Abstract
Until now the clinical investigation of cervical pain due to whiplash injury is mainly based on finger palpation. The present study introduces a PC-interactive pressure algesimetry to standardize cervical pain measurement. Pressure pain scores of the splenius and trapezius muscles of 23 patients with an acute cervical syndrome after whiplash injury were compared to those of 24 healthy subjects. The pressure painfulness of neck and shoulder muscles was significantly increased in whiplash patients. The splenii muscles showed an equally increased muscle pain whereas the trapezii muscles showed a left-sided preponderance of painfulness, possibly due to the seat belt position in this group of motor vehicle drivers. The computer-interactive pressure algesimetry enables a standardized and rater-independent quantification of the cervical syndrome with neck and shoulder pain caused by whiplash injury.
Introduction
Neck pain and neck stiffness, often accompanied by a post-traumatic headache, are the most common complaints after whiplash injury (1). In the standard clinical examination the amount of pressure painfulness in the strained muscles is mostly determined by pressure induction in the respective muscles by means of finger palpation (2). The second and third finger are pressed against the muscle of interest while performing small rotatory movements. The degree of muscle pressure pain is estimated by the examiner simply by means of observing the patients’ behaviour with respect to the degree of aversive motor reactions, pain induced mimic reactions, or comments about the pain (3–5). The clinical evaluation of current pressure painfulness thus remains subjective and qualitative in nature. It depends on subjective specifications of the patient as well as on different investigation techniques and assessment qualities of the investigator. The interrater reliability is known to be low (2, 4, 6). Therefore objectification and course assessment of post-traumatic muscle pain caused by whiplash injury is complicated. This applies especially to the acute cervical pain syndrome following whiplash injury (7). Considering the background of associated litigation, it is particularly difficult to draw conclusions from the patients’ statements or behaviour patterns about the pressure painfulness of the strained neck/shoulder muscles (8).
Keeping this in mind it was the aim of the present study to quantify apparatively the increased pressure painfulness of the neck and shoulder muscles after whiplash injury. This was achieved by application of a pressure algesimetry which does not depend on the examiners experience. In a preliminary study (6) our group showed that the pressure algesimetry used in the present investigation – a further developed technique first described by Göbel (9) – can psychophysically quantify muscle pain in a valid and reliable manner.
Methods
Subjects
Twenty-three patients (13 female, 10 male, mean age 28.9 ± 12.1 years SD, range of 18–53 years) with a post-traumatic cervicocephalic pain syndrome with symptoms of neck and mostly dull and bilateral occipital pain following a whiplash injury of grade 1–2 according to the Quebec Task Force (10) were included in the study. Grade 1 is associated with complaints about the neck such as pain, stiffness, or tenderness without objective physical signs, grade 2 includes decreased range of motion and cervical point tenderness. No patient showed focal neurological symptoms such as radicular pain sensation or radicular deficits in the clinical investigation to exclude confounding factors on the dependent variable. Further exclusion criteria were a direct head impact, a direct neck trauma or painful accompanying injuries as well as a previous neurological or psychiatric history and chronic pain syndromes, including different kinds of headache (i.e. chronic tension-type headache, migraine, cervical pain). All patients were drivers involved in rear-end collision accidents and suffered from painful tension of the neck muscles. Patients were investigated in the acute post-traumatic stage, on average seven days (± 2.9 SD, maximum 14 days) after the accident. They were recruited from four emergency rooms in the area.
The control group consisted of 24 healthy age- and sex-matched volunteers. The mean age was 28.8 years (± 10.2 SD, range 19–53 years); 11 female, 13 male. Exclusion criteria were as for the whiplash group with special consideration that none of healthy control subjects suffered from idiopathic headache. Subjects experiencing tension-type headache more than one day per month were not included in this study. All healthy control subjects were free of acute neck pain. Except for oral contraceptives, none of the controls took medication (especially central acting drugs or analgesics) seven days before the investigation.
There were no significant differences in age and sex distribution of the two groups (independent t-test, P = 0.95; χ2 test, χ2 (1) = 0.54, P = 0.56). All subjects were fully informed about the purpose of the examination and gave their informed consent.
Data acquisition
All patients underwent neurological assessment before the recordings. Previous medical history, actual symptoms and accident history were taken by means of standardized interviews and history questionnaires.
A pressure algometer (9) was utilized to apply quantifiable defined pressure stimuli in a standardized way using a stylus with a stamp diameter of 2 mm and a contact surface area of 3 mm2. A constant pressure of 1310 kPa (400 g) was applied to each muscle for a stimulus induction period of 180 s. Pain evocation was terminated prematurely if the maximal tolerable pain intensity was reached.
