Ketogenic Diet for

Type 2 Diabetes
Dr. Sarah Hallberg DO, MS
Purpose

 Examine the current treatment limitations

for patients with type 2 diabetes
 Research on carb restriction for diabetes
 Discuss a ketogenic diet
 Ongoing study at Indiana University Health
The Problem

2012
14.3% of all adults had diabetes
38% have pre-diabetes
Now OVER 50% (52.3%) have diabetes or pre-diabetes

JAMA. 2015;314(10):1021-1029. doi:10.1001/jama.2015.10029

The problem
The Solution……
Medical Nutrition Therapy Goals

1. Promote Healthy eating patterns

 achieve and maintain body weight goals
 attain individualized glycemic, blood pressure, and lipid goals
 delay or prevent the complications of diabetes

2. Address individual nutrition needs

 personal and cultural preferences
 health literacy
 access
 willingness
 Medical Nutrition Therapy Goals
3. Maintain pleasure of eating
4. To provide an individual with diabetes the practical
tools for developing healthy eating patterns rather
than focusing on individual macronutrients, micronutrients,
or single foods.
 ADA Nutrition Guidelines 2017

There is no single ideal dietary distribution of

calories among carbohydrates, fats, and proteins
for people with diabetes, macronutrient
distribution should be individualized while
keeping total calorie and metabolic goals in mind.
About those Macronutrients…….
Carbohydrates
ADA 2017 Guidelines
“Studies examining the ideal amount of carbohydrate intake
for people with diabetes are inconclusive, although
monitoring carbohydrate intake for people and considering
the blood glucose response to dietary carbohydrates are key
for improving postprandial glucose control. “
“Carbohydrate intake has a direct effect on postprandial
glucose levels in people with diabetes and is the primary
macronutrient of concern in glycemic management”

Carbohydrates will cause blood sugar to rise.

“Monitor Carbohydrate Intake”
Take away this
glass and then
“monitor”
this child
ADA 2017 Guidelines
“Couple insulin administration with
carbohydrate intake”

ok….
 This “couples” with a lot of insulin

This does not

Am J Clin Nutr March 2002 vol. 75 no. 3 505-510

 Nutrition Therapy Recommendations for
the Management of Adults with Diabetes

“total amount of carbohydrate eaten is

the primary predictor of glycemic
response.”

Diabetes Care Jan 2014, 37 (Supplement 1) S120-S143; DOI: 10.2337/dc14-S120

 ADA 2017 Guidelines
“For people whose meal schedules or carbohydrate
consumption is variable, regular counseling to help them
understand the complex relationship between carbohydrate
intake and insulin needs is important”

Complex???
Eat carbs.
Blood sugar rises.
Insulin needs go up.
Simple.
The
carbohydrate:protein:fat
ratio of the control diet
was 55:15:30
The test diet ratio was
20:30:50
 ADA 2017 Guidelines
“Whole grain consumption was not associated with
improvements in glycemic control in the 2 diabetes”

This makes
sense. It
still causes
glucose to
rise.

Katri S Juntunen et al. Am J Clin Nutr 2002;75:254-262

Proteins
ADA 2017 Guidelines
Some research has found successful
management of type 2 diabetes with meal plans
including slightly higher levels of protein (20 -
30%, which may contribute to increased satiety.
Fats
ADA 2017 Guidelines
“The ideal amount of dietary fat for individual with
diabetes is controversial. The IOM has defined an
acceptable macronutrient distribution for total fat for
ALL adults to be 20 - 35% of energy.”

