Alexandre Lebel (CRAD) The Geography of Overweight in Québec Robert Pampalon (INSPQ) Denis Hamel (INSPQ) Marius Thériault (CRAD) Halifax June 2 nd, 2008.

Alexandre Lebel (CRAD)

The Geography of Overweight in Québec

Robert Pampalon (INSPQ)

Denis Hamel (INSPQ)

Marius Thériault (CRAD)

Halifax

June 2nd, 2008

CPHA 2008 Annual Conference

Background

Objective et methodology

Results and analyses

Presentation plan

•Why obesity is a public health concern?

•Why geography should be considered?

•Why obesity is a public health concern?

•Why geography should be considered?

•Main concepts

•Multilevel analysis

•Main concepts

•Multilevel analysis

Linked to a large number of chronic diseases•Coronary artery disease

•Stroke

•Breast cancer

•Colorectal cancer

•Hyperlipidemia

•Ostéo-arthrite

•Endometrial cancer

•Type 2 diabetes

•Pulmonary embolism

•Hypertension

•Gallbladder disease

20%Are directly

attribuable to obesity

Birmingham et al. (1999)

Life expectancy going down ?

Background:

1) Why obesity is a public health concern?

Main associations with overweight: individual characteristics

•Sex

•Age

•Marital status

•Income

•Education

•Race

•Other genetic factors

•Smoking

•Diet

•Physical activity

Lots of associations

Many interventions

Few results

2) Why geography is important?

Contextual Contextual characteristicscharacteristics

•Urban design

•Land use

•Transport system

•Rural

•Security

•Socioeconomic characteristics

Few associations!

Inconsistent resultsDifferent methodologiesHard to compareGeography given low importance

(Region definition, scale)

0

5

10

15

20

25

30

1973 1983 1993 2003

% o

bé

sit

é

Québec

Canada

Angleterre

É-U

Chine

Australie

Brésil

Mexique

France

Sources : diverses

Geographic differences in obesity prevalence

In the WorldIn the World

In CanadaIn Canada

Source : CCHS, 2003 (in Mongeau, 2005)

Obesity prevalence by province

In QuébecIn Québec

Source : CCHS, 2003 (in Mongeau, 2005)

Overweight prevalance by health region

Québec 47 %

General objective:

To explore the spatial distribution of overweight in the Québec To explore the spatial distribution of overweight in the Québec population while taking into consideration individuals’ population while taking into consideration individuals’ characteristics, contextual indicators and an appropriate spatial characteristics, contextual indicators and an appropriate spatial frame of reference.frame of reference.

Objective and Methodology

Spatial frame of referenceSpatial frame of reference

15 Health Regions

Periphery b

Town (CA)d

Rural(MIZ)

CMA ofMontréal orQuébec City

Proximale

Remotef

City (CMA)c

Center a

Other urbanareas

Privileged or deprived 51 units

Multilevel AnalysisMultilevel Analysis

SampleSample

• CCHS 2003 (cycle 2.1)

• 20 000 Adults Living in Québec

•Dependant variable: BMI ≥ 25

• No pregnant women

• No individuals with BMI < 18,5

•Self-reported information

Results Results MenMen51 regions51 regions9 522 adults9 522 adults

Men

Rank Region1 Montréal Center Privileged (a)2 Québec Rural Remote (f)3 Laval Periphery (b)4 Québec Rural Proximal (e)6 Appalachians Towns (d)8 Estrie Center (c)

40 Montérégie Towns (c)43 Lanaudière Towns (d)44 Québec Center Deprived (a)45 Gatineau City (c)47 Gatineau Rural Proximal (e)48 Abitibi Towns (d)49 Gaspésie Rural Remote (f)50 Côte-Nord Rural Proximal (e)51 Côte-Nord Towns (d)

ResultsResultsWomenWomen51 regions51 regions10 927 adults10 927 adults

Women

Rank Region1 Montréal Centre Privileged (a)2 Saguenay Towns (d)3 Québec Rural Proximal (e)4 Québec Center Privileged (a)5 Québec Rural Remote (f)6 Saguenay City (c)8 Estrie City (c)

13 Québec Periphery Privileged (b)40 Laurentides Towns (d)45 Montréal Periphery Privileged (b)46 Outaouais City (c)47 Mauricie Towns (d)48 Gaspésie Rural Remote (f)49 Mauricie Rural Remote (f)50 Outaouais Rural Remote (f)51 Outaouais Rural Proximal (e)

As a summary, we found that:As a summary, we found that:

Independent of individual characteristics known to be associated with overweight, many regions present significantly higher or lower odds of overweight.

These differences are not randomly distributed through space.

The geographical patterns are different between genders.

There is an important variability within rural as well as urban areas.

The contextual effect is stronger for women.

Consequently: • Geography is important

• Studying at local scale

Financial support

And other partners

ÉSAD

We hope to publish details of this research in 2008

in The Canadian Journal of Public Health