Physical activity of baby boomers compared to older and younger generations in Canada: An age-period-cohort analysis Mayilee Canizares, Elizabeth Badley, Monique Gignac, Sheilah Hogg-Johnson, and Anthony Perruccio University Health Network, Toronto, Canada, University of Toronto, Toronto, Canada, Institute for Work and Health, Toronto, Canada Canadian Public Health Association May 2014
Background Physical activity has well-documented health benefits. Studies using repeated cross-sectional surveys indicate that participation in leisure time physical activities (LTPA) has increased over time. Also being sedentary in usual activities has increased over time. Cross-sectional studies suggest that participation in LTPA decline with age and is also lower in younger birth cohorts. It is not well documented if these changes in LTPA over time differ by birth cohorts.
To examine birth cohort and/or period effects (secular changes) in the age-trajectories of LTPA over 16 years among Canadian adults. To examine whether the age-trajectories of LTPA were associated with changes in usual activities after controlling for sex, educational attainment, income, and BMI. Objectives
What are age, period, and cohort effects? Age effects are the consequences of growing older. How things change with age irrespective of birth cohort and calendar time Period effects are the consequences of influences that vary through time. Changes that affect the entire population irrespective of age and birth cohort (e.g. new treatments, environmental disasters, changes in policies, the economy etc.) Cohort effects are the consequences of being born and growing up at different times. Changes according to year of birth, irrespective of age and calendar time Blackwell encyclopedia of sociology, Ed, Ritzer G. 2007
Data Sources Canadian National Population Health Survey (NPHS) National representative longitudinal household survey Biennial interviews from 1994/95 to 2010/11 9 cycles spanning 16 years Sample of 8,570 participants in one of the birth cohorts
1. Usual activities ‘Thinking back over the past 3 months, which of the following best describes your usual activities or work habits?’ ‘ Sit during the day and don’t walk much’, ‘Stand or walk quite a lot’, ‘Lift/carry light loads’, ‘Heavy work or carry heavy loads’ 2. Leisure time physical activity(LTPA): at least moderate Moderate active: >1.5 kcal/kg/day 1 3. Walking/cycling for commuting and errands (WCCE): at least moderate ‘In a typical week in the past 3 months, how many hours did you usually spend walking (cycling) to work or school or while doing errands?’ 1 Katzmarzyk and Tremblay (2007) Dimensions of physical activity
Statistical Analysis Multilevel growth models Age-trajectory of LTPA by cohort: o Not including period o Including period Added usual activities to the APC model of LTPA previously fitted, controlling for sex, education, income, and BMI
Results
Leisure time physical activity
Predicted age- trajectory by cohort LTPA: age trajectories by cohort Cohort effect: p< Age effect: p< Period effect: -
Predicted age- trajectory by cohort holding period constant Predicted age- trajectory by cohort LTPA: age trajectories by cohort Cohort effect: p=0.015 Age effect: p< Period effect: p<0.0001
Usual activities
Leisure time physical activity : effect of changes in usual activities Odd Ratio (95% CI) Usual activities: sedentary 0.73 ** (0.71 – 0.74) **Significant at p< Models included age terms, cohort, period. a Also included sex, education, income and BMI Further analysis show a significant interaction between usual activities and sex
GenXYBB OBBWWII Leisure time physical activity : by categories of usual activities by sex Solid/Dashed: Not Sedentary/ Sedentary in usual day Dark/Light: Women/Men
Physical activity measured in the NPHS Self-reported nature of the data Attrition due to drop-outs and mortality, particularly in the older cohort Limitations
In all cohorts, individuals are increasingly participating in LTPA as they grow older and this growth is due to period effect o health promotion initiatives have been somewhat successful Members of the younger cohorts are more likely to be physically active during leisure time than their older counterparts Generally, those who are sedentary in their usual activities are less likely to participate in LTPA o particularly women in all cohorts and men in the two oldest cohorts The results indicate that efforts to address LTPA also need to take into account usual activities Summary and Conclusions
Canadian Institutes of Health Research Operating Grant - Secondary Analysis of Databases ‘In sickness or in health? How well are the baby boomers aging compared to older generations: an analysis of age, period and cohort effects.’ P.I. Elizabeth Badley Acknowledgements
Predicted age- trajectory by cohort WCE: age trajectories by cohort
Predicted age- trajectory by cohort holding period constant Predicted age- trajectory by cohort
Usually sedentary: age trajectories by cohort
Predicted age- trajectory by cohort holding period constant Predicted age- trajectory by cohort Usually sedentary: age trajectories by cohort