3 Risk factors

While the average life time chance of getting cancer in the West for those born after 1960 is between 1 in 3 (women) to 1 in 2 (men) the actual risk for an individual will depend on a range of specific risk factors. Interestingly, there are significant difference in cancer incidence across the globe (Figure 3.1).5

Global overview of cancer types with the highest incidence by region

Figure 3.1: Global overview of cancer types with the highest incidence by region

Why do we see these differences?

Epidemiological studies aim to tease out such factors by studying large cohorts of individuals over long periods of time. By correcting for various factors which may confound findings independent of direct health impacts (e.g. geography, socioeconomic status, health care access,…) using statistical approaches these may suggests correlations between certain factors and an individual’s risk of getting cancer.

Some factors are linked with specific life style choices, including, for example diet, smoking, alcohol consumption and physical activity. These factors are in principle open to change e.g. supported by information and interventional campaigns. On the other hand there are factors that can not be easily changed – for example age or familial (genetic) risk. The prevalence of these various risk factors varies across the world, by region and country, and changes are often also related to other geographical or social factors.

Overall the risk factors are not dissimilar although the specific balance may differ. For example, in the West the most avoidable risks such as smoking, insufficient physical activity, alcohol, diet, overweight and obesity, are more relevant whereas infections contribute significantly to the risk in some of the developing countries.

3.1 ‘Risk factor’ age

Figures showing the development of cancer incidence with age clearly illustrate the importance of age. While there is quit a low cancer incidence in the first four to five decades the likelihood of a cancer diagnosis increases quite markedly over the following decades (Figure 3.2). The overall number of cases shows a maximum for those aged 60 - 70 years; the lower number of cases thereafter can be linked to overall decline in number of very old citizens.

Cancer incidence and number of cases for different age groups in England and Wales

Figure 3.2: Cancer incidence and number of cases for different age groups in England and Wales

3.2 The Effect of socio-economic status on incidence

For most types of cancers the incidence is higher than average for the more deprived groups in any population (positive percentage deprivation gap). This gap is particularly large for smoking-related cancers such as laryngeal, lung and oral cavity cancers. For a few types of cancers, including breast, prostate and malignant melanoma the incidence is lower in more deprived groups (negative percentage deprivation gap); this is potentially related to risk factors such as late first pregnancy, lower parity (number of children) and hormone replacement therapy which tend to be more prevalent in higher socio-economic groups.

3.3 Life style / avoidable risk factors

Clearly, some of these risk factors are based on an individuals choices and life style. In fact, it is thought that at least 40% of cancers cases in the UK can be linked to lifestyle and other avoidable factors.

The most important factor shown to affect the risk of cancer is clearly smoking:

  • Smoking

    • See below
  • Alcohol

    • Alcohol drinking causes an estimated 6% of deaths worldwide, around 1 in 8 of which are due to cancer (2012). Alcohol drinking prevalence is highest in Europe and America.
  • Diet

    • Unhealthy diets, e.g. low in fruit and vegetables and high in salt, are becoming more common in lower-resource countries.

    • Overweight and obesity are leading causes of death worldwide. Overweight and obesity prevalence is increasing particularly in low- and middle-income countries.

  • Infections

    • Globally around 18% of the cancer cases can be linked to infections with the effects being much more significant in low-income countries

3.3.1 Lung cancer

Lung cancer is one of the more common forms of cancer - and one of those most easily avoided, as 9 out of 10 lung cancers are caused by smoking and around 50% of smokers eventually die of smoking-related diseases. Other types of exposure can also cause lung cancer but are not typically as wide-spread, e.g. asbestos, radon, chemicals.

While the number of cigarettes smoked is important the duration of the smoking habit is most important. The exposure to smoke also affects those non-smokers that are exposed (passive smoking) e.g. through a spouse (25% increased risk) or high exposure in the work place (e.g. bars, 17% increased).

Tobacco smoke is an aerosol containing about 1010 particles/mL and around 4800 different chemical compounds of which at least 60 are known to be carcinogens. It also contains compounds that are known to be direct mutagens, such as polycyclic aromatic hydrocarbons (PAH), nicotine, or N-nitrosamines.6
It is also known to have indirect (enabling) mutagenic effects by enhancing the effects of other environmental toxins. While e-cigarettes avoid many of these factors it is important to be mindful of the fact that other components e.g. nicotine can also be harmful. ^[The role of nicotine in e-cigarettes - see e.g. here, or here. Smoking cessation programs are very important as the risks are reduced rapidly and are similar to non-smoker after 15 years.

