Study Links Very Low and High Carb Diets to Higher Mortality

55% carbohydrates: the narrow corridor recommended by the study

A study indexed in PubMed suggests that both very low and very high carbohydrate diets are associated with higher mortality. The optimal point identified is around 55% of caloric intake, adding a second conclusion: plant-based fats and proteins are preferable to animal sources. The data come with caveats and an underlying question: how much exercise did the studied individuals perform?

At face value, the arithmetic is stark. A 55% carbohydrate intake on a 2,000-calorie diet equals 1,100 calories, nearly 300 grams daily. Proponents of ketosis aim for under 50 grams. Between these extremes lies nearly every nutrition manual published in the last twenty years.

How many carbohydrates are too few?

No single figure closes the debate. The range spans from 50 grams daily for ketogenic diets to nearly 300 grams for official recommendations, with intermediate proposals distributing 40% carbohydrates, 35% protein, and 25% fat. None stands alone: it depends on each person's energy expenditure.

This introduces the variable the study does not break down. A sedentary worker consuming 55% carbohydrates does not burn them; someone with a physical job or twenty hours of cycling a week does. The most common criticism is exactly this: carbohydrate is quick fuel, and using quick fuel without expending it takes a toll. The seal as a destination.

Can you live without carbohydrates?

The body produces glucose from proteins through gluconeogenesis, as ketogenic diet defenders note. From there to claiming eliminating carbohydrates is harmless is a stretch: long-term evidence is disputed, and ketosis has clinical indication for epilepsy.

The theory also circulates that carbohydrates and fats compete for the same cellular transporters, the so-called Randle cycle, and that mixing them should be avoided. It is argued that in nature, foods come with fat or carbohydrates, almost never both at once. Breast milk remains the inconvenient exception.

The study's fine print: who funds the research

In the conflicts of interest section, the California Walnut Commission and Dairy Management Inc appear as funders for one of the authors, along with NIH funds and personal payments from Novartis and Zogenix. The latter company specializes in epilepsy drugs: the condition where the ketogenic diet has recognized clinical indication.

That an antiepileptic manufacturer appears in a study on carbohydrates and mortality gives ammunition to those who suspect sponsorship. It proves nothing on its own. Nor does it help sell neutrality.

The tribe eating potatoes: 90% carbohydrates

The most cited counterexample is the Tukisenta of Papua New Guinea, with a diet composed of around 90% carbohydrates. Their cardiovascular analyses and several biomarkers come out surprisingly well.

The nuance is that these populations do not eat as humans did thousands of years ago: they eat tubers and wild roots, not refined flours. And here appears the other argument, the increase in genes coding for amylase, the enzyme that digests starch. After thousands of years of agriculture, genetics are no longer the same.

Ancel Keys, the rabbits, and the blue zones

The shadow looming over half a century of nutritional recommendations has a name. Ancel Keys was not a doctor or veterinarian: he studied zoology, and his thesis on fat and heart attacks came from an experiment with rabbits. A herbivore forced to eat what it does not eat.

With the so-called blue zones, something similar occurs. Several of their promoters are staunch defenders of strictly plant-based diets, and blood analyses of centenarians point in less comfortable directions than the question of a questionnaire about what someone ate twenty years ago.

The toll of antinutrients

The final twist of the matter is less segarro and more chemical. Legumes and nuts come with lectins, oxalates, solanines, phytates, and cyanogenic compounds. These substances, in high quantities, do no one any favors. Processing—soaking with vinegar or lemon, prolonged cooking, discarding the water—eliminates around 80% of most of them.

A percentage remains. Low, but it remains. And none of these compounds excludes plants from the diet; it simply requires preparing them as they have always been prepared. Raw potato is not food.

---

There is one detail no one disputes: breast milk is rich in carbohydrates and is the reference food for a newborn. From there, everyone chooses their side. The study puts a figure in the center, and the two extremes still do not shake hands: who funds the research, what the tribe ate, and how much the subject moved. With these three variables open, 55% functions as a headline, not a conclusion.