Vasomotor reactions in the face and head of patients with migraine

Abstract

Objective:

To systematically review extracranial vasomotor reactions in patients with migraine, and to discuss potential overlaps with rosacea.

Background:

Autonomic manifestations are common in migraine and may reflect trigeminovascular system activation. Flushing during migraine attacks may indicate altered extracranial vasomotor reactivity which resembles that of the facial skin disorder rosacea.

Methods:

PubMed and EMBASE were searched for studies investigating extracranial vascular reactions in migraine and rosacea published until January 2018.

Results:

We uncovered 41 studies investigating extracranial circulation in migraine and 12 studies in rosacea. Skin temperature was generally lower on the forehead, nose and hands of migraine patients compared to controls. The superficial temporal artery (STA) showed greater amplitude variations in migraine patients compared to controls, and compression of either the STA or the common carotid artery during attacks led to transient pain relief in about one-third of the investigated patients. Facial skin blood flow was asymmetrical in migraine patients compared to healthy volunteers.

Conclusion:

Findings on extracranial microcirculation suggest an altered function of the autonomic nervous system in migraine patients. Similar signs of autonomic nervous system dysfunction are reported in rosacea. We suggest an overlap in autonomic and neurovascular pathophysiology in migraine and rosacea.

Keywords

autonomic symptoms cranial arteries facial skin microcirculation temperature

Introduction

Autonomic manifestations are common in migraine¹ and are thought to reflect activation of the trigeminal parasympathetic pathway during attacks. Research in the area has predominately focused on the cranial autonomic symptoms; conjunctival injection, lacrimation, nasal congestion, rhinorrhoea and eyelid oedema.¹ However, clinical observations have shown that patients may experience facial pallor during the initial stage of migraine attacks, followed by a sense of flushing and sometimes facial oedema.² A predisposition to severe flushing of the face, neck and chest following emotional stress has been described in migraine patients,³ and a ‘red migraine’ comprising flushing and other autonomous symptoms was first described in 1867.⁴

A link between migraine and the dermatologic disorder rosacea has been reported in recent studies. The conditions overlap in epidemiology⁵ and may also share pathophysiological aspects.⁶ Rosacea is a facial skin condition characterized by facial redness and flushing in the centrofacial area along with secondary symptoms such as facial oedema, papules and pustules, ocular symptoms and rhinophyma. Disease activity in rosacea has been connected to release of the neuropeptides calcitonin gene-related peptide (CGRP), vasoactive intestinal polypeptide and pituitary adenylate cyclase-activating polypeptide, which are also active in migraine.⁷ Infusion with CGRP induces flushing in both healthy individuals and migraine patients and interestingly, pretreatment with CGRP receptor antagonist in healthy volunteers blocks flushing in 100% of the participants.⁸

Clinical observations have prompted further research in the extracranial vasomotor reactions in migraine patients through facial skin thermography, forehead skin blood flow measurements, pulsation measurements of the superficial arteries and estimation of forehead sweating as markers of autonomic dysfunction.^9
–11 The wide range of investigations and methods have resulted in very different and occasionally conflicting results, complicating direct comparison.

The purpose of this systematic review was to identify and categorise studies that investigated vasomotor reactions in migraine patients and discuss the next steps necessary in further illuminating facial vasomotor mechanisms in migraine. Furthermore, we sought to identify studies in rosacea with the same focus, to reveal potential overlaps in vasomotor reactions between the two disease entities. Investigations of facial blood flow and temperature in rosacea patients are sparse and inconclusive, and we suggest that studying extracranial vasomotor reactions in both rosacea and migraine patients may help in understanding the possible link between the two disorders.

Methods

Systematic review on migraine

A systematic literature search was performed using PubMed with the keywords (Migraine*) AND (facial OR face) AND (flow OR temperature OR circulation OR flushing OR heat). We furthermore searched EMBASE using the following 15 strings: 1:’exp migraine//’; 2:’facial.mp.’; 3:’extracranial.mp.’; 4:’blood flow.mp.’; 5: ’circulation.mp.’; 6: ’temperature.mp.’; 7: ’(review or meta-analysis or metaanalysis).m.titl.’; 8: ’(animal/ or nonhuman/) not human/’; 9: ’flushing.mp.’; 10: ’2 or 3’; 11: ’4 or 5 or 6 or 9’; 12: ’10 and 11’; 13: ’1 and 12’; 14: ’13 not 7’; 15: ’14 not 8’.

