Abstract
Background/Aim: Solar radiation represents the most important environmental risk factor for skin cancer. Despite growing awareness of the importance of regular sun protection products, the incidence of malignant melanoma continues to rise. This raises the question of whether sunscreen use truly provides a protective benefit against the risk of melanoma. The aim of this study was to assess the relevance of sunscreen use for melanoma risk.
Materials and Methods: A systematic review and meta-analysis were conducted using PubMed database. The use of sunscreen was compared based on the categories “ever” vs. “never”. The study quality and level of evidence were assessed using recognized assessment tools. Crude odds ratios (ORs) with 95% confidence intervals (CIs) were calculated. Moderation analyses were conducted to explore the influence of sex, study quality, and geographical location on the overall effect estimate.
Results: A total of 23 relevant studies were identified. An OR of 0.98 (95%CI=0.79-1.21, p=0.83) for the categories sunscreen use “ever” vs. “never/rarely” and an OR of 0.95 (95%CI=0.75-1.20, p=0.66) for the categories sunscreen use ever vs. never indicate no significant association between sunscreen use and melanoma risk. This means that a protective effect for sunscreen use and melanoma risk could not be proven. There was significant heterogeneity between the studies, which could not be explained despite sensitivity analyses, subgroup analyses, and examination of funnel plots. As an asymmetric funnel plot and a significant Egger’s test occurred, the results should be interpreted with caution.
Conclusion: No association was found between melanoma risk and sunscreen use. Many of the studies showed methodological inaccuracies and further studies examining sunscreen application frequency and behavior towards sun exposure in more detail are needed.
Introduction
Skin cancer is a major public health concern, with malignant melanoma being the most aggressive form. The primary environmental risk factor for melanoma is ultraviolet (UV) radiation, which induces DNA damage and promotes carcinogenesis (1, 2). Sunscreen is widely recommended as a preventive measure, as it absorbs, reflects, or scatters UV radiation, reducing skin exposure (3). Despite increased sunscreen use, melanoma incidence continues to rise, raising questions about its protective efficacy (4). Studies present conflicting results regarding sunscreen use and melanoma risk. Some meta-analyses suggest that regular sunscreen use reduces melanoma risk, particularly when applied correctly and alongside other protective measures (5, 6). However, other studies report no significant reduction, possibly due to inconsistent application, behavioral factors leading to prolonged sun exposure, and differences in sun protection factor (SPF) levels (7, 8). Observational studies suggest that sunscreen users may engage in riskier sun exposure behaviors, increasing UV-related damage (9). This study systematically reviews and meta-analyzes the association between sunscreen use and melanoma risk, comparing individuals who have ever used sunscreen to those who have never (or rarely) used it. This study adds to the growing body of research examining the association between sunscreen use and melanoma risk and provides further insights to inform public health recommendations and skin cancer prevention strategies.
Materials and Methods
This meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines to guarantee a clear, structured, and thorough presentation of the existing research evidence (10). Furthermore, the methodology was guided by the approach used by Haddad et al. (11).
Search strategy for study identification. The systematic literature search in the PubMed database was finalized on October 31, 2024. The search strategy involved using a range of English terms and keywords, including both free text and MeSH terms, such as “sunscreening agents”, “sunblock”, “sun protection factor”, “suntan lotion”, “sun cream”, “solar protector”, “block out”, as well as terms related to skin cancer, including “skin neoplasm”, “skin tumor”, “melanoma”, “malignant”, and “human”. When uncertainties regarding study selection arose, guidance was sought from Professor Dr. Jörg Reichrath. Studies deemed potentially relevant for the meta-analysis were assessed in full text.
Selection and inclusion criteria. The inclusion criteria for studies were as follows: the publication had to be in either German or English; acceptable study designs included randomized controlled trials, cohort studies, and case-control studies; and the research had to examine the relationship between sunscreen use and the risk of developing melanoma. Reported effect measures – such as odds ratios (OR) or relative risks (RR) – were either directly used, converted as needed, or calculated from the study’s raw data. Priority was given to observational studies, both prospective and retrospective, within cohort or case-control frameworks. Studies lacking a clear control group were excluded from the analysis.
Data extraction. Relevant data from the included studies were compiled into several Excel spreadsheets, containing all key variables required for the meta-analysis. In cases where specific information was missing, the corresponding authors were contacted to request the necessary data.
Risk of bias. To evaluate the quality of the included studies, we applied the Newcastle-Ottawa Scale (NOS), a standard tool commonly used in meta-analyses. Additionally, each study was assessed for its level of evidence based on the criteria established by the Oxford Centre for Evidence-Based Medicine (OCEBM) (12).
Statistical analysis. To estimate the overall association, we calculated pooled odds ratios (ORs) with corresponding 95% confidence intervals (CIs). Given the low incidence rate of malignant melanoma, we applied the rare disease assumption as described by Cornfield, under which the OR approximates the risk ratio (RR) (13, 14). Accordingly, the pooled OR was interpreted as a relative risk measure throughout the meta-analysis.
