Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Physical activity, cardiorespiratory fitness and risk of cutaneous malignant melanoma: Systematic review and meta-analysis

  • Gundula Behrens ,

    Contributed equally to this work with: Gundula Behrens, Tobias Niedermaier

    Roles Conceptualization, Formal analysis, Investigation, Methodology, Supervision, Visualization, Writing – original draft, Writing – review & editing

    Gundula.Behrens@klinik.uni-regensburg.de

    Affiliation Department of Epidemiology and Preventive Medicine, University of Regensburg, Regensburg, Germany

  • Tobias Niedermaier ,

    Contributed equally to this work with: Gundula Behrens, Tobias Niedermaier

    Roles Conceptualization, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Department of Epidemiology and Preventive Medicine, University of Regensburg, Regensburg, Germany

  • Mark Berneburg,

    Roles Formal analysis, Investigation, Methodology, Writing – review & editing

    Affiliation Department of Dermatology, University Hospital Regensburg, Regensburg, Germany

  • Daniela Schmid,

    Roles Formal analysis, Investigation, Methodology, Writing – review & editing

    Affiliation Department of Epidemiology and Preventive Medicine, University of Regensburg, Regensburg, Germany

  • Michael F. Leitzmann

    Roles Conceptualization, Formal analysis, Investigation, Methodology, Supervision, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Department of Epidemiology and Preventive Medicine, University of Regensburg, Regensburg, Germany

Physical activity, cardiorespiratory fitness and risk of cutaneous malignant melanoma: Systematic review and meta-analysis

  • Gundula Behrens, 
  • Tobias Niedermaier, 
  • Mark Berneburg, 
  • Daniela Schmid, 
  • Michael F. Leitzmann
PLOS
x

Abstract

Background

Numerous epidemiologic studies have examined the relation of physical activity or cardiorespiratory fitness to risk of cutaneous melanoma but the available evidence has not yet been quantified in a systematic review and meta-analysis.

Methods

Following the preferred reporting items for systematic reviews and meta-analyses (PRISMA), we identified 3 cohort studies (N = 12,605 cases) and 5 case-control studies (N = 1,295 cases) of physical activity and melanoma incidence, and one cohort study (N = 49 cases) of cardiorespiratory fitness and melanoma risk.

Results

Cohort studies revealed a statistically significant positive association between high versus low physical activity and melanoma risk (RR = 1.27, 95% CI = 1.16–1.40). In contrast, case-control studies yielded a statistically non-significant inverse risk estimate for physical activity and melanoma (RR = 0.85, 95% CI = 0.63–1.14; P-difference = 0.02). The only available cohort study of cardiorespiratory fitness and melanoma risk reported a positive but statistically not significant association between the two (RR = 2.19, 95% CI = 0.99–4.96). Potential confounding by ultraviolet (UV) radiation-related risk factors was a major concern in cohort but not case-control studies.

Conclusions

It appears plausible that the positive relation of physical activity and cardiorespiratory fitness to melanoma observed in cohort studies is due to residual confounding by UV radiation-related risk factors.

Impact

Future prospective studies need to examine the association between physical activity, cardiorespiratory fitness and melanoma after detailed adjustment for UV radiation-related skin damage.

Introduction

Worldwide, marked annual increases in incidence rates of invasive cutaneous melanoma have been observed across several decades, in particular among populations with low skin pigmentation [14]. The most likely reason for the positive long-term trend in melanoma incidence is increased UV (ultraviolet) radiation-related skin damage caused by prolonged sun exposure [5].

UV radiation-related DNA damage is the most relevant modifiable melanoma risk factor, and oxidative stress, chronic inflammation, and impaired immune function all represent contributing factors [69]. In contrast, surprisingly little is known about the independent effects of physical activity and cardiorespiratory fitness on melanoma prevention, although physical exercise has a positive impact on numerous biological pathways, such as increased DNA repair capacity, decreased levels of oxidative stress, reduced inflammation, and enhanced immune function [1013].

Available evidence from epidemiologic studies of physical activity, cardiorespiratory fitness and melanoma risk has been inconsistent, which may be due to differences in study design and incomplete adjustment for potential confounding variables. We therefore conducted a systematic review and meta-analysis of physical activity, cardiorespiratory fitness and melanoma with careful consideration of the influence of study design and adjustment for potential confounding factors, including UV radiation-related skin damage, sun exposure, and sun sensitivity.

Methods

Literature search

Following the preferred reporting items for systematic reviews and meta-analyses (PRISMA) [14], two researchers (GB, TN) conducted independent systematic literature searches in PubMed and Web of Knowledge to identify English language epidemiologic studies which assessed physical activity or cardiorespiratory fitness on an individual level and examined the relations of physical activity or cardiorespiratory fitness to newly incident cutaneous melanoma (S1 and S2 Files).

Our systematic literature search in PubMed and Web of Knowledge was performed from inception of the databases to March 29, 2018, and it yielded 2,039 hits. Exclusion and inclusion criteria are shown in Fig 1. We removed 1,957 articles after screening titles and abstracts. After reading the manuscripts of the remaining articles, we excluded two cohort studies [15, 16] that did not assess physical activity on an individual level, two case-control studies [17, 18] that used cancer patients as controls, and 32 articles that were irrelevant. We included 1 cohort study of cardiorespiratory fitness and melanoma [19], 3 cohort studies of physical activity and melanoma [2022] and 5 case-control studies of physical activity and melanoma [2327] in our systematic review. We did not identify any other relevant studies by searching the reference lists of relevant articles. One [20] of the 3 included cohort studies of physical activity and melanoma combined information from 12 individual cohorts in a pooled analysis. For the main analysis, we used the pooled risk estimate from that study [20] because it was adjusted for adiposity. In sensitivity analyses and in sub-analyses stratified by gender and geographic region, we used adiposity-unadjusted estimates from that study [20] because published adiposity-adjusted estimates were not available.

thumbnail
Fig 1. Flowchart of literature search results, last conducted on March 29, 2018.

https://doi.org/10.1371/journal.pone.0206087.g001

Because there was only one investigation of cardiorespiratory fitness and melanoma [19], we were unable to perform a meta-analysis of cardiorespiratory fitness and melanoma risk. We did not combine studies of physical activity and cardiorespiratory fitness because physical activity and cardiorespiratory fitness represent independent risk factors for chronic diseases [2830].

