![[ Visit Client Website ]](images/banner.gif)
Tuesday, October 21, 2008
Columbus A-C (Hyatt Regency Penns Landing)
Purpose: We used growth-curve models to investigate health-related quality of life (HRQL) trajectories among the general adult population and to determine the association between the individual characteristics and variations among trajectories.
Methods: Longitudinal Canadian National Population Health Survey (Cycles 2 (1996/97) to 6 (2004/05)) was used. The target population was those aged 18 years and older in Cycle 2, including those who subsequently were institutionalized and/or died. We used information for 13,665 respondents with 53,151 records in total. Health Utilities Index Mark 3 (HUI3) was used as the HRQL measure. Separate HRQL trajectories were estimated for young (age 18-39), middle-aged (40-64) and seniors (65+). Socio-demographic and lifestyle factors were included in the model. Dummy variables indicating institutionalization, death and cohort membership were also included.
Results: Unconditional models for the three age groups were fitted separately. Significant cohort effects were observed in the older age groups. A typical life course trajectory was estimated as concave with a HUI3 score of around 0.95 at age 18 with a very slow decline until the age of 60 (HUI3 around 0.80), followed by a rapid decline. At the age of around 90, the predicted HUI3 was as low as 0.30. Results from independent conditional models showed that factors associated with trajectories differed substantially between the age groups. Receiving social assistance, not having a high school diploma and not being married had negative impacts on trajectories for young and middle-aged. These factors were unimportant for seniors. However, unfavorable lifestyle factors (i.e. abstaining from alcohol, smoking and physical inactivity) had important negative effects for seniors. In particular, the mean decrement in the trajectory when one became inactive was 0.05 for seniors, more than twice as great as was found for the young (0.02) and middle-aged (0.02) (p<0.01 for all parameters). In contrast, having more chronic conditions lowered the mean trajectories by similar magnitudes regardless of age.
Conclusions: We found important heterogeneities in life-course trajectories. Understanding differential impacts of the determinants of health factors on trajectories is important in developing effective health policy for various life stages.
Methods: Longitudinal Canadian National Population Health Survey (Cycles 2 (1996/97) to 6 (2004/05)) was used. The target population was those aged 18 years and older in Cycle 2, including those who subsequently were institutionalized and/or died. We used information for 13,665 respondents with 53,151 records in total. Health Utilities Index Mark 3 (HUI3) was used as the HRQL measure. Separate HRQL trajectories were estimated for young (age 18-39), middle-aged (40-64) and seniors (65+). Socio-demographic and lifestyle factors were included in the model. Dummy variables indicating institutionalization, death and cohort membership were also included.
Results: Unconditional models for the three age groups were fitted separately. Significant cohort effects were observed in the older age groups. A typical life course trajectory was estimated as concave with a HUI3 score of around 0.95 at age 18 with a very slow decline until the age of 60 (HUI3 around 0.80), followed by a rapid decline. At the age of around 90, the predicted HUI3 was as low as 0.30. Results from independent conditional models showed that factors associated with trajectories differed substantially between the age groups. Receiving social assistance, not having a high school diploma and not being married had negative impacts on trajectories for young and middle-aged. These factors were unimportant for seniors. However, unfavorable lifestyle factors (i.e. abstaining from alcohol, smoking and physical inactivity) had important negative effects for seniors. In particular, the mean decrement in the trajectory when one became inactive was 0.05 for seniors, more than twice as great as was found for the young (0.02) and middle-aged (0.02) (p<0.01 for all parameters). In contrast, having more chronic conditions lowered the mean trajectories by similar magnitudes regardless of age.
Conclusions: We found important heterogeneities in life-course trajectories. Understanding differential impacts of the determinants of health factors on trajectories is important in developing effective health policy for various life stages.
See more of: Poster Session III
See more of: 30th Annual Meeting of the Society for Medical Decision Making (October 19-22, 2008)