“Lowering total fat intake infrequently improved

glycemic control or CVD risk factors in clinical trials
involving patients with diabetes”
Mediterranean/MUFA
2017 Standards of Care
5 studies cited that show a Mediterranean diet can
improve A1c.
1 did not have glycemic control as primary endpoint
2 did not show a difference in glycemic control
1 showed low carb did better
Meta-analysis looked at developing diabetes only
Mediterranean/MUFA
Nutrition Therapy Recommendations for the
Management of Adults with Diabetes 2014
6 RCT that included people with diabetes reported improved
glycemic control
2 showed no difference
1 looked at only new diagnosis improvements
1 did not have glycemic control as a primary endpoint
1 showed that the low carb mediterranean is superior to both
traditional mediterranean and ADA standard
1showed low carb had better A1c improvements than
Mediterranean and low fat.
WHAT IS A MEDITERRANEAN DIETARY
PATTERN?
Dietary Patterns
A variety of eating patterns are
acceptable:
Mediterranean
DASH
 2017 - 2 references of which neither looked at glycemic control.
 2014 - Only 1 small study reviewed
Plant-based
 2017 - review showed that plant based is just as good as standard
treatment, not consistently better.
 2014 - “did not consistently improve glycemic control or CVD risk
factors”
The 2014 Nutrition review did look
at low carb as an eating pattern
11 studies reviewed
7 showed clear advantage for low carb
4 did not show a difference
 2 showed greater med reduction in LC “These limited data suggest
that changes in medication may have masked a potential in the lo-
carb dietary advice group to have m a more positive impact on
glycemic control”
 1 study was on its with and without diabetes and there was a
statistically significant advantage within the patients with diabetes
 1 was tried to do low fat and low carb
AACE
AACE
American Association of Diabetes Educators
Healthy Eating Patters

 My Plate
 DASH
 Mediterranean
 Vegetarian or Vegan
Common Thread

LOW FAT
Macronutrients

Fat – low
Protein - ??
Carbs – must be high(er)
“total amount of carbohydrate
eaten is the primary predictor
of glycemic response.”
 From ADA Website 3/3017

depends on many things including how active you are and what, if any, medicines you take. Some people are active and can eat more

depends on many things including how active you are and what, if any, medicines you take. Some people are active and can eat more

y, and lower your risk of diabetes complications.

ohydrate at a meal. You may need more or less carbohydrate at mea

much carb to eat at a meal, choose your food and the portion size to match.
The way it used to be
The origin of low carb for diabetes

The eighth patient described in his ledger had special

significance for Elliott Joslin: it was his mother, Sara
Proctor Joslin, diagnosed with diabetes in 1900. The
disease was considered uniformly fatal, but Sara Joslin
lived an astonishing 13 years after her diagnosis. She
followed her son's instructions to eat a low-carbohydrate,
high-fat diet.
“The discovery of insulin was a

severe setback to the advancement

of the science and art of

nutrition” Louis Newburgh 1936

OOPS IS RIGHT
sociation. “Grains, beans, and starchy vegetables form the foundation of the Diabetes Food Pyramid. The message is to eat more of the

rmful to them instead?

Another solution
Ketogenic Diet

 Low Carbohydrate (under 50gr of total carbs often under

30gr)
 Adequate Protein – NOT high
 High Fat including saturated fat but also high MUFA
Carb Sources

 Non-starchy Vegetables
 Nuts and Seeds
 Limited Berry Fruit
 Dairy
 NO Grains, potatoes or sugar
 A Day in the Low Carb Life
Breakfast Snacks
black coffee 2 oz mixed nuts, broth
4 sausages 2 oz soft cheese with
6 oz celery

Lunch Dinner
2 cups mixed greens 8 oz tomato bisque
6 oz water pack tuna 8 oz steak
10 black olives 4 oz buttered green beans
½ cup blue cheese dressing 4 oz sauteed mushrooms
(yogurt, olive oil) 4 oz maple walnut ice cream
(made w/ sucralose/xylitol)

Total: 2100 kcal fat, 600 protein, 150 carbs (74% fat, 5% carb, 21% protein)
We can use ketones for fuel
And the science says…..
 Weight Loss with a low-carbohydrate, Mediterranean,
or low-fat diet
• 2 year RCT with 322 moderately
obese subjects
• Adherence was 84.6% at 2 years
• Subgroups of 17 -19 met with
dietitians
• 1, 2, 5, 7 weeks and ten 6 week
intervals for 18 sessions of 90 min
each

