Abstract
Purpose
Previous studies from mostly Western populations have suggested possible associations between obesity and melanoma risk. This study aimed to investigate associations between obesity status and melanoma using a nationwide cohort of Koreans.
Materials and Methods
A total of 4,441,403 adults who received a national health examination in 2012 were included from the Korean National Health Insurance Service database, and followed until December 31, 2022. Obesity status was defined based on the body mass index at the baseline health examination. Cox proportional hazards analyses were performed to evaluate associations between obesity status and incident melanoma, with adjustment for confounders. Stratified analyses were performed by sex and menopausal status (in women).
Results
Overall, melanoma risk increased according to obesity status (p for trend=0.024); adjusted hazard ratios (95% confidence intervals) for melanoma risk were 0.766 (0.438–1.340) in underweight; 1.292 (1.072–1.557) in overweight; 1.202 (1.002–1.442) in obesity; and 1.191 (0.798–1.778) in severe obesity compared to normal weight (reference). In stratified analyses, similar trends to those of the overall study population were observed among men and premenopausal women (p for trend=0.052 in men and 0.036 in premenopausal women). Among premenopausal women, the risk of melanoma increased linearly with obesity status. Meanwhile, among postmenopausal women, melanoma risk showed no significant difference or trend according to obesity status.
Conclusion
Overweight and obesity were associated with increased risk of melanoma in a population-based cohort of Koreans. Obese individuals, especially men and premenopausal women, may require more thorough prevention and screening strategies for melanoma.
Keywords: Obesity, Overweight, Melanoma, Skin neoplasms, Cohort studies
Introduction
Malignant melanoma (hereafter referred to as melanoma) is the most serious type of skin cancer, with 59,000 estimated global deaths in 2022 [1,2]. The incidence of melanoma has steadily increased, especially in populations with light skin colors [2]. In Korea, the incidence of melanoma is lower than most Western countries, but the age-standardized annual incidence rate (per 100,000) increased from 2.6 to 3.0 during 2004-2017, and the prevalence rate increased from 3.6 to 6.2 [3]. In many countries with increasing obesity prevalence like Korea, the burdens of non-communicable diseases like metabolic syndrome and cancer are predicted to increase continually [4]. Therefore, prevention of melanoma is important for health promotion in both Western and Asian populations.
Obesity, a rapidly increasing global epidemic, is known to elevate risk of at least 13 types of cancers including those of the endometrium, colorectum, and postmenopausal breast [4]. Also, evidence linking obesity to other cancers is increasingly being uncovered [4]. Meanwhile, the association between obesity and melanoma has been suggested, but the results are inconclusive. Several studies have observed positive associations between obesity and melanoma risk [5-7], while others have reported no association [8,9]. S1 Table summarizes previous major investigations on the obesity-melanoma association. Some findings suggest differences in associations according to sex, menopausal status, and ethnicity [7,10,11]. To date, most previous studies on the relationship between obesity and melanoma are based on Western populations [6-10], and research on Asian populations is scarce. Also, large population-based studies using nationally representative samples are limited, especially in non-Caucasian ethnicities (S1 Table).
Thus, we aimed to investigate the association between obesity and risk of incident melanoma using a population-based nationwide cohort of Koreans. We also aimed to evaluate differences in associations by sex and menopausal status.
Materials and Methods
1. Data source
This retrospective cohort study used the database of the Korean National Health Insurance Service (NHIS). In Korea, the NHIS provides mandatory health insurance coverage to nearly the entire population (97%) as a single insurer, and Medical Aid covers the remaining 3%; data from both groups are included in the NHIS database. The database includes demographics and medical records, such as diagnostic codes by the International Classification of Diseases 10th revision (ICD-10), medical facility usage, and medication prescriptions based on claims data, along with the national health examination data of its beneficiaries.
The national health examination is a complimentary biennial (or annual) health screening program provided to NHIS enrollees aged 40 years or older or employed people of any age. The health examination data include information on anthropometric measurements, laboratory results, lifestyle factors, and medical history based on questionnaires [12].
2. Study population
From the NHIS database, 4,910,068 Korean adults aged 20 years or older who received a national health examination in 2012 were initially identified. Among them, 119,928 individuals who were previously diagnosed with cancer based on diagnostic codes were excluded. Individuals with missing data on baseline health examination results (n=303,739) and individuals who were diagnosed with cancer within a one-year lag period from baseline (n=44,998) were also excluded sequentially. Finally, 4,441,403 individuals were included as the study population (Fig. 1).
Fig. 1.
Flow chart of study population selection.
3. Exposure: obesity
Obesity was assessed by anthropometric measurements performed at the baseline national health examination. Body mass index (BMI) was calculated as body weight in kilograms divided by the square of height in meters. Obesity status at baseline was categorized into underweight (BMI < 18.5 kg/m2), normal weight (≥ 18.5 and < 23 kg/m2), overweight (≥ 23 and < 25 kg/m2), obesity (≥ 25 and < 30 kg/m2), and severe obesity (≥ 30 kg/m2) according to the World Health Organization’s criteria for the Asia-Pacific region [13].
