Low Carb Diets and Diabetes

Low carbohydrate diets
and diabetes
Alan Barclay
BSc(Nutrition); Grad Dip Dietetics; PhD
© Alan Barclay 2015 Australia other countries. All rights reserved
The Headlines
The facts
Carbohydrate definitions
What are carbohydrates1?
Class (Polymerisation)
Sugars (1-2)
Oligosaccharides (3-9)
Sub-Group
Components
Monosaccharides
Glucose, galactose, fructose
Disaccharides
Sucrose, lactose, maltose
Polyols
Erythritol, Glycerol, Lactitol, Maltitol,
Mannitol, Sorbitol, Xylitol
Malto-oligosaccharides
Maltodextrins
Other oligosaccharides
Raffinose, stachyose, fructooligosaccharides
Starch
Amylose, amylopectin, modified starches
Non-starch polysaccharides
Cellulose, hemicellulose, pectins,
hydrocolloids
Polysaccharides (>9)
glucose
starch
1. Carbohydrates in human nutrition. 1998. FAO Food and Nutrition Paper – 66.
Carbohydrates
are an essential nutrient 2,3
• Primarily an energy source for our bodies:
→ the preferred fuel for our brains and nervous
systems, and for our exercising muscles
• Structural (within cells)
• Genetic (RNA and DNA)
• Function of certain proteins (glycoproteins)
• Adds taste, texture and colour to our foods and
drinks.
2. Mann and Truswell. 2002. 3. Wahlqvist, 1988.
Glycemic index (GI)
• compares equal quantities of available
carbohydrate in foods
• is a measure of their effect on blood glucose
levels in 10+ healthy people over a 2 hr period
• is a percentage
• ISO 26642 2010
Glycemic Index (GI): Ranking
Individual food portion4:
Low
Moderate
High
55 or less
56 - 69
70+
Whole day5:
Low
Moderate
High
45 or less
46-59
60+
4. Australian Standard. Glycemic Index of Foods AS4694-2007. Standards Australia. 2007.
5. Brand-Miller, Nutrition & Dietetics 2009; 66: 136–137
Definitions
Glycemic load (GL)
• a function of a food’s glycemic index and its total
available carbohydrate content and defined as:
Glycemic Load = GI (%) x Carbohydrate (g)
• Using an apple as an example:
GI value = 38%; Carbohydrate per serve =15 g
GL = 0.38 x 15 = 6
The GL of a typical apple is 6
Glycemic Load (GL): Ranking
Individual food portion6:
Low
Moderate
High
0-10
11-19
20+
Whole day7:
Low
< 100 g 8,700 kJ/d diet
6. Brand-Miller JC, Holt SHA, and Petocz P. Glycemic load values:2002. Am J Clin Nutr. 2003; 77 (1): 993-5.
7. Livesey et al, AJCN. 2013.
Glycemic Load
Glycemic Load = GI (%) x Carbohydrate (g)
• 1 unit of GL ~ 1 g of glucose
• The higher the GL, the greater the elevation in
blood glucose and insulin levels8.
You can lower the GL of your foods, meals and diet, by:
• Eating less carbohydrate
• Lowering the GI of the carbohydrate you eat
• Both
8. Foster-Powell K, et al. Am J Clin Nutr. 2003; 76 (1): 5-56.
Definition of low → high carbohydrate diets9
• Very low-carbohydrate ketogenic diet:
20-50 g/d or less than 10% of a 8,400 kJ diet
• Low carbohydrate diet:
<130 g/d or <26% of energy from a 8,400 kJ diet
• Moderate Carbohydrate Diet:
130-230g/d, or 26 - 45% of energy from a 8,400
kJ diet
• High Carbohydrate Diet:
> 230g/d or 45% of energy from a 8,400 kJ diet
9. Feinman et al. Nutrition. 2015.
What is the minimum
carbohydrate requirement?
Minimum carbohydrate requirement10
• Adult brains require 140 g of glucose per day
• Red blood cells require 40 g of glucose per day
• Therefore minimum requirement is 180 g glucose / day
• However, gluconeogenesis (lactic acid, amino acids
and glycerol) can supply ~130 g glucose per day
• So absolute minimum is 50 g glucose per day
• However, judgement can be impaired and fetus may be
affected in short-medium term
• Long-term (≥ 2 years) effects not known
10. Mann and Truswell. Essentials of Human Nutrition. 2003
What do the evidence based
guidelines say?
Guidelines for people with diabetes
Goals of nutritional management of
diabetes11
1.) Achieve and maintain blood glucose levels, lipid
and lipoprotein profiles and blood pressure levels in
the normal range or as close to normal as is safely
possible
2.) Prevent, or at least slow, the rate of development
of the chronic complications of diabetes by modifying
nutrient intake and lifestyle
3.) Address individual nutrition needs, taking into
account personal and cultural preferences and
willingness to change
4.) Maintain the pleasure of eating by only limiting
food choices when indicated by scientific evidence
11. Evert et al. Diabetes Care, 2013.
Carbohydrate recommendations11
11. Evert et al. Diabetes Care. 2013
Fat recommendations11
11. Evert et al. Diabetes Care. 2013
Protein recommendations11
11. Evert et al. Diabetes Care. 2013
Dietary Patterns11
11. Evert et al. Diabetes Care. 2013
Macronutrient mix11,12
• US and Canadian Diabetes Associations, 2013
“Evidence suggests that there is not an ideal
percentage of calories from carbohydrate, protein, and
fat for all people with diabetes (B); therefore,
macronutrient distribution should be based on
individualized assessment of current eating patterns,
preferences, and metabolic goals.”