N Engl J Med 2008; 359:229-241

Weight Loss with a low-carbohydrate,
Mediterranean, or low-fat diet
 Low fat
 AHA Guidelines calorie restricted 30% fat
 Low Carb
 No calorie restriction
 20g carbs per day for 2 months
 Then increased to max of 120gr for maintenance
 Mediterranean
 Calorie restricted<35%fat
 Rich in veggies and low in red meat

N Engl J Med 2008; 359:229-241

Weight Loss with a low-carbohydrate,
Mediterranean, or low-fat diet
 36 patients had diabetes
 Fasting glucose decreased only in Mediterranean group
 Glucose went UP in the low fat group
 HOMA-IR decreased in low carb and Mediterranean – more
in Mediterranean
 At 24 months changes in A1c was only significant in the low
carb group 0.9+0.8%

N Engl J Med 2008; 359:229-241

Effect of a low-carbohydrate diet on appetite, blood glucose
levels, and insulin resistance in obese patients with type 2
diabetes

 10 Obese patients with Type 2 Diabetes

 Hospitalized for testing period
 Day 1 – 7 Usual diet
 Day 8 – 21 Low-Carbohydrate Diet 21 g/day with ad lib fat
and protein

Ann Intern Med 2005;142:403-411

Effect of a low-carbohydrate diet on appetite, blood
glucose levels, and insulin resistance in obese patients with
type 2 diabetes

Figure 1
Glucose and Insulin response for
patients with type 2 diabetes on low
carbohydrate diet vs. control. Data
(means ± SE) are for 9 patients with
type 2 diabetes after seven days on
their usual high-carbohydrate diet
(control) and after 2 weeks) on a
low-carbohydrate diet. Medication
was reduced in 4 patients and
discontinued in one during the low-
carbohydrate diet. Figure redrawn
from Boden, et al. [8].
Ann Intern Med 2005;142:403-411
A randomized pilot trial of a moderate carbohydrate diet
compared to a very low carbohydrate diet in overweight or
obese individuals with type 2 diabetes mellitus or prediabetes

 RCT 34 patients with type 2 diabetes or prediabetes

 ADA low fat diet calorie-restricted
 Low Carb High Fat Diet – 20 – 50 total grams carbs – no
calorie restriction
 Each group only lost 1 participant
 All attended 13 – 2 hour classes
. PLoS ONE 9(4): e91027
 Figure 2. Change in HbA1c by diet group.

PLoS ONE 9(4): e91027.

The effect of a low-carbohydrate, ketogenic
diet versus a low-glycemic index diet on
glycemic control in type 2 diabetes mellitus
 RCT 84 Patients with type 2 Diabetes and Obesity
 Very Low Carb <20 total gram per day – NO calorie restriction
 Low GI diet 55% carbs and 500kcal reduction from baseline intake
 Twenty of 21 (95.2%) LCKD group participants had an elimination or reduction
in medication, compared with 18 of 29 (62.1%) LGID group participants (p <
0.01).
 From baseline to 24 weeks, the reduction of mean ± SD hemoglobin A1c was
greater for the LCKD group
 LCKD - 8.8 ± 1.8% to 7.3 ± 1.5%, p = 0.009, within group change, n = 21
 LGID group - 8.3 ± 1.9% to 7.8 ± 2.1% p = NS, within group change, n = 29
 between groups comparison p = 0.03
Westman, Eric C et al. “The Effect of a Low-Carbohydrate, Ketogenic Diet versus a Low-Glycemic Index Diet on
Glycemic Control in Type 2 Diabetes Mellitus.”Nutrition & Metabolism 5 (2008): 36. PMC. Web. 29 Apr. 2016.
Beneficial Effects of Ketogenic Diet in Obese Diabetic Subjects