Due to these mutagenic and carcinogenic compounds smoking is the single most preventable cause of death in the world. About 30% of these deaths (or 19 % in total) are caused by cancer. While in the West smoking is in decline an increasing proportion of the populations in low- or middle-income countries now smoke. Importantly, nine out of ten lung cancers are caused by smoking. More specifically, the individual risk for developing smoking related cancer depends on the number of cigarettes smoked regularly and more importantly the duration of the smoking habit. It is the prolonged, repeat exposure in particular that increases the risk as mutations accumulate over time (see Mutations and colon cancer).

From the turn of the 20th century smoking and cigarette manufacturing increased dramatically. This was eventually associated with an increased incidence of lung cancer but was initially considered to be associated with improved diagnostics, although a potential causative role of smoking was discussed. This view shifted only very gradually and against the strong resistance from interest groups and lobbies - thanks largely to the efforts of British epidemiologists Richard Doll and Tony Bradford Hill:

Initial studies in Germany (1936) demonstrating that less non-smokers were developing lung cancer than would have been expected from their proportion in the general population and case–control studies in the US (1950) associating lung cancer with cigarette smoking. Then, some decades later in the 50’s, Doll and Hill reported based on a larger case–control study that smoking was “a cause, and an important cause” of lung cancer (Doll and Hill (1950)). The definite study by Doll and Hill was a prospective cohort study with questionnaires to more than 34,000 male British physicians. Reports from the study (1954, 1956) confirmed the earlier findings clearly demonstrating higher mortality in smokers; furthermore the risk was linked to a clear dose–response dependent on the number of cigarettes smoked. It took until 1964 before this was recognised through a report of the United States Surgeon General, which concluded that smoking is causally related to lung cancer – making smoking the first clearly recognised general cause of cancer. The study was regularly updated (most recent 2004), showing that prolonged cigarette smoking reduced average life expectancy by 10 years (Alberg, Shopland, and Cummings (2014)). It also established that smoking cessation at the age of 50 halved these risks while cessation at the age of 30 normalised the situation almost completely. About 1 billion men worldwide remain smokers with smoking being linked to 1.2 million deaths worldwide every year.

Considering the public health impact policy makers in many countries have been relatively slow to respond to these insights although a benefit could be demonstrated quite rapidly ( see link).

3.3.2 Stomach cancer

Worldwide incidence of stomach cancer.

Figure 3.3: Worldwide incidence of stomach cancer.

More than 70% of the cases of stomach cancer are found in developing countries with global incidence varying from between 1 (Botswana) to 62 (Korea) per 100,000. While smoking also increases the link of stomach cancer, more than 50% of stomach cancer cases are associated with (chronic) infections of H. pylori, a bacterium that colonizes the stomach. While the mode or route of infection has not been fully elucidated although prognosis when treated is typically very good. In addition, there are dietary risk factors thought to be linked to stomach cancer including a diet rich in salted foods, salted meat or fish, and pickled vegetables; smoked foods also appear to increase the risk. On the other hand, eating fresh fruits and vegetables appears to lower this risk.

3.3.3 Cervical cancer

Incidence of cervical cancer acrosss the globe

Figure 3.4: Incidence of cervical cancer acrosss the globe

Cervical cancer (or cancer of the cervix uteri) is the 4th leading cause of cancer death in women worldwide (2008). Again, the global distribution of cases is uneven with 90% of cervical cancer deaths occurring in developing countries. This death rate variability linked to cervical screening efforts and a number of risk factors. The major risk factor for this form of cancer is infection with human papilloma virus (HPV) which can increase the risk by more than 200-fold. HPV infection occur typically after sexual contact and sexual activity is therefore an important risk factor. However, typically infections remain symptomless and are cured spontaneously (no intervention). It is in cases where the infection persists that the cancer risk is increased most. Persistent infections are less common in healthy individuals and more commonly linked to to immunosuppression. This could be caused by underlying diseases (infections, parasites, HIV/AIDS) but also to the nutritional status, further combined with factors such high number of childbirths etc. The distribution of cervical cancers is thus strongly linked to regions where these challenges remain prevalent.

In addition, the immunisation program recently introduced in many countries will further contribute to a reduction as it provides protection against about 70% of viruses. Nevertheless, as a risk for persistent infections remains for other serotypes cervical screening remains important tool to improve diagnosis and treatment.

