The glycemic index or glycaemic index (GI) is a number associated with a particular type of food that indicates the food's effect on a person's blood glucose (also called blood sugar) level. The number typically ranges between 50 and 100, where 100 represents the standard, an equivalent amount of pure glucose.
The GI represents the total rise in a person's blood sugar level following consumption of the food; it may or may not represent the rapidity of the rise in blood sugar. The steepness of the rise can be influenced by a number of other factors, such as the quantity of fat eaten with the food. The GI is useful for understanding how the body breaks down carbohydrates  and only takes into account the available carbohydrate (total carbohydrate minus fiber) in a food. Although the food may contain fats and other components that contribute to the total rise in blood sugar, these effects are not reflected in the GI.
The glycemic index is usually applied in the context of the quantity of the food and the amount of carbohydrate in the food that is actually consumed. A related measure, the glycemic load (GL), factors this in by multiplying the glycemic index of the food in question by the carbohydrate content of the actual serving. Watermelon has a high glycemic index, but a low glycemic load for the quantity typically consumed. Fructose, by contrast, has a low glycemic index, but can have a high glycemic load if a large quantity is consumed.
GI tables are available that list many types of foods and their GIs. Some tables also include the serving size and the glycemic load of the food per serving.
A practical limitation of the glycemic index is that it does not measure insulin production due to rises in blood sugar. As a result, two foods could have the same glycemic index, but produce different amounts of insulin. Likewise, two foods could have the same glycemic load, but cause different insulin responses. Furthermore, both the glycemic index and glycemic load measurements are defined by the carbohydrate content of food. For example when eating steak, which has no carbohydrate content but provides a high protein intake, up to 50% of that protein can be converted to glucose when there is little to no carbohydrate consumed with it. But because it contains no carbohydrate itself, steak cannot have a glycemic index. For some food comparisons, the "insulin index" may be more useful.
Glycemic index charts often give only one value per food, but variations are possible due to variety, ripeness (riper fruits contain more sugars increasing GI), cooking methods (the more cooked, or over cooked, a food the more its cellular structure is broken with a tendency for it to digest quickly and raise GI more), processing (e.g., flour has a higher GI than the whole grain from which it is ground as grinding breaks the grain's protective layers) and the length of storage. Potatoes are a notable example, ranging from moderate to very high GI even within the same variety.
Most of the values on the glycemic index do not show the impact on glucose levels after two hours. Some people with diabetes may have elevated levels after four hours.
Determining the GI of a food
Foods with carbohydrates that break down quickly during digestion and release glucose rapidly into the bloodstream tend to have a high GI; foods with carbohydrates that break down more slowly, releasing glucose more gradually into the bloodstream, tend to have a low GI. The concept was developed by Dr. David J. Jenkins and colleagues in 1980–1981 at the University of Toronto in their research to find out which foods were best for people with diabetes. A lower glycemic index suggests slower rates of digestion and absorption of the foods' carbohydrates and may also indicate greater extraction from the liver and periphery of the products of carbohydrate digestion. A lower glycemic response usually equates to a lower insulin demand but not always, and may improve long-term blood glucose control and blood lipids. The insulin index is also useful for providing a direct measure of the insulin response to a food.
The glycemic index of a food is defined as the incremental area under the two-hour blood glucose response curve (AUC) following a 12-hour fast and ingestion of a food with a certain quantity of available carbohydrate (usually 50 g). The AUC of the test food is divided by the AUC of the standard (either glucose or white bread, giving two different definitions) and multiplied by 100. The average GI value is calculated from data collected in 10 human subjects. Both the standard and test food must contain an equal amount of available carbohydrate. The result gives a relative ranking for each tested food.
The current validated methods use glucose as the reference food, giving it a glycemic index value of 100 by definition. This has the advantages of being universal and producing maximum GI values of approximately 100. White bread can also be used as a reference food, giving a different set of GI values (if white bread = 100, then glucose ≈ 140). For people whose staple carbohydrate source is white bread, this has the advantage of conveying directly whether replacement of the dietary staple with a different food would result in faster or slower blood glucose response. A disadvantage with this system is that the reference food is not well-defined.