Muscle pain was induced in the left and right splenius muscle and in the left and right trapezius muscle. Each muscle was investigated separately. To avoid pressure application to tender points the position of the algometer stylus was determined in the following standardized manner. For the trapezius muscle the contact point was the half the distance between the processus spinosus of C7 and the acromion, for the splenius muscle the contact point was two thirds of the distance between the processus spinosus of C7 and the mastoid. Patients and healthy control persons were seated comfortably in a chair, the lower arms supported by arm rests. For stimulation of the splenius muscles the head was supported by an anatomically shaped polystyrene block in a slightly inclined position. Before the recordings were carried out the participants could acquaint themselves with the pressure algometer.
In order to exclude the possibility of subjects’ pressure pain reports being biased by a mediating investigator, we extended the pressure algesimetric method (9) by applying PC-aided response recordings. Furthermore the software yielded the possibility to acquire the pain perception and its development in a high temporal resolution. During the continuous pressure pain application subjects rated the perceived pain intensity via cursor keys onto a graduated scale on a PC-display. On that visual analogue scale (VAS) pain could be evaluated from 0 (no pain at all) to 10 (unbearable pain). Subjects were instructed to press the cursor pointing upward (downward) whenever they felt that the pain has reached an advanced (diminished) level on the VAS. Using a custom written software program their behaviour was recorded continuously and the degree of pressure pain as a function of time was calculated offline. The following parameters of the acquired graphs served for quantification of painfulness: (i) time until pain perception (pain threshold; in seconds) (ii) maximally experienced pain intensity in VAS units (pain maximum) (iii) time until pressure pain became intolerable (tolerance threshold; in seconds) (iv) rise of the slope of the pain intensity-time function (VAS units/s) and (v) area under the curve of the pain intensity-time function (integral; VAS units∗seconds).
Data analysis
For each group, mean values of the pain threshold, of the maximal pain intensity, of the tolerance threshold, of the slope of the pain intensity-time function and of the area under the curve were computed separately for each muscle. Multivariate analysis of variance (MANOVA) and subsequent t-tests, corrected for multiple comparisons according to Holm (11), to elucidate the significant overall effect were used to compare the muscle pressure painfulness of whiplash patients and healthy control subjects. T-tests for dependent samples were applied to investigate differences of pain perception (left vs. right side of the body) within the groups. Probability of type one error (α) was set to 5%. Data were analysed using the Statistical Program for Social Sciences (SPSS, Chicago, USA), version 9.
Results
Figure 1 shows the pressure pain intensity time course of the left and right splenius muscle. In many whiplash patients (top row) the muscle pressure pain intensity reaches the upper limit of maximally tolerated pressure pain with a pain strength of 10 on the VAS relatively fast. At that time pressure stimulus application was terminated. The fast achievement of the tolerance threshold determines the rapid rise of the graph with the high slope and the calculated large area under the curve up to the stimulus duration of 180 s. In the bottom row of Fig. 1 the corresponding pressure pain functions of the left and right splenius muscles of 10 healthy subjects are illustrated. In all cases pressure pain is tolerated during the whole stimulation period and the tolerance threshold is not reached. Consequently the course of the pain intensity-time function rises less sharply with lower values for the slope and for the integral below the curve.
Single recordings (normal line) and mean values (bold line) of the time functions of pressure painfulness in VAS units of the left and right splenius muscle. In a paradigmatic manner the response curves of 10 patients (WP; a, b) each with neck pain due to acute whiplash injury are compared with those of 10 individuals from the healthy control group (CG; c, d). There is an increased pain maximum, a steeper slope and a larger area under the curve (integral) of the pain intensity-time function in whiplash patients.
Similar results were observed in the algesimetric investigation of pressure painfulness of the left and right trapezius muscle (Fig. 2) with a steep rise of the pain intensity-time function indicating the raised pressure painfulness in the patients strained shoulder muscles. In contrast to the patient group, healthy subjects were less sensitive to pressure pain induction in these muscles as expressed by a corresponding flat course of the pain intensity-time function.
As for the splenius muscles (Fig. 1) single recordings (fine line) and mean values (bold line) of the time functions of pressure painfulness in VAS units of the left and trapezius splenius muscle are presented. WP, whiplash patients; CG, control group.
The statistical analysis (MANOVA) revealed a significant effect for group membership (Λ(20,26) = 0.000). Analysing the dependent variables separately, the following significant differences between patients and healthy control subjects were found (Table 1).
Intergroup comparison: mean ± SD of investigated parameters of pressure painfulness (pain threshold, maximal pain, tolerance threshold, slope and integral) and P-values of post hoc t-tests, calculated for the splenius and trapezius muscles, separated for left and right and in addition for the average of both sides
WP, whiplash patients;, CG, control group.
∗P < 0.05, corrected for multiple comparisons (11).
Within the patient group the maximal muscle pressure pain was significantly increased for the left (P < 0.000) and right (P < 0.000) splenius muscle and for the left trapezius muscle (P < 0.000). The VAS scores for the right trapezius muscle did not differ significantly between the groups (P = 0.091).
