I want to be direct with you about this topic, because it tends to provoke either dismissal or disproportionate alarm, and neither serves you well. The evidence linking obstructive sleep apnoea to cancer is real, it is growing, and the mechanism is well understood. At the same time, the absolute risk elevation for most individual cancer types remains moderate, and there are as yet no studies proving that treating sleep apnoea definitively prevents cancer in humans. What I can tell you is that the biology is coherent, the population-level data is consistent across many different studies and countries, and the finding adds yet another reason to take an untreated sleep apnoea diagnosis seriously.
The Mechanism: How Oxygen Deprivation Fuels Tumour Biology
To understand the cancer link, you need to understand what solid tumours do when they run short of oxygen. As a cancer grows, its centre outpaces the blood supply and becomes starved of oxygen. The tumour responds by activating a molecule called hypoxia-inducible factor 1-alpha (HIF-1a). Think of HIF-1a as a master switch that turns on a programme designed to help the tumour survive and spread: it stimulates the growth of new blood vessels to feed the tumour, promotes resistance to the cell death mechanisms that would normally destroy cancer cells, and favours the type of metabolism that lets cancer cells function at low oxygen. The most aggressive, treatment-resistant cancers are typically characterised by high HIF-1a activity.
This is precisely the molecular cascade that intermittent nocturnal hypoxia from sleep apnoea activates across the entire body, every night. The difference is that in the context of sleep apnoea, this tumour-promoting switch is not being flicked on locally within a single cancer. It is being switched on systemically, in every tissue, repeatedly. In animal studies, intermittent hypoxia accelerates tumour growth, increases metastasis (spread to other organs), and worsens survival compared with controls exposed to the same tumour but without the hypoxia. The mechanism is not theoretical. It is experimentally established.
Intermittent nocturnal hypoxia mimics, across the whole body, the exact low-oxygen microenvironment that aggressive tumours create locally to fuel their own growth. This is not a distant analogy. It is the same molecular pathway.
The Population Evidence
A 2022 analysis pooling results from 20 different observational studies covering more than 5.3 million people found that severe nocturnal hypoxia in OSA patients is associated with a 28 to 43 per cent higher cancer incidence compared with those without significant oxygen drops during sleep. Cancer mortality — the risk of dying from cancer — was nearly tripled in patients with the most severe nocturnal hypoxia: researchers calculated a hazard ratio of 2.66, meaning those patients were 2.66 times as likely to die of cancer as those without significant overnight oxygen drops.
A hazard ratio of 2.66 for cancer death is a large number. To translate it into real terms: if someone without OSA has a 5 per cent chance of dying from cancer over the next decade, someone with severe nocturnal hypoxia from OSA would face a risk closer to 13 per cent over the same period. That is not a marginal difference. Cancer incidence also increases in a clear dose-response pattern with OSA severity: those with the highest oxygen desaturation indices have the highest cancer risk. This dose-response pattern is one of the most convincing features of the evidence, because it suggests the relationship is genuinely driven by the severity of the breathing problem, not simply by coincidence or shared lifestyle factors.
Cancer by Type: What the Specific Evidence Shows
The overall cancer risk elevation is not uniform across tumour types. Researchers have been able to quantify the association for several specific cancers. Rather than presenting only the statistical ratios, I have included the real-world baseline UK lifetime risk and where that moves with OSA, so you can judge the practical significance for yourself.
| Cancer Type | Hazard Ratio with OSA | UK Lifetime Baseline | Estimated Risk with OSA |
|---|---|---|---|
| Thyroid cancer | 2.32 (rising to 3.27 with longer follow-up) | Relatively rare — about 1 in 500 | About 1 in 215, rising to 1 in 150 over time |
| Melanoma (skin) | 1.71 | Roughly 1 in 40 | Roughly 1 in 23 |
| Kidney cancer | 1.81 | Roughly 1 in 52 | Roughly 1 in 29 |
| Colorectal cancer | 1.70 | Roughly 1 in 22 | Roughly 1 in 13 |
| Breast cancer (women) | 1.36 | Roughly 1 in 8 | Roughly 1 in 6 |
| Lung cancer | 1.28 | Roughly 1 in 14 | Roughly 1 in 11 |
| Liver cancer | 1.19 | Relatively uncommon | Around 20% higher than baseline |
I want to be careful with how these numbers are read. A hazard ratio of 1.36 for breast cancer does not mean that 36 per cent of women with OSA will develop breast cancer. A hazard ratio describes the relative change in risk — so the 1-in-8 UK lifetime risk for breast cancer rises to roughly 1 in 6 with OSA. That is a meaningful shift in absolute terms. For colorectal cancer, 1 in 22 becomes roughly 1 in 13. For melanoma, 1 in 40 becomes roughly 1 in 23. The thyroid cancer finding is the most striking in relative terms, but since thyroid cancer has a low baseline incidence even a doubled risk remains a small absolute number.