GI values can be interpreted intuitively as percentages on an absolute scale and are commonly interpreted as follows:
|Low GI||55 or less||beans (white, black, pink, kidney, lentil, soy, almond, peanut, walnut, chickpea); small seeds (sunflower, flax, pumpkin, poppy, sesame, hemp); most whole intact grains (durum/spelt/kamut wheat, millet, oat, rye, rice, barley); most vegetables, most sweet fruits (peaches, strawberries, mangos); tagatose; fructose; mushrooms; chilis|
|Medium GI||56–69||not intact whole wheat or enriched wheat, pita bread, basmati rice, unpeeled boiled potato, grape juice, raisins, prunes, pumpernickel bread, cranberry juice, regular ice cream, sucrose, banana|
|High GI||70 and above||white bread (only wheat endosperm), most white rice (only rice endosperm), corn flakes, extruded breakfast cereals, glucose, maltose, maltodextrins, potato, pretzels, bagels|
A low-GI food will release glucose more slowly and steadily, which leads to more suitable postprandial (after meal) blood glucose readings. A high-GI food causes a more rapid rise in blood glucose levels and is suitable for energy recovery after exercise or for a person experiencing hypoglycemia.
The glycemic effect of foods depends on a number of factors, such as the type of starch (amylose versus amylopectin), physical entrapment of the starch molecules within the food, fat and protein content of the food and organic acids or their salts in the meal — adding vinegar, for example, will lower the GI. The presence of fat or soluble dietary fiber can slow the gastric emptying rate, thus lowering the GI. In general, coarse, grainy breads with higher amounts of fiber have a lower GI value than white breads. However, most breads made with 100% whole wheat or wholemeal flour have a GI not very different from endosperm only (white) bread. Many brown breads are treated with enzymes to soften the crust, which makes the starch more accessible (high GI).
While adding fat or protein will lower the glycemic response to a meal, the relative differences remain. That is, with or without additions, there is still a higher blood glucose curve after a high-GI bread than after a low-GI bread such as pumpernickel.
Fruits and vegetables tend to have a low glycemic index. The glycemic index can be applied only to foods where the test relies on subjects consuming an amount of food containing 50 g of available carbohydrate. But many fruits and vegetables (not potatoes, sweet potatoes, corn) contain less than 50 g of available carbohydrate per typical serving. Carrots were originally and incorrectly reported as having a high GI. Alcoholic beverages have been reported to have low GI values; however, beer was initially reported to have a moderate GI due to the presence of maltose. This has been refuted by brewing industry professionals, who say that all maltose sugar is consumed in the brewing process and that packaged beer has little to no maltose present. Recent studies have shown that the consumption of an alcoholic drink prior to a meal reduces the GI of the meal by approximately 15%. Moderate alcohol consumption more than 12 hours prior to a test does not affect the GI.
Many modern diets rely on the glycemic index, including the South Beach Diet, Transitions by Market America and NutriSystem Nourish Diet. However, others have pointed out that foods generally considered to be unhealthy can have a low glycemic index, for instance, chocolate cake (GI 38), ice cream (37), or pure fructose (19), whereas foods like potatoes and rice have GIs around 100 but are commonly eaten in some countries with low rates of diabetes.
The GI Symbol Program is an independent worldwide GI certification program that helps consumers identify low-GI foods and drinks. The symbol is only on foods or beverages that have had their GI values tested according to standard and meet the GI Foundation's certification criteria as a healthy choice within their food group, so they are also lower in kilojoules, fat and/or salt.
Recent animal research provides compelling evidence that high-GI carbohydrate is associated with increased risk of obesity. In one study, male rats were split into high- and low-GI groups over 18 weeks while mean body weight was maintained. Rats fed the high-GI diet were 71% fatter and had 8% less lean body mass than the low-GI group. Postmeal glycemia and insulin levels were significantly higher, and plasma triglycerides were threefold greater in the high-GI-fed rats. Furthermore, pancreatic islet cells suffered "severely disorganized architecture and extensive fibrosis." However, the GI of these diets was not experimentally determined. In a well controlled feeding study no improvement in weight loss was observed with a low glycemic index diet over calorie restriction. Because high-amylose cornstarch (the major component of the assumed low-GI diet) contains large amounts of resistant starch, which is not digested and absorbed as glucose, the lower glycemic response and possibly the beneficial effects can be attributed to lower energy density and fermentation products of the resistant starch, rather than the GI.
In humans, a 2012 study shows that, after weight loss, the energy expenditure is higher on a low-glycemic index diet than on a low-fat diet (but lower than on the Atkins diet). See also news coverage and reactions from other obesity researchers.
Several lines of recent  scientific evidence have shown that individuals who followed a low-GI diet over many years were at a significantly lower risk for developing both type 2 diabetes, coronary heart disease, and age-related macular degeneration than others. High blood glucose levels or repeated glycemic "spikes" following a meal may promote these diseases by increasing systemic glycative stress, other oxidative stress to the vasculature, and also by the direct increase in insulin levels. The glycative stress sets up a vicious cycle of systemic protein glycation, compromised protein editing capacity involving the ubiquitin proteolytic pathway and autophagic pathways, leading to enhanced accumulation of glycated and other obsolete proteins.
