Obesity and Risks for Malignant Melanoma and Non-Melanoma Skin Cancer: Results from a Large Danish Prospective Cohort Study
2014; Elsevier BV; Volume: 135; Issue: 3 Linguagem: Inglês
10.1038/jid.2014.438
ISSN1523-1747
AutoresCamilla Præstegaard, Susanne K. Kjær, Jane Christensen, Anne Tjønneland, Jytte Halkjær, Allan Jensen,
Tópico(s)Nonmelanoma Skin Cancer Studies
Resumobasal cell carcinoma body mass index confidence interval hip circumference hazard ratio malignant melanoma non-melanoma skin cancer squamous cell carcinoma UV radiation waist circumference waist:hip ratio Obesity has increased alarmingly during recent decades in the Western world and is a risk factor for several cancers (World Cancer Research FundWorld Cancer Research Fund/American Institute for Cancer Reseach: Food, Nutrition, Physical Activity, and the Prevention of Cancer: A Global Perspective. AICR, Washington, DC2007http://www.dietandcancerreport.org/cancer:resource:center/downloads/Second:Expert:Report:full.pdfGoogle Scholar; Lahmann et al., 2010Lahmann P.H. Cust A.E. Friedenreich C.M. et al.Anthropometric measures and epithelial ovarian cancer risk in the European Prospective Investigation into Cancer and Nutrition.Int J Cancer. 2010; 126: 2404-2415PubMed Google Scholar). However, evidence has not been fully established for an association between obesity and malignant melanoma (MM) and non-melanoma skin cancer (NMSC). Recently, results from a large meta-analysis showed that men with high body mass index (BMI) had an increased risk for MM, whereas no association between BMI and risk for MM was found among women (Sergentanis et al., 2013Sergentanis T.N. Antoniadis A.G. Gogas H.J. et al.Obesity and risk of malignant melanoma: a meta-analysis of cohort and case-control studies.Eur J Cancer. 2013; 49: 642-657Abstract Full Text Full Text PDF PubMed Scopus (114) Google Scholar). Fewer studies have addressed the associations between BMI and risk for NMSC and the results are conflicting (Gerstenblith et al., 2012Gerstenblith M.R. Rajaraman P. Khaykin E. et al.Basal cell carcinoma and anthropometric factors in the U.S. radiologic technologists cohort study.Int J Cancer. 2012; 131: 149-155Crossref PubMed Scopus (28) Google Scholar; Nagel et al., 2012Nagel G. Bjorge T. Stocks T. et al.Metabolic risk factors and skin cancer in the Metabolic Syndrome and Cancer Project (Me-Can).Br J Dermatol. 2012; 167: 59-67Crossref PubMed Scopus (32) Google Scholar; Pothiawala et al., 2012Pothiawala S. Qureshi A.A. Li Y. et al.Obesity and the incidence of skin cancer in US Caucasians.Cancer Causes Control. 2012; 23: 717-726Crossref PubMed Scopus (61) Google Scholar). However, the vast majority of studies used only BMI as a measure of obesity, although it is increasingly accepted that hip circumference (HC), waist circumference (WC), and waist:hip ratio (WHR) more accurately reflect a person's body fat distribution (Pischon et al., 2008Pischon T. Nöthlings U. Boeing H. Obesity and cancer.Proc Nutr Soc. 2008; 67: 128-145Crossref PubMed Scopus (270) Google Scholar). We here addressed the association between obesity, measured as BMI, WC, HC, and WHR, and risks for MM and NMSC, by the use of data from the prospective Danish Diet, Cancer, and Health Cohort study including 57,053 participants enrolled from 1993 to 1997. A detailed description of the cohort has been provided elsewhere (Tjonneland et al., 2007Tjonneland A. Olsen A. Boll K. et al.Study design, exposure variables, and socioeconomic determinants of participation in Diet, Cancer and Health: a population-based prospective cohort study of 57,053 men and women in Denmark.Scand J Public Health. 