U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. Some Aromatic Amines, Organic Dyes, and Related Exposures. Lyon (FR): International Agency for Research on Cancer; 2010. (IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, No. 99.)

Cover of Some Aromatic Amines, Organic Dyes, and Related Exposures

Some Aromatic Amines, Organic Dyes, and Related Exposures.

Show details

5Summary of Data Reported

5.1. Exposure data

5.1.1. Hair Dyes: Production, use, occupational exposure and exposure after personal use

Modern hair dyes may be classified as a) oxidative (permanent) hair dyes, b) semi-permanent and c) temporary dyes. Oxidative hair dyes represent approximately 80% of the market and consist of colourless primary intermediates (para-substituted aromatic amines) and couplers (meta-substituted aromatic amines and other compounds) that, in the presence of peroxide, form the dye by a chemical reaction. The concentration of oxidative hair-dye ingredients is approximately proportional to their degree of shade: dark colours tend to contain the highest concentrations of colouring ingredients (up to 3% primary intermediates during use) whereas light shades (blond) contain lower concentrations. Semi-permanent (direct) hair dyes contain colour molecules of low molecular weight, such as nitro-phenols, nitro-aminophenols and nitro-phenylenediamines. Temporary dyes contain direct hair dyes of high molecular weight, such as azo, triphenylamine and indamine colourants. A worldwide survey in 2005 showed the presence of 50 ingredients in oxidative, 43 in semi-permanent and 88 ingredients in temporary hair dyes. The most frequently used oxidative hair-dye ingredients are para-phenylenediamine, resorcinol, 2,5-diaminotoluene, para- and meta-aminophenol, 4-amino-2-hydroxytoluene, 4-amino-meta-cresol and 2-methyl-5-hydroxyethylaminophenol. The majority of oxidative hair-dye ingredients have been on the market since the 1930s. Most semi-permanent or temporary hair dyes have a much more limited use.

Occupational exposure studies found no or only traces of hair-dye ingredients in the air of hair salons, whereas measurable amounts were detected on hairdressers’ hands. The major occupational exposure pathway appears to be via skin contact, followed by dermal absorption. The same exposure pathway applies to personal (consumer) use of hair dyes.

Hair dyes have been subject to regulations in many countries, and the number of substances permitted for use in hair dyes has been restricted during the past 40 years; in 2007, 135 individual ingredients were no longer allowed in the European Union for use in hair dyes.

5.2. Human carcinogenicity data

5.2.1. Professional use of hair colourants

The Working Group reviewed the literature on cancer at several sites in hairdressers, barbers and beauticians. Many additional studies have been published since the previous review in 1993 (IARC Monograph 57). These include several case–control studies and a few cohort studies. Most data from cohort studies derive from linkage between census data and cancer registries in Scandinavian countries, with limited potential to adjust for confounding by important correlates of cancer risk, e.g., lifestyle and reproductive factors. The Working Group noted that the evidence mainly concerned exposures that occurred before 1980s, and often much earlier.

(a) Bladder cancer

The cohort studies indicated an increased risk for cancer of the urinary bladder among male hairdressers, but not among female hairdressers. In a large Scandinavian cohort of hairdressers, barbers, beauticians and other related workers identified in the 1970 census and followed-up for 20 years, there was a significant 50% increase in risk for bladder cancer in men and a non-significant 10% decrease in risk in women. Allowance for smoking was generally not possible, although results for lung cancer suggest that higher exposure to tobacco in hairdressers could not totally account for the bladder cancer excess.

More than 20 case–control studies were available for evaluation. Most of these, including three of the larger studies, reported increased risks in the range of 1.3–1.7 in male hairdressers. A pooled analysis of 11 studies conducted in six European countries found no significant increase in risk among male or female hairdressers. Overall, risks appeared generally lower for women than for men. The number of exposed subjects was generally small, and did not allow a reliable assessment of the risk by duration and period of exposure.

In view of the consistent yet modest increase in risk reported in studies of hairdressers and barbers, especially men, and in the absence of solid data on the relation between duration and period of exposure, the Working Group concluded that there was limited evidence of an increased risk for bladder cancer in hairdressers.

(b) Haematological malignancies

With regards to cancers of the haematological system, the heterogeneity in the diseases included and the differences in the classification used often hampered comparison between the results of different studies. Although one cohort study of barbers among male US veterans and an Italian case–control study reported significant increases in risk for multiple myeloma based on only few exposed cases, these results were not replicated in other studies. In the large Scandinavian cohort, no excess was found in either sex for multiple myeloma or for other haematological malignancies.

(c) Breast cancer

Many studies on breast cancer, including the largest case–control and cohort studies, did not show any increased risk for breast cancer associated with professional use of hair colourants.

(d) Childhood cancers

One cohort study and five case–control studies investigated the risk for childhood cancers in the offspring of hairdressers and barbers. Although some positive associations were reported, an overall evaluation is difficult because of the different sites and/or histologies investigated in various studies and the problems in the identification of the relevant period of exposure (before or around conception, or during pregnancy).

(e) Other sites

(i) Ovarian cancer

A modest increase in risk for ovarian cancer was reported in two cohorts of US cosmetologists and Scandinavian hairdressers, which was significant only in the latter study. The excess in risk appeared stronger in—or limited to—women exposed in earlier periods. No case–control study was available for evaluation. The lack of adjustment for potential confounders, especially reproductive history and oral contraceptive use, does not allow confounding to be ruled out.