Extraction of data

Two researchers (GB, TN) conducted independent data extractions from the identified studies. Any disagreement was resolved by discussion. We included relative risks (RR) with 95% confidence intervals (CI) for all available physical activity domains and for any timing in life of physical activity. We used the most comprehensively adjusted risk estimates provided that those risk estimates were also adjusted for age and, if applicable, gender. If available, we extracted separate risk estimates for men and women. In a sensitivity analysis, we included only one risk estimate per study and gender. For that sensitivity analysis, we preferred recreational physical activity and recent past physical activity over all other domains and timings in life of physical activity.

Classification of timing in life of physical activity

We classified levels of physical activity as recent past physical activity if they referred to a time period less than 10 years prior to the baseline assessment (cohort studies) or cancer diagnosis (case-control studies). If those physical activity levels referred to a period further back in time, the physical activity variable was classified as distant past physical activity. An exception was made if that variable captured more than one time period in the past and if the utmost time periods were ≥10 years apart. In that case, the physical activity variable was classified as a measure of consistent physical activity over time. We considered occupational physical activity to fall into the latter category because people usually hold occupations for more than one decade.

Study quality assessment and adjustments for UV radiation-related risk factors

We used the Monninkhof score [31] that was specifically developed for studies of physical activity and cancer incidence to assess the methodologic quality of the included studies, in particular the potential for selection bias (N = 5 items), misclassification (N = 11 items), and confounding (N = 3 items), which were weighted using a proportion of 2:2:1 with a maximum attainable score of 105 points. In one item (9 points), we chose adjustment for sun exposure and sun sensitivity as proxies for UV radiation-related skin damage [32, 33] on an individual level to assess the degree of control for major potential confounding factors. Studies adjusting for sun exposure on an individual level considered sun exposure during holidays 20 years prior to the interview [24], total lifetime sun exposure [27], recreational lifetime sun exposure [25] and occupational lifetime sun exposure [25]. Studies adjusting for sun sensitivity used one or more of the following variables: skin type [2325], hair color [23, 25, 27], eye color [23], and immediate skin reaction to <30 minutes of UV radiation exposure at the beginning of the outdoor season [24]. One case-control study [24] additionally adjusted for UV radiation-related skin damage by including history of sunburns in childhood, actinic cheilitis, actinic keratosis, and solar lentigo in the model.

Statistical methods

We interpreted all hazard ratios, incidence rate ratios, and odds ratios as relative risks. We summarized melanoma risk estimates comparing the highest and lowest physical activity categories on the log-scale in precision-weighted random effects meta-regression models [34] using the restricted maximum likelihood method (REML). We used funnel plots, Begg’s and Egger’s tests, and the Q- and I2-statistics to assess potential publication bias and between-study heterogeneity [3537]. All statistical tests were two-sided and they were based on a significance level of 5%. All analyses were conducted in R, version 3.3.2, using the metafor-package [38, 39].

Main analysis, stratified analyses, and sensitivity analyses

In the main analysis, we combined all available risk estimates of the association between physical activity and melanoma incidence, separately for cohort and case-control studies. In stratified analyses, we investigated the influence of the study quality score, gender, study geographic region, physical activity domain, timing in life of physical activity, physical activity measure, adjustments for UV radiation-related sun damage, sun sensitivity, sun exposure (on an individual level), smoking, adiposity, alcohol intake, and history of type 2 diabetes. In addition, we ran analyses stratified by adjustment for both sun sensitivity and sun exposure. To examine if any of the stratification variables explained the difference by study design, we repeated the stratified analyses for cohort and case-control studies combined. We compared summary estimates using likelihood ratio tests. In a sensitivity analysis, we included only one risk estimate per study and gender in the meta-analysis.

Results

Study characteristics

We considered 3 cohort studies [2022] (N = 12,605 cases, of which the pooled study [20] contributed 12,438 cases, Table 1) and 5 case-control studies [2327] (N = 1,295 cases, Table 1) of physical activity and melanoma risk, and 1 cohort study [19] (N = 49 cases, Table 2) of cardiorespiratory fitness and melanoma risk.

thumbnail
Table 1. Characteristics of the 3 cohort studies (N = 12,605 cases) and 5 case-control studies (N = 1,295 cases) on physical activity and melanoma risk included in the systematic review and meta-analysis grouped by study design.

https://doi.org/10.1371/journal.pone.0206087.t001

thumbnail
Table 2. Characteristics of the one cohort study (N = 49 cases) on cardiorespiratory fitness and melanoma risk included in the systematic review.

https://doi.org/10.1371/journal.pone.0206087.t002

The one melanoma risk estimate [19] for cardiorespiratory fitness was based on middle-aged men residing in Norway and it was positive but not statistically significant (RR = 2.19, 95% CI = 0.99–4.96). It was adjusted for smoking and adiposity but not for UV-radiation related risk factors.

