EMG and nerve conduction studies can only measure damage to larger, myelinated nerves. These include all motor nerves and the sensory nerves that detect mechanical touch, vibration, and friction.

There are much smaller/thinner nerves, however, that have only rudimentary or no myelin coverings at all. These so-called small-fibers subsume the sensations of pain and temperature, much of the autonomic system, and may be damaged without EMG/NCV studies being able to pick up on it--the technology is just not advanced enough at this juncture.

To measure dysfuction of those nerves--and diabetic neuropathy often presents as a small-fiber syndrome--such tests as quantitative sensory testing and sudomotro axon reflex testing are often done, but the current gold standrard is skin biopsy to determine intraepidermal small-fiber density and condition. Take a look at this:

http://www.medscape.com/viewarticle/563262_1

There's also some interesting research being done on using skin biopsies for certain types of demyelinating neuropathies:

http://brain.oxfordjournals.org/cgi/...ull/128/5/1168

Also take a look at this list of serological tests for neuropathy causes:

http://www.questdiagnostics.com/hcp/...eralNeurop.htm

The sad part about all this is that one may have neurological symptoms from both spinal issues and/or more peripheral nerve damage, and these symptoms can be exactly the same and appear in exactly the same bodily regions. EMG/NCV can help localize, or rule out, just where the damage may be occurring along the line, but its ability to do so is much diminished when the small fibers are involved.