緒論：規律運動是糖尿病照護的基礎，然政府推動之每日萬步對糖尿病患血糖控制 是否有實質效果，需進一步釐清及驗證，研究之目的為探討結構式有氧與萬步運動對胰 島β 細胞功能、身體組成及代謝指標之影響。方法：本研究為12 週橫斷式運動介入性研 究，樣本為北部某教學醫院新陳代謝門診之第二型糖尿病病患共109 位，隨機分派實驗 組為結構式有氧組36 位及萬步組37 位；及控制組36 位，維持原靜態生活形態。個案在 介入前後分別接受身體組成及血液指標之測量，另，透過2 小時葡萄糖耐糖試驗 (OGTT) 之多點血糖與胰島素值，運用BIGTT 的數學模式計算出胰島素敏感性 (BIGTT-SI) 及急 性胰島素反應 (BIGTT-AIR) 做為胰島細胞功能之指標。結果：血糖調控機制在12 週運 動介入後，結構式有氧運動相較控制組，降低空腹血糖值 (Δ = -32.63%, p < .001)、2 小時 葡萄糖耐糖值 (Δ = -24.64%, p = 0.047) 及糖化血色素 (Δ = -0.77%, p < .001)，並增加急性 胰島素反應分泌量Δ = 0.91%, p < .001)；萬步組相較控制組，降低空腹血糖值 (Δ = -33.43%, p < .001)、2 小時葡萄糖耐糖值 (Δ= -29.09%, p = 0.036) 及糖化血色素 (Δ = -0.69%, p < .0001)，並增加急性胰島素反應分泌量 (Δ = 0.63%, p < .001) 結論：本研究結 果證實每日累計萬步運動相似於結構式有氧運動在葡萄糖代謝上具正面效益，進而能改 善胰島β 細胞功能。因此，鼓勵坐式生活型態的個案能以累計萬步運動取代結構式有氧 運動，將運動融入生活並建立規律運動習慣。

Introduction: Regular exercise training has demonstrated beneficial effects on glycemic control. However, it is unknown if a pedometer-based intervention with a daily accumulation of a million steps would improve insulin action. This study’s aim was to evaluate the efficacy of two types of exercise training as compared with a control group on pancreatic β-cell function. Methods: We conducted a randomized controlled trial of 109 diabetes patients, who were recruited from an Endocrinology & Metabolism outpatient clinic. Patients were randomly assigned into training groups, including 36 in a structured based (SB) and 37 in a pedometer-based million step (PBMS), and 36 in unchanged lifestyle control group (CC). All participants in training groups followed the indicated exercise protocols during experimental period of 12 weeks. Anthropometric measurements and related biochemistry were assessed before and after treatments. The pancreatic β-cell function were mainly evaluated by BIGTT-AIR (Beta-cell function, Insulin sensitivity, and Glucose Tolerance Test, acute insulin response) or BIGTT-SI (insulin sensitivity). Results: After 12weeks exercise intervention, the SB group was significantly decreased in FG (Δ = -32.63%, p < .001), 2-h glucose (Δ = -24.64%, p = 0.047) and HbA1C(Δ = -0.77%, p < .001), and improved BIGTT-AIR (Δ = 0.91%, p < .001) as compared to control group; the PBMS group was significantly decreased in dreased (Δ = -32.63%, p < .001), 2-h glucose (Δ = -24.64%, p = 0.047) and HbA1C (Δ = -0.77%, p < .001), therefore significantly improved BIGTT-AIR (Δ = 0.91%, p < .001). Conclusion: The study proved that pedometer-based million step training had the same effect as structured based group training, were effective in enhancing insulin action and in improving pancreatic β cell function. Hence, the pedometer-based million steps exercise will be the best choice, in deed can instead of structured based exercise for the sedentary people to build up the regular exercise behavior.