GLYCEMIC INDEX
- ADA - American Diabetes Association
- CVD - Cardiovascular disease
- GI - Glycemic Index
- GL - Glycemic Load
- SBP - Systolic blood pressure
- USDA - United States Department of Agriculture
- DEFINITION OF GLYCEMIC INDEX (GI) AND GLYCEMIC LOAD (GL)
- Glycemic index (GI)
- The GI is a measure of how much blood sugar levels tend to rise after the ingestion of various carbohydrate-containing foods
- Carbohydrates have the most direct effect on blood sugar levels when compared to fats and proteins
- Carbohydrates come in different forms and their absorption in the intestine is affected by a number of factors
- The GI allows a comparison of different foods based on how they tend to affect blood sugar levels
- Glycemic load (GL)
- The amount of carbohydrate consumed in a meal will also affect the rise in blood sugar levels
- GL combines the GI and the amount of carbohydrates consumed to give a measure of total carbohydrate load
- By accounting for the amount of carbohydrates consumed, the GL can give a more accurate estimate of a food's effect on blood sugar levels than GI alone
- CALCULATING THE GLYCEMIC INDEX AND GLYCEMIC LOAD
- Glycemic index (GI)
- The GI of a food is calculated in the following manner:
- 1. Ten healthy people are fed an amount of test food so that they receive 50 grams of carbohydrate from the food
- 2. Blood sugars are then measured on the participants at different intervals over the next 2 hours
- 3. The area under the curve for blood sugar response is then calculated for each person
- 4. On a separate occasion, the same people are given a reference food (typically white bread or 50 grams of glucose) and the same calculations are performed
- 5. The area under the curve for the test food is divided by the area under the curve for the reference food
- 6. This ratio is then multiplied by 100 to get the Glycemic Index
- 7. The GI values for each food from each participant are then averaged to give a final GI for that food [1,2]
- Glycemic load (GL)
- GL is equal to the GI of a food multiplied by its carbohydrate content in grams divided by 100
- GL takes into account the amount of the food that is eaten [1,2]
- Example:
- Raw apple has a GI = 52
- One large raw apple has approximately 31 grams of carbs
- If a person eats 2 raw apples their GL = 52 X (31 X 2) = 3224 /100 = 32.24
- GLYCEMIC INDEX CATEGORIES
- Overview
- There is no universally accepted range for what constitutes a low, medium, and high GI food
- The table below shows values that are generally used
GI category |
GI value |
Low GI |
≤ 55 |
Medium GI |
56 - 69 |
High GI |
≥ 70 |
- DIABETES AND THE GLYCEMIC INDEX
- Overview
- A number of studies of varying design have evaluated the effect of the glycemic index on blood sugar control in diabetics. Most studies are very small and of suboptimal design.
- Professional guidelines:
- The ADA 2018 diabetes guidelines state the following about the glycemic index:
- The literature concerning glycemic index and glycemic load in individuals with diabetes is complex often yielding mixed results, though in some studies lowering the glycemic load
of consumed carbohydrates has demonstrated A1C reductions of –0.2% to –0.5%
- Studies longer than 12 weeks report no significant influence of glycemic index or glycemic load independent of weight loss on A1C; however, mixed results
have been reported for fasting glucose levels and endogenous insulin levels [63]
- StraightHealthcare analysis:
- Unfortunately, there are no good, long-term studies that have evaluated the effects of a low GI diet on diabetes outcomes
- The available larger studies have suboptimal designs that make it difficult to draw conclusions
- Two meta-analyses that we reviewed found that a low GI diet (average GI around 65) lowered A1C values by an average of 0.3 - 0.5% when compared to a high GI diet (average GI around 83)
- Based on the available evidence, it is not clear what effect the GI has on diabetic control and diabetes outcomes
- OTHER CONDITIONS AND THE GLYCEMIC INDEX
Diabetes prevention
- A number of cohort studies have evaluated the relationship of dietary GI and GL with risk of diabetes
- Some studies have found a positive relationship between higher GI/GL diets and diabetes incidence while others have not