Cohort and case-control studies of physical activity and melanoma risk differed with respect to their gender distribution. Specifically, case-control studies reported separate risk estimates for men [2527] and women [25, 27], whereas most cohorts examined melanoma risk in women [20] or in men and women combined [2022]. Cohort and case-control studies had in common that there were as many studies from Europe [20, 2224] as there were from North America [20, 21, 2527].

Most cohort [2022] and case-control studies [23, 24, 26, 27] of physical activity and melanoma risk examined recreational physical activity. Studies used various physical activity assessments including physical activity frequency [26, 27], physical activity duration [21, 23], energy expenditure [20, 25], and qualitative ratings [22, 24, 26].

None of the included cohort studies and four [2325, 27] of the five included case-control studies of physical activity and melanoma risk adjusted for sun sensitivity or sun exposure on an individual level. However, two cohort studies used geographic region [22] and ground-level solar UV radiation [20] (erythemal dose; for a sub-cohort) to assess sun exposure on an ecologic level. The latter study [20] observed a statistically significant positive association between physical activity and melanoma incidence in geographic areas of high UV exposure and a statistically non-significant positive association between physical activity and melanoma incidence in geographic areas of low UV exposure. The former study [22] was only conducted in geographic areas of low UV exposure and it also observed a statistically non-significant positive association between physical activity and melanoma incidence.

One third of the cohort and case-control studies of physical activity and melanoma risk adjusted for smoking and adiposity [20, 23, 26], and one case-control study adjusted for type 2 diabetes [23].

Main analysis

We observed a statistically significant difference between the summary melanoma risk estimates for physical activity from cohort studies and case-control studies (P-difference = 0.02, Fig 2). Cohort studies indicated a statistically significant risk increase of 27% for melanoma when comparing physically active participants with physically inactive participants (RR = 1.27, 95% CI = 1.16–1.40), while case-control studies showed a statistically non-significant risk reduction of 15% for melanoma for the same comparison (RR = 0.85, 95% CI = 0.63–1.14). We found no indication of publication bias among cohort and case-control studies (cohort studies: P-value for Begg’s test = 0.99; P-value for Egger’s test = 0.79, Fig 3A; case-control studies: P-value for Begg’s test = 0.48; P-value for Egger’s test = 0.07, Fig 3B). We observed between-study heterogeneity among case-control studies but not among cohort studies (cohort studies: I2 = 0%, P-heterogeneity = 0.92; case-control studies: I2 = 43%, P-heterogeneity = 0.03).

thumbnail
Fig 2. Forest plot of a random effects meta-analysis including 13 risk estimates of melanoma for a high versus low level of physical activity, grouped by study design; among cohorts: I2 = 0%, P-heterogeneity = 0.92; among case-control studies: I2 = 43%, P-heterogeneity = 0.03; P-difference by study design = 0.02.

Abbreviations: RR, relative risk; CI, confidence interval; RPA, recreational physical activity, OPA, occupational physical activity; TPA, total physical activity.

https://doi.org/10.1371/journal.pone.0206087.g002

thumbnail
Fig 3.

A) Funnel plot for random effects meta-analysis including 4 risk estimates of melanoma for a high versus low level of physical activity among cohort studies: P-value for Begg’s test = 0.99; P-value for Egger’s test = 0.79. B) Funnel plot for random effects meta-analysis including 9 risk estimates of melanoma for a high versus low level of physical activity among case-control studies: P-value for Begg’s test = 0.48; P-value for Egger’s test = 0.07.

https://doi.org/10.1371/journal.pone.0206087.g003

Stratified analyses and sensitivity analyses

We observed a stronger positive association between physical activity and melanoma in studies from Europe than studies from North America among cohort studies but not among case-control studies (Table 3). In contrast, study quality and timing in life of physical activity influenced the risk estimates for physical activity and melanoma risk among case-control studies but not among cohort studies. Case-control studies with a higher study quality score reported a null association between physical activity and melanoma risk. In contrast, case-control studies with a lower study quality score revealed a statistically significant inverse association between the two.

thumbnail
Table 3. Random effects summary estimates of melanoma risk for a high versus low level of physical activity by selected participant and design characteristics, evaluated separately for cohort and case-control studies and, in sensitivity analyses, for cohort and case-control studies combined.

https://doi.org/10.1371/journal.pone.0206087.t003

Because case-control studies of lifetime physical activity and melanoma risk scored higher in the methodologic quality assessment than did case-control studies of recent or distant past physical activity, adjustment for timing in life of physical activity removed the statistically significant influence of study quality (P-difference across quality score before and after adjustment for timing in life of physical activity among case-control studies = 0.03 and 0.30, respectively). Similarly, the between-study heterogeneity among case-control studies was no longer apparent after adjusting for timing in life of physical activity (I2 = 0%; P-heterogeneity = 0.17). Case-control studies of recent and distant past physical activity observed statistically significant and non-significant inverse relations between physical activity and melanoma, respectively, whereas case-control studies of lifetime physical activity revealed a statistically non-significant positive association between the two (P-difference = 0.02, Table 3). In contrast, cohort studies found a statistically significant positive relation between recent past physical activity and melanoma risk and no statistically significant variation in risk estimates across timing in life of physical activity (Table 3).

When we repeated the stratified analyses among cohort and case-control studies combined in a sensitivity analysis, we found that the summary risk estimate that was based on energy expenditure as a measure of physical activity was positive, whereas summary risk estimates that were based on physical activity frequency, physical activity duration, or qualitative ratings were inverse (P-difference = 0.001, Table 3). The energy expenditure-based risk estimate was driven by the pooled cohort study [20].

