In this guide
Somewhere between "eat the rainbow" and "never drink your calories", a new rule has taken over health feeds: vegetables first, protein second, bread last. The claim is that the order of a meal changes the shape of your blood-sugar curve, even when the food on the plate is identical. For once, this isn't a wellness-influencer invention. Small human studies have pointed the same way for about a decade. What's been missing is a convincing explanation of why.
On 8 October, a team led from Kyoto University published one in the journal Diabetes, and it has been reported widely since. We read the press materials, the journal summary and the human studies it builds on, including two gestational-diabetes studies from this year, and tried to separate a promising mechanism from a ready-made diet rule. The honest answer sits in between.
The short version: in mice, including mice with type 2 diabetes and almost no insulin, olive oil given 15 minutes before glucose blunted the early rise in blood sugar, slowed the stomach's emptying, and triggered earlier release of the gut hormones GLP-1 and GIP and of glucagon. Blocking glucagon alone didn't abolish the effect, so glucagon and GLP-1 appear to work together in a partly redundant, nerve-assisted way. It's animal data from one paper, but it supplies a plausible biological reason for something people have observed: eating vegetables, protein or fat before carbohydrate can soften the post-meal spike.
Why "Eat Your Veg First" Became a Health-Feed Obsession
Post-meal blood sugar has become the new health metric. Continuous glucose monitors, once a diabetes tool, are now sold to everyone, and "glucose spikes" have become shorthand for everything from energy crashes to ageing. In that climate, a free trick that needs no supplement, no app and no food group to be banned is irresistible. "Meal sequencing", also called "food ordering", fits that bill perfectly.
The danger is the usual one. A promising idea gets turned into a rule, the rule gets stripped of its conditions, and soon people are drinking spoonfuls of olive oil before dinner because a headline said to. So let's go through what was actually done, what the best human data say, and what a sensible, low-effort version looks like.
The Kyoto Study at a Glance
| Item | Details |
|---|---|
| Title | Olive Oil Before Oral Glucose Delays Gastric Emptying and Lowers Glycemia Through Partially Redundant Glucagon and GLP-1 Signaling in Mice |
| Journal and date | Diabetes, 8 October 2026 (doi:10.2337/db26-0531); news release from Kyoto University the same day |
| Team | An international group led from Kyoto University, Japan |
| Model | Mice fasted overnight, then split into three conditions |
| Conditions | Water before glucose; olive oil 15 minutes before glucose; glucose 15 minutes before olive oil |
| Measured | Blood glucose, plasma insulin, glucagon, GIP and GLP-1, plus the speed of gastric emptying |
| Key result | Olive oil first (not after) blunted the early glucose rise and slowed gastric emptying; early glucagon, GIP and GLP-1 release rose |
| Diabetes models | Benefits were retained in mice with type 2 diabetes and severe insulin deficiency |
| Mechanism tests | Blocking glucagon signalling alone didn't eliminate the effect; vagus-nerve pathways were involved |
| Human data | None in this paper |
What the Mice Actually Did
The design is simple and elegant. Overnight-fasted mice were given glucose by mouth, which is a standard way to test how the body handles a sugar load. Some had water first. Some had olive oil 15 minutes before the glucose. Some had the glucose first and the olive oil 15 minutes afterwards. Everything else was held constant, so any difference came down to timing.
The result was clear. Olive oil given before glucose blunted the early rise in blood sugar and slowed how fast the stomach emptied into the intestine. Olive oil given after glucose didn't do the same. And the benefit survived in mice with type 2 diabetes and severe insulin deficiency. That last point is important, because it suggests the mechanism doesn't depend on a normally working insulin system. It's a gut-level effect.
The Hormone Story: GLP-1, GIP and Glucagon in Plain English
When fat reaches the small intestine, specialised gut cells release signalling hormones. Two of them, GLP-1 and GIP, are incretins, hormones that help the body prepare for incoming sugar. GLP-1 also slows the stomach's emptying and dampens appetite, which is the same pathway that GLP-1 medications target at much higher strength.
