Strategies for improving access to information promoting dementia risk reduction

Dementia is a life-changing and life-limiting condition, disproportionately affecting disadvantaged people who experience high dementia risk factor prevalence and low risk reduction knowledge.
Dementia risk reduction initiatives primarily reach advantaged people who tend to be at lower risk of developing dementia. Strategies are therefore needed to ensure dementia risk information reaches those for whom it has most potential benefit. A review of knowledge dissemination theories identified interpersonal information sharing as an underexplored strategy for reaching high-risk groups of people with dementia risk information. The Wicking Centre’s Preventing Dementia Massive Open Online Course (PDMOOC) is a free global dementia risk education initiative that provides information directly to participants (PDMOOC participants), who in turn provide information to additional people through information sharing (information recipients). This body of work aimed to understand the PDMOOC’s ability to promote dementia risk reduction among groups of people who tend not to engage with this information. Specifically, this thesis addressed four aims:
i. Examine PDMOOC participants’ characteristics, PDMOOC experiences, and PDMOOC outcomes.
ii. Examine the audience, content, and impact of information sharing from the perspective of PDMOOC participants.
iii. Explore the drivers of PDMOOC participant information sharing.
iv. Examine the content and impact of information sharing from the perspective of
information recipients.
Method
Characteristics of 55,739 PDMOOC participants and their PDMOOC experiences and outcomes were examined across 7 PDMOOC iterations from 2016-2020 using fixed-response and free-text online survey questions. Categorical responses were analysed using generalised linear models and generalised additive models. Free-text responses were analysed using topic modelling. A pragmatic, complementary, mixed methods approach was adopted. Multiple methods and data sources produced multi-faceted understandings. Initial study results informed subsequent study design. Each subsequent result was interpreted in light of all previous results, and all results were integrated to reach final conclusions and recommendations.
PDMOOC-related information sharing content, audience, and perceived recipient impact were examined using responses to a free-text question asking how knowledge was applied. Reports of online or in-person information sharing were identified (2,010 reports among 8,473 responses) and categorised inductively into sub-themes within the pre-defined categories of content shared, information recipients, and information sharing impacts. Categorisation was compared across multiple coders to evaluate coding reliability.
To explore the factors that influenced information sharing, in-depth interviews were undertaken with 37 PDMOOC participants. Quantitative and qualitative data collected during interviews was analysed using social network analysis techniques and reflexive thematic analysis, respectively.
The perspectives of information recipients on shared information content and impact were collected using categorical and free-text online survey questions. Surveys were completed by PDMOOC enrolees who learnt about the PDMOOC through word-of-mouth (n=559). Categorical responses were analysed using summary statistics, paired t-tests, and Wilcoxon tests. Free-text responses were analysed by coding into pre-defined risk-factor categories, with additional inductive categories added as required. Categorisation was compared across multiple coders to evaluate coding reliability.
Results
The PDMOOC disproportionately attracted women (86% women), older people (mean age 48.9; standard deviation 15.0), and highly educated people (78% post-secondary educated). Respondents reported improved dementia risk knowledge after PDMOOC completion (98.4%, n=16,514), and 85.8% of respondents agreed that “The MOOC has had an impact on my behaviour and lifestyle choices”.
PDMOOC participants regularly shared PDMOOC-related information. A quarter of free-text knowledge application question respondents reported sharing information. Of 37 interview participants, 36 reported information sharing, providing information to 544 people through close and weak social connections, including 55% of reported close contacts.
The proportion of men was higher among information recipients than among PDMOOC participants (24% vs 14%), as was the proportion of people with less than post-secondary education (21% vs 16%).
In qualitative interview data, participants reported sharing information with network members they felt would be receptive: generally, women and older people, aligning with stereotypes around dementia and preventative health behaviour. This data also showed that dementia stigma influenced dementia risk information sharing. Quantitative data showed that PDMOOC participants disproportionately shared information with network members whose education level was the same as the majority of their close contacts. Coding reliability analysis showed that information sharing about the risk domains of physical activity, cognitive activity, and diet was reported most frequently by both PDMOOC participants and information recipients (respectively 55%, 42%, and 16% of PDMOOC participant reports; respectively 23%, 13%, and 23% of information recipient reports). Thematic analyses showed that participants preferentially shared information that personally resonated with them through strong agreement or disagreement, and preferentially shared information they anticipated would receive a positive response.
Increased recipient knowledge and improved recipient risk-related behaviour was reported by information recipients, and proxy reported by PDMOOC participants. Among information recipients who received dementia risk information (n=148), 53% reported increased knowledge, while 51% reported behaviour change.
Discussion
The PDMOOC, like other dementia risk reduction initiatives, primarily reaches people who are likely to already have high dementia risk knowledge and low levels of modifiable risk, limiting impact and equity. Information sharing transmitted dementia risk information to a larger and more diverse group of people, and impacted recipient knowledge and behaviour. However, information was preferentially shared with people similar to initiative participants (women, older people, and highly educated people; especially in highly educated networks). This was related to perceived audience receptivity, stereotypes, and dementia stigma. These three factors could be addressed in risk reduction programs to promote information sharing with broader groups of people.
Information was generally shared about a limited range of risk factors. Personal resonance and perceived network member interest impacted which risk factors were discussed. Increasing the personal resonance of under-shared risk factor information, and the social acceptability of discussing these factors, may promote the sharing of information about a wider range of risk factors. This would maximise the ability for shared information to promote dementia risk reduction amongst the people with whom it is shared.
These results suggest that further efforts are needed to promote dementia risk reduction among demographically diverse groups of people, despite information sharing beginning to achieve this goal. These results also highlight strategies for improving information delivery to promote broader information sharing, further improving initiative equity and impact.