Investigation of Sensory and Neuropsychological Parameters in Migraine Sufferers: A Cross-Sectional Study with Negative Findings.

Why carry out this study? Migraine is a common and disabling disorder lacking reliable biomarkers, highlighting the need for tools like quantitative sensory testing (QST) to better understand its pathophysiology This study hypothesizes that high-frequency episodic and chronic migraines are linked to changes in sensory functions and aims to evaluate this through QST, psychometric assessments and conditioned pain modulation (CPM) What was learned from the study? Our study shows that individuals with high-frequency or chronic migraine did not differ from healthy controls in sensory, psychosocial or pain inhibition measures These negative findings suggest that QST and CPM may have limited sensitivity to detect migraine-related alterations, despite prior evidence indicating abnormal nociceptive processing Larger, longitudinal and more selective studies are needed to identify vulnerable phenotypes or migraine subgroups with detectable sensory changes

From January 2 to August 30, 2022, individuals diagnosed by experienced neurologists specialized in the field of headache, with ≥ 10 headache days per month, were examined. Migraine diagnoses were made according to the current classification of the International Headache Society (International Classification of Headache Diseases-3rd edition [ICHD-3]) [ 28 ]. The number of headache days was taken from the Migraine Disability Assessment (MIDAS), a self-reporting questionnaire (see below). For CM (≥ 15 headache days with migraine features on ≥ 8 days per month), this is the more appropriate parameter compared to the number of migraine days. For HFEM, we applied the definition used by Lipton et al. (2014) with 10–14 headache days per month. However, classification based on frequency is challenging because headache frequency tends to vary over time. Furthermore, epidemiological data indicate that individuals with HFEM share many characteristics with those who have CM [ 29 ]. Therefore, these two groups were combined to facilitate recruitment. Both male and female individuals aged 18–65 years were included as this demographic represents the primarily affected working-age population [ 30 ]. We examined migraine sufferers experiencing ≥ 10 headache days per month, as sensitization was deemed most likely in this group. However, in these individuals, the current migraine phase cannot be reliably determined. Given that alterations in hypothalamic activity and sensory thresholds can occur up to 48 h before headache onset, pain-free individuals with HFEM or CM are more likely to be in the preictal or postictal phase [ 11 , 31 ]. During the mentioned period, a control group (CG) without headaches was tested using the same paradigm excluding three questionnaires. Participants were not required to be completely free of pain at the time of assessment. However, tests were conducted only if ≥ 24 h had passed since their most recent severe migraine attack as many migraineurs often report signs of tension and altered mechanical sensitivity in musculoskeletal regions like neck, temporomandibular joint, or pericranial structures [ 32 – 35 ]. The Graded Chronic Pain Scale (GCPS) was used to evaluate pain intensity. Details about participants’ medication intake were gathered via the demographic questionnaire (see Table  2 —confounders). Individuals were excluded from the study if they had severe psychiatric disorders, including schizophrenia, bipolar disorder, borderline personality disorder, active substance abuse, or dementia. Since depression and anxiety disorders frequently co-occur with chronic headaches, these conditions were not considered grounds for exclusion. Furthermore, anyone with a prior history of a major neurological event or trauma, including significant injuries to the brain or neck, was not eligible to participate. Since individuals with CM often present with additional chronic primary or secondary pain conditions such as endometriosis or low back pain [ 36 – 38 ], it was not possible to entirely exclude comorbid pain diagnoses. The migraine group (MG) was primarily recruited from the headache outpatient clinic at the University Clinic for Neurology, University Hospital Salzburg, Austria, with additional participants sourced from a neurology practice in Graz. The healthy control group was recruited through the University of Applied Sciences Graz email distribution list, targeting co-workers and students. All participants were initially informed about the study background and testing procedures via telephone and the informed consent letter. After that, further information including questionnaires and data sheets was sent by email. Participation was entirely voluntary, and individuals could withdraw from the study at any time without providing a reason. Prior to their participation in the study, written informed consent was obtained from all participants. After reviewing the requested information, the responsible ethics committee for the federal state of Salzburg issued a positive vote on the ethics application and approved it on 30 August 2021, with the authorization being valid for 1 year (EC no. 1122/2021). The study was conducted in accordance with the Helsinki Declaration of 1964 and its later amendments, and it was not pre-registered on any clinical study registry. All questionnaires used were license-free, could be downloaded free of charge and did not require approval. Most of the measurements for the test subjects in the migraine group were conducted at the Department of Neurology at the Christian Doppler University Hospital in Salzburg, Austria. The healthy control group and part of the migraine group ( n  = 7) were tested on the premises of the Department of Physiotherapy at the University of Applied Sciences in Graz. Both well-lit and quiet rooms were approximately 12 m 2 in size, including a window to ensure regular ventilation. Each room contained a desk for discussing the questionnaires and conducting tests using tablets or