2007; 35: 432-441Crossref PubMed Scopus (487) Google Scholar). At the time of study enrollment, i.e., at baseline, information on anthropometric measures was obtained by trained healthcare professionals who measured height, weight, WC, and HC by standardized procedures (Rinaldi et al., 2012Rinaldi S. Lise M. Clavel-Chapelon F. et al.Body size and risk of differentiated thyroid carcinomas: findings from the EPIC study.Int J Cancer. 2012; 131: 1004-1014Crossref PubMed Scopus (91) Google Scholar). Information on potential confounders, including skin reaction to the sun and numbers of freckles and moles on the arms, was obtained from questionnaires, whereas information on vital status and migration was obtained from linkage to the Central Population Register. MM cases diagnosed in the cohort during the study period were identified by linkage to the Danish Cancer Registry, whereas all NMSC cases were identified through linkage to an already established NMSC database (Birch-Johansen et al., 2010Birch-Johansen F. Jensen A. Mortensen L. et al.Trends in the incidence of nonmelanoma skin cancer in Denmark 1978-2007: Rapid incidence increase among young Danish women.Int J Cancer. 2010; 127: 2190-2198Crossref PubMed Scopus (171) Google Scholar), which contains all incident cases of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) diagnosed in Denmark registered in either the Danish Cancer Registry or in the Danish Registry of Pathology. Incident skin cancers from the Danish Cancer Registry were classified according to the International Classification of Diseases, 10th edition (ICD-10), codes C43 (MM) and C44 (NMSC) and to the International Classification of Diseases Oncology 3rd edition (ICD-O-3), as MM (morphology codes M-872–879), BCC (M-809), or SCC (M-807). Data on NMSC cases from the Danish Registry of Pathology were classified by Danish SNOMED classification topography codes T01 and T02 and relevant morphology codes as either BCC (M-809) or SCC (M-807). From both registries, only invasive cancers were included (morphology codes with '3' as the last digit). Each cohort member was followed from date of cohort entry until diagnosis of MM, BCC and/or SCC, date of death/emigration, or end of follow-up on 31 December 2010. In a Cox proportional hazard model, we examined associations between the various anthropometric measures and gender-specific rates of MM, BCC, and SCC. We used age as the underlying time axis to ensure that all analyses were based on comparisons of cohort members at the same age. Time under study was included as the time-dependent variable and modeled by a linear spline. Two-sided 95% confidence intervals (CIs) for estimated hazard ratios (HRs) were calculated with the Wald test of the Cox regression parameter. In all, 26,685 men and 29,243 women were eligible for analysis (Supplementary Figure S1 online). A total of 357 cohort members developed MM, 3,465 developed BCC, and 341 developed SCC within a median follow-up period of 14.4 years. Table 1 shows the distribution according to quartile of BMI of the study participants. In general, men had higher WC and WHR compared with women, whereas women had a higher HC compared with men. Download .pdf (.2 MB) Help with pdf files Supplementary MaterialTable 1Characteristics of 55,928 participants enrolled in the Danish Diet, Cancer, and Health study (1993–2007) according to BMICharacteristicQuartile of BMI (kg m-2)1234No. of participants (%)11,962 (21.4%)12,950 (23.2%)14,225 (25.4%)16,792 (30.0%) Men6,204 (23.3%)6,605 (24.8%)5,993 (22.5%)7,883 (29.5%) Women5,758 (19.7%)6,345 (21.7%)8,231 (28.2%)8,909 (30.5%)No. of participants with MM (%)76 (21.3%)81 (22.7%)93 (26.1%)107 (30.0%) Men37 (19.7%)42 (22.3%)44 (23.4%)65 (34.6%) Women39 (23.1%)39 (23.1%)49 (29.0%)42 (24.9%)No. of participants with