Association between hair dye use and human cancers: A systematic review.

Introduction
The global hair color market is valued over 23 billion dollars with over 2 billion in sales in the United States. Asia dominated the market accounting for over 35% of global revenue in 2021. Hair dyes are classified by permanence: temporary, semipermanent, and permanent hair dye (PHD).1,2 PHDs account for approximately 80% of hair dye products on the market and contain coupler m-aminophenols that oxidizes and penetrates the hair cortex. Whereas, temporary and semipermanent do not oxidize therefore weakly adhere to the hair shaft.2 As a result, more consumers use PHD because it provides a more natural hair color than the other hair dyes.
The International Agency for Research on Cancer classifies occupational exposure to hair dyes as a probable carcinogen, though personal use remains unclassified.1 Certain chemicals found in hair dyes: 2,4-diaminoanisole, 4-amino-2-nitrophenol, and 4-chloro-o-phyenylenediamine are banned in Europe. Studies suggest certain chemicals including paraphenylenediamine (PPD) may be carcinogenic.3 This systematic review assesses the association between hair dye use including type, color, and cancer risk along with public health implications.

Methods
We conducted a systematic search using PubMed and MEDLINE with the Medical Subject Headings: hair dyes, hair colorants, hair color agents, cancer, and neoplasm. This systematic review protocol was registered at PROSPERO (CRD420251076719). We limited the search to English, human studies published between January 1964 and March 2025. Inclusion and exclusion criteria were specified before the literature search. To be included, the original articles needed to fulfill the following criteria: full-text, human studies investigating hair dye exposure as a risk factor for any type of cancer. We included clinically relevant randomized control, case control, prospective, and cohort studies. Exclusion criteria were non-English studies, secondary articles, letters to editors, commentaries, and experimental studies. Study design and outcome data were extracted and summarized in accompanying tables. We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (Fig 1), and graded studies according to the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence.4 The Risk of Bias in Non-randomized Studies of Interventions tool was used to assess study risk, categorizing each as low, moderate, serious, critical, or no information (Table I, Table II, Table III, Table IV, Table V, Table VI).
Three reviewers (J.M.,C.N.,G.B.) read abstracts to determine eligibility for inclusion in this systematic review. For each study, we recorded the study year, study design, level of evidence (LoE), study population and study findings.

Results

Study selection
The literature search yielded 413 studies that met search criteria (Fig 1). After screening, 115 publications were reviewed for full-text eligibility with 2 more studies added based on bibliographic evaluation. Nineteen articles were excluded leaving 96 articles for inclusion. Thirteen articles are prospective cohort studies and 83 case-control studies. For the included studies, the LoE ranged from 2b to 3b (Fig 2).
Of the 96 studies, 87 evaluated adult patients, 2 included both adult and children, and 5 focused on maternal exposure and pediatric cancer risk. Analysis showed various cancers potentially associated with hair dye use: breast, bladder, lymphoma, hematopoietic, ovarian, childhood cancers, melanoma, prostate, salivary gland, glioblastoma, tenosynovial giant cell tumor, and soft tissue sarcoma (Table I).

Breast cancer
Most studies (n = 16, LoE:2b-3b) assessed the relationship between hair dye use and breast cancer (Table II). Overall, most studies report a direct association except for several cohort studies reporting associations in specific subgroups.
Cancer association with hair dye use varied by race particularly for PHD.5 Among African American (AA) women using hair dye every 5-8 weeks, breast cancer risk increased by 60% particularly in those using black colored PHD. In contrast, light colored-PHD was associated with significantly increased breast cancer among White women.5 Hair straightener use varies by race with AA women report using hair straighteners more frequently. Amongst this group, using hair straighteners every 5-8 week is associated with a 31% increased risk of breast cancer. Additionally, AA women have a 72% increased risk of estrogen receptor-positive disease using dark hair dye shades.6
Frequency and duration of hair dye use influences breast cancer risk. The cumulative frequency of semipermanent and PHD use is associated with increased risk of breast cancer, with odds ratios rising from 1.07 for 1-2 uses to 1.31 for 35-89 uses. Starting hair dye use at an earlier age, 20-29 years is associated with increased odds of breast cancer compared to those who began after age 40.7,8
The type and combination of hair coloring products are associated with increased breast cancer risk. Women using 2 or more types of hair dye products had a 3-fold increased risk if they accumulated 90 or more applications in their lifetime. Conversely, breast cancer risk is not strongly related among women who dyed their hair after bleaching or used frosting or rinses prior to dye application.8 Overall, breast cancer risk may be elevated among AA women, frequent and long-term users, and those using dark-colored or permanent dyes, particularly in relation to estrogen receptor-positive disease.

