The Portfolio diet tracked 10% lower death risk across 513,284 people in four cohorts

Most famous diets were spotted before they were designed. Somebody noticed that Cretans had good hearts, or that Okinawans lived a long time, and the research followed the population.

The Portfolio diet went the other way. It was assembled in a lab from four foods that each lower cholesterol a little, on the theory that four small effects stacked would rival a drug: plant sterols, viscous soluble fiber, soy protein and nuts, eaten together rather than one at a time. For two decades that was all it was, a cholesterol intervention with a good mechanism and no long-term outcome data.

Chan and colleagues, writing in Critical Reviews in Food Science and Nutrition, have now pooled 10 studies consisting of 513,284 participants to ask whether the people who happen to eat this way actually live longer. Each step up the adherence score went with a 10% lower risk of all-cause mortality.

What a standard deviation of diet looks like

The unit here needs unpacking, because it is not “switching to the Portfolio diet”.

Each 1-standard deviation increase in the Portfolio Diet Score was associated with a 6% lower risk of cardiovascular disease and a 10% lower risk of all-cause mortality. A standard deviation is roughly the typical spread across the population, so the comparison is between someone eating this way a bit more than average and someone eating it a bit less. It is a nudge along a scale, not a conversion.

That makes the mortality figure larger than it first appears. Ten percent for a nudge is a bigger claim than ten percent for a wholesale change, because the exposure contrast is so much smaller.

Where the numbers come from

Not all 513,284 people contributed to the headline. Data from 4 prospective cohort studies, covering 352,396 people, were pooled using random-effects models for disease outcomes; the rest of the studies informed the review without entering that calculation.

Four cohorts is a thin base for a pooled estimate, and it is the main reason to hold this result loosely. The intervals around both figures are tight, which reflects the enormous sample rather than a large number of independent findings. Tight intervals from few studies are a different thing from tight intervals from many.

Everyone in those cohorts also chose their own food. People who score high on a plant-sterol-and-soy-protein index are, on average, people who exercise, do not smoke, and can afford almonds. The authors adjust for what the original studies measured, and there is no adjusting for what they did not.

The two outcomes that did not get pooled

Qualitative synthesis suggested that higher adherence was associated with lower type 2 diabetes risk and lower odds of cancer. Qualitative synthesis is the phrase to notice: the studies were described side by side rather than combined into a number, usually because they measured things too differently to add up.

The authors are direct about what that is worth, calling those associations preliminary and hypothesis-generating. Treat them as reasons to run the next study, not as findings.

Why this diet is different from the others

Heart disease is the leading cause of death in the United States, and it is also a major cause of disability. The standard dietary advice against it is a list: limit saturated fats, foods high in sodium and added sugars, and eat plenty of fresh fruit, vegetables, and whole grains.

The Portfolio diet is narrower than that and more specific. It names four things and asks you to eat all of them, because the case for it was never that these foods are wholesome. It was that plant sterols block cholesterol absorption, viscous fiber traps bile acids, soy protein alters liver handling of cholesterol, and nuts supply unsaturated fat in place of saturated. Four different routes to the same number.

That design is why the mortality finding matters more than another observational diet result would. The mechanism was specified in advance, the cholesterol effect was demonstrated in trials first, and this analysis asks whether the downstream outcome followed. It did, in people who were never assigned to anything.

What to actually put in the trolley

Nothing here requires a new shopping habit so much as four of them at once: oats or barley most days, a handful of nuts, soy in some form, and a sterol-fortified spread if you use one.

Cholesterol is the reason. High levels of cholesterol can clog your arteries and raise your risk of coronary artery disease and heart attack, and this pattern was built to move that number without a prescription.

Two limits belong on it. Nobody has randomly assigned people to eat this way for a decade and counted deaths, so the strongest claim available is association. And a diet that lowers cholesterol is not a replacement for a drug that lowers cholesterol; for anyone already on a statin, this is a thing to do alongside, and a conversation to have before changing anything.