Glycemic Index & Glycemic Load: The Complete A-to-Z Guide

In 1981, Dr. David Jenkins and his colleagues at the University of Toronto published a paper in the American Journal of Clinical Nutrition that changed how nutritionists thought about carbohydrates (Jenkins, David J. A. and Wolever, Thomas M. S. and Taylor, Richard H. and Barker, Helen and Fielden, Hasib and Baldwin, Janet M. and Bowling, Ann C. and Newman, Helen C. and Jenkins, Alexandra L. and Goff, David V., 1981).

At the time, the conventional wisdom was simple: sugars are fast, starches are slow. Sweet foods spike blood sugar; starchy foods are "complex" and therefore safer.

Jenkins and his team tested this assumption directly. They fed test subjects different carbohydrate-containing foods and measured blood glucose responses over two hours. The results dismantled the conventional model.

Bread — a starch, supposedly "complex" — spiked blood glucose nearly as fast as pure glucose. Lentils — also a starch — barely moved the needle. Fructose (fruit sugar) had a surprisingly low blood glucose response.

The old categories of "simple" and "complex" carbohydrates were physiologically meaningless. What mattered was the measured response in the body — not the chemical classification of the food.

Jenkins called this measure the Glycemic Index.

The Glycemic Index is not a chemical property of food. It is a biological measurement of human response (Wolever, Thomas M. S. and Jenkins, David J. A. and Jenkins, Alexandra L. and Josse, Robert G., 1991).

To measure the GI of a food, researchers follow a standardized protocol:

1. A group of at least 10 healthy, fasted subjects consume a portion of the test food containing exactly 50 grams of available carbohydrate
2. Blood glucose is measured at regular intervals over two hours
3. The area under the blood glucose curve is calculated
4. This is compared to the blood glucose response to 50 grams of pure glucose (the reference food, set at 100)

The result is expressed as a percentage of the glucose response. A food with GI 50 produces half the blood glucose rise of pure glucose for the same amount of carbohydrate.

This is why GI values include confidence intervals — they are averages across multiple people, and individual responses vary. The methodology is well-established and internationally standardized (Augustin, Livia S. A. and Kendall, Cyril W. C. and Jenkins, David J. A. and Willett, Walter C. and Astrup, Arne and Barclay, Alan W. and Björck, Inger and Brand-Miller, Janette C. and Brighenti, Furio and Buyken, Anette E. and Ceriello, Antonio and La Vecchia, Carlo and Livesey, Geoffrey and Liu, Simin and Riccardi, Gabriele and Rizkalla, Salwa W. and Sievenpiper, John L. and Trichopoulou, Antonia and Wolever, Thomas M. S. and Baer-Sinnott, Sylvia and Poli, Andrea, 2015).

Category	GI Range	What It Means
Low GI	55 or below	Slow glucose release, sustained energy, lower insulin demand
Medium GI	56 to 69	Moderate response, context-dependent
High GI	70 and above	Fast glucose spike, rapid insulin response, potential crash

The reference point (glucose = 100) is sometimes replaced with white bread (also = 100) in older literature. This creates confusion when comparing values across studies. The modern standard uses glucose as the reference (Atkinson, Fiona S. and Foster-Powell, Kaye and Brand-Miller, Janette C., 2008).

GI tells you the speed of glucose release per gram of carbohydrate. It does not tell you how much carbohydrate is in a realistic portion.

This distinction matters enormously. Consider watermelon.

Watermelon has a GI of 76 — high GI, on paper alarming.

But a standard serving of watermelon is mostly water. That serving contains only about 6 grams of available carbohydrate. The actual blood glucose impact is minimal.

Meanwhile, white pasta has a GI of 49 — technically low GI, seemingly safe. But a typical pasta serving contains 40-50 grams of carbohydrate. The total blood glucose impact is substantial.

GI without portion context is half the picture.

Glycemic Load was introduced to correct GI's limitation by incorporating portion size (Salmeron, Jorge and Manson, JoAnn E. and Stampfer, Meir J. and Colditz, Graham A. and Wing, Alicia L. and Willett, Walter C., 1997).

The formula is straightforward:

\[ GL = \frac{GI \times \text{grams of available carbohydrate per serving}}{100} \]

Available carbohydrate = total carbohydrate minus fiber. Fiber does not raise blood glucose — it is subtracted from the calculation.

