Study finds low-carb diet outperforms low-fat diet in terms of weight and blood fat levels

It’s perhaps one of the most contentious issues in nutrition: should individuals who want to control their weight eat a diet low in fat, or low in carbohydrate? The ‘conventional’ view is that fat is ‘calorific’ and fattening, so low-fat is the way to go. However, some (e.g. Gary Taubes and the late Robert Atkins) have argued that it’s carbohydrate that, essentially, makes us fat, so we should be reigning in our consumption of such foods. One way to settle this argument is to study the effects of these diets on weight loss, and this week saw the publication of a study in which a low-fat was pitted against a low-carb diet over a period of two years. This study also assessed the effects of a so-called ‘Mediterranean diet’ too.

Participants eating the low-fat and Mediterranean diets were asked to restrict calories (1500 and 1800 calories per day for women and men respectively). Individuals on the low-carb diet could eat as much as they liked. For full details of the 3 diets in this trial, see the free full text article.

322 individuals were enrolled in the study, of whom 272 completed it. Of those who completed the study, the average weight losses were:

Low-fat diet group ” 3.3 kg
Mediterranean diet group ” 4.6 kg
Low-carb diet group ” 5.5 kg

The participants of the study also had certain blood parameters checked, including blood fat levels. In medicine, the ratio of total cholesterol (generally taken to be a bad thing) to so-called HDL (generally taken to be a good thing) cholesterol is believed to be a marker for the risk of cardiovascular disease: the higher the ratio, the greater the risk. Individuals on the low-carb diet saw a reduction in this ratio which exceeded that achieved on the low fat diet. Compared to the low-fat group, the low-carb group also saw a statistically significant drop in their levels of unhealthy blood fats known as ‘triglycerides’.

In short, compared to the low-fat group, the low-carb group lost more weight and saw improved changes in their blood fat levels.

At the start of the study, 36 participants had been diagnosed with diabetes. These (as well as other) participants had their fasting blood sugar levels checked as part of the study. Compared to the diabetics eating a low-fat diet, those eating a Mediterranean diet experienced a statistically significant reduction in their blood glucose levels, but the low-carb group did not. I find this result somewhat surprising, seeing as there is quite a lot of evidence now which shows carbohydrate restriction can be effective for helping to control blood sugar levels in diabetes.

However, it is perhaps worth pointing out that at the start of the study, more than three-quarters of the diabetics in the low-carb group were on medication for their diabetes, compared to 50 and 47 per cent of diabetics in the low-fat and Mediterranean diet groups respectively. In other words, it is possible that, overall, the diabetics in the low-carb group had more advanced disease which may not have been so amenable to a nutritional approach.

What this study does do, I think, is add to the body of evidence which suggests that lower carb/carb-controlled eating has distinct merit, and generally has the capacity to out-perform low-fat diets in the weight loss stakes (and without any conscious restriction of food intake, either). The results of this study suggest that such a diet may be superior in terms of cardiovascular disease risk too.


Shai I, et al. Weight Loss with a Low-Carbohydrate, Mediterranean, or Low-Fat Diet. NEJM 2008;359:229-241

10 Responses to Study finds low-carb diet outperforms low-fat diet in terms of weight and blood fat levels

  1. Tiggy 21 July 2008 at 2:59 am #

    If the diabetics in the low carb group were on medication, wouldn’t that increase their weight loss?

  2. SkepTicTacToe 21 July 2008 at 11:10 am #

    One point to note is that the low carb diet as stipulated has a poor nutrtional profile compared to the standard low carb model (by which I mean ‘Paleo’). This is evidenced by the line that “the participants were counseled to choose vegetarian sources of fat and protein”.

    Why on earth would you encourage a vegetarian diet – paticularly a low carb diet? I would wager that a Paleo low-carb diet involving free range meat and dairy products would wipe the floor with the other diets when it comes to weight loss.

  3. sokpuppet 21 July 2008 at 5:54 pm #

    I can’t see how many calories the low carb diet group consumed. I guess in line with other low carb diets it’ll be low due to the high amount of satiating proteins and fats?

  4. rob clark 21 July 2008 at 8:07 pm #

    There’s no such thing as ‘more advanced diabetes’, you either have it or you don’t.

    That said, this aspect of the study clearly wasn’t a level playing field as like was not being compared to like. I don’t think the results do anything to disprove your ideas ” of course carbohydrate restriction is an effective tool in controlling blood sugar levels.

  5. Dr John Briffa 21 July 2008 at 8:41 pm #


    “There’s no such thing as ‘more advanced diabetes’, you either have it or you don’t.”

    Not sure I agree with this. I would say that a type 2 diabetic with a HbA1c of 10 per cent has more advanced disease than one with a HbA1c of, say, 6 per cent (for instance).

