Drs. De Bellis and Keshavan Reply

To the Editor: We appreciate the comments by Mr. Tae and Dr. Hong. The amygdala and hippocampus are often hard to differentiate from each other, even at a histologic level. On the basis of earlier work in studies of child and adolescent psychiatry (1–4), we utilized a landmark method to differentiate these structures. In this method, the slice showing the most anterior mammillary bodies is used as the amygdala-hippocampus boundary. The disadvantage of this landmark method is that measures of the anterior hippocampus may include some of the posterior amygdala on a small number of slices. As described in our article and outlined in Figure 2, our measurement of the anterior hippocampus can potentially include a small part of the posterior amygdala. Therefore, we explored whether this difference is related to the few slices that contain both the amygdala and hippocampus. Secondary analysis of our results revealed that the smaller hippocampal volumes seen in subjects with adolescent-onset alcohol use disorders was not due to our decision to use this landmark method (total hippocampal volume adjusted for intracranial volume: F=6.27, df=1, 33,p<0.02).

As stated by Mr. Tae and Dr. Hong, recent studies in adult psychiatry show that visualizing the hippocampus perpendicular to its long axis, a modification of our landmark method, has the advantage of separating the posterior amygdala from the anterior hippocampus and the disadvantage of including the adjacent entorhinal cortex in the measurement of the anterior amygdala (5–7). We agree that MRI measurement methodology is an important issue to keep in mind when comparing data across different studies. Since standardizing hippocampal measures across development is an important issue, we have began utilizing both methods in our image analysis.