“The ideal macronutrient distribution for the
management of diabetes may need to be
individualized based on individual preferences and
perceived palatability, as several studies suggest that
wide variations can be effective”
11. Evert. Diab Care. 2013; 12. Dworatzek et al. Can Jour Diab, 2013
Is one pattern better than another?
• Systematic review and meta-analysis of 20
RCTs of 7 different dietary patterns following
3,073 people for ≥6 months13
• Low-CHO, low-GI, Mediterranean, and high
protein diets all effective in ↓ HbA1c by 0.120.47% points
“Dietary behaviors and choices are often personal,
and it is usually more realistic for a dietary
modification to be individualized rather than to
use a one-size-fits-all approach for each person.”
13. Ajala et al, . AJCN, 2013.
What about the recent CSIRO study14?
• 6 months RCT; 115 adults with T2DM; 9 dietetic visits
• 14% (57 g) of energy carbohydrate vs 53% (205 g)
• 24.7 g of fibre vs 31.1g
• Saturated : unsaturated fat ratio 0.23 vs 0.48
• No difference in weight, BMI, waist, FM, FFM
• No difference in fasting glucose, HbA1c, BP, TC, LDL
• Maximum glucose, glucose range, MAGE and other
measures of glucose variability improved on very low
CHO diet
• HDL and TG improved on very low CHO diet
14. Tay, et al. Diabetes Care. 2014
What about vegetarian diets?
• Systematic review and meta-analysis of 6 RCTs
including 255 people followed for an average of 24
weeks (range 4-74 weeks)15
• ~10% of energy from fat, 15% protein, 75%
carbohydrate
• Consumption of vegetarian diets was associated with a
significant reduction in HbA1c −0.39 % points
15. Yokoyama Y, et al. Cardiovasc Diagn Ther. 2014
What about studies in people with
Type 1 diabetes16?
• 4 week RCT; 10 adults with T1DM
• 39% (221 g) of energy carbohydrate vs 70% (363 g)
• 11 g of fibre vs 72 g
• Saturated : unsaturated fat ratio 0.46 vs 0.24
• No difference in weight, fasting glucose or HbA1c
• Carbohydrate disposal per unit of insulin, and basal
insulin decreased on the high carb, high fibre diet
• Total and HDL cholesterol decreased on the high carb,
high fibre diet
16. Anderson et al. AJCN. 1991
Why are we able to be flexible in
our carbohydrate
recommendations?
Both amount and type of carbohydrate
affect BGLs
• The total amount of carbohydrate ingested (whether in a
single food or as part of a meal) accounts for 57–65% of
the variability in blood glucose response.
• The type (GI) of carbohydrate explains a similar
amount (60%) of the variance.
• Together (GL), the amount and the type of carbohydrate
accounted for 90% of the total variability in blood
glucose response17, 18.
17. Wolever TM and Bolognesi C. J Nutr 1996;126:2798–2806
18. Wolever TM and Bolognesi C. J Nutr 1996;126:2807–2812
Carb-counting and glycated
hemoglobin19
6 RCT’s. 667 participants
Carb counting meta analysis
Model
Std diff in means and 95% CI
Statistics for each study
Study name
Std diff
in means
Standard
error
Variance
Lower
limit
Upper
limit
Z-Value
p-Value
Trento 2011
-0.551
0.272
0.074
-1.085
-0.017
-2.023
0.043
Scavone 2010
DAFNE 2002
-0.410
-1.015
0.143
0.179
0.020
0.032
-0.690
-1.366
-0.129
-0.665
-2.863
-5.673
0.004
0.000
Laurenzi 2011
-0.539
0.272
0.074
-1.073
-0.006
-1.982
0.047
Gilbertson 2001
0.262
0.201
0.040
-0.131
0.656
1.307
0.191
Kalergis 2000
Fixed
0.688
-0.396
0.376
0.085
0.141
0.007
-0.049
-0.563
1.424
-0.230
1.830
-4.671
0.067
0.000
Random
-0.300
0.226
0.051
-0.743
0.144
-1.325
0.185
-2.00
-1.00
Fav ours Carb count
0.00
1.00
2.00
Fav ours Control
HbA1c changes
• In adults using carbohydrate counting, there
was a -0.4% improvement in HbA1c (p = 0.048)
19. Bell, Barclay, Petocz, Colagiuri, Brand-Miller. Lancet Diabetes Endocrinol. 2014.
GI and BGLs in children with type 120
Randomised cross-over trial, 20 Australian children aged 7-17 yrs.
