Low Glycemic Index Diets Overview

Mainstream groups, such as the American Heart Association and the American Dietetic Association, endorse a unified set of guidelines for the optimum diet. According to these organizations, the majority of calories in the daily diet should come from carbohydrates (55% to 60%); fat should provide no more than 30% of total calories; and protein should be kept to 10% to 15%.

However, many popular diet books turn the standard diet on its head. As described in the entry on low-carbohydrate diets , the Atkins diet, the Zone diet, Protein Power, and other “alternative” dietary approaches turn thumbs down on carbohydrates and advocate increased consumption of fat and/or protein. According to theory, the low-carb approach aids in weight loss (and provides a variety of other health benefits) by reducing the body’s production of insulin.

The low glycemic index (low-GI) diet splits the difference between the low-carb and low-fat approaches. It maintains the low-carb diet’s focus on insulin, but it suggests choosing certain carbohydrates over others rather than restricting carbohydrate intake.

Evidence suggests that carbohydrates are not created equal. Some carbohydrates, such as pure glucose, are absorbed quickly and create a rapid, strong rise in both blood sugar and insulin. However, others (such as brown rice) are absorbed much more slowly and produce only a modest blood sugar and insulin response. According to proponents of the low-GI diet, eating foods in the latter category will enhance weight loss and improve health. However, as we shall see, despite some promising theory there is as yet no solid evidence that low-GI diets enhance weight loss.

Besides weight loss, preliminary evidence suggests that the low-GI approach (or, even better, a related method called low glycemic load) may help prevent heart disease. The low-GI approach has also shown promise for treating and possibly preventing diabetes.

The precise measurement of the glucose-stimulating effect of a food is called its glycemic index. The lower a food’s glycemic index, the less potent its effects on blood sugar (and therefore insulin).

The glycemic index of glucose is arbitrarily set at 100. The ratings of other foods are determined as follows. First, researchers calculate a portion size for the food to supply 50 g of carbohydrates. Next, they give that amount of the food to at least 8-10 people and measure the blood sugar response. (By using a group of people rather than one person, researchers can ensure that the idiosyncrasies of one individual don’t skew the results.) On another occasion, researchers also give each participant an equivalent amount of glucose and perform the same measurements. The glycemic index of a food is then determined by comparing the two outcomes. For example, if a food causes half of the blood sugar rise of glucose, it is assigned a GI of 50; if it causes one-quarter of the rise, it is assigned a GI of 25. The lower the glycemic index, the better.

When scientists first began to determine the glycemic index of foods, some of the results caused eyebrows to rise. It didn’t surprise anyone when jellybeans turned out to have a high glycemic index of 80—after all, jellybeans are mostly sugar. Nor was it unexpected that kidney beans have a low glycemic index of 27 because they are notoriously hard to digest. But when baked potatoes came back with a reading of 93, researchers were taken aback. This rating is higher than that of almost all other foods, including ice cream (61), sweet potatoes (54), and white bread (70). Based on this finding, low-GI diets recommend that you stay largely away from potatoes. (However, the concern regarding potatoes is probably unnecessary. See the discussion of glycemic load below.)

There are other surprises hidden in the glycemic index tables. For example, fructose (the sweetener in honey) has an extraordinarily low glycemic index of 23—lower than brown rice and almost three times lower than white sugar. Candy bars also tend to have a relatively good (low) glycemic index, presumably because their fat content makes them digest slowly.

It’s difficult to predict the glycemic index of a food without specifically testing it, but there are some general factors that can be recognized. Fiber content tends to reduce the glycemic index of a food, presumably by slowing down digestion. For this reason, whole grains usually have a lower GI score than refined, processed grains. Fat content also reduces GI score. Simple carbohydrates (such as sugar) often have a higher GI score than complex carbohydrates (such as brown rice).

However, there are numerous exceptions to these rules. Factors such as the acid content of food; the size of the food particles; and the precise mixture of fats, proteins, and carbohydrates can substantially change the GI measurement. For a measurement like the glycemic index to be meaningful, it has to be generally reproducible between people. In other words, if a potato has a glycemic index of 93 in one person, it should have pretty much the same glycemic index when given to another person. Science suggests that the GI passes this test. The glycemic index of individual foods is fairly constant between people, and even mixed meals have a fairly predictable effect according to most (but not all) studies. ^{1 2 3 4 5} Thus, the GI of a food really does indicate its propensity to raise insulin levels. Whether a diet based on the index will aid in weight loss, however, is another story.

