Risk factors/epidemiology

Risk categories

Skin cancer risk is categorised into average risk, increased risk, and high risk by the Royal Australian College of General Practitioners, as described below.^[1]

Melanocytic skin cancer

Individuals with medium/dark skin colour and no other risk factors are at average risk of melanocytic skin cancer.

Individuals with the following characteristics are at an increased risk of melanocytic skin cancer:

• Family history of melanoma in first-degree relative (relative risk [RR] = 1.7)
• Fair complexion, a tendency to burn rather than tan, the presence of freckles, high naevus count (>100), light eye colour, light or red hair colour
• Presence of actinic damage (RR = 2)
• Past history of non-melanocytic skin cancer (NMSC) (<40 years higher risk)
• People with childhood high levels of ultraviolet (UV) exposure and episodes of sunburn in childhood (RR = 2)

Those at high risk (that is, those at greater than 6 times the average risk of melanocytic skin cancer) have have >5 atypical (dysplastic) naevi (atypically shaped moles) and/ora history of melanoma in themselves or a first-degree relative.^[1]

Non-melanoma (keratinocytic) skin cancer

Individuals with fair to lighter than olive skin, and are less than 40 years of age without any risk factors are at an average risk of non-melanoma skin cancer. Individuals with the following characteristics are at increased risk:

• fair complexion, a tendency to burn rather than tan, the presence of freckles, light eye colour, light or red hair
• family history of skin cancer
• aged >40 years
• male
• presence of multiple solar keratoses
• high UV exposure (for example, outdoor workers)

Those at high risk will have characteristics associated with increased risk, as well as a previous history of non-melanoma skin cancer, past exposure to arsenic and/or immunosuppression.^[1]

Age

Melanoma is more common in older adults than younger people,^[2][3] with the mean diagnosis age being 66 years among men and 62.3 years among women in 2018.^[4] Older adults have had more cumulative sun exposure than younger people, with every additional decade of high sun exposure shown to further increase the risk of melanoma.^[5][6] However, by limiting recreational sun exposure a person is likely to decrease their risk of melanoma, regardless of their age.^[5]

Although early onset melanoma is comparatively rare, melanoma is the most commonly diagnosed cancer among Australians aged 15-29 years.^[4] There were 1,225 new melanoma cases among those aged under 40 years in 2018 (estimated to be 1,169 in 2022).^[4]

It is estimated that in 2015, just over two-thirds (68%) of NMSC treatments were provided to people aged 65 and over.^[7]

Familial and personal skin cancer history

An individual’s skin cancer history is a risk factor for subsequent melanoma^[1] (particularly following Lentigo maligna melanoma or nodular melanoma)^[8] and non-melanoma skin cancer.^[9] In retrospective studies, the percentage of melanoma patients that develop second primary melanoma ranges between 2% and 20%^[10], varying according to the design and measures used.^[11]

The risk of melanoma is increased in relatives of cutaneous malignant melanoma patients. Approximately 5% to 10% of cases of cutaneous melanoma occur in families with a hereditary predisposition.^[12] Most cases of familial melanoma are due to shared sun exposure experiences among family members with susceptible skin types.^[13] Meta-analysis of 60 studies assessing family history of skin cancer (melanoma diagnosed in a first-degree relative) estimated an almost 2-fold increase in melanoma risk.^[14]

A prospective study by Ferrone and colleagues estimated a 19% cumulative five-year risk of a second primary tumour for patients who were diagnosed with a primary melanoma and also had a familial history of melanoma.^[8]

Phenotype

Traits such as red or blond hair, light-coloured eyes, fair skin, sun sensitive skin (particularly Fitzpatrick type I skin) and propensity to freckle are genetic risk factors for developing melanoma and keratinocytic skin cancers when combined with UV exposure.^[6][14]

Those with fair skin possess little epidermal eumelanin, which absorbs UV radiation and acts to neutralise UV radiation-generated free radicals, and are thus at greater risk of skin cancer.^[15][16] However, it is the high pheomelanin to eumelanin ratio in the melanin of individuals with a red-hair and fair skin phenotype that greatly increases their melanoma risk. Mouse models suggest this genetic factor may be a UV radiation-independent carcinogen, with UV overexposure likely to exacerbate melanoma risk among this phenotype.^[17]

Light eye colour is also a significant risk factor for ocular melanoma.^[18]

Genetics

Those with particular genetic variants may be more sensitive to UV radiation than others, and may therefore only require modest levels of exposure to initiate development of melanoma.^[19][20]

Common variations in at least twenty genes are known to influence melanoma risk in the population.^[21] In 2020 Landi and colleagues identified 31 new genetic regions associated with melanoma risk.^[22]

Polygenic risk scores (PRSs) aggregate the effects of many genetic variants across the genome into a single score aiming to reflect an individual’s genetic risk of disease. A 2022 study by Steinberg and colleagues derived newer melanoma PRSs from a larger, more diverse meta-analysis than had previously been undertaken, which was found to have. better risk prediction performance than an earlier PRS.^[23]

Research is continuing in this area in order to further improve risk analysis for individuals.