Bladder cancer
Some studies (n = 11; LoE:2b-3b) found no significant association between bladder cancer and hair dye use (Table III). However, the literature presents inconsistent findings with many studies limited by small sample sizes and population-based design.
A larger population-based study reports an increased risk of bladder cancer in both women and men with ever using black hair dye.20 Among women using PHD for over 31 years or over 282 applications have an increased risk of bladder cancer. On the other hand, men using any hair dye was inversely associated with any bladder cancer (odds ratio = 0.5; 95% CI = 0.3-0.8).21
Genetic susceptibility appears to modify the association between hair dye use and bladder cancer. Two studies report increased risk among individuals carrying slow acetylation N-acetyltransferase 2 (NAT2) genotypes or specific CYP1A2 genotypes. Exclusive PHD use was strongly associated with increased bladder cancer risk among individuals with the NAT2 slow acetylation genotype and those lacking the NAT1∗10 genotype.22 In contrast, no increased risk was observed among NAT2 fast acetylators.22 Similarly, PHD use was associated with a 2.5-fold increased bladder cancer risk among women with the slow CYP1A2 genotype, an association not observed in the rapid genotype.23,91

Prostate cancer
Two studies (LoE = 3b) examine the association between hair dye use and prostate cancer (Table III). A study in Taiwan evaluated 296 prostate cancer cases with matched controls.37 Both studies did not distinguish between permanent, semipermanent, or temporary hair dyes. Exposure was assessed only as ever use versus never use. They report more than a twofold increased risk, particularly among men under 60 years of age who used hair dyes for over 10 years, applied them at least 6 times per year, and began use before 1980.37 Conversely, a Finnish cohort found no significant association between hair dye use and prostate cancer incidence or mortality.38

Gynecological cancer
Several studies (n = 3; LoE: 2b-3b) suggest an association between hair product use and gynecologic cancers (Table III).26 Women who used PHDs within the last 12 months had increased risk of nonserous ovarian tumors. Frequent use of hair straighteners was also reported within this cohort representing a potential confounding variable.40 Multiple studies included in this review identified a higher risk of ovarian cancer with frequent hair straightener use (≥4 times annually).39,40 Additionally, occupational exposure to hair chemicals may be relevant. Female hairdressers have a higher risk of ovarian cancer suggesting that cumulative exposure in a professional setting may contribute to gynecological cancer.40

Hematopoietic cancers

Leukemia
An association between hair dye use and leukemia risk has been reported in multiple studies (n = 11; LoE: 2b-3b) with variations in leukemia subtype and exposure patterns (Table IV).
PHD was associated with an increase in overall leukemia risk among long-term users (>20 years).24 Case-control studies reported stronger associations with chronic lymphocytic leukemia (CLL). Dark PHD increases CLL risk in females.62 Only CLL was significantly associated with the use of hair dyes and the risk of CLL increases with lifetime doses received.52 Furthermore, the risk of CLL increased among women using hair dyes before 1980.24,52 Hair dye use was associated with a 50% increased risk of developing acute lymphoblastic leukemia (ALL) with risks more than doubling among users with over 15 years of exposure.44
Conversely, some studies (n = 9; LoE:2b-3b) did not observe significant associations between hair dye use and certain leukemia subtypes including acute myeloid leukemia, and ALL highlighting inconsistencies in the current literature.25,45, 46, 47, 48, 49,60

