Pain, 51 (1992) 169-173 0 1992 Elsevier Science
169 Publishers
B.V. All rights reserved
0304-3959/92/$05.00
PAIN 02135
Cervicogenic headache, migraine, and tension-type headache. Pressure-pain threshold measurements Gunnar Bovim Department of Neurology, Trondheim University Hospitals, Regionsykehuset, 7006 Trondheim (Norway) (Received
7 January
1992, revision
received
1 May 1992, accepted
28 May 1992)
Summary
Pressure-pain threshold (PPT) measurements were performed with a pressure algometer, at 22 specified points in the head in patients with cervicogenic headache (n = 32), migraine (with and without aura) (n = 26) and tension-type headache (n = 17). Comparisons were made with a group of healthy controls (n = 20). The average PPT differed significantly between the groups (ANOVA, F = 9.5, P < O.OOOS),largely caused by the low threshold in cervicogenic headache patients. There were no significant differences between controls and the 2 other headache groups. In the cervicogenic headache group, the lowest PPT was found in the occipital part of the head on the side with pain predominance. The ratio between the dominant and non-dominant sides (all 11 points on each side) was 0.85 in cervicogenic headache, whereas it was 0.99 in migraine patients with side preponderance of the pain. The present results support the view that the pathogenesis of cervicogenic headache differs from that of migraine and tension-type headache. The results may further support the theory that fibres from the C2 level (innervating the occipital part of the head) may be included in the pathogenetic mechanism in cervicogenic headache. Key words: Cervicogenic
headache; Migraine; Tension-type
Introduction
Diagnostic criteria for cervicogenic headache are based on the history and findings at clinical examination (Sjaastad et al. 1990). On the basis of these criteria, cervicogenic headache may probably be separated from other headache categories with a reasonable degree of certainty (Headache Classification Committee 1988). The most difficult diagnostic task is probably to differentiate cervicogenic headache from migraine without aura (common migraine). Tension-type headache (tension headache) may also represent a differential diagnostic problem. Pressure-pain threshold (PPT) measurement has been used as a diagnostic procedure in different pain syndromes (Fisher 19861, and the method has also been used to evaluate headache patients (Jensen et al. 1988;
Correspondence to: Gunnar Bovim, Department of Neurology, Trondheim University Hospitals, Regionsykehuset, 7006 Trondheim, Norway. Tel.: (47-7) 998-420; Fax: (47-7) 997-581.
headache; Pressure-pain
threshold
Langemark et al. 1989). This method has not previously been used in cervicogenic headache. PPT measurements may possibly give information about pathophysiologic mechanisms. If PPT were different in cervicogenic headache, migraine without aura, and tension-type headache, it might point to different pathogenetic mechanisms in the 3 headache categories. In doubtful cases, PPT measurements might be of diagnostic value in differentiating clinical pictures.
Materials
and methods
Patients Thirty-two patients with cervicogenic headache, 26 with migraine (8 with and 18 without aura) and 17 with tension-type headache were included (Table I). As controls, 20 persons without headache or other pain problems were included. The diagnosis of cetvicogenic headache was established according to the present criteria (Sjaastad et al. 1990). A brief summary of these criteria has been published in PAIN recently (Bovim et al. 1992). All of the cervicogenic headache patients had, according to the criteria, unilateral pain without side shift. The diagnosis of migraine and tension-type headache followed
I 70 the IHS criteria (Headache Classification Committee, 1988). All the migraine patients had either unilateral pain with side shift (i.e., unilateral attacks sometimes on the left and sometimes on the right side) or bilateral pain. Sixteen of the migraine patients could, nevertheless. point out 1 of the sides as dominant. All the patients with tension-type headache had bilateral, global pain, generally with a maximum in the frontal part of the head. Fourteen patients had chronic and 3 had episodic tension-type headache. Pericranial muscle tenderness seemed increased in at least I3 of these patients. At the time of investigation, the patients were without analgetic medication for the last 24 h. Most of them had no or only weak headaches (less than 30% of their maximal). Two patients with tension-type headache had more than 50% pain.