The surprise in this study is glucagon. It's usually described as the hormone that raises blood sugar, the opposite number to insulin. Yet in this experiment, a fat preload triggered an early glucagon surge, and glucagon seemed to help lower glucose, working alongside GLP-1 in what the authors call a "partially redundant" way. When researchers blocked glucagon signalling alone, olive oil still lowered glucose, because GLP-1 could carry much of the load. Vagus-nerve signalling, the body's gut-brain wiring, was part of the picture too.
The takeaway isn't that you should memorise hormone names. It's that eating fat early seems to set off a coordinated slowdown: the stomach empties more slowly, glucose arrives in the blood more gradually, and the early spike is smaller.
Why Mice Aren't People (and Why This One Still Matters)
- The doses and the delivery were experimental. Mice received olive oil by tube and glucose in a standardised solution, not a meal with chewing, texture, fibre and mixed ingredients.
- The gap was a neat 15 minutes. Real meals unfold over a messy 10–30 minutes, with bites of everything.
- Mouse gut physiology isn't human gut physiology. Gastric emptying rates, hormone responses and nerve wiring differ in detail.
- The study measured glucose and hormones, not long-term health. A smaller spike after one test meal doesn't prove better weight, HbA1c or heart outcomes over years.
Why it still matters: the paper doesn't claim sequencing works in people. It explains how it could, which makes it more reasonable to run proper human trials. Mechanism studies like this are the kind of work that turns "people have noticed this" into "here's the biology, now let's test it".
What the Human Evidence Says So Far
| Study | What it did | What it found |
|---|---|---|
| Shukla et al., Diabetes Care, 2015 (small pilot) | 11 adults with type 2 diabetes ate the same meal (bread, orange juice, protein, vegetables and fat) on two occasions, with carbohydrate first on one day and last (15 minutes after the rest) on the other | Glucose was about 29%, 37% and 17% lower at 30, 60 and 120 minutes when carbohydrate came last; insulin was lower too |
| Oat-bran preload pilot, Nutrients, 2016 | 10 healthy adults drank oat bran (rich in β-glucan fibre) in water before a white-bread meal | A viscous fibre preload reduced the glucose response, likely because it thickens stomach contents and slows digestion |
| DEFI-GDM trial (registered, completion due September 2026) | Randomised crossover in gestational diabetes: egg before toast versus toast before egg | Results pending; the registry notes earlier work suggesting peaks may fall by roughly 40%, a claim to treat cautiously until published |
| Rosenberg et al., Pregnancy, 2026 | Randomised pilot crossover in 16 women with diet-controlled gestational diabetes comparing a high-fat meal with a high-protein meal, matched for calories and carbohydrate | Peak glucose was higher after the high-fat meal (146 mg/dL) than the high-protein meal (137 mg/dL; p = 0.03) |
Two things stand out. The human studies are small, mostly pilots, and mostly short. And almost all point the same direction: putting carbohydrate last, or putting something slow-digesting first, tends to flatten the curve. That's consistent, but "consistent in small studies" isn't "proven in large trials".
Fat First Isn't a Free Pass: The Gestational Diabetes Twist
The most useful nuance in this month's data comes from the pregnancy trial. In it, the high-fat meal produced a higher peak than the high-protein meal, even though calories and carbohydrate were identical. The authors suggested that protein may boost insulin secretion more strongly, while a high-fat meal may promote some insulin resistance.
That sounds like it contradicts the Kyoto mice, but it doesn't. They're different questions. Kyoto asked what happens to the early rise when fat comes before a sugar load. The pregnancy trial compared the overall response to two different complete meals. Fat can slow stomach emptying and blunt an early spike while still contributing to a later, longer elevation. The lesson is that order is one lever and composition is another, and neither makes a high-carbohydrate meal harmless.
Fibre and Vegetables: The Preload With the Best Track Record
If you're going to pick one thing to eat first, the evidence is most practical for vegetables and fibre. A fibre preload physically thickens the stomach contents, which slows digestion and glucose absorption, as in the oat-bran pilot. Non-starchy vegetables also take up room, add chewing time and slow down the pace of eating. They carry few calories and almost no rapid carbohydrate. And they're easy, cheap and already recommended for other reasons.
