Fluctuations in local and widespread mechanical sensitivity throughout the migraine cycle: a prospective longitudinal study

A prospective longitudinal observational study within and between people with migraine and healthy participants. The study was approved by the Medical Ethical Committee of the VU Medical Centre, Amsterdam, The Netherlands (METC-2015-551) and performed according to the Declaration of Helsinki 2013. The study is reported according to the STROBE guidelines [21].

People with migraine were recruited from primary healthcare centres between April 2017 and April 2018. During the same period, healthy participants were recruited from the general public via advertisements in local newspapers and social media. Eligibility criteria for people with migraine were: a medical diagnosis of migraine (i.e., migraine diagnosed by a general practitioner according to the International Classification of Headache Disorders (ICDH- III) [22], aged between 18 and 65 years and Dutch or English speaking. Exclusion criteria were: other types of headaches such as medication overuse headache, head or neck complaints within 2 months prior to the measurements, musculoskeletal painful conditions, psychiatric conditions, malignancy or other neuropathic pain states. Furthermore, participants who received treatment for headache 48 h before the measurements or those who received botulinum toxin injections were also excluded. Participants were not allowed to use analgesic medication within 24 h before the measurements. Healthy participants were people without current pain or a history of chronic pain, neurological disorders or headache in the last 12 months. Healthy participants were matched for age and sex. All participants provided written informed consent prior to participating in the study.

Baseline measurements were performed in the interictal phase and consisted of completion of various questionnaires and the assessment of mechanical sensitivity via determination of pressure pain thresholds (PPTs). All measurements were conducted at a primary healthcare centre. A questionnaire was completed to obtain sociodemographic information, medication use and migraine characteristics. The six-item Headache Impact Test (HIT-6) [23] was used to measure the impact of headache on daily functioning. Scores on the HIT-6 range from 36 to 78. The HIT-6 is a reliable (ICC = 0.77) and valid tool to discriminate headache impact in a migraine population [23]. The Central Sensitization Inventory was used to gain insight into the degree of central sensitization [24]. The inventory has good test-retest reliability in people with chronic pain (ICC = 0.88) and healthy participants (ICC = 0.91) [24]. It consists of 25 items, and the score ranges from 0 to 100. The Allodynia Symptom Checklist (ASC-12) is a 12-item validated questionnaire to identify cutaneous allodynia with a score ranging from 0 to 12 [25]. Additionally, all participants rated their headache pain intensity before each measurement on a numeric pain rating scale (NPRS). The scale ranges from 0 to 10, and is valid and reliable in several patient populations [26].

PPTs were measured bilaterally at five test locations in a fixed, cyclic order with a 20 s interval between the measurements: 1) temporalis muscle (i.e., 1 cm lateral to the external angle of the orbit) [27], 2) paraspinal muscles C1 (i.e., 2 cm lateral of the midline of the neck, below the occipital bone), 3) upper trapezius muscle (i.e., midpoint between acromion and spinous process C7) [16, 28], 4) extensor carpi radialis muscle (i.e, at 1/3 of the length of the underarm, distal to the elbow) and 5) tibialis anterior muscle (i.e., at 1/3 of the length of the lower leg, distal to the knee). The first three test locations are cephalic regions and the latter two are extra-cephalic regions. In totally, PPTs were determined three times at the same location; first on the dominant side of the migraine (defined as the most painful side of the current attack) and subsequently on the non-dominant side. In case of bilateral migraine, the most prevailing side of the headache was measured first. The patients’ dominant side was compared with the side of the dominant hand in healthy participants (and the patients’ non-dominant side with the side of the non-dominant hand in healthy participants). Participants were instructed to press a switch when they first felt the sensation of pressure change into pain. PPTs were measured with a digital algometer (Algometer type II, rubber probe area 1cm², Somedic Electronics, Solna, Sweden) with an application rate of 50 kPa per second. The algometer is a valid and reliable (ICC: 0.75–0.95) instrument to measure PPTs [29, 30].

