Abstract
BACKGROUND: Prior studies found that some groups have lower genetic consent rates than others. Participant consent for genetic studies enables randomized trials to examine effects of interventions compared to control in participants with different genotypes. METHODS: Unadjusted and multivariate associations between genetic consent rates and participant, study, and consent characteristics in 9573 participants approached for genetics consent in the multicenter Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial, which used a layered genetics consent. RESULTS: Eighty-nine percent of eligible participants consented to genetic studies ("Any Consent") and 64.7% consented to studies of any genes by any investigator ("Full Consent"), with similar rates in randomized groups. Controlling for multiple characteristics, African-Americans had lower consent rates than others (Any Consent Odds Ratio, OR = 0.62, p = 0.0004; Full Consent OR = 0.67, p < 0.0001). Those with high school or higher education level had higher rates than less than high school graduates (Full Consent ORs 1.41-1.69, p-values < 0.0001). Consent rates were lower when genetics consent was separate from the main trial consent on the same day (Any Consent OR 0.30; Full Consent OR 0.52, p values < 0.0001) or on a subsequent day (Any Consent OR 0.70, p = 0.0022; Full Consent OR 0.76, p = 0.0002). CONCLUSION: High rates of consent for genetic studies can be obtained in complex randomized trials, with lower consent rates in African-Americans, in participants with less than high-school education, and for sharing samples with other investigators. A genetics consent separated from the main trial consent was associated with lower consent rates.