notebooks, two armchairs and a treatment bench for performing the QST. All test subjects were tested by the first author of this study (BT). After the explanatory talk about the examination itself and the clarification of open questions on the part of the participants, the first four tests were carried out using an iPad tablet and notebook. The complete QST was then carried out, followed by the CPM. In total, the procedures took around 2 h per person. Between the individual test steps, the room was regularly ventilated 3–5 times, and the lighting was chosen so that it was not irritating, especially for the people in the migraine group. A total of six validated questionnaires, described in the following sections, were used for the study. The TAS-20 records alexithymia, i.e. the so-called emotional blindness, using a questionnaire with 20 items. The German translation of the original English scale [ 39 , 40 ] shows adequate validity and reliability in a sample analysis [ 41 ]. The scale is still recommended for the assessment of alexithymia [ 42 ]. However, in the course of an investigation into the psychometric properties in the application of the scale in somatoform pain disorders, the originally propagated three-factor model is not recommended and only the use of the total values is recommended for interpretation and recording [ 43 ]. The DASS is a self-report instrument designed to measure three related negative emotional states: depression, anxiety and stress. Each of these states is assessed by a set of 14 items, making a total of 42 items. The questionnaire uses a 4-point severity/frequency scale to assess the extent to which an individual has experienced each state over the past week. Overall, the DASS is both reliable and valid and is recommended in its short form for patients with persistent pain [ 44 , 45 ]. This questionnaire is available in a validated German version and contains 25 items, which are rated from 0–4 points. The cut-off value for the hypothesis of the existence of a central sensitization of pain is 40 points. In an empirical study on the validity of the German tool called “Zentrale Sensibilisierung Screeningtool” (ZSS), the Neck Disability Index (NDI) was used as a comparable gold standard for a population that suffered from persistent, non-specific neck pain. The Spearman rank correlation showed a statistically significant correlation between NDI and ZSS of ρ  = 0.5570 ( p  < 0.01). With a Cronbach's alpha of α  = 0.87 and a total of 25 items, the internal consistency of the ZSS and thus also its validity can be rated as good [ 46 ]. The MIDAS shows moderate to strong test-retest reliability and criterion validity for measuring headache-related disability over 3 months. It is responsive to clinical change but may be affected by recall bias [ 47 , 48 ]. The HIT-6 has strong internal consistency ( α  > 0.80) and high convergent validity with other quality-of-life and headache measures. It is sensitive to changes in headache burden and commonly used in both research and clinical settings [ 49 ]. The questionnaire on the severity of chronic pain, originally developed by von Korff et al. [ 50 ], has proven particularly useful in the field of orofacial symptoms [ 51 ] but is also used as an internationally comparable and established tool for other persistent pain disorders [ 52 , 53 ]. The German translation used in our study has proven to be a valid and reliable measurement instrument for hierarchically classifying the severity of chronicity and shows good comparability with the original English version [ 54 ]. The questionnaires TAS-20, DASS and the German version of the CSI were administered to both the migraine group and the control group, while MIDAS, HIT-6 and GCPS were used only in the migraine group. Quantitative sensory testing (QST) was performed according to the standardized protocol of the German Research Network on Neuropathic Pain (DFNS) [ 22 ] to assess thermal and mechanical sensory thresholds using the Q-Sense device (Medoc Ltd. Advanced Medical Systems, Israel), which quantifies the function of Aδ- and C-fibres. Thermal thresholds including cold detection (CDT), warm detection (WDT), cold pain (CPT), heat pain (HPT) and thermal sensory limen (TSL) were measured via a Peltier element thermode (30 × 30 mm) applied to the dominant hand (dorsum) and the right and left forehead (see Fig.  1 ). Starting at a baseline of 32 °C, temperatures were altered incrementally (1 °C/s) with participants signalling their perception. Safety limits restricted temperatures to 20–50 °C. Results were averaged from repeated measurements and exported to Excel. Mechanical thresholds including touch sensitivity and mechanical pain thresholds were assessed using von Frey hairs (0.25–256 mN) and pin-prick stimulators (8–512 mN) with thresholds calculated as geometric means. Dynamic mechanical allodynia and pain sensitivity were tested using mixed stimuli (cotton ball, Q-tip, brush) rated on a 0–100 scale. The wind-up ratio (WUR), a peripheral and central sensitization measure, was calculated from ratings of repeated pin-prick stimuli (256 mN on the hand, 128 mN on the forehead) applied at 1 Hz. Vibration detection thresholds (VDTs) were measured with a 64-Hz tuning fork applied to bony prominences. In contrast, pressure pain thresholds (PPTs) were recorded using a pressure algometer (Wagner Instruments™, Greenwich, CT) on muscle bellies (thenar muscles for the hand, temporalis muscle bilaterally for the forehead). These methods ensure reliable and reproducible assessment of sensory and pain thresholds grounded in published reference data [ 21 , 55 – 57 ]. Fig. 1 Procedures used in quantitative sensory testing (QST). Top row (left to right): thermal detection threshold testing with a thermode; mechanical detection threshold testing using von Frey filaments; pinprick pain threshold assessment using a weighted pinprick stimulator. Bottom row (left to right): dynamic mechanical allodynia testing with a soft brush; vibration detection threshold with a tuning fork; pressure pain threshold measurement using