Search terms and in-/exclusion criteria were predefined (NW, CEC and MA). The review was conducted in accordance with the PRISMA guidelines (available at: http://www.prisma-statement.org/). The PubMed search was done on 2 March 2017 and the EMBASE search was finished on 19 July 2017. The PubMed search was repeated on 15 January 2018 in order to uncover any new relevant research in the area published in the time since our initial search was performed. Duplicates were extracted from the search and results were reviewed independently by two reviewers (NW and CEC), first by title and abstract for relevance, and then full text to confirm eligibility for this review. Disagreements were resolved through discussion.

Eligible articles were defined as experimental studies in patients with headache or headache-prone subjects, with focus on facial and/or extracranial circulation or temperature. Articles were excluded if they described animal studies or were published in other languages than English. Review articles, meta-analyses, editorials and conference abstracts were also excluded. Reference lists of included papers were also reviewed to uncover potentially missed articles throughout the initial search.

Systematic review on rosacea

A systematic literature search was also performed for rosacea by searching PubMed for references published any time before 9 January 2018 with the following search string: ‘Rosacea AND (flow OR temperature OR circulation OR heat) AND english[language] NOT review[publication type]’. Only human studies were included, and reference lists were screened for additional papers.

Results

Migraine: Search results

The PubMed search on migraine returned 144 articles, and 219 articles were identified through EMBASE. After removing duplicates, we were left with 308 original articles of which 33 were eligible for this article as they examined extracranial vasomotor responses in migraine patients. Eight additional studies were included after reviewing reference lists, resulting in a total of 41 studies eligible for review. The repeated search in January 2018 uncovered five new articles, but all were excluded because they were either animal studies or did not investigate extracranial circulation (Figure 1). The publications included in this review studied extracranial circulation in headache patients. Outcome variables were grouped in the following four categories: facial temperature; measurements of the superficial temporal artery (STA), both during attacks and during stress; compression of the STA or common carotid artery (CCA) during migraine attacks; and facial skin blood flow measurements. Results of the individual studies are depicted in Table 1.

Figure 1.

Flow diagram of search and article selection for both migraine and rosacea. Modified from Moher et al.⁷¹

Table 1.

Overview of studies on migraine listed according to topic investigated; facial temperature, measurements in the STA, compression of STA/CCA during migraine attacks and facial skin blood flow.

