Introduction

Melanoma, a malignant neoplasm originating from melanocytes, ranks among the most aggressive forms of skin cancer, with a global incidence that has risen steadily over recent decades. In 2020, medical records showed approximately 324,635 new cases diagnosed globally, highlighting its significant impact on public health1. While ultraviolet (UV) radiation is universally recognized as an important factor linked to melanoma development2,3, recent studies also indicate that imbalances in metabolic processes might contribute to the pathogenesis of melanoma4,5. Metabolic syndrome (MetS) refers to a combination of health issues, including abdominal obesity, hypertension, hyperglycemia and dyslipidemia6. While research has shown connections between MetS and several types of cancer7, the relationship with melanoma is still not well understood.

Preliminary evidence indicates a potential link between MetS and melanoma. Nagel et al. reported an association between metabolic risk factors and melanoma incidence within the MetS and Cancer Project8, while Sha et al. identified a correlation with melanoma metastasis in a Chinese cohort9. However, over the past ten years, systematic research investigating the relationship between MetS and melanoma has remained insufficient. Notably, there is a particular lack of exploration into the synergistic effects of metabolic and circadian rhythm factors. Circadian disruption, exemplified by prevalent modern lifestyle issues such as sleep deprivation and depression, has been independently confirmed by studies to be associated with tumorigenesis10. In a review, Markova-Car et al. suggested that circadian rhythm may contribute to the progression of melanoma through mechanisms such as influencing immune regulation and melatonin secretion11. Nevertheless, comprehensive studies that integrate the analysis of metabolic syndrome and circadian rhythm factors are still lacking.

To address this gap, the concept of CircS has been proposed in recent years, offering a more holistic framework for understanding how metabolic and circadian dysregulation may influence cancer risk12. CircS not only encompasses the metabolic abnormalities associated with MetS but also integrates indicators of circadian rhythm disruption, such as short sleep duration and depression. Crucially, robust evidence demonstrates fundamental sex differences in metabolic pathophysiology: men exhibit significantly higher visceral adiposity and hepatic fat accumulation, while women show greater peripheral subcutaneous adipose tissue deposition13. Experimental studies further reveal sex-specific responses to circadian disruption: females display more pronounced disturbances in energy homeostasis (e.g., decreased leptin, increased ghrelin), whereas males exhibit heightened hedonic appetite responses14. These pathophysiological divergences necessitate gender-stratified analyses to accurately elucidate epidemiological associations between MetS, CircS and melanoma risk.

Additionally, practical metabolic markers, such as the triglyceride glucose (TyG) index and its derivatives, have gained attention as cost-effective surrogates for insulin resistance and have been extensively investigated15. However, their utility in assessing melanoma risk remains to be substantiated.

Despite the valuable insights provided by prior studies into the metabolic and circadian-related risks of melanoma, several critical questions remain unresolved: (i) Lack of Integrative Research: There is a dearth of comprehensive studies that combine analyses of MetS and CircS to elucidate their synergistic effects on melanoma risk. (ii) Underexplored Gender Differences: The influence of gender on these associations has not been sufficiently examined. (iii) Limited Generalizability: Existing research often relies on small sample sizes or specific populations, lacking the support of large-scale, representative databases. To address these issues, this study leverages data from the National Health and Nutrition Examination Survey (NHANES), a nationally representative health survey conducted by the Centers for Disease Control and Prevention in the United States. NHANES collects extensive health, nutrition, and medical data from a diverse population, providing a robust platform to investigate the interrelationships between MetS, CircS, the TyG index and its derivatives, and melanoma risk. Through gender-stratified analysis, we observed that the predictive power of MetS components for melanoma risk was significantly more pronounced in males. Building on this finding, we applied advanced statistical techniques to the male dataset, including forest plots and restricted cubic spline (RCS) curves, to explore the dose-response relationship between metabolic syndrome components and melanoma risk, as well as to identify potential thresholds.