The relationship between sunscreen use and melanoma risk was assessed using unadjusted (crude) ORs derived from each individual study. Although all included studies reported on the frequency of sunscreen application, the categorization schemes varied considerably. To improve comparability, we merged the response categories “never” and “never/rarely” into a unified reference group. Additionally, we conducted a separate meta-analysis limited to studies that included a distinct “never” use category, thereby enhancing consistency and minimizing heterogeneity.
All statistical analyses were carried out using the Metafor package within R version 4.3.3. Forest plots were generated to visually summarize the pooled estimates. Between-study heterogeneity was quantified using Cochran’s Q test (with statistical significance set at p<0.05, two-sided) and the I2 statistic. A random-effects model (RE model) was employed, utilizing restricted maximum likelihood estimation (REML) for parameter estimation.
To examine the potential for publication bias, we constructed funnel plots and applied Egger’s test. In instances of substantial heterogeneity, we performed sensitivity analyses to explore potential sources. Where the dataset permitted, moderator analyses were conducted based on study characteristics such as geographic region, participant sex, and study quality (as rated using the Newcastle-Ottawa Scale). When a moderator variable significantly influenced the effect estimate, targeted subgroup analyses were subsequently undertaken.
Results
Search results and study characteristics. The initial PubMed database search, using the pre-defined search terms, yielded a total of 2,439 records. All identified studies were subjected to title and abstract screening. After excluding 2,351 studies due to irrelevance or lack of focus on the research question, 87 full-text articles remained for detailed evaluation. Upon full-text review, an additional 64 publications were excluded due to methodological limitations or failure to meet the established inclusion criteria. Ultimately, 23 studies were deemed eligible and included in the meta-analysis (3, 15-36). Detailed information on individual study characteristics is available in the Supplementary Material. A PRISMA-compliant flow diagram illustrating the literature selection process is shown in Figure 1.
Flow chart summarizing the literature review process for the meta-analysis (source: www.prisma-statement.org).
The included studies spanned a publication window of 35 years, ranging from 1986 to 2021. In terms of geographic distribution, 14 studies were conducted in Europe, four in Australia, three in North America, and two in South America. While the majority of studies enrolled both sexes without distinction, some were sex-specific: Holly et al. (1995) and Savoye et al. (2018) included only female participants, whereas Stenehjem et al. (2017) focused exclusively on men (25, 26, 34). To assess study quality, we applied the Newcastle-Ottawa Scale (NOS), which evaluates non-randomized studies based on selection, comparability, and exposure or outcome. Most studies achieved a score of 5 or higher (out of 9), indicating acceptable methodological rigor. Two studies, by Stenehjem et al. and Vranova et al., scored four points each (26, 27). For clarity, we defined NOS scores of ≥5 as indicative of low risk of bias, and scores <5 as high risk of bias. Additionally, each study was classified using the Oxford Centre for Evidence-Based Medicine (OCEBM) criteria. The majority fell under evidence level 3b (individual case-control studies) with a recommendation grade B. A smaller number, categorized as level 2b, represented individual cohort studies with the same recommendation grade. Only three studies were assigned level 4 evidence with a grade C recommendation (26, 27, 33).
Meta-analysis. The first analysis. included 19 studies regarding sunscreen use (ever vs. never/rarely) among melanoma patients and healthy controls (Figure 2). The pooled odds ratio was 0.98 (95%CI=0.79-1.21) with a p-value of 0.83. The results are graphically presented in a forest plot with corresponding weights in a random-effects model. The results of the meta-analysis (n=19 studies) suggest that sunscreen use is neither a risk factor nor a protective factor for melanoma. However, the 95%CI included ORs in both directions at a level of clinical relevance. When restricting the included studies to the comparison of sunscreen use “ever” vs. “never” in n=17 studies an OR of 0.95 (95%CI=0.75-1.20) with a p-value of 0.66 was yielded indicating no significant effect in line with the first meta-analysis (Figure 3).
Forest plot showing no association between melanoma risk and sunscreen use (categories: ever vs. never/rarely). I2: 85.9%, Q: 86.9, p<0.0001.
Forest plot showing no association between melanoma risk and sunscreen use (categories: ever vs. never). I2: 89.43%, Q: 97.8, p<0.0001.
Sensitivity analyses. In the main analysis of the association between melanoma risk and sunscreen use (ever vs. never/rarely), substantial heterogeneity was observed across studies: Q(df=18)=86.8988, p<0.0001; I2=85.94%; H2=7.11 (Supplementary Material). Therefore, we conducted a sensitivity analysis to further investigate this heterogeneity. We examined potential moderators, including study quality, sex, and geographic location. However, none of the subgroup analyses revealed notable results that could sufficiently explain the heterogeneity. The coefficient of determination (R2), which indicates how well the moderation model (study quality, sex, or geographic location) explains the variation in the dependent variable (odds ratios across study groups), was below 7% in all analyses. This suggests that the moderators had little influence on the odds ratios. Additionally, we created a funnel plot and quantified its asymmetry using Egger’s test, which was significant (p=0.002) (Supplementary Material). The funnel plot indicated that studies with smaller sample sizes (and thus larger 95%CIs) tended to report smaller ORs <1 indicating a protective effect. However, it remains unclear whether publication bias or methodological differences are the underlying cause of these small-study effects. Therefore, our findings should be interpreted with caution. Taken together, our results highlight the need for large, high-quality studies to further investigate the association.
