Genetic associations in large versus small studies: an empirical assessment.

摘要: Summary Background Advances in human genetics could help us to assess prognosis on an individual basis and optimise the management of complex diseases. However, different studies same genetic association sometimes have discrepant results. Our aim was how often large arrive at conclusions than smaller studies, whether this situation arises more frequently when findings first published disagree with those subsequent research. Methods We examined results 55 meta-analyses (579 study comparisons) associations tested magnitude effect differs versus studies. Findings noted significant between-study heterogeneity 26 (47%) meta-analyses. The differed significantly ten (18%), 20 (36%), 21 (38%) tests rank correlation, regression SE, inverse variance, respectively. largest generally yielded conservative complete meta-analyses, which included all (p=0·005). In 14 (26%) metaanalyses proposed stronger Only nine (16%) replicated without hints or bias. There little concordance discrepancies, small discrepancies. Interpretation Genuine bias affect Genetic risk factors for diseases should be assessed cautiously and, if possible, using scale evidence.