News · Nutrition & Diet
The best diet-and-sleep combination tracked 43% less lung cancer across 192,034 people
Diet and sleep are usually studied one at a time. Scoring both together in the UK Biobank and following people for 14 years, the group with the best Mediterranean diet and the best sleep had the lowest lung cancer risk by a wide margin.
- People with both the best diet and the best sleep had 43% less lung cancer.
- Diet and sleep were scored together rather than one at a time, which is unusual.
- Smoking dominates lung cancer risk by far more than this, and travels with both habits.
- The analysis allowed for people dying of other causes first, which makes it conservative.
- 192,034 people over 14 years, and an association is not evidence of cause.
Almost every study of lifestyle and cancer takes one habit at a time. Diet in one paper, sleep in another, exercise in a third, and the reader is left to add them up.
Writing in the European Journal of Nutrition, researchers scored both at once. Mediterranean diet adherence was assessed using an adapted Mediterranean Diet Adherence Screener, and sleep was evaluated using a composite score based on five sleep behaviors, and the two were combined into nine joint levels.
Compared with participants in the lowest level, those in the highest level had a 43% lower risk of lung cancer.
Why the joint score is the point
Habits cluster. The person eating vegetables and olive oil is, on average, also the person going to bed at a sensible hour, and studying either in isolation attributes to one what may belong to both.
Combining them does not solve that. What it does is describe the gradient a person actually occupies, because nobody has a good diet and a bad life in every other respect. Nine levels from worst-worst to best-best is closer to how lifestyle presents itself than a single-exposure model.
The trade-off is precision. Splitting 192,034 people and 1,156 cases into nine boxes leaves modest numbers in the corners, which is why the interval on the headline figure runs from 0.41 to 0.79.
The confounder that governs everything here
Lung cancer is cancer that forms in tissues of the lung, usually in the cells that line the air passages, and it is the leading cause of cancer death in both men and women.
It is also, overwhelmingly, a smoking disease. Heavy long-term smoking multiplies the risk by tens; the difference reported here is a factor of about two between the best and worst lifestyle groups.
That matters because the two are not independent. People who score highly on a Mediterranean diet and a healthy sleep pattern are far less likely to have smoked, to have smoked as much, or to have lived with a smoker. Adjustment handles reported smoking status and pack-years imperfectly, because self-reported smoking histories are approximate and because passive exposure is rarely recorded at all.
No cohort study of diet and lung cancer escapes this. The honest reading is that some unknown share of a 43% difference is residual tobacco.
What the design does well
Two features are worth crediting.
The analysis used Fine-Gray proportional subdistribution hazards models that accounted for competing risks from all-cause mortality. In a cohort followed for fourteen years, people in the least healthy group die of other things before lung cancer can appear, and a model that ignores this exaggerates their apparent cancer risk. Accounting for it makes the estimate more conservative, not less.
The second is the siesta question, which was asked in advance. No evidence was found that siesta modified the association between Mediterranean diet adherence and lung cancer risk. A prespecified question answered in the negative is a small mark of discipline in a literature full of subgroup findings discovered after the fact.
What a plausible mechanism would look like
If the association is partly real, the candidates are unexciting and well-established.
A Mediterranean pattern delivers more antioxidants and polyphenols and less processed meat, and short sleep and irregular sleep are associated with higher inflammatory markers and disturbed melatonin, which has been studied in cancer for decades. Both plausibly affect how well cells repair damage.
None of that is tested here. The study measures an association between two lifestyle scores and a diagnosis, and the authors say so directly: this finding should be interpreted as an observational association rather than evidence of causality.
What to take from it
Not a diet for lung cancer prevention. There is no such thing, and presenting one would be a disservice.
What the study supports is narrower and still worth having: among people at similar smoking exposure, the ones eating and sleeping well appeared to fare better, and the two habits appeared to add to each other rather than substituting.
For anyone weighing what to change, the order of operations is unchanged. Tobacco first, by a distance nothing else approaches. Then, on this evidence, the diet and the sleep are worth treating as one project rather than two.
People also ask
What did the study find?
Compared with participants in the lowest level of the joint diet-sleep pattern, those in the highest level had a 43% lower risk of lung cancer (subdistribution hazard ratio 0.57, 95% CI 0.41-0.79; P=0.001) over a median 14 years, during which 1,156 incident lung cancer cases occurred.
How were diet and sleep measured?
Mediterranean diet adherence was assessed using an adapted Mediterranean Diet Adherence Screener, and sleep was evaluated using a composite score based on five sleep behaviors. Both were split into thirds and combined into nine joint levels.
What about smoking?
It is the dominant cause of lung cancer and the central problem for any study like this. People who eat well and sleep well also smoke less, and no statistical adjustment fully separates a lifetime of tobacco exposure from the habits that travel with it.
What is a competing-risk model?
One that accounts for the fact that people can die of something else before lung cancer has a chance to appear. In an older cohort followed for 14 years that matters, because ignoring it inflates the apparent risk in the least healthy group.
Did napping change anything?
No. No evidence was found that siesta modified the association between Mediterranean diet adherence and lung cancer risk, which was a prespecified question rather than something found afterwards.
How many cases is this built on?
1,156 across nearly 200,000 people. Lung cancer is common in absolute terms and rare in a cohort of relatively healthy volunteers, so the joint categories at the extremes contain modest numbers, which is why the interval is wide.
What should anyone do with this?
Nothing that displaces the one thing that matters. Not smoking, and not living with smoke, dominates lung cancer risk by an order of magnitude no diet approaches. This is an observational association rather than evidence of causality, and general information rather than medical advice.