The tolerance threshold (defined as a function of time) was reached more rapidly in all four investigated neck/shoulder muscles within the whiplash than the control group. Left (P < 0.000) and right (P < 0.000) splenius muscle and left (P < 0.000) and right (P = 0.003) trapezius muscle of the patients were significantly more susceptible to pressure pain. Correspondingly the rise of the pain intensity-time function's slope was significantly larger. Left (P < 0.000) splenius muscle as well as left (P = 0.001) and right (P = 0.006) trapezius muscle showed a significantly steeper slope for whiplash patients. The integral was also significantly larger for the left (P < 0.000) and right (P = 0.001) splenius muscle and for the left (P < 0.000) trapezius muscle in the patient group. Right splenius’ slope and (P = 0.018) and right trapezius’ integral (P = 0.031) did not differ significantly between the groups.
The results concerning the pain threshold diverged from the remaining pain related parameters. There were no statistically relevant differences in the pain threshold between the two examined groups neither for both splenius muscles (P = 0.006; P = 0.223) nor for the trapezius muscles (P = 0.115; P = 0.877).
Figure 3 illustrates the differences in calculated pain related parameters between whiplash patients and healthy persons. To give a general survey the average of each parameters left and right value was calculated.
Survey of the pain related parameters, average of both splenii and average of both trapezii muscles. Whiplash patients (WP, ) in comparison to the control group (CG, □). Logarithmically transformed mean values of the pain threshold (pt; in seconds), pain maximum (m; in VAS units), tolerance threshold (tt, in seconds), slope (s; VAS/s) and integral (i; VAS ∗ seconds ∗ 103) of the pain intensity-time function. PU, particular unit; l, left; r, right. ∗P < 0.05, ns, not significant; post hoc t-tests, corrected for multiple comparisons (11).
As the deviation in increased muscle pain in whiplash patients appeared to be more pronounced in the left than in the right investigated muscles, the question was raised whether the differences in pressure painfulness within the groups with regard to the side of the strained neck/shoulder muscles are statistically relevant. Some of the pain related parameters revealed that the patients experienced significantly more pressure pain in the left than in the right trapezius muscle. This holds true for the integral under the pain-intensity function (P = 0.028; Figure 4). Slope and maximal muscle pain intensity showed a tendency towards statistically relevant differences (P = 0.054; P = 0.055). However, no side differences could be seen for the pressure painfulness of the paravertebral muscles. There was only a tendency for the integral of the left splenius pain intensity-time function to differ significantly from the right one (P = 0.058). In the control group the side of pressure pain induction had no effect on the results.
Mean values of the area under the pain intensity-time curve, intragroup comparison. Within the patient group (WP) the left trapezius muscle was significantly more pressure sensitive than the right one. Such a side difference could not be observed for the control group (CG). The side difference is ascribed to the left sided seat belt as all whiplash patients were left seated drivers. ∗P < 0.05, ns, not significant; paired comparisons (t-tests).
Discussion
In whiplash injury neck and shoulder muscles are usually exposed to an intense strain because of a rapid flexion-extension movement of the head and neck due to the rear-end impact (8, 12). The present study investigated the quantity of muscle pressure pain after such a muscle strain.
Applying the PC-aided pressure algesimetry we could demonstrate an increased pressure painfulness of the left and right splenius and of the left and right trapezius muscles for almost all pain related parameters examined in acute whiplash injury. The shoulder muscles (trapezius) were more painful on the left than on the right side which can be ascribed to the seat-belt position. As all patients examined were motorcar drivers it is reasonable to assume that the higher pressure painfulness of the left trapezius is due to the wearing of seat belts. In contrast, the paravertebral splenius muscles showed an equally increased muscle painfulness on both sides.
According to Mense (13, 14) we assume that a possible explanation of increased pressure painfulness in acute whiplash injury may be a muscle nociceptor sensitization and an enhanced responsiveness to painful stimuli as a consequence of binding endogenous algesic ligands to the receptors of nociceptive free nerve endings. Even though in most cases of whiplash injury no structural damage can be detected (15) micro lesions beyond the maximal resolution of present imaging methods of the strained muscles in whiplash patients cannot be excluded. Algesic agents originating from damaged tissue are bradykinin (BK), prostaglandine, 5-hydroxytryptamin and adenosine triphosphate (ATP). ATP, which is present in muscle tissue, is released after strain induced muscular sarcomeric disrupture and plays a fundamental role in generation of increased muscle pain and in sensitization of muscle nociceptors (13, 16), even in whiplash injury (17). In case of painful stimulation further algesic agents such as substance P, calcitonin-related peptide or somatostatin are released from vesicles in receptive endings. All these algesic ligands, that in part interact, increase the excitability of the nociceptive endings.