In an additional sensitivity analysis, we allowed only one risk estimate per study and gender, which yielded comparable summary risk estimates to those from the main analysis (RR among cohort studies = 1.28, 95% CI = 1.16–1.40; RR among case-control studies = 0.91, 95% CI = 0.65–1.26). Similarly, using all 12 individual risk estimates rather than the pooled risk estimate from the pooled cohort study [20] did not materially alter the results (S1 Table, S1 Fig).

Discussion

The primary new finding of the current meta-analysis is that the association between physical activity and melanoma differs by study design. Specifically, we found a statistically significant positive association between physical activity and melanoma incidence among cohort studies, whereas we observed a statistically non-significant inverse relation among case-control studies. The one cohort study of cardiorespiratory fitness and melanoma reported a positive but statistically non-significant association between the two.

The most important risk factor for melanoma is UV radiation-related skin damage, which includes a history of sunburns, dysplastic and common naevi, actinic keratosis, and solar lentigo [4043]. Additional melanoma risk factors are those that enhance susceptibility to UV radiation-related skin damage, such as skin phototype, hair color, eye color, freckles, the ability to tan, and the skin’s immediate reaction to sun exposure at the beginning of the tanning season [42].

The presence of melanoma at sites that are rarely exposed to UV radiation and the relations of naevus patterns (site and count) to the histology and location of melanoma suggest that there are diverging pathways of melanoma incidence [4446]. One recent study [47] reported that the observed positive relation of bone mineral density to naevus count (and thereby to melanoma incidence) could be partly explained by leucocyte telomere length. However, it remains unclear whether that association is causal because that study did not adjust for sun exposure as a potential confounding factor and because leucocyte telomere length has been inversely related to sun exposure, naevus count, and melanoma incidence [47, 48]. Similarly, it is uncertain whether physical activity affects leucocyte telomere length and naevus count [4951]. A large cross-sectional study reported a positive association between physical activity and naevus count among 26,000 men but not among 67,000 women in age-adjusted analyses [51]. That observed gender difference could be due to confounding by sun exposure because women are less likely to participate in outdoor physical activities than men and if they do, they have lower vitamin D serum levels than men in each outdoor physical activity category [52, 53]. Carefully designed future studies are required to examine if physical activity affects leucocyte telomere length, naevus count, and melanoma incidence independently of sun exposure.

Moreover, various genetic factors have been related to the incidence of melanoma, including genes related to pigmentation and melanin production, the development of naevi, DNA repair, and family history of melanoma [8, 54, 55]. Immunosuppressive therapy or immunosuppressive UV radiation-related cell damage also increase the risk of melanoma [7]. In addition, adiposity, chronic inflammation, and oxidative stress have all been positively related to melanoma incidence [6, 8, 9, 56].

Mechanistically, physical activity and cardiorespiratory fitness may prevent melanoma through enhanced immune function, increased DNA repair capacity, reduced oxidative stress, decreased chronic inflammation, and weight control [1013, 5761]. Thus, a potential preventive role of physical activity and cardiorespiratory fitness for the development of melanoma is biologically plausible. However, it may be challenging for epidemiologic studies to detect a potential protective effect of physical activity and cardiorespiratory fitness without comprehensive adjustment for UV radiation-related skin damage. In fact, incomplete adjustment for UV radiation-related skin damage may residually confound the relations of physical activity and cardiorespiratory fitness to melanoma because physical activity and cardiorespiratory fitness are positively associated with time spent outdoors, sunburns, and UV radiation-related skin damage [6265].

Because none of the three cohort studies of physical activity and melanoma included in the present meta-analysis were adjusted for UV radiation-related skin damage, sun sensitivity, or sun exposure (on an individual level), the statistically significant positive summary risk estimate for physical activity and melanoma risk obtained from cohort studies may be a result of major confounding by UV radiation-related risk factors. The greater magnitude of the risk estimate among studies from Europe as compared to that from studies from North America indicates a greater degree of confounding of the physical activity and melanoma relation in European than North American populations, potentially due to lower levels of skin pigmentation [66] and greater levels of outdoor physical activity [67] among Europeans than North Americans. In line with this, one pooled prospective analysis [20] reported that the positive association between physical activity and melanoma was more pronounced in geographic areas of high UV radiation exposure than in areas of low UV radiation exposure. Interestingly, one case-control study [24] that was able to adjust for UV radiation-related skin damage observed a strong but statistically non-significant inverse association between melanoma and physical activity in childhood, a time period during which sunburns pose one of the strongest risk factors for the development of subsequent melanoma [68].

Four of the five included case-control studies attempted to minimize the potential for UV radiation-related confounding by detailed adjustment for UV radiation-related skin damage, sun sensitivity (skin phototype, hair color, eye color, and immediate skin reaction to UV radiation), and sun exposure (on an individual level). Thus, it is conceivable that the statistically non-significant inverse summary risk estimate for physical activity and melanoma risk obtained from the included case-control studies is closer to the true association between physical activity and melanoma than that obtained from the cohort studies.

Because the included studies used a large variety of physical activity assessments and because there was no single predominant type of physical activity assessment, in our main analysis we summarized risk estimates across all types of physical activity assessments to obtain a summary risk estimate of melanoma incidence for an “average” physical activity assessment. As a consequence, we included more than one risk estimate per gender in our main analysis for one [22] of the three cohort studies and for one [26] of the five case-control studies. We therefore verified that results from the main analysis combining all types of physical activity assessments were comparable to those from a rigorous sensitivity analysis allowing only one risk estimate per study and gender. In addition, we stratified analyses by physical activity domain, timing in life of physical activity, and type of physical activity measure but we did not find any statistically significant influence of those variables on the association between physical activity and melanoma incidence with one exception: among case-control studies, the risk estimate for consistent physical activity over time tended to be positive, while those for recent past physical activity and distant past physical activity tended to be inverse. This observation may be explained by the potential of residual confounding through incomplete adjustment for UV radiation-related risk factors in some of the case-control studies of consistent physical activity over time.

