We are going to rely more on this most excellent review paper by Dr. David Benton of Swansea, UK, Carbohydrate ingestion, blood glucose, and mood to continue our exploration of the question of how ingestion of carbohydrates affects mental state.
First off, some of the short-term studies. We all can guess the results, but a dozen or so studies have shown comparisons between a meal of say, turkey breast versus starch-based meals, or placebo versus a sucrose or glucose containing drink. Not surprisingly, mood/energy measurements at 14-30 minutes after the sugar drinks shows an increase in energy levels, followed by a slump at two hours. In the protein meals versus starchy meals, subjects report a decrease in subjective sense of energy two hours after eating the starchy meal compared to protein. All of these studies were done while subjects were resting quietly.
In studies done of subjects asked to do demanding cognitive tasks (presumably slurping up all available glucose as the noggin is burning nitro), falling levels of reported energy directly correlated with falling blood sugar levels. Subjects with low blood glucose at 30 minutes and 2 hours doing demanding cognitive tests had higher tension ratings than those with high blood glucose. In diabetics, blood glucose levels (measured continuously) have been associated with immediate mood changes, with low corresponding to a more negative mood state than high levels. (There are some interesting perturbations of this phenomena over measurements of days and weeks, however, which I will address in detail in the next post).
Now on to that theory I've actually already discussed at some length in previous posts - via insulin and albumin effects, ingestion of carbohydrate leads to the preferential shuttling of the rare amino acid, tryptophan, into the brain. Since tryptophan is the precursor to serotonin, eating carbohydrates will presumably make one more serene, sleepier, and more sated than someone whose brain is starving for serotonin (low cerebrospinal fluid (CSF) serotonin levels are associated with violence, insomnia, hunger, and suicide, partly because serotonin and its metabolite melatonin play a major role in the mood, appetite, and sleep centers).
Richard and Judith Wurtman of MIT are the ones who originally proposed the theory, and they suggested that humans (particularly premenstrual women and those afflicted with seasonal affective disorder) crave carbs for the serotonin-boosting pharmacologic effect (1)(3). I've seen this theory in textbooks and in research papers many, many times, and I rather took it as fact. Judith Wurtman is a big carb proponent and is likely to give rather strongly-worded quotes to the media damning low-carb diets for causing depression. In the comments, Zooko pointed me to this, er, interesting article by Dr. J. Wurtman for the Huffington Post. Also to her twitter feed, which advises the dubious practice of snacking on pure carbs in between meals for a "serotonin surge" - with specific recommendations for pretzels, rice cakes, and marshmallows. Long term I would say that anyone wanting to lose fat who also has mental health problems should really pretty much avoid adding any processed, micronutrient-poor calories to the mix. What is interesting about the good Dr. Benton's review paper is that he leaves the Wurtmans' tenuous theory crumbled into dust when we look at the context of carbohydrate ingestion in the real world.
It is absolutely true that a carbohydrate only meal after a fast will increase the entry of tryptophan into the brain in rodents, humans, and other primates. The problem is, the moment you eat a bit of food protein, this effect no longer takes place. Dr. Garner was able to increase serotonin in mice brains via a mixed diet, but he fed the little critters extra tryptophan supplements to do the trick (which resulted in increased mouse brain serotonin and a deadly increase in scratching and subsequent skin infections). Without the extra tryptophan boost, as little as 4% of protein in the meal will eliminate the carbohydrate tryptophan-boosting effect. To put that amount of protein into perspective, potatoes, rice, chocolate, and flour all have too much protein for them to increase tryptophan entry into the brain. Bascially, one has to eat pure starch (or a sugary drink) to achieve the pharmacologic carb-serotonin effect, long enough after a previous meal that no protein remains in the gut. That also means that pretzels and rice cakes wouldn't work for a "serotonin surge," and one must stick to marshmallows and jelly beans, which doesn't have much of an evolutionary precedent. This teensy amount of protein killing any "serotonin surge" is true in studies of rats, other primates, and humans (2).