N = 64
Demographics:
● Non-diabetic and diabetic obese subjects
● BMI > 30
● Two Groups:
○ Fasting BG > 6.1 mmol/l (Group I; n = 31)
○ Normal fasting BG (Group II; n = 33)
● Age: 46.4 ± 9.4 (Group I); 40.0 ± 11.4 (Group II)
Method:
● Outpatient; 56 weeks
● Ate 20 g of carbs and 80–100g protein for 12 weeks, then 40g
carbs for 12-56 weeks
Results as %∆ for Total (Group I, Group II):
● Fasting BG: –31.0% ± 25.0 (–50.9% ± 12.5; –7.4% ± 11.9)
● Triglycerides: –59.0% ± 32.0 (–40.8% ± 38.0; –40.8% ± 38.0)
● LDL: –28.2% ± 20.1 (–33.0% ± 20.4; –22.9% ± 18.7)
● HDL: 52.3% ± 43.8(63.4% ± 51.1; 39.8% ± 30.0
● Weight: –25.8% ± 6.4 (–24.4% ± 6.7; –27.2% ± 6.0)

Source: Beneficial effects of ketogenic diet in obese diabetic subjects; Dashti HM et al.; Mol Cell Biochem (2007) Aug; 302(1-2):249-56
Beneficial Effects of Ketogenic Diet in Obese Diabetic Subjects

Source: Beneficial effects of ketogenic diet in obese diabetic subjects; Dashti HM et al.; Mol Cell Biochem (2007) Aug; 302(1-2):249-56
Beneficial Effects of Ketogenic Diet in Obese Diabetic Subjects

Source: Beneficial effects of ketogenic diet in obese diabetic subjects; Dashti HM et al.; Mol Cell Biochem (2007) Aug; 302(1-2):249-56
A Very Low-Carbohydrate, Low–Saturated Fat Diet for
Type 2 Diabetes Management: A Randomized Trial

115 obese adults with T2DM

Randomized to low-carbohydrate or high carbohydrate diet
Both diets hypo-caloric
Both diets <10% saturated fat
Low carb = 14% carbs (<50gr), 28% protein
High carb 53% carbohydrate 17% protein
24 weeks
included structured exercise

Diabetes Care Nov 2014, 37 (11) 2909-2918; DOI: 10.2337/dc14-0845

A Very Low-Carbohydrate, Low–Saturated Fat Diet for
Type 2 Diabetes Management: A Randomized Trial

Diabetes Care Nov 2014, 37 (11) 2909-2918; DOI: 10.2337/dc14-0845

A Very Low-Carbohydrate, Low–Saturated Fat Diet for
Type 2 Diabetes Management: A Randomized Trial

LC reduced HbA1c to a greater extent among participants

with baseline HbA1c >7.8
No diet effect in participants with baseline HbA1c < 7.8
Med reduction different
Percentage weight loss was not different between the groups
with baseline HbA1c >7.8

Diabetes Care Nov 2014, 37 (11) 2909-2918; DOI: 10.2337/dc14-0845

A Very Low-Carbohydrate, Low–Saturated Fat Diet for
Type 2 Diabetes Management: A Randomized Trial

LC participants were 85% more likely to spend

higher proportions of time in the euglycemic range
LC participants were 56% less likely to spend higher
proportions of time in the hyperglycemic ranges
LC participants were 16% less likely to spend more
time in the hypoglycemic range

Diabetes Care Nov 2014, 37 (11) 2909-2918; DOI: 10.2337/dc14-0845

An Online Intervention Comparing a Very Low-Carbohydrate
Ketogenic Diet and Lifestyle Recommendations Versus a Plate
Method Diet in Overweight Individuals With Type 2 Diabetes:
A Randomized Controlled Trial
 25 patients randomized to
 Create Your Plate from ADA
 Ketogenic diet (25 - 50gr total carbs per day)
 32 weeks
 online intervention in both groups
 drop out rate for intervention 0% and 39% in control group
 twice the A1c reduction in the intervention group

J Med Internet Res 2017;19(2):e36

J Med Internet Res 2017;19(2):e36
Carbohydrate Restriction has a More Favorable Impact
on the Metabolic Syndrome than a Low Fat Diet

