November 12, 2009

Effect of The Atkins Diet, Long-term, on Mood

The latest study to compare low fat with low carb diets came up with the finding that long term the Atkins diet caused more mood problems than the low fat diet. This has caused a flurry of posts from LC bloggers and forum participants dissing the study. But I've had a good look at it and I think people are missing some very important things when they dismiss this study outright.

The full text is available online here:

Long-term Effects of a Very Low-Carbohydrate Diet and a Low-Fat Diet on Mood and Cognitive Function. Grant D. Brinkworth et al. Arch Intern Med 2009;169(20):1873-1880.

The most significant finding of this study, which seems to have escaped everyone who has written about it, is that it contradicted the earlier, and very heavily publicized, finding that the low carb diet caused problems with memory and thinking.

This year-long study found Working memory improved by 1 year (P < .001 for time), but speed of processing remained largely unchanged, with no effect of diet composition on either cognitive domain.

So that should put to rest any concerns you might have had about the impact of eating a very low carb diet on your ability to think clearly.

Now let's see what else the study found.

Unlike many studies of supposedly "low carb" diets, this diet was indeed a very low carb diet with a composition that matched that described in the most recent Atkins book. The nutrient breakdown was:
4% of total energy as carbohydrate, 35% as protein, and 61% as fat (20% saturated fat), with the objective to restrict carbohydrate to less than 20 g/d for the first 8 weeks and with an option to increase to less than 40 g/d for the remainder of the study.

This study also limited saturated fat which is not characteristic of classic Atkins, but the direction the Atkins brand has moved into as it has come to copy South Beach. Over the years the original "Atkins" diet was modified several times to incorporate the techniques found in other bestselling low carb diet books, so this shouldn't be a surprise.

In one major characteristic this diet differ from Atkins as described in the book. From the outset, calories were restricted to "approximately 1433 kcal/d for women and 1672 kcal/d for men."

Those who wish to ignore the findings of the study completely have pointed to this limitation as if it discredited the results. However, I do not believe this is fair. For overweight and slightly obese people, the calorie levels used here correspond very closely to what many people who have successfully lost weight on the Atkins diet report eating on the online diet support forums, for example, Low Carb Friends once the "easy pounds" lost in the first month or so are gone.

Atkins dieters almost always stall after the first 6 weeks and those who do not have a lot of weight to lose often find they do have to cut back on calories to continue on with weight loss.

So much for the "Atkins diet" used in the study. But when we look at the "low fat" diet, we see something even more interesting. The "low fat" diet is described as having "46% of total energy as carbohydrate, 24% as protein, and 30% as total fat (<8% saturated fat), with the objective to restrict saturated fat intake to less than 10 g/d for the study duration and with the inclusion of an approved food exchange (equivalent to the energy content of 20 g of carbohydrate) between weeks 9 and 52."

This is a very moderate low fat diet, very different from diets like Ornish. And more significantly, this is a diet that is actually quite low in carbohydrate. Working out the 46% ratio of carbohydrate against the 1433 calorie daily intake for women, we find that a woman on this diet would be eating only 165 grams of carbohydrate a day, NOT the 300 grams a day which is so often recommended by dietitians.

This is not an overwhelming carbohydrate intake for a person who has functioning beta cells. And everyone in this study DID have functioning beta cells, since the study excluded people with diabetes, and also eliminated people with cardiovascular disease--which would have probably got rid of participant with undiagnosed diabetes.

The study found that both groups of dieters lost the same average amount of weight at the end of the year, on average about 30 lbs. This full text of the mood study does not explain what the groups starting weights were. This data is probably available in an earlier publication about the same study that reported the physiological rather than psychological findings of this study. You can find it HERE. Unfortunately, free full text is not available for that study.

This earlier report on this same study found that the Atkins dieters ended up with a higher HDL and lower Triglycerides than the Low Fat group, but also higher LDL, though the particle size of the LDL was not investigated.

It also reports the LC group also seems to have lost more body fat: LC: –11.3 ± 1.5 kg; LF: –9.4 ± 1.2 kg; P = 0.3, though this may not be statistically significant.

The physiological study report also found that "Blood pressure, fasting glucose, insulin, insulin resistance, and C-reactive protein decreased independently of diet composition." This means that the choice of diet did NOT have any impact on these parameters though they improved similarly on both diets.

It is a very significant finding that the two diets ended up producing the same blood sugar outcomes, though this has to be viewed with the knowledge that this was a study that did not include people with diabetes.

So far I see nothing to make me question that this was a decently conducted study that has something to teach us about the impact of these diets on people who, if they had any blood sugar abnormalities had the mild ones described as "pre-diabetes"--a group that includes many people who have insulin resistance, but also who have normal beta cells and hence who will never progress to full fledged diabetes.

So what should we make of the finding that mood deteriorated more in people on the Atkins diet than on the low fat diet?