	Author	Year	Participants	Design	Main Findings
	Sargent et al.²⁸	1972	63 M10 TTH2 Cluster headacheData from 13 patients were not analysed for various reasons, not specified from which group	TestsPatients received instructions to focus on increasing temperature in the hands which should in turn help in clinically improving headachesWeekly visits for about 1 month and then every 3 months, up to 2–3 years. Daily exercises at home with the temperature trainerImprovement ratingsHeadache Improvement rated by two psychologists and one internist. Positive response defined as an increase in hand-to-forehead temperature ratioTemperatureMeasured with a ‘temperature trainer’ indicating difference in temperature between mid-forehead and right index finger. No absolute temperatures were recorded. Room temperature is not specified	Improvement in headaches with pre-trainingThe internist and the two psychologists rated headaches to be improved in 68–90% of patientsHand-warming exerciseImprovement in migraine coincided with increase in hand-to-forehead temperature ratio
	Lance et al.¹⁷	1973	10 M investigated interictally12 M during 13 spontaneous attacks6 M during reserpine-induced headache15 HC	ThermographyInfrared radiation from skin measured with an AGA Thermovision cameraRoom temperature is 20–22°CInduced attacksIntramuscular injection of 1.0 mg reserpine-induced headache in 6 M subjects	InterictallySymmetrical heat-loss in 6/10 M, corresponding to the majority of HC3/10 M were warmer on habitual side of headache, and 1/10 was coolerSpontaneous attacksSymmetrical heat-loss in 3/13 attacks. 8/13 were cooler ipsilateral to headache side, and 2/13 were warmer ipsilateral to headacheReserpine-induced headache4/6 showed asymmetry of forehead temperature2/6 were warmer (mild headache) and 2/6 were cooler (severe headache) on headache side
Facial skin temperature	McArthur and Cohen²⁰	1980	8 M	During attack and interictallyTestsBiofeedback sessions consisting of 10 min of rest followed by 20 min of feedback consisting of temperature, frontalis EMG or alpha brain wave feedback condition once every week for 8 weeksTemperatureMeasured with thermistors in the finger and forehead with 1-min intervals for 10 min	TemperatureNo difference between ictal and interictal observations in neither hand nor facial temperatures
	Drummond^18,b	1982	20 MIdentified via questionnaire among first year psychology students, confirmed during migraine-specific interview20 HCAge- and sex-matched	Participants investigated interictallyTemperatureMeasured with thermistorsRoom temperature 24–26°CSkin temperature measured over the frontal branch of the STA at 1-min intervalsTestsCarbon dioxide rebreathingCold pressor testMental arithmeticHead-up tiltRadiant heatIsometric exercise	BaselineSkin temperature is approximately 0.40°C lower in migraine patients contralateral to habitual side of headacheCarbon dioxide rebreathing testSkin temperature increased more in M, especially contralateral to usual side of headache, though still slightly lower temperatures in M compared to HC at the end of the testNo increase in HCCold pressor testDuring instructions for the test, skin temperature contralateral to habitual side of headache decreased more in M compared to HC, but during the actual test, there were no side differences in temperaturesMental arithmeticSkin temperature increased less in M (0.10°C) compared to HC(0.20°C), especially contralateral to habitual side of headacheHead-up tilt, radiant heat, isometric exerciseNo differences between groups in any of these tests
	Drummond and Lance^10,b	1983	57 MWith unilateral migraine50 HCAge- and sex-matched	Ictally, during a unilateral attack in the anterior part of the headThermographyMeasured with ‘AGA thermovision 680 camera’ (infrared)Room temperature 20–22°CFrontal and lateral views of the whole faceTemporal, frontotemporal, supraorbital and orbital temperatures studied for asymmetryAsymmetry was defined as a difference of 0.50°C or more between headache and non-headache side	Temporal temperatureSignificantly more M (37/57) than HC (22/50) had asymmetric heat lossOrbit temperatureSignificant difference in asymmetry between M (25/57) and HC (10/50)Eight patients were warmer ipsilateral to headache and 17 were cooler
Facial skin temperature	Gamble and Elder²²	1983	24 M	InterictallyThermographyMeasured with an ‘Autogen 1000 thermal FB monitor’Room temperature 20–22°CMeasured in the skin of dominant hand and temporal area of the faceTests- Frontalis electromyography biofeedback- Thermal biofeedback- Taped progressive muscular relaxationTests applied in a six-factor mixed design comparing the use of one of the three factors alone or together in subjects. Unclear whether all patients went through all combinations	ThermographyTemporal temperature decreased with thermal biofeedback, but no other tests showed effects on temperatureMultimodal feedback (successful performance of all three modalities had the greatest positive effect measured in hours per week of headache)No effect on hand temperature in either of the tests