4. Covariates
Demographic data were obtained from the NHIS eligibility database. Income level was dichotomized at the lowest 25%. Lifestyle factors were assessed by questionnaires at the national health examination. Smoking status was categorized into never smoker, ex-smoker, or current smoker, with intensity assessed by 20 pack-year intervals. Drinking status was categorized into non-drinker, mild drinker (< 30 g), or heavy drinker (≥ 30 g) according to average alcohol intake per day. Regular physical activity was defined as moderate physical activity ≥ 5 days per week or vigorous physical activity ≥ 3 days per week. Comorbidities including diabetes mellitus, hypertension, and dyslipidemia were defined by claim records. In Korea, all information on diagnostic codes at the time of medical institution usage and prescribed medications is collected in the NHIS claims database. Using this data, individuals were defined as having a specific condition if there was any record of medical facility visit under the corresponding ICD-10 diagnostic code with relevant medication prescription records (diabetes mellitus: E10-E14 with antidiabetic medication; hypertension: I10-I13, I15 with antihypertensive medication; dyslipidemia: E78 with antidyslipidemic medication) prior to the baseline health examination.
5. Study outcome: melanoma
The primary outcome was incident melanoma, defined by ICD-10 diagnostic code C43 and ascertained by cancer co-payment reduction program code V193. Reliability of the definition of cancer by primary diagnosis claims data along with co-payment reduction program claims from the NHIS was verified in previous research [14]. The co-payment reduction program of the Korean NHIS is established to cover 95% of the expenses of its beneficiaries diagnosed with severe diseases like cancer. Therefore, almost all Korean patients with cancer are registered in this program, and the ascertainment of cancer diagnosis by this program code is regarded as reliable [14].
The study subjects were followed from a one-year lag period after the baseline health examination, considering the increased chance for detection of prevalent cancer following the health examination, to the incidence of melanoma, death, or December 31, 2022, whichever came first.
6. Statistical analysis
Descriptive analyses were performed to determine baseline characteristics of the study population. Student’s t tests (for continuous variables) and chi-square tests (for categorical variables) were used for comparisons according to obesity. To evaluate incidence of melanoma according to obesity status, Cox proportional hazards analyses were conducted using the normal weight group as a reference. The proportional hazards assumption was tested using Schönfeld residuals. Model 1 was unadjusted, and Model 2 was adjusted for possible confounders of sex [2], age [2], income level [15], diabetes mellitus [16], smoking [17], drinking [17], and regular physical activity [18]. Additionally, a sensitivity analysis using BMI as a continuous variable to evaluate the melanoma risk increase per 1-unit (kg/m2) increment in BMI was conducted. Stratified analyses by sex and menopausal status (in women) were also performed. We classified menopausal status based on age of 50 years, as the NHIS general health screening database lacked data on menopausal status [19].
All statistical analyses were performed using SAS ver. 9.4 (SAS Institute Inc.). A p-value < 0.05 was considered statistically significant.
Results
Among the study population of 4,441,403 individuals, 161,377 (3.6%) were underweight; 1,717,116 (38.7%) were normal weight; 1,083,665 (24.4%) were overweight; 1,303,152 (29.3%) were obese; and 176,093 (4.0%) were severely obese. Baseline characteristics of the study population were compared according to the presence of obesity (BMI ≥ 25 kg/m2) (Table 1). Participants with obesity were more likely to be older, male, in a high-income level, current smoker, heavy drinker, perform regular physical activity, have more comorbidities (diabetes mellitus, hypertension, and dyslipidemia), and have larger waist circumference than participants without obesity (all p < 0.001).
Table 1.
Baseline characteristics of the study population
| Total (n=4,441,403) | Weight status |
p-value | ||
|---|---|---|---|---|
| Without obesitya) (n=2,962,158) | With obesity (n=1,479,245) | |||
| Age (yr) | 47.98±13.73 | 47.30±13.96 | 49.35±13.15 | < 0.001 |