The major strength of the present systematic review and meta-analysis of physical activity and melanoma risk is the large number of cases and the inclusion of both cohort and case-control studies. Considering both study types was essential because case-control studies were methodologically superior to cohort studies in their degree of adjustment for UV radiation-related risk factors. Although case-control studies are generally considered to be prone to recall and selection biases, case-control studies of melanoma risk factors have not been subject to notable recall bias [6974]. Also, the included case-control studies attempted to reduce the potential for selection bias by choosing population-based controls [23, 2527] or a wide variety of hospital-based controls [24] and by adjusting for variables predictive of participants’ response rates, including age, sex, education, and socio-economic status [2327, 7577]. We did not identify any publication bias. However, due to the limited number of studies available for meta-analysis, the power of the statistical tests for publication bias was low.

In summary, we observed significant heterogeneity between summary risk estimates from cohort and case-control studies of physical activity and melanoma incidence. It is likely that the statistically significant positive association between physical activity and melanoma risk and the positive but statistically not significant association between cardiorespiratory fitness and melanoma observed in cohort studies is the result of incomplete adjustment for UV radiation-related skin damage. Case-control studies of physical activity and melanoma generally adjusted for UV radiation-related risk factors and produced a statistically non-significant inverse relation between physical activity and melanoma incidence. To clarify the true nature of the relations of physical activity and cardiorespiratory fitness to melanoma incidence, future prospective studies are required to examine physical activity and cardiorespiratory fitness in relation to melanoma risk after careful adjustment for UV radiation-related skin damage.

Supporting information

S1 Table. Random effects summary estimates of melanoma risk for a high versus low level of physical activity by selected participant and design characteristics using the 12 individual risk estimates rather than the pooled risk estimate from Moore et al. 2016 [20] as sensitivity analysis to Table 3.

https://doi.org/10.1371/journal.pone.0206087.s003

(DOCX)

S1 Fig. Funnel plot for random effects meta-analysis including 15 risk estimates of melanoma for a high versus low level of physical activity among cohort studies in the sensitivity analysis including the 12 individual risk estimates rather than the pooled risk estimate from Moore et al. 2016 [20] as sensitivity analysis to Fig 3B: P-value for Begg’s test = 0.77; P-value for Egger’s test = 0.70.

https://doi.org/10.1371/journal.pone.0206087.s004

(EPS)

