Vitamin D status is influenced by well-known determinants, but factors associated with low 25-hydroxyvitamin D levels in the cutaneous melanoma population are not well defined. The aim of this study was to confirm the well-known determinants and to assess new determinants for 25-hydroxyvitamin D levels in a cutaneous melanoma population. In a prospectively included cohort of 387 patients with cutaneous melanoma the association of 25-hydroxyvitamin D levels with sex, age, body mass index, time of blood withdrawal, Fitzpatrick phototype, vitamin D supplementation, score for intensity of lifetime sun exposure, smoking, education level, hair and skin colour, eye colour, total number of benign naevi, freckles and parameters of chronic sun damage was investigated. In addition, 25-hydroxyvitamin D levels were correlated with pathological parameters of the primary tumour and melanoma stage (8th edition of the American Joint Committee on Cancer (AJCC). Univariate and multivariate logistic regressions were performed using R software. The following factors had a significant effect on vitamin D status: body mass index, seasonal time of blood sampling, vitamin D supplementation, and a subtype of skin, and hair colour.

Key words: melanoma; vitamin D; body mass index.

Accepted Mar 21, 2022; Epub ahead of print Mar 21, 2022

Acta Derm Venereol 2022; 102: adv00692.

DOI: 10.2340/actadv.v102.262

Corr: Julie De Smedt, School of Biomedical sciences, Department of Oncology, Catholic University of Leuven, Herestraat 49, BE-3000 Leuven, Belgium. E-mail: marjan.garmyn@uzleuven.be

SIGNIFICANCE

Low levels of 25-hydroxyvitamin D are associated with a worse outcome in patients with melanoma. Vitamin D status in healthy persons is influenced by well-known determinants. This study investigated the determinants of 25-hydroxyvitamin D levels in patients with cutaneous melanoma. The results showed a significant effect of the body mass index, seasonal time of blood sampling, use of vitamin D suppelements and a subtype of skin and hair color on 25-hydroxyvitamin D levels in melanoma patients. Current trials investigating the benefit of vitamin D supplementation in patients with melanoma, must take overweight and obesity into account.

INTRODUCTION

Cutaneous melanoma (CM) is a form of skin cancer with increasing incidence rates worldwide. In 2020, an estimated 9,272 deaths in central and eastern Europe were caused by CM (1). According to the National Cancer Institute in the USA, CM was responsible for 5.6% of all new cases of cancer and 1.2% of all cancer deaths in 2021 (2).

Vitamin D (VD) is a secosteroid produced mainly in the human skin upon ultraviolet B (UVB) radiation. The other source of VD is exogenous via dietary intake or supplements. Metabolization of VD is required to form biologically active products. The classical metabolic pathway is a 2-step process with a 25-hydroxylation in the liver to form 25-hydroxyvitamin D (25OHD), followed by 1-alpha hydroxylation in the kidney to form 1-alpha,25-dihydroxyvitamin D (3). An alternative metabolic pathway of VD is via cytochrome P450 side-chain cleavage enzyme (CYP11A1), whereby multiple hydroxymetabolites of VD are produced (4–8). Measurement of the level of 25OHD in blood serum is the best index of VD stores (9, 10). However, this may have some limitations, since not all metabolites are measured. 25OHD levels in the general population are influenced by well-known determinants, such as body mass index (BMI), age, gender, season of blood sampling, sun protection (clothing and use of sunscreen), incidence of several chronic illnesses, and skin type (11). Levels of 25OHD are also influenced by genetic variants of VD pathway genes (12). Levels of 25OHD below 20 ng/ml are associated with the greatest risk of cancer, infections, cardiovascular and metabolic diseases (13).

Besides its well-known function in calcium/phosphate homeostasis, VD has also pleiotropic anti-cancer effects (3). In vitro and in vivo studies have demonstrated an anti-melanoma activity of VD, with effects on cellular growth, differentiation, apoptosis, malignant cell invasion, and metastasis (14, 15).

Various observational studies have shown a beneficial effect of higher 25OHD levels on overall survival, relapse-free survival, and prognostic pathology parameters, such as Breslow thickness and ulceration of the primary tumour (16–22). In molecular and clinicopathological studies, defects in VD signalling pathways are correlated with CM progression and reduced disease outcome (23).