(ii) Lung cancer

Small increases in lung cancer risk of the order of 20–40% were found in most cohort studies, which did not, however, adequately adjust for smoking. A higher prevalence of smokers among hairdressers than in the general population was reported in Scandinavia and the USA No informative case–control study was available for evaluation. The Working Group concluded that tobacco smoking cannot be excluded as a likely cause of the modest excess in lung cancer observed in hairdressers.

5.2.2. Personal use of hair colourants

The Working Group revised and evaluated the epidemiological evidence of an association between cancer at several sites and personal use of hair dyes.

(a) Bladder cancer

Several studies with contradictory results have been published. Increased risks for bladder cancer were reported in two studies in the USA, while no association was found in three larger studies, two from the USA and one from Spain. These recent studies had similar characteristics in design and methodology.

One study from the USA suggested an increased risk for bladder cancer among users of hair colourants, in particular among those who exclusively used permanent hair dyes. Further, this study showed that exclusive use of permanent dyes among subjects with slow acetylation (NAT2 genotype) or among CYP1A2 slow metabolizers was associated with an increased risk for bladder cancer. The other studies did not confirm these results. In the Spanish study, there was no indication of an increased risk for bladder cancer associated with the NAT2 genotype, but there was a non-significant association with NAT1*10. The available cohort studies were largely negative for bladder cancer. The available meta-analyses did not show an association.

The Working Group considered that the available evidence for cancer of the bladder was overall inadequate.

(b) Haematological malignancies

The results for this tumour type were difficult to interpret: many different malignancies are involved, and many of the studies do not provide analyses for the different disease entities. Historically, the results have been inconsistent in identifying an increased risk. While cohort studies were largely negative for haematological malignancies, the results of case–control studies varied greatly. In those that showed an increased risk, the increase tended to be moderate. A recent pooled analysis was evaluated with particular interest because it was a large study evaluating hair-dye exposure in relation to single lymphoma entities including case–control data derived from Canada, the USA and six countries in Europe. The study showed an overall increased risk of 1.1 among women who were regular users, and of 1.3 among those women who had started regular hair-dye use before 1980. The risk was consistently elevated for follicular lymphoma and chronic lymphocytic leukaemia, but not for other types. For these lymphoma subtypes, the risk did not vary by intensity, years of use or type of exposure, remaining generally of the order 1.2–1.4. When the period of first use was considered, the increased risk for chronic lymphocytic leukaemia was mainly observed among those who started use before 1980, with a statistically significant increase in risk among those reporting use for more than 20 years. For follicular lymphoma, increased risks were observed throughout the two study periods. Overall, the Working Group considered this evidence to be inadequate.

(c) Breast cancer

For breast cancer the Working Group considered the evidence as inadequate based on several studies, none of which except one showed an association.

(d) Childhood cancers

For childhood cancers, the studies evaluated dealt with childhood brain tumours and Wilms tumours. The Working Group discussed in depth the potential biases and study limitations that could explain some of the increased risks observed for some brain tumours, and considered that some of the reported associations could not be simply explained by recall bias, as mothers may not have known about this hypothesis at the time of the studies. The Working Group considered that the evidence presented in the studies was inadequate.

(e) Other sites

Only a few studies were available to the Working Group, and no evaluation was made.

5.3. Animal carcinogenicity data

Various commercially available hair-dye formulations and various laboratory preparations of hair dyes were tested for carcinogenicity in mice or rats by skin application in 11 studies and by subcutaneous injection in a single study in rats.

In three studies by skin painting in mice, all using different formulations, increased incidences of lymphomas were observed in female mice compared with concurrent controls, for five different formulations. The increased incidences were not significant when compared with historical controls. In three studies by skin painting in rats, a significant increase in the incidence of mammary adenomas in females was observed for two formulations, and a significant increase in pituitary adenomas was seen in females for one formulation and in males for a different formulation. In the single subcutaneous injection study in rats, an increased incidence in mammary and uterine tumours was observed. The other studies either showed no increased incidence of tumours at any site or were inadequate for evaluation.

5.4. Other relevant data

Studies that investigated the induction of chromosomal aberrations in peripheral blood lymphocytes of professional hair-colourists or in volunteers who had their hair dyed reported no effect. The same is true for two studies that investigated sister chromatid exchange in lymphocytes of hairdressers. Two studies on the mutagenicity of urine collected from hair-dye users and cosmetologists were inadequate for evaluation. One study assessed sister chromatid exchange, DNA breakage (measured by single-cell gel electrophoresis) in lymphocytes and mutagenicity in urine in professional hair-colourists. No effect was seen for any of these three endpoints. A study on micronucleus formation in hair-dye users did not show a difference with non-user controls.

In an early study, 90% of a large number of commercial oxidative hair-dye formulations were mutagenic in bacteria. In later studies, this percentage dropped to 0–50%. The hair-colouring product henna did not show bacterial mutagenicity in one study. When tested separately, its active ingredient was mutagenic.

The urine of rats skin-painted with oxidative hair-colouring products was mutagenic when collected during the first 24 hours. The mutagenicity disappeared afterwards.

No effects were seen in a heritable translocation assay in rats skin-painted twice weekly for ten weeks with a semi-permanent dye formulation.

©International Agency for Research on Cancer, 2010.
Bookshelf ID: NBK385436

Views

  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (4.2M)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...