PLoS One. 2015 Aug 11;10(8):e0135190. doi: 10.1371/journal.pone.0135190. eCollection 2015.
Sanna Heikkinen 1, Janne Pitkäniemi 2, Tytti Sarkeala 1, Nea Malila 3, Markku Koskenvuo 4
PMID: 26263013 PMCID: PMC4532449 DOI: 10.1371/journal.pone.0135190
Abstract
Introduction: Role of hair dyes in the etiology of breast cancer has occasionally raised concern but previous research has concluded with mixed results. Remnants of prohibited aromatic amines have been found in many hair dye products, and elevated levels of DNA-adducts of these amines have been detected from breast epithelial cells of hair dye users. However, the IARC working group has concluded that there is inadequate evidence for carcinogenicity of personal hair dye use and limited evidence in experimental animals for carcinogenicity of hair colorants.
Material and methods: We investigated whether the use of hair dyes is associated with breast cancer risk in women. The study design was a retrospective population-based case-control study in Finland, with a self-administered questionnaire from 6,567 breast cancer patients, aged 22-60 years and diagnosed in 2000-2007, and their 21,598 matched controls. We report odds ratios (OR) with 95% confidence interval (95% CI) from a conditional logistic regression model applied to the frequency matched sets of cases and controls. Bias-adjusted odds ratios from the sensitivity analysis are also presented.
Results: After adjusting for potential confounders, the odds of breast cancer increased by 23% (OR: 1.23, 95% CI: 1.11-1.36) among women who used hair dyes compared to those who did not. In women born before 1950 an increase of 28% was noted (OR: 1.28, 95% CI: 1.10-1.48). We also observed a significant trend between the OR and cumulative use of hair dyes (P: 0.005). Bias-adjusted odds ratios varied between 1.04 and 2.50.
Conclusions: Our results suggest that use of hair dyes is associated with breast cancer incidence. The impact on public health may be substantial due to vast popularity of hair coloring in modern societies. It should be noted that regardless of all efforts, a possibility of bias cannot definitively be ruled out and use of a prospective design is warranted. Based on the present results, it may be concluded however that safety of hair dyes in relation to breast cancer cannot yet be fully acknowledged and lack of external safety assessment within the cosmetics industry is of major concern.