Algometer PPT measurement was performed with a pressure afgometer (Model PTH-AF2, Pain Threshold Meter, commercially available from Pain Diagnostic and Thermography Corporation). This apparatus consists of a force gauge connected with a hard rubber tip, I cm in diameter. The rubber disc was applied to the skin. When the pressure increased, the force was transmitted to the body of the algometer. where an indicator moved in a clockwise direction. It is calibrated in kg/cm* and the pressure range is O-11 kg/cm’. The algometer was calibrated at the Technical College of Norway before and in the middle of the series and found to be in perfect order.
Procedure The patients were examined sitting on a chair. Twenty-two specified points (11 on each side of the head) (Fig. 1) were marked and examined. The person studied was informed to give a sign immediately when perceiving the gradually increasing pressure as painful. Since the study especially focused on localized differences in PPT, the persons examined were carefully informed to always give the sign at the same degree of unpleasantness/pain. The examinations were carried out starting at point 1 fright and left) and ending up at point 11(Fig. I), and the procedure was made
TABLE
1
AGE AND GORIES
SEX DISTRIBUTION
For the headache groups, has also been tabulated.
Cervicogenic headache (n = 32) Migraine without aura (n = 18) Migraine with aura (n = 8) Tension-type headache (n = 17) Healthy controls (n = 20)
IN THE
the pain intensity
DIFFERENT
at time of investigation
Age
Percent
Headache
Mean 6.D.)
women
Mean 6.D.)
43 (11)
91
21 (24)
33
(8)
72
5 (14)
39
(l(J)
xx
0 (0)
37 (10)
47 **
42(15)
65
CATE-
intensity
*
17 (21)
* In percent of maximal headache. ** The frequency of women in this group is lower than in the tension-type headache population in general, due to the selection of patients living nearby and being available for studies.
Fig. I. Schematic drawing of the I I points on the left side. The locations of the points were: anterior (I), intermediate (2 and 3) and posterior (4) part of the temporal muscle; 3 cm above the superior margin of the ear (5); over the supraorbital nerve 2 cm above the supraorbital incisure 16): 2 cm from the midline at the vertex (71, the mastoid process (8). the superior insertions (9) of the sternocleidomastoid muscle; and trapezius muscle (10) over the greater occipital nerve, 2 cm inferior to the external occipital protuberance and 2 cm lateral to the midline (I I). Contralaterally, ‘mirror’ points were tested.
a total of 3 times. The average been used in the calculations.
of the 3 values
for each
point
has
Statistical analyses Average PPTs for the whole head (A), the forehead (AF) and the occipital region (AO) were computed for all patients and controls. The ratio AO/AF was used for a between-group comparison of the relationship between bilateral occipital and frontal pain thresholds. In the cervicogenic headache patients and in the I6 migraine patients with a side predominance, calculations were made with regards to the dominant hemicranium, and the values were compared to those on the other side. The values for the forehead and occipital areas on the dominant side were separately compared to the contralateral measurements. In order to investigate whether the PPT differed between headache groups, analysis of variance (ANOVA) was carried out. Exploratory data analysis (post hoc) was carried out with 2-sided Student’s I tests. A P value of 0.0s was regarded as significant. The correlations between PPTs, age. sex and pain level were computed (Pearson r). Pain-level was given as percent of the maximal pain.
Results Statistics concerning age and sex are displayed in Table I, together with pain level at the time of investigation. Data on migraine patients with and without aura (‘classic’ and ‘common’ migraine) were pooled because statistical differences were not observed between them. No systematic relationship was observed between pain level and PPT in cervicogenic headache fr = 0.06). An apparently significant correlation in tension-type
171
* 3t’liI”;
I
I
2'
C
co
M
I
04
T
Fig. 2. Average PPT of the whole head. Comparison between the different groups: C = cervicogenic headache; M = migraine, T = tension-type headache, Co = controls. Fifty percent of the sample is collected within the box, and the median is indicated. Twenty-five percent of the sample is located on each side of the box (bars and *). Outliers are represented by *.