	Drummond and Lance^11,b	1984	40 MA82 MO37 Tension vascular27 TTH50 HC	Investigated during 209 spontaneous attacks; some patients were investigated during two or three attacks, and interictallyTemperatureMeasured with AGA Thermovision 680 camera (infrared)Room temperature 20–22°CFrontal (for studying heat-loss from forehead, supraorbital region, orbits and cheeks) and lateral (for studying heat-loss from the temple and lower profile of the face) views of the whole face taken before and after compression of the STA for 15–20 sPainSubjective rating on a verbal rating scale of 1–10Pain relief assessed in percent relief and only >50% pain relief was considered significant.	TemperatureOrbit temperatures during attacks were greater in M compared to HC. Supraorbital, temple, cheeks and lower profile were not different from M to HC. Frontotemporal temperatures were slightly lower in M compared to HCCompression38/202 of attacks responded with headache relief to STA compression.Temperature in the frontotemporal region, investigated in 32 attacks of responders, was significantly warmer ipsilateral to headache (mean 0.20C). STA compression for several minutes would abolish thermographic asymmetry.No thermographic correlation with attacks responding only to CCA compression, however 74% of these patients reported effect of the vasoconstrictor ergotamine tartrate
	Mathew and Alvarez²³	1984	4 MA90 MO69 Mixed35 Cluster23 Nuscle contraction18 Post-traumatic19 HC	53/94 M investigated during attack, and 41/94 M interictallyTemperatureMeasured with an AGA Thermovision 720 camera (infrared)Room temperature is approximately 23°CInfrared thermography performed on forehead bilaterally.	MigraineAsymmetrical temperature pattern in 71% of M, ictally and interictally.Most common asymmetry was hypothermia in the supraorbital area ipsilateral to headache, both during attacks and interictallyIncreasing intensity of pain was correlated with a decrease in temperature in MMuscle contraction headache had opposite effect with pain being correlated with higher forehead temperatureHC78% of HC had symmetrical heat distribution of the face, remaining 22% had less than 1°C difference between sides
	Drummond^19,b	1985	10 MO10 MA10 TTH10 HC	InterictallyThermographyMeasured with thermistorsAssessed in cheeks, 3 cm below eyes and 3 cm from the midlineRoom temperature is 24–26°CTestsMental arithmeticReaction time testParticipants went through the same tests on three separate days with approximately 24 h between each session	BaselineSignificantly higher temperature in the three headache groups(34.10°C) compared to controls(33.50°C). No difference between headache groupsDuring testsResponses did not differ between or within the tests.Facial temperature in HC increased an average of 0.050°C compared to 0.040°C in the headache group. Difference significant
Facial skin temperature	Swerdlow and Dieter²⁴	1986	275 M45 HC	Ictally and interictallyThermographyMeasured with Inframetrics Model 520 (infrared)External carotid area measured with infrared thermography bilaterallyRoom temperature is 20–22°CCold patchAn area of the skin with a temperature 0.50°C lower or more than the surrounding skin	Thermography62.5% of M demonstrated vascular ‘cold patches’ that were present both during headache and interictallyOnly 3/45 HC demonstrated a vascular cold patch
	Swerdlow and Dieter³³	1987	15 M14 Mixed M20 HC	InterictallyThermographyMeasured with Inframetrics Model 520 (infrared)Evaluated in the area of the internal/external carotid territories (region from the bridge of the nose to the top of the forehead and spanning to both temples)Room temperature is 20–22°CTestsHyperoxia	HCResponded with an immediate drop in temperature in response to hyperoxiaHeadache groupsAll patients responded with a significant change during hyperoxia50% of migraine patients and 42.86% of mixed M had increase in temperature. This was present throughout 10-min follow-up.Five migraine patients and one mixed subject continued to elicit this change after follow-up50% of migraine patients and 57% of mixed type demonstrated a drop in temperature in response to hyperoxiaNo difference between headache (migraine vs. mixed) groupsHeadache4/15 M experienced pain relief following hyperoxiaPreventive62% of patients on preventive medication with either beta or calcium blockers had increased facial temperature after hyperoxia, whereas only 35% of patients without preventive treatment had an increased facial temperature (change not significant)
	Falkenstein and Hoormann^34,b	1987	30 M	InterictallyInfrared thermographyMeasured with two thermo elements (Cu/Cu-Constantane) connected to Gould thermocouplersMeasured in the forehead and in the right index fingerRoom temperature not mentionedBiofeedbackSelf-control exercises with and without real-time feedback to reduce BPV of STATen study days of repeated exercises with 2 min breaks in-between and less and less feedback.	ThermographyTemperature was significantly increased in both hand and forehead at the end of each task compared the end of each pause