Category	GL Range	Practical Meaning
Low GL	10 or below	Minimal blood glucose impact
Medium GL	11 to 19	Moderate impact
High GL	20 and above	Significant blood glucose impact

Daily total GL targets (for reference, not prescription):

• Low GL diet: under 80 per day
• Medium GL diet: 80-120 per day
• High GL diet: over 120 per day

Food	GI	Serving (g)	Available Carbs (g)	GL	Verdict
Watermelon	76	150	6	5	Low GL despite high GI
White pasta	49	200 cooked	40	20	High GL despite low GI
Sweet potato (boiled)	63	150	18	11	Medium GL
Lentils	32	150	18	6	Low GL
White rice	73	150	40	29	High GL
Apple	36	120	14	5	Low GL
Dates (3 pieces)	42	24	16	7	Low GL in small portion

The takeaway: use GL, not GI, to make food decisions. GI is the foundation. GL is the application.

Understanding this mechanism explains every GI/GL recommendation that follows (Ludwig, David S., 2002).

When you eat carbohydrates, your digestive system breaks them down to glucose, which enters the bloodstream. Your pancreas responds by releasing insulin — the hormone that moves glucose from the blood into cells (muscle, liver, fat) for use or storage.

High-GL meal:

Large glucose spike → large insulin surge → rapid glucose clearance
→ blood glucose drops below baseline → hunger signal fires
→ you eat again → caloric surplus → fat storage

Low-GL meal:

Gradual glucose rise → measured insulin response → stable blood glucose
→ satiety hormones active for longer → natural hunger signal
→ you eat when genuinely hungry

The physiological difference between these two cycles is not minor. The hunger triggered by a high-GL meal is not a failure of willpower. It is a hormonal signal generated by your own biology in response to the food you ate.

High-GL eating patterns do not make you fat because you lack discipline. They make you fat because they make you genuinely hungry more often than your body actually needs food (Brand-Miller, Janette C. and Holt, Susanna H. A. and Pawlak, David B. and McMillan, Joanna, 2002).

Worth noting: different foods trigger different insulin responses even with similar blood glucose profiles. Dairy, for example, has a low GI but a surprisingly high insulin demand (Holt, Susanna H. A. and Miller, Janette C. Brand and Petocz, Peter, 1997).

For most people eating mixed meals, the GI and GL framework provides sufficient guidance. The insulin index is useful for those managing insulin resistance or diabetes specifically.

Heat and water disrupt the starch granule structure, increasing digestibility and GI.

Food	Raw/Minimal	Cooked/Processed	GI Change
Sweet potato	~50 boiled	~94 baked	+44
Pasta	~40 al dente	~55 overcooked	+15
Oats	~42 steel cut	~79 instant	+37
Potato	~78 boiled	~87 mashed	+9

Practical rule: Shorter cooking time = lower GI. Al dente pasta, boiled rather than baked sweet potato, steel cut rather than instant oats.

When cooked starchy foods are cooled, a portion of the digestible starch converts to resistant starch — a form that behaves like fiber and is not broken down in the small intestine (Higgins, Janine A., 2004).

Food	Hot (freshly cooked)	Cooled and reheated	GI Reduction
White rice	~73	~53	~27% lower
Potato	~78	~56	~28% lower
Pasta	~49	~43	~12% lower

Practical rule: Rice and potato cooked the day before and reheated have meaningfully lower GI than freshly cooked hot servings. Meal prep has metabolic benefits beyond convenience.

The finer a grain is ground, the more surface area is exposed to digestive enzymes, and the faster glucose is released.

• Whole intact grain: lowest GI
• Cracked grain: moderate GI
• Stone-ground flour: moderate GI
• Finely milled white flour: highest GI

This is why "whole wheat bread" has nearly the same GI as white bread (74 versus 75) (Atkinson, Fiona S. and Foster-Powell, Kaye and Brand-Miller, Janette C., 2008). The grain is whole, but it has been ground into flour — particle size is the dominant variable, not the grain's original fiber content.

Intact grains (steel-cut oats, whole barley, bulgur, freekeh) have substantially lower GI than their flour equivalents.

As fruit ripens, starch converts to sugar, raising GI.

Fruit	GI (unripe/firm)	GI (ripe/soft)
Banana	~30 (green)	~62 (ripe yellow)
Mango	~41 (firm)	~60 (soft)

Practical rule: Buy fruit slightly underripe and eat it before it softens fully if managing blood glucose is a priority.