Following a low glycemic index diet is fairly easy. Basically, you follow the typical diet endorsed by authorities such as the American Dietetic Association, but you choose carbohydrates that fall toward the lower end of the glycemic index scale. Books such as The Glucose Revolutiongive a great deal of information on how to make these choices.

There are two primary theoretical reasons given why low-GI diets should help reduce weight. The most prominent reason given in books on the low-GI approach involves insulin levels. Basically, these books show that low-GI diets reduce insulin release, and then take almost for granted the idea that reduced insulin levels should aid in weight loss. However, there is little justification for the second part of this argument. Excess weight is known to lead to elevated insulin levels, but there is little meaningful evidence for the reverse: that reducing insulin levels will help remove excess weight. For more information on this very complicated subject, see the article on low-carb diets .

Books on the low-GI diet give another reason for using their approach as well: they state that low-GI foods fill you up more quickly than high-GI foods and also keep you feeling full for longer. Unfortunately, as we shall see, there is more evidence against this belief than for it.

The Satiety Index

A measurement called the Satiety Index assigns a numerical quantity to the filling quality of a food. ⁶ These numbers are determined by feeding people fixed caloric amounts of those foods, and then determining how soon they get hungry again and how much they eat at subsequent meals. The process is similar to the methods used to establish the GI index.

The results of these measurements do not corroborate the expectations of low-GI diet proponents. As it happens, foods with the worst (highest) GI index are often the most satiating—exactly the reverse of what low-GI theory proponents would say. ⁷ For example, the Satiety Index tells us that potatoes are among the most satiating of foods. However, as noted above, the GI analysis gave potatoes a bad rating. According to the low-GI theory, you should feel hunger pangs shortly after eating a big baked potato. In real life, that doesn’t happen—even without the sour cream!

There are numerous other contradictions between research findings and the low-GI/high-satiety theory. For example, one study found no difference in satiety between fructose (fruit sugar) and glucose when taken as part of a mixed meal, even though fructose has a GI more than four times lower than glucose. ⁸ A few studies do seem to suggest that certain low-GI foods are more filling than high-GI foods. ^{9 10} However, in these studies the bulkiness and lack of palatability of the low-GI foods chosen may have played a more important role than the food’s glycemic index. ¹¹ Thus, the satiety argument for low-GI diets doesn’t appear to hold up to scrutiny.

There’s another problem with the low-GI approach: It’s probably the wrong way to assess the insulin-related effects of food.

The glycemic index measures blood sugar response per gram of carbohydrate contained in a food, not per gram of the food. This leads to some odd numbers. For example, a parsnip has a glycemic index of 98, almost as high as pure sugar. If taken at face value, this figure suggests that dieters should avoid parsnips like the plague; however, if you’ve ever eaten a parsnip, you know that it’s not exactly candy. In fact, parsnips are mostly indigestible fiber, and you would have to eat a few bushels to trigger a major glucose and insulin response.

The reason for the high number is that the glycemic index rates the effects per gram of carbohydrate rather than per gram of total parsnip, and the sugar present in minute amounts in a parsnip itself is highly absorbable. The high glycemic index rating of parsnips is thus extremely misleading. Books such as The Glucose Revolutiontake care of issues like this on a case-by-case basis by saying, for example, that you can consider most vegetables “free foods” regardless of their glycemic index. But in fact the same considerations apply to all foods and distort the meaningfulness of the scale as a whole.

A different measurement, the glycemic load (GL), takes this into account. The GL is derived by multiplying the glycemic index by the percent carbohydrate content of a food. In other words, it measures the glucose/insulin response per gram of food rather than per gram of carbohydrate in that food. Using this system, the GL of a parsnip is 10, while glucose has a relative load of 100. And remember the potato problem, that terrible bogey of GI diets? The GL of a typical serving of potato is only 27—not such a bogeyman at all. Such numbers make a lot more sense.