| Sex | ||||
| Male | 2,410,087 (54.3) | 1,474,396 (49.8) | 935,691 (63.3) | < 0.001 |
| Premenopausal women (age < 50 yr) | 1,002,977 (22.6) | 820,136 (27.7) | 182,841 (12.4) | |
| Postmenopausal women (age ≥ 50 yr) | 1,028,339 (23.2) | 667,626 (22.5) | 360,713 (24.4) | |
| Income | ||||
| Q1, Medical Aid | 827,380 (18.6) | 563,290 (19.0) | 264,090 (17.9) | < 0.001 |
| Q2 | 953,374 (21.5) | 660,056 (22.3) | 293,318 (19.8) | |
| Q3 | 1,203,737 (27.1) | 798,437 (27.0) | 405,300 (27.4) | |
| Q4 | 1,456,912 (32.8) | 940,375 (31.8) | 516,537 (34.9) | |
| Smoking | ||||
| No | 2,647,612 (59.6) | 1,856,577 (62.7) | 791,035 (53.5) | < 0.001 |
| Ex, < 20 PY | 487,713 (11.0) | 296,532 (10.0) | 191,181 (12.9) | |
| Ex, ≥ 20 PY | 199,978 (4.5) | 108,055 (3.7) | 91,923 (6.2) | |
| Current, < 20 PY | 731,656 (16.5) | 463,969 (15.7) | 267,687 (18.1) | |
| Current, ≥ 20 PY | 374,444 (8.4) | 237,025 (8.0) | 137,419 (9.3) | |
| Drinking | ||||
| No | 2,243,718 (50.5) | 1,539,719 (52.0) | 703,999 (47.6) | < 0.001 |
| Mild | 1,867,697 (42.1) | 1,235,124 (41.7) | 632,573 (42.8) | |
| Heavy | 329,988 (7.4) | 187,315 (6.3) | 142,673 (9.6) | |
| Regular physical activity | 844,959 (19.0) | 550,104 (18.6) | 294,855 (19.9) | < 0.001 |
| Comorbidity | < 0.001 | |||
| Diabetes mellitus | 438,089 (9.9) | 224,394 (7.6) | 213,695 (14.5) | < 0.001 |
| Hypertension | 1,204,521 (27.1) | 614,893 (20.8) | 589,628 (39.9) | < 0.001 |
| Dyslipidemia | 900,046 (20.3) | 484,399 (16.4) | 415,647 (28.1) | < 0.001 |
| Height (cm) | 164.13±9.25 | 163.75±8.98 | 164.91±9.72 | < 0.001 |
| Body weight (kg) | 64.32±11.95 | 59.12±8.70 | 74.72±10.71 | < 0.001 |
| BMI (kg/m2) | 23.77±3.28 | 21.96±1.96 | 27.38±2.25 | < 0.001 |
| Waist circumference (cm) | 80.28±9.29 | 76.20±7.33 | 88.44±7.18 | < 0.001 |
Values are presented as number (%) or the mean±standard deviation. BMI, body mass index; PY, pack-years; Q, quartile.
Obesity was defined as BMI ≥ 25 kg/m2.
A total of 730 (0.02%) participants was diagnosed with incident melanoma during an average of 9.0 (±1.5) years of follow-up. Among the overall study population, risk of incident melanoma increased according to obesity status (p for trend=0.024 in adjusted analysis); compared to the normal weight group (reference), adjusted hazard ratios [aHR] (95% confidence intervals [CI]) for melanoma risk were 0.766 (0.438-1.340) in underweight; 1.292 (1.072-1.557) in overweight; 1.202 (1.002-1.442) in obesity; and 1.191 (0.798-1.778) in severe obesity (Table 2, Fig. 2). Nevertheless, the absolute risk remained low across all obesity status groups, with incidence rates ranging from 0.009 to 0.022 per 1,000 person-years. A sensitivity analysis using BMI as a continuous variable showed similar results to those of the main analysis: aHR (95% CI) for melanoma risk per 1-unit increment in BMI was 1.026 (1.003-1.050) (p=0.027).
Table 2.
Risk of incident melanoma according to obesity status in the overall study population
| Obesity status | No. | Event | Duration (PY) | Incidence rate (per 1,000 PY) | Model 1 |
Model 2a) |
|---|---|---|---|---|---|---|
| HR (95% CI) | aHR (95% CI) | |||||
| Underweight (BMI: < 18.5 kg/m2) | 161,377 | 13 | 1,445,575.4 | 0.009 | 0.618 (0.354-1.081) | 0.766 (0.438-1.340) |
| Normal weight (BMI: ≥ 18.5 and < 23 kg/m2) | 1,717,116 | 226 | 15,532,832.8 | 0.015 | 1 (reference) | 1 (reference) |
| Overweight (BMI: ≥ 23 and < 25 kg/m2) | 1,083,665 | 218 | 9,793,203.7 | 0.022 | 1.530 (1.270-1.843) | 1.292 (1.072-1.557) |
| Obesity (BMI: ≥ 25 and < 30 kg/m2) | 1,303,152 | 246 | 11,758,182.3 | 0.021 | 1.438 (1.201-1.723) | 1.202 (1.002-1.442) |
| Severe obesity (BMI: ≥30 kg/m2) | 176,093 | 27 | 1,590,455.0 | 0.017 | 1.168 (0.784-1.741) | 1.191 (0.798-1.778) |
| p for trend | < 0.001 | 0.024 | ||||
| BMI (kg/m2), per 1-unit increase | 4,441,403 | 730 | 40,120,249.2 | 0.018 | 1.044 (1.022-1.066) | 1.026 (1.003-1.050) |
| p-value | < 0.001 | 0.027 |
BMI, body mass index; CI, confidence interval; HR, hazard ratio; PY, person-years.
Adjusted for sex, age, income level, diabetes mellitus, smoking, drinking, and regular physical activity.
Fig. 2.
Adjusted hazard ratios (95% confidence interval) of melanoma according to obesity status in the overall population (A) and in subgroups by sex (B) and menopausal status (C, D). Adjusted for sex, age, income level, diabetes mellitus, smoking, drinking, and regular physical activity.
