新生儿高胆红素血症与催产素关系的病例-对照研究 1:1配比多级暴露水平分析

采用配比多级暴露水平分析方法,对200对(1:1)新生儿高胆红素血症与催产素关系的病例-对照研究资料进行分析,结果表明:分娩过程中使用催产素与新生儿高胆红素血症发生有关(χ²= 15.538,df=3,P=0.0014),三个暴露水平(催产素使用单位)的估计相对危险度分别为1~5.9(u):OR=1.085;6~8.9 (u):OR=5.5;9+(u):OR=3.687,并经对数线性趋势检验有极显著性(χ²=11.991,df=1,P=0.0005)。     

六号染色体短臂MHC区DRB3、DRB1基因多态性与精神分裂症症状的关系探讨

目的 探讨六号染色体短臂MHC区DRB3、DRBl基因多态性与精神分裂症症状的相关性。方法 采用聚合酶链反应(PCR)和限制性内切酶片段长度多态性(RFLP)方法检测两个基因位点上的单核苷酸多态性(SNPs)，并对116例精神分裂症患者家系进行连锁不平衡分析。结果 DRBl的SNP(rs707954：G／T碱基互换)等位基因与关系妄想症状呈显著相关(X^2=5．484，df=l，P=0．019)，GG、GT、TT三种基因型频率在关系妄想症状中呈显著相关(X^2=6．771，df=2，P=0．034)。rs707954的三种基因型频率与情感淡漠症状呈显著相关(X^2=12．110，df=4，P=0．017)。结论 DRBl位点等位基因与关系妄想和情感淡漠症状高度相关，DRBl基因型与关系妄想症状高度相关。     

Investigating the HLA component in rheumatoid arthritis: an additive (dominant) mode of inheritance is rejected, a recessive mode is preferred.

We examined the mode of inheritance of rheumatoid arthritis (RA) and estimated the genetic contribution of the HLA-linked locus to the development of RA using data from 111 multiplex families (54 London, 57 Cleveland), and 43 randomly ascertained patients (Seattle). HLA-DR4 was present in 78 multiplex probands (70%); a further 16 probands who were negative for DR4 were positive for DR1. Both DR4 and DR1 were significantly in excess when compared to control population frequencies (P less than 0.001); an additional finding was an excess of DR7, although the numbers of probands with DR7 were small. Despite the well-established HLA association with RA, neither recessive nor additive (dominant) modes of inheritance, nor any intermediate models have been ruled out using affected sib-pair and antigen genotype frequency among patients (AGFAP) methods. However, in our study the AGFAP data for HLA-DR4 and DR1 were close to recessive expectations (P = ns) while an additive (dominant) mode of inheritance was rejected (P less than 0.001). The same results were obtained by an independent method which considered HLA-DR transmission from affected parents to their affected children. The affected sib-pair haplotype sharing method showed deviation from random expectations but did not allow discrimination between recessive and additive (dominant) modes. The effect of the HLA-linked locus on familiarity accounted for only a 1.61-fold increased risk to sibs over the population prevalence, compared to an observed value of 3.90. This indicated that there could be at least one other non-HLA locus predisposing to RA with a weight that is slightly greater than that of HLA.     

New joint covariance- and marginal-based tests for association and linkage for quantitative traits for random and non-random sampling

We develop novel statistical tests for transmission disequilibrium testing (tests of linkage in the presence of association) for quantitative traits using parents and offspring. These joint tests utilize information in both the covariance (or more generally, dependency) between genotype and phenotype and the marginal distribution of genotype. Using computer simulation we test the validity (Type I error rate control) and power of the proposed methods, for additive, dominant, and recessive modes of inheritance, locus-specific heritability of the trait 0.05, 0.1, 0.2 with allele frequencies of P=0.2 and 0.4, and sample sizes of 500, 200, and 100 trios. Both random sampling and extreme sampling schemes were investigated. A multinomial logistic joint test provides the highest overall power irrespective of sample size, allele frequency, heritability, and modes of inheritance.     

