@article{ref1,
title="The U-shaped relationship between levels of bouted activity and fall incidence in community-dwelling older adults: a prospective cohort study",
journal="Journals of gerontology. Series A: Biological sciences and medical sciences",
year="2020",
author="Lu, Zhihui and Lam, Freddy M. H. and Leung, Jason C. S. and Kwok, Timothy C. Y.",
volume="ePub",
number="ePub",
pages="ePub-ePub",
abstract="BACKGROUND: It remains uncertain whether the association between physical activity (PA) and falls is U-shaped, and few studies have explored the potential mediation of PA accumulation pattern. <br><br>METHODS: We measured PA in 671 community-dwelling older adults (82.7±3.8 years) using wrist-worn accelerometer for 7 days. PA was further classified to bouted PA (≥10 min bout length) and sporadic PA (<10 min bout length) for sub-analysis. Fall incidence in the following 12-month was recorded through tri-monthly telephone interviews. Classification and Regression Tree analysis was used to identify two optimal cutoff values of each PA measurement to predict falls. Participants were then divided into &quot;inactive&quot;, &quot;moderately active&quot; and &quot;highly active&quot; groups accordingly. Negative binomial regression models were used to estimate the association between the PA measures and fall incidence. <br><br>RESULTS: 639 participants completed 12-month follow-up. Ninety-three (14.6%) experienced a total of 118 falls. Inactive and highly active older adults had higher falls per person month relative to the moderately active group (inactive: IRR=2.372, 95%CI=1.317-4.271; highly active: IRR=2.731, 95%CI=1.196-6.232). Sub-analyses found similar significant finding with bouted PA (p<0.001) but not sporadic PA (p≥0.221). The association between bouted PA and falls remained significant even after adjusting fall incidence for bouted activity time (inactive: IRR=3.636, 95%CI=2.238-5.907; highly active: IRR=1.823, 95%CI=1.072-3.1). Further adjustments for fall-related risk factors did not meaningfully change the results. <br><br>CONCLUSION: A U-shaped relationship was identified between bouted but not sporadic PA and fall incidence. There is an approximately two-fold increase in fall rate in highly active elderly even after adjusting for activity time.<br><br>© The Author(s) 2020. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.<p /> <p>Language: en</p>",
language="en",
issn="1079-5006",
doi="10.1093/gerona/glaa058",
url="http://dx.doi.org/10.1093/gerona/glaa058"
}