Spousal Loss and Physical Activity Trajectories Among Older Adults

Thursday, April 2, 2009
Exhibit Hall RC Poster Sessions (Tampa Convention Center)
Kin-Kit Li, The University of Hong Kong, Hong Kong, Hong Kong and Bradley J. Cardinal, Oregon State University, Corvallis, OR
Purpose

The principles of life-span development (i.e., impact of early development on later development) and timing of event (i.e., differential effects of an event according to its temporal appearance in the life course) in the life course perspective were applied to examine primarily 1) whether pre-spousal loss physical activity (PA) trajectories explained post-spousal loss PA trajectories, and 2) whether advanced age predicted less favorable post-spousal loss PA trajectories. A secondary purpose was to explore the multiple influences of contact with children, spousal care, and depression on post-spousal loss PA trajectories. Four PA trajectories were created including stable active, activity adopters, activity relapsers, and stable sedentary, which were developed from previously identified transitional shift patterns of physical activity (Levy & Cardinal, 2006).

Methods

The sample included 930 widowed participants (Mage=86.16, SDage=6.35) from the Australian Longitudinal Study of Aging. The analysis time metric was centered at the time of spousal loss and multiple imputations were used for the resulting missing values. Both pre- and post-spousal loss PA trajectories captured three years of duration with four measurement points. Sequential Latent Class Growth Analysis was used to extract the membership of the four PA trajectories for both pre- and post-spousal loss and to examine the hypothesized relationships. The membership of pre-spousal loss PA trajectories, age, contact with children, depression, spousal care, and other covariates were used to predict the membership of the post-spousal loss PA trajectories.

Analysis/Results

Results of the multinomial regression, as a part of the LCGA, showed that post-spousal loss PA trajectories were predicted by pre-spousal loss PA trajectories (Odds Ratios [OR]=7.90-16.59, ps<.01) and spousal loss (ORs=4.10-4.63, ps<.05). No significant effect of age was found. The findings suggest that people who decrease their PA levels before spousal loss might have a higher chance to increase their activities rather than to remain sedentary after spousal loss, and people who provide spousal care before spousal loss might be more likely to increase their activity levels after spousal loss.

Conclusions

Early intervention aimed at developing habitual PA may prevent the less favorable PA trajectories after spousal loss. More attention should be given to the primary care-givers of their spouses for physical activity promotion. For future studies, the social and environmental factors that contribute to changes in the PA trajectories in the transition of spousal loss should continue to be explored. Other life-course events and transitions should also receive more attention for PA promotion interventions.