Naevi

The number of common naevi (moles) increases risk of cutaneous melanoma. A meta-analysis of 46 studies showed having between 101-120 naevi present a highly significant risk for melanoma - almost seven times greater (pooled relative risk of 6.89) than people with very few naevi (0-15 naevi).^[24]

^{Figure 1: Dysplastic naevi}

^{Images generously provided by Dr Alvin Chong, Skin & Cancer Foundation Victoria}

The presence of dysplastic (atypical) naevi also indicates a high skin cancer risk.^[1] Having more than five dysplastic naevi, as compared with having none, results in a melanoma RR (relative risk) of 6.36.^[24] Furthermore, patients with a primary melanoma more than double their five-year risk of a second primary melanoma (24% cf. 11% increase in risk) if they have dysplastic naevi.^[8]

Solar lentigines

Solar lentigines, also known as solar lentigo, are the result of actinic (sun) damage and have the appearance of small, tan-brown to black pigmented, flat or slightly raised spots. In particular, their presence is strongly associated with risk of lentigo malignant melanoma (LMM), Compared with zero solar lentigines on the arm, having more than 10 is associated with a fifteen-fold risk of LMM (compared with an almost five-fold risk of superficial spreading melanoma).^[25]

Sun exposure

In Australia, up to 95% of melanomas are attributable to overexposure to UV radiation.^[26][27] Skin exposure to UV radiation can result in DNA damage and mutations. These most commonly occur to the p53 suppressor gene through the formation of cyclobutane pyrimidine dimers (CPDs) UV-induced bonding between adjacent pyrimidines, such as thymine or cytosine DNA bases,^[28] and other mutagenic photoproducts which are primarily caused by UVB radiation.^[29] Cutaneous UV absorption also causes damage indirectly, due to the generation of excess reactive oxygen species (ROS) causing oxidative DNA damage (primarily UVA-induced).^[29] However, this has been suggested as not being "sufficient to induce mutations in the normal skin genome".^[30]

UV radiation is classified as a complete carcinogen because it causes mutations as well as general damage, and inducts and promotes tumour growth.^[16] Excessive UV exposure increases the risk of both melanoma and keratinocytic skin cancers.^[31]

^{Figure 1: UV-damaged, distorted DNA molecule showing cyclobutane pyrimidine dimers}

^{Image retrieved from Wikimedia Commons}

Both UVA and UVB radiation are independently associated with melanoma mortality according to an ecological study by Garland and colleagues, which links age-adjusted mortality rates with the spectrophotometric UV measurements of 45 countries (adjusting for population average skin pigmentation).^[32] UVB radiation has been shown to be more carcinogenic than UVA in experimental induction of squamous cell carcinoma (SCC).^[33] However, some research suggests that UVA may be more carcinogenic than previously thought. Data suggests that the basal skin layer is particularly vulnerable to UVA-induced damage.^[34][35]

Skin cancer risk is not just related to the amount of sun exposure, but also to the pattern. Melanoma and basal cell carcinoma risk are particularly linked to habitual low sun exposure combined with high recreational sun exposure. For example, this is exactly the pattern of many indoor workers, with high sun exposure on sunny weekends or summer holidays. On the other hand, a more continuous pattern of exposure, such as that due to occupational exposure, is associated with increased SCC risk.^[31][36]

Melanoma risk is associated with cumulative, intermittent exposure to UV radiation.^[31] A meta-analysis of 57 studies by Gandini and colleagues found a positive association between melanoma risk and intermittent sun exposure, but an inverse association for continuous over-exposure.^[37]

Chronic exposure in later life has also been shown to be an important risk factor for BCC - risk is higher among outdoor compared with indoor workers according to a systematic review and meta-analysis by Bauer and colleagues.^[38] Wu et al reported that chronic sun exposure in adulthood, as assessed by cumulative UV flux over long durations, were associated with substantially increased risks of BCC and SCC.^[39]

Childhood exposure

High sun exposure in the first 10 years of life more than doubles melanoma risk,^[5] while intense, intermittent sun exposure (number of sunburns and sunbathing vacations) during each decade up to 29 years of age increases risk of melanoma by more than one-and-a-half times.^[6] Wu et al reported that melanoma risk is strongly associated with sun exposure in early life, as evidenced by the number of blistering sunburns between ages 15 and 20 years.^[39]

Childhood is a critical period for melanoma risk because of sun exposure, as shown by ecological studies, and, less consistently, case-controlled studies.^[40] Intensive childhood UV exposure increases the risk of developing BCC, while chronic childhood exposure increases the risk of SCC.^[31]

