Designs and analyses for exploring the relationship of magnetic fields to childhood leukaemia: A pilot project for the Danish National Birth Cohort

Abstract

Pooled analyses have consistently displayed an association of childhood leukaemia with residential magnetic fields, even after attempting to account for study problems. In light of the serious possibility that the link is indeed causal, a study of possible joint effects (interactions) of magnetic fields and genetic cofactors on childhood leukaemia may be justifiable. Such a study would face serious obstacles of limited numbers for subgroup analysis. To address these obstacles, we describe a design and analysis strategy that combines multiphase (multistage) sampling, measurement error modelling, and Bayesian methods for subgroup analysis that incorporate information from earlier pooled analyses. Special attention is given to prior specification, which would be the potentially controversial element. The approach could be pilot tested on data from the Danish National Birth Cohort, but an informative study would require augmenting these data with case-control sampling. We conclude that the approach outlined may be of value not only for this topic but also in other studies of effects of rare exposures and genetic factors on rare outcomes.

Keywords

Bayesian methods childhood leukaemia magnetic fields multiphase sampling shrinkage estimation

Get full access to this article

View all access options for this article.

References

Greenland S. , Sheppard AR , Kaune WT , Poole C. , Kelsh MA A pooled analysis of magnetic fields, wire codes, and childhood leukemia. Epidemiology 2000;11:624-34.

Greenland S. The impact of prior distributions for uncontrolled confounding and response bias: a case study of the relation of wire codes and magnetic fields to childhood leukemia. J Am Statist Assoc 2003;98:47-54.

Greenland S. Multiple-bias modeling for analysis of observational data (with discussion). J R Statist Soc Ser A 2005;168:267-308.

Greenland S. , Kheifets L. Leukemia attributable to residential magnetic fields: results from analyses allowing for study biases. Risk Analysis 2006;26:471-82.

Ahlbom A. , Day N. , Feychting M. , Roman E. , Skinner J. , Dockerty J. , et al. A pooled analysis of magnetic fields and childhood leukaemia . Br J Cancer 2000;83:692-8.

Schuz J. , Svendsen A. , Linet M. , McBride M. , Roman E. , Feychting M. , et al. Night-time exposure to electromagnetic fields and childhood leukemia: an extended pooled analysis. Am J Epidemiol 2007;166:263-9.

Mejia-Arangure JM , Ortega-Alvarez MC , Fajardo-Gutlerrez A. , Perez-Saldivar ML , Flores-Lujano, Bernaldez - Rios R. , et al. Magnetic fields and acute leukemia in children with Down syndrome . Epidemiology 2007;18:158-61.

Mezei G. , Kheifets L. Selection bias and its implications for case-control studies: a case study of magnetic field exposure and childhood leukemia. Int J Epidemiol 2006;35:397-406.

Greaves MF , Wiemels J. Origins of chromosome translocations in childhood leukaemia. Nat Rev Cancer 2003 ;3:639-49.

10.

Mori H. , Colman SM , Xiao Z. , Ford AM , Healy LE , Donaldson C. , et al. Chromosome translocations and covert leukemic clones are generated during normal fetal development. Proc Natl Acad Sci USA 2002;99:8242-7.

11.

Greaves M. Childhood leukaemia. BMJ 2002;324:283-7.

12.

Olsen J. , Melbye M. , Olsen SF , Sørensen TIA , Aaby P. , Andersen AN , et al. The Danish National Birth Cohort - its background, structure and aim. Scand J Public Health 2001;29:300-7.

13.

Greenland S. , Lash TL Bias analysis. In: Rothman KJ , Greenland S , Lash TL , editors. Modern epidemiology , 3rd edn. Philadelphia: Lippincott , 2008. p 345-80.

14.

Greenland S. Regression modeling. In: Rothman KJ , Greenland S , Lash TL , editors. Modern epidemiology , 3rd edn. Philadelphia: Lippincott , 2008. p 418-55.

15.

Greenland S. Tests for interaction in epidemiologic studies: a review and a study of power. Statistics Med 1983;2:243-51.

16.

Greenland S. Avoiding power loss associated with categorization and ordinal scores in dose-response and trend analysis. Epidemiology 1995;6:450-4.

17.

Flanders WD , Greenland S. Analytic methods for two-stage case-control studies and other stratified designs. Statistics Med 1991;10:729-47.

18.

Breslow NE , Holubkov R. Maximum likelihood estimation of logistic regression parameters under two-phase, outcome dependent sampling . J R Statist Soc Ser B 1997;59:447-61.

19.

Robins JM , Rotnitzky A. , Zhao LP Estimation of regression coefficients when some regressors are not always observed. J Am Statist Assoc 1994;89:846-66.

20.

Little RJA , Rubin DB Statistical analysis with missing data, 2nd edn. New York: Wiley, 2002.

21.

Poole C. , Greenland S. , Luetters C. , Kelsey JL , Mezei G. The relation of socioeconomic status to childhood leukemia: a review of the literature (with comment). Int J Epidemiol 2006;35:370-85.

22.

Kavet R. , Zaffanella LE , Pearson RL , Dallapiazza J. Association of residential magnetic fields with contact voltage. Bioelectromagnetics 2004;25:530-6.

23.

Kavet R. Contact current hypothesis: summary of results to date. Bioelectromagnetics 2005;26 (Suppl 7):S75-85.

24.

Carroll RJ , Ruppert D. , Stefanski LA , Crainiceanu C. Measurement error in nonlinear models, 2nd edn. Boca Raton, FL: Chapman & Hall , 2006.

25.

Greenland S. When should epidemiologic regressions use random coefficients? Biometrics 2000;56:915-21.

26.

Greenland S. Principles of multilevel modelling. Int J Epidemiol 2000;29:158-67.

27.

Carlin B. , Louis TA Bayes and empirical-Bayes methods of data analysis, 2nd edn. New York: Chapman & Hall, 2000.

28.

Efron B. , Morris CN Data analysis using Stein's estimator and its generalizations. J Am Statist Assoc 1975; 70:311-19.

29.

Greenland S. Generalized conjugate priors for Bayesian analysis of risk and survival regressions . Biometrics 2003;59:92-9.

30.

Greenland S. , Schwartzbaum JA , Finkle WD Problems from small samples and sparse data in conditional logistic regression analysis. Am J Epidemiol 2000;151: 531-9.

31.

Greenland S. Variable selection and shrinkage in the control of multiple confounders (invited commentary). Am J Epidemiol 2008;167:523-9. Erratum : Am J Epidemiol 2008;167:1142.

32.

Greenland S. Bayesian perspectives for epidemiologic research. I. Foundations and basic methods (with comment and reply). Int J Epidemiol 2006;35:765-78.

33.

Greenland S. Bayesian methods for epidemiologic research. II. Regression analysis. Int J Epidemiol 2007 ;36:195-202.

34.

Greenland S. Prior data for non-normal priors. Statistics Med 2007;26:3578-90.

35.

Greenland S. Methods for epidemiologic analyses of multiple exposures: a review and a comparative study of maximum-likelihood, preliminary testing, and empirical-Bayes regression. Statistics Med 1993;12:717-36.

36.

Copas JB Extra uncertainty from incomplete data. Proceedings of the 21st International Biometric Conference (invited papers, July 2002) . International Biometric Society, Freiburg, Germany, p 129-36.

37.

Brown RC , Dwyer T. , Kasten C. , Krotoski D. , Li Z. , Linet MS , et al. Cohort profile: the International Childhood Cancer Cohort Consortium (I4C). Int J Epidemiol 2007;36:724-30.