The validity, reliability, responsiveness and sensitivity/ specificity of EMG have to be discussed separately for the particular physiological information sought from EMG, and for various EMG techniques, types of recording, and applications. EMG is often falsely understood as ‘one method’, and as a method precisely measuring muscle ‘function’. Muscle function is, however, complex and different EMG techniques address different aspects of it, but never really cover all of it. Indeed, motor unit EMG techniques, for instance, are more useful in diagnosing denervation and reinnervation (i.e. helping in diagnosing a neurological lesion) than in diagnosing the functional deficit (i.e. quantifying the number of motor units and thus providing data that would be directly functionally relevant).
It has to be distinguished whether EMG is used to detect the pattern of muscle activity, or rather to detect muscle denervation/reinnervation. EMG methods are reasonably reliable, reproducible, sensitive and specific to diagnose muscle denervation/reinnervation – but this is mostly ‘expert opinion’ relying on long-term correlation of clinical and EMG findings in conditions affecting musculature in general. Correlation of EMG findings to muscle function (strength, power and endurance) is – in the individual – insecure (excluding instances of minor or severe/complete denervation).