Tag Archives: diet comparison

Low-Fat and Low-Carb Diets End Battle in Tie After Two Years, But…

Dieters on low-fat and low-carb diets both lost the same amount of weight after two years, according to a just-published article in Annals of Internal Medicine.  Both groups received intensive behavioral treatment, which may be the key to success for many.  Low-carb eating was clearly superior in terms of increased HDL cholesterol, which may help prevent heart disease and stroke.

The study was funded by the National Institutes of Health and was carried out in Denver, St. Louis, and Philadelphia.

How Was It Done?

Healthy adults aged 18-65 were randomly assigned to either a low-fat or low-carbohydrate diet.  Average age was 45.  Average body mass index was 36 (over 25 is overweight; over 30 is obese).  Of the 307 participants, two thirds were women.  People over 136 kg (299 lb) were excluded from the study—I guess because weight-loss through dieting is rarely successful at higher weights.  Diabetics were excluded. 

The low-carb diet:  Essentially the Atkins diet with a prolonged induction phase (12 weeks instead of two).  Started with maximum of 20 g carbs daily, as low-carb vegetables.  Increase carbs by 5 g per week thereafter as long as weight loss progressed as planned.  Fat and protein consumption were unlimited.  The primary behavioral goal was to limit carb consumption.

The low-fat diet:  Calories were limited to 1200-1500 /day (women) or 1500-1800 (men).  [Those levels in general are too low, in my opinion.]  Diet was to consist of about 55% of calories from carbs, 30% from fat, 15% from protein.  The primary behavioral goal was to limit overall energy (calorie) intake. 

Both groups received frequent, intensive in-person group therapy—lead by dietitians and psychologists—periodically over two years, covering such topics as self-monitoring, weight-loss tips, management of weight regain and noncompliance with assigned diet.  Regular walking was recommended.

Body composition was measured periodically with dual X-ray absorptiometry.

What Did They Find?

Both groups lost about 11% of initial body weight, but tended to regain so that after two years, both groups average losses were only 7% of initial weight.  Weight loss looked a little better at three months in the low-carb group, but it wasn’t statistically significant. 

The groups had no differences in bone density or body composition.

No serious cardiovascular illnesses were reported by participants.  During the first six months, the low-carb group reported more bad breath, hair loss, dry mouth, and constipation.  After six months, constipation in the low-carb group was the only symptom difference between the groups.

During the first six months, the low-fat group had greater decreases in LDL cholesterol (with potentially less risk of heart disease), but the difference did not persist for one or two years.

Increases in HDL cholesterol (potentially heart-healthy) persisted throughout the study for the low-carb group.  The increase was 20% above baseline.

About a third of participants in both groups dropped out of the study before the two years were up.  [Not unusual.]

My Comments

Contrary to several previous studies that suggested low-carb diets are more successful than low-fat, the study at hand indicates they are equivalent as long as dieters get intensive long-term group behavioral intervention. 

Low-carb critics warn that the diet will cause osteoporosis, a dangerous thinning of the bones that predisposes to fractures.  This study disproves that.

Contrary to widespread criticism that low-carb eating—with lots of fat and cholestrol— is bad for your heart, this study notes a sustained elevation in HDL cholesterol (“good cholesterol”) on the low-carb diet over two years.  This also suggests the low-carbers  followed the diet fairly well.  The investigators also note that low-carb eating tends to produce light, fluffy LDL cholesterol, which is felt to be less injurious to arteries compared to small, dense LDL cholesterol.

A major strength of the study is that it lasted two years, which is rare for weight-loss diet research.

A major weakness is that the investigators apparently didn’t do anything to document the participants’ degree of compliance with the assigned diet.  It’s well known that many people in this setting can follow a diet pretty well for two to four months.  After that, adherence typically drops off as people go back to their old habits.  The group therapy sessions probably improved compliance, but we don’t know since it wasn’t documented. 

How often do we hear “Diets don’t work.”  Well, that’s just wrong.

Overall, it’s an impressive study, and done well. 

