Cosmetic Adverse Reactions and Prognostic Factors: A Retrospective Cohort Study in Wuhan, China

ABSTRACT

Background

Despite the increasing prevalence of cosmetic‐related adverse reactions, their potential prognoses and associated risk factors have received limited attention within the medical community.

Methods

A retrospective analysis was conducted on 7794 cases of cosmetics‐related adverse reactions reported in Wuhan, China, from 2017 to 2023. Data were collected from multiple sources, including medical institutions, cosmetic manufacturers, distributors, individual consumers and regulatory agencies. Demographic characteristics, clinical diagnoses, symptom profiles and allergy histories were systematically analysed using descriptive statistics. Cases were stratified by recovery outcome into complete versus incomplete groups to identify prognostic factors. Comparative analyses were performed using χ2 tests, with statistically significant variables (p < 0.05) further evaluated through multivariate logistic regression to calculate odds ratios (ORs) and 95% confidence intervals (CIs). All statistical analyses were performed using SPSS version 25.0, with a significance level set at p < 0.05.

Results

Among the 7794 patients, the majority (76.56%) achieved complete recovery, while a substantial proportion (23.44%) experienced persistent or permanent symptoms, such as post‐inflammatory hyperpigmentation (PIH), atrophic scarring and other forms of skin damage. Statistical analysis identified age and a history of allergies as significant risk factors for long‐term skin damage. Patients under 20 years of age were 2.06 times more likely to suffer persistent or permanent symptoms compared to those over 50 years (p = 0.000). Additionally, individuals with a history of food allergy were 3.16 times more prone to long‐term skin damage (p < 0.05). A history of drug allergy also significantly increased the risk, with an OR of 1.887 (95% CI: 1.34, 2.66; p = 0.000).

Conclusion

This study identified younger age (< 20 years) and a history of food or drug allergies as significant risk factors for long‐term symptoms following cosmetic adverse reactions. These insights assist dermatologists in recognising high‐risk individuals and guiding tailored preventive strategies. Integrating these indicators into clinical evaluation and patient counselling may enhance early intervention, ultimately reducing the likelihood of persistent or irreversible cutaneous sequelae. Overall, the findings highlight the importance of individualised risk assessment and proactive management in ensuring cosmetic safety and optimising dermatological and public health outcomes.

This study reveals that being under 20 or having a history of allergies significantly increases the risk of long‐term symptoms after a cosmetic adverse reaction. These findings help dermatologists identify high‐risk patients for proactive management and tailored preventive strategies.

1. Introduction

Driven by the opening‐up policy and the rapid development of the market economy—particularly the influx of international cosmetic brands—China's cosmetic industry has undergone substantial growth, making the country the world's second‐largest consumer market for cosmetics [1]. Cosmetics are defined as chemical products, applied to the surface of the human body (e.g., skin, hair, nails, or lips) through topical application or spraying for cleansing, odour elimination, skin care and enhancing appearance [2]. However, as the cosmetic market continues to expand and consumer demand rises, health and social issues related to cosmetic use are becoming increasingly prominent. While cosmetics can enhance the quality of life, they can also lead to various cosmetic adverse reactions (CARs) [3]. CARs are defined as localised or systemic damage involving the skin, mucosa, appendages and other tissues caused by the use of cosmetics. In recent years, the growing diversity and prevalence of cosmetic products have led to a higher incidence of adverse skin reactions, with a growing diversity of disease types and clinical manifestations. While irritant reactions are the most common, other manifestations may include type IV hypersensitivity, contact urticaria, photosensitivity, pigmentary disorders, damage to hair and nails, paronychia, acneiform eruptions, folliculitis and exacerbation of pre‐existing dermatoses [4]. The exact incidence of cosmetic contact dermatitis remains unknown; however, based on the increasing number of reported cases, it appears to be rising in parallel with the rapid proliferation and diversification of cosmetic products [5]. The exact incidence is difficult to determine because mild cases typically resolve spontaneously after discontinuation of the cosmetic product, eliminating the need for medical consultation.