Methods

Study population and design

The NHANES is a cross-sectional survey that employs a complex, multi-stage probability sampling design to select a nationally representative sample of the civilian, non-institutionalized U.S. population. The sampling process involves selecting counties or groups of counties, followed by segments within those areas, households within segments, and finally, individuals within households. To ensure reliable estimates for specific subgroups, NHANES oversamples certain populations, such as Hispanic persons and non-Hispanic Black persons16,17.

The analysis encompassed six consecutive NHANES survey cycles spanning 2007–2018, with a total of 59,842 individuals included. All the participants gave handwritten agreements and passed the ethics reviews from the NCHS Ethics Review Board. Participants were excluded if they: (1) were under 20 years of age; (2) were pregnant; (3) had missing or indeterminate data for melanoma status; or (4) lacked sufficient data to classify all of the following variables: MetS, CircS, TyG, TyG-WC, TyG-WHtR, and TyG-BMI. Finally, 29,132 participants met the screening requirements and were enrolled in the analysis, including 209 melanoma patients and 28,932 non-melanoma participants. Details about the screening process are shown in Fig. 1. Other survey methods are available at NHANES websites (https://www.cdc.gov/nchs/nhanes/index.htm).

Fig. 1
figure 1

Sample selection process flow chart.

Full size image

Assessment of melanoma

Data on tumor diagnoses were extracted from the NHANES database using responses to the cancer history questionnaire. Participants were asked two sequential questions to ascertain cancer status: (i) “Have you ever been told by a doctor or other health professional that you had cancer or a malignancy of any kind?” and (ii) for those responding affirmatively, “What kind of cancer was it?” In the NHANES survey, participants reported their cancer type by selecting from a predefined list, which included “Melanoma”, “Skin (non-melanoma)”, and “Skin (don’t know what kind)” for skin cancer. Only participants who selected ‘Melanoma’ were included as participants with melanoma in our analysis. Based on these responses, participants were classified into two groups for inclusion in the study: those reporting a diagnosis of melanoma were designated as melanoma patients, while those reporting no history of cancer were categorized as non-melanoma group. Participants with an undetermined melanoma status, such as those who reported a history of cancer but did not specify “Melanoma” as a cancer type, or those with missing or unclear cancer history data, were excluded from the analysis.

Diagnosis of metabolic syndrome and circadian syndrome

The classification of participants into MetS, depression symptoms, and CircS categories was conducted using raw data from NHANES. These variables were not pre-defined in NHANES, instead we derived them based on established diagnostic criteria as described below.

MetS was defined according to the harmonized criteria established by the International Diabetes Federation Task Force on Epidemiology and Prevention, National Heart, Lung and Blood Institute, American Heart Association, World Heart Federation, International Atherosclerosis Society and International Association for the Study of Obesity. Participants were classified as having MetS if they met three or more of the following criteria18: (i) Elevated waist circumference (≥ 102 cm in males, ≥ 88 cm in females); (ii) Elevated blood pressure (systolic ≥ 130 mm Hg and/or diastolic ≥ 85 mm Hg) or current use of antihypertensive medication; (iii) Reduced high-density lipoprotein cholesterol (HDL-c) (< 40 mg/dL in males, < 50 mg/dL in females) or use of lipid-lowering therapy; (iv) Elevated triglycerides (≥ 150 mg/dL) or use of lipid-lowering therapy; (v) Elevated fasting glucose (≥ 100 mg/dL) or use of glucose-lowering medication. For each participant, we extracted the relevant variables from NHANES and computed the number of MetS criteria met. If three or more criteria were fulfilled, the individual was classified as having MetS. If fewer than three criteria were met and sufficient data were available to confirm this, the individual was assigned to the non-MetS group. Participants with insufficient data to determine MetS status were designated as having missing MetS classification.

Depression symptoms were assessed using the Patient Health Questionnaire-9 (PHQ-9)19, a validated nine-item screening tool that evaluates the frequency of depressive symptoms over the preceding two weeks. Participants with a PHQ-9 score of 5 or higher were categorized as exhibiting depression symptoms, consistent with established thresholds for mild depressive symptomatology. Participants with scores ≥ 10 were categorized as having depressive symptoms; those with scores < 10 were classified as having no depressive symptoms. Participants with incomplete or non-informative PHQ-9 responses received a missing depression symptom classification.

