To the Editor:

Dr. Kaye et al. suggested that "childhood anxiety represents one important genetically mediated pathway toward the development of anorexia nervosa" (p. 2215) and that this is reflected in onset of OCD before anorexia. Dr. Kaye et al. reported that of 94 patients with anorexia nervosa, 35% also had OCD. Upon examination of these, about 12% dropped out, and of those remaining, 23% had OCD before they had anorexia nervosa. Thus, 94×35%×88%×23%= seven individuals of 94 (7%) had OCD before they had anorexia nervosa. A French study on the same topic (1) reported that two (7%) of 29 patients had OCD before they had anorexia nervosa. Thus, the expression of OCD before anorexia nervosa is rare.

The lifetime prevalence of anxiety disorders has been reported to be very high, with estimates as high as 12%–18% and even 30% (see Dr. Kaye et al.) having been reported. Because the incidence of OCD increases exponentially when an individual approaches puberty and there are no sex differences in OCD (2), we must conclude that the major expression of the "genetically mediated pathway" for anxiety disorders that Dr. Kaye et al. suggested is a "pathway toward the development of anorexia nervosa" occurring much later than the onset of anorexia nervosa. Also, we must assume that these hypothetical genes are expressed as anorexia only in girls (the prevalence of anorexia is very low in boys) but that the same genes are expressed as OCD in both girls and boys. Whether genes with such time-dependent, sexually dimorphic phenotypic expression patterns exist remains to be investigated.

Furthermore, although it is apparently possible to diagnose OCD retrospectively at 5 years (see Kaye et al.) or even at 3 years of age (1), it appears that childhood OCD is 8–12 times more prevalent in the United States (see Dr. Kaye et al.) than in England (2). If OCD causes anorexia, one would expect anorexia nervosa to be about 10 times more prevalent in the United States than in England. There is no evidence that this is the case. The available evidence does not support the hypothesis that OCD is a risk factor for the development of anorexia nervosa.