PPTs were assessed during the four phases of the migraine cycle. The timing of the sessions was individualized. Patients contacted the research centre when their headache started and were measured within several hours for the preictal phase measurement. The ictal measurement in which they experienced the throbbing headache was routinely scheduled 1 day later and the postictal assessment was individually planned one till 3 days after the ictal measurement depending on individual experiences of ictal phase duration. Healthy participants were measured accordingly with equal time intervals between the measurements as the matched person with migraine.

A sample size calculation was performed based on the effect size and standard deviation (SD) of a study that compared PPTs at the trapezius muscle between people with migraine and healthy participants [31]. Based on a mean difference between groups of 56 kPa, a SD of 68, two-tailed α of 0.05, β of 0.80, 3 follow-up measurements, intra-person correlation coefficient of 0.6 and an anticipated dropout rate of 10%, 19 participants per group were required.

The mean from the three PPT measures at the same location was calculated for the dominant and non-dominant sides. A PPT value more than three standard deviations from the mean of the three measures, was considered as an outlier and was removed. In those cases, the mean was calculated based on the remaining two PPT measures. Participants who did not attend the four measurement sessions were excluded from the statistical analysis. Normality of continuous variables was visually inspected by Q-Q plots, box plots and histograms and checked by the Kolmogorov-Smirnov test. Characteristics were compared both visually and statistically with an independent sample t tests or Mann-Whitney U tests for continuous variables and Chi square tests for dichotomous variables. Linear mixed model analyses with fixed factor (time), covariate (group) and interaction (time*group) were used to detect differences between the groups at the four time points in the cephalic and extra-cephalic region. A linear mixed model analysis with fixed factor (time) was performed to detect differences within the migraine group between the four phases. A random intercept was chosen to account for the correlated nature of multiple measurements from the same individual. The regression coefficient (B), p-value, and confidence intervals (95%CI) were computed for the crude models, as well as for the models that were adjusted for age and baseline PPT values (i.e., PPT values in the interictal phase). A p-value of < 0.05 was considered to be statistically significant. Individual line graphs were created per participant to visualize individual fluctuations throughout the migraine cycle and visualize individual differences in PPTs. SPSS version 25.0 (IBM Corp., Armonk, New York, USA) was used for statistical analysis.

Fifty-one people with migraine were assessed for eligibility criteria of whom 29 commenced the study. Four patients had to be excluded during the study as they used analgesic medication and six patients did not attend one or more follow-up sessions. Nineteen people completed all four measurement sessions. There were no significant differences in participant characteristics between people with migraine who completed all sessions and those who were excluded during the study (p > 0.136). Figure 1 provides the flow diagram of the study, and Table 1 summarizes the participant characteristics at baseline.

PPTs in people with migraine were significantly lower in the preictal, ictal and postictal phase compared to the interictal phase in both the cephalic and extra-cephalic regions (Table 2). The strongest decrease in PPTs was found in the ictal phase compared to the interictal phase in the cephalic and extra-cephalic regions. The results at the non-dominant side were comparable with the dominant side in the cephalic regions with significantly lower PPTs in the preictal, ictal and postictal phase compared to the interictal phase (Appendix 1). For the extra-cephalic regions (non-dominant side), significantly lower PPTs were found in the ictal phase compared to the interictal phase, postictal phase (extensor carpi radialis muscle) and preictal phase (tibialis anterior muscle), though not in the preictal phase on the extensor carpi radialis muscle and the postictal phase on the tibialis anterior muscle (Appendix 1).

People with migraine showed significantly lower PPTs than healthy participants throughout the four phases of the migraine cycle at all five test locations at both the dominant side (Table 3) and non-dominant side (Appendix 2). There were no outliers.

These findings were observed in both the crude and the adjusted analyses, except for the extensor carpi radialis muscle in the preictal phase (non-dominant side).

To visualize individual fluctuations in PPT patterns throughout the migraine cycle, line graphs per participant were created in the cephalic and extra-cephalic regions in people with migraine and matched healthy participants at the dominant side (Fig. 2). Similar patterns with decreasing PPTs from the interictal to the preictal and subsequent to the ictal phase were observed in people with migraine and were most pronounced in the cephalic region at the dominant side. But large individual differences in magnitude and fluctuations in mechanical sensitivity were observed. No cyclic changes were visually and statistically detected in the healthy participants (p > 0.289).