Sample Size Calculation for Twin Concordance Studies with Ascertained Multinomial Data

雙胞胎相符度檢定鑑取多分類資料研究之樣本數計算

黃崑明(Jason Kun-Ming Huang) ； 戴政(John Jen Tai)

《中國統計學報》 51卷4期 (2013/12) Pp. 409-426

摘要

雙胞胎研究設計是一個探索遺傳因子對某一疾病影響的實用方法。常用於雙胞胎資料分析的統計方法，大多是以對同卵與異卵雙胞胎的相符度進行相等與否檢定，作為發展方法的方向。在同卵與異卵雙胞胎若具有相同的非遺傳因子的假設下，一個得到顯著結果的檢定，提供了所研究疾病於疾病發展過程中，遺傳因子扮演的可能角色的證據。雖然雙胞胎研究的施作通常是建立在已知個數的雙胞胎登錄群體資料庫（亦即，若如此，則無需進行樣本數估算），但在無資料庫可利用情形下，樣本數的計算仍是必要的。當雙胞胎研究需要處理的疾病分類超過二種，且取樣的方式是非隨機的鑑取程序時，則所面對要分析的資料成為統計學上所稱的截切多分類配對資料，此時有病個體被獨立鑑取到的機率（遺傳學稱之為鑑取率或首被鑑取者比例），必須被納入分析方法中。以之前發展出以概似函數為基礎，分析鑑取多分類資料的雙胞胎相符度檢定方法作為主體，本研究推導出了對應這些方法的樣本數估計方法，評估了數種首被鑑取者比例對樣本數的影響及其對應的檢定力的表現，並且利用二個已存在的資料，來舉例說明這裡所提出的樣本數計算方法。

關鍵字

相符度模型 ； 非中心參數 ； 首被鑑取者相符度 ； 重參數化 ； 截切多項分配

並列摘要

Twin study design is a useful genetic approach for exploring the influence of genetic factors on a disease. The statistical methods commonly used in analysis of twin pairs data intend to test the equality of concordances between monozygotic (MZ) and dizygotic (DZ) twins. A test of significant result offers the evidence for the potential role of genetic factors played in the development of an investigated disease under the assumption that nongenetic effects are shared to the same degree by MZ and DZ twin pairs. Although twin studies are often carried out based on existing large population-based twin registries of a fixed size (i.e., no need to estimate sample size), estimation of sample size for a twin study is still needed when registries are not available. In this paper, based on the likelihood-based concordance testing procedures developed for analysis of the ascertained data previously, the corresponding sample size estimation formulas are derived. The effects of various proband proportions on sample sizes and powers corresponding to the calculated sample sizes are evaluated. Two existing data sets are employed for illustration of our proposed formulas.

並列關鍵字

Concordance model ； Noncentrality parameter ； Probandwise concordance ； Reparameterization ； Truncated multinomial distribution

參考文獻

Allen, G.,Hrubec, Z.(1979).Twin concordance: a more general model.Acta Geneticae Medicae et Gemellologicae.28,3-13.

Bartfay, E.,Donner, A.(2001).Statistical inferences for a twin correlation with multinomial outcomes.Statistics in Medicine.20,249-262.

Boomsma, D.I.,Vink, J.M.,van Beijsterveldt, T.C.E.M.,de Geus, E.J.C.,Beem, A.L.,Mulder, E.J.C.M.,Oerks, E.M.,Riese, H.,Willemsen, G.A.H.M.,Bartels, M.,van den Berg, M.,Kupper, N.H.M.,Polderman, T.J.C.,Psothuma, D.,Rietveld M.J.H.,Stubbe, J.H.,Knol, L.I.,Stroet, T.,van Baal, G.C.M.(2002).Netherlands twin register: a focus on longitudinal research.Twin Research.5,401-406.

Donner, A.,Eliasziw, M.(1992).A goodness-of-fit approach to inference procedures for the kappa statistic: confidence interval construction, significance-testing and sample size estimation.Statistics in Medicine.11,1511-1519.

Hannah, M.C.,Hopper, J.L.,Mathews, J.D.(1985).Twin concordance for a binary trait. II. Nested analysis of ever-smoking and ex-smoking traits and unnested analysis of a "committed-smoking" trait.American Journal of Human Genetics.37,153-165.