Healthy Aging in Canada Vieillir en sant au Canada

1
Healthy Aging in Canada Vieillir en santé au
Canada

• Laurent Martel
• Division de la démographie
• Margot Shields
• Division de la santé

Études longitudinales et défis démographiques du
21e siècle Fédération Canadienne de
Démographie Université de Montréal, 18 19
décembre 2005
2
Research Question

• Why do some Canadians grow older in good health
  while others dont ?
• Given that a person was in good health in 1994,
  what were the predictors associated with staying
  healthy over the next 8 years ?
• Middle-age adults (45-64 years)
• Older adults (65 years and over)
• Factors investigated
• Demographic (age, sex, living arrangements,
  rural/urban)
• Socio-economic (education)
• Health behaviours (smoking, leisure-time physical
  activities, alcohol consumption, BMI)
• Psycho-social (sense of coherence, social
  support, financial stress, family stress)
• Chronic conditions

3
What is healthy aging ?

• No disabilities or fully corrected disability
• No dependency
• Good, very good or excellent self-perceived
  general health (compared to fair or poor)
• No depressive symptoms (as proxy for mental
  health)
• Still alive!
• Still living in a household (i.e., not living in
  an institution)

4
Data

• National Population Health Survey Longitudinal
  file with 5 cycles (1994-95 to 2002-03)
• Each respondent interviewed once every two years
• No new respondent added after 1994-95 (panel)
• Household and institutional residents covered in
  all provinces and territories
• 17 276 respondents aged 12 years and over in
  households
• 2 118 respondents aged 12 years and over in
  institutions
• Response rate (households) 92,8 for cycle 2,
  88,2 for cycle 3, 84,8 for cycle 4 and 80,6
  for cycle 5.

5
Sample

• Household population in provinces only
• 3 887 respondents aged 45-64 years
• 150 excluded because of non-response (4)
• final sample 3 737 respondents
• population under study 2 496 in good health in
  1994-95 (70)
• 141 respondents by proxy (5)
• 2 740 respondents aged 65 years and over
• 49 excluded because of non-response (2)
• final sample 2 691 respondents
• population under study 1 309 in good health in
  1994-95 (51)
• 112 respondents by proxy (6)

6
How many seniors were in good health in 1994 ?






Significantly lower than previous age group
(plt0,05)
7
How many males and females were in good health in
1994 ?


Significantly lower than males (plt0,05)
8
Percentage of people maintaining good health for
2, 4, 6 and 8 years










Significantly lower than previous age groups
for the same duration (plt0,05)
9
Percentage of people maintaining good health for
2, 4, 6 and 8 years
65 years and over
45-64
No significant differences between sexes (plt0,05)
10
Analytical method

• Cox proportional hazards model to estimate the
  hazards of staying healthy over the 8 year period
• Non-response
• at only one cycle if respondents were in good
  health in preceding and subsequent cycles, we
  imputed good health
• at two cycles or more censored
• Problem lots of ties in failure time because
  the survey is not fully longitudinal but rather
  repeated measures
• we cant use the survey weight with PhReg in SAS
• Solution use a logistic procedure with a
  cloglog link function
• exactly same results as PhReg
• requires to explode the data file
• One line per respondent to one line per event
  (event occur in the middle of the two- year
  interval between cycles)

11
Results demographic, socio-economic factors and
chronic conditions
Reference category
12
Results Health behaviours and psycho-social
factors
Reference category
13
The more, the better





Significantly higher than previous group
(plt0.05)
Source 2003 Canadian Community Health Survey
14
The more, the better
Adjusted for demographic, socio-economic and
chronic conditions
Reference category
Significantly different from the reference
category (plt0.05)
Source 2003 Canadian Community Health Survey
15
Summary

• Some healthy behaviours not significantly related
  to maintaining good health among the middle aged
  adults. Negative consequences not yet
  measurable, but eventually catch up later in
  life. Stress appears more important among this
  age group.
• All healthy behaviours significantly related to
  maintaining good health through senior years.
• Cumulative effect of healthy behaviours suggests
  that promoting them might be a way to increase
  healthy and successful aging.
• Some variables have been omitted
• family medical history
• nutrition

16
How to find this article ?

• Healthy aging among the elderly based on
  dependency
• Report on the demographic situation 2002,
  Demography Division
• Healthy aging among seniors (with 2003 CCHS data)
• to be released in January 2006 in the Annual
  Report on the Health of Canadians watch
  Statistics Canada web site