The prevalence of occupational asthma and rhinitis among woodworkers in south-eastern Nigeria

Wood dusts are known to cause respiratory disorders like rhinitis and asthma. This study was therefore done to determine the magnitude of the problem among woodworkers in south-eastern Nigeria exposed to high level of wood dust. Five hundred and ninety one woodworkers were selected using a stratified random sampling. The prevalence of woodwork-related rhinitis and asthma were then observed in the study population. Also the peak expiratory flow rate (PEFR) of each woodworker was obtained. The prevalence of occupational rhinitis was 78%, while that of asthma was 6.5%. As period of woodwork increased the prevalence of rhinitis and asthma increased (rhinitis: chi2 trend = 53.015, df = 1, P = 0.000). For asthma, chi2 trend = 19.721, df = 1, P = 0.000). Also the PEFR significantly became low with increasing years of exposure to woodwork (chi2 trend = 75.965, df = 1, P = 0.000). In conclusion the prevalence of rhinitis and asthma in woodworkers was high and significantly increased with years of working as a woodworker.     

Estimation of P‐value of MAX test with double triangle diagram for 2 × 3 SNP case‐control tables

Single nucle otide polymorphisms (SNPs) are the most popular markers in genetic epidemiology. Multiple tests have been applied to evaluate genetic effect of SNPs, such as Pearson's test with two degrees of freedom, three tests with one degree of freedom (χ² tests for dominant and recessive modes and Cockran‐Armitage trend test for additive mode) as well as MAX3 test and MAX test, which are combination of four tests mentioned earlier. Because MAX test is a combination of Pearson's test of two degrees of freedom and two tests of one degree of freedom, the probability density function (pdf) of MAX statistics does not match pdf of χ² distribution of either one or two degrees of freedom. In order to calculate P‐value of MAX test, we introduced a new diagram, Double Triangle Diagram, which was an extension of de Finetti diagram in population genetics which characterized all of the tests for 2 × 3 tables. In the diagram the contour lines of MAX statistics were consisted of elliptic curves and two tangent lines to the ellipses in the space. We normalized the ellipses into regular circles and expressed P‐value of MAX test in an integral form. Although a part of the integral was not analytically solvable, it was calculable with arbitrary accuracy by dividing the area under pdf into finite rectangles. We confirmed that P‐values from our method took uniform distribution from 0 to 1 in three example marginal count sets and concluded that our method was appropriate to give P‐value of MAX test for 2 × 3 tables. Genet. Epidemiol. 34:543–551, 2010. © 2010 Wiley‐Liss, Inc.     

Linear trend tests for case-control genetic association that incorporate random phenotype and genotype misclassification error

The purpose of this work is the development of linear trend tests that allow for error (LTT ae), specifically incorporating double-sampling information on phenotypes and/or genotypes. We use a likelihood framework. Misclassification errors are estimated via double sampling. Unbiased estimates of penetrances and genotype frequencies are determined through application of the Expectation-Maximization algorithm. We perform simulation studies to evaluate false-positive rates for various genotype classification weights (recessive, dominant, additive). We compare simulated power between the LTT ae and its genotypic test equivalent, the LRT ae, in the presence of phenotype and genotype misclassification, to evaluate power gains of the LTT ae for multi-locus haplotype association with a dominant mode of inheritance. Finally, we apply LTT ae and a method without double-sample information (LTT std) to double-sampled phenotype data for an actual Alzheimer's disease (AD) case-control study with ApoE genotypes. Simulation results suggest that the LTT ae maintains correct false-positive rates in the presence of misclassification. For power simulations, the LTT ae method is at least as powerful as LRT ae method, with a maximum power gain of 0.42 over the LRT ae method for certain parameter settings. For AD data, LTT ae provides more significant evidence for association (permutation p=0.0522) than LTT std (permutation p=0.1684). This is due to observed phenotype misclassification. The LTT ae statistic enables researchers to apply linear trend tests to case-control genetic data, increasing power to detect association in the presence of misclassification. If the disease MOI is known, LTT ae methods are usually more powerful due to the fact that the statistic has fewer degrees of freedom.     