Soluble fiber — found in oats, legumes, apples, and psyllium — forms a viscous gel in the digestive tract that slows glucose absorption.

This is the primary reason legumes have such low GI despite being high in carbohydrates. It is also why rolled oats (GI ~55) behaves very differently from rice (GI ~73) despite both being grains.

High-fiber foods slow everything down. This is physiologically protective and satiating.

Fat and protein slow gastric emptying — the rate at which food leaves your stomach and enters your small intestine (Wolever, Thomas M. S. and Jenkins, David J. A., 1986).

A carbohydrate eaten alone spikes faster than the same carbohydrate eaten with protein and fat.

Scenario	GI Impact
White bread alone	High
White bread + olive oil	Moderate
White bread + egg + olive oil	Lower still
Hummus on bread	Lower than bread alone

Practical rule: Never eat high-GI carbs in isolation. Adding fat and protein to any meal reduces the effective glycemic response. A date eaten alone hits differently than a date eaten after a meal containing protein and fat.

Acidic foods slow gastric emptying and starch digestion, reducing GI by up to 30-35%.

• Sourdough bread (long-fermented, high organic acid content): GI ~54 versus white bread GI ~75
• Adding vinegar dressing to a salad reduces the GL of the carbs eaten in the same meal
• Lemon juice on rice or grains measurably reduces glycemic response

Practical rule: Lemon on salad is not just flavor. The acidity slows glucose release from everything eaten alongside it.

Starch exists in two forms:

• Amylose: compact, tightly packed structure — digested slowly, lower GI
• Amylopectin: branched, open structure — rapidly digested, higher GI

Basmati rice has more amylose than standard white rice — which is why it has a GI of ~58 compared to ~73 for standard white rice. Waxy rice varieties (glutinous/sticky rice) are almost pure amylopectin — very high GI (~98).

This factor is largely outside your control as a consumer, but it explains why "rice is high GI" is not a universal statement.

Food	GI	Serving (g)	GL	Category
Cornflakes	81	30	20	High
White rice (boiled)	73	150	29	High
White bread	75	30	11	High
Instant oats	79	250 cooked	21	High
Whole wheat bread	74	30	9	High
Pita bread (white)	68	30	10	Medium
Couscous	65	150	23	Medium
Brown rice	68	150	16	Medium
Sourdough (long fermented)	54	30	8	Low
Rolled oats (cooked)	55	250	13	Low
Quinoa	53	150	13	Low
Pasta, white (al dente)	49	180	24	Low
Buckwheat	51	150	16	Low
Steel cut oats	42	250	11	Low
Bulgur wheat (boiled)	47	150	12	Low
Freekeh (boiled)	43*	150	11*	Low
Rye bread (whole grain)	41	30	8	Low
Barley (boiled)	28	150	9	Low

Food	GI	Serving (g)	GL	Category
Broad beans (fava, boiled)	40*	150	9*	Low
Green peas	51	80	4	Low
Chickpeas (boiled)	28	150	8	Low
Lentils, green (boiled)	32	150	5	Low
Lentils, red (boiled)	26	150	5	Low
Black beans	30	150	7	Low
Kidney beans	24	150	7	Low
Soybeans	16	150	1	Low
Hummus	6	30	0	Low

Food	GI	Serving (g)	GL	Category
White potato (baked)	85	150	26	High
White potato (mashed)	87	150	17	High
White potato (boiled)	78	150	15	High
Pumpkin (boiled)	75	80	3	High GI, Low GL
Beetroot (boiled)	64	80	5	Medium GI, Low GL
Sweet potato (boiled)	63	150	11	Medium
Sweet potato (baked)	94	150	17	High
Sweet corn (boiled)	52	80	9	Low
Carrots (boiled)	39	80	2	Low
Non-starchy vegetables	<15	Unlimited	<1	Low

Non-starchy vegetables include: broccoli, cucumber, tomato, lettuce, spinach, eggplant, zucchini, peppers, onion, garlic, mushrooms, cabbage, asparagus.