The first thing I note, before I look at mood, is that as is the case in so many diet studies, the drop out rate was high--41%, And slightly more people in the Low Carb arm dropped out than in the Low Fat arm.

This is not the first time we've seen this happen. The drop out rate of the Atkins dieters was only exceeded, very slightly, by that of the extreme Ornish dieters in the JAMA diet bakeoff published back in 2005. In that study the Atkins compliance rate was considerably lower than that of the dieters eating the conventional Weight Watchers diet, though compliance on all diets deteriorated.

This latest study did not measure compliance, but my guess is that by the end of the study, it wasn't very good because it never is on any diet and there is evidence suggesting that "carb creep" is a huge problem, long term, even for people who believe they are eating low carb diets.

But that high drop out rate, with the higher Atkins drop out rate suggests that people eating the diet were not universally thrilled and the mood indicators that the study presents seem to me to point to why.

The study used standardized questionnaires, which like all psychological tools are very fuzzy in concept, but since mood itself is a very fuzzy concept, there isn't any better way to measure it.

The study design has going for it that it used several different tools to measure mood, not just a single one. These were "the POMS,which measures 6 separate aspects of mood, including tension-anxiety, depression-dejection, anger-hostility, vigor-activity, fatigue-inertia, and confusion-bewilderment, and provides a global score of mood disturbance (total mood disturbance score [TMDS]) that is determined by subtracting the vigor-activity score from the sum of the 5 negative mood factors; (2) the Beck Depression Inventory (BDI) and (3) the Spielberger State-Trait Anxiety Inventory (SAI)."

This link will take you to the graphs describing the findings:

Questionnaire Data

What stands out is that the results of all three measurement tools come out with very similar results. At 8 weeks, all the dieters were considerably happier than when they started their diets. And at 8 weeks, the Low carb dieters were MUCH happier than the low fat dieters according to the Depression-Dejection score, probably because the early weight losses on Atkins are much more dramatic and motivating than the losses on the non-ketogenic diet.

By six months things started to change. Both groups were feeling increased Vigor and Activity at six months, probably because that was the time when they had made most of their weight loss and that weight loss made them feel much better about themselves and more prone to physical activity. But by that point, the measurement tools suggest that the mood of the Atkins dieters is beginning to deteriorate compared to that of the people on the low fat diet.

Anyone who has participated in the online Atkins support groups knows that by 6 months most Atkins dieters have either stalled out completely (usually those who are eating very high calorie intakes in the belief that only carbs matter) or they have gone from dramatic weight losses to losing very, very slowly.

You can see exactly how much real dieters lose on a very low carb diet with the analysis of dozens of real people's posted monthly low carb weight loss experiences you'll find HERE.

Slowing weight loss is characteristic of ALL long term diets no matter what the diet composition, but for the Atkins dieter who saw swift weight loss early on in the diet, the slowdown can be devastating, because most Atkins dieters assume that the extreme changes they have to make in what they eat will be rewarded with huge weight losses like those they experience in the first few weeks. When this doesn't happen, the limitations of the diet are much harder to endure.

Experience in the support group environment reinforces the finding that it is at 6 months into the very low carb diet that people run into serious problems with it. The excitement of eating steak, cheese, and avocados has worn off and unless a person is able to cook and willing to put time into hunting up recipes, the food allowed on the diet can become very boring indeed.

And there is another problem, one that is rarely discussed on the Low Carb support boards:. The early high energy level that enthuses people about the very low carb diet may start to fade out. Some people, in fact, experience thyroid slowing (so-called Euthyroid syndrome) characterized by drops in T3.

My guess is that the finding of this study is real, because it mirrors what I've seen in the support groups. After an initial burst of enthusiasm that lasts about 6 months, a large number of low carb dieters disappear. And even those who lose significant amounts of weight tend to disappear shortly after they stall out for a few months or reach goal, only to show up on the support groups a year later with tales of crashing off the diet and regaining all the lost weight.

Because I did the same thing myself, after 3 years on a very low carb diet, I have a very good appreciation of the process that leads from initial enthusiasm, to depression about the need to eat in a way that is so different from what one might want to eat, and the way that dropping energy levels can lead to giving up entirely.

Since I have diabetes and can't process carbs my choice was NOT to say to heck with it and just live with being fat. But people who don't have diabetes can and do.

So I think the mood issue is real and I think that until it is addressed, people will alway have trouble sticking with a low carb diet, long term, no matter what it's impact on their health if it is too extreme.

And it is that factor of extremeness that I think this study sheds some interesting light on.

If we ignore the issue of "fat" which is really a red herring in these two diets, what we see here is that for people WITHOUT diabetes, there may be a much better outcome in terms of mood with a diet that though it restricts carbs a lot compared to the Standard Diet, doesn't restrict them extremely.

