Dietary supplement use and cancer-preventive health behaviors among colorectal cancer survivors and a comparison population in North Carolina

This research examines dietary supplement use and cancer preventive health behaviors among people with and without colorectal cancer in NC STRIDES. Dietary supplement use categories were created using a normative approach. The categories were characterized by different demographic profiles and types of vegetable and fruit consumption. Individuals in Category 1 (multivitamins and single vitamins/minerals) and Category 2 (nonvitamin/nonmineral products) were more likely to be consuming more vegetables, and higher quality vegetable and fruits, than Category 5 (no dietary supplements); while Category 4 (multivitamins only) had patterns that were more consistently similar to Category 5. Collapsing dietary supplement use into a dichotomous variable (yes/no) does not fully capture the dietary and behavioral differences that exist among categories.; Qualitative interviews further revealed that behavioral motivations are complex, especially in an aging population where dietary supplement use is common. It may be that people who take multivitamins only are more similar to those who do not take any supplements because they are taking them for “nutrient insurance,” rather than for well-being or self-treatment of chronic diseases. Major themes included the role of significant life events in dietary change, concerns about contamination of the food supply, and the overall lack of guidance about supplements from physicians.; Five health behavior clusters were defined using cluster analysis. Most individuals in Cluster 1 met the goals: eating 5-A-Day, being physically active 30 minutes per day, regular multivitamin use, and a normal BMI. Nearly 90% of Cluster 2 ate 5-A-Day and 100% of Cluster 3 met the physical activity goals; both were low in all others. Cluster 4 was average and Cluster 5 did not meet any goals (and all were obese). Clusters 1 and 2 were already at 5-A-Day, and did not increase after participation in the NC STRIDES intervention. Clusters 3–5 were at four daily servings—and all increased by one serving after intervention. While these findings do not necessarily support tailoring on health behaviors (the intervention worked almost equally well among all clusters not already eating 5-A-Day), understanding multidimensional patterns is critical to our conceptualization of “cancer preventive” behaviors-and our ability to impact them.