Socio-demographic correlates of screening intention for colorectal cancer

OBJECTIVE: To assess the relationship between socio-demographic factors and screening intention for colorectal cancer (CRC). METHODS: A cross-sectional survey of a random sample of 884 Queenslanders aged 40-80 years was conducted using a computer-assisted telephone interviewing system. The factors measured included socio-demographic characteristics, personal history of CRC, knowledge of others with CRC and perceived symptom status. Chi-squared and Monte Carlo estimates of Fisher Exact Tests were performed to determine the associations between socio-demographic factors and screening intention. In multivariate analyses, multinomial logistic regression (MNLR) was utilised to examine potential determinants of screening intention. RESULTS: 77.5% (95% CI 74.0%-80.7%) of the respondents indicated their intention to participate in CRC screening if it were recommended by their doctor or health authorities. The likelihood ratio chi-squared tests in the MNLR analyses show that age (chi(df = 6)2 = 15.0; p = 0.02), education (chi(df = 8)2 = 19.4; p = 0.01), perceived symptom status (chi(df = 4)2 = 22.9; p = 0.00), sex (chi(df = 2)2 = 4.5; p = 0.11), income (chi(df = 14)2 = 19.6; p = 0.14) and personal history of CRC (chi(df = 2)2 = 4.3; p = 0.12) were potential determinants of screening intention. Other socio-demographic factors, including country of birth, private health insurance status, Socio-economic Index for Areas, and Rural and Remote Areas Classification codes, were not associated with screening intention. CONCLUSIONS AND IMPLICATIONS: The results indicate that a variety of socio-demographic factors are associated with screening intention and need to be considered in the future development of a population-based screening program for CRC.     

Health education to school children in Okpatu, Nigeria: impact on onchocerciasis-related knowledge

We assessed the influence of health education on the knowledge about onchocerciasis in school children in Okpatu, Nigeria. The children, aged between 11 and 17 y, received health education in both English and Igbo (local language) for three months on the transmission, clinical manifestations, treatment and prevention of onchocerciasis. Illustrated pictorial materials were used to support and enhance their understanding of the subject matter. Their level of knowledge was evaluated nine months later using a pre-tested personal interview administered questionnaire. A significantly higher proportion of these children knew about onchocerciasis (chi2=260.4, df=1, P<0.0001), and its causative agent (chi2=175.0, df=4, P<0. 0001), clinical manifestations (chi2=254.0, df=5, P<0.0001), diagnosis (chi2=123.9, df=2, P<0.0001), treatment (chi2=197.8, df=3, P<0.0001) and prevention (chi2=220.8, df=3, P<0.0001) in the post- than in the pre-educational intervention. It is therefore concluded that school-based health education showed an increase in knowledge about onchocerciasis and school children could provide a useful 'multiplier' resource for health education in the community.     

Group sequential large sample T2-like chi2 tests for multivariate observations

In many studies, a K degree of freedom large sample chi2 test is used to assess the effect of treatment on a multivariate response, such as an omnibus T2-like test of a difference between two treatment groups in any of K repeated measures. Alternately, a K df chi2 test may be used to test the equality of K+1 groups in a single outcome measure. Jennison and Turnbull (Biometrika 1991; 78: 133-141) describe group sequential chi2 and F-tests for normal errors linear models, and Proschan, Follmann and Geller (Statist. Med. 1994; 13: 1441-1452) describe group sequential tests for K+1 group comparisons. These methods apply to sequences of statistics that can be characterized as having an independent increments variance-covariance structure, thus simplifying the computation of the sequential variance-covariance matrix and the resulting sequential test boundaries. However, many commonly used statistics do not share this structure, including a Liang-Zeger (Biometrika 1986; 73: 13-22) GEE longitudinal analysis with an independence working correlation structure and a Wei-Lachin (J. Amer. Statist. Assoc. 1984; 79: 653-661) multivariate Wilcoxon rank test, among others. For such analyses, this paper describes the computation of group sequential boundaries for the interim analysis of emerging results using K df tests that are expressed as quadratic forms in a statistics vector that is distributed as multivariate normal, at least asymptotically. We derive the elements of the covariance matrix of multiple successive K df chi2 statistics based on established theorems on the distribution of quadratic forms. This covariance matrix is estimated by augmenting the data from the successive interim analyses into a single analysis from which the component sequential tests and their variance-covariance matrix can then be extracted. Boundary values for the sequential statistics can then be computed using the method of Slud and Wei (J. Amer. Statist. Assoc. 1982; 77: 862-868) or using the alpha-spending function of Lan and DeMets (Biometrika 1983; 70: 659-663) with a surrogate measure of information. An example is presented using the analysis of repeated cholesterol measurements in a clinical trial.     