References

  1. 1. Erdmann F, Lortet-Tieulent J, Schuz J, Zeeb H, Greinert R, Breitbart EW, et al. International trends in the incidence of malignant melanoma 1953-2008—are recent generations at higher or lower risk? International journal of cancer Journal international du cancer. 2013;132(2):385–400. pmid:22532371.
  2. 2. Geller AC, Clapp RW, Sober AJ, Gonsalves L, Mueller L, Christiansen CL, et al. Melanoma epidemic: an analysis of six decades of data from the Connecticut Tumor Registry. Journal of clinical oncology: official journal of the American Society of Clinical Oncology. 2013;31(33):4172–8. pmid:24043747; PubMed Central PMCID: PMC3906570.
  3. 3. Baade P, Meng X, Youlden D, Aitken J, Youl P. Time trends and latitudinal differences in melanoma thickness distribution in Australia, 1990–2006. International journal of cancer Journal international du cancer. 2012;130(1):170–8. pmid:21344376.
  4. 4. van der Leest RJ, Zoutendijk J, Nijsten T, Mooi WJ, van der Rhee JI, de Vries E, et al. Increasing time trends of thin melanomas in The Netherlands: What are the explanations of recent accelerations? European journal of cancer. 2015;51(18):2833–41. pmid:26589973.
  5. 5. Kojo K, Jansen CT, Nybom P, Huurto L, Laihia J, Ilus T, et al. Population exposure to ultraviolet radiation in Finland 1920–1995: Exposure trends and a time-series analysis of exposure and cutaneous melanoma incidence. Environmental research. 2006;101(1):123–31. pmid:16290819.
  6. 6. Rawson RV, Johansson PA, Hayward NK, Waddell N, Patch AM, Lo S, et al. Unexpected UVR and non-UVR mutation burden in some acral and cutaneous melanomas. Laboratory investigation; a journal of technical methods and pathology. 2017;97(2):130–45. pmid:28067894.
  7. 7. Kubica AW, Brewer JD. Melanoma in immunosuppressed patients. Mayo Clinic proceedings. 2012;87(10):991–1003. pmid:23036673; PubMed Central PMCID: PMC3538393.
  8. 8. Kyrgidis A, Tzellos TG, Triaridis S. Melanoma: Stem cells, sun exposure and hallmarks for carcinogenesis, molecular concepts and future clinical implications. Journal of carcinogenesis. 2010;9:3. pmid:20442802; PubMed Central PMCID: PMC2862505.
  9. 9. Halliday GM. Inflammation, gene mutation and photoimmunosuppression in response to UVR-induced oxidative damage contributes to photocarcinogenesis. Mutation research. 2005;571(1–2):107–20. pmid:15748642.
  10. 10. Fedewa MV, Hathaway ED, Ward-Ritacco CL. Effect of exercise training on C reactive protein: a systematic review and meta-analysis of randomised and non-randomised controlled trials. British journal of sports medicine. 2017;51(8):670–6. pmid:27445361.
  11. 11. Cao Dinh H, Beyer I, Mets T, Onyema OO, Njemini R, Renmans W, et al. Effects of Physical Exercise on Markers of Cellular Immunosenescence: A Systematic Review. Calcified tissue international. 2017;100(2):193–215. pmid:27866236.
  12. 12. de Sousa CV, Sales MM, Rosa TS, Lewis JE, de Andrade RV, Simoes HG. The Antioxidant Effect of Exercise: A Systematic Review and Meta-Analysis. Sports medicine. 2017;47(2):277–93. pmid:27260682.
  13. 13. Cash SW, Beresford SA, Vaughan TL, Heagerty PJ, Bernstein L, White E, et al. Recent physical activity in relation to DNA damage and repair using the comet assay. Journal of physical activity & health. 2014;11(4):770–6. pmid:25078521; PubMed Central PMCID: PMC3844056.
  14. 14. Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Bmj. 2009;339:b2535. pmid:19622551; PubMed Central PMCID: PMC2714657.
  15. 15. Sormunen J, Backmand HM, Sarna S, Kujala UM, Kaprio J, Dyba T, et al. Lifetime physical activity and cancer incidence—a cohort study of male former elite athletes in Finland. Journal of science and medicine in sport / Sports Medicine Australia. 2014;17(5):479–84. pmid:24239090.
  16. 16. Pukkala E, Kaprio J, Koskenvuo M, Kujala U, Sarna S. Cancer incidence among Finnish world class male athletes. International journal of sports medicine. 2000;21(3):216–20. pmid:10834356.
  17. 17. Brownson RC, Chang JC, Davis JR, Smith CA. Physical activity on the job and cancer in Missouri. American journal of public health. 1991;81(5):639–42. pmid:2014869; PubMed Central PMCID: PMC1405078.
  18. 18. Dosemeci M, Hayes RB, Vetter R, Hoover RN, Tucker M, Engin K, et al. Occupational physical activity, socioeconomic status, and risks of 15 cancer sites in Turkey. Cancer causes & control: CCC. 1993;4(4):313–21. pmid:8347780.
  19. 19. Robsahm TE, Falk RS, Heir T, Sandvik L, Vos L, Erikssen J, et al. Cardiorespiratory fitness and risk of site-specific cancers: a long-term prospective cohort study. Cancer medicine. 2017;6(4):865–73. pmid:28317282; PubMed Central PMCID: PMC5387170.
  20. 20. 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 internal medicine. 2016;176(6):816–25. pmid:27183032.
  21. 21. Paffenbarger RS Jr., Hyde RT, Wing AL. Physical activity and incidence of cancer in diverse populations: a preliminary report. The American journal of clinical nutrition. 1987;45(1 Suppl):312–7. pmid:3799521.
  22. 22. Veierod MB, Thelle DS, Laake P. Diet and risk of cutaneous malignant melanoma: a prospective study of 50,757 Norwegian men and women. International journal of cancer Journal international du cancer. 1997;71(4):600–4. pmid:9178814.
  23. 23. Gogas H, Trakatelli M, Dessypris N, Terzidis A, Katsambas A, Chrousos GP, et al. Melanoma risk in association with serum leptin levels and lifestyle parameters: a case-control study. Annals of oncology: official journal of the European Society for Medical Oncology / ESMO. 2008;19(2):384–9. pmid:17925285.
  24. 24. Kaskel P, Sander S, Kron M, Kind P, Peter RU, Krahn G. Outdoor activities in childhood: a protective factor for cutaneous melanoma? Results of a case-control study in 271 matched pairs. The British journal of dermatology. 2001;145(4):602–9. pmid:11703287.
  25. 25. Lee TK, MacArthur AC, Gallagher RP, Elwood MJ. Occupational physical activity and risk of malignant melanoma: the Western Canada Melanoma Study. Melanoma research. 2009;19(4):260–6. pmid:19531966.