headache (r = 0.61, P = 0.01) disappeared when the 2 patients with pain levels above 50% were removed from the analysis (r = 0.10, P = 0.72). Sex differences were not found. Average PPT seemed to increase with age, but the correlation was significant only in the migraine group (r = 0.42, P = 0.04). Differences within the tension-type headache group (episodic vs. chronic, with vs. without pericranial muscle tenderness) were not found, but the numbers in some of the subgroups were small. The average PPT of the whole head differed significantly between the 4 groups (ANOVA, F = 9.5, df = 3, P < O.OOOS),largely caused by the low threshold in the cervicogenic headache group (Fig. 2). As for migraine, tension-type headache, and control groups there were no reciprocal statistical differences (Table II). TABLE
’
I
I
C
co
THRESHOLDS
IN CERVICOGENIC
Average PPTs for the whole head, the occipital/ and analysed.
I
Fig. 3. Ratio between occipital and frontal parts of the head. The ratio is: average PPT of occipital points 4, 8, 9, 10, 11, divided by average PPT of the frontal points 1, 2, 3, 5, 6, 7 (Fig. 1). Comparison between the different groups: C = cervicogenic headache: M = migraine; T = tension-type headache; Co = controls. Fifty percent of the sample is collected within the box, and the median is indicated. Twenty-five percent of the sample is located on each side of the box (bars and *). Outliers are represented by *.
The ratio between occipital and forehead PPT (both sides combined) differed between the 4 groups (ANOVA, F = 6.8, df = 3, P < 0.0005). Both sides were included to make comparisons with tension-type headache possible. The cervicogenic headache patients had a lower ratio than the others (Table II). The overlap was, however, considerable (Fig. 3). In the cervicogenic headache group the mean occipital/ frontal ratio on the dominant side was 0.86, whereas it was 0.90 on the non-dominant side (P = 0.061, i.e., lowest on the dominant side. In the migraine patients with side predominance, the occipital/frontal ratio on the dominant side was 0.99 vs. 0.91 on the non-domi-
Whole head
frontal
HEADACHE
VS. THE OTHER
ratio and the dominant/
Occipital/frontal
non-dominant
GROUPS
ratio
(n = 17)
(n = 20)
* P < 0.01; ** P
169 Publishers
B.V. All rights reserved
0304-3959/92/$05.00
PAIN 02135
Cervicogenic headache, migraine, and tension-type headache. Pressure-pain threshold measurements Gunnar Bovim Department of Neurology, Trondheim University Hospitals, Regionsykehuset, 7006 Trondheim (Norway) (Received
7 January
1992, revision
received
1 May 1992, accepted
28 May 1992)
Summary
Pressure-pain threshold (PPT) measurements were performed with a pressure algometer, at 22 specified points in the head in patients with cervicogenic headache (n = 32), migraine (with and without aura) (n = 26) and tension-type headache (n = 17). Comparisons were made with a group of healthy controls (n = 20). The average PPT differed significantly between the groups (ANOVA, F = 9.5, P < O.OOOS),largely caused by the low threshold in cervicogenic headache patients. There were no significant differences between controls and the 2 other headache groups. In the cervicogenic headache group, the lowest PPT was found in the occipital part of the head on the side with pain predominance. The ratio between the dominant and non-dominant sides (all 11 points on each side) was 0.85 in cervicogenic headache, whereas it was 0.99 in migraine patients with side preponderance of the pain. The present results support the view that the pathogenesis of cervicogenic headache differs from that of migraine and tension-type headache. The results may further support the theory that fibres from the C2 level (innervating the occipital part of the head) may be included in the pathogenetic mechanism in cervicogenic headache. Key words: Cervicogenic
headache; Migraine; Tension-type
Introduction
Diagnostic criteria for cervicogenic headache are based on the history and findings at clinical examination (Sjaastad et al. 1990). On the basis of these criteria, cervicogenic headache may probably be separated from other headache categories with a reasonable degree of certainty (Headache Classification Committee 1988). The most difficult diagnostic task is probably to differentiate cervicogenic headache from migraine without aura (common migraine). Tension-type headache (tension headache) may also represent a differential diagnostic problem. Pressure-pain threshold (PPT) measurement has been used as a diagnostic procedure in different pain syndromes (Fisher 19861, and the method has also been used to evaluate headache patients (Jensen et al. 1988;
Correspondence to: Gunnar Bovim, Department of Neurology, Trondheim University Hospitals, Regionsykehuset, 7006 Trondheim, Norway. Tel.: (47-7) 998-420; Fax: (47-7) 997-581.