	Swerdlow and Dieter²⁶	1989	20 MO4 MA12 mixed headaches	InterictallyInfrared thermographyMeasured with Inframetrics Model 520 (infrared)Assessed bilaterally from the bridge of the nose to the top of the forehead and towards both temples‘Cold patch’An area in the face with a temperature that is at least 0.50°C lower compared to the surrounding areasTreatmentPatients were started in preventive therapy of either beta or calcium channel blockerInvestigated before start of preventives and after treatment of 6–15 months	Preventive therapy30/40 (75%) reported beneficial effect from preventive therapy, whereas 10/40 (25%) reported worsening of headaches‘Cold patch’ after 6–15 months of therapy47% with an improvement in headaches had a larger cold patch, whereas 40% of those with worsened headaches showed larger cold patchesIn both groups, 40% had unchanged patterns
Facial skin temperature	Drummond²¹	1990	72 M^a 8 TTH^a	35 interictally and 80 during spontaneous attacks of headache51 M examined during migraine attack21 M examined during attack not fulfilling criteria for migraineTemperatureMeasured with AGA Thermovision 680 camera (infrared)Assessed bilaterally in the upper forehead, supraorbital region (lower 2/3 of forehead), orbits and cheeks	Headache-free periodHeat distribution symmetrical bilaterally, no differences compared to during attacksDuring spontaneous attackNo consistent side difference in facial temperatureHeat-loss from orbital region greater ipsilateral to headache in those with throbbing pain. Upper part of the forehead slightly warmer ipsilateral to headache
	Dalla Volta et al.²⁷	1991	217 MO29 MA	TemperatureMeasured with an Inframetric 525 L electronic thermographic deviceAssessed from the bridge of the nose to the top of the forehead.Room temperature not known, but mentioned to be constant along with humidity.TreatmentParticipants randomized to initiation of preventive therapy with either beta blocker or calcium antagonistInvestigated with thermography before preventive therapy and after 6 months of therapy. Only patients demonstrating a cold area of 0.50°C less than the surrounding areas on initial investigation were studied again. This reduced the patients group from 246 to 206, not mentioned whether these patients had MO or MA	Full clinical recovery85.3% of the patients who reported full clinical recovery of symptoms also showed thermographic improvement with a smaller ‘cold patch’Partial remission52.2% of patients reporting partial recovery showed an improvement in ‘cold patch’No clinical improvement24/206 (11.7%) patients reported no clinical improvement from therapy, of these 83.3% had unchanged thermography, 12.5% had a larger ‘cold patch’ and 1 patient had a smaller ‘cold patch’
	Ford and Ford²⁵	1997	694 MO^a 202 MA^a 35 Cluster headache^a 33 Post-traumatic headache^a 29 Other headaches23 HC	Infrared thermographyMeasured with Bales Scientific Instruments BSI 700 Unit (infrared)Assessed in the whole face including the upper part of the chest and shouldersRoom temperature is 20°CAbnormality criteriaA thermal difference of 0.50°C from surrounding or a homologous region of the face. If difference was found in cheeks, a difference of 10°C was necessary	Migraine593/694 (85.4%) of patients with MO and 180/202 (89.1%) of patients with MA had abnormal thermograms.The most common location for abnormality was hypothermia in the supraorbital area, but abnormality was also found in the orbits, temples and other portions of forehead, cheeks and cervical regions
	Zaproudina et al.²⁹	2013-1	5 MA^a 7 MO^a 29 HC	InterictallyInfrared thermographyMeasured with FLIR A315 digital infrared camera systemRoom temperature is 22–24°CMeasured in the whole face and in the palm of both hands	ThermographyMean skin temperature in M significantly lower (>2°C) on the nose and hands compared to HCSpecific skin temperatures in left-sided and bilateral M did not differ from HCIn right-sided M, however, temperature was higher in the forehead and lower in the fingertips compared to controls and to the other MTemperature gradients between cheek–nose and forehead–nose higher in right-sided M compared to HCUnilateral MO related to lower temperature in the cheek ipsilateral to painThermography – hand50% of M displayed cold hands. No side difference between handsHand temperature inversely correlated with frequency of attacks50% of HC with cold hands had family history of migraine
Facial skin temperature	Zaproudina et al.³⁰	2014	11 HC (hereof 5 HC with a family history of M^a)	Headache-free periodInfrared thermographyMeasured with FLIR A315 digital infrared camera systemRoom temperature is 22–24°CMeasured in the whole faceTestSublingual administration of nitroglycerinThermography performed for 3 h post-administration	HeadacheAll participants experienced initial onset of light headache which quickly resolved but returned after 2.5–3 h in 45.5% (5/11). One patient experienced a worsened headache with nausea and vomiting after 6 hOut of the 5 experiencing headaches, 4/5 had a family history of migraine. In the non-headache group, only 1/6 had a family history of migraineThermographyLower pre-administration skin temperatures were found in participants who experience headache. This was especially prominent in the nose area