  26. 26. Parent ME, Rousseau MC, El-Zein M, Latreille B, Desy M, Siemiatycki J. Occupational and recreational physical activity during adult life and the risk of cancer among men. Cancer epidemiology. 2011;35(2):151–9. pmid:21030330.
  27. 27. Shors AR, Solomon C, McTiernan A, White E. Melanoma risk in relation to height, weight, and exercise (United States). Cancer causes & control: CCC. 2001;12(7):599–606. pmid:11552707.
  28. 28. Edwards MK, Blaha MJ, Loprinzi PD. Influence of sedentary behavior, physical activity, and cardiorespiratory fitness on the atherogenic index of plasma. Journal of clinical lipidology. 2017;11(1):119–25. pmid:28391877.
  29. 29. Wientzek A, Floegel A, Knuppel S, Vigl M, Drogan D, Adamski J, et al. Serum metabolites related to cardiorespiratory fitness, physical activity energy expenditure, sedentary time and vigorous activity. International journal of sport nutrition and exercise metabolism. 2014;24(2):215–26. pmid:24226038.
  30. 30. Franks PW, Ekelund U, Brage S, Wong MY, Wareham NJ. Does the association of habitual physical activity with the metabolic syndrome differ by level of cardiorespiratory fitness? Diabetes care. 2004;27(5):1187–93. pmid:15111543.
  31. 31. Monninkhof EM, Elias SG, Vlems FA, van der Tweel I, Schuit AJ, Voskuil DW, et al. Physical activity and breast cancer: a systematic review. Epidemiology. 2007;18(1):137–57. pmid:17130685.
  32. 32. Svobodova A, Vostalova J. Solar radiation induced skin damage: review of protective and preventive options. International journal of radiation biology. 2010;86(12):999–1030. pmid:20807180.
  33. 33. Kennedy C, Bajdik CD, Willemze R, De Gruijl FR, Bouwes Bavinck JN, Leiden Skin Cancer S. The influence of painful sunburns and lifetime sun exposure on the risk of actinic keratoses, seborrheic warts, melanocytic nevi, atypical nevi, and skin cancer. The Journal of investigative dermatology. 2003;120(6):1087–93. pmid:12787139.
  34. 34. Normand SL. Meta-analysis: formulating, evaluating, combining, and reporting. Stat Med. 1999;18(3):321–59. pmid:10070677.
  35. 35. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50(4):1088–101. Epub 1994/12/01. pmid:7786990.
  36. 36. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. Bmj. 1997;315(7109):629–34. Epub 1997/10/06. pmid:9310563; PubMed Central PMCID: PMC2127453.
  37. 37. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21(11):1539–58. Epub 2002/07/12. pmid:12111919.
  38. 38. R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria2016.
  39. 39. Viechtbauer W. Conducting Meta-Analyses in R with the metafor Package. Journal of Statistical Software. 2010;36(3).
  40. 40. Olsen CM, Carroll HJ, Whiteman DC. Estimating the attributable fraction for cancer: A meta-analysis of nevi and melanoma. Cancer prevention research. 2010;3(2):233–45. pmid:20086181.
  41. 41. Gandini S, Sera F, Cattaruzza MS, Pasquini P, Abeni D, Boyle P, et al. Meta-analysis of risk factors for cutaneous melanoma: I. Common and atypical naevi. European journal of cancer. 2005;41(1):28–44. pmid:15617989.
  42. 42. Caini S, Gandini S, Sera F, Raimondi S, Fargnoli MC, Boniol M, et al. Meta-analysis of risk factors for cutaneous melanoma according to anatomical site and clinico-pathological variant. European journal of cancer. 2009;45(17):3054–63. pmid:19545997.
  43. 43. Dennis LK, Vanbeek MJ, Beane Freeman LE, Smith BJ, Dawson DV, Coughlin JA. Sunburns and risk of cutaneous melanoma: does age matter? A comprehensive meta-analysis. Annals of epidemiology. 2008;18(8):614–27. pmid:18652979; PubMed Central PMCID: PMC2873840.
  44. 44. Ragnarsson-Olding BK. Spatial density of primary malignant melanoma in sun-shielded body sites: A potential guide to melanoma genesis. Acta oncologica. 2011;50(3):323–8. pmid:21299449.
  45. 45. Martin-Gorgojo A, Llinares M, Viros A, Requena C, Garcia-Casado Z, Traves V, et al. Cutaneous melanoma primary site is linked to nevus density. Oncotarget. 2017;8(58):98876–86. pmid:29228734; PubMed Central PMCID: PMC5716774.
  46. 46. Whiteman DC, Watt P, Purdie DM, Hughes MC, Hayward NK, Green AC. Melanocytic nevi, solar keratoses, and divergent pathways to cutaneous melanoma. Journal of the National Cancer Institute. 2003;95(11):806–12. pmid:12783935.
  47. 47. Ribero S, Glass D, Aviv A, Spector TD, Bataille V. Height and bone mineral density are associated with naevus count supporting the importance of growth in melanoma susceptibility. PloS one. 2015;10(1):e0116863. pmid:25612317; PubMed Central PMCID: PMC4303431.
  48. 48. Mangino M, Christiansen L, Stone R, Hunt SC, Horvath K, Eisenberg DT, et al. DCAF4, a novel gene associated with leucocyte telomere length. Journal of medical genetics. 2015;52(3):157–62. pmid:25624462; PubMed Central PMCID: PMC4345921.
  49. 49. Mundstock E, Zatti H, Louzada FM, Oliveira SG, Guma FT, Paris MM, et al. Effects of physical activity in telomere length: Systematic review and meta-analysis. Ageing research reviews. 2015;22:72–80. pmid:25956165.
  50. 50. Chilton W, O'Brien B, Charchar F. Telomeres, Aging and Exercise: Guilty by Association? International journal of molecular sciences. 2017;18(12). pmid:29186077; PubMed Central PMCID: PMC5751176.
  51. 51. Dai H, Sun Q, Zhang C, Zhang X, Li WQ, Manson JE, et al. Associations between benign cutaneous nevi and risk of Type 2 diabetes mellitus in men and women: results from two prospective cohort studies. Diabetic medicine: a journal of the British Diabetic Association. 2017;34(7):925–33. pmid:27917515; PubMed Central PMCID: PMC5459671.
  52. 52. Sport New Zealand. Sport and Active Recreation in the Lives of New Zealand Adults. 2013/14 Active New Zealand Survey Results. 2015: last accessed on 19 September 2018 under https://sportnz.org.nz/assets/Uploads/attachments/managing-sport/research/Sport-and-Active-Recreation-in-the-lives-of-New-Zealand-Adults.pdf.