Given the association of VD status and clinical outcome, the question arises as to whether VD supplementation is useful in the (adjuvant) treatment of patients with CM. If VD supplementation proves useful, as currently assessed in ongoing trials, one question will be which patients with CM are prone to have low 25OHD levels, and hence would potentially benefit more from supplemental therapy with VD. The aim of this study is to define which factors are associated with 25OHD levels < 20 ng/ml. Thus, in a well-defined prospective population of patients with CM, levels of 25OHD lower than 20 ng/ml were assessed, and 25OHD levels were correlated with basic patient demographics. A further aim was to investigate the relationship between 25OHD levels, tumour-node-metastasis (TNM) stage (8th AJCC) (24) and histopathological parameters in the study population.

MATERIALS AND METHODS

Study population, recruitment and study procedure

This study is part of a multicentre randomized double-blind placebo-controlled phase III trial investigating the effect of VD supplementation on melanoma outcome (ViDMe trial). The full protocol of the ViDMe trial has been published (25). Briefly, patients with stage IB to III CM (according to the 7th AJCC staging), age 18–80 years, were randomized in the ViDMe trial at the University Hospitals Leuven, the University Hospitals of Antwerp, the Centre Hospitalier Universitaire de Liège in Belgium and at the Clinical Center at the University of Debrecen in Hungary if the only treatment was surgery. Most patients were recruited in the study centres in Belgium, only 4 patients were recruited in Hungary. For inclusion and exclusion criteria we refer to the published study protocol (25). The ViDMe trial, which is still ongoing, was approved by the local ethics committees, and written informed consent was obtained from all participants. A total of 387 patients prospectively recruited from Q4 2012 to Q1 2020 with a baseline serum sample available were included in this study.

Determination of clinical parameters

Basic demographic data on age, sex, tumour site and clinical stage were collected. At randomization, patients were questioned on ethnicity, current smoking status, highest education level, personal medical history, previous and current sun exposure. Current (concomitant) drug intake, VD supplementation (via over-the-counter supplements before randomization) and calcium supplementation were registered. A full skin examination was performed by a physician to assess skin phototype, naevus phenotype, presence of freckles, and signs of chronic sun damage. The latter was determined by the presence of actinic keratosis, solar lentigines and guttate hypomelanosis. The patient’s height and weight were measured in order to estimate BMI.

Determination of pathological parameters

Pathological staging of the primary tumour was performed at the Department of Pathology of the participating centres and is based on the Breslow thickness of the tumour, mitotic rate and presence or absence of ulceration. In addition, the following histological parameters were registered: histological subtype, Clark level, Breslow thickness, ulceration, mitosis, vascular invasion, microsatellites, tumour infiltrating lymphocytes (TILs) and perineural invasion. Nodal micrometastasis were detected by routine H&E staining, supplemented by immunohistochemistry with a cocktail of antibodies directed against melan-A, tyrosinase and gp100 (HMB45). Processing of the sentinel node and reporting of the results was performed according to European Organisation for Research and Treatment of Cancer (EORTC) guidelines.

Determination of 25-hydroxyvitamin D serum levels

Upon inclusion in the ViDMe trial, serum was collected for determination of baseline 25OHD levels following standard procedures using liquid chromatography/tandem mass spectrophotometry in the diagnostic laboratory of Leuven University Hospital (26). This analytically complex method can be considered the gold standard for measurement of 25OHD, but is not available in most routine clinical laboratories. If accurate measurement or differentiation of VD metabolites is desired, this is the method of choice.

Statistical analyses

Clinical and pathological parameters were represented as means and standard deviations (SD) for continuous variables, and frequencies or proportions (%) for categorical variables between group A (25OHD < 20 ng/ml) and group B (25OHD ≥ 20 ng/ml).

Initial melanoma stage (AJCC 7th edition) was recalculated to AJCC 8th edition for statistical analysis of the pathological parameters. Univariate logistic regression analysis and multivariate logistic regressions were performed in order to assess the associations between VD status and the clinical parameters.

