Diet, Depression, and Anxiety

This post will focus on a single paper, published in March 2010 in the American Journal of Psychiatry: Association of Western and Traditional Diets With Depression and Anxiety in Women. (Thank you to Dr. Hale for pointing out the study). In the introduction, the authors make note that depression and anxiety are highly prevalent with other chronic dietary-related illnesses such as cardiovascular disease, obesity, and type 2 diabetes. At the same time, psychiatry lacks evidence-based prevention and treatment strategies based on dietary modification. That may be for the best, considering what has happened with obesity and type II diabetes over the past 30 years, but that is, in part, why I'm trying to get the information out, blog-wise.

So - the study! Well, these researchers in Australia hijacked an ongoing study, called the "Geelong Osteoporosis Study." Thousands of intrepid women were randomly picked from compulsory voting rolls, and all told, 1046 women ages 20-93 were followed for roughly 10 years for the diet and depression part. Each participant filled out yearly questionnaires about her diet. (Problem number one - imagine I gave you a questionnaire about your usual consumption of 74 foods, 6 alcoholic beverages, and the type of bread, dairy products, and fat spreads you used? How accurate could it be? The researchers say the "comprehensive food frequency questionnaire" was a validated instrument, but one must keep limitations in mind.)

Then, all participants were given a SCID (that's a standard structured clinical interview used to diagnose psychiatric disorders in research), and they looked for current diagnoses of major depressive disorder, dysthymia (a low-grade, ongoing depression), and anxiety disorders. Also, a wide variety of statistical analysis tools were used to account for so-called "covariates" such as socioeconomic status, physical activity, alcohol consumption, and smoking. (Problem number two - which is the major problem with any observational study - one can never really account for all the covariates. For example, people who drink moderate amounts of wine have less heart disease. Wine could be a factor in that. Or perhaps people who drink wine happen to exercise more. Or maybe people who drink wine also have magical hearts. We really don't know. If you take a group of people and force-feed some a glass of wine or two a day, and then tell another group to abstain, then see if there is a difference in heart disease between the two groups, then you have a prospective trial, and that's not the kind of data we're talking about with the current study I'm examining, which is an observational study. We end up with associations and correlations with such studies, which are interesting, but could be meaningless. Tom Naughton brings up this issue and the 2010 US dietary guidelines in this blog post. Don't click that link if you have extremely delicate sensibilities.) The researchers weighed and measured the height of all the participants also.

Results! Everyone's diet was analyzed and segregated into three basic groups - traditional, Western, and modern. Traditional diets were comprised of vegetables, fruit, beef, lamb, fish, and whole grain foods. A Western pattern was associated with meat pies (Australian fast food), processed meats, pizza, chips (I seriously do not know if they mean french fries or potato chips here - Australian readers, help me out!), hamburgers, white bread, sugar, flavored milk drinks, and beer. The "modern" diet consisted of fruits, salads, fish, tofu, beans, nuts, yogurt, and red wine (in other words, the people who read the news reports on all the observational studies out there...).

And the punch line? A traditional dietary pattern "was associated with a lower likelihood of depressive and anxiety disorders."

Traditional fruit, veggie, lamb, beef, fish and whole grain eaters had a 25% lower risk for major chronic disease (cardiovascular disease and cancer) after 10 years. They had 35% reduced odds for having major depression or dysthymia, and 32% reduced odds for anxiety disorders. The "Western" (junk) and "modern" (bean, fish, wine, and tofu) eaters fared about the same, but the Western eaters were slightly more depressed.

So there you have it! Beans and tofu and meat pies are correlated with depression and anxiety! Well, the authors of the study are pretty fair about the limitations of their design in the discussion. The do mention another study and how high-fat, high-sugar diets caused decreased hippocampal BDNF in animals, and that diets high in refined carbohydrates are associated with more inflammation. They also made note that a Mediterranean-style diet (which would roughly correspond to the traditional and modern diets) tends to decrease inflammatory markers.

I wonder what the data would look like without the whole grain eaters? Also, full fat dairy versus low fat dairy. Why do researchers never present the really interesting stuff to a paleolithic diet-inspired psychiatrist?

(A sobering thought is that I may be the only paleolithic-diet inspired psychiatrist).