BCC (%)861 (24.9%)896 (25.9%)898 (25.9%)810 (23.4%) Men429 (25.7%)475 (28.4%)384 (23.0%)383 (22.9%) Women432 (24.1%)421 (23.5%)514 (28.7%)427 (23.8%)No. of participants with SCC (%)85 (24.9%)90 (26.4%)78 (22.9%)88 (25.8%) Men53 (26.1%)54 (26.6%)40 (19.7%)56 (27.6%) Women32 (23.2%)36 (26.1%)38 (27.5%)32 (23.2%)Mean age at entry (years±SD) Men55.9±4.356.1±4.456.1±4.456.2±4.4 Women56.0±4.456.0±4.356.4±4.456.7±4.4Mean BMI (kg/m2±SD) Men22.4±1.325.0±0.627.0±0.630.9±2.8 Women20.5±1.223.0±0.625.4±0.930.8±3.6Mean waist circumference (cm±SD) Men85.9±5.692.2±4.996.9±5.0106.4±8.4 Women71±5.276.1±5.381.7±6.093.8±9.9Mean hip circumference (cm±SD) Men94.3±4.498.3±4.0101±4.1106.5±6.3 Women93±4.697.6±4.3102±4.7110.5±8.3Mean waist:hip ratio (ratio±SD) Men0.9±0.10.9±0.11.0±0.11.0±0.1 Women0.8±0.10.8±0.10.8±0.10.9±0.1Abbreviations: BCC, basal cell carcinoma; BMI, body mass index; MM, malignant melanoma; No., number; SCC, squamous cell carcinoma; SD, standard deviation. Open table in a new tab Abbreviations: BCC, basal cell carcinoma; BMI, body mass index; MM, malignant melanoma; No., number; SCC, squamous cell carcinoma; SD, standard deviation. No convincing associations were found between the anthropometric measures and risk for MM (Supplementary Table S1 online). Women with a BMI in the highest quartile had a decreased risk for BCC of 0.67 (95% CI: 0.54–0.82) in comparison with women with a BMI in the lowest quartile, whereas no association was found between higher BMI and risk for BCC in men. Among women, every 2 kg m-2 increase in BMI was associated with a decreased risk for BCC of 0.90 (95% CI: 0.86–0.94). Inverse linear dose–response associations were observed between increasing WC and WHR and risk for BCC in both genders and between increasing HC and risk for BCC among women but not among men (Table 2). No associations were observed between the anthropometric measures and risk for SCC among men. Among women, WC, HC, and WHR did not affect the risk for SCC, but every 2 kg m-2 increase in BMI was associated with a decreased risk of 0.80 (95% CI: 0.68–0.94) (Supplementary Table S2 online).Table 2HRs for BCC among 26,685 men and 29,243 women according to anthropometric measuresAnthropometric measureMenWomenCases/male cohortAdjusted HR (95% CI)1Adjusted for age, sun sensitivity (redness, pain, and blistering; redness, pain, and peeling; redness, then tan, or only tan), degree of freckling (none, few, moderate, or many), and number of nevi (none, few, moderate, or many). Further, BMI is adjusted for waist circumference, and waist circumference and hip circumference are mutually adjusted.Cases/female cohortAdjusted HR (95% CI)1Adjusted for age, sun sensitivity (redness, pain, and blistering; redness, pain, and peeling; redness, then tan, or only tan), degree of freckling (none, few, moderate, or many), and number of nevi (none, few, moderate, or many). Further, BMI is adjusted for waist circumference, and waist circumference and hip circumference are mutually adjusted.1,671/26,6851,794/29,243BMI2BMI quartiles, men: 1:?24?kg?m-2, 2:>24–? 26?kg?m-2, 3:>26–? 28?kg?m-2, 4:>28?kg?m-2; women: 1:?22?kg?m-2, 2:>22–? 24?kg?m-2, 3:>24–? 27?kg?m-2, 4:>27?kg?m-2. 1st quartile429/5,7751.00432/5,3261.00 2nd quartile475/6,1301.07 (0.93–1.23)421/5,9240.88 (0.76–1.01) 3rd quartile384/5,6091.00 (0.85–1.18)514/7,7170.83 (0.72–0.97) 4th quartile383/7,5000.85 (0.69–1.05)427/8,4820.67 (0.54–0.82) Per 2 kg m-20.96 (0.90–1.01)0.90 (0.86–0.94)Waist circumference3Waist circumference quartiles, men: 1:?89?cm, 2:>89–?95?cm, 3:>95–?101?cm, 4:>101?cm; women: 1:?73?cm, 2:>73–? 79?cm, 3:>79–? 