In stratified analyses by sex and menopausal status, similar trends to the overall study population were observed among men and premenopausal women (p for trend=0.052 in men and 0.036 in premenopausal women). Among premenopausal women, the risk of melanoma appeared to increase linearly with obesity status, although we could not confirm a significant difference according to obesity status, at least partly, due to the small event numbers (Table 3). Among men, aHR (95% CI) for melanoma risk was 0.923 (0.404-2.108) in underweight; 1.545 (1.188-2.009) in overweight; 1.313 (1.010-1.708) in obesity; and 1.316 (0.719-2.408) in severe obesity compared to the normal weight group (reference). Among premenopausal women, aHR (95% CI) for melanoma risk was 0.720 (0.170-3.052) in underweight; 1.356 (0.711-2.587) in overweight; 1.541 (0.806-2.946) in obesity; and 2.726 (0.949-7.832) in severe obesity. Meanwhile, among postmenopausal women, risk of melanoma showed no linear trend or significant difference according to obesity status (p for trend=0.618). When using BMI as a continuous variable, BMI was significantly associated with melanoma risk only among premenopausal women: aHRs (95% CIs) for melanoma risk per 1-unit increment in BMI were 1.027 (0.992-1.063) among men, 1.085 (1.013-1.162) among premenopausal women, and 1.014 (0.978-1.052) among postmenopausal women (p=0.130, p=0.019, and p=0.449, respectively) (data not shown).
Table 3.
Stratified analyses by sex and menopausal status regarding the risk of incident melanoma according to obesity status
| Obesity status | No. | Event | Duration (PY) | Incidence rate (per 1,000 PY) | Model 1 HR (95% CI) | Model 2a) aHR (95% CI) |
|---|---|---|---|---|---|---|
| Men | ||||||
| Underweight (BMI: < 18.5 kg/m2) | 49,005 | 6 | 424,168.6 | 0.014 | 1.005 (0.441-2.292) | 0.923 (0.404-2.108) |
| Normal weight (BMI: ≥ 18.5 and < 23 kg/m2) | 774,381 | 98 | 6,956,509.2 | 0.014 | 1 (reference) | 1 (reference) |
| Overweight (BMI: ≥ 23 and < 25 kg/m2) | 651,010 | 131 | 5,874,736.2 | 0.022 | 1.583 (1.218-2.057) | 1.545 (1.188-2.009) |
| Obesity (BMI: ≥ 25 and < 30 kg/m2) | 834,711 | 136 | 7,533,082.5 | 0.018 | 1.282 (0.989-1.662) | 1.313 (1.01-1.708) |
| Severe obesity (BMI: ≥ 30 kg/m2) | 100,980 | 12 | 914,909.4 | 0.013 | 0.931 (0.511-1.696) | 1.316 (0.719-2.408) |
| p for trend | 0.228 | 0.052 | ||||
| Premenopausal women (age < 50 yr) | ||||||
| Underweight (BMI: < 18.5 kg/m2) | 88,907 | 2 | 819,821.5 | 0.002 | 0.450 (0.107-1.890) | 0.720 (0.170-3.052) |
| Normal weight (BMI: ≥ 18.5 and < 23 kg/m2) | 563,666 | 28 | 5,165,389.7 | 0.005 | 1 (reference) | 1 (reference) |
| Overweight (BMI: ≥ 23 and < 25 kg/m2) | 167,563 | 14 | 1,527,766.5 | 0.009 | 1.690 (0.890-3.211) | 1.356 (0.711-2.587) |
| Obesity (BMI: ≥ 25 and < 30 kg/m2) | 151,832 | 14 | 1,380,163.0 | 0.010 | 1.871 (0.985-3.554) | 1.541 (0.806-2.946) |
| Severe obesity (BMI: ≥ 30 kg/m2) | 31,009 | 4 | 281,537.9 | 0.014 | 2.620 (0.919-7.469) | 2.726 (0.949-7.832) |
| p for trend | 0.003 | 0.036 | ||||
| Postmenopausal women (age ≥ 50 yr) | ||||||
| Underweight (BMI: < 18.5 kg/m2) | 23,465 | 5 | 201,585.3 | 0.025 | 0.847 (0.345-2.079) | 0.730 (0.297-1.795) |
| Normal weight (BMI: ≥ 18.5 and < 23 kg/m2) | 379,069 | 100 | 3,410,933.8 | 0.029 | 1 (reference) | 1 (reference) |
| Overweight (BMI: ≥ 23 and < 25 kg/m2) | 265,092 | 73 | 2,390,701.0 | 0.031 | 1.041 (0.770-1.408) | 1.002 (0.741-1.355) |
| Obesity (BMI: ≥ 25 and < 30 kg/m2) | 316,609 | 96 | 2,844,936.8 | 0.034 | 1.150 (0.869-1.522) | 1.069 (0.807-1.416) |
| Severe obesity (BMI: ≥ 30 kg/m2) | 44,104 | 11 | 394,007.7 | 0.028 | 0.952 (0.511-1.775) | 0.928 (0.497-1.734) |
| p for trend | 0.407 | 0.618 |
BMI, body mass index; CI, confidence interval; HR, hazard ratio; PY, person-years.
Adjusted for sex, age, income level, diabetes mellitus, smoking, drinking, and regular physical activity.
Discussion
Our nationwide cohort study showed that the risk of incident melanoma increased according to obesity status, with higher risk in participants who are overweight, obese, or severely obese than in normal-weight participants. In stratified analyses by sex and menopausal status, similar trends were observed in men, while a linearly increasing trend was observed in premenopausal women. Postmenopausal women showed no association between obesity and melanoma risk.