N = 40
Demographics:
● 40 overweight subjects with atherogenic dyslipidemia
● Age: 18 – 55 years
● BMI > 25 kg/m2
Method:
● Outpatient; 12 weeks
● Two randomly assigned groups:
○ Hypocaloric LCD: ~1,500 kcal; 12% carb, 59% fat, 28%
protein
○ Hypocaloric LFD: ~1,500 kcal, 56% carb; 24% fat; 20%
protein
Results as %∆ for CRD vs LFD:
● Fasting BG: -12% vs -2%
● Fasting Insulin: -49% vs -17%
● Insulin Resistance: -55% vs -18%
● TAG: -51% vs. -19%
● HDL-C: 13% vs -1%
● TAG/HDL-C ratio: -54% vs -20%
● Weight: -12% vs. -6%

Source: Carbohydrate Restriction has a More Favorable Impact on the Metabolic Syndrome than a Low Fat Diet; Volek et al.; Lipids. 2009 Apr; 44(4):297-309
Nitrogen metabolic and insulin requirements in
obese diabetic adults on a protein-sparing modified
fast
 6 obese diabetics requiring insulin were admitted as
inpatients for the study
 7th patients was treated as outpatient
 Protein sparing modified fast (0.8 – 1.5gr/kg)
 Nitrogen balance studies on 3 patients

Diabetes, 1976 vol 25, no 6; 494-504

Nitrogen metabolic and insulin requirements in
obese diabetic adults on a protein-sparing modified
fast
 Nitrogen balance was obtained with at least 1.3gr/kg of IBW
 Insulin stopped in all patients 0 – 19 days
 5 of the 7 sustained weight loss of over 40 pounds
 Hyperinsulinism will tend to establish a positive caloric balance in adipose
tissue, as insulin favors glucose and triglyceride uptake while inhibiting
mobilization of fat. The realization that hyperinsulinism converts adipose
tissue into a sink for calories implies that weight-reducing regimens should be
specifically designed to allow a sharp fall in insulin levels. Thus, the
avoidance of carbohydrates assumes special importance in the fasting obese
given this rend towards hyperinsulinism not only because of the anabolic effect
of insulin on adipose tissue but because hyperinsulinemia per se produces
insulin resistance.
Diabetes, 1976 vol 25, no 6; 494-504
Systematic review and meta-analysis of different dietary
approaches to the management of type 2 diabetes

 20 RCT’s
 n = 3073
 The low-carbohydrate, low-GI, Mediterranean, and high-
protein diets all led to a greater improvement in glycemic
control [glycated hemoglobin reductions of −0.12% (P =
0.04), −0.14% (P = 0.008), −0.47% (P < 0.00001), and
−0.28% (P < 0.00001), respectively] compared with their
respective control diets

Am J Clin Nutr March 2013 vol. 97 no. 3 505-516

Systematic review and meta-analysis of different
dietary approaches to the management of type 2
diabetes
 Conclusion: Low-carbohydrate, low-GI, Mediterranean, and
high-protein diets are effective in improving various
markers of cardiovascular risk in people with diabetes and
should be considered in the overall strategy of diabetes
management.

Am J Clin Nutr March 2013 vol. 97 no. 3 505-516

 Forest plots that show differences in Hb A1c between low-carbohydrate and other diets (A), low-GI
and other diets (B), Mediterranean and other diets (C), and high-protein and other diets (D).

Olubukola Ajala et al. Am J Clin Nutr 2013;97:505-516

©2013 by American Society for Nutrition

 Low-carbohydrate condition. Participants in the low-
carbohydrate condition were provided with the CalorieKing
Thirty grams of carbohydrate was specifically chosen as the
target intake, a goal we had used in our previous study (1).
Although the glycemic index was not specifically discussed,
participants were encouraged to select whole grain
products and foods with a high fiber content. Participants
were not instructed to restrict their total fat or caloric
intake, although general advice was provided on the various
types of dietary fat. They were encouraged to consume
healthy fats (e.g., monounsaturated and polyunsaturated)
and to minimize the intake of saturated and trans fats.
Urinary or plasma ketones were not measured to evaluate
dietary adherence.