AGFAP Method: Applicability under different ascertainment schemes and a parental contributions test

The antigen/allele genotype frequencies among patients (AGFAP) method has been powerful in discriminating between modes of inheritance, and detecting heterogeneity effects, for a number of diseases associated with the HLA system. The method is not dependent on the high level of polymorphism seen in the HLA system, but does require a marker allele association with disease. With recent rapid advances in mapping of the human genome, the method is increasingly relevant in all disease studies. Extension of the AGFAP method to ascertainment schemes other than random sampling of patients is presented here. The method is shown to be robust for distinguishing between incompletely penetrant recessive vs. additive or dominant models if affected children are obtained from nuclear families selected on the basis of at least two affected members: two affected sibs, or an affected parent and affected child. The method can lead to false conclusions for data from families ascertained for at least one affected parent and two affected children. A new test, termed the parental contributions test, applicable in families selected for the presence of an affected parent, and one or more affected children, is presented. The test, based on the expected symmetry (recessive) vs. asymmetry (additive and dominant) of parental marker allele contributions to an affected offspring in these pedigrees, is powerful in distinguishing between these modes of inheritance when there is a marker allele association with disease. Sporadic cases of disease are shown to cause deviations from AGFAP expectations for the recessive model, but not for the additive model. These results will aid in study of the genetics, and hence molecular basis, of complex diseases. © 1993 Wiley-Liss, Inc.     

Genetic model testing and statistical power in population-based association studies of quantitative traits

The optimal method for considering different genetic models in association studies is not clear. We compared analytical strategies that use different genetic models to analyze genotype-phenotype information from association studies of quantitative traits in unrelated individuals. We created simulated datasets where the minor alleles are causal with an additive, dominant, or recessive mode of inheritance over a range of allele frequencies. We then computed power to detect these causal alleles using one or a combination of statistical models in a standard regression framework, including corrections for the multiple testing incurred by analyzing multiple models. Our results show that, as expected, maximal power is achieved when we test a single genetic model that matches the actual underlying mode of inheritance of the causal allele. When the inheritance pattern of the causal allele is unknown, the co-dominant model, a single two degrees of freedom test, has good overall performance in any of the three simple modes of inheritance simulated. Alternatively, it is slightly more powerful to analyze all three genetic models together (additive, dominant, and recessive), but only if the significance thresholds used to correct for analyzing multiple models are appropriately determined (such as by permutation). Finally, a commonly employed approach, testing the additive model alone, performs poorly for recessive causal alleles when the minor allele frequency is not close to 50%. Our observations were confirmed by analyzing an existing genetic association dataset in which we detect the effect of a KCNJ11 variant on insulinogenic index in unrelated non-diabetic individuals.     

Affected kin-pair IBD methods: genetic models.

Cases of interest using affected sib-pair methods to distinguish between recessive and additive (dominant) modes of inheritance of a disease-predisposing gene involve goodness-of-fit tests with a small expected number in the "share-zero parental haplotypes" category, as well as an unknown parameter, the frequency of the disease-predisposing allele. Our simulations demonstrate that the real significance level of the chi-square test using the three-haplotype-sharing IBD values (share 2, 1, and 0 parental haplotypes) is close to the assumed (.05) level in these cases, so that the haplotype-sharing classes do not have to be lumped, which would leave no degrees of freedom for a statistical test. The validity of the chi-square approximation in cases of small expected frequencies has previously been described, but the situations that have been considered do not cover the very small values in the share-zero category that are often expected in the affected sib-pair analysis, nor do they involve estimation of an unknown parameter. Although including IBD values from affected kin pairs other than sibs can be a very powerful tool in demonstrating linkage of a marker and disease, these pairs do not add power, in fact they reduce the power, of the chi-square tests of goodness-of-fit of modes of inheritance.     

Evidence that BCL3 plays a role in the etiology of nonsyndromic oral clefts in Brazilian families

The BCL3 gene has been considered a susceptibility locus for nonsyndromic cleft lip with or without cleft palate (NSCL/P), based on association and linkage studies in some populations. We evaluated an intragenic marker at the BCL3 gene and the microsatellite D19S178 (1.1 cM distant from the BCL3 gene) among 98 infants born with NSCL/P and their parents, using the transmission disequilibrium test (TDT) and a method for haplotype analysis. Our analysis, based on BCL3 alleles, revealed the existence of a marginal association of allele 135pb of the BCL3 gene with NSCL/P (chi(2)=3.60; P=0.058; 1 df), with a major effect in female (chi(2)=5.77; P=0.016; 1 df) and in familial cases (chi(2)=3.79; P=0.051; 1 df). However, the haplotype analysis detected no significant segregation distortion, even if the alleles of the D19S178 were grouped into two classes. These findings support previous findings that BCL3 plays a role in the etiology of NSCL/P as an allele of low penetrance or as a modifier locus. We hypothesize that there might be more than one mutation in this gene associated with NSCL/P, or alternatively, that more than one mutation has arisen associated with the 135-bp allele. Genet. Epidemiol. 23:364-374, 2002 Copyright 2002 Wiley-Liss, Inc.     