	Tunis and Wolff⁴⁴	1953	75 M	Patients studied 2–5 times a week for a period of 5–30 weeks5000 pulse records in totalPulse amplitudeMeasured in the STA	Pulse amplitudeVariability and amplitude of the STA significantly greater in interictal M compared to HC.Increase in variability and amplitude 12–72 h before headache and during the headache phase.Decrease in variability and amplitude (constriction) in the 0–12 h before headache onset with constriction being most prominent immediately before onset of attack.
	Dalessio et al.⁴⁵	1979	10 M5 HC	Pulse amplitudeMeasured in frontotemporal regions (supraorbital and STA) and in the fingerTestAll participants were taught a relaxation technique that should be carried out for 45 min twice a week for 4 weeksPulse amplitude measured at the beginning of experiment and at the end of the sessions	Breathing exercise8/10 experienced improvement in headaches after breathing exercisesPulse amplitudePatients with improved conditions has an increase in finger pulse with a subsequent rise in finger temperature, along with a decrease in pulsation in the frontotemporal regionConstriction of the STA was significant in HC during breathing exercise, when comparing to rest
Measurements in the STA	Drummond and Lance⁴⁰	1981	107 M10 TTH23 HC	Pulse amplitudeMeasured with Narco Bio-Systems Korotkoff sounds microphone and a photoplethysmograph placed over the right STA with a Velcro strap. Microphones were attached with small air bags to ensure that the microphones were held to the head with equal pressureTestChanging positions from sitting to standingLight exerciseSTA measured 30 s before, during standing and 30 s after the exercise test (seated)	Position changingSTA amplitude decreased below baseline in all participants when changing position from sitting to standing. No difference between groupsLight exerciseSTA amplitude increased above baseline in unilateral headache, especially ipsilateral to usual headache side. No side difference in controls, tension type and bilateral migraine
	Drummond^18,b	1982	20 M20 HC	InterictallySTA amplitudeDetected with a Narco Bio-Systems Korotkoff sounds microphone ipsilateral to usual side of headache in patients with unilateral migraine, or right side with bilateral migraineTestsMental arithmeticHead-back tilt	Mental arithmeticSignificant change in amplitude of STA ipsilateral to usual headache in M (increased 2.5 mV) compared to HC (decrease 1.1 mV)Head-back tiltFirst trial: STA amplitude decreased 19% in HC compared to 3% in MSecond trial: STA amplitude decreased approximately 14% in both groups
	Drummond and Lance^10,b	1983	66 M	During attacks of unilateral headachePulsationDetected with Narco Bio-Systems Korotkoff sounds microphone.Bilaterally on STA in front of the ear.Compression testsManual compression of STA/CCA for 15–20 s on each sidePainSelf-evaluated pain increase, no change or decrease Decrease was defined as >50% decrease in pain with ipsilateral compression	STA amplitudeNo difference in mean amplitude between headache side and non-headache side in patients.39/62 with headache confined to one frontotemporal area had increased amplitude ipsilateral to the headache compared to non-headache sideCompressionRelief from compression associated with increased amplitude ipsilateral to headache
Measurements in the STA	Drummond⁴¹	1984-2	30 MDuring spontaneous unilateral headachePatients selected from Drummond 1984-1, and split into groups by(a) effects of STA/CCA compression on headache intensity, and(b) whether heat-loss increased during headache12 patients in prophylactic treatment with propranolol, pizotifen, methysergide, amitriptyline or monoamine oxidase inhibitorsNo prophylactic or acute treatment taken on the day of examination,but 16/30 had taken mediciation within 48 h of examination	Pulse amplitudeRecorded with a Narco Bio-Systems reflectance photopleshysmograph fastened with an adhesive washer. Pulse amplitude amplified by a Type 7173 transducer couplerMeasurement over frontal branch of STA ipsilateral to headacheRoom temperature 20–22°CTestsMental arithmeticLight exercisePatients were evaluated in the following groups(1) patients with signs of extracerebral vascular involvement during headache(2) patients with heatloss from frontotemporal region during headache compared with the remainder(3) patients with extracerebral vascular headache compared to those with intracerebral source of pain according to compression test	Mental arithmetic(1) Increase in STA greatest in this group, amplitude averaging 29% during test compared to 5% in the remainder of patients(2) Increase in STA during test greater in this group compared to the other groups(3) Increase greater during test in patients with extracerebral vascular involvementLight exercise(1) 32% increase in STA compared to 8% in the other groups. No difference before exercise(2) increase in STA after exercise greater in this group