Sunburn

While sunburn does not directly cause melanoma, it is often used as an indicator of melanoma risk. This is because it is memorable and is a marker of the type of sun exposure most associated with melanoma (intermittent intense exposure) combined with the type of skin that is most susceptible to melanoma (skin that is sensitive enough to UV to burn).^[36][40] Recalling childhood sunburn is a good indication that an individual was over-exposed to UV and had sun-sensitive skin – two factors that increase melanoma risk. Reductions in the frequency of sunburn are also often used as a practical and appropriate short-term outcome measure to assess intervention programs, which are aimed at reducing excess sun exposure and ultimately preventing melanoma.^[41][42][43]

A study by Pfahlberg and Kölmel found that having more than five sunburns doubled the risk of melanoma, with risk similar for childhood exposure before 15 years of age, and exposure after 15 years of age.^[44] However, other studies found that childhood and youth were critical periods of increased melanoma risk due to sunburn. The number of recalled sunburns prior to 30 years of age was found to significantly increase melanoma risk - however the positive relationship with melanoma risk was weaker for sunburns incurred between the ages of 30-39, and not evident for 40-49 years of age.^[6] Having a severe sunburn in childhood more than doubled melanoma risk in people aged 18-39 years in a case control study by Cust and colleagues.^[45] Painful sunburns before the age of 20 are associated with increased risk of melanoma (1.4 times increase in risk), squamous cell carcinoma (1.5 times increase risk), and certain basal cell carcinoma subtypes (1.6 times increase in risk).^[46] Savoye and colleagues^[47] found severe sunburns under the age of 25 was associated with increased risk of all skin cancer types. However, severe sunburns obtained at age 25 or older were not associated with an increased melanoma risk, only BCC and SCC.

Solarium use

A systematic review has shown a dose-response relationship between sunbed use and melanoma risk. Sunbed use increases melanoma risk by 20%, with an increase of 59% if used before 35 years of age.^[48] A study in the USA found that the more frequently people used solariums, the higher their risk of melanoma - regardless of the age at which they began using solariums.^[49]

A meta-analysis by Wehner and colleagues found that indoor tanning increased risk of both SCC and BCC, with risk particularly high among people who had used indoor tanning before 25 years of age (SCC relative risk=2.02; BCC RR=1.40).[50] Indoor tanning is a strong risk factor for early-onset BCC among people under 40 years of age, particularly among women.^[51]

Outdoor work

It is estimated that around 200 melanomas and 34,000 keratinocytic skin cancers per year are caused by occupational exposures in Australia.^[52]

Outdoor workers receive between five to ten times the annual dose of UV than indoor workers.^[53] Meta-analyses have found that the risk of SCC among outdoor workers is nearly double that of indoor workers^[54] while risk of BCC is increased by almost 1.5 times.^[38] These reviews conclude that outdoor work constitutes an independent and robust risk factor for the development of cutaneous SCC and BCC.^[54][38]

A meta-analysis by Gandini and colleagues did not find an association between high occupational sun exposure and melanoma.^[37] However there is evidence that melanomas of the head and neck are associated with chronic sun exposure.^[55] Melanoma risk is particularly linked to habitual low sun exposure combined with high recreational sun exposure.^[36]

Melanoma risk at different body sites is associated with different patterns of sun exposure at different latitudes. Recreational (i.e. intermittent) sun exposure is a strong predictor of melanoma on the trunk and limbs at all latitudes, whereas occupational exposure appears to predict melanoma on the head and neck - albeit weakly - predominately at low latitudes. Total sun exposure is associated with melanoma on the limbs at low latitudes.^[56]

Risk factors for second primary melanoma or keratinocytic skin cancer

A prospective study by Ferrone and colleagues estimated the cumulative five-year risk of a second primary tumour for patients diagnosed with a primary melanoma is 11%, with almost half that risk within the first year. Risk of a second primary melanoma increased to 19% for patients with a family history of melanoma, and 24% for patients with dysplastic naevi^[8] In 2020 Villani and colleagues conducted a retrospective study and found the risk to be 8.2%.^[11] Phenotypic characteristics (fair skin and inability to tan) also increased risk of second primary melanomas.^[57]

First primary lentigo or nodular melanomas had higher risks of multiple melanomas compared with the more common superficial spreading melanoma.^[57] High exposure to UV radiation (childhood exposure and lifetime recreation sun exposure) was also shown to increase risk of multiple primary melanomas.^[5]

Registry-linked data shows the cumulative incidence of second primary melanomas varies across populations, and is estimated at 6.4% at 10 years after diagnosis in Queensland^[58] - higher than in Switzerland or the US.^[59]

A meta-analysis of 17 studies by Marcil and Stern estimated that of patients with a history of non-melanoma skin cancer 44% would develop a subsequent basal cell carcinoma and 18% a squamous cell carcinoma within 3 years. The risk of developing a BCC from a prior SCC or BCC were similar, but the risk of developing a SCC following a BCC diagnosis was relatively low.^[9]

Back to top

Last modified: 5 August 2022