Besides peripheral, central components might also contribute to the present findings. The permanent nociceptive input of the sensitized nociceptors is able to induce central sensitization. Central neuron populations involved in pain processing might become more excitable and a larger number of neurons might receive afferent input from the hurt muscle (13, 18, 19) thus increasing and expanding pain perception. A possible mechanism for central sensitization is described by Mense and Hoheisel (20). Furthermore a transient dysfunction of the descending pain control system is suggested to influence pain perception (7, 21). Keidel et al. (22) found abnormalities in the brainstem mediated antinociceptive inhibitory reflexes of the temporalis muscle with a reduced duration of the late exteroceptive suppression period in patients with acute headache following whiplash injury.
To sum up, processes of early sensitization may occur peripherally as well as centrally. Sensitization of muscle nociceptors is the neurophysiological basis of local abnormal painfulness and of an enhanced reaction to painful stimulation. The computer-aided pressure algesimetry is an adequate instrument to study the altered pain perception of the strained musculature in whiplash injury. Processes of sensitization have been demonstrated to occur within two hours in rats by Sperry & Goshgarian (23). Persistent neuroplastic changes as described by Mense (24) and Bendtsen (25, 26) may provide the basis for the transition from acute pain into longer lasting complaints which clinically are seen in 10–20% of whiplash patients (7).
The fact that we did not find a consistent diminution of whiplash patients’ pain threshold with our method in contrast to the plausible alterations of the further pain related parameters is likely due to a methodological constraint of our experimental proceeding. We did not start with a pressure intensity of zero and did not increase it until every subjects’ individual pain threshold was detected. Instead of this, our purpose was to apply a constant pressure of 1310 kPa right from the beginning of our recordings. The application of a defined and constant pressure intensity in each subject narrowed the reliable determination of individual pain thresholds and led to high variability in pain detection. Both, whiplash patients (as expected) and some control subjects have experienced the applied pressure intensity immediately as painful, so that no statistical differences between the groups could be observed.
Symptomatology after whiplash injury can be susceptible to examiner-dependent questioning and investigation techniques – especially with regard to pain examination. Therefore a need to develop standardized diagnostic tools exists. Generally, muscle pain is evaluated by palpation, but little is known about the validity of this commonly used method for examining muscle pain after whiplash injury (15, 27, 28). In the present study we could quantify and standardize increased muscle painfulness in traumatically strained muscles with the recording technique described on the basis of numerical analysis. The computerized interactive pressure algesimetry is independent of varying examination techniques and assessment qualities of different investigators. The pressure application rate and duration are controllable and objective. These advantages do not completely apply to the pressure controlled palpation suggested by Bendtsen et al. (3, 4), which still remains dependent on individual investigation criteria.
In contrast to the digital pressure dolorimetry described by Bendtsen et al. (3) or the pressure algesimetry described by Brennum et al. (29), the applied pressure pain in the present study was kept constant to obtain the time course of the induced pain. The calculated parameters (pain threshold, pain maximum, tolerance threshold, slope, and integral of the pain intensity-time function) were chosen to reflect this time course. However, many studies analysed only the pressure pain threshold as a measure of altered pain sensitivity after whiplash injury (29–33). None of them acquired the whole dynamic range of pain perception – from just noticeable perception up to intolerability regarding the time course of increase as a measure of pain processing. Above all the pressure pain threshold has proven to be a rather unstable measuring instrument because of its high degree of variability (34). Examiners prior knowledge as well as measurement order could bias pressure pain thresholds in patients with facial and temporal area pain (34).
Another approach to investigate the increased muscle pain is the evaluation of muscle tension by surface electromyographic (EMG) recordings. However, since EMG recordings also register various artefacts (2), a wide range within patient and control group samples is described (23, 35), and stable EMG recordings are found in increasing pain (35) the relationship between muscle pain and increased muscle tension is possible but not yet proven. Carlson et al. (36) did not find any difference in pain and nonpain subjects’ average EMG activity at all suggesting that muscle pain is not accompanied by muscle hyperactivity as expressed by greater EMG activity. For the given reasons the frequently used EMG techniques fail to represent a valid and easy to handle diagnostic instrument measuring muscle pain. In contrast, sufficient mechanical stimulation of muscle tissue, which can be achieved by pressure application, has proven to excite nociceptors causing muscle pain (16). Thus it can be concluded that the pressure algesimetry is a valid pain measuring instrument.
By using the computer-aided pressure algesimetry we have been able to measure muscle pressure painfulness following whiplash injury in a standardized manner. We could quantify the increased painfulness of the neck/shoulder muscles in whiplash injury which depends on assessment qualities and interpretations of the investigator in the standard clinical examination. Applying this method to further prospective controlled investigations, it may be useful to better standardize pain levels in outcome measures.