- There are no randomized, controlled trials that have evaluated the association between dietary GI values and diabetes risk
- In general, there is some observational evidence that diets higher in GI and GL may increase the risk of developing diabetes
- The evidence is not conclusive [11,12,13,14,15,16,17,18,19,20,21]
Weight loss
- A handful of trials have compared diets with different glycemic loads and indexes in weight loss
- Glycemic Index in Weight Loss, Archives of Internal Medicine (2006)
[PubMed abstract]
- A study in the Archives of Internal medicine enrolled 129 patients with an average BMI of 31
- Main inclusion criteria: age 18 - 40 years; BMI ≥ 25; body weight < 330 lbs (150 kg); weight fluctuations < 11 lbs in previous 2 months
- Main exclusion criteria: chronic illness; medications other than birth control pills; special diets
- Baseline characteristics: average age 32 years; female sex - 75%; average weight - 191 lbs (87 kg); average BMI - 31;
- Patients were randomized to 1 of 4 diets:
- Diet 1 (32 patients) - High carbohydrate/High Glycemic index diet (GI - 70, GL - 129)
- Diet 2 (32 patients) - High carbohydrate/Low Glycemic index diet (GI - 45, GL - 89)
- Diet 3 (32 patients) - High protein/High glycemic index (GI - 59, GL - 75)
- Diet 4 (33 patients) - High protein/Low glycemic index (GI - 44, GL - 59)
- Diets for women and men were designed to provide 1400 kcal/day and 1900 kcal/day, respectively
- All key carbohydrate and protein foods were provided along with some preprepared meals
- Change in body composition was measured with dual energy x-ray absorptiometry
- PRIMARY OUTCOME: Change in body weight and body composition at 12 weeks
- At 12 weeks, the following was seen:
- Primary outcome (weight loss): Diet 1 - 8.1 lbs (3.7 kg), Diet 2 - 10.6 lbs (4.8 kg), Diet 3 - 11.7 lbs (5.3 kg), Diet 4 - 9.7 lbs (4.4 kg) (p=0.17 for difference among groups)
- Primary outcome (percent weight loss): Diet 1 - 4.2%, Diet 2 - 5.5%, Diet 3 - 6.2%, Diet 4 - 4.8% (p=0.09 for difference among groups)
- Decrease in fat mass: Diet 1 - 6.2 lbs (2.8 kg), Diet 2 - 9.9 lbs (4.5 kg), Diet 3 - 9.5 lbs (4.3 kg), Diet 4 - 8.1 lbs (3.7 kg) (p=0.08 for difference among groups)
- Dropout rate: Diet 1 - 16%, Diet 2 - 6.3%, Diet 3 - 3.1%, Diet 4 - 15.2% [33]
- Low-Glycemic Load vs Low-Fat Diet, JAMA (2007)
[PubMed abstract]
- A study in JAMA enrolled 73 overweight patients
- Main inclusion criteria: age 18 - 35 years; BMI ≥ 30
- Main exclusion criteria: body weight > 308 lbs (140 kg); smoker; diabetes; use of medications that could affect weight loss or insulin sensitivity; major medical condition
- Baseline characteristics: average age 27 years; average weight - 227 lbs (103 kg); average % body fat - 40%
- Patients were randomized to 1 of 2 diets:
- Diet 1 (36 patients) - Low-glycemic load diet
- Diet 2 (37 patients) - Low-fat diet
- For Diet 1, participants were counseled to consume low–glycemic load foods (particularly nonstarchy vegetables, legumes, and temperate fruits) and to limit intake
of high–glycemic load foods (such as refined grains, starchy vegetables, fruit juices, and sweets)
- For Diet 2, participants were counseled to consume low-fat grains, vegetables, fruits, and legumes and to limit intake of added fats, sweets, and
high-fat snacks
- Participants were not given specific calorie goals
- PRIMARY OUTCOME: Change in body weight and body fat percentage as determined by dual energy x-ray absorptiometry at 6, 12, and 18 months
- During follow-up, the following was seen:
- Primary outcome (weight loss at 18 months): Diet 1 ∼ 4.4 lbs (2 kg), Diet 2 - ∼ 4.4 lbs (2 kg) (p=0.99)
- Primary outcome (decrease in % body fat at 18 months): Diet 1 - 1.5%, Diet 2 - 1.1% (p=0.50)
- Fasting glucose and fasting insulin levels did not differ significantly between the two groups at 6 and 18 months [34]
- StraightHealthcare analysis:
- In these 2 small trials, glycemic index/load did not have a significant effect on weight loss or body composition
Cardiovascular disease risk factors
- The OmniCarb trial detailed below evaluated the effects of low and high GI diets on cardiovascular risk factors
- OmniCarb Trial - Effect of High vs Low GI on CVD Risk Factors, JAMA (2014)
[PubMed abstract]
- The OmniCarb trial enrolled 163 overweight adults. The trial was a crossover trial where patients consumed 4 different diets.