Food	GI	Serving (g)	GL	Category
Watermelon	76	150	5	High GI, Low GL
Banana (ripe)	62	120	16	Medium
Grapes	59	120	11	Medium
Mango	51	120	8	Low
Kiwi	53	120	6	Low
Orange	43	120	5	Low
Dates (dried)	42	60	18	Low GI, Medium GL
Apple	36	120	6	Low
Pear	38	120	4	Low
Strawberries	40	150	1	Low
Blueberries	34	150	6	Low
Cherries	22	120	3	Low
Grapefruit	25	120	3	Low

Note on dates: Per 60g (roughly 4-5 dates), GL reaches medium range. In small amounts (2-3 dates), GL stays low. Traditional portions are well within the safe range (Atkinson, Fiona S. and Foster-Powell, Kaye and Brand-Miller, Janette C., 2008).

Food	GI	Serving (g/ml)	GL	Category
Full fat milk	39	250	5	Low
Skim milk	37	250	4	Low
Plain yogurt (full fat)	35	200	3	Low
Labneh	~30*	100	~2*	Low
Ice cream (full fat)	57	50	6	Medium GI, Low GL

Food	GI	GL	Note
All meats (chicken, lamb, beef, fish)	0	0	No carbohydrates
Eggs	0	0	No carbohydrates
Olive oil	0	0	Pure fat
Nuts (almonds, walnuts)	0-15	0-1	Negligible carbs
Cheese (all types)	0	0	Negligible carbs

Flour	GI	Practical Use
White wheat flour	~85	Reference — what you are replacing
Almond flour	~5	Baked goods — very crumbly, needs binder
Chickpea flour (besan)	~39	Savory dishes, flatbreads, pancakes
Oat flour	~55	Closest texture to regular flour
Whole spelt flour	~52	Lower than wheat, handles similarly
Almond + oat flour (50/50)	~30	Best balance of texture and GI

Before eating a carbohydrate, replace the question "Is this healthy?" with:

"Will this spike me and leave me hungry in two hours, or will it release slowly and hold me?"

This is a more honest question, and it points directly to GL.

Build your personal food list in three categories:

Green Light — Eat Freely: Foods with low GL per realistic serving. These are your foundation.

• All non-starchy vegetables
• Legumes (lentils, chickpeas, foul, beans, hummus)
• Intact whole grains (steel cut oats, bulgur, barley)
• Berries, citrus, apples
• Full-fat dairy (yogurt, labneh)
• All proteins and healthy fats

Yellow Light — Portion Awareness: Foods with moderate GL — fine in controlled portions, problematic in excess.

• Sweet potato (boiled, not baked)
• Rolled oats (not instant)
• Whole fruit (banana, mango, grapes)
• Brown rice
• Sourdough bread (genuine, long-fermented)
• Dates (2-3 maximum)

Red Light — Minimize or Remove: Foods with high GL in typical servings.

• White rice
• White bread, pita, wraps
• White potato (all preparations)
• Instant oats
• Pasta in large portions
• Sugary foods, sweetened drinks
• Refined flour products (pastries, crackers, cereals)

Before writing off or green-lighting any food, apply the 8 modifiers from Part 5:

• Is it cooked al dente or overcooked?
• Was it cooled and reheated? (Lower GI)
• Is it intact grain or ground into flour? (Intact = lower)
• Is it ripe or firm? (Firm = lower for fruit)
• What is it eaten with? (Fat and protein lower the response)
• Is there acid in the meal? (Vinegar or lemon reduces GL)

A bowl of rice eaten hot and alone is a different metabolic event than the same rice eaten cold and reheated alongside salmon, olive oil, and a lemon-dressed salad.

GI and GL values are population averages. Individual responses vary significantly based on gut microbiome composition, insulin sensitivity, and personal physiology (Augustin, Livia S. A. and Kendall, Cyril W. C. and Jenkins, David J. A. and Willett, Walter C. and Astrup, Arne and Barclay, Alan W. and Björck, Inger and Brand-Miller, Janette C. and Brighenti, Furio and Buyken, Anette E. and Ceriello, Antonio and La Vecchia, Carlo and Livesey, Geoffrey and Liu, Simin and Riccardi, Gabriele and Rizkalla, Salwa W. and Sievenpiper, John L. and Trichopoulou, Antonia and Wolever, Thomas M. S. and Baer-Sinnott, Sylvia and Poli, Andrea, 2015).

The only way to know your personal response is to test it.

2-Week Personal Test Protocol:

Week 1 — Baseline: Eat only green-light carbohydrates for 7 days. No bread, rice, pasta, sugar. Track daily:

• Energy level (1-10)
• Hunger stability (1-10)
• Morning weight
• Training performance (if applicable)

Week 2 — Reintroduce one food at a time: Add one yellow or red light food for 3 days. Same tracking.