In the overall study population, overweight and obesity were associated with increased risk of melanoma. Nevertheless, the absolute risk remained low across all obesity status groups, including the severe obesity group, which is consistent with the reported incidence rate of melanoma in Korea [3]. As melanoma is a rare cancer in Korea, absolute risk increase due to obesity may not be substantial; however, given the rapid rise in obesity and the anticipated increase in melanoma incidence and mortality [2,4], the increase in melanoma risk associated with obesity may pose a growing health concern.
The aHRs for melanoma risk were 1.29 in overweight, 1.20 in obesity, and 1.19 in severe obesity group. Similar result was observed in a previous meta-analysis of case-control studies, which incorporated 2,304 melanoma cases and 2,468 controls from mostly Western populations, presenting odds ratio of 1.36 for melanoma in individuals with overweight and obesity, defined as BMI ≥ 25 kg/m2 [20]. Large cohort studies are rare, and mixed results have been observed [5-7,9,10].
Interestingly, a prospective study of United States veterans has shown a positive relationship between obesity and melanoma, with a far more evident association in black men (relative risk [RR], 2.4) than in white men (RR, 1.3) [7]. In that previous study, obesity was based on discharge diagnosis, and the result was adjusted only for age and calendar year, which may have led to a larger magnitude of association than the actual relationship. Nevertheless, the degree of association differed distinctly between black and white men. In the same vein, several cohort studies among Caucasian populations have observed null associations [8,9,21], whereas a study among a cohort of Korean men has observed an increased melanoma risk in obese individuals (RR, 2.0 in men with BMI 25-26.9 kg/m2 and 2.8 in men with BMI 27-29.9 kg/m2), as in our study [5]. In a previous Korean study, the risk estimates were higher than ours, but were based on a small event number (n=51) in a study population limited to civil servants, private school staff, and their dependents. Additionally, a multi-center study of 261 melanoma patients in Korea observed a significant association between obesity and increased Breslow thickness, suggesting that obesity may promote melanoma growth in Koreans [22]. It is possible that ethnic difference might exist in the association between obesity and melanoma risk. Indeed, previous studies have reported that melanoma has different clinical and genetic characteristics in Asians, African-Americans, and Hispanics relative to Caucasians [23]. Skin color is the main reason for different tumor characteristics by race, resulting from factors like melanin type [2]. As populations with light skin color have greater susceptibility of melanocytes to damage from ultraviolet (UV) radiation [2], and common melanoma subtypes in these populations (such as superficial spread melanoma and lentigo maligna melanoma) mainly arise in regions of high sun exposure [2,23], the increased risk of melanoma associated with UV exposure may overshadow the impact of obesity on melanoma risk. Meanwhile, in populations with dark skin color including African-Americans and Asians, acral lentiginous melanoma is the most common subtype, for which sunlight exposure is not considered a risk factor [2,16,23]. However, as studies of obesity and melanoma risk from non-Caucasian populations, especially Asians, are limited, further research is necessary to elucidate ethnic differences.
Several factors have been postulated to mediate melanoma risk increase in obese individuals. Chronic inflammation with abnormal cell signaling and proliferation is the main suggested molecular mechanism for risk increase in obesity, as in other cancers [24,25]. An inflammatory state may induce aberrant secretion of adipokines (especially leptin), inadequate production of hormone or growth factors, and increase oxidative stress and reactive oxygen species in melanocytes, linking to the tumorigenesis of melanoma [24,25]. Factors like increased body surface area, which can lead to difficulty in skin examination and overlooked skin folds, genetic susceptibility, and low serum vitamin D level are also suggested to contribute to increased melanoma risk in obesity [24,25]. Additionally, considering our hypothesis that the obesity-melanoma association may be more evident in populations where acral lentiginous melanoma is the most common subtype, it is possible that increased plantar pressure in obese individuals [26] might contribute to elevated melanoma risk. Several studies have suggested that physical stress like increased pressure, particularly in soles, may increase the risk of acral lentiginous melanoma [27,28]. Considering a recent prospective study that showed a significant reduction in melanoma risk (57%) among subjects who received bariatric surgery [29], obesity-related increase in melanoma risk may, at least partially, be attributable to reversible mechanisms.
Notably, our analyses showed differences in associations between obesity and melanoma risk by sex and menopausal status, presenting increasing trends in melanoma risk according to obesity status among men and premenopausal women. Previous meta-analysis that stratified data by sex has shown significant associations between overweight or obesity with melanoma risk only in men and not in women [11]. That study analyzed 11 case-control studies and 10 cohort studies and reported a pooled effect estimate of 1.3 for both overweight and obesity in males, which is similar to our result [11]. In that previous research, obesity was not associated with melanoma among females in overall meta-analysis. However, post hoc analysis of studies considering sunlight exposure has shown mutual confounding between exposure to sunlight and obesity in females, suggesting that obese women may be less likely to expose themselves to sunlight, resulting in an obscuration of obesity-melanoma association [11]. As estradiol and estrogen receptor β agonist are known to suppress melanoma cell proliferation, increased estrogen sensitivity in women compared to men was also hypothesized to mitigate the effect of obesity and increased leptin level on melanoma risk [30,31]. Nonetheless, further studies are warranted to explain the mechanisms underlying the association between obesity and melanoma, as our study shows that premenopausal women may be at increased risk of melanoma with increasing obesity category.