The diagnosis of CARs is based on the national ‘General Guideline: Diagnostic Criteria and Principles of Management of Skin Diseases Induced by Cosmetics’ (GB17149.1‐7‐1997), established by the State Health Department of China [2]. According to this guideline, a definitive diagnosis requires a history of cosmetics use, with symptoms appearing on the exposed areas and/or surrounding regions where cosmetics were applied; symptoms that disappear or improve after discontinuing the suspected cosmetics, and symptoms that recur or worsen upon re‐exposure to similar products, along with typical clinical manifestations. The term CARs has long been recognised in China; as early as the 1990s, the government had already initiated systematic monitoring of cosmetic‐related adverse reactions, with regulatory measures closely resembling those applied to drugs [6]. Designated dermatology departments in state‐appointed medical institutions are responsible for the diagnosis, data collection and reporting of CARs.

While the prognosis for most CARs is favourable—with spontaneous remission commonly observed after discontinuation of the offending product—a subset of patients may not recover fully without intervention. In such cases, appropriate treatments such as topical medications and oral antihistamines can facilitate resolution [7]. Nevertheless, a small proportion of patients may experience persistent cutaneous adverse effects from cosmetics. Despite this clinical concern, previous studies have paid limited attention to the irreversible skin damage that CARs may induce [8]. Based on patient follow‐up data, we observed complete resolution of cutaneous lesions in the majority of cases, while a subset exhibited incomplete resolution, characterised by persistent erythema, hypopigmentation, hyperpigmentation and scarring. For these individuals, the adverse effects not only failed to meet cosmetic expectations but also resulted in varying degrees of cutaneous compromise, leading to considerable physical and psychological distress. These findings highlight the urgent need for early identification of high‐risk patients and the implementation of targeted management strategies.

This cross‐sectional study systematically characterises the clinical and demographic profiles of 7794 patients with CARs in Wuhan, China and investigates prognostic risk factors through multivariate regression analysis.

2. Methods

2.1. Sample Source and Data Collection

All cases came from the National Cosmetics Adverse Reaction Monitoring System in Wuhan City from 2017 to 2023, which contains 7794 cases of adverse reactions to cosmetics. The primary sources of adverse reaction reports on medical and health institutions (primarily dermatology departments of hospitals) (77.9%), cosmetics manufacturing enterprises and operating enterprises (9.9%), individuals (8.0%), as well as other entities such as regulatory departments and monitoring agencies (4.2%). The data from the reporting forms was recorded in a database. Notably, 77.9% of cases (6072/7794) were diagnosed and reported by dermatologists, while a small proportion (8.0%) were self‐reported by individual consumers. We retrospectively reviewed the clinical characteristics and outcomes based on the follow‐up visits of each patient suspected of CARs. Follow‐up evaluations were performed 2–4 weeks after the initial visit to assess the clinical outcomes.

2.2. Diagnostic Criteria of Cosmetic Skin Diseases

The diagnosis of adverse skin reactions to cosmetics refers to the current standard in China, namely ‘Guidelines for the diagnosis and treatment of adverse skin reactions to cosmetics (2024 edition)’.

2.3. Statistics

Descriptive statistics (frequency and percentage) were used to summarise demographic characteristics. These include demographic information, dermatological diagnosis, clinical signs and symptoms, history of cosmetic allergy, food allergy and drug allergy and type of cosmetics. Based on the patient's recovery progress, they were divided into a recovery group and an incomplete recovery group. Differences in the proportions of the characteristics between two groups were tested for significance with the χ ² test, and variables with statistical significance were further included in a binary logistic regression to determine the odds ratio (OR) and 95% confidence intervals (CI). p < 0.05 was considered to be statistically significant. Statistical analyses were performed using SPSS version 25.

3. Results

3.1. Characteristics of Patients With CARs

From 2017 to 2023, a total of 7794 suspected cases of CARs were reported in Wuhan through the Chinese Adverse Drug Reaction Monitoring System. Table 1 displays the basic characteristics profile of the patients. As shown in Table 1, most patients were young to middle‐aged adults, with the highest proportion in the 21–30‐year age group. Females accounted for the vast majority of cases (90.3%) (Table 1). Most reports were submitted by medical and health institutions (77.9%), indicating that surveillance of CARs largely depends on clinical settings. In contrast, contributions from consumers, enterprises and regulatory agencies were relatively limited. In terms of severity, 99.73% of cases were classified as mild. However, 23.44% of these showed incomplete resolution of skin lesions during follow‐up, predominantly presenting as PIH and scarring. Among the 25 severe cases, 17 developed long‐term cutaneous damage.