Theory is one thing and practice is another. It is certainly possible that making sure to focus on low-GI or low-GL foods will help you lose weight, even if the theoretical justification for the idea is weak. However, there is only preliminary positive evidence to support this possibility, and the largest and longest-term trial failed to find benefit. In one of the positive studies, 107 overweight adolescents were divided into two groups: a low-GI group and a low-fat group. ¹² The low-GI group was counseled to follow a diet consisting of 45% to 50% carbohydrates (preferably low-GI carbohydrates), 20% to 25% protein, and 30% to 35% fat. Calorie restriction was not emphasized. The low-fat group received instructions for a standard low-fat, low-calorie diet divided up into 55% to 60% carbohydrates, 15% to 20% protein, and 25% to 30% fat. Over a period of about 4 months, participants on the low-GI diet lost about 4.5 pounds, while those on the standard diet lost just fewer than 3 pounds.

Unfortunately, this study does not say as much about the low-GI approach as it might seem. Perhaps the most obvious problem is that the low-GI diet used here was also a high-protein diet. It is possible that high-protein diets might help weight loss regardless of the glycemic index of the foods consumed. (In fact, that is precisely what proponents of high-protein diets claim.)

Another problem is that participants were not assigned to the two groups randomly. Rather, researchers consciously picked which group each participant should join. This is a major flaw because it introduces the possibility of intentional or unintentional bias. It is quite possible, for example, that researchers placed adolescents with greater self-motivation into the low-GI group, based on an unconscious desire to see results from the study. This is not an academic problem, and modern medical studies always use randomization to circumvent it.

Finally, researchers made no effort to determine how well participants followed their diets. It might be that those in the low-fat diet group simply didn’t stick to the rules as well as those in the low-GI diet group because the rules were more challenging.

Despite these many flaws, the study results are still promising. Losing weight without deliberately cutting calories is potentially a great thing.

In another study, 30 overweight women with excessively high insulin levels were put on either a normal low-calorie diet or one that supplied the same amount of calories but used low-GI foods. ¹³ The results over 12 weeks showed that women following the low-GI diet lost several pounds more than those following the normal diet.

In yet another small study, this one involving overweight adolescents, a conventional reduced calorie diet was compared against a low-GL diet that did not have any calorie restrictions. ¹⁴ The results showed that simply by sticking to low-GI foods, without regard for calories, the participants on the low-GI diet were able to lose as much or more weight as those on the low-calorie diet.

However, in the largest and longest-term study published to date, an 18-month trial of 203 Brazilian women, use of a low-GI diet failed to prove more effective than a high-GI diet. ¹⁵ Additionally, a smaller study failed to find a low-GI diet more effective for weight loss than a low-fat diet except in people with high levels of circulating insulin. ¹⁶

There is some evidence that a low-GI diet (or, even better, a low-GL diet) might help prevent cancer and heart disease. The low-GI approach has also shown promise for preventing or treating diabetes.

Heart Disease Prevention

One large observational study evaluated the diets of more than 75,000 women and found that those women whose diets provided a lower GL had a lower incidence of heart disease . ¹⁷ In this study, 75,521 women aged 38 to 63 years were followed for 10 years. Each filled out detailed questionnaires regarding her diet. Using this data, researchers calculated the average GL of each participant. The results showed that women who consumed a diet with a high GL were more likely to experience heart disease than those who consumed a diet of low GL.

Other observational studies suggest that consumption of foods with lower GL may improve cholesterol profile—specifically, reduced triglyceride levels and higher HDL (“good” cholesterol) levels. ¹⁸ These effects, in turn, might lead to decreased risk of heart disease. However, yet other observational studies have found little or no relationship between heart disease and GI or GL. ¹⁹ These contradictory results are not surprising. But, even if the observational study results were entirely consistent, it wouldn’t prove the case for a low-GI approach. Conclusions based on observational studies are notoriously unreliable due to the possible presence of unidentified confounding factors. For example, because there is an approximate correlation between fiber in the diet and GL, it is possible that benefits, when seen, are really due to fiber intake instead. Factors such as this one may easily obscure the effects of the factor under study, leading to contradictory or misleading results.