a mechanical pressure algometer (all photographs © by the author [Bernhard Taxer]—depiction of the person shown permitted. Figures adapted from the PhD thesis of Bernhard Taxer) Procedures used in quantitative sensory testing (QST). Top row (left to right): thermal detection threshold testing with a thermode; mechanical detection threshold testing using von Frey filaments; pinprick pain threshold assessment using a weighted pinprick stimulator. Bottom row (left to right): dynamic mechanical allodynia testing with a soft brush; vibration detection threshold with a tuning fork; pressure pain threshold measurement using a mechanical pressure algometer (all photographs © by the author [Bernhard Taxer]—depiction of the person shown permitted. Figures adapted from the PhD thesis of Bernhard Taxer) The ice water test was employed to investigate conditioned pain modulation (CPM) in this study. The water was cooled to 8 °C using ice cubes [ 58 , 59 ], and the temperature was measured with a conventional bathtub thermometer. Participants immersed their non-dominant hand in the water for 1 min, with the option to remove it earlier if they experienced severe pain. Pain perception was recorded on a 0–100 scale. Pressure algometer measurements on the thenar muscles, applied at the palm side of the thenar muscle on the point connecting the longitudinal axis of the thumb and index finger [ 60 ] and the temporal muscles on both sides, determined at a fixed location 2 cm behind the lateral canthus of the eye and 2 cm above the orbito-meatal line where palpation had revealed the anterior part of the temporal muscle to be most prominent [ 35 , 61 ], served as test stimuli. The results were calculated using a ratio by dividing the second value after the ice water immersion by the first value with higher ratios indicating activated conditioned pain modulation. The study population in the present analysis is identical to that described in Taxer et al. [ 62 ]. As the current work represents a secondary analysis of the same dataset, no new sample size calculation was performed. Although the sample size calculation was originally based on parametric assumptions, reflecting common practice in prior QST studies [ 63 – 66 ], the final choice of non-parametric analysis was guided by the actual data distribution and measurement level of the variables. Group differences in outcome measures were summarized descriptively using boxplots, which depicted the minimum, lower quartile, median, upper quartile and maximum. For statistical inference, non-parametric ANOVA tests were performed, employing a two-sided alpha level of 0.05 [ 67 ]. The Bonferroni-Holm correction was applied to account for multiple testing. While QST typically yields interval-scaled data (e.g. temperature or pressure thresholds), the subjective nature of pain ratings and sensory thresholds—especially in chronic pain populations—raises concerns about their interval properties and robustness to outliers. In our dataset, several variables showed skewed distributions and non-constant variance, which motivated a conservative, non-parametric approach. Accordingly, Spearman's rank correlation coefficients were used to assess monotonic relationships. This choice reflects both the ordinal nature of some outcome variables (e.g. pain intensity ratings) and the distributional characteristics of the data. To standardize the evaluation of the QST scores, a Z-score was calculated for each subject in the migraine group based on the mean and standard deviation of the control group. The mean Z-score was then used to observe the average deviation from the control group. The Z-transformation was performed according to Rolke et al. to create profiles of sensory changes [ 22 ]. All analyses were conducted using the statistical software package R (version 4.1.3) [ 68 ]. Several QST-based studies in orofacial pain research [ 69 ] have employed parametric methods based on the assumption of interval-scaled data. While our research question is conceptually aligned—focusing on trigeminal sensitivity and somatosensory processing in chronic migraine—the present dataset exhibited characteristics that called for non-parametric methods. We therefore refrained from re-calculating sample size for a non-parametric design retrospectively. However, future studies could build on these insights to design power analyses based on more conservative statistical assumptions or robust estimators.

Baseline characteristics of individuals including origin, educational background and the extent of headache burden are summarised in Table  1 . Table 1 Baseline characteristics of all test subjects MG ( n  = 45) CG ( n  = 25) Age Median (range) 38 (43) 38 (32) Sex f = 39 f = 21 Nationality n  = 45 AT n  = 25 AT Employed Yes: n  = 33 No: n  = 6 Academic education: n  = 6 Yes: n  = 15 No: n  = 0 Academic education: n  = 10 Highest level of education Primary school: n  = 14 Graduate: n  = 17 University: n  = 14 Primary school: n  = 0 Graduate: n  = 12 University: n  = 13 Months with HA Median (range) 252 (552) – Median (range) number of monthly HA days (derived from MIDAS) 12.3 (26) – AT Austrian, MG migraine group, CG control group, m male, f female, n number of subjects, HA headache (Table adapted from the PhD thesis of Bernhard Taxer and translated into English for Neurology and Therapy) Baseline characteristics of all test subjects Age Median (range) Yes: n  = 33 No: n  = 6 Academic education: n  = 6 Yes: n  = 15 No: n  = 0 Academic education: n  = 10 Primary school: n  = 14 Graduate: n  = 17 University: n  = 14 Primary school: n  = 0 Graduate: n  = 12 University: n  = 13 Months with HA Median (range) AT Austrian, MG migraine group, CG control group, m male, f female, n number of subjects, HA headache (Table adapted from the PhD thesis of Bernhard Taxer and translated into English for Neurology and Therapy) In addition to the general characteristics of the participants, Table  2 attempts to capture potential confounders that must be considered when interpreting the results. Table 2 Possible confounders to be considered MG ( n  = 45) CG ( n  = 25) Current medication NOPA: n  = 34 Triptans: n  = 22 Tricyclic AD: n  = 9 MAB: n  = 9 Anticonvulsive: n  = 5 (of which TPM: n  = 3) Beta-blockers: n  = 2 SNRI: n  = 2 – Current and previous non-drug treatment(s) Current: Physiotherapy: n  = 6 Acupuncture: n  = 4 Massage: n  = 3 Relaxation techniques: n  = 2 Psychotherapy: n  = 2 Osteopathy: n  = 1 Orthodontics: n  = 1 Previous: Physiotherapy: n  = 7 Acupuncture: n  = 7 Relaxation techniques: n  = 3 Osteopathy: n  = 3 Botulinum toxin: n  = 3 Massage: n  = 2 Orthodontics: n  = 1 Biofeedback: n  = 1 Migraine surgery: n  = 1 – Additional diagnoses Cervical spine shoulder syndrome: n  = 10 Depression: n  = 3 Anxiety disorder: n  = 1 Mb. Hashimoto: n  = 3 Hypothyroidism: n  = 3 Post-COVID syndrome: n  = 2 Chronic gastritis: n  = 2 Mb. Basedow: n  = 1 Polycystic ovarian syndrome: n  = 1 Psoriasis: n  = 1 Melkerson-Rosenthal syndrome: n  = 1 Tension-type headache: n  = 1 Restless leg syndrome: n  = 1 Irritable bowel syndrome: n  = 1 Mb. Raynaud: n  = 1 Arterial hypertension: n  = 1 Chronic sinusitis: n  = 1 Endometriosis: n  = 1 Adipositas: n  = 1 Temporomandibular joint pain: n  = 1 Mb. Basedow: n  = 1 Polycystic ovarian syndrome: n  = 1 Endometriosis: n  = 1 Persistent pain (> 3 months) in other body regions Spine: n  = 12 Shoulder: n  = 5 Hip: n  = 1 Spine: n  = 2 Knee: n  = 1 Lower thigh: n  = 1 Nicotine Yes: n  = 6 No: n  = 39 Yes: n  = 1 No: n  = 24 Time since last menstrual period  < 1 week: n  = 5  < 2 weeks: n  = 9   4 weeks respectively menopause: n  = 9 No information: n  = 15  < 1 week: n  = 3  < 2 weeks: n  = 4   4 weeks respectively menopause: n  = 5 No information: n  = 5 Hormonal treatment Thyroid medication: n  = 9 Oral contraception: n  = 6 Other: n  = 5 Hormone coil: n  = 3 None: n  = 23 Thyroid medication: n  = 1 Oral contraception: n  = 4 Other: n  = 0 Hormone coil: n  = 1 None: n  = 20 MAB monoclonal antibody therapy, NOPA non-opioid analgesics, n number of subjects, SNRI serotonin noradrenaline reuptake inhibitors, TPM topiramate, tricyclic, AD tricyclic antidepressants, MG migraine group, CG control group (Table adapted from the PhD thesis of Bernhard Taxer and translated into English for Neurology and Therapy) Possible confounders to be considered NOPA: n  = 34 Triptans: n  = 22 Tricyclic AD: n  = 9 MAB: n  = 9 Anticonvulsive: n  = 5 (of which TPM: n  = 3) Beta-blockers: n  = 2 SNRI: n  = 2 Current: Physiotherapy: n  = 6 Acupuncture: n  = 4 Massage: n  = 3 Relaxation techniques: n  = 2 Psychotherapy: n  = 2 Osteopathy: n  = 1 Orthodontics: n  = 1 Previous: Physiotherapy: n  = 7 Acupuncture: n  = 7 Relaxation techniques: n  = 3 Osteopathy: n  = 3 Botulinum toxin: n  = 3 Massage: n  = 2 Orthodontics: n  = 1 Biofeedback: n  = 1 Migraine surgery: n  = 1 Cervical spine shoulder syndrome: n  = 10 Depression: n  = 3 Anxiety disorder: n  = 1 Mb. Hashimoto: n  = 3 Hypothyroidism: n  = 3 Post-COVID syndrome: n  = 2 Chronic gastritis: n  = 2 Mb. Basedow: n  = 1 Polycystic ovarian syndrome: n  = 1 Psoriasis: n  = 1 Melkerson-Rosenthal syndrome: n  = 1 Tension-type headache: n  = 1 Restless leg syndrome: n  = 1 Irritable bowel syndrome: n  = 1 Mb. Raynaud: n  = 1 Arterial hypertension: n  = 1 Chronic sinusitis: n  = 1 Endometriosis: n  = 1 Adipositas: n  = 1 Temporomandibular joint pain: n  = 1 Mb. Basedow: n  = 1 Polycystic ovarian syndrome: n  = 1 Endometriosis: n  = 1 Spine: n  = 12 Shoulder: n  = 5 Hip: n  = 1 Spine: n  = 2 Knee: n  = 1 Lower thigh: n  = 1 Yes: n  = 6 No: n  = 39 Yes: n  = 1 No: n  = 24 < 1 week: n  = 5 < 2 weeks: n  = 9  4 weeks respectively menopause: n  = 9 No information: n  = 15 < 1 week: n  = 3 < 2 weeks: n  = 4  4 weeks respectively menopause: n  = 5 No information: n  = 5 Thyroid medication: n  = 9 Oral contraception: n  = 6 Other: n  = 5 Hormone coil: n  = 3 None: n  = 23 Thyroid medication: n  = 1 Oral contraception: n  = 4 Other: n  = 0 Hormone coil: n  = 1 None: n  = 20 MAB monoclonal antibody therapy, NOPA non-opioid analgesics, n number of subjects, SNRI serotonin noradrenaline reuptake inhibitors, TPM topiramate, tricyclic, AD tricyclic antidepressants, MG migraine group, CG control group (Table adapted from the PhD thesis of Bernhard Taxer and translated into English for Neurology and Therapy) Of the three questionnaires that were compared between the MG and the CG (CSI, DASS, TAS-20), only the questionnaire for recording central sensitization (CSI) showed a statistically significant difference between these groups with p  < 0.001 (Fig.  2 ). The remaining parameters from the TAS-20 and DASS questionnaires show no statistically significant differences between the two groups in this study ( p  = 0.24 and p  = 0.39). Fig. 2 Box-plot. Comparison of the Central Sensitization Inventory (CSI), Depression Anxiety and Stress Scale (DASS) and Toronto Alexithymia Scale-20 items (TAS-20) questionnaires between migraine (MG) and control group (CG) (figures in points). Numbers next to the boxes represent the median. (Figures adapted from the PhD thesis of BT and translated into English for Neurology and Therapy) Box-plot. Comparison of the Central Sensitization Inventory (CSI), Depression Anxiety and Stress Scale (DASS) and Toronto Alexithymia Scale-20 items (TAS-20) questionnaires between migraine (MG) and control group (CG) (figures in points). Numbers next to the boxes represent the median. (Figures adapted from the PhD thesis of BT and translated into English for Neurology and Therapy) There were no significant differences in the QST parameters between the two groups tested across the entire test battery. However, the descriptive analysis suggested a trend toward altered heat pain sensitivity of the forehead and minimal variations in the cold detection threshold on the left forehead, although these differences did not reach statistical significance. To facilitate comparison, Z-scores were calculated from the mean values of the measurements (see Figs.  3 and 4 ). A Z-score of 0 indicates a value identical to the mean of the healthy control group, while negative or positive Z-scores reflect a loss or