Association of serum vitamin D and serum vitamin C with the risk of melanoma among adults in the United States: a cross-sectional analysis of 2003-2006 and 2017-2018 NHANES data.

To analyze the relationship between serum Vitamin D (VD) and Vitamin C (VC) levels and melanoma risk in American adults using NHANES data. The exposure variables were serum VD and VC. The outcome variable was melanoma presence. Covariates included sex, age, race, BMI, poverty-income ratio, alcohol consumption, smoking, diabetes, hypertension, sunscreen use, and sunburn history. We use weighted logistic regression and restricted cubic spline (RCS) analyses to evaluate the relationship and receiver operating characteristic (ROC) curve to assess the prediction performance. Data from 12,743 (VD) and 12,615 (VC) participants were included. Higher serum VD levels (≥75 nmol/L) and the highest VC quartile levels (OR = 1.91, 95% CI: 1.24-2.97) were associated with an increased melanoma risk. The ROC analysis indicated that the combination of VC or VD with covariates exhibited excellent predictive efficacy (VD and VC model 3: AUC = 0.863). RCS analysis revealed a non-linear correlation between VD and melanoma, and a significant non-linear dose-response relationship between serum VC levels and melanoma. Stratified analysis indicated the adverse effect of higher VD on melanoma was more pronounced in men (OR = 3.55, 95% CI: 1.99-6.33; p<0.001), non-Hispanic white individuals (95% CI: 1.19-2.95; p = 0.009), and diabetic patients (OR = 18.81, 95% CI: 3.44-102.79; p = 0.002). Sensitivity analysis confirmed results stability. High serum VD and VC levels are associated with increased melanoma risk. Personalized prevention considering ultraviolet exposure and genetic factors are recommended to avoid excessive supplementation.