The variable selection for the final multivariate model was carried out using a forward stepwise variable selection based on Akaike Information Criterion (AIC) using the data omitting missing values (n = 357). The model with the lowest AIC was choosen as a final model and re-fitted with the whole dataset (n = 387). Multicollinearity was checked by variance inflation factors for each variable with a threshold < 5. The results of the logistic regression models were represented as odds ratios (ORs) on 25OHD levels < 20 ng/ml and 95% confidence intervals (95% CI). All analyses were performed with R software (version 4.0.3).

RESULTS

In total, 387 serum samples were analysed. Of these, 155 (40%) had 25OHD levels < 20 ng/ml and 232 (60%) had 25OHD levels ≥ 20 ng/ml. Mean 25OHD plasma concentration for the whole study population was 23.33 (SD 8.77) ng/ml.

Clinical characteristics of group A (25OHD <20 ng/ml) vs group B (25OHD ≥ 20 ng/ml)

Median age was 56 years in group A and 54 years in group B. In general, inclusion in the ViDMe study occurred more in the winter period (December to February). In group A, most patients (46%) had a Fitzpatrick phototype III and, in group B, most patients (42%) had a Fitzpatrick phototype II. The majority of the general population (86%) was not taking VD supplements at the time of inclusion in the ViDMe study. Fifty-six percent of patients in group A and 52% in group B indicated having normal sun exposure in work and in their free time. The majority of patients in both groups were not smoking at the time of inclusion in the ViDMe study or had smoked in the past. In group A the majority (35%) of patients completed secondary school, compared with in group B the majority (32%) completed vocational university. In both groups, most patients had a light skin, blond or light-brown hair and blue/green/grey eye colour. Upon physical examination in both groups, most patients had a total number of benign naevi < 25, no presence of actinic keratosis, presence of solar lentigines (head and neck region, on the back of the hands and on the shoulders), no freckles, and no guttate hypomelanosis (Table I).

Pathological characteristics of group A (25OHD < 20 ng/ml) vs group B (25OHD ≥ 20 ng/ml)

In both groups the most dominant histological subtype was superficial spreading melanoma, with 58% of patients in group A and 39% of patients in group B having this subtype. Regarding primary tumour thickness, the mean Breslow thickness for group A was 1.48 mm and for group B 1.30 mm, and most tumour tissue in both groups was Clark level 4. For other pathological parameters, both groups showed the same trend, with, mostly, no presence of ulceration, mitosis, vascular invasion, or microsatellites, but presence of TILs. For perineural invasion, information on primary tumour tissue was, in the majority, unknown or missing. In the group with 25OHD levels < 20 ng/ml, most patients were stage IB followed by stage III. In comparison in the group with 25OHD levels ≥ 20 ng/ml, most patients were stage IB, followed by stage IA (Table SI).

Univariate analysis between 25OHD status and clinical parameters

Having a high BMI ≥ 25–30 kg/m2 or ≥ 30 kg/m2 was a significant risk factor for 25OHD levels < 20 ng/ml compared with patients with CM with a normal BMI (≥ 18.5–< 25 kg/m2) (OR 2.21, 95% CI 1.36–3.63, p = 0.002 and OR 5.03, 95% CI 2.73–9.48, p < 0.001, respectively). Patients included during winter and spring had a significantly higher risk for 25OHD levels < 20 ng/ml than patients included during summertime (OR 2.14, 95% CI 1.20–3.91, p = 0.011; OR 2.09, 95% CI 1.11–4.02, p = 0.024, respectively) (Table II).

Patients with CM with light skin and blond/light-brown hair had a higher risk of 25OHD levels ≥ 20 ng/ml compared with patients with CM with light skin and dark-brown/black hair (OR 0.47, 95% CI 0.29–0.77, p = 0.003). Compared with patients with idiopatic guttate hypomelanosis as a sign of chronic sun damage, patients with no such lesions were at risk of lower 25OHD levels (OR 2.05, CI 1.05–4.25, p = 0.043).