TABLE 1.

Demographics of subjects.

Characteristics	N = 7794
Gender, n (%)
Females	7039 (90.3%)
Males	755 (9.7%)
Age, n (%)
≤ 20	585 (7.16%)
21–30	2839 (36.43%)
31–40	2183 (28.01%)
41–50	1427 (18.31%)
> 50	787 (10.10%)
Primary sources
Medical and health institutions	6072 (77.9%)
Individuals	623 (8.0%)
Cosmetics manufacturing enterprises and operating enterprises	772 (9.9%)
Regulatory departments and monitoring agencies	327 (4.2%)

3.2. Cosmetic Classification and Diagnosis of CARs

The main classes of cosmetics giving rise to CARs were skin care products (68.94%), mainly leave‐on products such as creams, emulsions, oils, as well as rinse‐off cleansing products. This was followed by products for hair treatments, notably hair dyes and styling products (Figure 1). The majority (91.41%) of the patients were diagnosed with cosmetic contact dermatitis, followed by cosmetic acne, hair disease, contact cheilitis and contact urticaria (Figure 2).

FIGURE 1.

Cosmetic products category causing CAR in patients.

FIGURE 2.

Diagnosis, clinical signs and symptoms of CAR and relative distribution.

3.3. Clinical Manifestations of CARs

The spectrum of clinical manifestations observed in this cohort corresponded to the dermatological diagnoses reported in Section 3.2. The fundamental skin lesions constituting these diagnoses were analysed. Among the 7794 patients, the five most common clinical symptoms include erythema, papule, oedema, plaque and scale. Some patients exhibit multiple symptoms simultaneously. The most common symptom that patients experienced was itching (78.49%), followed by burns, tautness, dryness of skin and pain (Figure 2). It is worth knowing that two‐thirds of patients had more than one cutaneous manifestation.

3.4. The Prognosis of CARs

After excluding patients with incomplete follow‐up data or undocumented allergy histories, we conducted an analysis of the follow‐up recovery outcomes in 5239 patients who presented with adverse reactions to cosmetic products. Recovery status was classified into two categories: Recovery group, defined as complete resolution or marked improvement in the absence of sequelae; and incomplete recovery group, defined as no significant improvement and/or the presence of sequelae such as post‐inflammatory hyperpigmentation or scarring. The degree of lesion improvement was evaluated and determined by dermatologists. We found that the majority of patients (76.56%) had complete recovery, but some patients (23.44%) exhibited long‐term cutaneous compromise, manifested as PIH, atrophic scarring, scarring and so forth (Table 2). Therefore, we divided the patients into two groups: the recovery group and the incomplete recovery group. We conducted a difference analysis on the gender, age and history of allergies (including food allergies, drug allergies and cosmetic allergies) of the patients in both groups. The results showed that there were statistically significant differences in the histories of food allergies, drug allergies and cosmetic allergies between the two groups (Table 2). In other words, patients with a history of food allergies, drug allergies, or cosmetic allergies were more prone to experiencing long‐term cutaneous damage after CARs. Specifically, penicillin (n = 68) and cephalosporins (n = 45) were the most common drug allergens, while seafood (n = 73) and mango (n = 52) were the most frequent food‐related triggers. Candidate risk factors showing significant differences between groups in the univariate analysis were included in the multivariate analysis (Figure 3). Our data indicated the risk of experiencing adverse outcomes of adverse reactions to cosmetics significantly varied across different ages, as patients under 20 years old were 2.06 times more than patients over 50 years old (p = 0.000), while patients with a history of food allergy were 3.16 times more likely to exhibit persistent or permanent symptoms (p < 0.05). A history of drug allergy is also an independent risk factor for persistent or permanent damage of the skin [OR 1.887 (1.34, 2.66), p = 0.000]. The logistic regression model demonstrated a good fit as indicated by the Hosmer–Lemeshow test (χ ² = 6.32, p = 0.61).

TABLE 2.

Univariate analysis of risk factors for CARs.