Efficiency robust statistics for genetic linkage and association studies under genetic model uncertainty

When testing genetic linkage and association, test statistics that follow a normal or Chi‐square distributions are often used. These statistics are usually derived under a specific mode of inheritance (genetic model). Common genetic models include, but not limited to, the recessive, additive, multiplicative, and dominant models. For many diseases, their underlying genetic models are often unknown. Instead, a family of scientifically plausible genetic models may be available, which includes the four commonly used models. Hence, the optimal test is not available. Employing a single test statistic which is optimal for one model may suffer from substantial loss of power when the model is misspecified. In this situation efficient robust tests are useful. In this tutorial, we first review several commonly used robust statistics, including maximin efficency robust tests, maximal tests, and constrained likelihood ratio tests for three common designs in genetic studies: (i) linkage analysis using affected sib‐pairs, (ii) association studies using parents–offspring trios, and (iii) case–control association studies (unmatched and matched). Codes in the R statistical language for applying these robust statistics to test for linkage and association are presented with examples. We also provide some comparisons of the performance of the various robust tests via simulation studies. Guidelines for applications are also given for each study design. Finally, applications of robust tests to genome‐wide association studies and meta‐analysis are discussed. Copyright © 2009 John Wiley & Sons, Ltd.     

Does employment improve the health of lone mothers? The ALSPAC Study Team. Avon Longitudinal Study of Pregnancy and Childhood.

In Britain the government is currently proposing legislation that will encourage welfare recipients to gain employment. A central tenet of this 'welfare to work' policy is that employment will not only reduce the poverty of welfare recipients, but also improve their health. This research assessed the extent to which the movement from 'welfare to work' is likely to benefit the mental and physical health of lone mothers with preschool children. The sample was 719 lone mothers and a comparison group of 8779 women with partners drawn from the Avon Longitudinal Study of Pregnancy and Childhood (ALSPAC). Data collected by self completion questionnaire at 33 months postpartum provided information about average weekly take home family income and the mother's employment status. The health outcomes measured were general well being, both minor and major depression (using the Edinburgh Postnatal Depression Scale), self report of respiratory symptoms (cough/cold, wheeze, influenza) from 18-33 months postpartum and self report of symptoms common in the childbearing years (backache, haemorrhoids) also from 18-33 months postpartum. Lone mothers who were not employed were the poorest group in the sample; 94% of this group (402) had a family income of less than pound sterling 200 per week, compared with 72% (188) of lone mothers who were employed, 25% (905) of partnered women who were not employed and 12% (466) of partnered women who were employed. Lone mothers were significantly more likely than women with partners to report poorer well being (chi2 = 11.7, df = 3, P = 0.01), to have a major depressive disorder (chi2 = 92.6, df = 1, P = 0.0001) and to report wheeze (chi2 = 31.1, df = 1, P = 0.0001), but significantly less likely to report cough/cold (chi2 = 9.9, df = 1, P = 0.0001) or haemorrhoids (chi2 = 16.6, df = 1, P = 0.0001). Lone mothers who were unemployed and living on less than pound sterling 100 per week were significantly more likely to be depressed (chi2 = 3.9, df = 1, P = 0.05) than those who were employed and living on pound sterling 200 or more per week, and significantly less likely to report cough/cold (chi2 = 3.8, df = 1, P = 0.05). Logistic regression analyses showed no significant independent association between employment and better health for lone mothers. Rather, when compared with lone mothers who were not working, those who were employed were more likely to report minor respiratory symptoms such as cough/cold (OR = 1.51, 95% CI = 1.00,2.31). Overall, the results suggested that the movement from 'welfare to work' is unlikely to improve the health of lone mothers.     

