Kate Neuhausen, MD
Family and Community Medicine
Resident at UCSF/SFGH
September 3, 2008
Intro – Low-Fat Diet
American Heart Association guidelines
Energy intake 1500 kcal per day for women and 1800
kcal per day for men
Distribution of calories: 30% from fat, 10% from
saturated fat
Intake of 300 mg cholesterol per day
Dietary recommendations:
Consume low-fat grains, vegetables, fruits, and legumes
Limit additional fats, sweets, high-fat snacks
Source: http://www.vpul.upenn.edu/ohe/library/fitness/foodpyramid.htm
Intro – Mediterranean Diet
Moderate-fat, restricted-calorie
Energy intake 1500 kcal per day for women and 1800
kcal per day for men
Goal distribution less than 35% calories from fat
Dietary recommendations:
Consume vegetables, fish, and poultry
Limit red meat in diet
Consume 30 to 45 g of olive oil and handful of nuts (5-7
nuts) per day
http://www.revolutionhealth.com/healthy-living/food-nutrition/food-groups-pyramids-overviews.com
Intro – Low-Carbohydrate Diet
Non-restricted calorie diet (Atkins diet)
Goal consumption: 20 g of carbohydrates per day for 2-
month induction phase with gradual increase to max
of 120 g per day to maintain weight loss
No limit on intake of total calories, protein, and fat
Dietary recommendations:
Choose vegetarian sources of fat and protein
Avoid trans fat
Source: http://thedivinelowcarb.blogspot.com/2007/07/anti-food-dilemma.html
Source: http://www.revolutionhealth.com/healthy-living/food-nutrition/food-groups-pyramids-overview
Source: http://www.revolutionhealth.com/healthy-living/food-nutrition/food-groups-pyramids-overviews.com
Methods - Patients
July 2005 – July 2007
On-site workplace medical clinic in Dimona, Isreal
Men and women ages 40-65
BMI at least 27 or presence of Type 2 DM or coronary heart
disease
Exclusion Criteria:
Pregnancy or breastfeeding
Cr > 2
Liver dysfunction (greater 2x increase AST/ALT)
GI problems preventing adherence to test diet
Cancer
Methods – Study Design
Two-year Randomized Controlled Trial
Efficacy trial – optimal circumstances
Participants randomly assigned to three diet groups:
low-fat, low-carbohydrate, and Mediterranean
Block randomization by sex, age, BMI, history of
coronary heart disease, history of Type 2 DM, and
current use of statins
Methods – Study Design
Intensive dietary counseling with nutritionists for total
of 18 session of 90 minutes each
Motivational phone calls lasting 10-15 min by dietitians
six times during 2-year interval with participants
having difficulties adhering to diet
Labeling of all meals in cafeteria with number of
calories, grams of carbs, fat, and saturated fat with
color coding according to diet group
Adherence evaluated by validated electronic food-
frequency questionnaires self-administered at
workplace at baseline, 6, 12, 24 months of follow-up
Methods – Study Design
Participants were weighed without shoes every month
BMI was calculated, and waist circumference
measured monthly
BP was measured every 3 months with automated
system after 5 minutes of rest
Blood samples obtained at 8 AM after 12-hour fast at
baseline, 6, 12, and 25 months
Methods - Outcomes
Weight Loss
BMI
Waist circumference
BP
HbA1c
Fasting lipids: LDL, HLD, triglycerides, cholesterol
Fasting plasma glucose and plasma insulin
Biomarkers: CRP, adiponectin, leptin
Results – Adherence
Baseline characteristics:
Mean age 52 years
Mean BMI 31
86% men
Rates of adherence:
Overall: 95.4% at 12 months; 84.6% at 24 months
Low-fat: 90.4% at 24 months
Mediterranean: 85.3% at 24 months
Low-carbohydrate: 78.0% at 24 months
Medication changes during study not significantly different
among groups:
