The problem was never "carbs" — it's fiber, processing, and portion#
There is no moral category of "bad carbs." The same gram of carbohydrate can arrive wrapped in fiber, water, and a whole-food structure, or stripped down to fast-digesting starch in a hyper-palatable package — and those two behave very differently in your body and your appetite. But the thing that differs is not "carbohydrate." It is how much fiber rides along, how processed the food is, and how easy it is to overeat. Sort carbs by those, and the old labels — simple versus complex, good versus bad — turn out to be a bad map.
Say the settled part plainly, because it is settled: whole, minimally processed carbohydrate foods — vegetables, fruit, legumes, whole grains — are better for long-term health than refined and ultra-processed ones. That is not in serious dispute. What this article does is replace the moralizing binary with the two or three lines that actually predict how a carbohydrate food treats you, and explain why a bowl of lentils and a can of soda deserve different treatment without either being a sin. The question of how many carbs to eat is separate and lives in how many carbs per day; this is about which.
Why "simple vs complex" was always a bad map#
The oldest version of the good/bad split is "simple" sugars (bad) versus "complex" carbohydrates (good), on the logic that longer starch molecules digest slowly and sugars digest fast. It falls apart on contact with real food. Plenty of "complex" starches — white bread, cornflakes, instant mashed potato — are digested and spike blood glucose about as fast as table sugar, because refining and cooking have already done the work your gut would have done. Meanwhile plenty of "simple" sugar foods — an apple, a carrot, plain milk — come with fiber, water, and nutrients that slow digestion and blunt the spike. The molecule is not the point; the matrix around it is.
This is exactly the gap the glycemic index was invented to close — a way to rank carbohydrate foods by how much they actually raise blood glucose, rather than by the shape of their molecules. It is a real improvement over "simple vs complex." It is also, on its own, not the answer either.
Glycemic index vs glycemic load — and why neither is the finish line#
Two terms get used interchangeably and shouldn't be. Glycemic index (GI) ranks a food by how much 50 grams of its digestible carbohydrate raise blood glucose against a reference. Its weakness is that it ignores portion: watermelon has a high GI, but a normal serving contains so little carbohydrate that its real-world blood-sugar effect is small. Glycemic load (GL) fixes that by multiplying GI by the carbohydrate actually in a serving (GI × grams ÷ 100), so it reflects what a real portion does. GL is the more useful of the two, and where they disagree, trust GL.
But here is the part the glycemic-index industry skips. When the big Lancet review weighed diets by glycemic index or load against health outcomes, it found smaller or no risk reductions — and graded the certainty of that evidence as low to very low, against moderate for dietary fiber and whole grains1. In plain terms: glycemic load beats glycemic index, but both are weak predictors of health next to a far simpler signal — how much fiber a food carries and how whole it is. If you were going to track one thing, it would be fiber, not GI (fiber: how much and why it matters). The fuller GI story has its own explainer; the headline is that it is a second-tier tool.
The two lines that actually matter: whole vs refined, and processing#
If not GI, then what? Two axes carry most of the signal.
The first is whole versus refined. Whole grains keep the bran and germ — and with them the fiber, minerals, and structure — that refining strips out. The payoff is well quantified: in a dose-response meta-analysis of 45 prospective studies, each additional 90 grams of whole grains a day (about three servings) tracked with a 22% lower risk of cardiovascular disease, a 19% lower risk of coronary heart disease, and a 17% lower risk of dying from any cause, with benefits continuing up to around 210–225 grams a day2. Refined grains keep little of that structure and little of the fiber.
The second axis is processing — and this is the one with the sharpest evidence, because it comes from a controlled trial rather than observation. The NOVA system sorts foods by how industrially processed they are, with "ultra-processed" foods (group 4) defined by industrial ingredients and cosmetic additives you would never find in a kitchen — the sodas, packaged snacks, and reconstituted starches that most "bad carbs" actually are3. To test whether processing itself matters, researchers fed 20 adults an ultra-processed and an unprocessed diet for two weeks each, matched for calories offered, macronutrients, sugar, sodium, and fiber, and let them eat as much as they liked. On the ultra-processed diet people ate 508 more calories a day and gained about a kilogram; on the unprocessed one they lost about the same4.
Match two diets for sugar, fiber, and every macro, and people still ate 500 calories a day more of the ultra-processed one. The processing was the difference — not the carbohydrate.
That is the cleanest evidence there is that the villain in a "bad carb" is usually its processing, not its carbohydrate. Ultra-processed foods are engineered to be eaten fast and in quantity, and that — not the sugar molecule as such — is what makes them easy to overeat.
So, is there such a thing as a bad carb?#
No — but there is a useful spectrum, and it runs along fiber and processing, not virtue. The reframe is the whole point:
| Stop asking | Ask instead |
|---|---|
| Is this carb "good" or "bad"? | How much fiber does it carry, and how processed is it? |
| Is it simple or complex? | Is it a whole food or ultra-processed? |
| Does it have a high glycemic index? | Would a normal portion be easy to overdo? |
Run any carbohydrate food through those questions and it sorts itself without moralizing. Lentils, oats, fruit, and vegetables land on the roomy end: high fiber, minimally processed, hard to overeat. Soda, white bread, and packaged snacks land on the other: low fiber, ultra-processed, easy to inhale. And context still matters — a refined carbohydrate like white rice around a hard workout is genuinely useful fuel, not a moral failure, which is where how many carbs per day meets this. The move that does the most good is not cutting carbs; it is shifting them toward whole and minimally processed, one swap at a time — the kind of durable change covered in sustainable weight-loss habits. The same food can even carry very different calories depending on how it is made, its own rabbit hole in calories in common foods.
FAQ#
Are carbs bad for you?#
No. Carbohydrate is a category of nutrient, not a health verdict, and the acceptable range runs to 45–65% of calories (macronutrients explained). What varies is quality: whole, high-fiber, minimally processed carbohydrate foods are consistently linked to better health, while refined and ultra-processed ones are easy to overeat and carry little fiber. The problem is never "carbs" as such — it is processing and fiber.
Is glycemic index a good way to choose carbs?#
It is a second-tier tool. Glycemic load (which accounts for portion) is more useful than glycemic index (which doesn't), but a major Lancet review found both to be weak predictors of health outcomes compared with dietary fiber and whole-grain intake1. Choosing whole, high-fiber foods captures most of the benefit without a GI chart.
Is the sugar in fruit bad because it's a "simple" carb?#
No. Fruit's sugar arrives packaged with fiber, water, and nutrients that slow its digestion, which is why whole fruit behaves nothing like soda despite both containing sugar. The "simple vs complex" label misleads here: what matters is the whole package, not the sugar molecule. Whole fruit is one of the easiest fiber upgrades you can make.
Sources#
- Reynolds A, Mann J, Cummings J, et al. Carbohydrate quality and human health: a series of systematic reviews and meta-analyses. Lancet. 2019;393(10170):434-445.
- Aune D, Keum N, Giovannucci E, et al. Whole grain consumption and risk of cardiovascular disease, cancer, and all cause and cause specific mortality: systematic review and dose-response meta-analysis. BMJ. 2016;353:i2716.
- Monteiro CA, Cannon G, Levy RB, et al. Ultra-processed foods: what they are and how to identify them. Public Health Nutr. 2019;22(5):936-941.
- Hall KD, Ayuketah A, Brychta R, et al. Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake. Cell Metab. 2019;30(1):67-77.