	Recovery (n = 4011) n%	Incomplete recovery (n = 1228) n%	Total (n = 5239)	p
Gender				0.265
Females	3629 (90.5%)	1124 (91.5%)	4753
Males	382 (9.5%)	104 (8.5%)	486
Age				0.043
≤ 20	272 (6.78%)	111 (9.04%)
21–30	1480 (36.90%)	457 (37.21%)
31–40	1098 (27.37%)	342 (27.85%)
41–50	753 (18.77%)	209 (17.02%)
> 50	408 (10.17%)	109 (8.88%)
History of cosmetic allergy	188 (4.7%)	75 (6.1%)	263	0.046
History of food allergy	127 (3.2%)	100 (8.1%)	227	0.000
History of drug allergy	103 (2.6%)	74 (6.0%)	177	0.000

Note: Bold values indicate significance at p < 0.05.

FIGURE 3.

Multivariate analysis of risk factors for Incomplete recovery. Note: The circle shows the relative risk of each covariate, and the bar shows the range of 95% confidence interval of each covariate, binary logistic regression.

4. Discussion

With the rapid expansion of the cosmetics market, the incidence of CARs has shown a progressive annual increase. Although most CARs are associated with favourable prognoses, a subset of patients may develop long‐term cutaneous damage, including persistent erythema, hyperpigmentation, hypopigmentation and scarring. Consistent with previous studies on CARs [9, 10, 11], our analysis of CARs reports in Wuhan from 2017 to 2023 revealed a clear female predominance (90%), with 64.44% of patients aged 20–40 years, closely aligning with the core demographics of cosmetic consumers. Among product categories, skincare products accounted for the highest proportion of adverse reactions (68.94%), followed by hair treatment products (3.97%). This is likely attributable to both the complex chemical composition of skincare formulations and their high application frequency—a finding consistent with previous reports [12]. The most common diagnostic profiles in our cohort included contact dermatitis, cosmetic acne, contact urticaria, contact cheilitis and hair disease, mirroring established CARs classifications [4]. Predominant cutaneous manifestations comprised erythema, papules and oedema, with approximately 82% of cases reporting pruritus—findings concordant with earlier epidemiological studies [4, 9, 13].

Although CARs typically present as acute and self‐limiting conditions, a proportion of patients experience persistent or permanent damage due to individual susceptibility or inappropriate treatment. In our study, 99.73% of cases were classified as mild; however, 23.44% of these exhibited incomplete resolution of skin lesions during follow‐up, predominantly presenting as PIH and scarring, which caused considerable psychological distress. Notably, among the 25 severe CARs cases, 17 developed long‐term cutaneous damage.

Using longitudinal follow‐up data, we examined clinical factors associated with persistent or irreversible symptoms. Our analysis showed no significant gender differences in disease prognosis, nor any associations with specific diagnoses, clinical manifestations, or cosmetic product categories.

We found a significantly higher incidence of long‐term symptoms in patients under 20 years of age (28.98%) compared with those over 50 years of age (21.08%) (adjusted OR 1.515, 95% CI 1.113–2.061). Previous studies have suggested a greater susceptibility to CARs in younger populations [11, 14, 15], largely due to their higher rates of cosmetic consumption and increased self‐care awareness. We propose two primary mechanisms underlying the age‐related disparity in long‐term CARs symptoms: age‐related immune variation and behavioural exposure patterns. First, developmental immunological immaturity renders children and adolescents more vulnerable to allergens, due to smaller body mass, immature organ systems, heightened vascularity, fragile collagen networks and underdeveloped epidermal barrier function [3]. These features facilitate increased antigen penetration and IgE‐mediated hypersensitivity [16, 17]. In contrast, older adults experience immunosenescence, characterised by diminished innate and adaptive immune functions. Reduced allergen sensitivity—evidenced by lower total/specific IgE levels and attenuated skin prick reactivity—results in fewer IgE‐mediated reactions. However, allergic manifestations in this group may present atypically, such as chronic eczema or drug‐induced hypersensitivity [9, 18, 19, 20]. Second, behavioural patterns critically influence exposure risks. Dose–response relationships have been observed between the severity of CARs and both the daily quantity of cosmetic products used (≥ 5 products: OR 1.70) and application frequency (> 3 times/day: OR 1.91) [14, 21]. Younger individuals tend to use a wider variety of products and apply those more frequently, which may increase their exposure and subsequent risk. Furthermore, heightened aesthetic concerns among adolescents drive stronger adherence to follow‐up and higher treatment expectations, which paradoxically coexist with increased psychosocial burden due to residual skin damage. These findings highlight the importance of age‐specific CARs management strategies that integrate biological vulnerabilities with practical interventions to mitigate adverse reactions.