20 participants initiated BP treatment
5 initiated meds for glycemic control; 1 reduced dosage
4 started and 3 stopped cholesterol-lowering therapy
Results – Dietary Intake at 24 months
Low-fat group: lower intake of saturated fat
Energy change:-572.5 kcal/day from baseline
% Carbs: 50.7% (-82.8 g/day)
% Protein: 19.0% (-19.8 g/day)
% Fat: 30.0% (-18.9% g/day)
% Saturated fat: 9.6% (-6.2 g/day)
Change in monounsaturated: saturated fat ratio: 0.02
Change in cholesterol: -94.2 mg/day
Change in dietary fiber: -4.7 g/day
Level of physical activity: 21.4 MET/wk
Detectable urine ketone bodies: 4.8% participants
Results – Dietary Intake at 24 months
Mediterranean group: higher ratio of mono-saturated
to saturated fat and higher intake of dietary fiber
Energy change:-371.9 kcal/day from baseline
% Carbs: 50.2% (-50.5 g/day)
% Protein: 18.8% (-17.5 g/day)
% Fat: 33.1% (-10.5 g/day)
% Saturated fat: 9.6% (4.6 g/day)
Change in monounsaturated: saturated fat ratio: 0.11
Change in cholesterol: -94.6 kg/day
Change in dietary fiber: +0.29 g/day
Level of physical activity: 15.6 MET/wk
Detectable urine ketone bodies: 2.8% participants
Results – Dietary Intake at 24 months
Low-Carb group: lower intake of carbs, higher intakes
of protein, total fat, saturated fat, and total cholesterol
Energy change: -550 kcal/day from baseline
% Carbs: 40.4% (-129.8 g/day)
% Protein: 21.8% (-6.9 g/day)
% Fat: 39.1% (-1.7 g/day)
% Saturated Fat: 12.2% (+0.56 g/day)
Change in monounsaturated: saturated fat ratio: -0.01
Change in cholesterol: +6.51 kg/day
Change in dietary fiber: - 10.o g/day
Level of physical activity: 16.3 MET/wk
Detectable urine ketone bodies: 8.3% participants
Results – Weight Loss at 24 months
Phase of maximum weight loss from 1-6 months
Maintenance phase from 7-24 months
Overall weight loss among 322 participants at 24 months:
Low-fat group: -2.9 +/- 4.2 kg
Mediterranean: -4.4 +/- 6.0 kg
Low-carb group: -4.7 +/- 6.5 kg
Weight loss among 277 male participants at 24 months:
Low-fat group: -3.4 kg
Mediterranean : -4.0 kg
Low-carb group: -4.9 kg
Weight loss among 45 female participants at 24 months:
Low-fat group: -0.1 kg
Mediterranean: -6.2 kg
Low carb group: - 2.4 kg
Weight Change during 2 years according to diet group
Source: Shai I et al. N Engl J Med 2008;359:229-241
Results - BMI, Waist Circumference,
BP at 24 months
Mean BMI changes
Low-fat group: -1.0 +/- 1.4
Mediterranean: -1.5 +/- 2.2
Low-carb group: -1.5 +/- 2.1
Mean Waist Circumference Changes
Low-fat group: -2.8 +/- 4.3 cm
Mediterranean: -3.5 +/- 5.1 cm
Low-carb group: -3.8 +/- 5.2 cm
Mean BP Decrease
Low-fat group: -4.3/0.9
Mediterranean: -5.5/2.2
Low-carb group: -3.9/0.8
Results - Fasting Lipid Profile
at 24 months
HDL Cholesterol: increased in all groups with greatest
increase in low-carb group
Low-fat group: +6.4 mg/dl
Mediterranean: +6.3 mg/dl
Low-carb group: +8.4 mg/dl
Triglycerides decreased significantly in low-carb and
Mediterranean groups compared to low-fat group
Low-fat group: -2.8 mg/dl
Mediterranean: -21.8 mg/dl
Low-carb group: -23.7 mg/dl
Results – Fasting Lipid Profile
at 24 months
LDL Cholesterol: no significant changes within or
between groups
Low-fat group: -0.05 mg/dl
Mediterranean: -5.6 mg/dl
Low-carb group: -3.0 mg/dl
Ratio of Total Cholesterol HDL: decreased in all groups
with greatest improvement in low carb group
Low-fat group: -0.6 mg/dl
Mediterranean: -0.9 mg/dl
Low-carb group: -1.1 mg/dl
Change in Fasting Lipid Profile over 2 yrs
Low-fat group: -2.8 +/- 4.3 cm
Mediterranean: -3.5 +/- 5.1 cm
Low-carb group: -3.8 +/- 5.2 cm
Source: Shai I et al. N Engl J Med 2008;359:229-241
Discussion
Low-fat, Mediterranean and low-carb (Atkins) diets
