Ênio Pereira Telles, Caroline Martins Leal de Almeida, Marina Dos Santos Barreto, Jussara Secundo Dos Santos, Danielle Gomes Santana, Guilherme Rodolfo Souza de Araujo, Ana Amélia Moreira Lira, Paulo Ricardo Martins-Filho, Adriana Gibara Guimarães, Mairim Russo Serafini, Adriano Antunes de Souza Araújo
The increasing use of hair dyes reinforces the need to identify potential toxicological hazards associated with their constituents and formulations. These products may release aromatic amines that, after dermal absorption, can induce oxidative stress and cellular damage, with potential cytotoxic, mutagenic, and carcinogenic effects. We reviewed in vivo evidence on toxicological hazards associated with isolated hair dye ingredients and finished commercial formulations in experimental animal models. Searches were conducted in PubMed, Web of Science, Embase, and Scopus, without language or year restrictions. Eligible studies assessed toxicity related to hair dye ingredients or formulations; secondary studies were excluded. Risk of bias was assessed using the Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE) tool. Twenty-five articles were included. The evidence indicated multifactorial toxicity affecting cardiac, hepatic, renal, testicular, respiratory, dermatological, hematological, genotoxic, hormonal, and developmental endpoints. However, relevance varied according to exposure route, dose, species, model, publication period, and whether the tested material was an isolated ingredient or complete formulation. Topical studies were the most relevant to cosmetic exposure, whereas oral, subcutaneous, intravenous, and embryo immersion models primarily supported hazard identification under non-usual or high-bioavailability conditions. SYRCLE assessment showed a predominance of unclear risk of bias, with methodological weaknesses in nearly all studies. Hair dye ingredients and finished formulations can produce toxicological hazard signals under specific experimental conditions. Nevertheless, these findings should not be interpreted as direct estimates of real-world consumer risk. More standardized, exposure-relevant, and robust studies are needed to improve hazard characterization and support risk interpretation, particularly for occupational exposure scenarios.