Prognostic factors for the survival time in gallbladder carcinoma]

Survival of 74 patients with proven gallbladder carcinoma was studied retrospectively. Initial treatment consisted of cholecystectomy, cholecystectomy and common bile duct exploration, bypass procedure (hepatojejunostomy) or laparotomy and biopsy or percutaneous biliary drainage. Staging (according to the American Joint Committee on Cancer) was the most important determinant for survival (chi 2 = 29.34; df 3; p less than 0.001). Most individual complaints such as pain (chi 2 = 1.66;), nausea/vomiting (chi 2 = 0.35), and palpable mass (chi 2 = 2.62) were not related significantly with survival. However cachexia (chi 2 = 17.12; df I; p less than 0.001) was correlated with decreased survival. Establishing the diagnosis preoperatively was associated with significantly shorter survival (chi 2 = 22.67; df I; p less than 0.001) than establishment at the moment of hospitalisation. In a hospital where a radical surgical treatment is not performed, less invasive non surgical therapy (insertion of an endoprosthesis) could probably be considered for this group of patients. Selected patients should be referred to a centre for a radical surgical approach.     

An efficient integrative resampling method for gene–trait association analysis

Genetic association studies are popular for identifying genetic variants, such as single nucleotide polymorphisms (SNPs), that are associated with complex traits. Statistical tests are commonly performed one SNP at a time with an assumed mode of inheritance such as recessive, additive, or dominant genetic model. Such analysis can result in inadequate power when the employed model deviates from the underlying true genetic model. We propose an integrative association test procedure under a generalized linear model framework to flexibly model the data from the above three common genetic models and beyond. A computationally efficient resampling procedure is adopted to estimate the null distribution of the proposed test statistic. Simulation results show that our methods maintain the Type I error rate irrespective of the existence of confounding covariates and achieve adequate power compared to the methods with the true genetic model. The new methods are applied to two genetic studies on the resistance of severe malaria and sarcoidosis.     

易感基因与环境因素的交互作用对女性系统性红斑狼疮发病的影响

目的 探讨中国长江以南汉族女性系统性红斑狼疮(SLE)患者细胞毒性T淋巴细胞相关抗原-4(CTLA-4)及程序性细胞凋亡-1(PDCD-1)基因多态与环境因素的交互作用.方法 采用单纯病例研究设计,收集258例符合条件的患者,应用聚合酶链反应-限制性片段长度多态(PCR-RFLP)技术检测CTLA-4与PDCD-1基因的单核苷酸多态性(SNPs),分别在显性、隐性、相加及相乘遗传模式下,选择Poisson对数线性模型估计基因-基因、基因-环境交互作用.结果 在PDCD-1基因的PD1.6位点,隐性及相加遗传模式下均存在GG基因型与紫外线暴露史的交互作用,OR值分别为3.714(95%CI:1.235～11.179)和3.199(95% CI:1.023～10.004).在CTLA-4-1722位点,显性遗传模式下,存在TT或TC基因型与紫外线暴露史的交互作用,OR值为4.874(95%CI:1.119～21.242);相乘遗传模式下存在PDCD-1基因的T等位基因与紫外线暴露史的交互作用,OR值为1.470(95%CI:1.047～2.065);而在相加遗传模式下,TT和TC基因型分别与紫外线暴露史存在交互作用,OR值分别为4.744(95% CI:1.037～21.737)和4.973(95% CI:1.110～22.287).结论 在中国长江以南汉族女性SLE患者中,紫外线暴露史分别与CTLA-4及PDCD-1基因多态存在交互作用,提示基因-环境交互作用可能与SLE有关.     

Maximizing association statistics over genetic models

The assessment of the association between a candidate locus and a disease may require the assumption of an inheritance model. Most researchers select the additive model and test the association with the Cochran-Armitage trend test. This test assumes a dose-response effect with regard to the number of copies of the variant allele. However, if there is reason to expect dominance or recessiveness in the effect of the variant allele, the heterozygous genotype may be grouped with one of the two homozygous, depending on the inheritance model, and a simple test on the 2 x 2 table can be used to assess independence. When the underlying genetic model is unknown, association may be assessed using the max-statistic, which selects the largest test statistic from the dominant, recessive and additive models. The statistical significance of the max-statistic has been previously addressed using permutation or Monte Carlo simulation approaches. We aimed to provide simpler alternatives to the max-test to make it feasible in large-scale association studies. Our simulations show that this procedure has an effective number of tests of 2.2, which can be used to correct the significance level or P-values. We also derive the asymptotic distribution of max-statistic, which leads to a simple way to calculate the significance level and allows the derivation of a formula for power calculations in the design of studies that plan to use the max-statistic. A simulation study shows that the use of the max-statistic is a powerful approach that provides safeguard against model uncertainty.