In short, for people who do not have diabetes, a diet of 165 g a day is a huge improvement on one of 300 g a day and may be all that is needed. The women eating that 165 grams of carbs a day did lose the same amount of weight and more importantly, their blood sugar profile and blood pressure did not vary from that of the people eating at much lower levels.

For people WITH diabetes this finding is negated by the fact that few of us can tolerate 165 grams of carbohydrate without seeing very poor blood sugars, ones guaranteed to produce complications.

But what I would take from this study is that it supports the strategy I have been promoting for the past 5 years--one that suggests you cut your carbs down ONLY to the level that gives you safe blood sugars, and no lower.

What that level will be varies from person to person and can only be determined using the strategy described HERE.

More importantly, what this study suggests is that if, like me, you find you can only control your blood sugar by eating at extremely low carbohydrate intake levels--for me it was no more than 50 grams a day with no single meal being higher than 12 g--if you start feeling depressed or rundown, it's time to look into finding a medication that will let you raise your carb intake a bit, but still keep hitting your blood sugar targets without making yourself miserable.

For me, the difference between eating at 100 g a day and 50 g a day is that I can do it, year in and year out, happily. (I'm in year 7 of maintaining a 17% loss of body weight.) But I can only eat 100 g a day using fast acting insulin at some meals.

Not EVERYONE gets depressed or exhausted on a long term low carb diet. The people who stay on these diets for years at a time and write the enthusiastic LC blogs are those who feel better on the diet, not worse. There are quite a few people that match that description, especially among those whose blood sugar is hard to control.

But if you run into problems as your low carb diet hits 8 months, or a year, or two, don't fear turning to medication for help. Insulin, metformin, and for some people Byetta, can make a big difference in how easy it is to control your blood sugar and your weight.

Our goal, after all is health but there is little point in purchasing "health" at the cost of your happiness. If after 6 months of eating a very low carb diet your energy level is low and your mood deteriorating, it's time to start tweaking. Talk to your doctor about adding the safe drugs to your regimen.

Too many people treat using medication as if this were a personal failure. It isn't. The safe medications are is just another set of tools to allow you to live a healthy and happy life.



dampmop said...

Thanks again for a nice, clear explanation of these studies. I've been searching online for a connection between the low carb diet and low thyroid (whether the diet reveals thyroid problems or causes them). What is your take on this? Also, do you see this as more of a problem for women than men?

Jenny said...

As explained in Lyle Macdonald's book, The Ketogenic Diet, the body responds to a ketogenic state similarly to the way it responds to starvation since the body runs on ketones during starvation. So over time it may downregulate the thyroid when a person is in ketosis as a self-preservation measure that makes sense in a true famine situation (which is a situation most humans encountered during their lives through most of the millions of years of hominid life.)

The Eades discussed this years ago and admitted to having to supplement T3 in some patients, but then they later said they hadn't. I found the web page on Wayback Machine where they had. So yes, it is an issue.

I don't think it is gender-specific, though it sometimes seems like more women have borderline thyroids.

Bernstein also observed a lot of Thyroid problems in his patients but attributes it to autoimmunity, since many are Type 1s.

I'm not a Type 1 and I have zero autoimmune markers but I have experienced this and I don't think it is the same as regular thyroid disease.

It resolves if you up your carbs a bit and get out of ketosis.

Stargazey said...

Just to provide another bit of data:

When I started Atkins, I had been taking 125 mg of Synthroid (T4) for about 20 years. So presumably my thyroid gland was making no thyroid hormone at all when I began to do low-carb. Any thyroid hormone I had, came from the drug. About a year ago, I began to experience symptoms of hypothyroid, even though my dose of T4 had not changed. Labwork revealed that I was not changing very much of the T4 to T3 (which is the active form of the hormone). Now I have to take 100 mg of T4 plus 25 mg of T3 to be symptom-free. So the total exogenous thyroid hormone dose remained the same, but the form of it had to be changed.

Anyway, it may be that low-carb changes the body's ability to convert T4 to T3. Or it may simply be an issue of getting older. But for the people who study this issue, I thought I'd throw that out there.

Unknown said...

Way back in ancient times before agriculture, we would have eaten carbs when they were available. That might be the way our thyroids are meant to work, eating carbs from time to time but not large amounts of sugar and grain all the time.

Since the people in the study didn't have diabetes, if they had followed the carb ladder in the Atkins book, instead, might they not have ended up as high as 165g/day or more in many cases?

I think I lost most of the weight I lost after dx at about that carb level, it was just because the carbs had been driving my appetite. I was still on one of the ß-cell stimulators when I ate like that, though. I reduced both carbs and medication gradually. I don't count anything now, I just avoid starch and sugar most of the time and have little splurges every so often, counting gets old.

ItsTheWooo said...

