Effects of a Long-Term Physical Activity Program on Activity Patterns in Older Adults

Purpose To examine the effect of a long-term structured physical activity (PA) intervention on accelerometer-derived metrics of activity pattern changes in mobility-impaired older adults. Methods Participants were randomized to either a PA or health education (HE) program. The PA intervention included a walking regimen with strength, flexibility, and balance training. The HE program featured health-related discussions and a brief upper body stretching routine. Participants (n = 1341) wore a hip-worn accelerometer for ≥10 h·d^-1 for ≥3 d at baseline and again at 6, 12, and 24 months postrandomization. Total PA (TPA) - defined as movements registering 100+ counts per minute - was segmented into the following intensities: low-light PA (LLPA; 100-759 counts per minute), high light PA (HLPA; 760-1040 counts per minute), low moderate PA (LMPA; 1041-2019 counts per minute), and high moderate and greater PA (HMPA; 2020+ counts per minute). Patterns of activity were characterized as bouts (defined as the consecutive minutes within an intensity). Results Across groups, TPA decreased an average of 74 min·wk^-1 annually. The PA intervention attenuated this effect (PA = -68 vs HE: -112 min·wk^-1, P = 0.002). This attenuation shifted TPA composition by increasing time in LLPA (10+ bouts increased 6 min·wk^-1), HLPA (1+, 2+, 5+, and 10+ bouts increased 6, 3, 2, and 1 min·wk^-1, respectively), LMPA (1+, 2+, 5+, and 10+ bouts increased: 19, 17,16, and 8 min·wk^-1, respectively), and HMPA (1+, 2+, 5+, and 10+ bouts increased 23, 21, 17, and 14 min·wk^-1, respectively). Conclusions The PA intervention increased PA by shifting the composition of activity toward higher-intensity activity in longer-duration bouts. However, a long-term structured PA intervention did not completely eliminate overall declines in total daily activity experienced by mobility-impaired older adults.