Evidence has been insufficient regarding menopausal status, but one large cohort study of women in the United Kingdom examined the effect of BMI on melanoma risk by menopausal status [10]. That study showed no association between BMI and melanoma in overall women, but an increasing trend in melanoma risk by BMI was observed in premenopausal women (RR per 10 unit increase in BMI, 1.62), while no significant difference was observed in postmenopausal women who never used hormone therapy (RR, 0.92), which is similar to our result [10]. The exact mechanism remains unclear, but considering the reports on increased estrogen exposure according to adiposity in postmenopausal women [29,32], it is plausible that estrogen can more evidently counterbalance the effects of obesity on melanoma in postmenopausal women compared to premenopausal women. However, another cohort study in the United Kingdom has observed risk increase only in obese men, with no difference by menopausal status observed in associations between BMI and melanoma among women [30]. Future research is warranted in regard to the obesity-melanoma relationship by detailed menopausal status or sex hormone level, considering confounding from sunlight exposure.
Our study has several limitations. First, information on UV exposure, which may substantially affect melanoma risk [2], was unavailable because it is not included in the NHIS database. Confounding from UV exposure may exist in the obesity-melanoma association, as several previous investigations suggested that obesity might be negatively associated with sunlight exposure time [11,33]. The association between obesity and melanoma risk may have been attenuated in our study than actual association, as obese individuals are likely to have lower levels of UV exposure compared to non-obese individuals. Nevertheless, as acral lentiginous melanoma, for which UV exposure is not considered a risk factor, is the most common type of melanoma in Korea (accounting for approximately 60%-70% of melanomas) [16,34,35], the bias from this potential confounding is unlikely to be substantial. Further studies adjusting for potential confounding from UV exposure are warranted to clarify the obesity-melanoma association in Asians. Second, we lacked precise information on menopausal status or data regarding hormone therapy. Instead, we defined women aged 50 years or older as postmenopausal, based on the reported mean age at menopause in Koreans [19]. However, this approximated definition may have resulted in misclassification from the actual menopausal status in some study subjects. Further studies incorporating precise information on menopausal status are needed. Third, as our study population was limited to those who underwent a national health examination, selection bias may exist, and the study population may have been more conscious about health management than individuals who did not receive a health examination. Additionally, we were unable to examine the effect of time-varying BMI or competing risks due to a lack of the corresponding data. Given that BMI may change over time, such changes may have affected the observed association between obesity and melanoma. Nevertheless, considering that previous reports have shown that significant changes in obesity status over time are uncommon in Korean adults [36], and that there is usually a time lag before such changes affect the development of tumors, the time-varying effect of BMI is presumed to be limited. Competing risks, such as death from other causes, may also have affected the observed melanoma risk. Future research should incorporate additional analyses that consider the potential effects of time-varying BMI or competing risks on the obesity-melanoma association.
Strengths of our study include the large, population-based nationwide cohort; Most of the previous studies had smaller sample size than ours. Our large sample size provided sufficient statistical power to accurately examine the association between obesity with melanoma and enabled stratified analyses by sex and menopausal status. Also, our study provides valuable data on Asians, a population that has been scarcely studied in previous research. Considering the clinical features of melanoma in Asia, which is less influenced by sunlight [16], our study may have an advantage over Western studies in examining the effects of non-sunlight factors like obesity on melanoma. We also incorporated various covariates that were not commonly considered in previous studies, such as income level and lifestyle factors.
In conclusion, overweight and obesity were associated with increased risk of incident melanoma in our population-based cohort of Koreans. When stratified by sex and menopausal status, similar trends were observed in men and premenopausal women, whereas postmenopausal women showed no significant association between obesity and melanoma. Obese individuals, especially men and premenopausal women, may require more thorough prevention and screening strategies for melanoma than non-obese individuals.
Footnotes
Ethical Statement
This study was approved by the Institutional Review Boards of Soongsil University (IRB number: SSU-202007-HR-236-02) and Samsung Medical Center (IRB number: SMC 2024-08-019). The requirement for written informed consent was waived as a de-identified database was used.
Author Contributions
Conceived and designed the analysis: Koo HY, Han K, Cho IY, Shin DW.
Collected the data: Han K.
Contributed data or analysis tools: Han K.
Performed the analysis: Han K, Jung J.
Wrote the paper: Koo HY, Han K, Park J, Kim S, Cho H, Cho IY, Shin DW.
Conflict of Interest
Conflict of interest relevant to this article was not reported.
Funding
This study was supported by a grant from the Korean Foundation for Cancer Research (KFCR) (grant number: CB-2022-A-1).
Electronic Supplementary Material
Supplementary materials are available at Cancer Research and Treatment website (https://www.e-crt.org).