In the past, postprandial hyperglycemia has been considered a risk factor associated mainly with diabetes. However, more recent evidence shows that it also presents an increased risk for atherosclerosis in the non-diabetic population and that high GI diets, high blood-sugar levels more generally, and diabetes are related to kidney disease as well.
Conversely, there are areas such as Peru and Asia where people eat high-glycemic index foods such as potatoes and high-GI rice without a high level of obesity or diabetes. The high consumption of legumes in South America and fresh fruit and vegetables in Asia likely lowers the glycemic effect in these individuals. The mixing of high- and low-GI carbohydrates produces moderate GI values.
A study from the University of Sydney in Australia suggests that having a breakfast of white bread and sugar-rich cereals, over time, may make a person susceptible to diabetes, heart disease, and even cancer.
A study published in the American Journal of Clinical Nutrition found that age-related adult macular degeneration (AMD), which leads to blindness, is 42% higher among people with a high-GI diet, and concluded that eating a lower-GI diet would eliminate 20% of AMD cases.
The American Diabetes Association supports glycemic index but warns that the total amount of carbohydrate in the food is still the strongest and most important indicator, and that everyone should make their own custom method that works best for them.
The International Life Sciences Institute concluded in 2011 that because there are many different ways of lowering glycemic response, not all of which have the same effects on health, "It is becoming evident that modifying the glycemic response of the diet should not be seen as a stand-alone strategy but rather as an element of an overall balanced diet and lifestyle."
A systematic review of few human trials examined the potential of low GI diet to improve pregnancy outcomes. Potential benefits were still seen despite no ground breaking findings in maternal glycemia or pregnancy outcomes. In this regard, more women under low GI diet achieved the target treatment goal for the postprandial glycemic level and reduced their need for insulin treatment. A low GI diet may also provide greater benefits to overweight and obese women. Interestingly, intervention at an early stage of pregnancy has shown a tendency to lower birth weight and birth centile in infants born to women with GDM. 
The number of grams of carbohydrate can have a bigger impact than glycemic index on blood sugar levels, depending on quantities. Consuming fewer calories, losing weight, and carbohydrate counting can be better for lowering the blood sugar level. Carbohydrates impact glucose levels most profoundly, and two foods with the same carbohydrate content are, in general, comparable in their effects on blood sugar. A food with a low glycemic index may have a high carbohydrate content or vice versa; this can be accounted for with the glycemic load (GL). Consuming carbohydrates with a low glycemic index and calculating carbohydrate intake would produce the most stable blood sugar levels.
Criticism and alternatives
The glycemic index does not take into account other factors besides glycemic response, such as insulin response, which is measured by the insulin index and can be more appropriate in representing the effects from some food contents other than carbohydrates. In particular, since it is based on the area under the curve of the glucose response over time from ingesting a subject food, the shape of the curve has no bearing on the corresponding GI value. The glucose response can rise to a high level and fall quickly, or rise less high but remain there for a longer time, and have the same area under the curve. For subjects with type 1 diabetes who do not have an insulin response, the rate of appearance of glucose after ingestion represents the absorption of the food itself. This glycemic response has been modeled, where the model parameters for the food enable prediction of the continuous effect of the food over time on glucose values, and not merely the ultimate effect that the GI represents.
Although the glycemic index provides some insights into the relative diabetic risk within specific food groups, it contains many counter-intuitive ratings. These include suggestions that bread generally has a higher glycemic ranking than sugar and that some potatoes are more glycemic than glucose. More significantly, studies such as that by Bazzano et al. demonstrate a significant beneficial diabetic effect for fruit compared to a substantial detrimental impact for fruit juice despite these having similar “low GI” ratings.
From blood glucose curves presented by Brand-Miller et al. the main distinguishing feature between average fruit and fruit juice blood glucose curves is the maximum slope of the leading edge of 4.38 (mmol/L)/hr for fruit and 6.71 (mmol/L)/hr for fruit juice. This raises the concept that the rate of increase in blood glucose may be a significant determinant particularly when comparing liquids to solids which release carbohydrates over time and therefore have an inherently greater area under the blood glucose curve.
- "Glycemic Index Defined". Glycemic Research Institute. Retrieved 2012-08-01.
- "Glycemic Index Food List & Chart". Retrieved 2014-08-08.
