2014-08-012024-05-14https://scholars.lib.ntu.edu.tw/handle/123456789/657656摘要：創傷性脊髓損傷(traumatic spinal cord injury)會引發全身性的免疫功能的失調。此外，脊髓損傷也可能會誘發系統性的淋巴細胞活化反應及自體抗體的合成，進而引發系統性自體免疫反應。多發性硬化症（multiple sclerosis）是一種免疫系統攻擊中樞神經系統的疾病，先前研究指出多發性硬化症，可能與中樞神經系統的創傷有關。然而以往有關於多發性硬化症的研究大多以分子生物學或神經影像構造分析為主，對於多發性硬化症與中樞神經系統創傷相關性之流行病學研究卻非常有限。截至目前為止，回顧文獻尚未有針對脊髓損傷與多發性硬化症關聯性的大規模流行病學研究，因此，脊髓損傷與多發性硬化症風險之間的關係仍有待釐清。由於脊髓損傷與多發性硬化症均為發生率較低之疾病，所以若要探討兩者之間是否有時序相關，使用大型健康資料庫進行縱向追蹤研究是較為可行的方法。 此外，因為多發性硬化症其發生率相當低，因此雖然以大型全民健保資料庫進行研究，可以預期的是這些多發性硬化症 outcome事件的發生數目會相當低，如果要以複回歸分析調整干擾因子的方式去分析有限的 outcome 事件時就會有許多限制。因此，本研究計畫將採用傾向分數配對（propensity score matching）使分析較為可行，把干擾因子的效應降到最低。 本計畫將使用行政院衛生福利部健康資料加值應用協作中心之全民健保資料庫及死亡登錄檔，使用世代研究設計探討創傷性脊髓損傷對於後續發生多發性硬化症風險的影響。為了控制來自許多潛在的干擾因子，包括年齡、性別、醫學共病症、社經地位、以及地理位置等，本研究使用傾向分數配對建立可比較的脊髓損傷與其對照世代研究分組。有鑑於全民健保資料庫資料量龐大，本研究將使用兩階段傾向分數配對，根據傾向分數挑選與脊髓損傷病患配對的唯一對照個案。使用 stratified Cox proportional hazard regression評估以創傷性脊髓損傷對多發性硬化症發生風險的影響。 <br> Abstract: Traumatic spinal cord injury may be associated with dysfunction in the regulation of immune system. In addition, spinal cord injury has been suggested to promote chronic systemic lymphocyte activation and autoantibody synthesis, which in turn, may result in systemic autoimmunity. Multiple sclerosis is a chronic inflammatory demyelinating disease of the central nervous system. Most evidence supports the autoimmune pathogenesis of this disease. Previous study has shown multiple sclerosis may be associated with traumatic injury in central nervous system. However, epidemiological research on the link between multiple sclerosis and traumatic injury in central nervous system was very limited. To the best of our knowledge, no population-based epidemiological study on association between spinal cord injury and multiple sclerosis has been published. The purpose of our proposal is therefore to investigate the impact of spinal cord injury on the risk of developing subsequent multiple sclerosis. Since both traumatic spinal cord injury and multiple sclerosis are diseases with low incidence, a population based, longitudinal follow-up study would be required to investigate the temporal association between these two diseases. Furthermore, given the low incidence of multiple sclerosis, it can be expected that the outcome events would be scarce, which may increase the difficulty of minimizing confounding effects through covariate adjustment in multiple regression analysis. We therefore apply propensity score matching to minimize the potential confounding effects of various demographic, clinical, and socioeconomic variables. This study will use the National Health Insurance database and mortality registry from the Department of Health, Taiwan. We will use propensity score matching to establish comparable spinal cord injury group and reference group. Due to the large sample size of the insurance database, a two-stage sampling method with 8-to-1 greedy matching algorithm will be used to identify the unique propensity score-matched reference subject. Stratified Cox proportional hazard regression matched on propensity score is used to estimate the effect of spinal cord injury on the risk of developing multiple sclerosis.脊髓損傷多發性硬化症危險因子傾向分數Spinal cord injuryMultiple sclerosisRisk factorsPropensity score.