References
- 1.Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74:229–63. doi: 10.3322/caac.21834. [DOI] [PubMed] [Google Scholar]
- 2.Lopes J, Rodrigues CM, Gaspar MM, Reis CP. Melanoma management: from epidemiology to treatment and latest advances. Cancers (Basel) 2022;14:4652. doi: 10.3390/cancers14194652. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kim T, Yoon S, Shin DE, Lee SC, Oh J, Lee SY, et al. Incidence and survival rates of cutaneous melanoma in South Korea using Nationwide Health Insurance claims data. Cancer Res Treat. 2022;54:937–49. doi: 10.4143/crt.2021.871. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Avgerinos KI, Spyrou N, Mantzoros CS, Dalamaga M. Obesity and cancer risk: Emerging biological mechanisms and perspectives. Metabolism. 2019;92:121–35. doi: 10.1016/j.metabol.2018.11.001. [DOI] [PubMed] [Google Scholar]
- 5.Oh SW, Yoon YS, Shin SA. Effects of excess weight on cancer incidences depending on cancer sites and histologic findings among men: Korea National Health Insurance Corporation Study. J Clin Oncol. 2005;23:4742–54. doi: 10.1200/JCO.2005.11.726. [DOI] [PubMed] [Google Scholar]
- 6.Samanic C, Chow WH, Gridley G, Jarvholm B, Fraumeni JF Jr. Relation of body mass index to cancer risk in 362,552 Swedish men. Cancer Causes Control. 2006;17:901–9. doi: 10.1007/s10552-006-0023-9. [DOI] [PubMed] [Google Scholar]
- 7.Samanic C, Gridley G, Chow WH, Lubin J, Hoover RN, Fraumeni JF Jr. Obesity and cancer risk among white and black United States veterans. Cancer Causes Control. 2004;15:35–43. doi: 10.1023/B:CACO.0000016573.79453.ba. [DOI] [PubMed] [Google Scholar]
- 8.Praestegaard C, Kjaer SK, Christensen J, Tjonneland A, Halkjaer J, Jensen A. Obesity and risks for malignant melanoma and non-melanoma skin cancer: results from a large Danish prospective cohort study. J Invest Dermatol. 2015;135:901–4. doi: 10.1038/jid.2014.438. [DOI] [PubMed] [Google Scholar]
- 9.Rapp K, Schroeder J, Klenk J, Stoehr S, Ulmer H, Concin H, et al. Obesity and incidence of cancer: a large cohort study of over 145,000 adults in Austria. Br J Cancer. 2005;93:1062–7. doi: 10.1038/sj.bjc.6602819. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Reeves GK, Pirie K, Beral V, Green J, Spencer E, Bull D, et al. Cancer incidence and mortality in relation to body mass index in the Million Women Study: cohort study. BMJ. 2007;335:1134. doi: 10.1136/bmj.39367.495995.AE. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sergentanis TN, Antoniadis AG, Gogas HJ, Antonopoulos CN, Adami HO, Ekbom A, et al. Obesity and risk of malignant melanoma: a meta-analysis of cohort and case-control studies. Eur J Cancer. 2013;49:642–57. doi: 10.1016/j.ejca.2012.08.028. [DOI] [PubMed] [Google Scholar]
- 12.Shin DW, Cho J, Park JH, Cho B. National General Health Screening Program in Korea: history, current status, and future direction: a scoping review. Precis Future Med. 2022;6:9–31. [Google Scholar]
- 13.World Health Organization, Regional Office for the Western Pacific . The Asia-Pacific perspective: redefining obesity and its treatment. Health Communications Australia; 2000. [Google Scholar]
- 14.Yang MS, Park M, Back JH, Lee GH, Shin JH, Kim K, et al. Validation of cancer diagnosis based on the National Health Insurance Service Database versus the National Cancer Registry Database in Korea. Cancer Res Treat. 2022;54:352–61. doi: 10.4143/crt.2021.044. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Larsen IK, Myklebust TA, Babigumira R, Vinberg E, Moller B, Ursin G. Education, income and risk of cancer: results from a Norwegian registry-based study. Acta Oncol. 2020;59:1300–7. doi: 10.1080/0284186X.2020.1817548. [DOI] [PubMed] [Google Scholar]
- 16.Soh BW, Kim YC, Choi JW. Epidemiology of cutaneous melanoma and melanoma in situ: a 10-year nationwide retrospective cohort study in South Korea. Melanoma Res. 2020;30:297–302. doi: 10.1097/CMR.0000000000000547. [DOI] [PubMed] [Google Scholar]
- 17.Freedman DM, Sigurdson A, Doody MM, Rao RS, Linet MS. Risk of melanoma in relation to smoking, alcohol intake, and other factors in a large occupational cohort. Cancer Causes Control. 2003;14:847–57. doi: 10.1023/b:caco.0000003839.56954.73. [DOI] [PubMed] [Google Scholar]
- 18.Moore SC, Lee IM, Weiderpass E, Campbell PT, Sampson JN, Kitahara CM, et al. Association of leisure-time physical activity with risk of 26 types of cancer in 1.44 million adults. JAMA Intern Med. 2016;176:816–25. doi: 10.1001/jamainternmed.2016.1548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Park CY, Lim JY, Park HY. Age at natural menopause in Koreans: secular trends and influences thereon. Menopause. 2018;25:423–9. doi: 10.1097/GME.0000000000001019. [DOI] [PubMed] [Google Scholar]