- Scheiner, Gary (2013). Until There is a Cure: The Latest and Greatest in Diabetes Self-Care. Spry Publishing LLC. ISBN 978-1-938170-13-3.
- "GI Database". Web.archive.org. Retrieved 2012-08-01.
- Freeman, Janine (September 2005). "The Glycemic Index debate: Does the type of carbohydrate really matter?". Diabetes Forecast. Archived from the original on February 14, 2007.
- Jenkins DJ, Wolever TM, Taylor RH et al. (March 1981). "Glycemic index of foods: a physiological basis for carbohydrate exchange". Am. J. Clin. Nutr. 34 (3): 362–6. PMID 6259925.
- Jenkins DJ, Kendall CW, McKeown-Eyssen G et al. (December 2008). "Effect of a low-glycemic index or a high-cereal fiber diet on type 2 diabetes: a randomized trial". JAMA 300 (23): 2742–53. doi:10.1001/jama.2008.808. PMID 19088352.
- Brouns F, Bjorck I, Frayn KN et al. (June 2005). "Glycaemic index methodology". Nutr Res Rev 18 (1): 145–71. doi:10.1079/NRR2005100. PMID 19079901.
- "What is a Glycemic Index?". Angelika Christie. 2009-09-21.
- Glycemic Index: From Research to Nutrition Recommendations?. Copenhagen: Nordic Council of Ministers. 2005. ISBN 92-893-1256-4. TemaNord2005:589.
- Atkinson FS, Foster-Powell K, Brand-Miller JC (December 2008). "International tables of glycemic index and glycemic load values: 2008". Diabetes Care 31 (12): 2281–3. doi:10.2337/dc08-1239. PMC 2584181. PMID 18835944.
- Brand-Miller, Jennie; Foster-Powell, Kaye (2005). The Low GI Diet Revolution: The Definitive Science-Based Weight Loss Plan. Marlowe & Company. p. 139. ISBN 978-1-56924-413-5.
- Brand-Miller JC, Fatema K, Fatima K et al. (June 2007). "Effect of alcoholic beverages on postprandial glycemia and insulinemia in lean, young, healthy adults". Am. J. Clin. Nutr. 85 (6): 1545–51. PMID 17556691.
- Godley R, Brown RC, Williams SM, Green TJ (May 2009). "Moderate alcohol consumption the night before glycaemic index testing has no effect on glycaemic response". Eur J Clin Nutr 63 (5): 692–4. doi:10.1038/ejcn.2008.27. PMID 18398423.
- "Nutrisystem". Web.archive.org. 2008-03-06. Retrieved 2012-08-01.
- John A. McDougall, "The McDougall Newsletter", June 2006.
- Foster-Powell K, Holt SH, Brand-Miller JC (July 2002). "International table of glycemic index and glycemic load values: 2002". Am. J. Clin. Nutr. 76 (1): 5–56. PMID 12081815.
- Pawlak DB, Kushner JA, Ludwig DS (2004). "Effects of dietary glycaemic index on adiposity, glucose homoeostasis, and plasma lipids in animals". Lancet 364 (9436): 778–85. doi:10.1016/S0140-6736(04)16937-7. PMID 15337404.
- Raatz, Susan K., et al. "Reduced glycemic index and glycemic load diets do not increase the effects of energy restriction on weight loss and insulin sensitivity in obese men and women." The Journal of Nutrition 135.10 (2005): 2387-2391.
- Ebbeling CB, Swain JF, Feldman HA et al. (June 2012). "Effects of dietary composition on energy expenditure during weight-loss maintenance". JAMA 307 (24): 2627–34. doi:10.1001/jama.2012.6607. PMC 3564212. PMID 22735432.
- Wanjek, Christopher (27 June 2012). "When Dieting, Not All Calories Are Created Equal". Scientific American.
- Bray GA (June 2012). "Diet and exercise for weight loss". JAMA 307 (24): 2641–2. doi:10.1001/jama.2012.7263. PMID 22735436.
- Kolata, Gina (9 July 2012). "In Dieting, Magic Isn't a Substitute for Science". The New York Times.
- Chiu CJ, Liu S, Willett WC et al. (April 2011). "Informing food choices and health outcomes by use of the dietary glycemic index". Nutr. Rev. 69 (4): 231–42. doi:10.1111/j.1753-4887.2011.00382.x. PMC 3070918. PMID 21457267.
