Mapping Community-Level Prevalence of Modifiable Risk Factors for Dementia in Canada

A large proportion of dementia risk is attributable to modifiable factors such as physical inactivity, hypertension, and social isolation. Prevention strategies will be essential to mitigate the expected increased number of people living with dementia. Data on the distribution of risk factors can help support these efforts.
The objective of this study was to derive community-level prevalence estimates for dementia specific modifiable risk factors.
Statistics Canada Canadian Community Health Survey (CCHS; 2001-2020) data were used to develop prediction models for several mid-life (age 45-64; heavy drinking, obesity, hypertension) and late-life (age 65+; smoking, physical inactivity, social isolation, diabetes) risk factors. Prevalence was estimated from the prediction model using age and sex stratified Census (2001-2016) population counts for communities across New Brunswick. Spatial-temporal models were used to increase the robustness of predicted prevalence estimates.
The risk factors with the highest prevalence were physical inactivity (67%), obesity (34%), and hypertension (31%). These three risk factors, in addition to risk factors for social isolation and smoking, were also found to have highest variability across communities. The prevalence of obesity, hypertension and diabetes increased over time, whereas smoking and social isolation remained consistent. While physical inactivity had the highest prevalence, this was found to decrease over time.
National population-based survey and Census data can be used to inform of the burden of dementia risk factors at the community-level. Community-level risk factor data may be helpful in directing resources to communities with the highest burden and to monitor changes in risk for these communities.

1. Department of Health
Knowledge Translation Event
May, 2024
Sandra Magalhaes
1,2
, Paramdeep Singh
1
, Simon
Youssef
1
, Samuel Cookson
1
, Pamela Jarrett
3,4
,
Andrew Sexton
5
, Karla Faig
6
, Chris McGibbon
5,7
1 New Brunswick Institute for Research, Data and Training (NB-IRDT),
University of New Brunswick, Fredericton, Canada. 2 Department of
Sociology, University of New Brunswick, Fredericton, Canada. 3 Horizon
Health Network, Saint John, Canada. 4 Faculty of Medicine, Dalhousie
University, Saint John, Canada. 5 Institute of Biomedical Engineering,
University of New Brunswick, Fredericton, Canada. 6 Horizon Health
Network, Fredericton, Canada. 7 Faculty of Kinesiology, University of New
Brunswick, Fredericton, Canada
MAPPING COMMUNITY-LEVEL
PREVALENCE OF
MODIFIABLE RISK FACTORS
FOR DEMENTIA IN CANADA

2. BACKGROUND
•Prevalence of dementia will increase due to aging populations
•Numbers of older adults living with dementia expected to triple by 2050
•Caring for individuals with dementia requires substantial resources
•Society- wide attention to prevention is needed
•Leading dementia researchers suggest a third of cases could be prevented
•Mid-and late- life risk factors are common to other chronic diseases

3. DISTRIBUTION OF DEMENTIA OVER TIME IN NB
6
6.2
6.4
6.6
6.8
7
2004-2008 2009-2013 2014-2018
Prevalence per 100
9.0
9.5
10.0
10.5
11.0
11.5
12.0
2004-2008 2009-2013 2014-2018
Incidence per 1000

4. REGIONAL DISTRIBUTION OF DEMENTIA IN NB

5. NB-PALM
Objective #1
Engagement with NB Seniors
Project team: Bryn Robison (HHN), Linda Yetman(HHN),
Stephanie Crapoulet(Vitalite), Vanshika Khaitan(CCNA
Trainee, McMaster University)
Seniors Engagement Strategy
Survey (N=245, 20% French,
80% English)
Development of a KT toolkit
Objective #2
Programs and Interventions
Project team: Chris McGibbon (UNB), Pam Jarrett (HHN,
DalMed-NB), Grant Handrigan(UdeM), Carole Tranchant
(UdeM), Danielle Bouchard (UNB), Karla Faig(HHN), Alana
Gullison(UNB), Molly Gallibios (UNB), Jose Hache (Vitalite)
Synergic@Home:
Feasibility trial, 4mo 2x2 RCT with 10mo f/u
(N=60 in NB)
Brain Health Support Program:
3mo pilot trial (N=8 in NB); 12mo full study
(N=60 in NB)
CognicitiFrancophone Validation:
(N=200 in NB)
Objective #3
Evaluation and Monitoring
Project team: Sandra Magalhaes(UNB, NB-IRDT), Pam
Jarrett (HHN, DalMed-NB), Chris McGibbon (UNB)
Evaluation & Monitoring Strategy
Development of a community-level
dementia risk model

6. RESEARCH OBJECTIVES
•To support scaling and sustainability planning for brain health interventions
•Where are interventions most needed and resources best directed?
•To better understand the burden of dementia risk factors
•Where is risk of dementia greatest and how is it changing over time?

