Stereology: A practical primer for neuropathology

We find the biggest advantage of stereology to be the emphasis on total number rather than density. We have used stereological approaches to assess neuronal number in human entorhinal cortex (6) and the superior temporal sulcus area during aging and in Alzheimer disease (8), in the rodent substantia nigra after MPP+ injection (13) in developmental studies of Down syndrome (14), and in transgenic animals to assess possible alterations due to amyloid deposition (15, 16). Each of these applications required different sampling schemes o provide he precision of estimates we felt necessary to answer the biological question posed. Is stereology worth the difficulty and expense? In many instances we would answer 'yes.' Does stereology replace and make obsolete all other methods? We think that this is an overstatement, and certainly in instances of fairly homogeneous structures where atrophy is not a major confound, or where the size of change is large compared with the size of errors introduced by compromising stereological principles, rigid adherence to stereological approaches may actually be disadvantageous in terms of time and effort to get to the same answer. Most importantly, samples of human brain are extraordinarily precious and unique resources used to help understand neural structure and disease processes. In some instances, an obligation to follow all the rules of stereology may make carrying out important studies impossible or impractical, and the opportunity to learn may be lost. Certainly varying from established stereological protocols raises concerns about potential bias, but even stereological techniques can have bias if the experiment is not well designed. Investigators can overcome biases to a great extent with an appropriate understanding of the limitations of their chosen techniques and by providing clear descriptions of techniques used in individual studies. The bottom line when deciding whether or not to use stereology is what counts is what one has to count.