Myelodysplastic syndrome
Two case-control studies (LoE:2b) found an association between the risk of myelodysplastic syndrome and hair dye use(Table IV). Both studies did not distinguish between permanent, semipermanent, or temporary hair dyes. Exposure was assessed only as ever use versus never use. Risk was further elevated among individuals with both hair dye use and occupational exposure to organic solvents suggesting a synergistic effect.58 Neither study specified the color or type of dye used. The association between hair dye and myelodysplastic syndrome remained after adjusting for confounders like smoking and benzene exposure.59

Hodgkin’s lymphoma
The association between hair dye and risk of Hodgkin’s lymphoma (HL) (n = 5; LoE:2b-3b) largely reported no significant relationship (Tables IV and V). A 32-year prospective cohort study of women with biennial follow-up showed hair dye use was not associated with increased HL risk.46 Similarly, a case-control study reported elevated HL risk among PHD users with a comparable risk increase observed for those using dark dye shades odds ratio 3.4 vs 3.6, respectively.67 These findings suggest that while isolated studies have reported elevated risks with older formulations, the overall evidence does not support a consistent association between hair dye use and HL risk.25,42,50,51,60,61

Non-Hodgkin’s lymphoma
Some studies (n = 11; LoE:2b-3b) suggest a potential elevated risk of non-Hodgkin’s lymphoma (NHL) with hair dye use (Table V). Eight found no significant association while 4 studies noted increased risk among users before the 1980 formulations.52,54,60, 61, 62,64, 65, 66, 67, 68, 69,92
Black dye use and semi-PHD exposure were associated with elevated risk, particularly for diffuse large B-cell lymphoma.9,48,54,92 Using hair dye for at least 15 years is correlated with increased lymphoid malignancy risk among women.48 Follicular lymphoma, a subtype of NHL has been associated with the use of dark hair dyes. Increased risk was observed among individuals with exposure prior to 1980 and those with over 20 years of cumulative use.54,68,69,92
Four studies explored how genetic variation may modify the risk of NHL associated with hair dye use.68,70, 71, 72 Women with rapid or intermediate NAT2 genotype have significantly elevated NHL risk.68 Likewise, BRCA2 polymorphisms similarly increased follicular lymphoma risk.70 Several subtypes of NHL showed hair dye use associated with t(14;18) translocation and bcl-2+ lymphoma.71,72 Overall, NHL risk elevates with darker hair shades, pre-1980 formulations, long term use, and in genetically predisposed individuals.

Multiple myeloma
Most studies (n = 9; LoE:2b-3b) found no association between hair dye use and multiple myeloma (MM) (Table VI). However, the largest cohort study found an increased MM risk with dark hair dye use (relative risk = 4.39; 95% CI: 1.1-18.3). Although the confidence intervals of the study are wide. Another study reported elevated MM risk with semi-PHD use in women and men.25 The literature presents inconsistent findings with many studies limited by small sample sizes.

Cancers at other sites
Limited studies (n = 4; LoE:3b) have assessed PHD use and other cancers (Table VI). Using hair dyes for more than 15 years was linked to salivary gland carcinoma.87 Another study found a gene variant may be a key factor in renal cell carcinoma development when combined with hair dye use.54,61 Neither soft tissue sarcoma nor tenosynovial giant cell tumor noted a nonsignificant increased risk when using hair dye.54 A large cohort study observed ever usage of hair dyes reduced overall cancer mortality but elevated respiratory cancer risk among women with more than 20 years of use.93

