Association Between Physical Activity Levels in the Hospital Setting and Hospital-Acquired Functional Decline in Elderly Patients

IMPORTANCE: The effects of in-hospital physical activity (PA) on outcomes among elderly patients has seldom been assessed. OBJECTIVES: To assess PA levels among elderly patients hospitalized for acute medical illness and to examine the association between PA levels and functional decline and other clinical outcomes at discharge. DESIGN, SETTING, AND PARTICIPANTS: This monocentric cohort study was performed among patients 65 years or older who were admitted for acute medical illness to the internal medicine ward of Lausanne University Hospital, Lausanne, Switzerland, from February 1 through November 30, 2018. Data were analyzed from January 1 through December 2, 2019. EXPOSURES: Daytime and 24-hour PA levels assessed via wrist accelerometers and measured in millig units (mG; 1 mG = 9.80665 × 10(−3) m/s(2)). MEAN OUTCOMES AND MEASURES: Functional decline (defined as a ≥5-point decrease in the modified Barthel Index), risk of bedsores, length of stay (LOS), and inability to return home. RESULTS: A total of 177 patients (106 [59.9%] men; median age, 83 [interquartile range, 74-87] years) were included. Lower mean (SD) PA levels were found in patients using walking aids before admission (daytime, 12 [5] vs 15 [7] mG; 24-hour, 10 [3] vs 11 [5] mG), those admitted for a reason associated with functional decline (daytime, 12 [6] vs 14 [7] mG; 24-hour, 10 [4] vs 11 [4] mG), or those prescribed physiotherapy (daytime, 12 [5] vs 15 [7] mG; 24-hour, 10 [4] vs 12 [5] mG). At discharge, functional decline was found in 63 patients (35.6%; 95% CI, 25.6%-43.1%), bedsore risk in 78 (44.1%; 95% CI, 36.6%-51.7%), and inability to return home in 82 (46.3%; 95% CI, 38.8%-54.0%). After multivariate analysis, no association was found between PA levels and functional decline (multivariable-adjusted mean [SE], 13 [1] vs 13 [1] mG for daytime levels [P = .69] and 10 [1] vs 11 [1] mG for 24-hour PA levels [P = .45]) or LOS (Spearman rank correlation, ρ = −0.06 for daytime PA levels [P = .93] and −0.01 for 24-hour PA levels [P = .52]). Patients at risk of bedsores had significantly lower PA levels than those not at risk (multivariable-adjusted mean [SE], 12 [1] vs 15 [1] mG for daytime PA levels [P = .008]; 10 [1] vs 12 [1] mG for 24-hour PA levels [P = .01]). Patients able to return home had significantly higher PA levels than those institutionalized (multivariable-adjusted mean [SE], 14 [1] vs 12 [1] mG for daytime PA levels [P = .04]; 11 [1] vs 10 [1] mG for 24-hour PA levels [P = .009]). CONCLUSIONS AND RELEVANCE: In this study, lower in-hospital daytime and 24-hour PA levels were associated with risk of bedsores and inability to return home on discharge. These findings are important given that one-third of elderly patients present with hospital-acquired functional decline.