Iqbal, N., et al (2010), Effects of a Low-intensity Intervention That Prescribed a Low-carbohydrate vs. a Low-fat Diet in
Obese, Diabetic Participants. Obesity, 18: 1733–1738. doi:10.1038/oby.2009.460
 The Canadian Trial of Carbohydrates in Diabetes (CCD), a 1-y controlled
trial of low-glycemic-index dietary carbohydrate in type 2 diabetes: no
effect on glycated hemoglobin but reduction in C-reactive protein

 The high-GI, low-GI, and low-CHO diets contained, respectively, 47%, 52%, and
39% of energy as carbohydrate and 31%, 27%, and 40% of energy as fat;
 For the low-CHO diet, key foods consisted of olive or canola oils or spreads,
nuts, and other foods low in SFAs and high in MUFAs and known to be
associated with reduced risks of diabetes and CVD (33-35) or known to reduce
blood lipids (16, 36, 37). These foods replaced carbohydrate foods normally
consumed and were prescribed in amounts sufficient to raise total fat intake
by ≈10%.

Am J Clin Nutr January 2008 vol 87 no. 1 114 - 125

 Macronutrients, food groups, and eating patterns in the management
of diabetes: a systematic review of the literature, 2010

 In studies reducing total carbohydrate intake, markers of

glycemic control and insulin sensitivity improved, but studies
were small, of short duration , and in some cases were not
randomized or had high dropout rates.

Diabetes Care. 2012 Feb;35(2):434-45

Prevention and management of type 2 diabetes:
dietary components and nutritional strategies

A meta-analysis of RCTs suggested that various dietary patterns

such as low-carbohydrate, low-GI, Mediterranean, and high-protein
diets were effective in improving glycemic control and CVD risk
factors compared to diets in diabetic patients.87 These results
provide a range of dietary options for diabetes management,
paying attention to overall diet quality, treatment goals, and
personal and cultural food preferences.

Several low-fat vegetarian or vegan diet trials have

been conducted in people with diabetes,87 but
improved glycemic control or CVD risk was not
consistently reported in these studies

Ley SH et al. Lancet (London, England). 2014;383(9933):1999-2007. doi:10.1016/S0140-6736(14)60613-9.

Ongoing study at Indiana University
Health
 400 treatment patients: All treated with a ketogenic diet. Blood
ketones obtained initially daily then decreasing frequency.
 200 patients treated “live” in the clinic with weekly group
meetings for the first 3 months then decreasing frequency over 2
years
 200 patients treated “virtually” with all education being done via
portal. Follow ups only at 3 months, 1yr and 2yrs or as needed

 100 controls
 Treated by the IU Health dietitians with ADA protocol
Ongoing study at Indiana University
Health
 All treatment patients given health coach
 Biomarker tracking in “app”
 Medication adjustment by supervising physician based on daily blood
glucose and blood pressure levels
 Online support community
Ongoing study at Indiana University
Health

Primary Outcomes
 Body Weight
 Metabolic Syndrome Criteria
 Type 2 Diabetes Status
Ongoing study at Indiana University
Health

Secondary Outcomes
 • Carotid intima media thickness (cIMT) will be measured
by ultrasound 3 times over 2 years (baseline, 12, and 24
months).
 • Serum lipids analyzed by NMR to determine LDL particle
size and number
 • Full Body DEXA
 Banked Samples
Time for a Paradigm Change
Review BHOB SGLT-2 data
 Multiple randomized
controlled trials including patients
with type 2 diabetes have reported that
a Mediterranean-style eating pattern
(75,79–82), rich in monounsaturated
fats, can improve both glycemic control
and blood lipids.

l in type 2 diabetes1,2,3,4
of a high-monounsaturated fatdiet with a high-carbohydrate diet in type 2 diabetes.Diabetes Care 2009;32:215–220 The study showed

hed in monounsaturated fattyacids and conventional diet on weight loss and metabolic parameters in obese non-diabetic and type 2 di
Macronutrients

Registered dietitians (RDs) should encourage consumption

of macronutrients based on the Dietary Reference
Intakes for healthy eating as research does not support
any ideal percentage of energy from macronutrients in
meal plans for persons with diabetes.
Of the 5 studies included in the initial assessment:
3 clearly showed an advantage for low carb
1 did not test low carb
1 should not have been included in the review as it did not meet the specified inclusion criteria.
1 study was left out that should not have been