Abstract TMP103: Associations Between Trust in Health Information Sources and Perceptions on the Modifiability of Stroke and Dementia Risks Using a U.S. Based Cohort

Background: Trust in healthcare information sources, specifically in healthcare professionals (HCPs) influences people’s health-related knowledge and behaviors. Epidemiological studies demonstrate that approximately 45% of dementia and 60% of stroke risks are attributable to modifiable risk factors. However, limited data exist on associations between trust levels in HCPs and people’s perceptions on the modifiability of dementia and stroke risks from a United States cohort. Methods: We developed a survey based on validated questionnaires and distributed it to a cohort of all U.S. states via the online Prolific platform in 2023. First, we described cohort characteristics and levels of trust in health information sources and mediums. We performed multivariable regression analyses between high trust of all HCPs (i.e. primary care physicians, specialists, and nurses) with the perceptions that dementia and stroke risks are modifiable, adjusting for age, sex assigned at birth, race/ethnicity, level of education, and status of knowing someone with stroke/dementia. Lastly, we performed hierarchal cluster analyses to characterize clusters of trust patterns and assessed their differences. Results: Our cohort consisted of 1,478 participants (52% females, median age 46 years [IQR:32-60], and 75% non-Hispanic Whites) with levels of trusts shown in Figure 1 . Following multivariable regression analyses, participants who highly trusted all HCPs were statistically more likely to perceive that maintaining and changing health habits can reduce the risks of stroke or dementia (adjusted odds ratios presented in Figure 2 ). Lastly, three clusters of trust patterns emerged: (i) those who highly trust most sources (n=781), (ii) those who only trust official health sources (n=540) and (iii) those who have low trust of all sources (n=103) ( Figure 3 ). Participants who have low trust of all sources (Cluster iii) were less likely to have a post-secondary degree (57% vs 67-71%, p=0.04) and to perceive that stroke (80% vs 93-94%, p<0.01) and dementia (65% vs 76-81%, P<0.01) risks are modifiable, compared to the other clusters. Conclusion: Our study identified highly trusted sources of health information, characterized trust patterns in a large U.S. cohort, and assessed their association with perceiving stroke/dementia risks as modifiable. This novel data could assist us in developing targeted interventions for risk reduction using trusted sources and mediums.