  6. cynic 22 July 2008 at 1:24 pm #

    An interesting note Dr B
    Maybe it’s worth pointing out that the 300odd participants in the study had an average age of 52, average BMI of 31 and 86% of them were male. Now (like you) I don’t think BMI is a perfect measure at all, but such a high number is telling you something useful:

    Consider if there was an average height of 1.78m, then their average weight works out at about 98kg
    Now a bmi of 25 (which is the high-end of the “healthy range”) would suggest an average weight of about 79kg.
    Ergo the study participants were on average about 19kg or about 25% overweight – Wow
    I hope I have done my sums correctly but two thoughts occur:

    1) Low fat & med group had a restricted their daily calorie intake to 1500 for women & 1800 for men and yet they only managed to lose an average 3.3kg and 4.6kg over two years.
    Wow!. I am not being mean here, it just makes me concerned about how well these diets were stuck to (which of course is always a risk, one the authors acknowledge and probably common across all THREE groups)

    2) These results are obtained on a group of obese middle-aged men, so how well can they be extrapolated to other groupings

  7. rob clark 22 July 2008 at 3:51 pm #

    I don’t think so, John. He (or she) has less well controlled diabetes ” which is where your dietary ideas for better control come in, of course.

    Personally, I don’t think the term ‘advanced’ is quite right, because the A1c can be lowered by judicious changes to diet and lifestyle, as you yourself advocate. Does this mean that person’s diabetes is now ‘less advanced’? I would suggest that it isn’t the diabetes that’s changed so much as their control of it.

    Ultimately, though, perhaps we’re just arguing over minor semantics ” hardly the most important aspect… 🙂

  8. Paul Anderson 25 July 2008 at 9:38 pm #

    Despite a huge amount of research in to type 2 diabetes, it seems to me that the condition is not very well understood.

    It appears that the transition from normal blood sugar levels to type 2 diabetes takes place over a considerable period of time, and is preceeded by a long period of time where insulin levels are elevated, and during which blood glucose levles often rise beyond normal, and safe levels. Both the elevated insulin levels and the abnormal sugar levels can cause long term complications which may be evident long before an official diagnosis is confirmed.

    In my opinion, there is no doubt that for many, a low carb diet can bring about normal or near normal BG levels. Whereas failure to control BG levels will result in further deterioration in the underlying condition and, almost certainly, to a need to use insulin.

    Research has shown that type 2 diabetics, and indeed many type 1’s (surprisingly) still produce new beta cells. High glucose levels increase the rate of beta cell loss through a process termed glucotoxicity. I believe it is possible to stop the progressive loss of beta cells through the maintenance of tight BG control. It may even be possible for the underlying condition to improve, as beta cell loss may be halted.

    I am not suggesting that its possible to completely reverse the condition to such an extent that an individual can return to a high carb diet. At worst good control will slow the progression of the underlying condition.

    The cut off point for the diagosis of diabetes is somewhat arbitary – a fasting level above 7 (126) in plasma blood. To suggest someone who has a fasting blood sugar level of 125 is not diabetic and someone who has a fasting level of 127 is may be technically correct but it has little relevance to the health of the individual concerned.

    Many readers of this blog will confrim that it is entirely feasible to maintain BG levels below 7 (126) at all times on a low carb diet, without medication, by following a low carb diet. I know of prople who have achieved this over an extended perio, of over 10 years, and who achieve hba1c’s below 5 ona consistent basis. It would not be possible to achieve this following the conventional advice. If anyone can point to a patient who has achieved this, without mediation, by following the standard low fat high carb dietary adivce I would be extremely surprised.

    My personal experience has been that the the underlying condition can be improved by following a low carb diet – perhaps helped by weight loss, and following an exercise regime. There have been many cases documented where patients have been able to discontinue medication, cease taking large doese of insulin and acheived non diabetic numbers by switching to a very low carb diet.

    I think diabetes advances due to beta cell loss if chronically high BG levels persist. This can be measured by a deterioration in the hba1c and a need for greater medication. The attainment of tight control can slow, possibly halt and in some cases reverse this deterioration.

    What is really needed is morfe research in to the long term effect of follwoinga low carb diet: in terms of general health and in terms of the progression or otherwise of the udnerlying condition.

    Its quite clear that patients following the standard dietary advice will see a progrssive deterioration in the underlying condition. Following a low carb approach offers hope, and for many a much better quality of life.

    Paul Anderson.

  9. Anna 1 August 2008 at 2:08 pm #

    My experience with BG control echos what Paul Anderson has said.

  10. Tony 6 September 2009 at 10:48 pm #

    Hmmmm, I was diagosed type 2 a year ago, & now a year later I seem to be not diabetic (in Turkey) technically (but not statistically) in the UK still diabetic.

    After being diagnosed I decided to research the subject, & found loads of bullsh*t & contrary information. Even the advice from my UK Doctor (or his ‘specialist) diabetic nurse would never help me. My Doctors first reaction was to offer Metformin for the diabetes & statins ‘just in case’ both were rejected.

    I bought a blood sugar checker & a blood pressure tester & kept a daily blood sugar, weright ,loss & blood pressure chart for a year.

    My research convinced me that this was a numbers game & finding Barry Groves website I discovered a pathway that enabled me to research the subject of diet thouroughly.

    Low carb, high fat & protein with a modicum of moderate excercise has basically cured my diabetes (which was really carbohydrate overload)

    My body weight is now withing the BMI (just)
    My fasting blood is 4.7-5.3
    My blood sugars never exceed 6 even after/before meals
    I have no eye/foot or other diabetic linked symptoms.
    my blood pressure is typically 126/63 @55bpm
    I have total control over my diabetes (if thats what I have!)

    I eat like horse, I drink alcahol moderately (white wine & fizzy water) never feel hungry & generally have more energy than before.

    Like Banting & others, if it works for me, it will work for everyone else & save our health service a fortune & let us live our natural span!


Leave a Reply