7. METHODS
Study Design
•Secondary analysis of Statistics Canada survey and Census data
•Canadian Community Health Survey (CCHS) 2000- 2020
•Canadian Census 2001, 2006, 2011 and 2016
•Prevalence of established modifiable risk factors were mapped to 242 Statistics
Canada Census Sub- Divisions (CSDs) in New Brunswick
•Mid-life (45- 64 yrs.): heavy drinking, obesity, hypertension
•Late-life (65+ yrs.): smoking, physical inactivity, social isolation, diabetes

8. METHODS
Risk Factor Definitions
Risk factor Risk period Risk definition
Heavy Drinking Mid-life Drink 4-6 times per week or everyday
Obesity Mid-life Current BMI ≥ 30
Hypertension Mid-life Diagnosed with high blood pressure
Smoking Late-life Currently smoking
Physical inactivityLate-life Sedentary/inactive on a weekly basis
Social isolationLate-life (i) Somewhat or very weak sense of belonging in local community
(ii) Single/never married marital status
Diabetes Late-life Diagnosed with type 1 or 2 diabetes

9. METHODS
Risk Factor Analysis
•Small area risk factor prevalence estimates were derived for four time periods
(2000- 2004, 2005- 2009, 2010- 2014, 2015- 2020)
•CCHS data were used to develop a prediction model for each risk factor
considering age, sex, socioeconomic status and census sub- division (CSD)
•Census data for middle- and older- aged population size within each CSD
were used as inputs in prediction models to estimate risk factor prevalence
•Spatial-temporal models were used to adjust for spatial correlation between
communities and to increase robustness of prevalence estimates

10. METHODS
DementiaVulnerability Index for Communities (D-VIC)
•DIV-C is a measure of vulnerability that combines all 8 risk factors
•Provides an opportunity to identify which communities are at increased
vulnerability to dementia
•Higher score indicates community has among the highest risk factor burden
•Each risk factor ranked to identify those in the top 20% of all communities
•Ranges from 0- 8 to indicate sum of the number of top ranked risk factors

11. RESULTS
Health zones and y-axis

12. MID-LIFE RISK FACTOR PREVALENCE
Prevalence in 2016 across NB communities
•Prevalence of obesity
and hypertension were
similar
•Prevalence of heavy
drinking is much lower
•Heterogeneity across
communities is observed

13. MID-LIFE RISK FACTOR PREVALENCE
Identifying Outlying Communities (2016)
•Substantial overlap in the
distribution of risk factor
across health zones
•More outliers identified
for communities with
higher prevalence than
lower prevalence
•More outliers identified
for heavy drinking

14. LATE-LIFE RISK FACTOR PREVALENCE
Prevalence in 2016 across NB communities
•Prevalence of physical
inactivity was
substantially higher than
other risk factors
•Prevalence of diabetes,
smoking and social
isolation in similar range
•Single marital status was
least prevalent

15. LATE-LIFE RISK FACTOR PREVALENCE
Identifying Outlying Communities (2016)
•Substantial overlap in the
distribution of risk factor
across health zones
•More outliers identified
for communities with
higher prevalence than
lower prevalence, except
for physical inactivity
•Most outliers identified for
diabetes and smoking

16. CHANGES OVER TIME
•Slight increase in prevalence
•Shift in outliers from lower to
higher prevalence
Diabetes
Obesity
Hypertension

17. MORE VARIABILITY AND MORE OUTLIERS
Changes over time
•Heavy drinking and smoking are becoming more variable and more
communities with high prevalence are outliers
Heavy Drinking Current Smoking

18. LESS OUTLIERS
Changes over time
•More variability over time and less communities with lower prevalence were
outliers for physical inactivity
Physical Inactivity

19. DEMENTIA VULNERABILITY INDEX FOR COMMUNITIES (D-VIC)
Combining Risk Factors
•Maximum score was 6
•Substantial overlap in the
distribution of D- VIC
across health zones
•Number of communities
with higher scores is
increasing over time
•Health zones 3, 5 and 6
had highest scores in 2016

20. OVERVIEW OF FINDINGS
•Existing data can be leveraged to inform on burden of dementia risk at the
community-level
•Outlying communities represent those that may be most vulnerable
•Physical inactivity was most prevalent and outliers with lower prevalence
•Outliers for other risk factors were more common for higher prevalence
•Increase in prevalence over time, with greater variability and more outliers
•Possibly suggesting future prevalence of dementia may be greater than
expected using current projections
•No community was ranked in top for all eight risk factors
•Number of communities with highest scores are increasing over time

21. STRENGTHS AND LIMITATIONS
Strengths
•Population based data sources
•Complex statistical modelling
•Focus on established modifiable
risk factors
•Prevalence in at- risk population for
small communities
Limitations
•Model based estimates
•Self-reported behavioural data
•Not all risk factors were possible to
quantify

22. NEXT STEPS
•Derive prevalence estimates for additional established dementia risk factors
•Depression, air pollution, traumatic brain injury
•Validate D-VIC using dementia cases in communities across NB
•Expand risk factor and D-VIC estimates to Canada-wide
•Disseminate research findings to relevant stakeholders

23. ACKNOWLEDGEMENT
Body Level One
Body Level Two
Body Level Three
Body Level Four
Body Level Five