The threshold assessment procedure includes non-invasive electrocutaneous stimulation of the skin of the thumb and forefinger of one hand by pressing the electrodes of the electrical stimulation unit. When reaching the respective threshold level the subjects release there fingers from the electrodes. The detection threshold (DT) was defined as the first pricking sensation and the pain threshold (PT) when the pricking sensation was altered to the first sensation of pain.

The threshold values at the respective levels were automatically recorded immediately when the fingers were released from the electrodes, but in that moment not shown to the subjects making them blind to the ongoing assessments.

The threshold assessment unit is controlled by a microprocessor (PainMatcher^®, Cefar Medical AB, Sweden). The generated current is distributed with a monophasic rectangular pulse of 15 mA and 10 Hz. The output intensity increases by gradually widening the pulse duration in steps of 4 μs to a maximum of 396 μs, i.e. in a total of 99 steps directly related to the output. The maximum electrical charge per pulse is 5.9 μC. The contact surface area, and hence the resulting current density, is ensured by a certain load of minimum finger pressure against the electrodes. Loads between 0 and 13 kΩ secured the output of 15 mA. The numerical cut-off range is 0–99.

Both electrical DTs and PTs were recorded on four occasions separated by 10 minutes: before, during and after vibration stimulation.

The vibratory stimulation (Vitamed, Germany) was applied with a rectangular probe of 13 × 20 cm to the dorsal aspect of the forearm, covering the dermatomes C5-8, with 3000 Hz, and a constant and moderate pressure for 20 minutes. The therapist was well-known to the experimental subjects.

The mean value and standard deviation (SD) were calculated for age. The threshold assessments were regarded as subjective estimations and the produced threshold data as ordinal data, here presented as the median and range (minimum to maximum) for the numerical units of the PainMatcher (PM) values 2829.

The proportions of subjects with increased, unchanged, and decreased threshold values on the second occasion were calculated. The hypotheses of no change in threshold assessments between before and after vibration were analyzed by the non-parametric sign test. Mann Whitney U-test was used to analyse gender differences. A p-value less than 0.05 was regarded as significant.

Higher baseline DTs and PTs to electrical stimulation in men compared to women seems to be a frequent finding in previous studies, e.g. 2224.

Previous studies have shown that in order to get optimal pain relief, the best site to apply the vibration probe is either in the area of pain or in dermatomes no more than two segmental levels from the pain 131030, or in the same dermatome on the contralateral side 10, or in a trigger point outside the painful area using moderate pressure 30. Vibration distal to the site of threshold measurement also seems to be more effective, at least during the stimulation 31. Furthermore, findings in a previous report indicate that a larger size of the stimulus probe results in more pain relief 3. In this study, the vibration was applied across up to four dermatomes, proximal to and in the same arm as the test stimulus, and the vibratory probe was relatively large (13 × 20 cm) with the aim to cover a reasonably large area. The 20 minutes used as standard application duration in this study is within the time interval used in previous studies, and was shown to be the shortest duration to elicit maximal pain relief for the most patients in a group with myalgia 6, but the most efficient application time for various pain states seems to be unknown. It could be that a longer application time than 20 minutes would have been more effective. Another factor which could be of importance is cyclic variation in thresholds to pain stimulations linked to the menstrual cycle 32. We did not include this variable in the present study.

The finding of this study that vibration may increase the threshold for experimentally induced pain is in accordance with earlier findings 101516. It is unclear how afferent signals elicited by vibration interfere with central transmission of nociceptive stimuli even though there are indications of mechanisms that include purinergic mechanisms 3334, lowering of spinal substance P 35, but not, however, ligands to naloxone sensitive μ-opioid receptors 3637.

Clinically, the effect of vibration stimulation in patients with different pain states varies between studies. Some demonstrated a pain lowering effect 131030 whereas others were unable to show a statistically significant effect 3839.

The reason why women tend to have lower thresholds to some types of sensory stimulation including painful stimuli is obscure. The sensitivity to vibration as such does not appear to be different between men and women 40, indicating that differences in the intensity of the afferent signal was not the cause.

Another possibility is gender related differences in temporal summation. Temporal summation of painful stimuli has previously been shown to be larger in women 4142, see however 43. Whether this is also true for non-painful stimulation does not seem to be known. It could be that longer vibration duration in men would have cancelled out the observed threshold differences between men and women.

Even though gender related differences in pain perception have been reported 2427, the gender differences appear to be rather small. The differences have been attributed to experimental, social, psychological and physiological factors, including the experimental setup and mode of stimulation, attention, emotional reactions including anxiety, willingness to report pain and gender of experiment assistant, and to catstrophizing 13202344. Some of the complexity of the issue is also illustrated by the finding that repetitive mechanical painful stimuli were rated equal for the first stimulus, but higher in women than in men for the fifth and tenth stimulus, respectively 41, indicating central mechanisms. Also in line with this, less habituation in women than in men was found after intramuscular glutamate injections 26, and greater temporal summation following repetitive noxious stimulation 42.