CircS was defined based on the presence of four or more of the following seven components20. These components consist of the five criteria for MetS, plus two additional factors: (i) Self-reported short sleep duration (< 6 h / day); (ii) Depression symptoms (PHQ-9 score of 5 or higher). It combines metabolic factors with elements related to circadian rhythms to assess circadian disruption thoroughly. For each participant, we calculated the number of meeting CircS criteria. If four or more were met, the individual was defined as having CircS. If fewer than four criteria were met and sufficient data were available, the individual was classified as not having CircS. Individuals with insufficient data for reliable classification were categorized as having missing CircS status.

Definitions of TyG, TyG-WC, TyG-WHtR and TyG-BMI

The TyG index is a well-established measure of insulin resistance, calculated by combining fasting blood sugar and triglyceride levels21. The vast majority of NHANES participants underwent comprehensive physical examinations at Mobile Examination Centers (MECs), which are standardized facilities designed for consistent clinical assessment across survey locations. During these mandatory visits, trained personnel collected blood samples for laboratory analysis and performed standardized measurements including body weight, height, and waist circumference. These values were used for calculating body mass index (BMI) and waist-to-height ratio (WHtR) using simple, standard formulas. The calculating formulas are as follows: (i) TyG = ln [triglycerides (mg/dl) × glucose (mg/dl) / 2]21; (ii) BMI = body mass (kg) / height2(m2)22; (iii) WHtR = waist circumference (cm) / height (cm)23; (iv) Triglyceride glucose-waist circumference (TyG-WC) = TyG × waist circumference (cm)24; Triglyceride glucose-waist height ratio (TyG-WHtR) = TyG × WHtR25; Triglyceride glucose-waist height ratio (TyG-BMI) = TyG × BMI26.

Covariates assessment

Potential confounding variables linked to MetS, CircS and melanoma, as noted in prior studies27,28,29, were adjusted for in the final analyses. These variables spanned three domains: demographic (age, sex, race/ethnicity), socioeconomic (marital status, education level, the ratio of family income to poverty (PIR)), and lifestyle (alcohol consumption habits and tobacco use history).

Sex was categorized as male or female. Race/ethnicity was classified into five groups: Mexican American, non-Hispanic Black, non-Hispanic White, other Hispanic and other races. Marital status was dichotomized into married or living with a partner versus others. Education level was stratified into three categories: less than high school, high school graduate and more than high school. The PIR was included as a continuous variable to reflect socioeconomic status.

The categorization of alcohol consumption was based on self-reported drinking patterns, which were classified into three distinct groups according to standardized criteria30. Heavy drinking was defined as daily consumption exceeding three standard drinks for females or four standard drinks for males. Moderate alcohol use encompassed individuals reporting daily intake of at least two drinks (female) or three drinks (male). The remaining participants who did not meet these thresholds were classified as mild drinkers. Due to the structure of the NHANES questionnaire—where participants were instructed to record “1” even if their consumption was less than one drink, and where a “zero drinks” response option was not available. As such, individuals who reported any drinking in the past year but did not meet the criteria for moderate or heavy drinking were grouped together as mild drinkers. Smoking status was defined as never smokers (individuals with fewer than 100 cigarettes smoked in total), former smokers (those with ≥ 100-lifetime cigarettes but no current tobacco use) and current smokers (those with ≥ 100-lifetime cigarettes and maintained regular smoking habits at the time of assessment).

Statistical analyses

For continuous variables, median estimates accompanied by 95% confidence intervals (CIs) were calculated, whereas categorical variables were analyzed using frequency distributions and relative percentages. Group differences were evaluated using the Chi-square test for categorical variables and analysis of variance for continuous variables.

For Table 1, NA values were labeled as “Unknown” to show the proportion of missing data. In the regression analyses, these ‘Unknown’ values were treated as NA, and observations with NA in the variables used in each model were excluded.