Patients with CM not taking VD supplements were at higher risk of low 25OHD levels compared with patients with CM taking VD supplements (OR 7.83, 95% CI 3.33–23.0, p < 0.001). None of the other investigated correlations of low VD levels with the other parameters were statistically significant.

Univariate analysis between 25OHD status, pathological parameters and TNM staging

Univariate analysis was performed with histological subtype, Clark level, mitosis, regression, micosatellites, vascular invasion, perineural invasion, TILs and pTNM 8th edition of AJCC staging. A significant correlation was found between higher tumour staging and probability of having low 25OHD levels (OR 1.14, 95% CI 1.01–1.29). No statistically significant correlation was found between other histopathological parameters and 25OHD levels (Table SII).

Multivariate analysis between 25OHD status and clinical parameters

The final multivariate model is represented in Table III. As all the variance inflation factors were < 5, it is assumed that there is no multicollinearity. Higher BMI (≥ 25–30 kg/m2 or > 30 kg/m2) was associated with higher risk of 25OHD levels < 20 ng/ml compared with CM patients with normal BMI (≥ 18.5–25 kg/m2) (OR 2.23, 95% CI 1.32–3.81 and OR 5.80, 95% CI 2.99–11.6, respectively). Patients whose serum was taken in wintertime for baseline 25OHD serum levels had a significantly higher risk of lower 25OHD levels compared with patients whose serum was taken in summertime (OR 2.74, 95% CI 1.44–5.38). Patients with CM taking no VD supplementation had a higher risk of low 25OHD levels compared with patients taking VD supplements (OR 8.69, 95% CI 3.53–26.4). Similarly to the univariate analysis, patients with light skin and blond or light-brown hair in the multivariate analysis had higher chances of normal 25OHD levels compared with patients with light skin and brown or black hair (OR 0.48, 95% CI 0.28–0.83) (Table III). Other parameters were not selected in the multivariate model.

DISCUSSION

This study investigated the correlation of socio-demographic variables (sex, age and educational level), clinical variables (BMI, Fitzpatrick phototype, hair/skin/eye colour, naevus phenotype, and signs of chronic sun damage), behavioural variables (smoking and sun exposure habits), season, and VD supplementation with 25OHD status in a well-defined carefully phenotyped and assessed CM population.

To our knowledge, this is the first study to investigate a wide range of potential determinants of VD status (measured as 25OHD levels in serum) simultaneously in a prospectively recruited CM population.

Low 25OHD levels (< 20 ng/ml) were found in 40% of patients with CM, and a mean 25OHD level of 23.33 ng/mL) in the CM population in this study. Low 25OHD levels (with the strongest association for levels < 20 ng/ml) are associated with cancer in general (13), but also with CM, as established in previous studies that showed higher 25OHD levels in the serum of healthy controls than in patients at the time of CM diagnosis (27). Higher 25OHD levels at the time of diagnosis of the primary melanoma have been shown to predict a lower risk of relapse and increased survival, independent of Breslow thickness (16, 22). In contrast, in another study, a change in 25OHD levels during follow-up after CM diagnosis, and not the 25OHD level at diagnosis, was ascribed as contributing to the effect on CM survival (28). Further research is necessary in patients with CM to explore the utility of determination of 25OHD status and to decipher the impact of baseline 25OHD levels and VD dynamics.

In the current study population with CM, a clear and significant correlation of overweight (BMI ≥ 25–< 30 and obesity BMI ≥ 30 kg/m2; p < 0.001) with 25OHD levels < 20 ng/ml was observed. This is in line with the literature demonstrating an association between low VD status and obesity in the general population (29). An inverse correlation between serum 25OHD levels, BMI, body weight and fat mass is a consistent finding in both adults and childeren of different ethnicities in a range of geographical locations (30).

It is not known if this association is causative, or is just a bystander effect (i.e. a not yet defined underlying condition may lead to low VD status), or if it simply originates from an association with common lifestyle factors (13, 29).

Multiple possible mechanisms are proposed as the cause of low 25OHD levels in obesity (30): lower sun exposure, low dietary intake, impaired synthesis in the skin, variance in VD metabolism (due to alterations in the vitamin D binding protein (VDBP) or more rapid metabolic clearance), or increased volumetric diluton into a larger tissue volume.

