RESUMO
Objective: We aimed to summarize the up-to-date epidemiology evidence on biomarkers of long-chain (LC) n-3 fatty acid (FA) intake in relation to breast cancer (BC).Methods: Epidemiology studies determining FA levels in biospecimen (circulating blood or adipose tissue (AT)) were identified from PubMed, EMBASE, and Cochrane Library databases until March 2018. Multivariate-adjusted risk ratios (RRs) with 95% confidence intervals (CIs) were pooled using a random-effect model. Difference in biospecimen proportions of LC n-3 FA between BC cases and non-cases were analyzed as a standardized mean difference (SMD).Results: Thirteen cohort and eleven case-control studies were eligible for the present meta-analysis. The estimated SMD was -0.14 (95% CI: -0.27, -0.11) for LC n-3 FA and -0.27 (95% CI: -0.42, -0.11) for LC n-3/n-6 FA ratio. When comparing the top tertiles with the bottom baseline levels, circulating LC n-3 FA was significantly associated with a lower risk of BC (RR: 0.84, 95% CI: 0.74, 0.96), but not AT (RR: 1.02, 95% CI: 0.70, 1.48). Significant inverse dose-response associations were observed for each 1% increment of circulating 20:5n-3 and 22:6n-3.Conclusion: This meta-analysis highlights that circulating LC n-3 FA as a biomarker of intake may be an independent predictive factor for BC, especially 20:5n-3 and 22:6n-3.