Hair dyes use in pregnant women and incidence of childhood cancer
Several studies (n = 5; LoE:2b-3b) explored maternal hair dye use and childhood cancers (Table VI). During the first trimester, hair dye use significantly increased ALL risk in offspring with even higher risks associated with use during breastfeeding.43 Two studies reported increased neuroblastoma risk: 1 showed 60% higher risk with use the month of conception and/or during pregnancy.78 Another study found a twofold increase of neuroblastoma associated with pregnant women using hair dye and work-related exposure during preconception.81 The 3 other studies found no significant links with Wilms tumor, astroglial tumors, or primitive neuroectodermal tumors.77,79,80

Brain tumors in adults
Two case-control studies (LoE:3b) found hair dye use associated with glioma and glioblastoma multiforme (Table VI).79,80 Risk of developing glioma and glioblastoma multiforme was associated with PHD exposure for 21 years or more.82 Use of dark-colored hair dye for over 20 years has been associated to an elevated risk of acoustic neuromas.79,80

Melanoma
A 37-year prospective study reports an increased melanoma risk with PHD.82 In contrast, 2 earlier case-control studies found no significant associations between melanoma and hair dye exposure (Table VI).82,83 Too few studies are available on melanoma to make a conclusion whether personal hair dye use is associated with melanoma.

Discussion
Prior to 1980, hair dye formulations contained known carcinogens including aromatic amines used as dye intermediates. In response, manufacturers changed the ingredients of PHD to remove carcinogenic ingredients: 4-methoxy-m-phenylenediamine, 2,4-toluenediamine, and 2-nitro-p-phenylenediamine. Despite these changes, some current hair dyes contain PPD which may be carcinogenic.94, 95, 96 Additionally, DMDM hydantoin a formaldehyde releasing agent and known carcinogen remains in use for some current hair dye products.97 Occupational exposure also reflects this risk. Individuals working as a hairdresser for more than 10 years were nearly twice as likely to develop bladder cancer compared to those who never worked as a hair dresser.98 However, a recent study including Swedish hairdressers found no increased risk of bladder cancer in recent decades.99 This may suggest that modern hair dye formulations, which are free of aromatic amines may not confer the same level of risk. More research is needed to fully understand the long-term health effects of current hair dye formulations.
The relationship between hair dye use and cancer risk appears variable by cancer type (Table I). While the evidence for breast cancer, hematopoietic, and gynecologic cancers suggests elevated risks in specific groups. Certain populations may be disproportionately affected. Breast cancer risk was elevated among AA women and frequent users of dark PHDs. AA women demonstrated higher risks across breast and gynecological cancers, possibly reflecting combined exposures to hair dyes and chemical straighteners. Similarly, leukemia studies suggest subtype-specific risks with CLL risk rising with greater cumulative lifetime exposure. The greater number of studies on breast and bladder cancers may reflect early hypotheses from the 1970s linking aromatic amines and PPD in hair dye to endocrine and urothelial carcinogenesis.1 The relative paucity of studies on other malignancies represents the need to assess potential associations with less commonly studied cancers.
Genetic polymorphisms in detoxification enzymes such as NAT2 and CYP1A2 were shown to modify bladder cancer and NHL risk in susceptible individuals.68 This points to the interplay between genetic susceptibility and environmental exposure that may underlie cancer risk disparities and warrants further research.68 Among women rapid or intermediate NAT2 genotype have been associated with elevated NHL risk.68 While, BRCA2 polymorphisms have similarly been linked to an increased risk of follicular lymphoma.70 In bladder cancer, the use of PHDs was associated with increased risk among individuals with the NAT2 slow acetylation genotype and those lacking the protective NAT1∗10. It may be beneficial to incorporate genetic screening for patients with significant or long-term hair dye exposure. This review is limited by heterogeneous endpoints, small sample sizes, recall bias, retrospective data and older dye formulations.

Conclusion
Despite inconsistencies among the studies, the link between hair dye use and cancer risk merits further study. Additional studies are needed to evaluate the safety of current hair dye formulations and their long-term health effects.

Conflicts of interest
None disclosed.