F62 Validation of digital assessment of physical activity in Huntington's disease: comparing fitbit charge 4 step count with research-grade accelerometers. (12th September 2022)

• Record Type: Journal Article
• Authors: Mann, Elliot; Drew, Cheney JG; Busse, Monica; Doheny, Emer; Lowery, Madeleine
• Is Part Of: Journal of neurology, neurosurgery and psychiatry. Volume 93(2022)Supplement 1
• Page Start: A58
• Page End: A59
• Publication Date: 2022-09-12
• Subjects: Fitbit Charge 4 -- Step count -- Validity -- Reliability -- Digital assessment; Neurology -- Periodicals; Nervous system -- Surgery -- Periodicals; Psychiatry -- Periodicals; 616.8
• DOI: 10.1136/jnnp-2022-ehdn.153
• Languages: English
• ISSNs: 0022-3050

Abstract

Abstract : Background: Physical activity has been implicated in improving symptom management and quality of life in Huntington's disease (HD). Verifying the role of physical activity in HD requires accurate quantification of exercise metrics. Aim: To assess the reliability and validity of Fitbit® Charge 4 step-count compared to research accelerometers and observer count in HD patients. Methods: 17 manifest HD participants completed two indoor 2-minute walking tests (2MWT) while wearing a Fitbit® charge 4, and an ActiGraph GTX9 on their non-dominant wrist, ActivPAL4™ on both anterior thighs, and ActiGraph GTX9 on both anterior shanks. Steps were manually counted from video recordings of the 2MWTs. Step-counts for devices were obtained from their proprietary software. Intraclass correlation coefficients (ICC) determined Fitbit and observer reliability, whilst Bland-Altman (B-A) analysis demonstrated monitor agreement. Results: Intra-rater reliability of researcher count from repeat trials was excellent (ICC 0.95/n = 17) meanwhile, Fitbit yielded good reliability (ICC 0.81/n = 16). B-A plots displayed greatest agreement between Fitbit and manual count (bias -4.7/SD 36.2). Average monitor count revealed largest consistencies with ActivPAL4 and Fitbit (bias -3.4/SD 37.4) whereas, wrist-worn ActiGraph deviated markedly (bias 4.9/SD 66.9). Motor symptom severity did not influence average Fitbit count although, upper extremity devices occasionally underestimated steps produced. Conclusions: Whilst superior step-count accuracy was seen with devices placed on lower limbs, these data show that Fitbit charge 4 may be a reliable device for monitoring steps in manifest HD.