Carcinogenesis, Vol. 23, No. 12, 2135,
December 2002
© 2002 Oxford University Press
LETTER TO THE EDITOR |
Proper controls for SNP studies?
Unit of Environmental Cancer Epidemiology, International Agency for Research on Cancer, 150 cours Albert-Thomas, 69008 Lyon, France
Email: brennan{at}iarc.fr, boffetta{at}iarc.fr
Dear Sir,
Hemminki and Försti are correct in commenting that ethnicity and age differences are unlikely to result in large biases in SNP frequencies between case and control groups. At least up until now, there are few instances of false positive findings which were later identified as having arisen from either of these sources of bias. We therefore concur with their recommendation not to waste resources in tight matching of controls to cases. However, their proposal in favour of using large series of control groups recruited by some convenient fashion (e.g. blood donors) is also not the answer to reducing the high levels of inconsistency in results from SNP studies.
The extreme lack of consistency in study findings from SNP studies is well documented (1,2). This lack of consistency is likely to be due to (i) a mixture of publication bias and subgroup analysis leading to preferential publication of false positive findings; and (ii) the plethora of small underpowered studies leading to false negative results. The availability of large convenient control groups will do nothing to reduce the proportion of false positive results, and may even exacerbate the problem. Conversely, without increasing the sample size of cases it is unlikely that the prevalence of false negative results will be much reduced. Instead what is required are large well-conducted studies including representative series of incident cases and controls with good-quality lifestyle and environmental information to help investigate potential gene–environment interaction. In summary, we need better designed, larger studies, not control groups of convenience.
References
- Vineis,P., d'Errico,A., Malats,N. and Boffetta,P. (1999) Overall evaluation and research perspectives. In Vineis,P., Malats,N., Lang,M., d'Errico,A., Caporaso,N., Cuzick,J. and Boffetta,P. (eds) Metabolic Polymorphisms and Susceptibility to Cancer. IARC Scientific Publication No. 148, IARC, Lyon, pp. 403–408.
- Brennan,P. (2002) Gene–environment interaction and aetiology of cancer: what does it mean and how can we measure it? Carcinogenesis, 23, 381–387.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||