In monkeys, Grimes et al. contrasted protein and carbohydrate breakfasts and found no impact on the levels of cerebrospinal fluid (CSF) tryptophan or serotonin turnover from individual carbohydrate-heavy meals. He went from protein being 25% to 6% of calories, and as protein declined, so did plasma and CSF levels of tryptophan due to lower amounts of raw material protein consumed. There was no immediate change in serotonin turnover or CSF tryptophan levels to individual meals. In the case of the monkeys, higher protein chronically is associated with higher serotonin, not higher carbohydrate, and blood changes of the ratio of tryptophan to other amino acids did not predict CSF tryptophan levels as it does in rats (which are what the Wurtmans studied). I must say the Wurtmans' studies are cited a good deal more often than Benton or Teff or Grimes.
So the carbohydrate = serotonin-increasing theory is likely bogus for humans, especially in the real world (yes, even on Thanksgiving), unless you happen to have a glass of lemonade for breakfast every morning, excluding all other food. BUT, that doesn't mean that carbs don't affect mood. What do studies without such ridiculous attention to actual mechanisms show?
In one study, a drink with 48 grams of carbohydrate and no protein decreased depression, anger, and confusion in women suffering PMS. Judith Wurtman also found that women increased their intake of carbohydrate and fat in the premenstrual stage, and experimental meals with only 4% of the energy as protein did improve the mood of women with PMS. (She did find these women had no higher ratios of tryptophan in their blood than women given control meals, and that women who had improved mood also had no increase in blood levels of tryptophan, suggesting that is not the mechanism for the improved mood) (4).
What do other researchers find? In several reviews, it was found that women generally report an increased craving for sweets and increased appetite in the pre-menstrual period. In humans and animals, there are also changes in basal metabolic rate in the different parts of the cycle, corresponding to the appetite changes, which makes sense. However, a review of studies of what women actually eat showed no difference in the total amount of carbohydrate consumed in any part of the menstrual cycle. Some studies have shown an increase in sweets, chocolate, and cake consumption before and during menses compared to ovulation (which occurs mid-cycle). So there is definite evidence that appetite increases just before and a bit during menstruation, but this corresponds to a higher metabolic rate, and it is foods generally containing a combination of carbohydrate and fat that are craved, not strictly carbohydrates.
In other studies, carbohydrate ingestion has been associated with better mood. De Castro had subjects keep food diaries and mood journals for 9 days, and found that on the days when more carbohydrates were consumed, mood was better (5), and people felt more energetic. He also found a cumulative effect - the more carbohydrate consumed over the week, the happier the person tended to be. In a study of 686 folks who reported their mood at mid-day and what they ate that morning, the more absolute carbohydrate the male subjects consumed that morning, the happier their morning mood. Absolute and relative amounts of protein and fat had no impact on mood in this unpublished study by Benton.
Finally, studies of experimental low carb vs. high carb diets of one week (6), three weeks (7), six weeks (8), and one year (9) show a better comparative mood and increased serenity in the high-carb dieters. The one year study was complicated by the fact that twice as many people in the low-carb group were on antidepressants, and the other studies may have been of too short a duration to bypass the "low carb flu" - and all of those low carb diets may have been high in omega 6. I don't have the energy to chase down all the full texts of all these studies right now (I've seen the year-long one, however), and the one-week one was done in cyclists, who might be cranky in the first week of a low-carb diet while training!
So what have we learned? Well, the Wurtmans and the carbohydrate-tryptophan-serotonin boost theory are scuttled, for now*. But carbohydrates don't seem to cause craziness, and for some people, they seem to help. Is the story any different for folks who have insulin resistance? We'll find out next time (though there might be a quick intervening review of a related topic in between that post and this one).
Please see the later post, Carbs and Serotonin, A Connection After All