The Breast Cancer Finding in Detail
The breast cancer association deserves particular attention given its clinical relevance to a large patient group. A 2022 meta-analysis specifically examining OSA and breast cancer across multiple large long-term studies found a 36 per cent higher risk in women with OSA — that is where the 1-in-8 to 1-in-6 figure comes from. This association was independent of BMI (body mass index) and other confounders in studies that controlled for them adequately, meaning the OSA itself, not simply the weight that often accompanies it, is contributing to the elevated risk.
The proposed mechanisms include not only the HIF-1a pathway described above but also OSA-related disruption of melatonin production. Melatonin has documented anti-tumour properties, and its overnight secretion is suppressed by the sleep fragmentation OSA causes. Systemic inflammation, which OSA drives through elevated inflammatory markers, may also contribute. I would encourage oncology teams and breast surgeons to consider sleep apnoea screening as part of their assessment of patients with new diagnoses, because the condition is treatable and almost universally overlooked in that clinical setting.
A Note on Causation
Throughout this page I have used the word association, which I think is the honest choice. The observational studies establish correlation between OSA severity and cancer risk. The animal studies establish that the mechanism is plausible and experimentally demonstrable. So-called Mendelian randomisation studies — which use people's genetic variants as a kind of natural experiment to test whether one thing actually causes another, rather than simply occurring alongside it — have produced suggestive results for some cancer types, but have not yet definitively confirmed causality for all of them in humans.
There are also no human studies yet showing that treating OSA reduces subsequent cancer incidence. This is partly a practical problem: cancer takes years to develop, and adequately powered trials with long enough follow-up are extremely difficult to conduct. The absence of that evidence is not the same as evidence of absence.
My clinical position is this: sleep apnoea is worth treating for many well-established reasons covered elsewhere in this series. If treating it also reduces cancer risk, that is a welcome additional benefit. The evidence is enough to take seriously, but not enough to use as the sole argument for treatment, and certainly not enough to cause disproportionate anxiety in someone who has already received a cancer diagnosis.
References
[1] Tan BKJ et al. Association of obstructive sleep apnea and nocturnal hypoxemia with all-cancer incidence and mortality: a systematic review and meta-analysis of observational studies. Journal of Clinical Sleep Medicine. 2022. 20 observational studies, 5.3 million participants; severe nocturnal hypoxia associated with 28-43% higher cancer incidence and HR 2.66 for cancer mortality.
[2] Cheng Y et al. Obstructive sleep apnea and the risk of cancer: a systematic review and meta-analysis. Sleep Medicine. 2021. Meta-analysis; moderate-to-severe OSA associated with 2.62-fold elevated cancer risk versus mild OSA; clear dose-response with AHI and oxygen desaturation index.
[3] Tan BKJ et al. Obstructive sleep apnea and thyroid cancer risk: a systematic review and meta-analysis. Thyroid. 2022. Meta-analysis; HR 2.32 for thyroid cancer overall, rising to HR 3.27 in studies with 5+ year follow-up duration.
[4] Tan BKJ et al. Obstructive sleep apnea and risk of melanoma: meta-analysis. British Journal of Dermatology. 2021. Meta-analysis; HR 1.71 for melanoma in OSA patients.
[5] Yap DWT et al. Association between obstructive sleep apnea and breast cancer: a systematic review and meta-analysis. Sleep Medicine Reviews. 2022. Meta-analysis; HR 1.36 for breast cancer in women with OSA, independent of BMI.
[6] Teo YH et al. Obstructive sleep apnea and gastrointestinal cancers: a systematic review and meta-analysis. Sleep Medicine. 2022. Meta-analysis; HR 1.70 for colorectal cancer, elevated risk for other GI cancers.
[7] Cheong HH et al. Obstructive sleep apnea and risk of lung cancer: a systematic review and meta-analysis. Journal of Thoracic Oncology. 2021. Meta-analysis; HR 1.28 for lung cancer in OSA patients.
Contact Professor Vik Veer
If you would like to discuss a sleep assessment or arrange a consultation, please use the details below.
Private secretary: 0207 458 4584