86?cm, 4:>86?cm. 1st quartile465/6,3481.00497/6,2341.00 2nd quartile488/6,4141.05 (0.92–1.21)445/6,5740.90 (0.79–1.03) 3rd quartile382/5,7220.97 (0.83–1.13)397/6,3770.86 (0.74–0.99) 4th quartile336/6,5300.83 (0.68–1.01)455/8,2640.85 (0.72–1.01) Per 5 cm0.94 (0.90–0.98)0.96 (0.93–0.99)Hip circumference4Hip circumference quartiles, men: 1:?96?cm, 2:>96–? 100?cm, 3:>100–?104?cm, >104?cm; women: 1:?95?cm, 2:>95–?100?cm, 3: >100–?105?cm, 4: >105?cm. 1st quartile496/6,8431.00487/6,3701.00 2nd quartile454/6,4930.99 (0.87–1.14)482/6,9340.95 (0.83–1.08) 3rd quartile401/5,6881.08 (0.92–1.25)402/6,2700.92 (0.80–1.06) 4th quartile320/5,9900.94 (0.77–1.14)423/7,8750.86 (0.72–1.02) Per 5 cm0.98 (0.92–1.04)0.96 (0.92–1.00)Waist:hip ratio5Waist:hip ratio quartiles, men: 1:?0.91, 2:>0.91–? 0.95, 3:>0.95–? 0.99, 4:>0.99; women: 1:?0.75, 2:>0.75–? 0.79, 3:>0.79–?0.85, 4:>0.85. 1st quartile468/6,3071.00494/6,8091.00 2nd quartile467/6,5950.96 (0.85–1.09)476/6,3411.04 (0.91–1.18) 3rd quartile415/6,0640.94 (0.82–1.07)447/8,0530.78 (0.69–0.89) 4th quartile321/6,0480.78 (0.68–0.91)377/6,2460.88 (0.77–1.01) Per 0.05 unit0.93 (0.89–0.97)0.94 (0.91–0.98)Abbreviations: BCC, basal cell carcinoma; BMI, body mass index; CI, confidence interval; HR, hazard ratio.1 Adjusted for age, sun sensitivity (redness, pain, and blistering; redness, pain, and peeling; redness, then tan, or only tan), degree of freckling (none, few, moderate, or many), and number of nevi (none, few, moderate, or many). Further, BMI is adjusted for waist circumference, and waist circumference and hip circumference are mutually adjusted.2 BMI quartiles, men: 1:?24?kg?m-2, 2:>24–? 26?kg?m-2, 3:>26–? 28?kg?m-2, 4:>28?kg?m-2; women: 1:?22?kg?m-2, 2:>22–? 24?kg?m-2, 3:>24–? 27?kg?m-2, 4:>27?kg?m-2.3 Waist circumference quartiles, men: 1:?89?cm, 2:>89–?95?cm, 3:>95–?101?cm, 4:>101?cm; women: 1:?73?cm, 2:>73–? 79?cm, 3:>79–? 86?cm, 4:>86?cm.4 Hip circumference quartiles, men: 1:?96?cm, 2:>96–? 100?cm, 3:>100–?104?cm, >104?cm; women: 1:?95?cm, 2:>95–?100?cm, 3: >100–?105?cm, 4: >105?cm.5 Waist:hip ratio quartiles, men: 1:?0.91, 2:>0.91–? 0.95, 3:>0.95–? 0.99, 4:>0.99; women: 1:?0.75, 2:>0.75–? 0.79, 3:>0.79–?0.85, 4:>0.85. Open table in a new tab Abbreviations: BCC, basal cell carcinoma; BMI, body mass index; CI, confidence interval; HR, hazard ratio. Our null findings for MM correspond with results from a meta-analysis by Olsen et al., 2008Olsen C.M. Green A.C. Zens M.S. et al.Anthropometric factors and risk of melanoma in women: a pooled analysis.Int J Cancer. 2008; 122: 1100-1108Crossref PubMed Scopus (41) Google Scholar. Results from two other meta-analyses (Renehan et al., 2012; Sergentanis et al., 2013Sergentanis T.N. Antoniadis A.G. Gogas H.J. et al.Obesity and risk of malignant melanoma: a meta-analysis of cohort and case-control studies.Eur J Cancer. 2013; 49: 642-657Abstract Full Text Full Text PDF PubMed Scopus (114) Google Scholar), observed an increased risk for MM among men with higher BMI but no association among women. In line with our results, most previous studies also observed an inverse association between BMI and risk for BCC. In one of the largest cohort studies to date, Pothiawala et al., 2012Pothiawala S. Qureshi A.A. Li Y. et al.Obesity and the incidence of skin cancer in US Caucasians.Cancer Causes Control. 2012; 23: 717-726Crossref PubMed Scopus (61) Google Scholar observed that obese participants with a BMI >30 kg m-2 had a 19% lower risk for BCC compared with participants with a BMI <25 kg m-2. The only previous study of the association between measures of obesity other than BMI and risk for BCC found no convincing associations between WC and WHR and risk for BCC (Olsen et al., 2006Olsen C.M. Hughes M.C. Pandeya N. et al.Anthropometric measures in relation to basal cell carcinoma: a longitudinal study.BMC Cancer. 