Archway fat-free oatmeal raisin. ⢠Cape Cod reduced fat ... Example: apple butter FII = 52. d f l d. Derived from reci
BEYOND GLYCEMIC INDEX: NEW FOOD INSULIN INDEX Laura Sampson Kent* Mary Franz* Jennie Brand Miller** Walter Willett* *Department of Nutrition, Harvard School of Public Health **University of Sydney Acknowledgement: This research was funded by grant CA55075, National Cancer Institute, National Institutes of Health.
RELEVANCE OF INSULIN SECRETION • Prevention and management of weight gain • Hyperlipidemia • Non‐insulin dependent diabetes mellitus (NIDDM) • Long‐term exposure to high insulin concentrations may be associated with increased risk of cancer y
WHAT IS FOOD INSULIN INDEX (FII)? WHAT IS FOOD INSULIN INDEX (FII)? Directly quantifies the postprandial insulin Directly quantifies the postprandial insulin response to an isoenergetic portion of a test food in comparison to a reference test food in comparison to a reference food Allows testing of foods with no or low Allows testing of foods with no or low carbohydrate content FII has been shown to predict the relative h b h d h l insulin demand evoked by mixed meals
FOOD INSULIN INDEX (FII) FOOD INSULIN INDEX (FII) • Measures Measures blood insulin response to a food blood insulin response to a food compared to a reference food glucose (FII=100) • Measures incremental insulin area under the curve (AUC) over two hours in response to curve (AUC) over two hours in response to consumption of a 1000 kJ portion of the test food divided by the AUC after ingestion of a food divided by the AUC after ingestion of a 1000 kJ portion of the reference food
How is FII Different from Glycemic Index? How is FII Different from Glycemic Index? FOOD INSULIN INDEX FOOD INSULIN INDEX • Measures postprandial increase in insulin secretion of a whole food • Dependent on d carbohydrate, quantity and quality of protein and fat q y p and their interactions
GLYCEMIC INDEX GLYCEMIC INDEX • Measures effect of carbohydrate‐containing foods on postprandial increase in blood‐glucose levels levels • GI is not always proportional to the insulin response
METHODS • APPROXIMATELY APPROXIMATELY 100 FOODS AND 16 CEREALS 100 FOODS AND 16 CEREALS WERE ANALYZED USING GLUCOSE AS THE REFERENCE FOOD • 407 FOODS, 147 MARGARINES, 58 OILS, 111 CEREALS WERE IMPUTED USING VARIOUS ALGORITHMS • 166 FOODS AND 89 CEREALS WERE RECIPE‐ DERIVED • 44 FOODS WERE ASSUMED TO BE 0
FOODS SENT TO UNIVERSITY OF SYDNEY FOODS SENT TO UNIVERSITY OF SYDNEY
MUFFIN MIXES READY‐MADE READY MADE SWEET ROLL
• SunMaid honey raisin bran muffin mix • Duncan Hines blueberry D Hi bl b muffin mix • Krusteaz fat‐free Kr stea fat free blueberry muffin mix • Entenmann cinnamon Entenmann cinnamon buns
Cake Mix Cake Mix Pancake Mix Cookies
Chips p Candy Bars
• Betty Crocker golden cake w/ Betty Crocker choc frosting • Aunt Jemima pancake mix • Chips Ahoy regular chocolate chip • Chips Ahoy reduced fat chocolate Chips Ahoy reduced fat chocolate chip • Archway fat‐free oatmeal raisin • Cape Cod reduced fat potato chips • Snickers candy bars • Hershey candy bars y y
U.S. Analyzed Cereals U.S. Analyzed Cereals CEREALS
• • • • • • • • • •
Cheerios Shredded Wheat Corn Flakes Quaker 100% Natural Cracklin Oat Bran L k Ch Lucky Charms Frosted Flakes Great Grains Great Grains Honey Bunches of Oats Wheaties
Matching U.S. Foods Purchased in Sydney Dairy • • • • • • • •
Cream cheese Low fat cheddar cheese Low fat processed cheese Low fat cottage cheese Low fat ice cream Sherbet 1% milk Butter
F it Fruits or fruit drinks f it d i k • • • • •
Peaches canned in juice Peaches canned in syrup Orange juice g j Prunes Punch
Cereals • •
Kashi 7 grain puffs K hi i ff Rice Bubbles
Meat, fish, mixed dishes • • • • • • •
Tacos Lasagna Roast chicken Shrimp Tofu Tuna in oil Tuna in oil Bacon
Miscellaneous items • • • • • • •
Olive oil Navy beans Raspberry jam Pretzels Walnuts Corn tortillas Jatz crackers (similar to Ritz)
Recipe‐derived Recipe derived FII Values FII Values • Used analyzed foods as ingredients • Used ingredients that were imputed from analyzed foods. y Example: apple butter FII = 52 Derived from recipe using apple, a.j., and df l d sugar as ingredients (all analyzed foods)
IMPUTED OR CALCULATED FII VALUES IMPUTED OR CALCULATED FII VALUES FOODS Cereal grains, baked goods, sweets and snack foods, beverages fruits and dairy beverages, fruits, and dairy
CALCULATION Similar analyzed foods and a carb/1000 kj adjustment carb/1000 kj imputed food x FII carb/1000 kj analyzed food
Breakfast cereals
Used analyzed cereals adjusting for carbohydrate/amount
Meats, fish, and poultry
Directly from similar analyzed foods
IMPUTED FII VEGETABLE VALUES IMPUTED FII VEGETABLE VALUES VEGETABLES • NON STARCHY
CALCULATIONS Recipe from avocado using an h2o ratio
• STARCHY
Recipe from potatoes or carrots using fiber ratios carrots using fiber ratios
• PEAS and LIMA BEANS
Recipe from beans using an fiber ratio
FII PUBLICATIONS FII PUBLICATIONS Bao J, de Jong , g V, Atkinson F, Petocz , , P, Brand‐Miller , JC. Food insulin index: physiologic basis for predicting insulin demand evoked by composite meals Am J Clin Nutr 2009;90:986‐992. meals. Am J Clin 2009;90:986 992 Holt SH, Miller JC, & Petocz Holt SH Miller JC & Petocz P. An insulin index of foods: P An insulin index of foods: the insulin demand generated by 1000‐kJ portions of common foods. Am J Clin Nutr, 66:1264‐67; 1997.
