Skin cancer is a significant global health concern, with three main types: basal cell carcinoma (BCC), squamous cell carcinoma (SCC), and melanoma. Ultraviolet radiation (UVR) is a key risk factor, and variations in sun exposure behavior and healthcare engagement between men and women may partly explain the higher incidence of melanoma in men in some Western countries. However, differences in immune responses and hormone levels may also contribute. Male-pattern baldness (MPB), or androgenetic alopecia, is characterized by hair follicle miniaturization and hair loss, often associated with dihydrotestosterone and testosterone level changes. Genetic studies have linked genes associated with both testosterone levels and MPB. Importantly, MPB has been associated with a significantly increased risk of scalp melanoma. The increased incidence of skin cancers among men with MPB may be due to higher chronic UV damage to the exposed scalp, a direct causal role of male-specific factors like testosterone, or other factors related to androgenic regulation of the immune response. This study aimed to investigate the relationship between testosterone levels, MPB, and skin cancers using Mendelian randomization (MR), a powerful technique to assess causal relationships from genetic data. Using data from large-scale genetic cohorts like UK Biobank and QSkin, we aimed to comprehensively assess the relationship between genetic instruments for MPB, sex hormones, and skin cancer susceptibility, also considering anatomical location.

Previous studies have shown associations between MPB and increased risk of skin cancers, particularly scalp melanoma and SCC. These observational studies suggested a potential link between MPB and higher UV exposure on the scalp. There have been studies exploring the role of sex hormones, particularly testosterone, in skin cancer development, with some suggesting a potential role in melanoma pathogenesis. However, these studies mostly rely on observational data, which is susceptible to confounding factors. Mendelian randomization (MR) offers a powerful tool to disentangle these complex relationships by using genetic variants as instrumental variables, addressing some of the limitations of observational studies.

This study employed a two-sample Mendelian randomization (MR) framework to investigate the relationships between testosterone levels, MPB, and the risk of keratinocyte cancers (KCs) and melanoma. Data from several large-scale genome-wide association studies (GWAS) meta-analyses and cohorts (UK Biobank, QSkin, GenoMEL) were utilized. Genetic instruments for MPB, total testosterone, free testosterone, and sex hormone-binding globulin (SHBG) were identified from GWAS summary statistics. Univariable MR analysis was initially performed to assess direct associations between these factors and skin cancer risk. Multivariable MR (MVMR) analysis was conducted to account for potential confounding and mediation by incorporating multiple traits simultaneously. Site-specific analysis was also performed to evaluate whether the associations differed depending on the anatomical location of the skin cancer. Sensitivity analyses were conducted to address potential pleiotropy (where genetic variants influence multiple traits) and to assess the robustness of the findings. Specifically, MR-PRESSO was used to identify and adjust for potential outlier SNPs. Phenome-wide association studies (PheWAS) were conducted to explore potential pleiotropic effects of MPB SNPs on pigmentation traits, examining whether the relationship between MPB and skin cancer could be partly explained by pigmentation.

The study found strong associations between MPB and the risk of KCs, particularly SCC, but not with androgen levels. Univariable MR analysis suggested that the MPB-KC association was significantly confounded by MPB SNPs involved in pigmentation pathways. Site-stratified MR analysis revealed strong associations between MPB and the risk of head and neck melanoma and squamous cell carcinoma, supporting the hypothesis that increased sun exposure on the scalp due to hair loss is a major driver. In contrast, no strong association was found between endogenous sex hormone levels (total testosterone, free testosterone, SHBG) and any KC outcomes or melanoma risk. Sensitivity analyses, including those using outlier-robust MR methods, confirmed the associations between MPB and head and neck skin cancers while mitigating concerns about pleiotropy. The inclusion of pigmentation-related traits in multivariable MR models attenuated the observed association between MPB and overall KC risk, further supporting the role of pigmentation as a confounder. The results demonstrated that the association between MPB and head and neck melanoma was primarily driven by melanoma on the scalp, underscoring the importance of increased sun exposure.

The findings of this study provide strong genetic evidence suggesting that the increased risk of skin cancers in men with MPB is primarily driven by increased sun exposure of the scalp rather than a direct effect of testosterone or other sex hormones. While previous observational studies have hinted at such a relationship, this study provides robust evidence using MR to overcome limitations of observational data and account for confounding. The strong association observed between MPB and head and neck skin cancers highlights the importance of sun protection measures for men with MPB, particularly focusing on the scalp. The lack of association between endogenous sex hormones and skin cancer risk does not support previous hypotheses suggesting a direct causal role for androgens. The confounding effect of pigmentation-related genes highlights the complexity of the genetic architecture of skin cancer risk and the importance of carefully considering pleiotropic effects in genetic studies.

This large-scale Mendelian randomization study provided strong evidence that the association between male-pattern baldness and skin cancer risk, particularly for head and neck cancers, is primarily attributable to increased sun exposure on the scalp. The study did not find evidence to support a causal role for testosterone or other sex hormones in skin cancer risk. Future research could explore other potential mediating factors, examine the interaction between genetics, sun exposure, and skin cancer risk, and investigate the generalizability of these findings across different populations and ethnicities.

The study relied on self-reported MPB in the UK Biobank, which might introduce some bias. The analyses were primarily conducted in individuals of European ancestry, limiting the generalizability of findings to other populations. The site-specific analyses had limited sample sizes for certain cancer types and locations. While sensitivity analyses were conducted to address pleiotropy, the possibility of residual confounding remains. Future research incorporating more granular data on sun exposure, histological subtypes of skin cancers, and diverse populations would strengthen the conclusions.