To ensure nationally representative estimates and account for the complex sampling structure of the NHANES dataset, all descriptive analyses and regression models were weighted using the appropriate survey design elements. Using the R package “survey” (version 4.4.2), we constructed a survey design object incorporating sampling weights (WTINT2YR), strata (SDMVSTRA), and primary sampling units (SDMVPSU). This approach accounted for the stratified, clustered, and multistage nature of the NHANES sampling design. Weighted analyses included baseline characteristics (Table 1) and both univariate and multivariate logistic regression models. Graphical analyses, including receiver operating characteristic (ROC) curves, forest plots stratified by sex, and RCS curves exploring the association between the number of MetS components and melanoma in males, were performed without applying weights, as these were intended for visualization rather than population-level inference.

Continuous variables—including age, total cholesterol, HDL-c, triglycerides, fasting glucose, TyG, TyG-WC and TyG-WHtR—were categorized into quartiles, ranging from the lowest (first quartile, Q1) to the highest (fourth quartile, Q4). Due to the uneven distribution of melanoma patients, specific adjustments were applied to the quartile groupings: for age, quartiles Q1 and Q2 were combined; for HDL-c in females, quartiles Q1 and Q2 were merged; and for both TyG-WC and TyG-WHtR, quartiles Q1 and Q2 were combined, as were Q3 and Q4. To assess the predictive value of metabolic factors for melanoma, ROC curve analysis was performed with gender-specific stratification via the R package “pROC” (version 1.18.5).

Relationships between MetS component counts and melanoma risk were then examined using multivariable logistic regression frameworks adjusted for survey weights, across subgroups defined by key variables. Three models were employed using the “rms” package (version 7.0.0): Model 1: No covariates adjusted. Model 2: Adjusted for age, race and gender. Model 3: Adjusted for age, gender, race, education, PIR, marital status, smoking status and alcohol consumption level.

In the male subgroup, further stratified analyses were conducted to explore variations in the association between MetS components and melanoma across distinct populations, accompanied by interaction tests to assess potential effect modification. P-values for effect modification were derived from likelihood ratio tests comparing Firth-penalized models with and without interaction terms (implemented via “logistf” package (version 1.26.0)). Forest plots were generated using the “forestplot” R package (version 3.1.6), with each estimate derived from fully adjusted weighted logistic regression models. Additionally, RCS plots were generated with R package “ggplot2” (version 3.5.1) to visualize the association between MetS components and melanoma risk in males, facilitating the examination of potential nonlinear relationships.

Data management utilized the “tidyverse” R package (version 2.0.0), with “dplyr” package (version 1.1.4) for manipulation and “broom” package (version 1.0.7) for model output tidying. All analyses were conducted in R version 4.4.0, with statistical significance defined as p < 0.05 (two-sided).

Results

Baseline characteristics of study participants by melanoma

Table 1 Characteristics of NHANES samples (2007–2018).
Full size table

Table 1 summarizes the baseline characteristics of 29,132 participants from the NHANES (2007–2018). The median age was significantly higher in the melanoma group compared to the non-melanoma group. The melanoma group included a higher proportion of males and Non-Hispanic Whites. Participants with melanoma were more likely to have education beyond high school and a higher median PIR.

Lifestyle factors differed between groups. The melanoma group had fewer heavy alcohol consumers and current smokers, but more former smokers. Hypertension prevalence was elevated in the melanoma group, as was median fasting plasma glucose. Abnormal waist circumference was more common among those with melanoma.

Metabolic and circadian profiles also varied significantly. The melanoma group exhibited higher prevalence of CircS and MetS. The number of MetS components was skewed toward three and four in the melanoma group. Median TyG index, TyG-WC and TyG-WHtR were significantly higher in the melanoma group, while TyG-BMI showed no notable difference. No significant differences were observed in triglycerides, HDL-c, total cholesterol or BMI.

Univariate logistic regression for factors associated with melanoma

Supplementary Table S1 presents the results of univariate logistic regression analyses evaluating factors associated with melanoma risk. Among demographic factors, older age and Non-Hispanic White race were strongly linked to increased melanoma odds, with all other racial groups demonstrating significantly lower risks. Higher education levels and greater PIR were also associated with elevated risk.

For lifestyle factors, moderate and heavy alcohol consumption were inversely related to melanoma risk compared to mild consumption, while former smoking status was associated with higher odd.