headache; Pressure-pain
threshold
Langemark et al. 1989). This method has not previously been used in cervicogenic headache. PPT measurements may possibly give information about pathophysiologic mechanisms. If PPT were different in cervicogenic headache, migraine without aura, and tension-type headache, it might point to different pathogenetic mechanisms in the 3 headache categories. In doubtful cases, PPT measurements might be of diagnostic value in differentiating clinical pictures.
Materials
and methods
Patients Thirty-two patients with cervicogenic headache, 26 with migraine (8 with and 18 without aura) and 17 with tension-type headache were included (Table I). As controls, 20 persons without headache or other pain problems were included. The diagnosis of cetvicogenic headache was established according to the present criteria (Sjaastad et al. 1990). A brief summary of these criteria has been published in PAIN recently (Bovim et al. 1992). All of the cervicogenic headache patients had, according to the criteria, unilateral pain without side shift. The diagnosis of migraine and tension-type headache followed
I 70 the IHS criteria (Headache Classification Committee, 1988). All the migraine patients had either unilateral pain with side shift (i.e., unilateral attacks sometimes on the left and sometimes on the right side) or bilateral pain. Sixteen of the migraine patients could, nevertheless. point out 1 of the sides as dominant. All the patients with tension-type headache had bilateral, global pain, generally with a maximum in the frontal part of the head. Fourteen patients had chronic and 3 had episodic tension-type headache. Pericranial muscle tenderness seemed increased in at least I3 of these patients. At the time of investigation, the patients were without analgetic medication for the last 24 h. Most of them had no or only weak headaches (less than 30% of their maximal). Two patients with tension-type headache had more than 50% pain.
Algometer PPT measurement was performed with a pressure afgometer (Model PTH-AF2, Pain Threshold Meter, commercially available from Pain Diagnostic and Thermography Corporation). This apparatus consists of a force gauge connected with a hard rubber tip, I cm in diameter. The rubber disc was applied to the skin. When the pressure increased, the force was transmitted to the body of the algometer. where an indicator moved in a clockwise direction. It is calibrated in kg/cm* and the pressure range is O-11 kg/cm’. The algometer was calibrated at the Technical College of Norway before and in the middle of the series and found to be in perfect order.
Procedure The patients were examined sitting on a chair. Twenty-two specified points (11 on each side of the head) (Fig. 1) were marked and examined. The person studied was informed to give a sign immediately when perceiving the gradually increasing pressure as painful. Since the study especially focused on localized differences in PPT, the persons examined were carefully informed to always give the sign at the same degree of unpleasantness/pain. The examinations were carried out starting at point 1 fright and left) and ending up at point 11(Fig. I), and the procedure was made
TABLE
1
AGE AND GORIES
SEX DISTRIBUTION
For the headache groups, has also been tabulated.
Cervicogenic headache (n = 32) Migraine without aura (n = 18) Migraine with aura (n = 8) Tension-type headache (n = 17) Healthy controls (n = 20)
IN THE
the pain intensity
DIFFERENT
at time of investigation
Age
Percent
Headache
Mean 6.D.)
women
Mean 6.D.)
43 (11)
91
21 (24)
33
(8)
72
5 (14)
39
(l(J)
xx
0 (0)
37 (10)
47 **
42(15)
65
CATE-
intensity
*
17 (21)
* In percent of maximal headache. ** The frequency of women in this group is lower than in the tension-type headache population in general, due to the selection of patients living nearby and being available for studies.
