市售穿戴式裝置估計輕度至費力運動強度的效果與閾值

The effect of commercially available wearable device estimates the threshold of light to vigorous exercise intensity

緒論：先前研究發現以加速度估計運動強度的效果尤佳。因技術發展突破，加速規已普遍內建於穿戴式裝置中，但其效益是否與過去研究一致，仍有待釐清。因此，本研究探討以市售穿戴式裝置作為監測儀器的可行性，並由運動加速度數值量化不同強度的閾值。方法：預備性實驗以 9 名體育系健康大學生，先行檢定裝置的信度與效度。正式實驗招募 18 名體育系健康大學生 (平均年齡20.89 ± 0.83歲、身高172.13 ± 5.17公分、體重70.93 ± 12.14公斤、最大心跳率192.38 ± 0.58 次/分鐘)。全部研究對象須先進行 5 分鐘坐姿安靜心跳率測量，隨後進行漸增負荷連續運動，速度階段分別訂為走路運動 3.2、4.0、4.8、5.6 公里/小時，跑步運動 6.4、9.6、11.3、12.9 公里/小時，每階段測驗時間為5分鐘。全程記錄 X、Y、Z 軸加速度與心跳率。將所得三軸加速度值依 √(X^2+Y^2+Z^2 ) 轉換為每分鐘加速度計數值，再以最後一分鐘穩定加速度計數值與保留心跳率百分比進行分析。資料處理以皮爾森積差相關與接受者操作特徵曲線進行分析，統計水準訂為α= .05。結果：儀器信度與效度檢定結果，分別為效標關聯效度ICC = .89 ~ .99, p <.05，再測信度 ICC = .58 ~ .98，p < .05。然而每分鐘加速度計數值與保留心跳率的相關程度為 r = .69，p < .05。不同運動強度的每分鐘加速度計數值閾值最佳切點，分別為輕度31,149 counts/min、中度32,451 counts/min、費力34,819 counts/min。結論：市售穿戴式裝置能作為有效且穩定的監測儀器，並降低取得三軸加速度參數的費用與技術門檻。定義輕度至費力運動強度的加速度計數值，可提供簡便、有效且長時間量化日常生活身體活動與運動強度的評估工具。

Purpose: Previous studies have reported an excellent effect of estimating exercise intensity by acceleration. Due to the breakthrough of technology, accelerometer is commonly included by the wearable devices in nowadays. Does the effectiveness still consistent with previous studies? Further clarification is needed. The aim was to explore the feasibility of a commercially available wearable devices serve as a monitoring instruments. And quantified the thresholds of different exercise intensities by motion accelerations. Methods: The preliminary experiment recruited 9 healthy college students from the department of physical education in which to test the reliability and validity of the devices.. Formal experiments recruited 18 healthy males of college students from the department of physical education (age: 20.89 ± 0.83 years, height: 172.13 ± 5.17 cm, weight: 70.93 ± 12.14 kg; HRmax: 192.38 ± 0.58 beat/min). Subjects performed the 5 mins seated resting heart rates measurement at the beginning of the formal experiments. Followed by incremental continuous exercise. The speeds of continuous exercise phase were set at 3.2, 4.0, 4.8, 5.6 km/h in walking, 6.4, 9.6, 11.3, 12.9 km/h in running. Each phase lasts for 5 mins. The X, Y, Z axes accelerations and heart rates were recorded during entire experiment. The acceleration counts per min (ACC counts / min) were converted by triaxial acceleration values by √(X^2+Y^2+Z^2 ). The analyzing data derived from the last minute’s (4-5 min) steady-state %HRR and ACC of each speed phase. Data were analyzed by Pearson’s r coefficient and ROC curves. Significant level was set at α = .05. Results: The reliability and validity of the device criterion-related validity were ICC = .89 - .99, p<.05 and test-retest reliability coefficient were ICC = .58 - .98, p<.05, respectively. However, the correlation between ACC and %HRR was r = .69, p<.05. The best ACC cut-off point of different exercise intensities thresholds were light 31,149 counts/min, moderate 32,451 counts/min, vigorous 34,819 counts/min, respectively. Conclusions: The commercially available wearable device could be served as an effective and stable monitoring instrument. Reducing the cost and the technical access level for obtaining three-axis acceleration parameters. The definitions of light to vigorous exercise ACC provides an easy and effective quantifying tool for the evaluation of physical activity and long-term exercise intensity in daily life.