Critical appraisal of methodological aspects in the recent meta-analysis evaluating the association between cyclin D1 G870A polymorphism and risk of breast cancer.