	Drummond^19,b	1985	10 MO10 MA10 TTH10 HC	InterictallyPulse amplitudeDetected by a Narco Bio-Systems Korotkoff sounds microphone and monitored on a Beckman R611 Dynograph. Signal amplified by a Beckman Type 9853A universal coupler.Measured in STA on side habitually affected by headache with unilateral headache, or on the right sideTestsMental arithmeticReaction time test	STA amplitudePeak in amplitude during initial part of test, no difference from one session to another and no difference between mental arithmetic and reaction time testIncrease greater in MA and MO than in TTH, significantly so in MA (3.1 mV) compared to MO (0.4 mV)
	Falkenstein and Hoormann^34,b	1987	30 M	Pulse amplitudeMeasured using a plethysmograph placed over the zygomatic facial branch of STATestSelf-control exercises with and without real-time feedback to reduce BPV of STA10 study days of repeated exercises with less and less feedback	STA amplitudeImprovement in ability to decrease pulse amplitude from baseline with self-control and biofeedback over the course of 10 study days
	Iversen et al.³⁷	1990	25 MUnilateral attack	Ictally and interictallySTA diameterMeasured with high resolution ultrasound (‘Dermascan C’)Assessed bilaterally during attacks approximately 5 h (2–58 h) after onset. Assessor blinded to side of headacheMeasurements repeated after at least 1 week free of migraine attacks	STA diameterNo side differences in diameter of STA interictallyIctally STA diameter decreased (constriction) to 90.6% of interictal diameter contralateral to headache side
Measurements in the STA	Drummond³⁹	1991	38 MO#	InterictallyPulse amplitudeDetected with a Narco Bio-Systems photoplethysmograph attached with adhesive washers bilaterally in the forehead 2 cm above the eyebrows in line with the pupils, and to the cheeks 3 cm below the eyes and 5 cm from the midlineTest (22 interictal MO)Mental arithmeticTest (38 interictal MO)External body heating until sweating on forehead (from 11 min to 73 min, average 44 min)	Mental arithmeticPulse amplitude increased during hard (17%) and extremely hard (13%) tasks No side differenceExternal body heatingIncrease in pulse amplitude of forehead (155–223%) and cheeks (319–346%), especially contralateral (223–346%) to usual headache side. Side difference in cheeks not significantFacial sweating significantly related to increase in STA amplitude contralateral, but not ipsilateral to headache
	Passchier et al.³²	1993	37 MO37 HC	Pulse amplitudeDetected by reflectance photoelectric transducers.Measured in STA and in the finger during testsTestsQuiet readingExamination from patient’s curriculumDuring IQ test	ThermographyMO had significantly higher temperatures during IQ test than HCPulse amplitudeAmplitude of STA smaller in M compared to HC, both during IQ test and examination from curriculum, but not during quiet reading
	Drummond and Granston⁴²	2003	23 M#22 HC	Pulse amplitudeRecorded with an MP100 Biopac Systems Analogue/digital data acquisition systemMeasured in the temples bilaterallyTestPainful stimulation by immersion of non-dominant hand into 20°C ice-water	Pulse amplitudeIncrease in amplitude greater in M than HCIncrease significant contralateral to painful stimulation
Measurements in the STA	Avnon et al.⁴³	2003	30 MO#9 MA#28 HC	Pulse amplitudeDetected by optic pulse sensors at each side of the forehead 1 cm above the eyebrows and 3 cm lateral to the midline, and a third sensor on the palmar side of the middle phalanx of the right middle finger.Response in pulse was assessed with photoplethysmographyTestSoapy drop or sterile saline was administered in each eye at random in a crossover mannerProcedure was repeated to investigate adaptation	Soapy eye dropSTA amplitude increased bilaterally immediately after administration in all migraine patients. Response significantly higher and persisted longer on habitual side of headache compared to contralateral sideNo difference between first and second administration of eye dropNo difference between side of administrationSaline eye dropSimilar vascular response to soapy eye drop, however faster decline in the responseWhen comparing to HC, there was no difference in peak ipsilateral vascular response. However, when dividing patients by pain side, there were significant differences when comparing to controlsBilateral M had the highest peak vascular response, whereas patients with unilateral headache at alternating side had a similar response to controls, and the lowest peak vascular response was found in unilateral migraine