Metabolic factors exhibited strong associations with melanoma. Both MetS and CircS significantly increased risk, with a dose-response pattern observed as the number of components for each syndrome rose. Elevated levels of fasting glucose, TyG, TyG-WC and TyG-WHtR were also associated with higher odds.

Gender-specific predictive performance of metabolic variables for melanoma

Fig. 2
figure 2

ROC curves for metabolic variables in predicting melanoma risk in males.

Full size image
Fig. 3
figure 3

ROC curves for metabolic variables in predicting melanoma risk in females.

Full size image

Univariate ROC analysis evaluated the standalone discriminatory capacity of metabolic biomarkers for melanoma risk, stratified by gender. In males, based on the ROC curves presented in Fig. 2, MetS components exhibited the highest predictive ability among the evaluated variables (AUC = 0.706), followed by the CircS components (AUC = 0.678), whereas triglycerides demonstrated the weakest predictive capacity (AUC = 0.503). HDL-c exhibited an inverse association with melanoma risk (AUC = 0.477). In females, as Fig. 3 shows, the TyG index and Triglycerides both exhibited the highest predictive performance for melanoma risk (AUC = 0.655). In contrast, TyG-BMI showed the lowest predictive ability (AUC = 0.497). These findings indicate gender-specific differences in the utility of metabolic variables for predicting melanoma risk.

Multivariable logistic regression models for the association between metabolic syndrome components and melanoma

Multivariable logistic regression analysis was performed to evaluate the association between MetS components and melanoma risk in the overall population, adjusting for demographic and metabolic covariates across three models (Supplementary Table S2). Supplementary Table 3 presents the quartile information of each continuous variable in multivariate logistic regression analysis. Age showed a significant trend across all models, with the second quartile (Q2) showing a significant association with increased melanoma risk in the fully adjusted Model 3 (OR: 2.61, 95% CI: 1.10–6.23, p = 0.031). No significant association was observed in gender-specific analysis and alcohol consumption level. In racial/ethnic groups, other race exhibited a significant association across three models. Among the metabolic variables examined, total cholesterol levels in the third quartile (Q3) was associated with a reduced risk in Model 1 (OR: 0.57, 95% CI: 0.35–0.95, p = 0.032), but this effect diminished in Model 3. Triglycerides, fasting glucose, TyG, TyG-WC, and TyG-WHtR showed no significant associations in the fully adjusted Model 3.

Subgroup analysis of metabolic syndrome components and melanoma risk in males

Fig. 4
figure 4

Subgroup analysis of the association between Metabolic Syndrome components and melanoma risk in males. Note: Age (years): Q1 (≦ 33.00), Q2 (33.00 ~ 48.00), Q1 + Q2 (≦ 48.00), Q3 (48.00 ~ 62.00), Q4 (> 62.00); Total cholesterol (mg/dl): Q1 (≦ 161.00), Q2 (161.00 ~ 187.00), Q3 (187.00, 216.00), Q4 (> 216.00); Triglycerides (mg/dl): Q1 (≦ 38.00), Q2 (38.00 ~ 46.00), Q3 (46.00 ~ 55.00), Q4 (> 55.00); HDL-c (mg/dl): Q1 (≦ 75.00), Q2 (75.00 ~ 107.00), Q3 (107.00 ~ 160.00), Q4 (> 160.00); Fasting glucose (mg/dl): Q1 (≦ 96.00), Q2 (96.00 ~ 103.00), Q3 (103.00 ~ 114.00), Q4 (> 114.00); TyG: Q1 (≦ 8.25), Q2 (8.25 ~ 8.66), Q3 (8.66 ~ 9.10), Q4 (> 9.10); PHQ-9 score: Q1 (0.00 ~ 0.00), Q2 (0.00 ~ 1.00), Q3 (1.00 ~ 4.00), Q4 (4.00 ~ 27.00), Q3 + Q4 (> 1.00); Sleep hours: Q1 (≦ 6.00), Q2 (6.00 ~ 7.00); Q3 (7.00 ~ 8.00); Q4 (> 8.00).

