研究背景：穿戴式運動監測裝置越來越受歡迎，然而除了追蹤運動量之外通常缺乏其他配套功能與資 訊再利用，導致身體活動量的改善效果有限。目前較少研究探討該如何跳脫穿戴式裝置的傳統記錄功 能，進一步使個人培養運動生活型態；這項突破將關係到未來穿戴式裝置在運動行為介入上的發展 性。本研究將建立一結合穿戴式裝置、手機微型程式、雲端資料庫網站之運動健康物聯網，由穿戴式 裝置之功能層面以及影響長期使用穿戴式裝置的因子部分的雙角度切入，探索運動健康物聯網對大學 生的身體活動改變量之最佳方案設計。 研究目的：(1) 評估以穿戴式運動偵測裝置(運動手錶)為主軸之健康物聯網，應包含哪些附加功能才能 達到改善身體活動量之最佳效果； (2) 評價在以穿戴式運動偵測裝置為主軸之健康物聯網中增添線上 社群及利用不同屬性(內部/外部促進者之有無)之社群，對於提升身體活動量之影響差異；(3) 檢定穿 戴式運動偵測裝置之健康物聯網對於身體活動量和健康指標上的作用，使否會依人格特質而有所不 同；(4) 探討由穿戴式運動偵測裝置為主軸之健康物聯網是否能間接影響大學生之飲食行為、睡眠習 慣、心理健康、和代謝心理指標；(5) 建立以線上社群網絡和人格特質預測生身體活動量之模型。 材料與方法：第一階段我們將設置一項以穿戴式運動偵測裝置為主的運動健康物聯網，招攬受試者240 人進行為期4 個月之穿戴式裝置運動介入方案，並隨機分配進入「自我監測功能介入組」(60 人)、「自 我監測功能+線上競賽遊戲介入組」(60 人)、「自我監測功能+線上競賽遊戲+資訊回饋介入組」(60 人)、 和「控制組」(60 人)。此實驗設計提供我們比較物聯網中不同的功能(紀錄與設定目標、線上競賽遊戲、 資訊回饋)對於身體活動量之縱向效果。在第二階段，我們在物聯網中增設Facebook 網路社交平台， 預計180 人將被招攬進行8 個月的追蹤，在接受人格測驗後將被隨機分配到「網絡社群介入組(無促 進介入者)」(60 人)、「網絡社群+外部促進介入者」(60 人)、和「網絡社群+內部促進介入者」(60 人)。 資料探勘技術將被用來量化社群間之互動和角色定位，加以探討這些關係如何影響身體活動量及健康 生理指標；多因子共變數分析將探討使用穿戴式運動裝置與人格特質對於大學生在參與介入前後之運 動改變量、心理健康、以及代謝生理指標之長期縱向影響。 預期成效: 藉由了解以穿戴式偵測裝置為主的物聯網對於增加身體活動量和改善生理指標之成效，可 供未來衛生單位擬定以創新科技策略達到提升活動量的本土化衛教策略。本研究結果將可提供後續穿 戴式裝置相關之運動研究多項建議，包含該結合哪些物聯網功能、是否該將社群網絡互動以及個人特 質納入介入方案之考量。同時，本研究探勘之資料將可建構出以社群網絡指標和人格特質來預測大學 生身體活動量程度的模型，供日後篩選出可能的高危險族群來提早參與介入並降低健康風險。鑒於穿 戴式運動裝置之普遍性，本研究所開發出的運動健康物聯網將預期達到比傳統運動衛生教育介入更具 深度與廣度的健康成效。 Background: Challenges continue to exist in long-term adoption of wearable activity devices. Without appropriate expanded features and meaningful information processed from collected data, the effectiveness of utilizing wearable devices to produce positive activity behavioral changes is lower. More studies on how to transform wearable activity devices from “wearable” to “actable” is needed. We propose a wearable activity devices-centered Internet of Things (IoT) network including wearable devices (sport watches), mobile applications, website, and cloud servers; we will use a physical activity intervention associated with the IoT to investigate the efficacy of wearable activity devices to promote physical activity for college students. Objectives: (1) to evaluate which features of a wearable activity device-driven intervention has the most beneficial effects on promoting physical activity; (2) to identify if online social network have impacts on effectiveness of an activity device-based intervention in promoting physical activity; (3) to examine whether or not the efficacy of a wearable activity device-based intervention in physical activity differed by traits of personality; (4) to assess if changes in activity levels via device-driven interventions lead to longitudinal improvements in diet, sleep, emotional well-being, and metabolic profile; (5) to build the prediction model of physical activity based on a device-based intervention with IoT. Methods: In the first stage, a randomized 4-arms repeated device-based interventions will be delivered to increase engagements in physical activity for 4 months. A total of 240 college students will be recruited and randomly assigned into 3 intervention groups (self-monitoring alone, self-monitoring + online competition game, and self-monitoring + online competition game+feedbacks) and the control group (N=60 for each group). In the second stage, 180 college students will particpate a 3-arm (N=60 for each group) group-randomized trial will be conducted for 8 months. Students will be randomly assigned into 2 intervention groups (Facebook group with external facilitation, Facebook group with internal facilitation) and a control group (N=40 for each group (Facebook group with no facilitation). Data mining techniques will be used to assess how social network and online interactions affect efficacy of the device-based intervention in activity levels. Repeated-measures analyses of variance will be conducted to assess whether or not intervention efficacy differ by personality traits. Expected research contributions: This study will be one of the very few that implement device-based interventions to increase activity levels for Taiwanese young adults. Our results could provide comprehensive investigations and translatable findings on finding effective additional features for wearable activity device to increase activity behaviors, including social network and individual characteristics. Moreover, the prediction model we built for physical activity and health is crucial for identifying the risk of an inactive lifestyle and the subsequent health problems for young adults, it also effectively screens for high risk population to be targeted for physical activity programs. With the rising popularity of the wearable technology, the utilization of a device-based IoT intervention may be a more affordable and practicable strategy to promote physical activity over time.
