The Effect of Personalized Risk Feedback On Mexican American's Risk Perceptions for Heart Disease, Diabetes, and Colon Cancer

Theoretical Background and research questions/hypothesis: Accurate risk perceptions are crucial to help people understand their risk for chronic diseases and to initiate or maintain health protective behaviors. The present study examines Mexican American’s risk perceptions for heart disease, diabetes, and colon cancer prior to and following receipt of personalized risk assessments (RA). Since health messages often do not reach Mexican Americans, personalized approaches may be an effective way to communicate risk information and to reduce risk perception biases.

Methods: To assess the effectiveness of a family-based intervention to communicate disease risk information, 497 Mexican Americans (comprising 162 households) were recruited from a longitudinal population-based study of cancer-risk factors among Mexican Americans established by and housed at the MD Anderson Cancer Center. Households were randomized into one of four feedback conditions characterized by the receiver of the feedback (one individual in the family versus all members of the family) and the content of the feedback (a risk assessment and behavioral recommendations versus risk assessment only). Participants completed a baseline survey and follow-up phone survey approximately 3 months after receiving personalized RAs or a pedigree only. Generalized estimating equations were used to predict baseline and follow-up risk perceptions controlling for familial risk, participant characteristics, risk factors, and receipt of personalized RAs.  

Results: Participants perceived moderate to high levels of risk at baseline for heart disease, diabetes, and colon cancer. These estimates were higher than familial risk estimates generated by the CDC’s Family Healthware™. Nearly 53% rated their heart disease risk as higher than Family Healthware™, nearly 40% for diabetes and 62% for colon cancer. At baseline, family history was the most consistent predictor of risk for heart disease (β=0.16, p<.01), colon cancer (β=0.25, p<.01), and diabetes (β=0.65, p<.01). Following receipt of risk feedback mean risk perceptions significantly decreased (p<.05) for all three conditions, although minimally. For nearly 50% of participants no change in risk perception occurred.  Baseline perceived risk was the most consistent predictor of perceived risk at follow-up for heart disease (β=0.26, p<.01), colon cancer (β=0.32, p<.01), and diabetes (β=.18, p<.01). Personalized RAs were also positively associated with perceived risk for heart disease (β=0.10, p<.01) and diabetes (β=0.07, p<.01) at follow-up, controlling for baseline levels of perceived risk.

Conclusions:  Mexican Americans perceive themselves to be moderately at-risk for health conditions that impact them at high rates. Although family health history was the most consistent predictor of risk, results suggest the receipt of personalized RAs may help Mexican Americans appropriately re-assess their risk for heart disease and diabetes.

Implications for research and/or practice: Receiving personalized risk feedback may help individuals identify an appropriate level of risk.