Sunlight habits differ geographically and between cultural groups, but previous studies did not show a difference in sunlight exposure and dietary intake between normal weight and obese adults (32, 33). A similar cutaneous production of VD was observed between normal weight and obese people (33).

Circulating VD is protein bound to VDBP and albumin. Alterations in the concentrations of these proteins and genetic variations in the binding affinity of VDBP affect 25OHD measurement. No differences were found between obese and normal weight people concerning those proteins, and no differences in VDBP genotypes were observed (31, 34).

VD is fat soluble and distributed in the liver, muscle, fat and, in smaller amounts, in other tissues. The volume of these compartments is increased in obese patients. A logical consequence of more widely distributed VD in a larger tissue volume is a lower level of distributed VD in the serum. The volumetric dilution has clinical implications if oral VD supplementation is given to obese patients. A smaller increase in 25OHD levels is seen in obese patients compared with normal weight patients, therefore loading doses of VD will need to be larger to achieve the same 25OHD levels (31).

In the current study population, intake of VD supplements was a significant determinant of 25OHD level (ORno VD supplementation=8.69, p-value < 0.001). Further research is required into the effect of VD supplementation on 25OHD levels and, subsequently, the effect on outcome in patients with CM, especially those in the obese and overweight group.

Univariate and multivariate analyses showed a significant correlation between seasonal collection of blood and 25OHD levels. Blood sampling in winter had a lower chance of normal 25OHD levels compared with the summer period (OR 2.74; p = 0.022). This is in line with the fact that VD is produced in the skin upon UVB-exposure, which is more pronounced during the summer period.

In both univariate and multivariate analysis, a protective effect of having a light skin in combination with blond or light-brown hair was seen in terms of normal 25OHD levels, compared with patients with light skin and brown or black hair. We cannot explain this finding, nor have we found any similar documentation in the literature.

We further investigated sociodemographic, and other clinical and behavioural parameters as possible determinants for 25OHD levels, but there were no statistically significant correlations with regard to sex, age, Fitzpatrick phototype, score for intensity of lifetime sun exposure, smoking, education level, eye colour, total number of benign naevi, solar lentigines, freckles, idiopathic guttate hypomelanosis, and actinic keratosis. In the univariate analysis, there was a significant correlation of idiopathic guttate hypomelanosis with 25OHD levels. Idiopathic guttate hypomelanosis is a marker of chronic sun damage and thus can be linked with higher 25OHD levels. However, the correlation with idiopathic guttate hypomelanosis was no longer significant in the multivariate analysis. Other parameters of chronic sun damage, such as solar lentigines and actinic keratosis, were not statistically significantly associated with 25OHD levels.

Univariate analysis demonstrated a statistically significant association between higher tumour stage and lower 25OHD levels, in line with previous studies (16, 17, 35, 36). We found no significant correlation between 25OHD level and histological subtype of primary tumour tissue, Clark level, mitosis, vascular invasion, perineural invasion, regression, or TILs.

In conclusion, this study found a clear significant association between 25OHD levels and the following factors: BMI, taking VD supplements, season of blood drawing, hair and skin colour. Low levels of 25OHD were associated with a higher tumour stage. Further research is ongoing to investigate whether VD supplementation after removal of the primary melanoma has a protective effect on melanoma outcome and could be used as adjuvant therapy in patients with CM. It will be of interest to investigate the benefit of VD supplementation in the overweight and obese CM population.

ACKNOWLEDGEMENTS

This research project was funded by Kom op tegen Kanker (Stand up to Cancer), the Flemish cancer society, Anticanderfunds, and agency for Innovation by science and technology (IWT) – applied biomedical research program (TBM).

The authors thank Tine Vanhoutvin and Dorien Hunin for their work and help with the ViDMe trial.

The authors have no conflicts of interest to declare.

