Refined grains are said to spike blood sugar more than sugar itself. What did the study that built the glycemic index actually find?
The Short Answer
Because “refined” and “sugary” are not the same axis, and only one of them predicts anything. In the study that built the glycemic index, white bread scored 69 ± 5% and table sugar 59 ± 10%: bread at least as high as sugar, on 5-10 volunteers per food.[1] Milling is part of why: it destroys the physical structure that slows digestion. But the same study found fiber content did not predict a food’s glycemic response, and neither did sugar content.[1] Structure did.
Something to Sit With
How much of your daily eating is built on refined grains? Breakfast cereal, sandwich bread, pasta, crackers, pizza: these often form the foundation of modern meals.
What would change if you treated refined grains the way you treat sugar?
Learn More
The Glycemic Index Surprise.
The glycemic index does not measure how fast a food raises blood glucose. It measures the two-hour blood-glucose area under the curve, the total response, as a percentage of the same carbohydrate load taken as pure glucose.[1] A rate and an area are different things, and most of the folklore around this measure comes from reading one as the other.
Jenkins and colleagues built the index in 1981 by testing 62 foods on groups of 5-10 fasting volunteers. Three of their results are the ones worth carrying:[1]
- White bread 69 ± 5%, wholemeal bread 72 ± 6%. Wheat fiber did not reduce the glycemic response.
- Sucrose 59 ± 10%. Bread at least as high as table sugar, which surprises people who think of bread as a “complex carbohydrate” and sugar as the obvious villain.
- Wholemeal spaghetti 42 ± 4%, against wholemeal bread’s 72 ± 6%. Same grain, same fiber, different physical form.
Small samples with real error bars; the modern published tables give different numbers for specific products, and these figures come from the founding paper rather than those tables.
Why This Happens.
Starch is glucose chains. The starch in refined flour is long chains of glucose molecules. When you eat white bread, enzymes in your saliva and intestines rapidly break these chains into individual glucose molecules, which enter your bloodstream quickly.
Sugar is different. Table sugar (sucrose) is one glucose molecule bonded to one fructose molecule.[2] When you eat sugar, it splits into glucose (which spikes blood sugar) and fructose (which doesn’t directly spike blood sugar: it goes to the liver).[3] So only half of sugar’s calories produce an immediate glucose spike.
Refinement destroys structure. Whole grains have intact cell walls that slow digestion. Milling removes them, leaving starch that enzymes reach easily. Note the scope of what the founding study found: across its 62 foods, fiber content did not predict glycemic response at all, and neither did sugar content.[1] The spaghetti-versus-bread comparison is the clean demonstration: identical grain and fiber, half the response, purely because of form.
Processing compounds the problem. Many refined grain products are also puffed, flaked, or extruded, processes that further break down the starch structure. The founding study’s own spaghetti result is the evidence for this: physical form moved the response by half, with grain and fiber held constant.[1]
The Insulin Response.
When blood glucose rises steeply, the pancreas releases a correspondingly large insulin response.[4] Ludwig and Ebbeling describe this as a key driver in the carbohydrate-insulin model of obesity, a model whose underlying physiology is not disputed and whose evidence in humans they themselves call inconclusive:[4]
- Refined carbs cause rapid glucose spike
- Body releases large insulin response
- Insulin drives glucose into cells and promotes fat storage[4]
- Blood sugar crashes as insulin overshoots
- Crash triggers hunger and cravings
- Cycle repeats
Part of that pattern is measured, the later steps rather than the fat storage: after a high-glycemic meal blood sugar falls to a trough around four hours later, hunger rises with it, and reward-region activity rises too.[5]
What This Means Practically.
“Complex carbohydrates” is misleading. The term was meant to distinguish starchy foods from simple sugars, implying they were better. But sugar content did not predict a food’s glycemic response at all in the founding study: the simple/complex split predicts nothing.[1]
Whole-grain flour is not the fix people think it is. Wholemeal bread scored no lower than white in the founding study; across all 62 foods, fiber content did not predict a food’s glycemic response.[1] Read that carefully: it is a comparison between foods. Take the fiber out of a given meal and its glucose response does rise: the paper’s two areas under the curve are 226 against 160 mmol/min/dl, a difference of about 29% on those numbers rather than a percentage it reports.[6] The two findings are not in conflict, and what flour destroys is structure. Intact whole grains (steel-cut oats, whole barley) are a different proposition from flour products of any kind.
The real culprits are often invisible. Bread, pasta, crackers, cereals, pizza crust, tortillas: these refined grain products are staples of modern eating. People avoiding “sugar” may consume large amounts of foods whose glycemic response is no lower.
The Practical Takeaway.
If you’re trying to manage insulin and blood sugar:
- Refined grain products (white bread, white rice, many crackers and cereals) tend to produce large glucose responses
- Switching from sugar to refined grains doesn’t help: it may make things worse
- Intact whole grains are better than flour-based products
- Any grain product should be eaten with protein, fat, and fiber to slow absorption
The distinction isn’t “carbs vs. no carbs,” and it isn’t simple versus complex either, and that split predicts nothing. It is how much structure survived the processing.
Further reading:
- Ludwig DS & Ebbeling CB, “The Carbohydrate-Insulin Model of Obesity: Beyond ‘Calories In, Calories Out’”[7] — JAMA Internal Medicine, 2018
- Foster-Powell K et al., “International table of glycemic index and glycemic load values: 2002”[8] — American Journal of Clinical Nutrition, 2002
- Atkinson FS et al., “International tables of glycemic index and glycemic load values: 2008”[9] — Diabetes Care, 2008
- Fung J, The Obesity Code[10] (2016) — Chapter on refined carbohydrates
Written by AI, directed and approved by me. How this is made →