gain of function, respectively, in the tested area (detailed in the supplementary material). Fig. 3 Z-score value quantitative sensory testing (QST), including cold detection thresholds (CDT), cold pain threshold (CPT), heat pain threshold (HPT), mechanical pain threshold (MPT), pressure pain threshold (PPT), stimulus response function (SRF), thermal sensory limen (TSL), vibration detection threshold (VDT) and wind-up ratio (WUR) for each region. (Figures adapted from the PhD thesis of Bernhard Taxer and translated into English for Neurology and Therapy) Fig. 4 Z-score value quantitative sensory testing, including cold detection thresholds (CDT), cold pain threshold (CPT), heat pain threshold (HPT), mechanical pain threshold (MPT), pressure pain threshold (PPT), stimulus response function (SRF), thermal sensory limen (TSL), vibration detection threshold (VDT) and wind-up ratio (WUR) for all regions together. (Figures adapted from the PhD thesis of Bernhard Taxer and translated into English for Neurology and Therapy) Z-score value quantitative sensory testing (QST), including cold detection thresholds (CDT), cold pain threshold (CPT), heat pain threshold (HPT), mechanical pain threshold (MPT), pressure pain threshold (PPT), stimulus response function (SRF), thermal sensory limen (TSL), vibration detection threshold (VDT) and wind-up ratio (WUR) for each region. (Figures adapted from the PhD thesis of Bernhard Taxer and translated into English for Neurology and Therapy) Z-score value quantitative sensory testing, including cold detection thresholds (CDT), cold pain threshold (CPT), heat pain threshold (HPT), mechanical pain threshold (MPT), pressure pain threshold (PPT), stimulus response function (SRF), thermal sensory limen (TSL), vibration detection threshold (VDT) and wind-up ratio (WUR) for all regions together. (Figures adapted from the PhD thesis of Bernhard Taxer and translated into English for Neurology and Therapy) The prominent Z-scores shown for the mechanical detection threshold can be explained by two clear outliers in the measurements of the migraine group. The non-representable value of the vibration detection threshold (VDT) in the measurement area of the dominant hand is due to the missing variance from the mean values in the control group. As all existing values are 8, there is no scattering of the results and a Z-score cannot be derived in this case. Testing the CPM using the ice water test did not reveal any statistically significant difference between the two groups during these investigations. A presentation of the detailed results is available within the supplementary material. The study allowed examination of correlations between various parameters including the Depression Anxiety and Stress Scale, Conditioned Pain Modulation, Quantitative Sensory Testing, Migraine Disability Assessment, Headache Impact Test and Wind-up ratio in participants with migraine and healthy controls. In the migraine group, weak to moderate monotonic correlations were identified between DASS scores and specific QST parameters such as increased cold pain sensitivity on the left forehead ( ρ  = 0.3586) and decreased mechanical pain thresholds on both sides of the forehead ( ρ  = − 0.3563 right, ρ  = − 0.4144 left). For CPM, a weak negative correlation was found between DASS and the pressure pain threshold ratio of the right temporalis muscle ( ρ  = − 0.2103) with a similar trend in the control group ( ρ  = − 0.4167). MIDAS severity levels did not correlate with CPM, but weak negative correlations with QST parameters including CDT (e.g. ρ  = − 0.2005 on the left forehead) and HPT of the dominant hand ( ρ  = − 0.2748) were observed. The median number of 12.3 (range 26) monthly headache days showed weak positive correlations with the WUR in several areas such as the dominant hand ( ρ  = 0.3275) and the left forehead ( ρ  = 0.2573). HIT-6 scores correlated weakly with CPT on the left forehead ( ρ  = 0.2235) and CPM in the thenar muscle of the dominant hand ( ρ  = 0.2910). In the migraine group, WUR demonstrated weak to moderate correlations with CPM across test areas, indicating possible central sensitization processes that were absent in controls. Finally, the WUR correlated weakly positively with the Central Sensitization Inventory in the right forehead region for both groups (e.g. ρ  = 0.2347 in the migraine group). These findings suggest a role of central mechanisms and sensitization in migraine as reflected in specific sensory and psychological parameters.

This cross-sectional study is, to the authors’ knowledge, the first to apply a multimodal test battery combining quantitative sensory testing, psychometric assessments and conditioned pain modulation in HFEM and CM. This approach generated a rich dataset enabling clinically relevant correlation analyses. Nevertheless, possibly due to the limited sample size and other above mentioned methodological constraints, hardly any statistically significant results were observed between migraine groups and healthy controls in somatosensory, psychometric or CPM measures, except for one self-reporting questionnaire assessing central sensitization (CSI). QST is considered a valid psychophysical procedure in the context of neuropathic pain syndromes, but in the present study, it did not provide any statistically significant differences between the tested groups. However, testing the WUR stands out given the correlations recorded and could play a role for future test clusters, whether in the context of studies using cross-sectional design or during therapeutic or pharmacological intervention studies. However, this parameter would first have to be tested on a larger study population to establish its validity. In conclusion, our findings suggest that, within the present sample, HFEM and CM are not consistently associated with altered somatosensory perception, psychosocial burden—particularly depression, anxiety, stress and alexithymia—or impaired endogenous pain inhibition, underscoring the need for longitudinal and subgroup-specific research to identify phenotypes at risk.