Individuals wishing to lose weight on their own can’t replicate these study conditions because of the in-person behavioral intervention component.  There are lots of self-help calorie-restricted balanced diets (e.g., Sonoma Diet, The Zone,  Advanced Mediterranean Diet) and low-carb diets (e.g., Atkins Diet, Banting’s Letter on Corpulence, Low-Carb Mediterranean or Ketogenic Mediterranean Diets).  On-line support groups—e.g. Low Carb Friends and SparkPeople and 3 Fat Chicks on a Diet—could supply some necessary behavioral intervention strategies and support.  

Choosing a weight-loss program is not as easy as many think.  [Well, I’ll admit that choosing the wrong one is easy.]  I review the pertinent issues in my “Prepare for Weight Loss” page.

Steve Parker, M.D.

Reference: Foster, Gary, et al.  Weight and metabolic outcomes after 2 years on a low-carbohydrate versus low-fat diet: a randomized trial. Annals of Internal Medicine, 153 (2010): 147-157   PMID: 20679559


Filed under Carbohydrate, Fat in Diet, ketogenic diet, Overweight and Obesity, Weight Loss

Comparison of Mediterranean, Low-Carb, and Low-Fat Weight-Loss Diets

The July 17, 2008, issue of the New England Journal of Medicine has a well-done study comparing the Mediterranean, low-carb, and low-fat weight-loss diets in an Israeli population over the course of two years.  The researchers conclude that “Mediterranean and low-carbohydrate diets may be effective alternatives to low-fat diets.  The more favorable effects on lipids (with the low-carbohydrate diet) and on glycemic control (with the Mediterranean diet) suggest that personal preferences and metabolic considerations might inform individualized tailoring of dietary interventions.”

How was the study set up?

Moderately obese participants (322) were randomly assigned to one of the three diets: 1) low-fat, calorie-restricted, 2) Mediterranean, calorie-restricted, or 3) low-carbohydrate, non-restricted.  Calories in the low-fat and Mediterranean diets were “restricted” to 1800 per day for the men, 1500 for the women.  Average age of participants was 52, and average body mass index was 31.  [A 5-foot, 10-inch man weighing 216 pounds (98.2 kg) has a BMI of 31.]  Nearly all participants – 277 or 86% of the total – were men.  So there were only 45 women.  Forty-six participants had type 2 diabetes.

The low-fat diet was based on the American Heart Association guidelines of 2000: 30% of calories from fat [this isn’t very low], 10% of calories from saturated fat, cholesterol limited to 300 mg/day.  [The AHA revised their guidelines in 2006.]  Low-fat dieters ”were counseled to consume low-fat grains, vegetables, fruits, and legumes and to limit their consumption of additional fats, sweets, and high-fat snacks.”

The Mediterranean diet was based on the recommendations of Walter Willett and P.J. Skerrett as in their book, Eat, Drink, and be Healthy: The Harvard Medical School Guide to Health Eating.  Mediterranean dieters ate 2 fish meals per week, a handful of nuts daily, 30-45 grams of extra virgin olive oil per day, etc.  [One tablespoon of olive oil is 14 grams.]  The AHA states that “this diet reflects the current recommendations from the American Heart Association.”  There were no specific recommendations regarding alcohol in any of the diets.

The low-carb diet was based on  Atkins’ New Diet Revolution of 2002.  The goal was to provide 20 grams of carbohydrate per day for the 2-month induction phase, with a gradual increase to a maximum of 120 grams daily to maintain weight loss.  Total calories, protein, and fat were not limited.  “Participants were counseled to choose vegetarian sources of fat and protein….”

Whole grains were recommended for the low-fat and Mediterranean cohorts.

All participants worked at the same nuclear research facility in Dimona, Israel.  They were given careful instructions, initially and periodically, regarding the diet to which they were assigned.  Lunch is the main meal of the day in Israel, and they all ate lunch at the facility’s self-service cafeteria, which prompted them to choose the proper food items.  I assume they were told to maintain the diet when off-duty.  Adherence to the diets was assessed by a food-frequency questionnaire.


