Sex Differences in Cancer Risk

Men and women often have different risks of the same diseases, including cancer. Considering and investigating them provides key insights into intervention strategies that can reduce disease burden.

My doctoral research investigated this for 28 cancer endpoints in over 470,000 people from the UK Biobank study by considering risk factors that may contribute to sex differences, making it one of the first comprehensive analyses on this topic. Here’s what I found.

Read the paper

Methodological Overview

Participants

Data from over 470,000 participants (54% women) aged 37-73 at recruitment were analysed.

On average, both men and women were aged 56 (SD 8) at recruitment.

They were followed up for an average of 10.5 (SD 2.2) years.

Cancer sites investigated

Over 32,300 incident cancer cases (58.1% in women) were identified in the cohort at shared anatomic sites.

28 cancer endpoints (15 sites and 13 subtypes) were selected for analyses.

Most common cancers

In men: colon, lung, melanoma, rectum, and non-Hodgkin lymphoma.

In women: breast (treated as a positive control), lung, colon, melanoma, and non-Hodgkin lymphoma.

Exposure: sex

Sex was recorded at recruitment and checked against genetic data for nearly all participants.

Participants were excluded where the recorded and genetically determined sex did not match.

Modelling approach

Cox proportional hazards models were used to estimate hazard ratios and 95% confidence intervals comparing cancer risk in men with women for each cancer site. Attained age was used as the underlying timescale so comparisons were made between participants at similar attained ages.

Analyses were stratified by age, deprivation index, and region. The considered extrinsic risk factors included socio-demographic factors (education, ethnicity) and health-related behaviours (smoking, alcohol consumption, physical activity) and intrinsic risk factors included biological characteristics (height, body mass index) for all cancer endpoints, and additionally site-specific risk factors.

Research background

Cancers with higher risk in

Men

compared to women

Oesophageal adenocarcinoma

5.5x

The risk for OAC was 5.45 times greater for men compared to women, stratifying by age, deprivation index, region and considering risk factors such as socio-demographic factors (education, ethnicity) and health-related behaviours (smoking, alcohol consumption, physical activity) and intrinsic risk factors included biological characteristics (height, body mass index) and gastro-oesophageal reflux disease.

Measures of central adiposity were also evaluated for their contributions to observed sex disparity. Waist circumference and waist-to-hip ratio were partially contributory to attenuating the relative risk from five-fold higher to 4.4- and 3.5-fold higher, respectively; however, a large male predominance remained.

Bladder cancer

3.5x

The risk for bladder cancer was 3.5 times greater for men compared to women, stratifying by age, deprivation, and region, and considering risk factors such as socio-demographic factors (education, ethnicity) and health-related behaviours (smoking, alcohol consumption, physical activity) and intrinsic risk factors included biological characteristics (height, body mass index).

The high incidence of bladder cancer in men compared to women is not fully understood but may be due to a combination of differences in carcinogenic exposures (through occupations), sex hormones, and genetic differences by sex.

Kidney cancer

1.8x

The risk for kidney cancer was 1.8 times greater for men compared to women, stratifying by age, deprivation, and region, and considering risk factors such as socio-demographic factors (education, ethnicity) and health-related behaviours (smoking, alcohol consumption, physical activity) and intrinsic risk factors included biological characteristics (height, body mass index) and hypertension.

Adjustment for measures of central adiposity instead of BMI, particularly waist circumference and waist-to-hip ratio strongly attenuated the association closer to 1.36-fold and 1.58-fold higher in men, respectively. It may be inferred central adiposity contributes strongly towards the sex differences of kidney cancer risk.

Rectal cancer

1.7x

The risk for rectal cancer was 1.7 times greater for men compared to women, stratifying by age, deprivation, and region, and considering risk factors such as socio-demographic factors (education, ethnicity) and health-related behaviours (smoking, alcohol consumption, physical activity) and intrinsic risk factors included biological characteristics (height, overall and central adiposity measures), and diabetes, inflammatory bowel disease, and processed meat intake.

Sex differences in rectal cancer remained after accounting for established risk factors, indicating that other rectal cancer-specific factors may be involved. Some residual confounding by alcohol is still possible because sensitivity analyses focused on light drinkers rather than never drinkers.

Cancers with higher risk in

Women

compared to men

Lung adenocarcinoma in never smokers

1.4x

Lung adenocarcinoma risk was higher in women compared to men (approximately 1.3-fold) and remained consistently elevated among never smokers (nearly 1.4-fold). In unadjusted analyses, the relative risk between men and women was close to the null; however, adjustment for established risk factors, particularly smoking, strengthened the association away from the null. This may reflect higher smoking prevalence among men masking an underlying higher risk of lung adenocarcinoma in women (50.0% men versus 39.9% women self-reported as current or former smokers).

Consequently, sex differences for lung adenocarcinoma in never smokers were investigated and the higher risk in women remained upon evaluation of contributions from household heating and cooking fuel exposure, a well-established risk factor for lung cancer, and asthma and menopausal status, candidate risk factors hypothesised to contribute to the observed sex differences. Altogether, the sex difference observed for lung adenocarcinoma for women was relatively modest compared to the sex differences for cancer sites with a male predominance.

What about the other cancer sites?

More data needed:

Thyroid: Thyroid cancer had a higher risk in women despite consideration of several risk factors. The observed differences are widely attributable to diagnostic differences. Women are observed to experience earlier and more frequent diagnoses due to higher healthcare seeking behaviours and referrals for gynaecological concerns.

Melanoma: For melanoma the differences in incidence by sex varied by age; under the ages of 65 women had a higher risk compared to men, whereas over the age of 65 men had a higher risk than women. Women in younger age groups tend to have higher melanoma rates than men, while men surpass women in mid-life to older age groups. More broadly, the age at which this shift occurs depends on the exposure to UV radiation levels specific to regions.

Breast: Breast cancer was included in the analyses as a positive control, demonstrating expected and stable sex differences across models.

Similar risk between men and women was observed for the following cancer sites after accounting for risk factors: