--but here's the short version.

When the norms for intraepidermal skin density number were established, after a lot of normal and symptomatic individuals were tested (mostly at Johns Hopkins), the researchers there decided--rather arbitrarily, from my view--to set the fifth and ninety-fifth percentiles as "abnormal" and indicative of small fiber neuropathy. (There is a lot of precedent for picking those numbers in medical research, though, that has a lot to do with normal--bell curves shaped--statistical distributions.)

There was a very high standard deviation noted in the research samples, though--that is, the absolute range of number of fibers per cubic centimeter of skin was quite broad. So a large range of numbers would be considered "normal". There was a tendency noted for absolute numbers to decrease with age.

The problem is that since most normal, asymptomatic people don't have skin biopsies done, we don't know what number they start at. A person could be symptomatic, have a skin biopsy performed, show up at the twentieth percentile and be told "no neuropathy". Problem is we don't know if that person was ALWAYS around that level--it's possible s/he could have once been at the fiftieth percentile, say, and if that had been known this biopsy figure would represent a diminishing of fiber density.

This is why the condition of the fibers is supposed to be examined as well. As enbloc notes, fiber swelling, along with excessive branching, may be noticed in some asymptomatic people with no symptoms, but it is not a "normal" finding and usually would prompt some more investigation or at least monitoring. Fortunately, skin biopsy is non-invasive and repeatable, so one can have these done serially (I have) and compared for further diminishing of density or deterioration of condition--or, hopefully, for improvement of density and condition. (This is how we were able to clinically document that my acute-onset body-wide small-fiber syndrome, whatever it was, was healing slowly over time--though no one knows if I will ever get back to whatever my original nerve fiber density numbers were, as we don't know these. But I have gone from the third and fifth percentiles to the sixteenth and twentieth percentiles and the nerve fibers are much less swollen and branched than they were during my first skin biopsy.)