  • After 24 months, how many participants were still involved?  90% in the low-fat group, 85% in the Mediterranean, 78% in the low-carb.
  • There was little change in the usage of medications, and no significant differences among the groups.
  • Daily energy intake (calories or kcal) decreased from baseline levels significantly – about 450 calories – in all groups at 6, 12, and 24 months compared with baseline, with no significant differences among the groups in the amount of decrease.
  • All groups started with 51% of energy intake (calories) from carbohydrate.
  • At 24 months, the low-carb dieters were getting 40% of their daily calories as carbohydrates.  The other two groups were eating 50% of energy intake from carbs. [This still seems like a lot of carbs on the Atkins diet.  A gram of carbs has 4 calories.  The stated carbohydrate goal was a maximum of 120 grams of carbs daily, on a diet of 1800 calories.  So 120 grams of carbohydrate should be 27% of total daily calories.  At no point did the low-carb group reduce their average percentage of calories from carbohydrates under 40%.  OK, maybe be in the first two weeks but those data are not reported.  On an 1800 calorie diet, 40% of calories from carbs would be 180 grams.]
  • At 24 months, the low-carb dieters were getting 39% of their daily calories as fat.  The other two groups were in the 30-33% range.
  • Baseline fat intake for all groups was 31-32% of total calories, with saturated fat being 10% of the fat calories.
  • The low-fat cohort dropped their fat calories from 31 to 30% of total calories, which is essentially no change from baseline percentage.
  • At 24 months, the low-carbers were getting 22% of their daily calories from protein.  The other groups were at 19%.  [The low-carb Atkins diet is often criticized as having too much protein.]
  • Only the low-carb group made major changes in macronutrient composition of their diet.  Macronutrients are protein, fat, and carbohydrates.  This Atkins group increased saturated fat from 10 to 12% of total calories, reduced carbs from 51 to 40% of calories, increased protein from 19 to 22% of calories, and increased total fat from 32% to 39% or total calories.
  • All cohorts lost weight, but losses were greater in the low-carb and Mediterranean groups.  For the 272 participants who completed the full 24 months of intervention, the losses averaged 3.3 kg (7.3 lb) for the low-fat group, 4.6 kg (10.1 lb) for the Mediterraneans, and 5.5 kg (12.1 lb) for the low-carb group.
  • Among the 45 women, the low-fat group lost only 0.1 kg (0.22 lb), the Mediterraneans lost 6.2 kg (13.6 lb), and the low-carbers lost 2.4 kg (5.3 lb).  There were only 15, 20, and 10 women in these groups, respectively.
  • All groups had significant blood pressure reductions: about 4 mmHg systolic and 1 mmHg diastolic.
  • HDL cholesterol (the “good cholesterol”) increased in all groups, 8.4 mg/dl in the low-carb group, about 6.3 in the others.
  • LDL cholesterol (the “bad cholesterol”) fell 5.6 mg/dl in the Mediterraneans, 3.0 mg/dl in the low-carbers, and none in the low-fat group.  But these were not statistically significant differences between the groups.
  • The ratio of total to HDL cholesterol decreased for all groups, but the relative 20% decrease in the low-carb group was statistically significant compared to the 12% relative decrease in the low-fat group.  The ratio fell 16% in the Mediterranean group.  [The total/HDL ratio is thought to reflect risk of developing atherosclerotic complications.  You want it under 5 to 1, and 3.5 to 1 may be ideal.]
  • The level of high-sensitivity C-reactive protein decreased significantly only in the Mediterranean and low-carb cohorts.  [C-reactive protein is felt to be a marker of the systemic inflammation that has a role in atherosclerosis or hardening of the arteries.]
  • Thirty-six of the diabetics had adequate lab studies for analysis – about 12 in each diet group.  Only those in the Mediterranean group had a significant decrease in fasting glucose – 33 mg/dl.  The low-fat group had an increase.  Glycated hemoglobin decreased in all three groups although to a significant degree (0.9%) only in the low-carb group.  [High glycated hemoglobin levels reflect poor control of blood sugar levels in diabetics.]
  • Insulin levels decreased significantly in all three groups, diabetic or not.  [Abnormally high insulin levels are felt to have adverse health effects.]

Limitations of the study

  • Relatively few women, making it difficult to reliably generalize results to women.
  • Relatively few people with diabetes, making it difficult to reliably generalize results to people with diabetes.
  • Israeli gene pool?  Results not applicable to others?
  • No change in physical activity recommended to participants.  Increased exercise should enhance weight loss.

