Premium
Aspirin use and breast cancer risk: a meta‐analysis and meta‐regression of observational studies from 2001 to 2005
Author(s) -
Mangiapane Sandra,
Blettner Maria,
Schlattmann Peter
Publication year - 2008
Publication title -
pharmacoepidemiology and drug safety
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.023
H-Index - 96
eISSN - 1099-1557
pISSN - 1053-8569
DOI - 10.1002/pds.1503
Subject(s) - medicine , aspirin , breast cancer , meta analysis , relative risk , observational study , cohort study , random effects model , oncology , epidemiology , population , covariate , pharmacoepidemiology , cancer , demography , confidence interval , statistics , environmental health , pharmacology , mathematics , sociology , medical prescription
Purpose To examine the recent epidemiological studies on aspirin use and breast cancer risk published from 2001 to 2005 within a meta‐analysis, to investigate reasons for heterogeneity between the individual studies and to analyse a dose‐response‐relationship considering frequency and duration of use. Methods We systematically searched for cohort‐studies and case‐control‐studies from 2001–2005, which evaluated the association between aspirin and breast cancer risk. We calculated a pooled estimate for the relative risk (RR) and investigated reasons for heterogeneity between the individual studies and analysed a dose‐response‐relationship using random effects mixed models. Results We identified 10 studies which met the inclusion criteria. The combined estimate of the RR was 0.75 (95%CI: 0.64, 0.88) using the random effects model. Heterogeneity between the studies could not be explained by the covariates study‐type and study‐population. The combination of frequency and duration of aspirin use resulted in a significant dose‐response‐relationship between aspirin use and breast cancer risk. Each additional pillyear reduced the breast cancer risk to about 2%. Conclusion Our meta‐analysis supports the current evidence that aspirin may reduce breast cancer risk. Moreover, a dose‐response‐relationship seems to exist. However, results have to be interpreted carefully, as exposure categories were defined very heterogeneously among the studies which weakens the validity of the pooled estimates. Copyright © 2007 John Wiley & Sons, Ltd.