[Skip to Navigation]
Citations 0
Letters to the Editor
March 2002

Prolonged Brainstem Auditory Evoked Potentials: An Autism-Specific or Autism-Nonspecific Marker

Arch Gen Psychiatry. 2002;59(3):289-290. doi:

In reply

We thank Drs Nagy and Loveland for raising 2 important issues. First, they underline a major point discussed in our article1 (ie, the likely genetic heterogeneity in autism). We wish to point out that the proportion of 52% they raised refers to those families in which neither the proband nor the parents (mother and father only) showed abnormally prolonged interpeak latency (IPL). The percentage of families in which neither the proband nor the relatives showed abnormally prolonged IPL was 27%. They are right to mention that if there are 2 subgroups of autism characterized by a dichotomous brainstem auditory evoked potentials (BAEP) characteristic (impaired and nonimpaired), then the correlation between IPL and autistic symptoms that we performed in the whole sample may not be truly informative. So, as suggested, we compared the group of non–BAEP-impaired autistic probands who also had non–BAEP-impaired parents (N = 17) with the remaining group of probands (N = 16). Again, we found no significant phenotypic differences on the Vineland Adaptive Behavior Scales (P = .25) and the Autistic Diagnostic Interview (P = .67; P = .52; P = .08). Hence, these analyses suggest either an absence of heterogeneity or that the clinical phenotypes we are using in autism are less than valid for distinguishing subsets of probands, an obstacle also encountered in the genetics of other major psychiatric disorders such as schizophrenia.2-4 We also compared the proportion of subjects exceeding the 95th percentile in the group of relatives of BAEP-impaired probands vs the group of relatives of non–BAEP-impaired probands on the I-III IPL and found no statistically significant differences (P = .20). Our study indicated that the I-III–impaired relatives were not seen only in families of BAEP-impaired probands. However, we might not have sufficient power to assess whether I-III–impaired relatives occurred more frequently in families of I-III–impaired probands.