News · Nutrition & Diet
Anti-inflammatory diet linked to longer life after cancer
In 7,686 Swedish patients with cancer of the prostate, bladder or kidney, those eating a more anti-inflammatory diet beforehand were less likely to die of anything. The link to dying of the cancer itself was weak.
- 7,686 Swedish patients with prostate, bladder or kidney cancer, mostly prostate.
- Diet was scored years before diagnosis, from food questionnaires, not after it.
- A one-step better diet score tracked with 9% lower death from any cause.
- For death from the cancer itself the signal was weak and faded under stricter analysis.
- The authors list four non-dietary explanations they cannot rule out.
Diet and cancer survival is a field crowded with overstatement. A study in European Urology Open Science is unusually careful about it, and the care is what makes it worth reading.
Two Swedish cohorts, 7,686 people who went on to develop a urologic cancer - meaning cancer of the prostate, bladder or kidney - and a diet score measured years before any of them knew.
How the anti-inflammatory diet score was built
The exposure is an Anti-Inflammatory Diet Index, built from food questionnaires and scored on how strongly a person’s eating pattern tracks with inflammation markers in blood. Higher means more vegetables, fruit, whole grains, fish, tea and coffee, and less processed and red meat.
Crucially it was assessed at least 2 yr before diagnosis, which removes a common trap. People change how they eat once they are told they have cancer, and studies that measure diet afterward end up measuring the diagnosis.
Among 7686 patients with valid AIDI, most had prostate cancer.
Two survival results, pointing different ways
Here the study splits, and the split is the point.
For death from any cause, the association was clear. Each 1-SD higher AIDI was associated with lower all-cause mortality, about 9% lower for each step up the scale.
For death from the cancer itself, it was not. Evidence for an association with urologic cancer-specific mortality was weak and borderline, and it faded further under an analysis designed to account for people dying of other things first.
So the people eating this way lived longer. What they were less likely to die of was mostly not their cancer.
That is a smaller claim than the headline version, and it is the one the data supports. It is also not nothing: dying less of heart disease and stroke while living with cancer is a real outcome that matters to real patients.
What else could explain the survival gap
This is the part that deserves credit. Rather than closing with a call to eat more vegetables, the paper names what else could produce its result.
Whether this reflects a causal effect of diet or residual confounding, healthy-lifestyle clustering, health care engagement bias, or differences in competing noncancer mortality requires further study.
Translated: the diet score may be standing in for smoking less, moving more, and going to the doctor sooner. People who eat this way differ from people who do not in ways a questionnaire cannot fully capture, and adjusting for the differences you can measure never removes the ones you cannot.
Health care engagement bias is the subtlest of the four. People attentive enough to eat carefully are often attentive enough to get symptoms checked early, and early detection changes survival regardless of anything on the plate.
What the diet score cannot show
Anything about changing your diet after a diagnosis, which is the question most patients actually have. This measured what people ate years before, in people who did not know what was coming.
The population is also narrow: Swedish, largely older, and dominated by prostate cancer, which has a survival profile unlike most other cancers. Whether the same holds in bladder or kidney cancer, or outside Sweden, is untested here.
What this means for eating after a cancer diagnosis
As a modest, honest result. Eating in this pattern before a cancer diagnosis went along with living longer afterward, mostly through causes other than the cancer, and the researchers cannot say the diet is why.
Diet is one of several things that get studied after a diagnosis rather than before one. The mental fog that follows treatment, usually called chemo brain, now has pooled evidence behind exercise as a treatment for it.
That is roughly what most nutrition findings should sound like. It is rarer than it should be that they do.
What gets studied least after a diagnosis is the rest of the body. Men on androgen deprivation therapy for prostate cancer carry a cardiovascular risk that, in a trial across eight countries, largely nobody was managing.
People also ask
What is an anti-inflammatory diet?
Here it is a score, the Anti-Inflammatory Diet Index, built from how often people reported eating foods that track with higher or lower inflammation markers in the blood. In practice a high score looks like more vegetables, fruit, whole grains, fish, tea and coffee, and less processed and red meat. It is not a branded diet plan, and no food in it is anti-inflammatory in a pharmaceutical sense.
Why does the difference between the two death measures matter so much?
Because it changes the story. Dying of anything covers heart disease, stroke, other cancers and infection. Dying of the urologic cancer, meaning the one they were diagnosed with, is the measure that would say diet acted on the tumor. The first was clear (HR 0.91; 95% CI, 0.88 to 0.95). The second was not (HR 0.95; 95% CI, 0.90 to 1.00), and it weakened further under an analysis that accounts for people dying of other causes first.
So did the diet help or not?
The most defensible reading is that people who ate this way before diagnosis died less of everything, and that most of that benefit was probably not about the cancer. That is still a real benefit. It is just a different claim from the one a headline would prefer, and the paper is careful not to make the stronger one.
Could it be that healthy eaters are just healthier generally?
Yes, and the authors say so directly. They list residual confounding, healthy-lifestyle clustering, health care engagement bias and differences in competing noncancer mortality as explanations they cannot exclude. People with high diet scores also tend to smoke less, move more and see doctors sooner, and no adjustment fully separates those.
Should someone with cancer change their diet based on this?
This is general information rather than medical advice, and the study measured diet years before diagnosis, so it says nothing about eating differently after one. Anyone in cancer treatment should take dietary advice from their own clinical team, who know their treatment and their nutritional needs.