This connects directly to our recent piece on how a structured high-fibre approach compared with metformin in newly diagnosed type 2 diabetes. Fibre works at several points: before the meal (viscosity), during digestion (slower absorption) and afterwards (gut bacteria and short-chain fatty acids). If you want one habit to try, a side salad or a bowl of vegetables at the start of your main meal ticks all three boxes.
A Practical Plate Order, With Realistic Expectations
- Start with non-starchy vegetables. A salad dressed with olive oil, a bowl of soup full of vegetables, or crudités. A dressing with olive oil adds the "fat first" element in a normal, food-shaped way.
- Move to protein. Fish, eggs, tofu, chicken, beans or yoghurt.
- Leave starch and sugar for last. Rice, pasta, bread, potato or dessert, eaten within the same 10–20 minutes, not hours later.
- Keep the portion of starch sensible. Order is a modifier, not a licence. Total carbohydrate still drives the size of the curve.
- Move afterwards if you can. A 10-minute walk after meals lowered blood sugar in a 2016 study in adults with type 2 diabetes (Reynolds et al., Diabetologia), and it stacks neatly with food order.
A realistic expectation: sequencing may shave the peak and make the curve gentler, not erase it. Think of it as a small, free upgrade to a meal you're already eating, not a fix for a diet that's heavy in refined carbohydrate.
One thing not to do: don't take shots of olive oil. The mouse study was about physiology, not a recipe, and a tablespoon of olive oil carries about 120 calories. If you add a dressing, count it. Our recipe nutrition calculator will show you what a salad with dressing really adds.
Who Should Be Careful With Meal Sequencing
| If you… | Take care because… |
|---|---|
| Take insulin or sulfonylureas | Changing meal patterns can change your glucose and your dose needs. Tell your care team before experimenting |
| Have gastroparesis or slow stomach emptying | Fat slows gastric emptying further, which could worsen symptoms. Get individual advice |
| Use GLP-1 medication | These already slow gastric emptying. Heavy fat early in a meal may add to nausea or fullness; ask your prescriber |
| Have reflux | Large amounts of fat early may worsen symptoms for some people |
| Have a history of disordered eating | Rigid rules about order and "forbidden" timing can feed unhelpful patterns |
| Are pregnant or have gestational diabetes | Individual targets matter; follow your obstetric and diabetes team's advice |
Where Macros and Calories Fit In
Meal sequencing is a timing trick inside a diet that still has to add up. Three practical points keep it honest:
- Total carbohydrate still matters most. Use the carb calculator or the macro calculator to set a sensible daily carbohydrate range for your goal, then use order to smooth the curve within it.
- Calories aren't free just because glucose is flat. Added fat early in a meal is energy-dense. If you're in a deficit, plan for it.
- Net carbs and fibre interact. If you eat low-carb, our net-carbs guide explains why fibre is subtracted and why vegetables-first is easy to fit in. For flexible approaches, see our macro calculator guide and our piece on carb cycling.
A Two-Week Test You Can Run on Your Own Plate
Population averages can't tell you how you respond. A short, honest self-experiment can, particularly if you have access to a glucose meter or sensor through your own healthcare.
- Pick one meal you eat often, such as a rice-and-chicken dinner or a pasta lunch.
- Eat it "normally" on day 1. If you have a meter or sensor, check your glucose before and about two hours after. Note hunger and energy.
- Eat the same meal, same portions, in sequence on day 3: vegetables, then protein, then starch within about 15 minutes. Repeat the readings.
- Repeat each version a couple of times to account for day-to-day noise: sleep, stress, activity and timing all move glucose.
- Judge by the pattern, not one reading. If sequencing gives you a smaller rise and you feel fuller, keep it. If nothing changes, drop it without guilt.
This isn't medical advice, and it isn't a substitute for your clinician's guidance if you have diabetes. It's a way to find out whether a plausible, low-cost habit does anything for you.