- Main inclusion criteria: ≥ 30 years old; SBP 120 - 159 mmHg; DBP 70 - 99 mmHg; BMI ≥ 25
- Main exclusion criteria: cardiovascular disease; diabetes; chronic kidney disease; taking medication that lowers blood pressure or lipids; fasting blood
glucose level ≥ 125 mg/dL
- Baseline characteristics: average age 53 years; average BMI - 32; average triglycerides - 105 mg/dl; average LDL - 153 mg/dl; average HDL - 58 mg/dl; average SBP - 132 mmHg; average
insulin sensitivity - 7.3
- Patients crossed over between 4 different diets. Each diet was consumed for 5 weeks with a 2-week washout between diets.
- 1. High-carbohydrate, high-glycemic index
- 2. High-carbohydrate, low-glycemic index
- 3. Low-carbohydrate, high-glycemic index
- 4. Low-carbohydrate, low-glycemic index
- Low carbohydrate = 40% of calories from carbs; High carbohydrate = 58% of calories from carbs
- Low glycemic index = average GI ≤ 45; High glycemic index = average GI ≥ 65
- All of the meals were prepared for the patients and they were consumed under staff supervision
- Calorie intake was adjusted to maintain initial body weight
- PRIMARY OUTCOME: Coprimary outcomes were insulin sensitivity (measured by Matsuda index); systolic blood pressure;
low-density lipoprotein (LDL) cholesterol; high density lipoprotein (HDL) cholesterol; and triglyceride levels. Outcomes were measured at the end of each 5-week diet period.
- The following outcomes were seen:
- Insulin sensitivity (lower numbers indicate less sensitivity): Diet 1 - 8.9, Diet 2 - 7, Diet 3 - 7.9, Diet 4 - 8.1 (1 vs 2, p≤0.01; all other p>0.05)
- LDL (mg/dl): Diet 1 - 139, Diet 2 - 146, Diet 3 - 138, Diet 4 - 138 (1 vs 2, p≤0.01; 3 vs 2, p≤0.01; 4 vs 2, p≤0.01; all other p>0.05)
- HDL (mg/dl): Diet 1 - 57, Diet 2 - 57, Diet 3 - 59, Diet 4 - 58 (1 vs 3, p≤0.01; all other p>0.05)
- Triglycerides (mg/dl): Diet 1 - 111, Diet 2 - 107, Diet 3 - 90, Diet 4 - 86 (3 vs 1 or 2, p≤0.01; 4 vs 3, p=0.02; 4 vs 1 or 2, p≤0.01; all other p>0.05)
- SBP (mmHg): Diet 1 - 124, Diet 2 - 124, Diet 3 - 123, Diet 4 - 123 (p>0.05 for all comparisons) [62]
- StraightHealthcare analysis:
- This OmniCarb study surprisingly found that at high carbohydrate intake, a low GI diet decreased insulin sensitivity when compared to a high GI diet. The low GI diet
also increased LDL cholesterol levels. These findings run contrary to the entire premise of a low GI diet.
