3185 — Trajectories in Physical Activity and Inactivity among Women Veterans in the Women's Health Initiative

Washington DL, VA Greater Los Angeles Healthcare System, VA HSR&D Center of Innovation; University of California, Los Angeles School of Medicine; Gray K, VA Puget Sound Health Care System; University of Washington School of Public Health; Katon JG, VA Puget Sound Health Care System; VA Office of Patient Care, Women's Health Services; University of Washington School of Public Health; Cochrane BB, University of Washington School of Nursing; Fred Hutchinson Cancer Research Center; Hoerster KD, VA Puget Sound Healthcare System, Seattle Division; Bastian L, VA Connecticut Healthcare System; Weitlauf JC, VA Palo Alto Health Care System, Sierra Pacific MIRECC and Center for Innovation to Implementation; Bean-Mayberry B, VA Greater Los Angeles Healthcare System, VA HSR&D Center of Innovation; University of California, Los Angeles School of Medicine; Richardson CR, Ann Arbor VA Medical Center; Tinker L, Fred Hutchinson Cancer Research Center; University of Washington School of Nursing

Objectives:
Individuals entering the military are more physically fit than similarly aged civilians ("healthy warrior effect"). However, trajectories of physical activity after military separation have not been described or compared to non-Veterans. Understanding these trajectories is particularly important for women Veterans (WVs) given their high prevalence of cardiovascular risk factors. Our objective was to compare trajectories of physical activity and sedentary behavior between Veteran and non-Veteran women.

Methods:
We included 3,240 WVs and 123,423 non-Veteran postmenopausal Women's Health Initiative participants aged 50-79 years at baseline. Self-reported physical activity, converted to MET-hours/week was assessed over 8 years in both the Clinical Trials and Observational Study participants. Self-reported sedentary behavior (hours sitting or lying down/day) was collected from Observational Study participants at baseline and years 3 and 6. We examined differences between Veterans and non-Veterans at baseline. We also assessed adjusted trajectories using a general estimating equations approach to linear regression, including an interaction term between visit and Veteran status to examine differences in trajectories.

Results:
At baseline WVs had higher physical activity than non-Veterans (13.1 vs. 12.5 MET hours/week; difference 0.70; 95% confidence interval [CI] 0.24, 1.16). Physical activity declined over time in Veterans (adjusted change per visit year -0.19 MET-hours/week; 95% CI -0.12, -0.27) and non-Veterans (-0.03 MET-hours/week; 95% CI -0.005, -0.03), but the decline was greater in Veterans (p < 0.001 for interaction). At baseline WVs and non-Veterans had similar levels of sedentary behavior (15.3 vs. 15.1 sedentary hours/day; difference 0.15; 95% CI -0.02, 0.33). Trajectories of sedentary behavior differed between Veterans and non-Veterans (p = 0.01 for interaction) with sedentary behavior remaining stable among Veterans (adjusted change per visit year -0.03 hours/day; 95% CI -0.06, 0.004), but slightly declining among non-Veterans (-0.07 hours/day; 95% CI -0.07, -0.08).

Implications:
WVs had healthier baseline physical activity and sedentary behavior compared to non-Veterans. However, over time WVs had more adverse trajectories than non-Veterans.

Impacts:
The less favorable physical activity trajectories we found among WVs may contribute to their excess mortality later in life. Physical activity should be monitored and promoted in WVs. Research should be directed toward identifying predictors of activity trajectories and facilitators of physical activity maintenance.