- Temelkova-Kurktschiev TS, Koehler C, Henkel E, Leonhardt W, Fuecker K, Hanefeld M (December 2000). "Postchallenge plasma glucose and glycemic spikes are more strongly associated with atherosclerosis than fasting glucose or HbA1c level". Diabetes Care 23 (12): 1830–4. doi:10.2337/diacare.23.12.1830. PMID 11128361.
- Uchiki T, Weikel KA, Jiao W et al. (February 2012). "Glycation-altered proteolysis as a pathobiologic mechanism that links dietary glycemic index, aging, and age-related disease (in nondiabetics)". Aging Cell 11 (1): 1–13. doi:10.1111/j.1474-9726.2011.00752.x. PMC 3257376. PMID 21967227.
- Balkau B, Shipley M, Jarrett RJ et al. (March 1998). "High blood glucose concentration is a risk factor for mortality in middle-aged nondiabetic men. 20-year follow-up in the Whitehall Study, the Paris Prospective Study, and the Helsinki Policemen Study". Diabetes Care 21 (3): 360–7. doi:10.2337/diacare.21.3.360. PMID 9540016.
- Allan L said at October 23, 2006 6:20 AM: (2006-10-21). "High Glycemic Index Diet Kidney Cancer Risk?". FuturePundit. Retrieved 2012-02-21.
- "Kidney Disease (Nephropathy)". American Diabetes Association. Retrieved 2012-07-29.
- "Diabetes and kidney failure". Better Health Channel. State Government of Victoria. Retrieved 2012-02-21.
- "White bread breakfast unhealthy?".[dead link] The Times of India., 10 March 2008.
- Chiu CJ, Milton RC, Gensler G, Taylor A (July 2007). "Association between dietary glycemic index and age-related macular degeneration in nondiabetic participants in the Age-Related Eye Disease Study". Am. J. Clin. Nutr. 86 (1): 180–8. PMID 17616779.
- Sheard NF, Clark NG, Brand-Miller JC et al. (September 2004). "Dietary carbohydrate (amount and type) in the prevention and management of diabetes: a statement by the american diabetes association". Diabetes Care 27 (9): 2266–71. doi:10.2337/diacare.27.9.2266. PMID 15333500.
- "Glycemic Index and Diabetes". American Diabetes Association. Retrieved 8 June 2011
- Sadler, Michele (2011). Food, Glycaemic Response and Health. Brussels, Belgium: ILSI Europe. pp. 1–30. ISBN 9789078637318.
- Mohd Yusof BN, Firouzi S, Mohd Shariff Z, Mustafa N, Mohamed Ismail NA, Kamaruddin NA (March 2014). "Weighing the evidence of low glycemic index dietary intervention for the management of gestational diabetes mellitus: an Asian perspective". Int J Food Sci Nutr 65 (2): 144–50. doi:10.3109/09637486.2013.845652. PMID 24517860.
- "The Glycemic Index and Diabetes". Joslin Diabetes Center. Retrieved 2012-08-01.
- "David Mendosa. Insulin Index. July 13, 2003". Mendosa.com. Retrieved 2012-08-01.
- Worthington DR (Jan–Mar 1997). "Minimal model of food absorption in the gut.". Med Inform (Lond) 22 (1): 35–45. doi:10.3109/14639239709089833. PMID 9183779.
- Bazzano LA, Li TY, Joshipura KJ, Hu FB (July 2008). "Intake of fruit, vegetables, and fruit juices and risk of diabetes in women". Diabetes Care 31 (7): 1311–7. doi:10.2337/dc08-0080. PMC 2453647. PMID 18390796.
- Brand-Miller JC, Stockmann K, Atkinson F, Petocz P, Denyer G (January 2009). "Glycemic index, postprandial glycemia, and the shape of the curve in healthy subjects: analysis of a database of more than 1,000 foods". Am. J. Clin. Nutr. 89 (1): 97–105. doi:10.3945/ajcn.2008.26354. PMID 19056599.
|Wikimedia Commons has media related to Glycemic index.|
- Human Nutrition Unit, School of Molecular Bioscience. "Glycemic Index". international GI database. University of Sydney. — Searchable database of over 2600 foods with their glycemic index and load values.
- List of low GI foods — Provided by the University of Sydney with some additional foods.
- Revised International Table of Glycemic Index (GI) and Glycemic Load (GL) Values—2008 Table of about 2500 food items with their glycemic index and load values.
- Foster-Powell K, Holt SH, Brand-Miller JC (July 2002). "International table of glycemic index and glycemic load values: 2002". Am. J. Clin. Nutr. 76 (1): 5–56. PMID 12081815. 750 items
- Michel Montignac: Glycemic Index