- 20.Mehta A, Hasija Y. Meta-analyses of associations of obesity with risk of melanoma and acne based on case-control studies. Biomed Dermatol. 2019;3:3. [Google Scholar]
- 21.Wolk A, Gridley G, Svensson M, Nyren O, McLaughlin JK, Fraumeni JF, et al. A prospective study of obesity and cancer risk (Sweden) Cancer Causes Control. 2001;12:13–21. doi: 10.1023/a:1008995217664. [DOI] [PubMed] [Google Scholar]
- 22.Kim JE, Chung BY, Sim CY, Park AY, Lee JS, Whang KU, et al. Clinicopathologic features and prognostic factors of primary cutaneous melanoma: a multicenter study in Korea. J Korean Med Sci. 2019;34:e126. doi: 10.3346/jkms.2019.34.e126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Higgins S, Nazemi A, Feinstein S, Chow M, Wysong A. Clinical presentations of melanoma in African Americans, Hispanics, and Asians. Dermatol Surg. 2019;45:791–801. doi: 10.1097/DSS.0000000000001759. [DOI] [PubMed] [Google Scholar]
- 24.Karimi K, Lindgren TH, Koch CA, Brodell RT. Obesity as a risk factor for malignant melanoma and non-melanoma skin cancer. Rev Endocr Metab Disord. 2016;17:389–403. doi: 10.1007/s11154-016-9393-9. [DOI] [PubMed] [Google Scholar]
- 25.Neagu M, Dobre EG. In: Obesity and lipotoxicity. Engin AB, Engin A, editors. Springer International Publishing; 2024. New insights into the link between melanoma and obesity; pp. 851–67. [DOI] [PubMed] [Google Scholar]
- 26.Birtane M, Tuna H. The evaluation of plantar pressure distribution in obese and non-obese adults. Clin Biomech (Bristol) 2004;19:1055–9. doi: 10.1016/j.clinbiomech.2004.07.008. [DOI] [PubMed] [Google Scholar]
- 27.Lee JH, Choi YD, Hwang JH, Shin MH, Yun SJ. Frequency of trauma, physical stress, and occupation in acral melanoma: analysis of 313 acral melanoma patients in Korea. Ann Dermatol. 2021;33:228–36. doi: 10.5021/ad.2021.33.3.228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Jung HJ, Kweon SS, Lee JB, Lee SC, Yun SJ. A clinicopathologic analysis of 177 acral melanomas in Koreans: relevance of spreading pattern and physical stress. JAMA Dermatol. 2013;149:1281–8. doi: 10.1001/jamadermatol.2013.5853. [DOI] [PubMed] [Google Scholar]
- 29.Hetemaki N, Savolainen-Peltonen H, Tikkanen MJ, Wang F, Paatela H, Hamalainen E, et al. Estrogen metabolism in abdominal subcutaneous and visceral adipose tissue in postmenopausal women. J Clin Endocrinol Metab. 2017;102:4588–95. doi: 10.1210/jc.2017-01474. [DOI] [PubMed] [Google Scholar]
- 30.Christakoudi S, Tsilidis KK, Riboli E. Prospective associations of leucocyte subtypes and obesity with the risk of developing cutaneous malignant melanoma in the UK Biobank cohort. BMC Cancer. 2024;24:625. doi: 10.1186/s12885-024-12344-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Marzagalli M, Casati L, Moretti RM, Montagnani Marelli M, Limonta P. Estrogen receptor beta agonists differentially affect the growth of human melanoma cell lines. PLoS One. 2015;10:e0134396. doi: 10.1371/journal.pone.0134396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Baglietto L, English DR, Hopper JL, MacInnis RJ, Morris HA, Tilley WD, et al. Circulating steroid hormone concentrations in postmenopausal women in relation to body size and composition. Breast Cancer Res Treat. 2009;115:171–9. doi: 10.1007/s10549-008-0069-3. [DOI] [PubMed] [Google Scholar]
- 33.Dos Santos EA, Cavalheiro LA, Rodrigues D, Machado-Rodrigues A, Silva MR, Nogueira H, et al. Is more prolonged sun exposure associated with healthier life habits and obesity indicators in Portuguese children? Am J Hum Biol. 2023;35:e23918. doi: 10.1002/ajhb.23918. [DOI] [PubMed] [Google Scholar]
- 34.Seo SH, Oh CK, Ko HC, Kim MB, Kwon KS, Bae YC. A 10-year review of cutaneous melanomas in Korea: clinicopathologic features and changes by multidisciplinary team approach. J Eur Acad Dermatol Venereol. 2009;23:1463–5. doi: 10.1111/j.1468-3083.2009.03258.x. [DOI] [PubMed] [Google Scholar]
- 35.Won YH, Hur SG, Lee SC, Chuun IK, Kim YP. Clinical and histopathologic study of malignant melanoma (1980-1990) Korean J Dermatol. 1991;29:193–202. [Google Scholar]
- 36.Kim ES, Jeong JS, Han K, Kim MK, Lee SH, Park YM, et al. Impact of weight changes on the incidence of diabetes mellitus: a Korean nationwide cohort study. Sci Rep. 2018;8:3735. doi: 10.1038/s41598-018-21550-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.