Fig. I. Schematic drawing of the I I points on the left side. The locations of the points were: anterior (I), intermediate (2 and 3) and posterior (4) part of the temporal muscle; 3 cm above the superior margin of the ear (5); over the supraorbital nerve 2 cm above the supraorbital incisure 16): 2 cm from the midline at the vertex (71, the mastoid process (8). the superior insertions (9) of the sternocleidomastoid muscle; and trapezius muscle (10) over the greater occipital nerve, 2 cm inferior to the external occipital protuberance and 2 cm lateral to the midline (I I). Contralaterally, ‘mirror’ points were tested.
a total of 3 times. The average been used in the calculations.
of the 3 values
for each
point
has
Statistical analyses Average PPTs for the whole head (A), the forehead (AF) and the occipital region (AO) were computed for all patients and controls. The ratio AO/AF was used for a between-group comparison of the relationship between bilateral occipital and frontal pain thresholds. In the cervicogenic headache patients and in the I6 migraine patients with a side predominance, calculations were made with regards to the dominant hemicranium, and the values were compared to those on the other side. The values for the forehead and occipital areas on the dominant side were separately compared to the contralateral measurements. In order to investigate whether the PPT differed between headache groups, analysis of variance (ANOVA) was carried out. Exploratory data analysis (post hoc) was carried out with 2-sided Student’s I tests. A P value of 0.0s was regarded as significant. The correlations between PPTs, age. sex and pain level were computed (Pearson r). Pain-level was given as percent of the maximal pain.
Results Statistics concerning age and sex are displayed in Table I, together with pain level at the time of investigation. Data on migraine patients with and without aura (‘classic’ and ‘common’ migraine) were pooled because statistical differences were not observed between them. No systematic relationship was observed between pain level and PPT in cervicogenic headache fr = 0.06). An apparently significant correlation in tension-type
171
* 3t’liI”;
I
I
2'
C
co
M
I
04
T
Fig. 2. Average PPT of the whole head. Comparison between the different groups: C = cervicogenic headache; M = migraine, T = tension-type headache, Co = controls. Fifty percent of the sample is collected within the box, and the median is indicated. Twenty-five percent of the sample is located on each side of the box (bars and *). Outliers are represented by *.
headache (r = 0.61, P = 0.01) disappeared when the 2 patients with pain levels above 50% were removed from the analysis (r = 0.10, P = 0.72). Sex differences were not found. Average PPT seemed to increase with age, but the correlation was significant only in the migraine group (r = 0.42, P = 0.04). Differences within the tension-type headache group (episodic vs. chronic, with vs. without pericranial muscle tenderness) were not found, but the numbers in some of the subgroups were small. The average PPT of the whole head differed significantly between the 4 groups (ANOVA, F = 9.5, df = 3, P < O.OOOS),largely caused by the low threshold in the cervicogenic headache group (Fig. 2). As for migraine, tension-type headache, and control groups there were no reciprocal statistical differences (Table II). TABLE
’
I
I
C
co
THRESHOLDS
IN CERVICOGENIC
Average PPTs for the whole head, the occipital/ and analysed.
I
Fig. 3. Ratio between occipital and frontal parts of the head. The ratio is: average PPT of occipital points 4, 8, 9, 10, 11, divided by average PPT of the frontal points 1, 2, 3, 5, 6, 7 (Fig. 1). Comparison between the different groups: C = cervicogenic headache: M = migraine; T = tension-type headache; Co = controls. Fifty percent of the sample is collected within the box, and the median is indicated. Twenty-five percent of the sample is located on each side of the box (bars and *). Outliers are represented by *.
The ratio between occipital and forehead PPT (both sides combined) differed between the 4 groups (ANOVA, F = 6.8, df = 3, P < 0.0005). Both sides were included to make comparisons with tension-type headache possible. The cervicogenic headache patients had a lower ratio than the others (Table II). The overlap was, however, considerable (Fig. 3). In the cervicogenic headache group the mean occipital/ frontal ratio on the dominant side was 0.86, whereas it was 0.90 on the non-dominant side (P = 0.061, i.e., lowest on the dominant side. In the migraine patients with side predominance, the occipital/frontal ratio on the dominant side was 0.99 vs. 0.91 on the non-domi-
Whole head
frontal
HEADACHE
VS. THE OTHER
ratio and the dominant/
Occipital/frontal
non-dominant
GROUPS
ratio
(n = 17)
(n = 20)
* P < 0.01; ** P
