住院中風病患睡眠指標與入院短期日常生活活動進程之探討

背景：在腦中風患者中，普遍存在睡眠呼吸障礙和睡眠障礙，這兩個睡眠問題的客觀指標會涉及日常生活能力；因此，睡眠呼吸障礙和睡眠障礙的客觀多頻道睡眠生理檢測(Polysonography, PSG)指標與中風住院病人日常生活能力恢復是否有關是值得探討的議題。目的：本研究以入出院的巴氏量表(Barthel index, BI)評估分數變化為日常生活能力恢復指標，並對腦中風住院患者進行睡眠呼吸障礙和睡眠障礙的多頻道睡眠生理檢測(PSG)。本研究將探討PSG指標與中風住院病人日常生活能力恢復的關連性。方法：本研究為資料回溯法，採用病歷資料收集病患社會人口學資料、中風特性、併發症、巴氏量表(BI)、美國國衛院腦中風評估表(National Institutes of Health Stroke Score, NIHSS)、多頻道睡眠生理檢測(PSG)指標。以 Pearson相關和t檢定來進行變項和巴氏量表(BI)之間的單因子分析；進一步以逐步迴歸分析找出巴氏量表(BI)的關鍵影響因子。結果：123位受試者中，平均年齡為62歲，平均身體質量指數(Body Mass Index, BMI)為24，女性佔33％，中風後至本次入院平均為91天，平均每小時的呼吸中止和通氣不足次數(Apnea hypopnea index, AHI)為47次，有103位(92％)為中度至重度睡眠呼吸障礙，住院期間使用連續陽壓呼吸器(Continuous positive airway pressure, CPAP)超過14天者有24位(19.5％)。經過多頻道睡眠生理檢測發現，平均總睡眠時間(Total sleep time)為259分鐘、平均睡眠效率(Sleep efficiency)為70％、平均睡眠潛伏期(Sleep latency)為24分鐘、平均入睡後醒來(Wakefulness after sleep onset)為93分鐘。巴氏量表前測與前後測差值平均分數為34和10分。57位(46％)缺血型中風受試者的美國國衛院腦中風評估表(NIHSS)平均分數為10分。單變量分析顯示，NIHSS和巴氏量表前測(p<0.020)與前後測差值(p=0.002)均呈負相關。同時，在向後逐步迴歸分析中，巴氏量表前測與年齡(p=0.026)和入睡後醒來時間(p<0.001)均呈現負相關，高血壓(p<0.001)、睡眠潛伏期(p=0.020)、非快速動眼期第一期百分比(p=0.001)及血氧飽和度下降≥4％(p=0.046)均與巴氏量表前後測差值呈現負相關。結論：本研究顯示，多頻道睡眠生理檢測(PSG)應該包含在中風病患的入院常規評估中。我們推測，失眠的中風病患將會成為更高風險的族群，但中風也可能會加重病患的睡眠呼吸障礙，但可透過戴連續陽壓呼吸器(CPAP)治療來改善睡眠障礙以穩定病情。

Background and Purpose—While prevalent sleep-disordered-breathing (SDB) and sleep disturbances found in subacute stroke patients, scarce researches with objectivemeasures of these two sleep problemstogether were found relating to activities of daily living (ADL) levels. We investigated that objective polysomnographic measures of SDB and sleep disturbance are correlated to ADL outcome over inpatients’stroke rehabilitation programs.Methods—123 eligible subjects were retrospectively enrolled with data of socio-demography, medical comorbidities, stroke characteristics, National Institutes of Health Stroke Score (NIHSS), BI scores, and parameters of one night polysomnography study.Results—Of overall 123 eligible patients (61.6±13.1 years, 23.8±3.4 kg/m2, 33% female; 90.5±36.7 days post-stroke; 46.7±25.1 events/hr in apnea-hypopnea index), 103 (92%) were moderate to severe SDB and 24 patients (19.5%, ≥14 days before discharge) had better CPAP adherence. Diverse polysomnographic values were found over total sleep time (259±71 min), sleep efficiency (69.5±19.3%), sleep latency (24.3±30.9 min) and wakefulness after sleep onset (93.1±74.2 min) as well as the percentage proportions of various ultradian sleeps. Admission levels and admission-discharge difference of Barthel Index (BI) scores were 33.8±23.2, and 10.1±9.2, respectively. NIHSS (10.2±5.6) were available in 57 (46%) ischemic patients. On univariate analysis, NIHSS were negatively associated with admission levels and changes of BI scores (p<0.001, =0.002, respectively). On stepwise backward logistic regression testing, while, age (p=0.026) and wakefulness after sleep onset (p<0.001) were both negatively associated (adjusted R2 =0.234) with admission levels of BI score; comorbidity hypertension, sleep latency, percentage-proportion of stage one sleep, and desaturation events ≥4% (p values <0.001, 0.020, 0.001 and 0.046, respectively; adjusted R2 =0.227) were all negatively independently associated with changes of BI scores.Conclusions—Our data support that PSG should be included in recommendation for management of subacute stroke patients. Our speculation that insomnia in subacute stroke individuals exhibit more urgent than SDB, in terms of stroke outcome, possibly alters the therapeutic strategy for these patients with SDB but resistant to CPAP therapy.