Monday, September 15, 2008

Food and Behavior

A colleague of mine who works in the public school system recently remarked that one of the parents with an ADHD child wanted to know whether her physician was correct in prescribing fish oil instead of a stimulant medication for her pre-adolescent son. This immediately prompted a deja vu flashback to my years studying the role of food additives (the Feingold diet), aspartame, coffee, and other foods as cures or causes of ADHD. Everything I learned back then was summarized in several journal articles and 2 books (Food Additives and Hyperactive Children, Springer, 1980; and Feeding the Brain, Perseus Publishing, 1989). What I learned from the Feingold/food additive controversy was the following:
  • Ben Feingold was an honest scientist and allergist with a sincere belief that Food Additives caused hyperactivity; a white-haired, persuasive charismatic figure who convinced thousands of his theory.
  • Just as he alleged, it was possible to "turn the hyperactivity on and off by giving or removing food additives." An effect we replicated in an ABAB design;
  • But all of the effect could be accounted for by placebo once a proper control group was added.
  • The only exception appeared to be for pre-school children where there was a slight indication that food additives increased hyperactivity (replicated by the Wisconsin group of nutritionists).

The power of placebo was strikingly illustrated by the very first child in our double-blind trial where food additve-loaded chocolate cookies were compared with additive-free chocolate cookies. On the first day of the trial the parent of the 6 yr old boy called, angrily threatening to sue us: "I don't know what you guys put in those cookies, but Kevin grabbed a knife and tore up our couch; he then took a hammer next door and destroyed the neighbor's motorcycle." Oh, oh...a looming court suit with big dollars flashed before me, but as fate would have it, you guessed it, he was on the placebo cookies. (Incidentally, this confirms Denny Cantwell's dictum that any red-haired boy named Kevin will grow up to be hyperactive.)

As Martin Orne once showed, the first thing a subject learns in an experiment is to obey what they think the experimenter wants to find.

Now back to feeding the brain. While at G. Washington and Children's Hospital in Washington DC, I carried out a number of experiments based on the Wurtman-Fernstrom hypothesis. They had demonstrated that a carbohydrate challenge elicits an insulin reaction which causes an efflux of large neutral amino acids (LNAA) from the blood, with the exception of tryptophan, which is lightly bound to albumin. Since LNAA cross the blood-brain barrier by competitive transport, tryptophan is given preferential input to the brain. Since tryptophan is a precursor for serotonin, the result is an increase in serotonin output, which is not regulated by positive feedback, so a burst of it can produce a sedative effect or other effects on the neurotransmitter systems.

We carried out several experiments in which we pre-fed normal and ADHD children with either a high protein breakfast, a carbohydrate breakfast, or a fasting condition, and tested them following a high level of carbohydrate beverage or aspartame placebo. An indwelling catheter was used to monitor hormonal responses, and tests of attention (CPT), cardiac response to warned reaction time (RT), and event-related potentials (ERP), monitored throughout the day. There were several results compatible with the Wurtman-Fernstrom hypothesis:

  • Baseline levels of carbohydrate were significantly higher in the ADHD than normals. (Judy Rapaport at the NIMH had reported no differences in serum carbohydrate levels, but the breakfast the children ate beforehand was not monitored, thus invalidating the results).
  • Attentional performance on the CPT (continuous performance task) and evoked cardiac response was impaired in the ADHD, but only when they had a carbohydrate breakfast, not a protein or fasting breakfast. The effect aoppeared to be in the early processing phase.
  • ERPs showed marked lowering of amplitude and slowed latency in the carb condition but only for the ADHD children.
  • Two hormones important in regulating carb levels (cortisol and growth hormone), were under-reactive in response to the carb challenge compared with controls.

I don't think we can conclude that high carbohydrate levels and increased serotonin cause ADHD. Instead, we propose that the recognized deficits in dopamine production or dopamine receptor function may be responsible for the hormonal effects, which in turn lead to a dysregulation of carb levels and the tryptophan/serotonin effect. This hypothesis has not to my knowledge been tested, and remains in my mind as an important future research problem.

Now, back to the fish oil issue. There is a small number of studies on the role of essential fatty acids (EFA), particularly omega-3, and ADHD, and some related areas like bipolar disorder and depression. In my next posting I will review this evidence. As a preview I can say that unlike neurofeedback and megavitamins, it is not the dictum of "Let the buyer beware" that applies, but the Scottish one, "Not proven."