Although the most commonly used psychiatric classification system in the United States (i.e., the DSM-IV) employs an exclusively categorical approach, many researchers have hypothesized that at least some psychological disorders exist along a continuum. The taxometric method allows researchers to examine empirically the latent structure of disorders to determine whether they are categorical or dimensional in nature. The latent structure of restrictive eating behaviors will be examined using the taxometric method in both a non-clinical and clinical sample of women, and follow-up construct validation analyses will be conducted. I hypothesized that restrictive eating would exist along a continuum that also encompasses normality and obesity. Furthermore, I predicted that individuals with relatively higher levels of restrictive eating pathology would have higher rates of the short allele of the serotonin transporter gene, the A allele of the serotonin 2a receptor gene, perfectionism, compulsivity, obsessionality, and harm avoidance, and lower levels of novelty seeking, than those with relatively lower levels of eating pathology. As hypothesized, restrictive eating behaviors were dimensional in nature (i.e., individuals with extreme restrictive eating patterns differed in degree, but not in kind, from those with less restrictive eating patterns). In each sample, a follow-up cluster analysis identified two clusters of women in the data: one labeled "restrictors" and one labeled "non-restrictors." Contrary to predictions, individuals in the "restrictors" clusters did not reliably demonstrate higher frequencies of the short allele of the serotonin transporter gene or the A allele of the serotonin 2a receptor gene than "non-restrictors." Relatedly, they exhibited higher levels on some, but not all, of the personality measures that were expected to be positively correlated with restrictive eating behaviors. In sum, restrictive eating behaviors appeared to be dimensional in nature, although certain constructs previously demonstrated to reliably covary with restrictive eating behaviors were not linearly associated with the degree of restrictive eating pathology. Implications for etiology, prevention, assessment, diagnosis, and treatment are discussed.