Full size image

Subgroup analysis was conducted to examine the association between the number of MetS components and melanoma risk in males across various demographic and metabolic variables (Fig. 4). In the forest plot, significant interactions were observed with total cholesterol, triglycerides, HDL-c, fasting glucose and TyG index, indicating that the relationship varies across quartiles of these metabolic markers. Borderline interactions were noted for age, race, sleep hours and PHQ-9 score, suggesting potential heterogeneity across these subgroups. No significant interaction was detected for alcohol consumption level. These findings highlight the influence of specific metabolic factors on the MetS-melanoma association in males, with varying effects across subgroup strata.

Nonlinear association of metabolic syndrome components with melanoma risk in males

Fig. 5
figure 5

Restricted Cubic Spline curve of the association between Metabolic Syndrome components and melanoma risk (OR) in the males.

Full size image

RCS analysis was utilized to assess the association between the number of MetS components and melanoma risk in males, as presented in Fig. 5. The RCS curve indicated an overall association between the number of MetS components and melanoma risk. A stable OR was observed for individuals with 0 to 2 components, followed by a marked increase beyond this threshold, with the OR exceeding 2.0 at four components. However, the test for nonlinearity did not reach statistical significance suggesting that the relationship may not substantially deviate from a linear pattern. Confidence intervals widened at higher component counts, indicating variability that warrants further exploration with larger cohorts.

Discussion

Using data from NHANES spanning 2007 to 2018, this study explores the connections between MetS, CircS and the risk of melanoma, focusing particularly on differences between males and females. The results show a strong link between components of MetS and a higher chance of developing melanoma in males, where these components prove more effective at predicting risk. Including CircS, which accounts for sleep and mood issues, the study highlights the significance to explore how metabolic and sleep-wake cycle problems together influence melanoma. These findings help fill key gaps in current research and lay the groundwork for improving melanoma risk assessment and prevention.

Gender-specific associations and biological plausibility

A key part of this study’s findings is that MetS components strongly predict melanoma risk in males more than in females. This gender difference should be studied further. It might stem from biological factors, such as sex hormones. Androgens, prevalent in males, have been implicated in modulating metabolic pathways and promoting oncogenic processes, including melanoma progression31,32. A cohort study revealed that females with localized melanoma showed a decreased propensity for metastatic spread33. Alternatively, behavioral factors, such as higher occupational UV exposure or lower engagement in preventive health measures among males, could contribute to this disparity34. These observations align with emerging evidence of sex differences in cancer epidemiology, as seen in hepatocellular carcinoma, where male-specific metabolic risks predominate35,36. Our findings thus advocate for gender-tailored screening protocols, prioritizing metabolic health assessments in male populations to mitigate melanoma risk.

Circadian syndrome as an emerging risk factor

The integration of CircS into our study represents a potential advancement, capturing circadian disruption indicators like sleep insufficiency and depression, which reflect the broader implications of modern lifestyle stressors. CircS, by encompassing circadian disruption indicators like sleep insufficiency and depression, captures the broader implications of modern lifestyle stressors. Chronic circadian misalignment, as experienced by night-shift workers, has been linked to impaired immune surveillance and increased oxidative stress, both of which foster tumorigenesis10,37,38. Notably, melatonin—whose suppression is a hallmark of circadian disruption—has been shown to reduce proliferation in a variety of melanoma cell lines and plays a role by reducing mitochondrial function, glycolysis and glucose uptake39. Besides, A. Alvarez-Artime demonstrated that melatonin induces G2/M phase cell cycle arrest and inhibits proliferation and migration of melanoma cells in the murine B16-F10 melanoma model40. Our findings indicate that CircS is associated with melanoma risk in univariate analysis, suggesting it may serve as a potential composite risk marker integrating metabolic and circadian influences, though this association was not confirmed in multivariable models adjusting for MetS components. This holistic approach resonates with recent calls to incorporate circadian health into cancer risk models, as evidenced by studies linking night-shift work to breast and colorectal cancers41,42. Incorporating CircS into clinical risk stratification could enhance the precision of melanoma prevention, particularly for high-risk groups exposed to circadian stressors.