REFERENCES

International Agency for Research on Cancer, World Health Organization. Factsheets melanoma, globocan 2020. [accessed January 4th, 2022] Available from: https://gco.iarc.fr/today/data/factsheets/cancers/16-Melanoma-of-skin-fact-sheet.pdf.
Melanoma of the skin – Cancer Stat Facts. National Cancer Institute SEER program. [accessed January 4th, 2022] Available from: https://seer.cancer.gov/statfacts/html/melan.html.
Jeon SM, Shin EA. Exploring vitamin D metabolism and function in cancer. Exp Mol Med 2018; 50: 1–14.
Slominski A, Semak I, Zjawiony J, Wortsman J, Li W, Szczesniewski A, et al. The cytochrome P450scc system opens an alternate pathway of vitamin D3 metabolism. FEBS J 2005; 272: 4080–4090.
Guryev O, Carvalho RA, Usanov S, Gilep A, Estabrook RW. A pathway for the metabolism of vitamin D3: unique hydroxylated metabolites formed during catlysis with cytocrhome P450scc (CYP11A1). Proc Natl Acad Sci USA 2003; 100: 14754–14759.
Slominski AT, Kim TK, Shehabi HZ, Semak I, Tang EK, Nguyen MN, et al. In vivo evidence for a novel pathway of vitamin D- metabolism initiated by P450scc and modified by CYP27B1. FASEB J 2012; 26: 3901–3915.
Slominski AT, Li W, Kim TK, Semak I, Wang J, Zjawiony JK, et al. Novel activities of CYP11A1 and their potential physiological significance. J Steroid Biochem Mol Biol 2015; 151: 25–37.
Slominski AT, Chaiprasongsuk A, Janjetovic Z, Kim TK, Stefan J, Slominski RM, et al. Photoprotective properties of vitamin D and lumisterol hydroxyderivatives. Cell Biochem Biophys 2020; 78: 165–180.
Holick MF. Vitamin D status: measurement, interpretation, and clinical application. Ann Epidemiol 2009; 19: 73–78.
Kowalówka M, Główka AK, Karaźniewicz-Łada M, Kosewski G. Clinical significance of analysis of vitamin d status in various diseases. Nutrients 2020; 12: 2788.
Tsiaras WG, Weinstock MA. Factors influencing vitamin D status. Acta Derm Venereol 2011; 91: 115–224.
Vaughan-Shaw PG, O’Sullivan F, Farrington SM, Theodoratou E, Campbell H, Dunlop MG, et al. The impact of vitamin D pathway genetic variation and circulating 25-hydroxyvitamin D on cancer outcome: systematic review and meta-analysis. Br J Cancer 2017; 116: 1092–1110.
Bouillon R, Van Schoor NM, Gielen E, Boonen S, Mathieu C, Vanderschueren D, et al. Optimal vitamin D status: a critical analysis on the basis of evidence-based medicine. J Clin Endocrinol Metab 2013; 98: E1283–E1304.
Slominski AT, Brożyna AA, Zmijewski MA, Jóźwicki W, Jetten AM, Mason RS, et al. Vitamin D signaling and melanoma: role of vitamin D and its receptors in melanoma progression and management. Lab Invest 2017; 97: 706–724.
Slominski AT, Brożyna AA, Skobowiat C, Zmijewski MA, Kim TK, Janjetovic Z, et al. On the role of classical and novel forms of vitamin D in melanoma progression and management. J Steroid Biochem Mol Biol 2018; 177: 159–170.
Newton-Bishop JA, Beswick S, Randerson-Moor J, Chang YM, Affleck P, Elliott F, et al. Serum 25-hydroxyvitamin D3 levels are associated with Breslow thickness at presentation and survival from melanoma. J Clin Oncol 2009; 27: 5439–5444.
Newton-Bishop JA, Davies JR, Latheef F, Randerson-Moor J, Chan M, Gascoyne J, et al. 25-Hydroxyvitamin D2/D3 levels and factors associated with systemic inflammation and melanoma survival in the Leeds Melanoma Cohort. Int J Cancer 2015; 136: 2890–2899.
