Abstract
Dear Sir I would like to comment on certain aspects of the paper entitled ‘Peripheral autonomic potentials in primary and drug-induced headache’ (by S. Evers et al.) published in Cephalalgia (1998; 18:216–21).
‘Peripheral Autonomic Potentials’ (PAPs), more commonly known as Sympathetic Skin Response (SSR), are used in clinical practice to assess the reflex activity of the sympathetic sudomotor pathway. This method has been shown to be useful in peripheral autonomic neuropathies (1), and has also been used in a variety of central diseases, such as stroke, multiple sclerosis, Parkinson’s disease, and multiple system atrophy (2–5).
It is known that SSR (PAPs) can be influenced by emotional state, mental stress, ‘surprise effect’ of stimulation, skin temperature, and habituation. For latency, measurement of single responses has proved of little value because of the wide variability, whereas the mean or median latency has shown better reproducibility and reliability in most studies (6). Unlike latency, the mean amplitude has failed to show good reproducibility. The largest amplitude of a series of stimulations is less influenced by the aforementioned factors and is therefore considered to be the most valuable parameter (7). In this study, however, only a single SSR (or PAP) was elicited in each studied subject.
Since efferent conduction times are primarily dependent upon conduction in small unmyelinated fibres, measurement of latency is not generally useful in studies aimed at investigating the central nervous system (8); amplitude can provide far more information about the sympathetic sudomotor pathway at the central level (9).
The authors of this article found increased latencies and normal amplitudes of PAPs in patients suffering from different types of headaches, a result that cannot be explained by central sympathetic hypofunction. The discrepancy between latency and amplitude may suggest involvement of the peripheral nervous system. Alternatively, this discrepancy may be due to the technical problems encountered by the authors (see Results).
The authors state that ergotamine abuse induced polyneuropathy in a subgroup of patients. According to this observation, one might presume that peripheral problems occurred also in other patients as a consequence of the frequent and long-lasting intake of drugs that can induce polyneuropathies. An objective nerve conduction study needed to be performed in all patients to rule out this possibility.
The authors also found that SSR was absent or not reproducible in 16 patients and two controls, who arbitrarily were excluded from the study. They claim that SSR absence and lack of reproducibility in these cases were due to technical reasons (e.g. movement artefacts, too low a skin temperature). In this regard, it should be borne in mind that movement artefacts do not abolish SSR, but may induce artefacts, that do not usually affect this long-lasting response. Since these subjects should at least have been instructed to relax before and during testing, it is not clear why they moved and why the test was not repeated in these cases. It should, moreover, be considered that SSR studies have been performed in patients with Parkinson’s disease, cerebellar degeneration and multiple system atrophy, who do have involuntary movements (4, 8).
The authors used the air conditioning (see Methods) and subsequently claim (see Results) that too low a skin temperature could be responsible for the absence or non-reproducibility of the response in certain cases. How, I wonder, can the results be reliable if the authors did not follow the basic principles of neurophysiology? Temperature is so important in SSR (PAP) studies that the latencies may well have been increased as a result of low temperature.
Milena De Marinis MD, Department of Neurological Sciences, ‘La Sapienza’ University, Rome, Italy
Dear sir We appreciate the interest of Dr De Marinis in our paper and in peripheral autonomic potentials (PAP). However, we did not understand most parts or her criticism and the intention of the letter.
It is not true that we elicited only a single PAP. In the methods section, we say that PAP were reproduced twice separately for both hands (resulting in a total of six PAP).
We did analyse the subgroup of patients with drug-induced headache as a separate group. The other patients did not have a history of continuous drug intake (please read the first part of the Methods section).
The patients did certainly not move voluntarily during the test but were sitting with closed eyes and relaxing (see the second part of the Methods section). It is a common finding that some patients show involuntary limb movement after receiving an electrical stimulus. This can lead to movement artefacts which make the PAP unable to be analysed, Repeating the test sometimes is useless in these patients since they, for example, have a psychological fear of the stimulus. These patients must be excluded from analyses. However, it is good scientific practice to mention even those patients who could not be analysed.
An air-conditioned room was used to have a standardized room and body temperature. We certainly did follow the ‘basic principles of neurophysiology’ and tried to standardize skin temperature. As Dr De Marinis should know, it is impossible to obtain exactly the same skin temperature in all subjects during a neurophysiological study without using a heater which can influence PAP in several other ways. Furthermore, PAP are influenced by so many anatomic regions of the body that the simple skin temperature does not sufficiently reflect the influence of temperature at all. This is the reason why we were very keen on a constant room temperature and think that we followed the ‘basic principles of neurophysiology’.
S Evers, Department of Neurology, University of Münster, Münster, Germany