以每日步數作為高齡者商用行動裝置測量之效度初探

Abstract

隨著科技及資訊世代的來臨，在2021-2023年間，臺灣智慧型手機持有率高達89%，臺灣高齡者商用行動裝置使用率達37.2%，國內過往研究較少針對不同商用行動裝置做效度研究。據此，本研究之主要研究目的為以三軸加速規測量之步數為黃金校標，比較其與商用行動裝置測量每日步數之關係。本研究以便利取樣的方式，招募社區高齡者198名進行研究，刪除未使用商用行動裝置的高齡者，最後有175位高齡者納入分析，每名受試者一天以一筆數據計算，總共有1400筆資料，排除不完整的資料410筆，最後共有990筆資料納入後續分析。本研究之研究工具如下1.以ActiGraph GT3X+三軸加速規測量高齡者每日步數，做為黃金校標。2.高齡者主要使用的商用行動裝置 (如智慧型手機及穿戴式裝置) 所測量高齡者之每日步數3.使用自填式問卷獲取社會人口學變項。分析方法包括描述統計、皮爾森相關、獨立樣本t檢定以及Bland-Altman plot圖。本研究之結果發現，所有商用行動裝置與三軸加速規所測之步數之皮爾森相關性皆大於0.8，表示裝置皆具有良好的效度。進一步區分商用行動裝置後，智慧型手機測量之每日步數平均會低估790步/天；而穿戴式裝置測量之每日步數平均則會高估1012步/天。綜上所述，智慧型手機會低估高齡者每日步數，而穿戴式裝置則會高估每日步數。有鑑於商用行動裝置可能對個人、臨床醫生和研究人員進行身體活動監測有實際用途，建議未來可以納入更多的樣本以及針對更多品牌的裝置進行效度研究，給予高齡者自我監測步數的依據準則，促進他們採取更加健康的生活方式和行為習慣。With the advent of the technology and information generation, the smartphone ownership rate in Taiwan reached 89% in 2021-2023, and the utilization rate of commercial mobile devices among the elderly in Taiwan reached 37.2%. It is noticed that there are few studies on the validity of different commercial mobile devices in Taiwan. Therefore, the main objective of this study was to compare the relationship between the accelerometer-measured steps and the steps measured by commercial mobile devices. In this study, we recruited 198 elderly people in the community using convenience sampling, and 175 elderly people who had commercial mobility devices were included in the analysis. With a total of 1,400 day-to-day data, and 410 incomplete data of them were excluded, and finally 990 data were included in the subsequent analysis. The research tools were as follows: 1. ActiGraph GT3X+ 3-axis accelerometer was used to measure the daily step counts of the elderly as a golden calibration; 2. the daily step counts of the elderly were measured by the commercial mobile devices that were mainly used by the elderly (e.g., smart phones and wearable devices); and 3. the socio-demographic variables were obtained by using a self-administered questionnaire. The methods of analysis included descriptive statistics, Pearson's correlation, independent sample t-test and Bland-Altman plot. Our results showed that the Pearson's correlation between the number of steps measured by all commercial mobility devices and the 3-axis accelerometer was all greater than 0.8, which indicated that these devices had good validity. After further differentiating the commercial mobile devices, the daily steps measured by smartphones were underestimated by 790 steps/day on average, while the daily steps measured by wearable devices were overestimated by 1012 steps/day on average. In conclusion, smartphones underestimated the number of daily steps among elderly people, while wearable devices overestimated. For commercial mobile devices for physical activity monitoring by individuals, clinicians and researchers, it is suggested that future studies should include more samples and more brands of devices, in order to inform the elderly for self-monitoring their steps and to promote the adoption of healthier lifestyles and behaviors.