• Did hunger stability drop?
• Did energy crash?
• Did weight jump?

The data tells you where you personally sit on the spectrum. Some people handle rice without issue. Some people respond sharply to it. The numbers guide you to the test. The test gives you the personal answer.

The goal is not a GI-obsessed eating protocol that consumes mental energy at every meal. The goal is a set of default food choices that are naturally low GL and require no calculation at mealtime.

Once your defaults are set — what you eat on a regular Tuesday, your go-to meal structures, your grocery list staples — you can stop thinking about numbers entirely. The numbers got you to the habits. The habits carry you forward.

No. Whole wheat bread has a GI of 74. White bread is 75. Nearly identical.

The fiber in whole wheat is real and provides some benefit, but the grain has been ground into flour — particle size dominates the glycemic response. Intact whole grains (bulgur, barley, steel cut oats) are genuinely low GI. Whole wheat flour products are not.

Most fruit is low GL. The fiber, water content, and micronutrient profile of whole fruit makes it a fundamentally different metabolic event from drinking fruit juice or eating refined sugar.

Fruit juice removes the fiber and concentrates the sugar. Whole fruit with intact fiber is almost universally low GL at normal serving sizes.

The glycemic difference between lentils (GI 32, GL 5) and white bread (GI 75, GL 11) is not minor. These foods produce dramatically different biological responses even at equal carbohydrate quantities. Treating all carbs as equivalent ignores decades of research.

GL is dose-dependent. A low-GI food consumed in very large quantities can still produce a high GL. Lentils at GL 5 per serving become GL 20 at four servings. Volume matters.

It is not. Calories, protein intake, micronutrient density, food quality, sleep, stress, and training all influence body composition and metabolic health. GI and GL are one lens among several. They are a particularly useful lens for carbohydrate selection, but they do not replace other fundamentals.

The same food can produce different blood glucose responses in different people — sometimes dramatically so. Research shows that individual variation in glycemic response can be greater than the variation between different foods (Augustin, Livia S. A. and Kendall, Cyril W. C. and Jenkins, David J. A. and Willett, Walter C. and Astrup, Arne and Barclay, Alan W. and Björck, Inger and Brand-Miller, Janette C. and Brighenti, Furio and Buyken, Anette E. and Ceriello, Antonio and La Vecchia, Carlo and Livesey, Geoffrey and Liu, Simin and Riccardi, Gabriele and Rizkalla, Salwa W. and Sievenpiper, John L. and Trichopoulou, Antonia and Wolever, Thomas M. S. and Baer-Sinnott, Sylvia and Poli, Andrea, 2015).

This is why Step 4 (personal experiment) matters. The tables give you the population average. Your body gives you the personal answer.

GI and GL values are measured with foods in isolation, under fasted conditions, in standardized portions. Real meals are mixed — they contain protein, fat, fiber, acid, and multiple carbohydrate sources simultaneously.

This context reliably lowers glycemic response compared to isolated food testing. The tables are conservative guides, not exact predictions for your mixed iftar plate.

The same food category can have significantly different GI depending on brand, variety, and preparation. Pasta GI ranges from ~40 to ~60 depending on grain variety and cooking time. White rice ranges from ~64 to ~93 depending on variety and preparation method.

Use table values as directional guidance, not precise measurements.

GI and GL are well-supported tools for blood glucose management and short-term food choice (Livesey, Geoffrey and Taylor, Richard and Livesey, Helen F. and Buyken, Anette E. and Jenkins, David J. A. and Augustin, Livia S. A. and Barclay, Alan W. and Liu, Simin and Wolever, Thomas M. S. and Willett, Walter C. and Brighenti, Furio and Salas-Salvadó, Jordi and Björck, Inger and Rizkalla, Salwa W. and Riccardi, Gabriele and La Vecchia, Carlo and Ceriello, Antonio and Trichopoulou, Antonia and Poli, Andrea and Kendall, Cyril W. C. and Ha, Mary Ann and Baer-Sinnott, Sylvia and Brand-Miller, Janette C., 2019). The evidence for long-term cardiovascular and weight outcomes is positive but more variable across populations. The framework is reliable for its primary purpose — guiding carbohydrate choices — and should be understood as that.