The Bottom Line
The Kyoto study is a neat piece of basic science that gives a plausible gut-hormone explanation for why eating fat, protein or vegetables before carbohydrate might soften the post-meal spike. It was done in mice, with a precisely timed olive-oil preload, so it can't tell you what to eat tonight. The small human studies agree in direction but not in scale or certainty, and a pregnancy trial this year reminds us that meal composition still matters alongside order.
If you want a sensible takeaway, it's gentle and easy: start meals with vegetables and protein, leave the starch for last, keep portions of refined carbohydrate in check, and take a short walk afterwards. No shots of oil, no rigid rules, and no panic if you eat the bread first at a dinner party.
Frequently Asked Questions
Does eating fat before carbs lower blood sugar?
In mice, olive oil given 15 minutes before glucose blunted the early rise in blood sugar and slowed stomach emptying. In people, small studies suggest that eating protein, vegetables or fat before carbohydrate can reduce the post-meal peak, but the studies are small and mostly short.
Why does meal order change blood sugar?
Fat and fibre slow the stomach's emptying and trigger gut hormones such as GLP-1, so glucose reaches the bloodstream more gradually. The Kyoto study found early release of GLP-1, GIP and glucagon, with vagus-nerve involvement, after a fat preload in mice.
Does the order work for people without diabetes?
Possibly. The Kyoto authors say their results support testing meal sequencing in people with and without diabetes. A smaller, gentler post-meal rise may still be a plus for general health, though long-term benefits haven't been shown.
Should I eat vegetables or protein first?
Vegetables first has the strongest practical support, since fibre thickens stomach contents and slows absorption. Protein next is sensible, and starch last. All three orderings tested in small studies tended to flatten the curve versus eating carbohydrate first.
Is olive oil before a meal a good idea?
Not as a ritual. The mouse study was about physiology, not a recommendation, and a tablespoon of olive oil is about 120 calories. A salad dressed with olive oil at the start of a meal is a more normal way to include fat early.
How long should I wait between courses?
The studies used about 15 minutes between the first food and the carbohydrate. In real life, eating in order within the same 10–20 minutes is a practical approximation. You don't need a stopwatch.
Does a fat-first meal lower the whole-meal glucose response?
Not necessarily. A 2026 pilot in gestational diabetes found a high-fat meal led to a higher peak than a high-protein meal with the same calories and carbohydrates. Order can soften an early spike, but overall composition still matters.
Can people on diabetes medication try meal sequencing?
Only with their care team's knowledge. If you use insulin or sulfonylureas, changes in meal pattern can change your glucose and dose needs.
Will this help me lose weight?
It isn't a weight-loss method by itself. Vegetables-first may help you feel fuller and eat a little less, but total calories and diet quality still drive weight. Use a calorie calculator to set your target.
References & Sources
- Olive Oil Before Oral Glucose Delays Gastric Emptying and Lowers Glycemia Through Partially Redundant Glucagon and GLP-1 Signaling in Mice (2026). Diabetes. doi:10.2337/db26-0531; Kyoto University news release (8 October 2026).
- ScienceDaily (11 October 2026), News-Medical (8 October 2026) and Scientific Frontline (9 October 2026) summaries of the Kyoto study.
- Shukla, A. P. et al. (2015). Food order has a significant impact on postprandial glucose and insulin levels. Diabetes Care, 38(7), e98–e99.
- Effect of consuming oat bran mixed in water before a meal on glycemic responses in healthy humans: a pilot study (2016). Nutrients, 8(9), 524.
- Rosenberg, J. et al. (2026). Impact of fat and protein on postprandial glycemia in gestational diabetes: a randomized pilot crossover trial. Pregnancy.
- ClinicalTrials.gov NCT06896799. Determining the effect of food ordering on blood glucose in gestational diabetes mellitus (DEFI-GDM).
- Reynolds, A. N. et al. (2016). Advice to walk after meals is more effective for lowering postprandial glycaemia in type 2 diabetes mellitus than advice that does not specify timing: a randomised crossover study. Diabetologia, 59(12), 2572–2578.
- MacrosMeasure (2026). Free Carb Calculator.
- MacrosMeasure (2026). Free Recipe Nutrition Calculator.
Calculate your own numbers — free.
18 science-backed calculators covering everything in this article and more.