The absence of significant differences in QST parameters between individuals with migraine and healthy controls observed in our study represents a relevant negative finding. This result invites critical reflection on the sensitivity of QST to detect altered somatosensory perception in migraine and may point to the influence of interictal measurements, clinical heterogeneity or sample-specific characteristics. Previous studies have demonstrated that individuals with migraine often exhibit abnormalities in nociceptive detection and processing as revealed through quantitative sensory testing. These differences, which can be specific to sensory modalities, measurement methods, test locations and migraine phase examined, have been reported to distinguish individuals with migraine from healthy controls [ 23 ]. Such findings highlight the potential of QST to detect somatosensory-related changes including altered recognition thresholds and hyper-responsiveness in chronic pain conditions. These gain-of-function phenomena are also the main characteristics in conditions underlying a nociplastic pain mechanism and part of its proposed descriptive pathway [ 25 ]. Sensory thresholds have been shown to differ between the ictal and interictal phases of the migraine cycle [ 70 ]. Evidence from functional magnetic resonance imaging (fMRI) further indicates that these thresholds begin to decline as early as 48 h prior to the onset of clinically apparent symptoms. These early changes suggest that sensory dysfunction precedes the headache phase and may reflect one of the initial manifestations of migraine-related pathophysiological processes [ 9 , 11 , 31 ]. A reduction in temperature and electrical pain thresholds during the preictal phase has recently been demonstrated by Peng and May in menstrual migraine also using QST in a longitudinal case study [ 10 ]. On the other hand, these findings imply that in individuals with HFEM or CM (12.5 ± 6 headache days per month in our study), distinguishing between ictal and interictal phases may be considerably more challenging because of the temporal overlap of symptoms. Nevertheless, this diagnostic ambiguity—alongside the epidemiological relevance of this subgroup—served as a key rationale for investigating HFEM or CM in the present study. Although we are aware of the limitations involved, we would still like to present these negative results as we are convinced of their validity based on a thorough and complete DFNS protocol and appropriate questionnaires. The CPM examination, on the other hand, is in line with the studies by Nahman-Averbuch et al. [ 23 , 71 ], where no significant differences were found in a classical CPM procedure as performed in our study. Our study adopts a cross-sectional design, which has inherent limitations. Simultaneous assessment of both the disease and associated abnormalities restricts the ability to establish clear causal relationships or consider pre-existing conditions [ 72 ]. The inclusion of rare pathologies necessitates large sample sizes for meaningful results, posing challenges in this study. However, such designs provide foundational data for future prospective or retrospective studies like cohort or case-control designs [ 72 ]. The most common additional musculoskeletal disorder in the migraine group ( n  = 10) was cervical spine/shoulder syndrome. These results go hand in hand with studies on the overlap of headache syndromes and other musculoskeletal complaints, explicitly in the shoulder-neck region [ 73 – 76 ]. Surprisingly, only one participant reported a diagnosis of temporomandibular joint pain (TMJ pain) [ 77 ]—a condition frequently observed in migraine, particularly in its chronic form [ 78 ]. The relative scarcity of reported TMJ pain in our sample may reflect underdiagnosis or sample-specific characteristics, yet it does not weaken the broader pattern suggesting that coexisting musculoskeletal and systemic conditions may amplify peripheral and central sensitization. Such sensitization is thought to contribute to nociplastic pain mechanisms [ 79 ] and is also observed in other persistent syndromes identified in our cohort, including irritable bowel syndrome ( n  = 1) and chronic gastritis ( n  = 2) [ 80 ]. These associations reinforce the concept that migraine often exists within a network of overlapping pain-related disorders, potentially sharing common underlying mechanisms. As demonstrated by Peng and May (2025) in their recent study on menstrual components and migraine, a deeper exploration of this topic would have been valuable. We recorded the time from the last menstruation in our participants (see Table  2 ) and, like Peng and May, applied QST—although in contrast to their study, we included a control area and assessed all components of the DFNS protocol. Interestingly, Peng and May found no significant menstruation-dependent influence on sensory thresholds [ 10 ]. Our control group also exhibits confounding factors that could influence the results. The initial characteristics in terms of age and gender distribution are comparable, but there are differences in the use of medication and additional complaints and illnesses. Comparing the recorded values of the control group with reference values from Rolke et al. [ 56 ], there are no identifiable differences in the mean values, and they showed comparable mean values for the region of the dominant hand. Additionally, it should be noted that the control group was recruited from the university environment and the personal network of the first author (BT). This recruitment strategy may have introduced a selection bias, as participants in the control group might differ from those in the migraine group regarding educational background and other sociodemographic characteristics, potentially confounding group comparisons. Both MIDAS and the HIT-6 appear to be suitable questionnaires for looking at the domains for headache assessment as they adequately evaluate both the intensity of the headache itself (days, frequency) and the impact of the headache on everyday life. The primary method to assess alexithymia is using the TAS-20 or TAS-26, which are considered the optimal instruments for this purpose. The DASS was utilized to encompass a more extensive array of psychological dimensions with a single questionnaire, thereby circumventing potential non-compliance issues during the completion of the questionnaires. Furthermore, it is imperative to emphasize that within the DASS, the depression scale alone, with a sensitivity of 77% and a specificity of 83%, attains results comparable to, and in certain instances even superior to, those described for the effective detection of depression using a questionnaire [ 44 , 45 ]. The absence of disparities in the domain of psychological components is noteworthy, given the findings of extensive reviews that suggest an increased prevalence of these phenomena [ 81 , 82 ], and research has demonstrated the efficacy of psychological interventions in the management of migraine [ 83 ]. Although this result may be due to the small comparison groups, it might be advisable to tighten the inclusion criteria in future studies, especially for the non-affected participant group, so that multidimensional questionnaires such as the DASS are used in advance for better filtering and comparison with a respondent group without any psychological abnormalities. Quantitative sensory testing is a valid and well-studied psychophysical procedure and has already been used in various