Risk Analysis in Cohort Studies with Heterogeneous Strata. A Global χ2-Test for Dose-Response Relationship, Generalizing the Mantel-Haenszel Procedure

W. Ahlborn; H.-J. Tuz; K. Überla

doi:10.1055/s-0038-1636366

Methods of Information in Medicine, Table of Contents

Methods Inf Med 1990; 29(02): 113-121
DOI: 10.1055/s-0038-1636366

Statistical Analysis

Schattauer GmbH

Risk Analysis in Cohort Studies with Heterogeneous Strata. A Global χ²-Test for Dose-Response Relationship, Generalizing the Mantel-Haenszel Procedure

W. Ahlborn

¹Institute for Statistics and Econometrics, University of Göttingen, F.R.G

,

H.-J. Tuz

¹Institute for Statistics and Econometrics, University of Göttingen, F.R.G

,

K. Überla

²Institute for Medical Information Processing, Biometrics and Epidemiology, University of Munich, F.R.G

› Author Affiliations

Abstract

In cohort studies the Mantel-Haenszel estimator OR̂ ^MH is computed from sample data and is used as a point estimator of relative risk. Test-based confidence intervals are estimated with the help of the asymptotic chi-squared distributed MH-statistic χ² _MHS. The Mantel-exten-sion-chi-squared is used as a test statistic for a dose-response relationship. Both test statistics – the Mantel-Haenszel-chi as well as the Mantel-extension-chi – assume homogeneity of risk across strata, which is rarely present. Also an extended nonparametric statistic, proposed by Terpstra, which is based on the Mann-Whitney-statistics assumes homogeneity of risk across strata.

We have earlier defined four risk measures RR ^k _j (k = 1,2,...,4) in the population and considered their estimates and the corresponding asymptotic distributions. In order to overcome the homogeneity assumption we use the δ-method to get “test-based” confidence intervals. Because the four risk measures RR ^k _j are presented as functions of four weights g_ik we give, consequently, the asymptotic variances of these risk estimators also as functions of the weights g _ik in a closed form. Approximations to these variances are given.

For testing a dose-response relationship we propose a new class of χ²(1)-distributed global measures Ĝ^k and the corresponding global χ²-test. In contrast to the Mantel-extension-chi homogeneity of risk across strata must not be assumed. These global test statistics are of the Wald type for composite hypotheses. The Mantel-extension-chi is a special case of the global-chi statistic and it is further shown that the Mantel-extension-chi can be expressed as a special weighted Terpstra statistic. Formulas for computing estimators of the global measures are provided. Three elaborated examples with hypothetical data of varying structure show, that the Mantel-extension-chi is systematically biased, generally overestimates the dose-response relation and leads to wrong conclusions, when heterogeneity is present. This is consistent with our theoretical considerations. In case of heterogeneity, when one wants to test an association between exposure and effect, or a dose-response relationship, the new global-χ²-test should be used in cohort studies.

Key-Words

Cohort Studies - Risk Measures - Heterogeneous Strata - Mantel-Haenszel-chi - Global-χ²-Test - Dose-Response Relation

Full Text

References

REFERENCES
1 Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959; 22: 719-48.
2 Mantel N. Alternative estimators of the common odds ratio for 2 × 2 tables. Biometrics 1986; 42: 199-201.
3 Ahlborn W, Tuz HJ, Überla K. Estimating relative risk from heterogeneous strata. Meth Inform Med 1988; 27: 118-24.
4 Greenland S, Robins JM. Estimation of a common effect parameter from sparse follow-up data. Biometrics 1985; 41: 55-68.
5 Breslow NE. Odds ratio estimators when data are sparse. Biometrika 1981; 68: 73-84.
6 Breslow NE, Day NE. Statistical Methods in Cancer Research (Vol. 2) – The Design and Analysis of Cohort Studies. Lyon: IACR Scientific Publications No; 82 1987
7 Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic Research: Principles and Quantitative Methods. Belmont Calif: Lifetime Learning Publications; 1982
8 Kilpatrick SJ, Viren J. Age as a modifying factor in the association between lung cancer in non-smoking women and their husbands smoking status. In: Perry R, Hirk PW. eds. Indoor and Ambient Air Quality. London: Selper Ltd; 1988: 195-202.
9 Rao CR. Linear Statistical Inference and its Applications. 2nd ed.. New York: Wiley; 1973
10 Bishop YMM, Fienberg SE, Holland PW. Discrete Multivariate Analysis: Theory and Practice. London: The M.I.T. Press; 1975
11 Heuser H. Lehrbuch der Analysis (Teil 2). Stuttgart: Teubner; 1983
12 Tuz HJ. Die Analyse von Kontingenzta-feln mit AssoziationsmaBen bei Heterogenitat der Schichten. Göttingen: Unitext-Verlag; 1988
13 Kendall MG, Stuart A. The Advanced Theory of Statistics. Vol. 1, 3rd ed.. London: Griffin; 1977
14 Terpstra TJ. The asymptotic normality and consistency of Kendall’s test against trend, when ties are present in one ranking. Indagationes Math 1952; 14: 327-33.
15 Wald A. Test of statistical hypothesis concerning several parameters when the number of observations is large. Trans Am Math Soc 1943; 54: 426-82.
16 Hayakawa T. The likelihood ratio criterion for a composite hypothesis under a local alternative. Biometrica 1975; 62: 451-60.
17 Mantel N. Chi-square tests with one degree of freedom: Extension of the Mantel-Haenszel procedure. J Am Stat Assoc 1963; 58: 690-700.
18 Karon JM. The effect of scoring methods on the chi-square for contingency tables with ordinal classes. San Diego Calif: Annual Meeting of the American Statistical Association; August 1978
19 Lehmann EL. Nonparametrics. New York: McGraw-Hill; 1975
20 Hemelrijk J. Note on Wilcoxon’s two-sample test when ties are present. Ann Math Stat 1952; 23: 133-5.
21 Kendall MA. Rank Correlation Methods. London: Griffin; 1970