Take-Home Points

  • Caloric restriction leads to weight loss.
  • Mild degrees of weight loss reduce blood pressure.
  • In this study, the low-carb/Atkins and Mediterranean diets were more effective than the “low-fat” diet.
  • Atkins dieters can lose weight without counting calories, by limiting carbohydrate intake.
  • You gotta wonder if the low-carb group would have been even more successful if they had actually limited carbs to 120 grams daily.
  • It’s possible a lower-fat diet may have been more efficacious than the one utilized here.
  • This study did not enroll enough women to prove that a calorie-restricted Mediterranean diet is superior to low-fat and Atkins diets.  The greater weight loss – 13.6 pounds for Mediterranean versus 5.3 with Atkins – is suggestive and requires further study.
  • The average amounts of weight loss are not much when you think about the effort expended over 24 months of intervention.
  • These dieters reportedly reduced their daily caloric intake from baseline levels by about 450 calories, over the course of two years.  Yet they lost relatively little weight.  The numbers do not jive.  Most likely there is a problem with the methodology.  I doubt the average daily calorie deficit was as high as 450.
  • The Mediterranean diet seems to have been better for the people with diabetes.  Confirmatory studies are imperative.  Insulin resistance is an important factor in type 2 diabetes.  Monounsaturated fats, which are prominent in olive oil and the Mediterranean diet, are linked to improvement in insulin resistance in other studies.
  • For people who need to lose excess fat yet refuse to consciously restrict overall caloric  intake, the low-carb Atkins diet is a reasonable option.
  • The traditional Mediterranean diet has demonstrable long-term health benefits: longer lifespan, less cancer (colon, prostate, breast, uterus), reduction of cardiovascular disease, less dementia, and prevention of type 2 diabetes.  The Atkins diet cannot make those claims in 2008.

Steve Parker, M.D.

Reference:  Shai, Iris, et al.  Weight Loss with a Low-Carbohydrate, Mediterranean, or Low-Fat Diet.  New England Journal of Medicine, 359 (2008): 229-241.

Additional information and critical analysis for health nuts like me:

Dr. Dean Ornish’s analysis in Newsweek online   Dr. Ornish is a leading low-fat diet advocate.

American Heart Association comments on the study in a July 19, 2008, news release

My additional comments:

The Mediterranean diet used in this study is based on Walter Willett’s 2001 book, Eat, Drink, and Be Healthy: The Harvard Medical School Guide to Healthy Eating.  From the author:

I wrote this book to show you where the USDA Pyramid is wrong and why it is wrong.  I wanted to offer a new healthy eating guide based of the best scientific evidence, a guide that fixed the fundamental flaws of the USDA Pyramid and helps you make better choices about what you eat.  I also wanted to give you the latest information on new discoveries that shuould have profound effects on our eating patterns. 

Dr. Willett made a conscious decision not to call his new eating plan a Mediterranean diet.  Elsewhere in the book he notes that the Mediterranean diet pyramid promoted by Oldways Preservation and Exchange Trust is a good, evidence-based guide for healthy eating.  The entire book promotes Harvard’s Healthy Eating Pyramid, not the Mediterranean diet per se.

Harvard’s Healthy Eating Pyramid:

Harvard's Healthy Eating Pyramid

So were the Mediterranean dieters in the study at hand even following the Mediterranean diet?  The most glaring difference is Harvard’s lack of emphasis on olive oil.  Of lesser note is Harvard’s recommendation to eat white rice, white bread, potatoes, and refined-flour pasta only sparingly.  However, the researchers for this study directed Mediterranean diet participants to ingest 30-45 grams of olive oil per day.  After comparing the Harvard pyramid with the Oldways Mediterranean pyramid and other Mediterranean diet descriptions, it is fair to say the dieters here were indeed instructed on a Mediterranean diet.  In fact, the Mediterranean diet in this study is quite similar to the Advanced Mediterranean Diet.

Traditional healthy Mediterranean diet pyramid of Oldways Preservation and Exchange Trust:

Traditional healthy Mediterranean diet pyramid of Oldways Preservation and Exchange Trust


1 Comment

Filed under Mediterranean Diet, Weight Loss