RESULTS of the mixed meal study: Group 1
A
M1
M7
Group 2
B
M8
M2 500
M9 500
M3
M10
Insulin (pmol/L)
M4 400
M5 M6
300
M11
400
M12 M13
300
White Bread
White Bread 200
200
100
100
0
0
0
15
30
45
Tim e (m in)
60
90
120
0
15
30
45
Tim e (m in)
Bao J et al. Am J Clin Nutr 2009;90:986-992.
60
90
120
RESULTS: Correlations A
140 Obse rved insulin respons se (Rela tive white bread =10 0)
r = 0.78, P = 0.0016 120 100 80 60 40 20 0 0 20 40 60 80 100 120 Insulin demand predicted by Food Insulin Index
RESULTS: Correlations B
140 Observe ed insulin n response e (Relativ ve white brread =100))
r = 0.53, P = 0.064 120 100 80 60 40 20 0 0
20
40
60
80
Available a ab e ca carbohydrate bo yd ate co content te t (g)
100
Highest versus Lowest FII Values Highest versus Lowest FII Values • Highest Highest analyzed FII analyzed FII Jelly Beans (120) Pancakes (112) Pancakes (112) Honeydew Melon (95) Potatoes (90) Potatoes (90) Breakfast Cereals (70‐113)
• Lowest Lowest analyzed FII analyzed FII Olive Oil (3) Avocado (5) Avocado (5) Walnuts (6) Tuna (16) Tuna (16) Chicken (20)
MEAN ANALYZED VS MEAN IMPUTED FII DATA
RE AL S
TS
KF AS T CE
S OD FO
SW EE BR EA
AR
CH Y
FR UI T ST
IR Y DA
N EI PR OT
NU TS & LE
GU M
ES
100 90 80 70 60 50 40 30 20 10 0
ANALYZED IMPUTED
LIMITATIONS Although Although imputed FII data are similar to imputed FII data are similar to analyzed data, algorithms used for imputing FII have not been validated. FII FII variability within some food groups variability within some food groups complicates the assignment of imputed values. For example, analyzed FII for dairy foods range from 18 for cream cheese to 86 for fruited yogurt.
CONCLUSION Most vegetables produce negligible insulin response, g p g g p , as measured by FII. Even fats, and protein sources produce largely variable insulin responses, as measured by FII. Imputed FII values track closely to analyzed values. Refined cereals, sweets, and potatoes produce the greatest insulin response, as measured by FII.
DERIVED INSULIN LOAD DERIVED INSULIN LOAD Insulin Load, Glucose = (calories*FII)/100 • the individual average insulinogenic load th i di id l i li i l d (IL) during the past year was estimated f from FFQs by multiplying the FII of each FFQ b lti l i th FII f h food by its energy content and the consumption frequency and summing ti f d i over all reported food items.
DIETARY INSULIN INDEX, DII DIETARY INSULIN INDEX DII
The average dietary DII was calculated by dividing the average Insulin Load, IL, by the I li L d IL b th total daily energy intake total daily energy intake.
RESULTS FROM COHORT STUDIES RESULTS FROM COHORT STUDIES MEN
WOMEN
• Median IL was 840 Median IL was 840 • Median IL was 677 Median IL was 677 • Median DII was 41.7 • Median DII was 42.7
TOP 8 FOODS CONTRIBUTING TO IL TOP 8 FOODS CONTRIBUTING TO IL MEN
WOMEN
Cold Cereal (6.1%) Potatoes (5 7%) Potatoes (5.7%) Dark Bread (4.4%) Skimmed Milk (3 2%) Skimmed Milk (3.2%) Bananas (3.2%) English Muffin (3 1%) English Muffin (3.1%) White Bread (2.9%) Orange Juice (2.8%) Orange Juice (2.8%)
Mashed Potatoes (6.2%) Skimmed Milk (5 7%) Skimmed Milk (5.7%) Cold Cereal (5.3%) D k B d (4 6%) Dark Bread (4.6%) Beef (3.6%) Y Yogurt (3.3%) (3 3%) White Bread (3.1%) English Muffin (2.7%)
TOP 4 FFQ ITEMS explaining the inter‐ individual variation of insulin load d d l f l l d MEN 49% LIQUOR COLD CEREAL BEER WINE
WOMEN 45% LIQUOR WINE YOGURT YOGURT COLD CEREAL
COHORT STUDY RESULTS COHORT STUDY RESULTS In multivariate adjusted regression models, dietary II and IL were not significantly associated with plasma C‐peptide. i ifi tl i t d ith l C tid Participants in the highest quintile of both II and IL had 26% higher triacylglycerol concentrations than participants in the lowest triacylglycerol concentrations than participants in the lowest quintile (ptrend