3.3.4 Sun exposure and melanoma

Malignant melanoma is the most aggressive form of skin cancer and the 6th most common form of cancer in the UK. The incidence has increased at least 5-fold since 1975. Melanoma is the most common form of cancer for young adults (25-29 years old). Melanoma is increasing faster in females (15-29 years old) than males in the same age group (see also Melanoma brochure. While malignant melanomas are more common it is important that other form of malignant skin cancers exist (skin cancer gallery).

Sun exposure is essential for the production of vitamin D but the increased risk of malignant melanoma is clearly linked to life style changes (more exotic holiday) and changed beauty ideals (sun bathing) that particular for populations with light skin types/ fair hair lead to accumulative damage through UV radiation (and thus risk of mutations). In addition to causing cancer sunlight causes other forms of photo damage with symptoms similar to extreme ageing dermatoheliosis. Ultraviolet A (UvA) rays can penetrate window glass and thus affect the epidermis and dermis with chronic exposure leading to photoaging (dermatoheliosis). This is not directly linked to carcinogenesis but UvB and to a lesser extent UvA are linked to photocarcinogenesis, via DNA mutations and direct toxicity.

3.4 Genetic risk factors

The variation of life styles of individuals and across the globe significantly shape the risk of developing a specific type of cancer. This variability also implies that these factors lead to exposure to agents that cause cancer and help to unravel the complex causes of cancer. However, risk factors are not only linked to external and avoidable causes but the observation that sometimes the risk of cancer in increase in specific families suggests that there may be a genetic predisposition that may contribute to an individuals overall cancer risk.

3.4.1 BRAC1 and BRAC2

An example for such a genetic predisposition has been shown to be linked to the DNA repair genes BRAC1 and BRAC2. The BRCA1 and BRCA2 genes produce a protein that helps to repair damaged DNA. This reduces the risk of incorrect DNA repair leading to further DNA damage and therefore reduces the downstream risk of DNA alterations promoting cancer. Therefore BRAC1 and BRAC2 are also considered “tumour suppressors” (see [Tumour suppressor genes]).

Their role is to maintain integrity of DNA by repairing specific types of damage. If the function for either of these genes is reduced as a consequence of a mutation the inability repair leads to a higher rate of errors and thus potential mutations in the affected cells.

As a heritable predisposition these defective gene are passed on in the germ line and affect all cells. Harmful mutations to the BRCA1 and BRAC2 genes can therefore be inherited from father and mother (heterozygous phenotype) and carry an increased risk of female breast, ovarian and some other cancers. The women affected have an increased risk of breast cancer (between 10 - 60%, depending on other factors) but also for other types of cancers e.g. ovarian. These cancers also tend to develop at a younger age. BRCA1&2 are involved with about a quarter of all hereditary breast cancers. The risk of developing breast cancer by the age of 70 increases from around 12 percent (general population) to 45% (BRCA2) to 55 - 65 % (BRCA1). For ovarian cancer the risk increases from 1.3% (general population) to an estimated 15% (BRCA2) or 39% (BRCA1). In addition other risk factors will affect an individuals overall risk. BRCA1&2 mutations can also increase the risk of some other forms of cancer in men and women, for example pancreatic cancer or prostate cancer.

3.4.1.1 Management7

Often one of the ethically and medically difficult questions is how to deal with heritable predispositions like the BRAC1 and BRAC2 mutations as there is not necessarily always a satisfactory medical treatment to alleviate the risk.
BRCA1 and BRAC2 mutations can be diagnosed in DNA samples from saliva or blood. Overall the mutations are relatively rare and testing tends to be only recommended if there are reasons e.g. familial cancer history to suspect such alterations. In general counselling should be considered to help to identify cases where testing should be considered but also to help patients to put the positive and negative results into the appropriate genetic context.

There are three principle approaches to managing the increased risk associated with harmful BRCA1 and BRAC2 mutations:

  • Enhanced screening (earlier, more frequent, additional modalities, biomarkers)

  • Prophylactic surgery (bilateral mastectomy, salpingo-oophorectomy)

  • Chemoprevention


  1. Further maps for exploration at Global cancer observatory↩︎

  2. Mutagens are agents know to lead to changes in the DNA of a cell such changes increase the risk of cancer. Carcinogens on the other hand will lead to a rapid development of cancer; this typically would be due to DNA changes but could also include other causative factors (e.g. epigenetic); more detail later.↩︎

  3. BRCA fact sheet, NICE guidance↩︎