Intervention trials (studies in which researchers actually intervene in participants' lives) are more reliable, and some have been conducted to evaluate the low-GI diet. For example, in the large weight-loss trial mentioned earlier, the low-GI diet failed to prove more effective than a high-GI diet regarding weight loss. The results did suggest, though, that a low-GI diet can improve cholesterol profile. ²⁰ However, this study was not primarily designed to look at effects on cholesterol. A study that primarily focused on this outcome followed 30 people with high lipid levels for three months. ²¹ During the second month, low-GI foods were substituted for higher-GI foods, while other nutrients were kept similar. Improvements were seen in total cholesterol, LDL cholesterol, and triglycerides, but not in HDL. Interestingly, a close analysis of the results showed that only patients who had high triglycerides at the beginning of the study showed benefit. Another controlled trial found that a high carbohydrate, low-GL diet optimized lipid profile as compared to several other diets. ²² But another study found that both low-fat and low-GI diets were about equally effective regarding lipid profile. ²³ Another approach to the issue involves analysis of effects on insulin resistance. Evidence suggests that increased resistance of the body to its own insulin raises the risk of heart disease. ^{24 25 26 27 28} One study found that use of a low-GI diet versus a high-GI diet improved the body’s sensitivity to insulin in women at risk for heart disease. ²⁹ Similar results were seen in a group of people with severe heart disease ³⁰ as well as in healthy people. ³¹ While these results are preliminary, taken together they do suggest that consumption of low-GI foods might have a beneficial effect on heart disease risk.

Low-GI Diet and Diabetes

Two large observational studies, one involving men and the other involving women, found that diets with lower GLs were associated with a lower rate of diabetes . ³² For example, one trial followed 65,173 women for a period of 6 years. ³³ Women whose diets had a high GL had a 47% increased risk of developing diabetes compared with those whose diets had the lowest GL. Fiber content of diet also makes a difference. People who consumed a diet that was both low in fiber and high in GL had a 250% increased incidence of diabetes.

However, as always, the results of these observational studies have to be taken with a grain of salt. It’s quite possible that unrecognized factors are responsible for the results seen. For example, magnesium deficiency is widespread and may contribute to the development of diabetes; whole grains contain magnesium and are also low-GI foods. Therefore, it could be that the benefits seen in these studies are actually caused by increased magnesium intake in the low-GI group, rather than effects on blood sugar and insulin. ³⁴ Furthermore, one observational study found no connection between the glycemic values of foods and the incidence of diabetes. ³⁵ Another observational study did find a correlation between carbohydrate intake (especially pastries) and the onset of diabetes, but no consistent relationship with glycemic index. ³⁶ Other studies have found no relationship between sugar consumption (a high-GI food) and diabetes onset. ^{37 38} Thus, reducing dietary GL may help prevent diabetes, but we don’t know this for sure.

Whether or not low-GI diets can prevent diabetes, for people who already have diabetes, going on a low-GI diet might improve blood sugar control. ³⁹ However, the benefits seem to be small at most. ⁴⁰

As you have seen, the evidence that a low-GI diet will help you lose weight is not yet very impressive. Its theoretical foundation is weak, and it appears to be using the wrong method of ranking foods regarding their effects on insulin. Conversely, however, there’s no reason to believe a low-GI diet causes harm. If you find that you lose weight with a low-GI diet, stick with it.

Note:While the most popular low-GI diet books ( The Glucose Revolution, Sugar Busters) recommend a diet that is generally reasonable and should be safe, it is easy to design some fairly extreme low-GI diets. For example, a diet consisting of nothing but lard would be a very, very low-GI diet, since the glycemic index of lard is 0. While it no longer seems that saturated fat is as harmful as once thought, a pure lard diet is probably not a good idea. If you run across a diet book that recommends achieving a low-GI by consuming an extreme diet, approach it with caution.