- The two diets with lower carbohydrate content were associated with lower triglyceride levels
- A number of other studies have looked at the effects of a low GI diet on CVD and CVD risk factors. Results from these studies have been mixed. [32,33,34,35,36,37,38]
- In conclusion, there is no evidence a low GI diet reduces CVD or improves CVD risk factors
Cancer risk
- Insulin and insulin growth factors are thought to play a role in cancer development
- Higher GI foods tend to stimulate higher amounts of insulin secretion
- Because of this relationship, a number of observational studies have looked at the association of GI with the risk of various cancers
- Results from studies have been mixed
- In general, there is no conclusive evidence that higher GI diets increase the risk of any type of cancer [39,40,41,42,43,44,45,46,47,48,49,50]
- The University of Sydney
- The University of Sydney provides a free online, searchable database of GI values for various foods
- The American Journal of Clinical Nutrition
- The American Journal of Clinical Nutrition issued a table in 2002 that lists the GI for over 750 foods
- The table is available as a free downloadable pdf file
- The table lists two GI values for foods because some labs use a 50 gram glucose load as the reference food and other labs use white bread as a reference food [52]
- CRITICISMS OF THE GLYCEMIC INDEX
- The glycemic index has been criticized for the following:
- Variability in the GI value for individual foods can be high
- The GI of a food is affected by other foods that are consumed with that food
- Following a strict, low GI meal plan is difficult and impractical
- Other dietary interventions (ex. carbohydrate counting, decrease calorie intake) have had better results in trials
- Beneficial effects of a low GI diet are mostly unproven in clinical trials [6,54]
- StraightHealthcare analysis:
- Many claims have been made about low GI diets and their effect in various diseases
- In general, these claims are largely unfounded and not supported by good clinical trials
- Diabetics and people who are trying to lose weight should focus on total calorie intake first and foremost
- Diabetics who have volatile, uncontrolled blood sugars despite good calorie and portion control may want to explore the GI of foods they consume. Foods with high GI can be replaced with foods of lower GI.
- What is PMID?
- PI = Manufacturer's Package Insert
- # PMID
- 1 - PMID: 1951155
- 2 - PMID: 17029903
- 3 - PMID: 19160276
- 4 - PMID: 12882846
- 5 - PMID: 11237934
- 6 - PMID: 14963048
- 7 - PMID: 18175767
- 8 - PMID: 8116560
- 9 - PMID: 18202243
- 10 - PMID: 19167958
- 11 - PMID: 12081851
- 12 - PMID: 9020271
- 13 - PMID: 12351467
- 14 - PMID: 15277155
- 15 - PMID: 15505008
- 16 - PMID: 17712022
- 17 - PMID: 17921375
- 18 - PMID: 18039988
- 19 - PMID: 18039989
- 20 - PMID: 17487328
- 21 - PMID: 18326601
- 22 - PMID: 12081852
- 23 - PMID: 10980801
- 24 - PMID: 11988062
- 25 - PMID: 17636786
- 26 - PMID: 16177201
- 27 - PMID: 18308431
- 28 - PMID: 17823436
- 29 - PMID: 19057524
- 30 - PMID: 21105792
- 31 - USDA 2010 recs for CARBS
- 32 - PMID: 15495112
- 33 - PMID: 12081850
- 34 - PMID: 11988062
- 35 - PMID: 12081853
- 36 - PMID: 17556687
- 37 - PMID: 18091767
- 38 - PMID: 19631353
- 39 - PMID: 12488297
- 40 - PMID: 19088152
- 41 - PMID: 12815005
- 42 - PMID: 14652274
- 43 - PMID: 16702366
- 44 - PMID: 17118965
- 45 - PMID: 17921397
- 46 - PMID: 17951595
- 47 - PMID: 18400721
- 48 - PMID: 19336549
- 49 - PMID: 19095757
- 50 - PMID: 18541570
- 51 - PMID: 19056599
- 52 - PMID: 12081815
- 53 - Institute of Medicine Dietary Reference 2005
- 54 - PMID: 12882880
- 55 - PMID: 21193625
- 56 - PMID: 15333500
- 57 - PMID: 11423492
- 58 - PMID: 11023137
- 59 - PMID: 16864756
- 60 - PMID: 17507345
- 61 - PMID: 24357209
- 62 - PMID: 25514303
- 63 - PMID 29222370 - American Diabetes Association Standards of Medical Care in Diabetes 2018