Can Diet and Sleep Lower Your Lung Cancer Risk?

Can diet and sleep lower your lung cancer risk?

They track with it. In 192,034 UK adults followed for a median of 14 years, people with both the best Mediterranean diet scores and the healthiest sleep had a 43% lower lung cancer risk than people with the worst of both (subdistribution hazard ratio 0.57). That is an observed association in a large group of people, not proof that changing your diet or sleep changes your odds.

The study pulled from the UK Biobank, a research project that recruited more than 500,000 British adults between 2006 and 2010 and has tracked their health ever since. The researchers included 192,034 people who had filled out at least one detailed diet questionnaire and who had no cancer diagnosis at the start. Median age was 57 and 54% were female. Follow-up ran through September 2023, and during that time 1,156 people were diagnosed with lung cancer.

Each person got two scores. The diet score was an adapted version of the Mediterranean Diet Adherence Screener, or MEDAS, built from 11 food categories: olive oil, vegetables, fruits, red meat, sweetened or fizzy drinks, wine, legumes, fish and shellfish, sweets and pastries, nuts, and white meat. The sleep score was simpler. Five yes-or-no questions, one point each: sleeping seven to eight hours a night, rarely or never having insomnia, not snoring, rarely dozing off during the day, and being a morning person rather than an evening person. Five points meant the healthiest sleep pattern.

What the data show

Diet and sleep each mattered on their own. People in the top third for Mediterranean diet adherence had about a 21% lower lung cancer risk than people in the bottom third (hazard ratio 0.79). The real reduction is very likely somewhere between 8% and 31% lower, and there is roughly a 2 in 1,000 chance the finding is coincidence. People in the top third for sleep had about a 26% lower risk (hazard ratio 0.74), very likely between 12% and 38% lower, with about a 1 in 1,000 chance of coincidence.

Combining them produced a step-by-step gradient. The researchers cross-classified diet and sleep into nine levels, from worst-worst to best-best. Compared with the worst level, the best level carried about a 43% lower risk (hazard ratio 0.57), very likely between 21% and 59% lower.

A simplified version of the analysis gives the plainest numbers. When the researchers cut both scores at the halfway mark and dropped anyone diagnosed in the first two years, 540 of 78,804 people in the worst combined group developed lung cancer, against 77 of 22,024 in the best group. Those are raw counts, before any adjustment for smoking or anything else, but they show the scale honestly: only a small fraction of either group was ever diagnosed.

Dr. Kumar’s Take

Lung cancer is a smoking story first, and nothing here changes that. The authors adjusted for smoking status, but only in three broad buckets: never, previous, and current. That does not capture how much someone smoked, for how long, or how long ago they quit. The authors say so themselves and call residual confounding by smoking a real limitation. People who eat well and sleep well are also, on average, lighter smokers, and no statistical adjustment fully separates those things.

What makes the result harder to dismiss is the shape of it. The risk fell step by step as diet and sleep both improved, rather than jumping around, and it held up when the researchers changed the definitions and dropped early diagnoses. The signal also did not show up for other common cancers, which cuts against a generic “healthy people are healthy” explanation.

The sleep finding is the more novel half. Diet and cancer has been studied for decades, but a five-part sleep score linked to lung cancer specifically, after adjusting for smoking, is newer ground. Whether that reflects melatonin, inflammation, immune surveillance, or simply that poor sleep marks something else about a person, this study cannot tell you.

Where the effect was strongest

Splitting by sex changed the picture. Among women, the two highest combined levels showed clear reductions, with the very best group at about a 51% lower risk (hazard ratio 0.49). Among men, the reductions showed up in the middle-to-high levels but not the top ones. The authors point to the obvious explanation: only 20 or fewer men in those top groups developed lung cancer at all, which is too few to measure anything reliably. The honest reading is that the direction looks similar in both sexes and the precision is better in women.

The researchers also tested whether a siesta habit changed the diet result, since afternoon naps are part of the Mediterranean cultural package. It did not.

Practical Takeaways

Nothing in this study competes with not smoking, which remains by far the largest modifiable driver of lung cancer risk.
The five sleep habits scored here are ordinary ones you can check yourself: seven to eight hours, falling asleep without a fight, no snoring complaints, no daytime dozing, and an earlier schedule.
The diet side was measured across everyday categories, so shifting toward vegetables, fruit, legumes, fish, nuts, and olive oil while pulling back on red meat, sweets, and sugary drinks is the whole intervention.
Treat the 43% figure as a comparison between two groups of people, not a promise about what changing your habits will do to your own risk.

FAQs

Does this mean a Mediterranean diet prevents lung cancer?

No. This was an observational study, which means the researchers watched what happened to people rather than assigning anyone to a diet. The authors are explicit that the finding is an association and does not establish a causal effect. Diet was also measured with 24-hour recall questionnaires, and 76,523 of the 192,034 participants had data from only a single one of them, while just 5,760 had data from all five rounds, so for many people the diet score is a rough snapshot of habits rather than a precise record of years of eating.

Why did the researchers use a different kind of statistical model?

Most studies of lifestyle and cancer use standard survival models that ignore a practical problem: people can die of something else before cancer has a chance to appear. This team used Fine-Gray subdistribution hazard models, which account for death from any cause as a competing event. That matters in a group followed for 14 years starting at a median age of 57, where deaths from heart disease and other causes are common. It produces more conservative estimates than the usual approach.

Should smokers or former smokers take this seriously?

The researchers tested whether smoking status changed the diet-and-sleep association and found no significant difference, meaning the pattern looked broadly similar across never, former, and current smokers. That is worth knowing, but it comes with the same caveat as the rest of the study: smoking was recorded in three coarse categories, so the test was not a fine-grained one. For anyone who currently smokes, quitting still dominates every other lever by a wide margin.

Bottom Line

Across 192,034 UK adults and 1,156 lung cancer diagnoses over 14 years, better Mediterranean diet adherence and healthier sleep each tracked with lower lung cancer risk on their own, about 21% and 26% lower respectively, and the two together tracked with a 43% lower risk. The gradient was consistent, survived sensitivity analyses, and did not appear for other common cancers. It is still observational, smoking was adjusted for only crudely, and lung cancer remained an uncommon diagnosis in every group. That is a reason to take diet and sleep seriously as part of a long game, not a reason to believe either one is protective on its own.

Read the full study