GI Values and How You Chew: How Far the Evidence on the Blood Sugar Curve Actually Goes

This article previously stated that “GLP-1 secreted by chewing well puts the pancreas’s insulin on standby in advance, constructing a gentle blood sugar curve,” and that swallowing food without chewing “causes an intense blood sugar spike followed by a crash.” The studies this site cites do not support that. The article has been rewritten.

From convenience-store packaging to the bands around health books, “low-GI, so it won’t make you gain weight” is everywhere. The glycemic index is a relative number for how much a food raises blood sugar compared with white rice or glucose.

So does changing how you chew the same food change the blood sugar curve? That is the question.

The trial that compared chewing directly with not chewing

E20 (Kamemoto et al. 2024, Scientific Reports) is a randomised crossover trial in 19 healthy young men. On one day they ate shredded cabbage while chewing to a metronome; on another they ate the same amount pureed and swallowed without chewing. Each was followed by an identical energy jelly, with 180 minutes of follow-up.

The results include findings that support chewing and findings that do not — in the same trial.

• Total GLP-1 was higher in the chewing condition at 45 (p=0.002), 60 (p=0.017) and 90 minutes (p=0.049), and the incremental areas under the curve for insulin and GIP were also higher (p=0.019 / p=0.009)
• But GLP-1 was lower in the chewing condition at 120 (p=0.027) and 180 minutes (p=0.023), and the overall GLP-1 area under the curve did not differ
• Glucose showed no difference in area under the curve, and at 30 minutes it was actually higher in the chewing condition (p=0.007)

So this trial cannot be used to say that chewing flattens the blood sugar curve. For glucose, either there was no difference, or the difference ran the other way.

The study was funded by Kewpie Corporation, and 3 of the 9 authors are employees of that company. What was compared is food form (shredded versus pureed), not the number of chews — the same design as E05 (whole walnuts versus walnut paste).

What about trials that specified the number of chews

E17 (30 chews per bite) and E19 (chewing gum while fasting) manipulated the chew count instead. Both found a significant rise in GLP-1. But in those same trials, neither glucose nor insulin changed.

GLP-1 rising and the blood sugar curve changing are two different things. Within what this site can cite, the first has been observed in several trials and the second has not.

That does not make texture pointless

E04 (the Suita Study) is a cross-sectional study finding more metabolic syndrome among people with lower chewing ability. Being cross-sectional, it cannot say which came first. This design cannot separate “metabolism worsened because chewing ability declined” from “people already in poorer health became less able to chew.”

E02 is a doctoral thesis using mice, not a peer-reviewed paper. It cannot serve as evidence for human blood sugar dynamics.

What can reasonably be said is roughly this:

• Pairing a meal with firm side dishes does lengthen the time the meal takes
• Shifting towards pastes, smoothies and liquid meals removes the opportunity to chew at all
• But no trial cited on this site has shown that doing so flattens the blood sugar curve

What this article does not assert

• “Chewing prevents blood sugar spikes” — in E20 the glucose area under the curve did not differ, and at 30 minutes it ran the other way
• “GLP-1 puts insulin on standby in advance” — this site cites no study that measured such a mechanism
• “It prevents the crash (hypoglycaemia) two to three hours after a meal” — no cited trial measured hypoglycaemia

There is no reason to start worshipping the GI number on the packet. But there is also, for now, no basis for replacing it with “your chew count is the real GI value.”

---

This article is for general information and is not a substitute for diagnosis or treatment.