Hair Dye and Cancer Risk: Who May Face Higher Exposure

Hair dye is a mass exposure: the U.S. market generates about $23 billion annually, more than 2 billion products are sold each year, and permanent formulations account for roughly 80% of sales. Yet the best current evidence does not support a simple verdict that hair dye either causes cancer or is harmless. A 2025 systematic review of 96 studies found a more useful, less comfortable answer: risk signals vary sharply by cancer type, exposure pattern, and population.

The clearest concern involves Black women who use permanent dyes, especially alongside chemical straighteners. Occupational exposure among hairdressers also carries an International Agency for Research on Cancer (IARC) classification as a probable human carcinogen. For personal users overall, however, the evidence remains unresolved. Association is not causation, and six decades of observational research cannot erase that distinction.

A broad review, but not a final answer

Greene et al.'s 2025 systematic review covered studies published from 1964 through 2025 on bladder cancer, breast cancer, leukemia, gynecologic cancers, and other malignancies. The review rated most breast and bladder cohorts as having moderate risk of bias using ROBINS-I and assigned the evidence Oxford levels 2b to 3b.

Those labels matter because every included study was observational. Researchers cannot ethically randomize people to years of potentially carcinogenic cosmetic exposure, so the field relies on cohorts and case-control studies. Cohorts can track exposure before diagnosis, while case-control studies often depend on participants remembering products used years earlier. A cancer diagnosis can change how carefully someone recalls past dye or relaxer use, which can inflate associations.

Product chemistry adds another problem. Hair dyes studied in the 1960s and 1970s were not the same products sold today. Manufacturers removed confirmed bladder carcinogens 4-aminobiphenyl and 2-naphthylamine from formulations before 1980. Older positive studies may therefore measure hazards that do not transfer cleanly to contemporary dyes. Pooling chemically different eras can blur a real effect, create one, or do both at once.

Breast cancer: the most actionable signal

The review's most decision-relevant finding came from the Sister Study, a prospective cohort of 46,709 women followed for an average of 8.3 years, during which 2,794 developed breast cancer. Permanent hair dye use was associated with a 45% higher breast cancer hazard among Black women (hazard ratio [HR] 1.45), compared with a 7% higher hazard among White women (HR 1.07). Among Black women who used permanent dye frequently, the estimate rose to HR 1.60.

These are relative estimates, not an individual woman's absolute chance of developing breast cancer. The supplied results do not provide absolute risks or confidence intervals for these subgroup estimates, so they cannot tell a reader how many additional cases occurred per 1,000 users. They do show that the association differed substantially within the same prospective study rather than appearing uniformly across all users.

Chemical straighteners were associated with increased breast cancer risk independently: HR 1.18 overall and HR 1.31 with frequent use, defined as at least every five to eight weeks. A separate population-based case-control study by Llanos and colleagues included 2,280 breast cancer cases and 2,005 controls. African American women using dark permanent dye had higher odds of estrogen receptor-positive breast cancer (odds ratio [OR] 1.51); combined dye and relaxer use was associated with an OR of 2.40.

Another case-control study, by Heikkinen and colleagues, included 6,567 cases and 21,598 controls. Permanent dye use was associated with breast cancer at OR 1.23 (95% confidence interval 1.11 to 1.36), with a statistically significant dose-response trend (p=0.005). The confidence interval excludes 1.0, so this specific association met the usual threshold for statistical significance. It still cannot establish that dye caused the cancers. Product choice, other cosmetic exposures, occupational chemicals, socioeconomic conditions, and access to screening could help explain part of the pattern.

The racial disparity deserves careful attention without speculation disguised as certainty. Differences in formulation, use patterns, scalp health, dermal absorption, hormone-related pathways, metabolic susceptibility, or unmeasured social exposures could contribute. The available studies do not separate those explanations. They do justify a more serious clinical conversation than a blanket reassurance.

Bladder cancer remains genuinely uncertain

Bladder cancer has long dominated public worries about hair dye because aromatic amines can be processed by the body and concentrated in urine. Of 18 bladder cancer studies in the review, seven reported statistically significant positive associations and 11 reported no significant association. That is not a hidden consensus waiting to be discovered with better adjectives.

The negative studies carry weight. The Gubéran occupational cohort followed 20,336 men and 21,903 women for 28.9 years and found no causal relationship between hair dye exposure and excess bladder cancer risk. Across several studies, adjustment for smoking and occupational co-exposures weakened the apparent signal. Smoking matters greatly here: Pan et al. describe smoking as accounting for about half of bladder cancer cases and discuss aromatic amines among established bladder carcinogens.

The mechanism is plausible, but plausibility is not proof that modern personal dye use causes bladder cancer. Aromatic amine metabolites can reach the bladder through urine and contact the urothelium during storage. Bellamri et al. showed that aristolochic acid can form DNA adducts and cause mutational damage in human bladder cells. That study examined aristolochic acid, not PPD or hair dye. It supports the general biological vulnerability of bladder tissue to urinary carcinogens; it does not demonstrate a hair-dye-specific pathway in people.

IARC classifies occupational exposure as Group 2A, meaning probably carcinogenic to humans, while personal hair dye use remains unclassified. The distinction fits the exposure contrast: hairdressers encounter products repeatedly and at higher cumulative doses, whereas personal use is intermittent. Unclassified means the evidence cannot support a formal carcinogen classification. It does not mean proven safety.

Leukemia and the problem of uneven evidence

For leukemia, 11 of 20 studies reported positive associations and nine reported null results. Aromatic hydrocarbons have established hematotoxic potential, so the signal has biological credibility. But heterogeneous designs and inconsistent documentation of confounders prevent a reliable estimate of the magnitude of any risk.

Small studies can also distort the picture. One study with 200 participants reported an OR of 8.6 (95% CI 3.3 to 22.3), an estimate too imprecise to treat as a sensible guide to typical consumer risk. The review did not provide a formal pooled meta-analysis weighted by study size, so dramatic outliers should not set the emotional temperature of the evidence.

Practical choices that match the evidence

For Black women who use permanent dark dyes and chemical straighteners, reducing cumulative exposure is a reasonable precaution. Consider semi-permanent or temporary products, longer intervals between applications, and avoiding application to an irritated or damaged scalp. These steps have practical logic, but no study here proves how much they reduce cancer incidence.

Hairdressers have a clearer occupational-health case for nitrile gloves, effective ventilation, and less direct skin contact. Those are proportionate responses to a chronic exposure that IARC classifies as probably carcinogenic.

For other personal users, the evidence supports informed restraint rather than panic. Lower application frequency and less scalp contact are low-cost ways to reduce exposure while the cancer risk from current formulations remains hard to quantify. Eating cruciferous vegetables is also reasonable for general health. Kennelley et al. review evidence that sulforaphane activates phase II detoxification enzymes in bladder tissue, but no human study shows that broccoli or sulforaphane prevents hair-dye-related cancer. Nutrition is not a chemical permission slip.

The defensible conclusion is stratified: personal hair dye use has uncertain cancer risk overall, while breast-cancer associations in Black women and chronic occupational exposure deserve more attention than the usual one-size-fits-all reassurance.