clinical areas [ 22 , 56 , 57 , 84 – 86 ]. Concerning migraine, the systematic review by Nahman-Averbuch et al. provides the largest overview of this topic to date [ 23 ]. In addition to the different degrees of severity of the migraine present, the contributing confounders, such as medication or perceived pain at the time of the test, could also influence the results of the sensory tests. The results of the QST conducted in our study show no significant differences between individuals with migraine headaches and the healthy control group. The review by Nahman-Averbuch et al. (2018) describes that people suffering migraine showed a higher pain sensitivity to PPT, HPT, suprathreshold electrical and suprathreshold cold stimuli. However, no differences were found for other QST paradigms such as electrical detection thresholds, electrical, cold and mechanical pain thresholds, or suprathreshold noxious heat [ 23 ]. The present study carried out QST according to the complete DFNS protocol. Nahman-Averbuch et al. [ 23 ] criticize the fact that many studies conducted to date have not followed this protocol. In particular, the parameters PPT and HPT are repeatedly emphasized and their clinical relevance during migraine is stressed. This appears to apply primarily to the craniocervical region rather than to distant or extratrigeminal areas, leading to the assumption that the examination and possible treatment of sensitized myofascial regions could play a significant role in the reduction of pain. The examination of these regions was recommended alongside other musculoskeletal examination methods for the primary headache types [ 74 , 75 , 87 ]. To date, additional sensory examination methods have only been described for 2-point discrimination. It has been shown that people suffering from migraine also exhibit significantly altered interictal values for high cervical 2-point discrimination compared to a healthy group of test subjects [ 88 ]. The uniform diagnosis of migraine according to the ICHD-3 criteria and the structured and standardized implementation of the QST in our study must be positively emphasized. Thus, data from all items of the QST are available in the present study, which differs greatly from the work of Nahman-Averbuch et al. (2018) where only parts of the DFNS protocol were included throughout. The complete implementation required more time, but from a methodological point of view, it provides added value in identifying which methods can be used specifically for CM and which are less relevant. The wind-up ratio appears to play a role in higher headache frequencies. The more headache days there are, the more conspicuous the WUR is in our study. The WUR in the context of QST is considered a possible clinical representation of central sensitization processes during nociception and pain. In the present study, a weak to moderate correlation between the ratio of the wind-up test and the calculated ratio of conditioned pain modulation was found in almost all areas measured in the migraine group. A weak statistical correlation was found between the WUR and the CPM in the migraine group, meaning that as one value increased or decreased, the other tended to change in the same direction—nevertheless, this pattern was not observed in the healthy control group. This could be clinically relevant when testing affected individuals for these parameters and trying to use this therapeutically. Clinically, this could support the inclusion of this examination procedure, as it is already used in the course of neuropathies or complex regional pain syndromes (CRPS) but also shows indifferent results there [ 89 – 91 ]. Different approaches are used to investigate CPM. In addition to the ice water test, which was used in this study, the phenomenon of endogenous pain inhibition could also be investigated using warm water tests, pressure cuffs or capsaicin applications. These tests consistently demonstrate adequate validity [ 92 ]. For this study, a standard protocol was used that meets the quality criteria. It could be relevant to carry out this examination using isometric exertion, as has already been investigated in musculoskeletal syndromes with varying results, showing inconsistent but also significant effects [ 93 , 94 ]. This procedure would be relevant as it might have added clinical value. Nevertheless, in the present study, no significant differences were found in the investigation of CPM. These results are consistent with the studies by Nahman-Averbuch et al. (2019) and Kisler et al. (2018) whereby the former involved a group of test subjects aged 12–17 years [ 95 ] and the second mainly included subjects with EM [ 96 ]. In addition, the methods of test stimuli and conditioning stimuli differ fundamentally from the ice water test carried out in our study [ 95 , 96 ]. In contrast, other studies indeed show a limitation of CPM compared to healthy control groups, although it must also be pointed out here that the inclusion criteria and study designs are not comparable, especially concerning the CPM method [ 71 , 97 ]. The weak monotonic correlations of CPM and increased DASS scores in our study apply to both tested groups. This therefore appears to be independent of headache and suggests an association between psychological factors and extent of endogenous pain inhibition. Finally, our study identified significant differences between the groups based on the results of the German version [ 46 ] of the CSI questionnaire [ 98 ]. While the CSI is commonly used to identify symptoms associated with central sensitization, it primarily reflects self-reported experiences rather than direct physiological measurements of nociception. Research suggests that the CSI shows only weak or no correlations with experimental measures of nociceptive sensitivity such as pain thresholds, temporal summation or conditioned pain modulation [ 99 ]. This raises questions about its ability to accurately represent the physiological processes of central sensitization. Instead, the CSI may be more indicative of psychological hypervigilance and associated factors rather than increased responsiveness of nociceptive neurons [ 99 ]. Using the CSI to record central sensitization is not a customary component of migraine studies. However, intervention studies are increasingly employing this tool to record an additional objectifiable outcome parameter beyond restrictions related to headache frequency or medication use [ 100 ]. The purpose of this questionnaire is to identify the underlying mechanism by which the condition manifests. This is achieved by determining the potential overlap with other pain syndromes, including restless leg syndrome, TMJ pain and cranial autonomic symptoms associated with migraine [ 101 – 103 ]. Consequently, the implementation of this questionnaire could be pertinent in both a scientific context and in clinical practice and would be a valuable addition to any treatment courses. Although the questionnaire is principally established as a screening tool, it nevertheless has the potential to be of use as a follow-up parameter during studies and treatments. In addition, the questionnaire provides clinical indications of other physical and psychologically relevant symptoms that enable a holistic approach to affected individuals. The presentation of the confounders (additional diagnoses) to be considered also suggests that ongoing central sensitization can be associated with various other syndromes. Among other things, overlaps with clinical pictures