1. Hollenbeck CB, Coulston AM, Reaven GM. Comparison of plasma glucose and insulin responses to mixed meals of high-, intermediate-, and low-glycemic potential. Diabetes Care. 11(4):323-9.
2. Laine DC, Thomas W, Levitt MD, Bantle JP. Comparison of predictive capabilities of diabetic exchange lists and glycemic index of foods. Diabetes Care. 10(4):387-94.
3. Wolever TM, Jenkins DJ, Vuksan V, et al. Glycemic index of foods in individual subjects. Diabetes Care. 1990;13:126-132.
4. Wolever TM, Jenkins DJ, Vuksan V, Jenkins AL, Wong GS, Josse RG. Beneficial effect of low-glycemic index diet in overweight NIDDM subjects. Diabetes Care. 15(4):562-4.
5. Wolever TM, Nuttall FQ, Lee R, Wong GS, Josse RG, Csima A, Jenkins DJ. Prediction of the relative blood glucose response of mixed meals using the white bread glycemic index. Diabetes Care. 8(5):418-28.
6. Holt SH, Miller JC, Petocz P, Farmakalidis E. A satiety index of common foods. Eur J Clin Nutr. 49(9):675-90.
7. Holt SH, Brand Miller JC, Petocz P. Interrelationships among postprandial satiety, glucose and insulin responses and changes in subsequent food intake. Eur J Clin Nutr. 50(12):788-97.
8. Rodin J, Reed D, Jamner L. Metabolic effects of fructose and glucose: implications for food intake. Am J Clin Nutr. 47(4):683-9.
9. Ludwig DS. The glycemic index: physiological mechanisms relating to obesity, diabetes, and cardiovascular disease. JAMA. 287(18):2414-23.
10. van Amelsvoort JM, Westrate JA. Amylose-amylopectin ratio in a meal affects postprandial variables in male volunteers. Am J Clin Nutr. 1992;55:712-718.
11. Holt SH, Miller JC, Petocz P. Interrelationships among postprandial satiety, glucose and insulin responses and changes in subsequent food intake. Eur J Clin Nutr. 1996;50:788-797.
12. Spieth LE, Harnish JD, Lenders CM, Raezer LB, Pereira MA, Hangen SJ, Ludwig DS. A low-glycemic index diet in the treatment of pediatric obesity. Arch Pediatr Adolesc Med. 154(9):947-51.
13. Slabber M, Barnard HC, Kuyl JM, Dannhauser A, Schall R. Effects of a low-insulin-response, energy-restricted diet on weight loss and plasma insulin concentrations in hyperinsulinemic obese females. Am J Clin Nutr. 60(1):48-53.
14. Ebbeling CB, Leidig MM, Sinclair KB, Hangen JP, Ludwig DS. A reduced-glycemic load diet in the treatment of adolescent obesity. Arch Pediatr Adolesc Med. 157(8):773-9.
15. Sichieri R, Moura AS, Genelhu V, Hu F, Willett WC. An 18-mo randomized trial of a low-glycemic-index diet and weight change in Brazilian women. Am J Clin Nutr. 86(3):707-13.
16. Ebbeling CB, Leidig MM, Feldman HA, Lovesky MM, Ludwig DS. Effects of a low-glycemic load vs low-fat diet in obese young adults: a randomized trial. JAMA. 297(19):2092-102.
17. Liu S, Willett WC, Stampfer MJ, Hu FB, Franz M, Sampson L, Hennekens CH, Manson JE. A prospective study of dietary glycemic load, carbohydrate intake, and risk of coronary heart disease in US women. Am J Clin Nutr. 71(6):1455-61.
18. Liu S, Manson JE, Stampfer MJ, Holmes MD, Hu FB, Hankinson SE, Willett WC. Dietary glycemic load assessed by food-frequency questionnaire in relation to plasma high-density-lipoprotein cholesterol and fasting plasma triacylglycerols in postmenopausal women. Am J Clin Nutr. 73(3):560-6.
19. van Dam RM, Visscher AW, Feskens EJ, Verhoef P, Kromhout D. Dietary glycemic index in relation to metabolic risk factors and incidence of coronary heart disease: the Zutphen Elderly Study. Eur J Clin Nutr. 54(9):726-31.
20. Sichieri R, Moura AS, Genelhu V, Hu F, Willett WC. An 18-mo randomized trial of a low-glycemic-index diet and weight change in Brazilian women. Am J Clin Nutr. 86(3):707-13.