Our study also revealed that patients with a history of allergies are more susceptible to chronic cutaneous complications. Drug hypersensitivity (adjusted OR 1.887, 95% CI 1.339–2.660, 6.0% vs. 2.6%) and food allergy (adjusted OR 2.338, 95% CI 1.730–3.161, 8.1% vs. 3.2%) emerged as independent prognostic predictors. Penicillin and cephalosporins were the most common drug allergens, while mango and seafood were frequent food‐related triggers. Supporting these findings, Lucca et al. identified gender, age, drug and food allergies, a family history of allergies, use of mixed cosmetics and frequent brand switching as predictive factors for CARs [11]. Multivariate analysis confirmed significant associations between CARs and drug allergy (adjusted OR: 3.9) and family history of allergies (adjusted OR: 1.91). The consistent correlation between allergy histories and both the incidence and prognosis of CARs likely reflects an underlying allergic predisposition, characterised by persistent Th2 cytokine polarisation and mast cell hyperreactivity. This evidence highlights the clinical necessity of obtaining comprehensive allergy histories—including drug, food and cosmetic exposures—during CARs evaluations to assess the risk of persistent or irreversible damage.

By addressing a critical gap in the literature, our study provides predictive evidence to aid dermatologists in the early identification of high‐risk patients, enabling timely intervention and reducing long‐term sequelae. Increased clinical vigilance should be directed toward paediatric patients and individuals with allergy histories, with prompt implementation of targeted strategies. These include immediate cessation of the suspected product, administration of anti‐allergic therapy, barrier restoration protocols and systemic immunomodulation when warranted.

Clear communication of possible long‐term skin damage risk is essential to improve adherence, and follow‐up should be conducted regularly to monitor outcomes. Adolescents are advised to choose products with minimalistic formulations that are fragrance‐free and have a low potential for sensitisation.

As a retrospective study based on patient‐reported data and follow‐up records, potential information bias must be considered. Inadequate data limited quantitative analysis of cosmetic usage frequency and dosage. Furthermore, the absence of consistent information on the anatomical localisation of skin reactions—a challenge inherent to standardising such data in a large cohort (n > 7900) with frequent multi‐site involvement—limited our ability to perform detailed anatomoclinical correlations, which are often critical for strengthening causal links to topically applied products. Residual confounding may exist due to unmeasured factors such as product type, duration of exposure and individual susceptibility. In addition, the absence of detailed compositional information prevented analysis of associations between specific chemical substances and persistent CARs outcomes, although certain ingredients (e.g., preservatives and fragrances) are well‐recognised triggers of adverse skin reactions. Notably, information on pre‐existing skin conditions, particularly atopic dermatitis (AD), was not collected; as AD is associated with impaired skin barrier function, atopy and higher risk of drug and food allergies, its absence may confound the observed association between age under 20 and persistent CARs. Another important limitation is the regional specificity of our cohort, which was drawn exclusively from Wuhan. This restricts the generalizability of our findings both within China, where cosmetic use patterns vary across regions, and internationally, given the substantial differences in cosmetic regulations and ingredient restrictions. Future studies with nationwide sampling, standardised exposure assessments, and detailed product composition data are warranted to validate and extend these findings.

Despite these limitations, this study contributes valuable clinical evidence to guide CARs epidemiology and eventually also its management and prognostic outcomes. The implementation of these findings could enhance dermatologic care and public health while promoting improved cosmetic safety surveillance to protect skin integrity across populations, particularly the more susceptible ones.

Author Contributions

Ayan Hasen: Conceptualization, Methodology, Software, Writing – original draft. Mengdi Yao: Methodology, Data curation. Tian Li: Data curation, Methodology. Ying Jiang: Methodology, Data curation. Yan Li: writing – review and editing, conceptualization, methodology, resources.

Funding

This study was supported by the National Natural Science Foundation of China (grant no. 82273536).

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data supporting the findings of this study are available from the National Cosmetics Adverse Reaction Monitoring System in Wuhan City from 2017 to 2023.