Gambichler T, Bindsteiner M, Hoxtermann S, Kreuter A. Serum 25-hydroxyvitamin D serum levels in a large German cohort of patients with melanoma. Br J Dermatol 2013; 168: 625–628.
Bade B, Zdebik A, Wagenpfeil S, Gräber S, Geisel J, Vogt T, et al. Low serum 25-hydroxyvitamin d concentrations are associated with increased risk for melanoma and unfavourable prognosis. PLoS ONE 2014; 9: e112863.
Wyatt C, Lucas RM, Hurst C, Kimlin MG. Vitamin D deficiency at melanoma diagnosis is associated with higher Breslow thickness. PLoS ONE 2015; 10: e0126394.
Tierman D, McEnery-Stonelake M, Joyce CJ, Nambudiri VE, Hodi FS, Claus EB, et al. Vitamin D deficiency is associated with a worse prognosis in metastatic melanoma. Oncotarget 2017; 8: 6873–6882.
Fang S, Sui D, Wang Y, Liu H, Chiang YJ, Ross MI, et al. Association of Vitamin D levels with outcome in patients with melanoma after adjustment for C-reactive protein. J Clin Oncol 2016; 34: 1741–1747.
Brożyna AA, Hoffman RM, Slominski AT. Relevance of vitamin D in melanoma development, progression and therapy. Anticancer Res 2020; 40: 473–489.
Gershenwald JE, Scolyer RA. Melanoma Staging: American Joint Committee on Cancer (AJCC) 8th Edition and Beyond. Ann Surg Oncol 2018; 25: 2105–2110.
De Smedt J, Van Kelst S, Boecxstaens V, Stas M, Bogaerts K, Vanderschueren D, et al. Vitamin D supplementation in cutaneous malignant melanoma outcome (ViDMe): a randomized controlled trial. BMC Cancer 2017; 17: 562.
Thomas RL, Jiang L, Adams JS, Xu ZZ, Shen J, Janssen S, et al. Vitamin D metabolites and the gut microbiome in older men. Nat Commun 2020; 11: 5997.
Cattaruzza MS, Pisani D, Fidanza L, Gandini S, Marmo G, Narcisi A, et al. 25-Hydroxyvitamin D serum levels and melanoma risk: a case-control study and evidence synthesis of clinical epidemiological studies. Eur J Cancer Prev 2019; 28: 203–211.
Saiag P, Aegerter P, Vitoux D, Lebbé C, Wolkenstein P, Dupin N, et al. Prognostic value of 25-hydroxyvitamin D3 levels at diagnosis and during follow-up in melanoma patients. J Natl Cancer Inst 2015; 107: djv264.
Pereira-Santos M, Costa PR, Assis AM, Santos CA, Satnos DB. Obesity and vitamin D deficiency: a systematic review and meta-analysis. Obes Rev 2015; 16: 341–349.
Walsh JS, Bowles S, Evans AL. Vitamin D in obesity. Curr Opin Endocrinol Diabetes Obes 2017; 24: 389–394.
Walsh JS, Evans AL, Bowles S, Naylor KE, Jones KS, Schoenmakers I, et al. Free 25-hydroxyvitamin D is low in obesity, but there are no adverse associations with bone health. Am J Clin Nutr 2016; 103: 1465–1471.
Macdonald HM, Mavroeidi A, Aucott LA, Diffey BL, Fraser WD, Ormerod AD, et al. Skin color change in Caucasian postmenopausal women predicts summer-winter change in 25-hydroxyvitamin D: findings from the ANSAViD cohort study. J Clin Endocrinol Metab 2011; 96: 1677–1686.
Wortsman J, Matsuoka LY, Chen TC, Lu Z, Holick MF. Decreased bioavailability of vitamin D in obesity. Am J Clin Nutr 2000; 72: 690–693.
Winters SJ, Chennubhatla R, Wang C, Miller JJ. Influence of obesity on vitamin D-binding protein and 25-hydroxy vitamin D levels in African American and white women. Metabolism 2009; 58: 438–442.
Lim A, Shayan R, Varigos G. High serum vitamin D level correlates with better prognostic indicators in primary melanoma: a pilot study. Aust J Dermatol 2018; 59: 182–187.
Lombardo M, Vigezzi A, Ietto G, Franchi C, Lori V, Masci F, et al. Role of vitamin D serum levels in prevention of primary and recurrent melanoma. Sci Rep 2021; 11: 5815.