21. Jenkins DJ, Wolever TM, Kalmusky J, Guidici S, Giordano C, Patten R, Wong GS, Bird JN, Hall M, Buckley G. Low-glycemic index diet in hyperlipidemia: use of traditional starchy foods. Am J Clin Nutr. 46(1):66-71.
22. McMillan-Price J, Petocz P, Atkinson F, O'neill K, Samman S, Steinbeck K, Caterson I, Brand-Miller J. Comparison of 4 diets of varying glycemic load on weight loss and cardiovascular risk reduction in overweight and obese young adults: a randomized controlled trial. Arch Intern Med. 166(14):1466-75.
23. Ebbeling CB, Leidig MM, Feldman HA, Lovesky MM, Ludwig DS. Effects of a low-glycemic load vs low-fat diet in obese young adults: a randomized trial. JAMA. 297(19):2092-102.
24. Saydah SH, Loria CM, Eberhardt MS, et al. Subclinical states of glucose intolerance and risk of death in the US. Diabetes Care. 2001;24:447-453.
25. Job FP, Wolfertz J, Meyer R, Hübinger A, Gries FA, Kuhn H. Hyperinsulinism in patients with coronary artery disease. Coron Artery Dis. 5(6):487-92.
26. Després JP, Lamarche B, Mauriège P, Cantin B, Dagenais GR, Moorjani S, Lupien PJ. Hyperinsulinemia as an independent risk factor for ischemic heart disease. N Engl J Med. 334(15):952-7.
27. Pyörälä K, Savolainen E, Kaukola S, Haapakoski J. Plasma insulin as coronary heart disease risk factor: relationship to other risk factors and predictive value during 9 1/2-year follow-up of the Helsinki Policemen Study population. Acta Med Scand Suppl. 701():38-52.
28. Lamarche B, Tchernof A, Mauriège P, Cantin B, Dagenais GR, Lupien PJ, Després JP. Fasting insulin and apolipoprotein B levels and low-density lipoprotein particle size as risk factors for ischemic heart disease. JAMA. 279(24):1955-61.
29. Frost G, Leeds A, Trew G, Margara R, Dornhorst A. Insulin sensitivity in women at risk of coronary heart disease and the effect of a low glycemic diet. Metabolism. 47(10):1245-51.
30. Frost G, Keogh B, Smith D, Akinsanya K, Leeds A. The effect of low-glycemic carbohydrate on insulin and glucose response in vivo and in vitro in patients with coronary heart disease. Metabolism. 45(6):669-72.
31. Jenkins DJ, Wolever TM, Collier GR, Ocana A, Rao AV, Buckley G, Lam Y, Mayer A, Thompson LU. Metabolic effects of a low-glycemic-index diet. Am J Clin Nutr. 46(6):968-75.
32. Salmerón J, Ascherio A, Rimm EB, Colditz GA, Spiegelman D, Jenkins DJ, Stampfer MJ, Wing AL, Willett WC. Dietary fiber, glycemic load, and risk of NIDDM in men. Diabetes Care. 20(4):545-50.
33. Salmerón J, Manson JE, Stampfer MJ, Colditz GA, Wing AL, Willett WC. Dietary fiber, glycemic load, and risk of non-insulin-dependent diabetes mellitus in women. JAMA. 277(6):472-7.
34. Meyer KA, Kushi LH, Jacobs DR Jr, Slavin J, Sellers TA, Folsom AR. Carbohydrates, dietary fiber, and incident type 2 diabetes in older women. Am J Clin Nutr. 71(4):921-30.
35. van Dam RM, Visscher AW, Feskens EJ, Verhoef P, Kromhout D. Dietary glycemic index in relation to metabolic risk factors and incidence of coronary heart disease: the Zutphen Elderly Study. Eur J Clin Nutr. 54(9):726-31.
36. Feskens EJ, Bowles CH, Kromhout D. Carbohydrate intake and body mass index in relation to the risk of glucose intolerance in an elderly population. Am J Clin Nutr. 54(1):136-40.
37. Feskens EJM, Kromhout D. Cardiovascular risk factors and the 25-year incidence of diabetes mellitus in middle-aged men. The Zutphen Study. Am J Epidemiol. 1989;130:1001-1008.
38. Baird JD. Diet and the development of clinical diabetes. Acta Diabetol Lat. 1972;9(suppl 1):621-639.
39. Wolever TM. The glycemic index: flogging a dead horse? Diabetes Care. 20(3):452-6.
40. Coulston AM, Reaven GM. Much ado about (almost) nothing. Diabetes Care. 20(3):241-3.