produced weight loss in a group of moderately obese
participants
Women lost significantly more weight with Mediterranean
diet vs. low-fat and low-carb diets
Mediterranean and low-carb diets had beneficial metabolic
effects
Similar caloric deficit achieved in all diet groups
No decrease in need for medications among participants
Continued improvement of biomarkers over 24 months
despite achievement of maximum weight loss by 5 months
suggests that healthy diet has benefits beyond weight
reduction
JAMA Weight Loss Study
Randomized control trial conducted over 12 months
of 311 overwight/obese (BMI 27-40) non-diabetic,
premenopausal women recruited from community
Compared Atkins (very low-carb), Zone (low carb),
LEARN (low-fat, high carb), Ornish (very high carb)
Results: Women in Atkins diet group lost more weight
and had more favorable metabolic effects (higher
HDL and lower triglycerides) after 12 months than
women in Zone, LEARN, and Ornish diet groups
Key Similarities
Studies compared different, limited diets with different
compositions of fat, carbs, and protein
All diets (low-fat, low-carb, etc) helped participants
lose weight but Atkins worked best and resulted in
greatest weight loss
Atkins has favorable metabolic effects, resulting in
greatest increase in HDL and greatest decrease in
triglycerides in both studies
Key Differences
In NEJM article, women lost most weight with
Mediterranean diet while in JAMA article, women lost most
weight with Atkins diet
JAMA article did not include Mediterranean diet
NEJM article measured more sustained weight loss over
two years vs. one year in JAMA article
JAMA article statistically adjusted for weight loss
differences among groups and found a combined effect of
benefit for Atkins diet attributable to both increased
weight loss and dietary composition
Strengths
Long duration of study (follow-up over 2 years)
One-phase design – all participants started
simultaneously
Excellent adherence to diets
Consistent data, participants successfully changed
dietary intake
No missing data; used intention to treat for patients
lost to follow-up
Multiple measurements over time
Frequency matched for effect possible effect modifiers
during randomization (BMI, CAD, Type II DM, statin
use)
Limitations
Unclear primary vs. secondary outcomes
Homogenous population of Israelis
Limited number of women enrolled in study
Intense intervention: extremely difficult to replicate 24
hours counseling by nutritionist
Workplace environment; difficult to generalize results
to other outpatient settings
Difficult to control for effect of different amounts of
exercise between different groups
Bottom Line
Significant weight loss over 2-years with all three diets:
low fat, Mediterranean, and low-carb
Low-carb and Mediterranean diets could be discussed w/
patients in clinical practice as options if they are more
consistent w/ cultural or personal dietary practices
Diets should be individualized according to personal
preferences and metabolic needs
Mediterranean diet shows high efficacy in weight loss for
women and should be discussed w/ female patients
Key may be encouraging patients to decrease calorie
consumption (smaller portions, etc)
Low-carb, non-restricted calorie diet is an option for patients
who won’t follow low-fat or restricted calorie diet
Citations
Shia, I., et al. Weight Loss with a Low-Carbohydrate,
Mediterranean, or Low-Fat Diet. N Engl J Med
2008;359:220-41.
Gardner, C.D., et al. Comparison of the Atkins, Zone,
Ornish, and LEARN Diets for Change in Weight and
Related Risk Factors Among Overweight Menopausal
Women. JAMA. 2007;297:969-77.