancer risk are not necessarily consistent with epidemiological evidence. Providing insight into the factors associated with these perceptions may inform future cancer prevention strategies. Aim: This study explored whether differences in perceptions of importance of a range of factors in cancer risk could be explained by different factor attributes, with a specific focus on "Control" and "Dread" which feature prominently in the risk perception literature. Methods: A sample of convenience (n = 168; 17-57 years) comprised of students (n = 130) and non-academic staff (n = 38) from two South Australian Universities completed an online survey assessing perceived importance of 12 factors in cancer risk and six attributes for each risk factor (control, voluntariness, knowledge, dread, catastrophic potential and severity of consequences; assessed by psychometric paradigm scales-modified; Fischhoff et al., 1978). Results: Cigarette smoking was most commonly rated as very or extremely important in cancer risk (89%), followed by sun exposure, family history, pesticides and overweight (48-67%). Less than 40% rated other factors (e.g., alcohol, exercise, stress) as very or extremely important. Principal components analyses confirmed two attribute factors: "Control" (perceptions of controllability and voluntariness) and "Dread" (knowledge, dread and perceptions of severity of consequences and catastrophic potential). Logistic regression analyses, adjusted for gender, staff/student status, age and language spoken at home, indicated that "Dread" was positively associated with perceived importance for all risk factors (p < .001), and "Control" was positively associated with five risk factors (sun, diet, mobile phones, exercise and overweight; p < .01). Conclusions: The degree of "Dread" associated with risk factors, and to some extent "Control" contributed to explaining variation in perceptions of importance of a number of cancer risk factors. However, the variation explained differed considerably between risk factors, indicating that other potential influencing factors should be explored in future studies