such as restless leg syndrome and TMJ pain, as well as cranial autonomic syndromes, and a heightened sensitivity to pain in test subjects with EM and CM are connected to the neurophysiological phenomenon of central sensitization in the literature [ 101 – 104 ]. This study has several limitations. First, the number of participants was limited because of the extensive measurements, and a comprehensive analysis of the QST parameters was hindered because of significant variance in the target variables. Despite these challenges, the study benefitted from comparable migraine and healthy control groups [ 105 – 107 ]. Although speculative, regarding the non-significant results between the two groups, the number of test subjects represents a limitation of this work and must be considered for further work and meaningful results. Second, the cross-sectional design prevents reaching any conclusions about causality among migraine, sensory alterations and psychological variables. Third, the migraine group was heterogeneous regarding comorbid conditions and medication use, which may have confounded the results. For example, the number of days with acute medication use was not documented; therefore, a potential confounding influence of such medication on the results cannot be excluded. Lastly, the use of self-report questionnaires such as the CSI may reflect psychological distress rather than direct physiological markers of central sensitization. The isolated observation of the descriptive data for heat pain sensitivity in the QST indicates a tendency towards abnormality, even if this was not statistically significant in our study. This parameter should be further investigated together with heat detection threshold, cold pain threshold and the thermal sensory limen, which are also further from 0 in the Z-score, to be able to generate parameters that can be used independently for future QST investigations. Surprisingly, the PPT in the present study was much less noticeable in the descriptive data or the Z-scores than previous publications would suggest [ 74 , 108 – 110 ]. The inclusion of the CSI could be relevant regarding physiotherapeutic options but also extended specific examination tools. The WUR could become more important as a sensory test, especially in higher headache frequencies. In addition, the WUR provides indications of impairments during conditioned pain modulation. This results in a series of tests that should be systematically evaluated in further investigations to derive sound and evidence-based recommendations. During daily practice, these findings may guide practical bedside assessments. For example, simple thermal stimuli (e.g. warm/cold metal tools, coins) could be used qualitatively to explore altered thermal pain thresholds in migraine sufferers. Likewise, applying a repeated pinprick stimulus (as a surrogate for WUR) can reveal increased temporal summation, helping clinicians to screen for central sensitization. Even without formal equipment, these observations—when carefully documented—can support treatment decisions, particularly when evaluating strategies in movement therapies or responses to pain modulation interventions. Using the CSI as a baseline and follow-up tool in clinical care may also help to track symptom development beyond mere headache frequency.

Migraine is a common, disabling primary headache disorder with high socio-economic and personal impact [ 1 – 3 ]. According to the latest evaluation of the Global Burden of Disease Study in 2021, migraine is the most disabling neurological disorder in Western Europe and, with 43.38 million disability-adjusted life years (DALYs) [ 4 ], the third leading condition worldwide [ 5 ]. Over the past 3 decades, the global prevalence of migraine has increased by 48% among women of childbearing age, making it the leading cause of disability in this demographic, surpassing back pain and depression [ 6 ]. Chronic migraine (CM) affects 1–2% of the global population, with about 3% of episodic migraine (EM) sufferers developing CM annually [ 7 ]. Migraine diagnosis relies primarily on clinical features, as no definitive biomarkers are currently available. Additional investigations, such as magnetic resonance imaging (MRI) or lumbar puncture, are typically employed to exclude secondary causes of headaches [ 8 ]. According to current knowledge, individual migraine attacks follow a cyclic progression, consisting of preictal, periictal and postictal phases, as well as a symptom-free interictal interval. Notably, hypothalamic activation is observed as early as 48 h prior to the onset of headache symptoms [ 9 – 11 ]. Chronic headache is one of the most demanding and disabling pain syndromes [ 12 , 13 ]. Besides pain, sufferers report severe limitations in daily activities and social interactions, leading to mood disorders and depression [ 14 , 15 ]. Alexithymia, defined as difficulty recognizing and processing emotions, has also been linked to chronic headaches, although its specific role in ongoing nociceptive processing remains unclear [ 15 , 16 ]. Previous studies have shown that individuals with chronic pain, including migraine, often exhibit sensitization characterized by altered recognition thresholds, hyper-responsiveness and impaired pain modulation [ 17 – 20 ]. These phenomena, reflecting changes in somatosensory perception, can be assessed through quantitative sensory testing (QST), which measures sensory and pain thresholds [ 21 , 22 ]. In individuals suffering from migraine, abnormalities in ongoing nociceptive processing associated with nociplastic mechanisms, i.e. changes in the central nervous system, have been reported to vary depending on the sensory modality, measurement method, test location and migraine phase (preictal, periictal, postictal, interictal) examined, thus offering valuable insights for distinguishing them from healthy controls [ 23 ]. We hypothesize that individuals suffering from high-frequency episodic (HFEM) or CM share clinical features related to nociplastic pain mechanisms, as described by Kosek et al. (2021). For example, some migraine sufferers describe mechanical allodynia during a headache attack as pain when combing their hair or even in other regions of the body [ 24 ]. These clinical signs might be detected sufficiently by QST methods as already described above and might be correlated to neuropsychological parameters [ 25 , 26 ]. Reflecting these ideas, this cross-sectional study aims to investigate changes in somatosensory functions using QST to possibly identify specific profiles or gain-of-function phenomena in individuals with HFEM or CM. In addition, psychometric questionnaires are used to assess stress, depression, anxiety and alexithymia. These domains are assessed together to capture the complex interplay between peripheral and central sensory processing and psychosocial factors, which may contribute to migraine chronification and treatment response. Additionally, we presumed to evaluate the participants responses to conditioned pain modulation (CPM), a measure of the human correlate of diffuse noxious inhibitory control (DNIC) utilizing a “pain inhibits pain” paradigm [ 27 ]. By integrating multimodal sensory and psychometric assessments, this work addresses existing gaps in understanding the sensory and affective dimensions of migraine and their potential role in impaired pain modulation.

Below is the link to the electronic supplementary material. Supplementary file1 (PDF 264 KB) Supplementary file1 (PDF 264 KB)