2010-08-012024-05-13https://scholars.lib.ntu.edu.tw/handle/123456789/650234摘要：Rothman 提出的因果圓派模式是流行病學家非常熟悉的方法。每一個因果圓派以一系列的組成因子組合而成，只要集合所有因果圓派內的組成因子，就意味著疾病的產生。在疾病的成因探討上，利用因果圓派模式能夠同時考慮多因子作用且釐清危險因子對於疾病錯綜複雜的共同作用。然而，此模式一直以來只是一個概念示意圖，並未被實際應用在實證資料的分析。有鑑於此，本計畫著手發展統計方法進行因果圓派的建構。我們將建立一個因果圓派量化指標，稱其為因果圓派權重。由於Rothman 所提出的因果圓派模式中指出，各組成因子應皆為危險因子且各組成因子間的交互作用假定為加成性及協同性，為了符合該假設，我們採用非負係數加成性羅吉斯迴歸來進行危險勝算比的估計，並計算族群可歸因危險性。因果圓派權重可經由可歸因危險性矩陣方程式求得，另採用拔靴法估計其信賴區間。本計畫同時也探討因果圓派模式分析之相關方法學議題，包括干擾因子的調整、多等級危險因子之分析方法、加成模式之應用、閾值分析、隨時間變動之危險因子分析方法、基因與環境交互作用的因果圓派等。並進行完整之電腦模擬研究以及進行實證研究。<br> Abstract: Epidemiologists are familiar with Rothman’s model of causal pies. A causal pie containscombination of component causes and people will inevitably develop disease with thecompletion of it. Rothman’s causal-pie model helps to clarify the multi-factorial and complexinteractive nature in disease causation. However, the model has been more of a theoreticalconcept than a practical tool of data analysis. The aim of this project is to develop thestatistical methodology for causal-pie modeling. We will construct a new index, thecausal-pie weights’ (CPWs). The CPWs serve to quantify the relative importance of each andevery class of causal pies. To be conforming to Rothman’s model, we will apply anonnegative linear odds model to constrain all the odds ratios to be equal to or greater thanone, and the interactions between them to be additive or supra-additive. Based on theseconstrained odds ratios, we will then calculate the population attributable fractions (PAFs).The CPWs can then be calculated from the PAFs using a matrix equation. We will use thebootstrap technique to calculate the confidence intervals of CPWs. We will address a numberissues related to causal-pie modeling, such as confounding adjustment, polychotomous riskfactors, additive regression model, threshold analysis, time-dependent risk factors analysis,gene-environment interaction, etc. And we will also conduct computer simulation as well asapply our causal-pie modeling technique to real data.