2006; 6: 82Crossref PubMed Scopus (20) Google Scholar). Two cohort studies have investigated the association between BMI and SCC: In agreement with our study, Pothiawala et al., 2012Pothiawala S. Qureshi A.A. Li Y. et al.Obesity and the incidence of skin cancer in US Caucasians.Cancer Causes Control. 2012; 23: 717-726Crossref PubMed Scopus (61) Google Scholar found that increasing BMI was associated with a decreasing risk for SCC among women but not among men, whereas Nagel et al., 2012Nagel G. Bjorge T. Stocks T. et al.Metabolic risk factors and skin cancer in the Metabolic Syndrome and Cancer Project (Me-Can).Br J Dermatol. 2012; 167: 59-67Crossref PubMed Scopus (32) Google Scholar found no association with BMI in either gender. Hence, despite the fact that several biological mechanisms link obesity to cancer risk (Calle and Kaaks, 2004Calle E.E. Kaaks R. Overweight, obesity and cancer: epidemiological evidence and proposed mechanisms.Nat Rev Cancer. 2004; 4: 579-591Crossref PubMed Scopus (2706) Google Scholar), our results, which are in accordance with the majority of other studies, indicate an inverse association between obesity and risk for NMSC even after adjustment for potentially confounding factors related to UV radiation (UVR) susceptibility. However, the main limitation of our study was that we were unable to adjust for the direct UVR exposure acquired by each individual as we had no information of sun seeking behavior or use of tanning beds. As it has been suggested that obese people have a different UVR seeking behavior compared with lean people (e.g., spend less time outdoors and avoid sunbathing) (Gerstenblith et al., 2012Gerstenblith M.R. Rajaraman P. Khaykin E. et al.Basal cell carcinoma and anthropometric factors in the U.S. radiologic technologists cohort study.Int J Cancer. 2012; 131: 149-155Crossref PubMed Scopus (28) Google Scholar), the inverse associations observed between obesity and risk for NMSC may therefore at least partly be influenced by unmeasured confounding by UVR exposure. Several strengths of this study are noteworthy. First, the vast majority of earlier studies only included BMI, whereas we also addressed the association between other anthropometric measures and risk for skin cancer. However, as we found no conspicuous patterns of associations between the different anthropometric measures and risk for skin cancer, we were unable to draw firm conclusions about whether general, abdominal, or gluteofemoral obesity is the best measure for analyzing potential associations with skin cancer. Another strength was that the anthropometric measures were obtained prospectively by healthcare professionals using standardized methods, thus minimizing the possibility of differential recall bias. The study population was randomly selected from the background population, and there was virtually no loss to follow-up because of the precise linkage between the study cohort and various Danish health registries. In conclusion, the results of this large population-based cohort study showed no convincing association between obesity and risk for MM, whereas inverse associations were found between various anthropometric measures and risk for BCC in both genders and for SCC among women only. Additional studies with more accurate data on exposure to UVR are needed to investigate the role of obesity in skin cancer pathogenesis. This research was supported by the Danish Cancer Society, which was not involved in the study design or data collection, analyses, interpretation of the results, the decision to submit the manuscript for publication, or writing the manuscript. Supplementary material is linked to the online version of the paper at http://www.nature.com/jid
Referência(s)