Discussion
The results of this meta-analysis and, consequently, the melanoma risk associated with sunscreen use are subject to various influences. The application of sunscreen may be inadequate in terms of both quantity and frequency of reapplication, leading to suboptimal protection from the outset. Additionally, individuals may not use sunscreen daily based on the sun exposure in their region but rather on sunny days or during vacations (3).
Furthermore, skin type influences melanoma risk, as the Celtic skin type is associated with an increased risk of skin cancer (4, 37). These highly sensitive skin types may not receive sufficient protection from sunscreen alone. In contrast, it can be assumed that individuals with darker skin types (mixed/Mediterranean type), the generally healthier lifestyle of Southern Europeans (e.g., a fish- and vegetable-rich Mediterranean diet, higher levels of daily physical activity), adherence to rest periods during peak UV intensity (“Siesta” tradition, “Pranzo”), differing sun exposure patterns, and other potential factors may have a protective effect against malignant melanoma as described in Haddad et al. (6, 11).
Other influential factors likely include: Variations in study designs and methodologies, large differences in sample sizes, measurement inaccuracies due to self-reported data from patients (which is inherently subjective) (7), differences in the amount of sunscreen applied, as this is a key determinant of its effectiveness (8, 38), the tendency for sunscreen to be insufficiently reapplied after swimming or sweating (9), the type of sunscreen used, particularly the sun protection factor (SPF), as higher SPF offers better protection than lower SPF (1), the potential for individuals to develop a false sense of security after applying sunscreen, leading to prolonged sun exposure and increased UV radiation exposure (2), additional protective measures taken by individuals who use sunscreen, such as seeking shade or wearing protective clothing, which may confound the observed melanoma risk reduction (5), genetic predisposition, which could play a crucial role in the effectiveness of sunscreen, as different genetic backgrounds may impact melanoma risk and the protective capacity of sunscreen (16, 40) and lastly the geographical location and regional UV radiation intensity, which could influence sunscreen efficacy (24).
Moreover, it is important to note that younger study participants may develop melanoma later in life, which could lead to false-negative results in the analysis (26).
Our analysis showed substantial heterogeneity. The asymmetric funnel plot and the significant Egger’s test suggest the presence of small-study effects or publication bias (10). However, it is important to recognize that this could be due to a mathematical artifact rather than actual publication bias or small-study effects.
This occurs due to the mathematical relationship between the odds ratio (OR) and its standard error (SE): The SE of a study’s effect estimate is dependent on the sample size (n). Smaller studies naturally have larger SEs. This effect is particularly noticeable in the studies of Espinosa Arranz et al. (1999), Green et al. (2011), Ródenas et al. (1996) which seem to drive the asymmetry in the funnel plot. The SE of the logOR is mathematically linked to the OR itself, meaning that more extreme ORs tend to have higher SEs. As a result, even if no real small-study effect exists, smaller studies may exhibit a wider spread in effect sizes, making the funnel plot appear asymmetric. Thus, Egger’s test may falsely indicate asymmetry, even in the absence of true bias (10). However, the fact that the smaller included studies show contradictory results, also suggests methodological differences. The underlying reasons, however, cannot further be examined in this study.
In addition, the presence of publication bias is supported by the observation that small-sized studies are predominantly located on the left side of the forest plot and tend to suggest a protective effect of sunscreen use. In contrast, studies with larger sample sizes mostly fail to demonstrate a protective effect. This suggests that small studies that do not indicate a protective effect are less likely to be published.
The fact that melanoma can appear anywhere on the skin, not just in sun-exposed areas, and in other organs, suggests the presence of other risk factors that need to be identified (39, 40). Studies based on the frequency of sunscreen application would be useful to assess how melanoma risk is affected when sunscreen is applied more frequently and in greater amounts.
Footnotes
Authors’ Contributions
S.B. designed the research and conducted the literature search. S.H. wrote the manuscript. Both serve as co-first authors. The statistical analyses were implemented by J.W. The Authors S.B., S.H. and J.W. interpreted the data, J.W. S.W., J.R. and T.V. revised the subsequent draft for important intellectual content, read and approved the final manuscript.
Supplementary Material
The Supplementary Material is uploaded on GitHub. It contains the Study Characteristics (Excel-Chart) and the full output of the analyses from “R” Available at: https://github.com/SinanHaddad1/Sunscreen-Use-Melanoma-A.Brunner-S.Haddad/blob/c67daeefe14d6a890717c43f4710295de7dca6bd/Supplementary%20MaterialMelanoma%20x%20Sunscreens%20Study%2C%20Brunner%20Haddad.pdf
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received May 17, 2025.
- Revision received June 6, 2025.
- Accepted June 13, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
References
Jump to section
Related Articles
Cited By...
- No citing articles found.