Predictive power of tyg-based metrics

Our study highlights the TyG and its derivatives (TyG-WC, TyG-WHtR) as potent predictors of melanoma risk, particularly in males. These cost-effective surrogates for insulin resistance showed stronger associations than traditional lipid markers, based on ROC analysis, underscoring the centrality of insulin signaling in melanoma pathogenesis. Insulin resistance is known to drive chronic inflammation and upregulate growth factors like IGF-143,44, which promote melanocyte transformation45,46. The superior performance of TyG-based metrics aligns with their established utility in other malignancies, such as pancreatic cancer, where they predict disease risk with high sensitivity47,48. The accessibility of TyG indices positions them as practical tools for population-level screening, especially in resource-constrained settings. Our findings thus reinforce the metabolic-cancer nexus and advocate for the integration of TyG-based assessments into melanoma risk profiling, offering a scalable approach to identify at-risk individuals.

Dose-response relationship and preventive implications

The RCS analysis in males reveals a dose-response relationship between the number of MetS components and melanoma risk, with a marked increase beyond two components. This threshold effect suggests that cumulative metabolic derangements exponentially heighten susceptibility, offering a critical insight for prevention: modest improvements in metabolic health could yield substantial risk reductions. This pattern mirrors dose-response dynamics observed in other cancers, such as colorectal cancer, where escalating metabolic risk factors correlate with disease incidence49,50,51. The identification of a risk threshold at two MetS components provides a tangible target for clinical interventions, such as lifestyle modifications addressing obesity or hyperglycemia. These findings underscore the urgency of early metabolic management to curb melanoma risk, particularly in males and highlight the potential of population-based strategies to alleviate the growing melanoma burden.

Limitations

As noted in Methods section, melanoma was identified through the NHANES cancer questionnaire. While standard, this introduces possible misclassification. Detailed data on UV exposure, the principal environmental risk factor for melanoma, were not available and may confound the observed associations.

The high proportion of missing data of some variables, such as triglycerides, and alcohol consumption level, may affect the interpretation of these findings. Future studies with more complete data or imputation could help address this.

Moreover, the cross-sectional design of this study limits our ability to draw causal inferences, necessitating longitudinal research to establish temporal relationships. Although melanoma diagnoses were collected using standardized NHANES protocols, reliance on self-reported data introduces the potential for recall bias. Nevertheless, the survey’s methodological rigor helps to mitigate this limitation. Furthermore, unmeasured confounding variables—such as detailed UV exposure histories and genetic predispositions—may have influenced our findings34. Future studies should incorporate these factors to strengthen and refine the observed associations.

Future research directions

Our findings open several avenues for further investigation. Mechanistic studies are needed to elucidate the pathways linking metabolic and circadian dysregulation to melanoma, particularly the male-specific predominance. For instance, exploring the role of insulin signaling or melatonin pathways could clarify these associations39,52,53. Prospective cohort studies across diverse populations could validate the predictive utility of CircS and TyG metrics, while also exploring gender-specific differences in these associations to address potential variations in risk profiles. Randomized controlled trials targeting metabolic or circadian interventions—such as weight loss programs or sleep optimization—could test their efficacy in reducing melanoma risk. Additionally, examining interactions between metabolic factors and environmental exposures, such as UV radiation, could provide a more comprehensive understanding of melanoma etiology.

Conclusion

Analysis of NHANES data from 2007 to 2018 reveals links between MetS, CircS and melanoma risk, with a focus on gender differences. In males, MetS strongly predicts melanoma risk, showing a clear gender-specific pattern. A dose-response trend appears in males, where melanoma risk increases sharply with more MetS components, indicating that metabolic health may be relevant for identifying individuals at higher melanoma risk. CircS also shows an association with melanoma risk, though its predictive ability is less than that of MetS components. In females, TyG and triglyceride levels outperform both MetS and CircS in predicting melanoma risk, yet these markers remain less effective than MetS components in males. These findings connect metabolic and circadian factors to melanoma, deepening knowledge of its causes. Further research should explore the biological reasons for these links and test them in diverse groups.