High GI (84) diet
Low GI (48) diet
20. Ryan et al. Diabetes Care. 2008 ;31(8): 1485-90
GI and glycated hemoglobin21
11 RCT’s. 402 participants
• Significantly less hypos (difference of -0.8 episodes per patient per month, P < 0.01).
21. Thomas and Elliot. The Cochrane Collaboration, 2009.
GL is the most important predictor of
insulin response22
10 healthy young people consuming 121 types of single
foods (study 1) and 13 mixed meals (study 2):
For foods: GL was a more powerful predictor of insulin
response than GI or available carbohydrate - explaining
59%, 55% and 49%, respectively (P=0.005).
For mixed meals: GL was the only predictor of
postprandial insulin response, explaining 46% (P = 0.01)
of the variation.
Carbohydrate content alone predicted the insulin
responses to single foods (P<0.001) but not to mixed
meals.
22. BAO et al. AJCN. 2011;93:1-13
Food Insulin Index23
23. Bao et al. AJCN. 2009 ; 89:97–105.
Are we eating too much
carbohydrate?
Macronutrient consumption in Australia 1995–2011/224,25
National representative surveys of Australians:
14,000+ in 1995 and 12,000+ in 2011/2
24. ABS. National Nutrition Survey. Com of Aust, 1998; 25 ABS. Australian survey. Com of Aust, 2014..
People with diabetes in Australia26
People with diabetes
Energy (kJ)
Protein (g)
Fat (g)
Mean
8,350
97*
78
Saturated fat (g)
30
Poly fat (g)
14
Mono fat (g)
27
Total carbohydrate (g)
214
% energy
19.7
34.6
13.3
41
Starch (g)
102
Sugars (g)
109
21
Alcohol (g)
4
1.4
Fibre (g)
29
GI
55
GL
122
26. Barclay et al. BJN. 2006; 96, 117–123
*1.67 times Australia RDI
Too many carbs, or wrong type?
We already eat a “moderate carbohydrate” diet:
45 % of kJs = 230 g a day for average weight stable
Australian adult in 1995 NNS
43.5% of kJs = 222 g a day in 2011-12 NHS
We do eat the wrong type27, 9
Average daily GI 55 and GL 132
Aim for GI ~ 45 and GL <100
Lower the populations GL by ↓GI not carbs?
27. Barclay, Petocz, Flood, Brand-Miller, AJCN. 2008;87:627-37. 8. Livesey et al, AJCN, 2013
Recommendations into
practice
Healthy eating principles
• Choose the healthiest option within each food
group or category
• Key nutrients to consider:
Energy (kJs)
< 2,200 kJ per serve for mains
< 600 kJ per serve for snacks
Saturated fat (g) < 28% of total fat
Carbohydrate (g) < 60 g per serve for mains
< 30 g per serve for snacks
Fibre
> 4 g per serve
GI
≤ 55
Use the “swap it, don’t stop it” approach to choose
the right type of carbohydrate
• Simply swap
low GI carbs for
high GI carbs
within each food
group or
category
• This principle
will lower the
dietary GL
Sample meal plan for overweight adult with
type 2 diabetes
Breakfast
Lunch
Dinner
45g Traditional muesli
2 x toast
1 1/3 Cup Spaghetti
1/2 cup milk ( fat)
1 x toast
+ Canola margarine
+ Canola margarine
210 g NAS Baked
Beans
Bolognaise Sauce
+ 2 cups salad
+ vinaigrette
Morning Tea
Afternoon Tea Supper
Fruit biscuits (2 small)
200 mL yoghurt
(NNS)
Fruit (eg, 1 sml Apple)
Nutrient analysis of sample meal plan
• Comparison to key recommended intakes:
Protein
=
64 g (0.76g/kg or 19% of energy)
Total Fat
=
27 g (18 % of energy)
Saturated Fat
=
6 g (4% of energy)
Carbohydrate
=
200 g (56% of energy)
Fibre
=
30 g (5g/1000kJ)
GI
=
45 (~45 / day)
GL
=
90 (<100 / day)
Sodium
=
1,137 mg (920 – 2300)
Sample Cantonese menu plan – modified
Nutrient analysis of Cantonese menu plan
Sample Greek menu plan - modified
Nutrient analysis of Greek menu plan
Sample South Indian menu plan
Nutrient analysis of South Indian menu plan
Summary
• Healthy eating is an important part of diabetes
management
• The amount and type of carbohydrate consumed
are the strongest predictors of an individuals BGLs
• On average, Australians do not consume high
carbohydrate diets, and there is evidence that people
with diabetes consume less, but the GL is too high
• Help people to maintain the pleasure of eating by
only limiting food choices when indicated by strong
scientific evidence
• The amount and type of carbohydrate consumed
should be influenced by the individuals cultural
background and personal food preferences
Declaration of interests
I am a Board member and
consultant to the GI Foundation.
I am a co-author of the Diabetes
and Pre-diabetes Handbook /
Managing Type 2 Diabetes and
The Ultimate Guide to Sugars and
Sweeteners.
Further information
Alan Barclay
+61 (0)2 9785 1037
+61 (0)416 111 046
[email protected]