  53. 53. Scragg R, Camargo CA Jr. Frequency of leisure-time physical activity and serum 25-hydroxyvitamin D levels in the US population: results from the Third National Health and Nutrition Examination Survey. American journal of epidemiology. 2008;168(6):577–86; discussion 87–91. pmid:18579538; PubMed Central PMCID: PMC2727193.
  54. 54. Olsen CM, Carroll HJ, Whiteman DC. Familial melanoma: a meta-analysis and estimates of attributable fraction. Cancer epidemiology, biomarkers & prevention: a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology. 2010;19(1):65–73. pmid:20056624.
  55. 55. Nikolaou V, Stratigos AJ. Emerging trends in the epidemiology of melanoma. The British journal of dermatology. 2014;170(1):11–9. pmid:23815297.
  56. 56. 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. European journal of cancer. 2013;49(3):642–57. pmid:23200191.
  57. 57. Vissers D, Hens W, Taeymans J, Baeyens JP, Poortmans J, Van Gaal L. The effect of exercise on visceral adipose tissue in overweight adults: a systematic review and meta-analysis. PloS one. 2013;8(2):e56415. pmid:23409182; PubMed Central PMCID: PMC3568069.
  58. 58. Ramos JS, Dalleck LC, Tjonna AE, Beetham KS, Coombes JS. The impact of high-intensity interval training versus moderate-intensity continuous training on vascular function: a systematic review and meta-analysis. Sports medicine. 2015;45(5):679–92. pmid:25771785.
  59. 59. Lavie CJ, Church TS, Milani RV, Earnest CP. Impact of physical activity, cardiorespiratory fitness, and exercise training on markers of inflammation. Journal of cardiopulmonary rehabilitation and prevention. 2011;31(3):137–45. pmid:21427600.
  60. 60. Mehanna E, Hamik A, Josephson RA. Cardiorespiratory Fitness and Atherosclerosis: Recent Data and Future Directions. Current atherosclerosis reports. 2016;18(5):26. pmid:27005804.
  61. 61. Loprinzi PD. Cardiorespiratory Capacity and Leukocyte Telomere Length Among Adults in the United States. American journal of epidemiology. 2015;182(3):198–201. pmid:26153476.
  62. 62. Holman DM, Berkowitz Z, Guy GP Jr., Hartman AM, Perna FM. The association between demographic and behavioral characteristics and sunburn among U.S. adults—National Health Interview Survey, 2010. Preventive medicine. 2014;63:6–12. pmid:24589442; PubMed Central PMCID: PMC4535173.
  63. 63. Lai JK, Lucas RM, Armstrong M, Banks E. Prospective observational study of physical functioning, physical activity, and time outdoors and the risk of hip fracture: a population-based cohort study of 158,057 older adults in the 45 and up study. Journal of bone and mineral research: the official journal of the American Society for Bone and Mineral Research. 2013;28(10):2222–31. pmid:23609238.
  64. 64. Jardine A, Bright M, Knight L, Perina H, Vardon P, Harper C. Does physical activity increase the risk of unsafe sun exposure? Health promotion journal of Australia: official journal of Australian Association of Health Promotion Professionals. 2012;23(1):52–7. pmid:22730941.
  65. 65. Schaefer L, Plotnikoff RC, Majumdar SR, Mollard R, Woo M, Sadman R, et al. Outdoor time is associated with physical activity, sedentary time, and cardiorespiratory fitness in youth. The Journal of pediatrics. 2014;165(3):516–21. pmid:25043155.
  66. 66. Savage SA, Gerstenblith MR, Goldstein AM, Mirabello L, Fargnoli MC, Peris K, et al. Nucleotide diversity and population differentiation of the melanocortin 1 receptor gene, MC1R. BMC genetics. 2008;9:31. pmid:18402696; PubMed Central PMCID: PMC2324112.
  67. 67. Bassett DR Jr., Pucher J, Buehler R, Thompson DL, Crouter SE. Walking, cycling, and obesity rates in Europe, North America, and Australia. Journal of physical activity & health. 2008;5(6):795–814. pmid:19164816.
  68. 68. Pustisek N, Sikanic-Dugic N, Hirsl-Hecej V, Domljan ML. Acute skin sun damage in children and its consequences in adults. Collegium antropologicum. 2010;34 Suppl 2:233–7. pmid:21302727.
  69. 69. Lazovich D, Vogel RI, Berwick M, Weinstock MA, Warshaw EM, Anderson KE. Melanoma risk in relation to use of sunscreen or other sun protection methods. Cancer epidemiology, biomarkers & prevention: a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology. 2011;20(12):2583–93. pmid:22016471.
  70. 70. Parr CL, Hjartaker A, Laake P, Lund E, Veierod MB. Recall bias in melanoma risk factors and measurement error effects: a nested case-control study within the Norwegian Women and Cancer Study. American journal of epidemiology. 2009;169(3):257–66. pmid:19011116.
  71. 71. Han J, Colditz GA, Hunter DJ. Risk factors for skin cancers: a nested case-control study within the Nurses' Health Study. International journal of epidemiology. 2006;35(6):1514–21. pmid:16943234.
  72. 72. Berwick M, Chen YT. Reliability of reported sunburn history in a case-control study of cutaneous malignant melanoma. American journal of epidemiology. 1995;141(11):1033–7. pmid:7771439.
  73. 73. Weinstock MA, Colditz GA, Willett WC, Stampfer MJ, Rosner B, Speizer FE. Recall (report) bias and reliability in the retrospective assessment of melanoma risk. American journal of epidemiology. 1991;133(3):240–5. pmid:2000841.
  74. 74. Malagoli C, Vinceti M, Pellacani G, Sieri S, Krogh V, Seidenari S, et al. Diet and melanoma risk: effects of choice of hospital versus population controls. Tumori. 2008;94(5):669–73. pmid:19112938.
  75. 75. May AM, Adema LE, Romaguera D, Vergnaud AC, Agudo A, Ekelund U, et al. Determinants of non- response to a second assessment of lifestyle factors and body weight in the EPIC-PANACEA study. BMC medical research methodology. 2012;12:148. pmid:23006680; PubMed Central PMCID: PMC3502143.
  76. 76. Van Loon AJ, Tijhuis M, Picavet HS, Surtees PG, Ormel J. Survey non-response in the Netherlands: effects on prevalence estimates and associations. Annals of epidemiology. 2003;13(2):105–10. pmid:12559669.
  77. 77. Tjonneland A, Olsen A, Boll K, Stripp C, Christensen J, Engholm G, 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. Scandinavian journal of public health. 2007;35(4):432–41. pmid:17786808.