I'm only remotely happy and sane if I am eating <60 carbs per day, preferably around 40-60. I suspect the reason so many people on the atkins diet were feeling bad relative to the more balanced diet is probably because they were eating too few carbs. If I eat less than 40 for any length of time I start feeling like crap. But if I eat more than 60 I feel like crap in a different way. 40-60 is the best place to be because it is ketogenic enough to keep me sane but high enough to prevent my body from really going "omg starvation". Or so I think.

It is so paradoxical to me that people get "happy" eating carbohydrate. I have never experienced this, and never will probably. After eating carbohydrate I just start to develop one of the following states;
1) atypical depression (if more lethargic)
2) irritation and anger and rage (if more energetic)

The only time I feel good, happy, motivated, clear, balanced, is when eating lower calories, very high fat, and low carb. I feel best right before a meal (hypoglycemia excluded, which sometimes happens when I delay food too much). I feel best before my first meal. I feel best when I haven't eaten much all day. I feel best if I don't eat much all day and it is late at night.

If I take things too far and start starving myself then I start feeling like crap and getting hypoglycemia a lot. But presuming I am eating adequate amounts of food on average, I feel BEST when avoiding food. And carbs. My mood is much much better.

I wonder what it is like to be one of those people who feels happy after carbs. Never experienced it ever.

Matt Stone said...

A very low carb diet is great for mood for the first couple of months. I call this the low-carb honeymoon period.

The problem is that most people who gravitate towards low-carb have typical neurotransmitter addiction profiles, and very low-carb diets, in addition to lowering thyroid hormones (which is one primary reason why people become progressively LESS carbohydrate tolerant, i.e. weaker and more unhealthy on a low carb diet) develop serotonin deficiencies.

Kathleen DesMaisons is much more on top of this than the low-carb community, as am I.

I appreciate your ability to pay attention to the study though Jenny, and not just simply write it off.

If I were on a low-carb diet and defending it I would have been intensely angered by the study, yelled at it, and thrown something. My mood eroded steadily on such a diet, and came to a head after 1 month of zero carb, where I had legitimately become unstable.

Red Sphynx said...


Thanks so much for this analysis. You have a real gift for reading a set of papers, drawing some non-obvious conclusions, and summarizing it all in plain language. (Yes, non-obvious. Even if I'd read the paper, I wouldn't have made the connection to the memory and thinking work.)

I've been following your postings for, what..10+ years now (first on a.s.d. and now your blog.) You've made an enormous difference in my ability to manage my Type II diabetes.

Your blog and websites are true labors of love. Thanks for all the time you put in.

Still in the 5% club after all these years.

Stargazey said...

I may be a minority of one, but the lower my carbs are (down to somewhere under ten a day), the better I feel. And I've been doing this for over six years.

If I eat too many carbs, I get depressed and I start thinking about food all the time.

Yes, the food is boring. Yes, I've had to start taking T3. Yes, I take 5-HTP if I start feeling the need for comfort food. But I am so much happier and more productive when I am doing low-carb that I never want to go back to higher-carb eating ever again. :-)

Jenny said...

Red Sphynx,

Thanks for the very kind words. I really miss a.s.d. Too bad that the spammers and maniacs overwhelmed it.

I learned so much participating there, plus I miss interacting with the regulars. said...

For some reason, in recent years, I have tended to not have mood swings, or only very insignificant changes in mood.

Many years ago, I did definitely have daily mood swings, almost as if I had a daily on-off mood switch. I would wake up each morning, and I would know immediately whether I would be in a cranky mood all day, or whether nothing could ruffle me. I have no idea what was going on, but it stopped, and I am now almost totally moodless.

So, I don't seem to experience any diet-related moods.

And, no, I don't take any medications of any type. None whatsoever.

Serenity is good.

LHL said...

Thank you Jenny for such a nice clear precis of these papers.
As a low carbing T2 diabetic of just over 1 year, I try to manage on around 3x20g carb meals a day. I can find all the low fat diet products in our UK supermarkets but nothing for low carb diets at all. This is what depresses me most - although I can and do cook meals daily it would sometimes be nice to have something quick from off the supermarket shelf. Oh, and I am also a vegetarian to complicate matters.

Jenny said...


Your market is full of LC products: meat, cheese, nuts and veggies.

The products marketed in the US as "Low Carb" are almost without exception dishonestly labeled. They are full of sugar alcohols which metabolize to glucose or resistant starches which are digested by gut flora. They also contain many chemicals. The "protein" in them is often poor quality collagen.

You are much more likely to succeed on a diet eliminating all "low carb diet" foods.

LHL said...

Point taken, mostly what I buy & cook is as you described, except for the meat. I also grow some of my own vegetables.

However, I also cater for 2 elderly pensioners in my household and would like to have time off from cooking just occasionally by popping something into the microwave or oven that I did not spend hours preparing.