When you review the literature for "evidence-based" treatments, there really are none for EDNOS (1), simply because research has focused on AN and BN. In anorexia, there are a few firm things to recommend - if severely underweight, refeed under supervision, and other treatments don't work particularly well until the starvation is to some extent reversed. For young people, family therapy is recommended. In bulimia, there is a particular type of cognitive behavioral therapy that is specific for the disorder and has good evidence base.
Eating disorders are among the most deadly of psychiatric disorders in the short term (anorexia is probably fatal in 5-10% of cases, though this number is dropping as acute treatment is getting better, whereas BN and EDNOS have a mortality rate of about 3.5%, and people with all the disorders, including BED, have a higher risk of suicide) - but the good news is at 10 year follow-up, between 70-80% of people with the disorders no longer have symptoms.
What is particularly frustrating about eating disorder literature is that no one has a real clue (backed by solid evidence) as to what causes them. There are some genetic links, some common psychologic features and environmental risk factors, and it is likely that people with low amounts of serotonin and dopamine in certain areas of the brain are more vulnerable to developing eating disorders.
My own suspicion (speculative, but sensible) is that the combination of societal pressures to be thin and our modern industrial diet's terrible track record of putting on fat lead to very common restricting behaviors, which basically makes everyone a member of a little Minnesota Starvation Experiment, only we are not locked down, so we continuously restrict, worry about exercise, then begin to obsess about food and have anxiety and depressive side effects, then we (quite sensibly) cave in to the evolutionary imperative to eat, regain the weight, etc. etc. A particular subset of people, often young women who perhaps face the most societal pressure, will develop extreme behaviors and become eating disordered. Thus my conjecture is that the industrial food is the problem, and wholesome unprocessed food (get rid of the excess sugar, excess linoleic acid, and gluten grains) a large part of the solution. My stance is actually as unproven as it is controversial - as many people with eating disorders restrict varieties of food, during recovery it is common practice to teach people to eat "everything in moderation" (though focusing on healthy, nourishing food, of course), so to suggest restricting whole categories of food again without a solid evidence base could be seen as feeding into the disorder.
But there is an issue with "everything in moderation," I think, putting aside the entire "paleolithic" nutritional paradigm and the neolithic agents of disease causing obesity in the first place. And that is that eating disorders are addictive (2). While some folks with addiction can actually, eventually, moderate their intake of the addictive substance, most people do better, at least for a while, with complete abstinence. Obviously one cannot completely abstain from food - but if we suspect the neolithic agents as building an addictive cycle neurochemically in the brain (and this is definitely speculation, but an educated guess, I would say), then it would seem that getting rid of them in the diet would be of obvious help. Once the situation is stabilized, and a more natural and realistic relationship with food and body and diet are adopted, then it would make sense to adopt more moderation to make life easier and less food-obsessed. But everyone is different - some people get very despondent with the idea that anything is off-limits. Treating eating disorders like every psychiatric condition requires some common sense, flexibility, and innovation.
Let's get back to my addiction angle for a minute, and, while I'm going to spend some time in a second post focusing on the neurobiology of binge eating, today I'm going to drop "eating disorders" and look at a new paper called Neural Correlates of Food Addiction. This study is actually about food and obesity in general.
The bottom line in America - 1/3 of adults are obese. Obesity-related disease is the second-leading cause of preventable death, and most obesity treatments are unsuccessful, with patients regaining all the lost weight within 5 years. Food and addictive drug use both result in the release of dopamine in the mesolimbic regions of the brain (at least in animal studies), so one gets a sense of "reward" and happiness from eating. Obese versus lean individuals show greater activation in the addictive behavior centers of the brain in anticipation of receiving "palatable" foods (obesity and food researchers here typically mean sugar and fat and salt - my still-unproven conjecture is that the massive excess of linoleic acid, fructose, and the easy, cheap availability of fatty-grain chips and the like and sweetened grain desserts are the issue, and while a gorgeous bone marrow dressing on a salad and a delicious steak would also light up the reward areas of the brain, it is part of the natural order of things to shut the reward cycle off if our bodies reach a certain level of fat stored on hand.) And, interestingly, one finds this to be the case in obese individuals and people addicted to drugs - obese people anticipate food intensely, but when the actual food comes, they have less enjoyment (in general, the reward brain areas light up less) than someone who is lean. The same is true for drug addiction - the brain seems to like the anticipation much more than the actual event.
In the study, 48 young (human) women with an average BMI of 28 were selected from women enrolled in a program "developed to help people maintain a healthy weight on a long term basis." Those with history of eating disorders, head injury, or current smoking, illicit drug use, or psychiatric diagnosis were excluded. Each participant was measured and weighed and took some tests assessing their level of food addiction (the research standard is a 25 item Yale Food Addiction Scale.) The test among these young women had a normal distribution.
Then the participants were put into an MRI for a baseline measurement after fasting 4-6 hours (no caffeine either). Then the women were exposed to pictures of a chocolate milkshake or a picture of a glass of water for a few seconds. Thereafter they received one of two "deliveries" - half a milliliter of chocolate milkshake* or a calorie-free tasteless solution meant to mimic saliva. All of this was randomized - some women got a picture of water then saliva, some milkshake picture then saliva, some water picture then milkshake, etc. And 40% of the time, no "taste delivery" was made to give the researchers an idea of what a scan would look like without the paired stimulus. The taste delivery occurred via some sort of tubing system within the MRI scanner itself, which is kind of a neat trick.
Results! The women (lean or obese) with a high level of food addiction on the scale test showed more activation of brain centers associated with addiction and reward in anticipation with the milkshake picture, and less activation with the actual milkshake. This is pretty solid human evidence that there are certain people who are truly addicted to chocolate milkshakes.
Interesting tidbits from this study - the addiction scores were not associated with BMI of the participants (average age of 20.8). However, those with high scores on the addiction scale test were far more likely to have periods of binge eating, emotional eating, and "problematic eating attitudes." The researchers felt that the young age of the sample might mean that the lean women with a high food addiction score would have increased risk of obesity, as older obese adults tend to have higher food addiction ratings. I suppose time will tell.
The main problem with the study is that the researchers did not measure hunger in the participants. Though they all fasted about the same amount of time, some may have been hungrier than others, and hungrier people have greater food reward signals in the brain. The other limitation I see is the selection in the first place - rather than selecting random young women from a college campus, for example, they chose those enrolled in a healthy eating seminar.
Well. What did we learn, really? "Highly palatable" food can be addictive to certain susceptible individuals. Usually things are highly palatable because they are somehow good for us. I have the feeling that in an environment without the neolithic agents of disease, the food reward system would work as planned, just like the sex reward and the exercise reward. Jack up the system with weird chemicals in excess of anything we ever experienced in evolution - like heroin, tobacco, or ho-hos, and some people will develop a problem.
* Obesity researcher milkshake recipe: 4 scoops Haagen-Dazs vanilla ice cream, 1.5 cups of 2% milk, and 2 tbs of Hershey's chocolate syrup. Personally I would recommend using some variety of artesian or homemade reduced sugar chocolate ice cream and whole milk without the syrup ;-) Also whipped cream and chocolate shavings on top.