	Graham and Wolff⁴⁶	1938	10 M	During 24 attacksTestsDigital compression of the CCA (both ipsilateral and contralateral)Digital compression of the STA (no specific side reported)Digital compression of the occipital artery (no specific side reported)	Compression of the CCATransient pain relief from ipsilateral compression ipsilateral in all attacks, and headache could be abolished ‘in many’ (no exact number reported)No headache relief by compression on contralateral sideCompression reduced pulsations in the STA ipsilateral to compressionCompression of the STAPressure alleviated pain in the anterior half of the headCompression of the occipital arteryPressure alleviated pain in the posterior half of the head‘Several’ (no exact number reported) participants reported that compression of STA or occipital artery to ‘took a layer off the headache’ but that some pain persisted. Further, ipsilateral pain of the CCA would alleviate pain
Compression of the STA or the CCA
	Blau and Dexter⁴⁷	1981	30 MO20 MA9 TTH2 M not included in analysis as they experienced relief from contralatera but not ipsilateral compression.	IctallyTestsDigital compression of the STA and occipital arteryCircular compression of the head just above eyebrows	Facial colour before tests27/50 were pale, 3/50 flushed and 20/50 had normal complexionTemporal artery compressionPartial pain relief in 16/48, pain increased in 15/48, no effect in 16/48, and 1/50 did not go through with this investigationCircular compression14/48 increase in pain, 21/48 decrease in pain and 12/48 no effect
	Drummond and Lance^10,b	1983	63 M	IctallyPain ratingPain relief of 25%, 50%, 75% or 100% after ipsilateral compression Only significant if >50% pain relief by compressionSTAManual compression both ipsilateral and contralateral to painful side for 15–20 sCCAThe 40 patients who did not experience relief by STA compression were tested with manual compression of the CCA for 15–20 s	Ipsilateral STA compressionNear-complete relief in 11/63 (17.5%) and partial relief in 12/63 (19%)No relief in 40/63 (63.5%)Ipsilateral CCA compressionNear-complete relief in 9/40 (22.5%) and partial relief in 10/40 (25%)No relief in 21/40 (52.5%)
Compression of the STA or the CCA	Drummond and Lance¹¹	1984-1	82 MO40 MA37 tension vascular27 TTH50 HC	During spontaneous attacks of headacheCompressionManual compression of either STA or CCA for 15–20 s, during attack	Compression of STA was generally slightly more effective in relieving headache than compression of CCAMO38% experienced relief from compression of either STA or CCAMA46% experienced relief from compression of either STA or CCACompression38/202 (19%) of attacks (MA/MO/tension vascular) responded with headache relief to STA compression.
Facial skin blood flow	Elkind and Friedman⁹	1964	72 M10 HC	Interictally(18 during spontaneous attacks)Tissue clearanceIntradermal injection of radioactive sodium in the frontotemporal areaMeasured in forehead and forearm	Frontotemporal blood flowBlood flow was not different between interictal M and HCDuring attack, however, blood flow was significantly increased compared to both asymptomatic patients and controls, especially ipsilateral to headacheForearmBlood flow lower than in the foreheadNo change during headache
			6 M	Seven spontaneous attacks in 6 patientsSkin blood flowMeasured before and after intramuscular injection/intravenous infusion of ergotamine tartrate	Headache disappeared in 5/7 after infusion of ergotamine tartrate. Frontotemporal skin flow remained unchanged in the 2/7 who did not experience relief
			25 M	InterictallyInfusion of norepinephrine	NorepinephrineDecrease in frontotemporal skin blood flow in 18/25. No change in 5/25 and increase in 2/25There was a positive correlation between headache intensity and forehead blood flow
			5 M2 HC	InterictallyIntracutaneous administration of bradykinin	BradykininTransient pain at injection site in all participants, but no continuous pain.Cutaneous blood flow increased in 58% of participants
			5 M	InterictallyIntravenous injection of methysergide	MethysergideNo consistent change in blood flow of forehead or forearm. One patient experienced pain relief
	Sakai and Meyer⁴⁹	1978	71 M32 HC	Ictally and interictallySkin blood flowMeasured in the face by the 133XE inhalation method	Skin blood flowIncreased 50% during headache in M compared to HC
Facial skin blood flow	Jensen and Olesen⁵¹	1984	59 MO	Ictally and interictally133Xe injected into the temporal muscle and washout detected externallyTestsTeeth-clenchingchewing	Ictal35% increase in blood flow ipsilateral to headache, 15% contralateralTeeth clenching302% increase in blood flow between attacks, 142% increase during attacks (no side difference)Chewing347% increase in blood flow between attacks, 239% increase during attacks (no side difference)
	Jensen and Olesen⁵²	1985	61 MO	Ictally and interictallyBlood flow133Xe injected into the anterior part of the temporal muscle and washout determined externallyTestsIsometric work (teeth-clenching at 35–50% of maximum strength)	Headache-free periodBlood flow significantly higher on right side compared to the left.During isometric work (headache-free)Significant increase in blood flow (287%)During attackBlood flow tended to be slightly increased compared to when headache-freeDuring isometric work (during attack)Significant increase in blood flow (173%)
	Jensen⁵⁷	1987	21 MA22 MO	Ictally and interictallyBlood flowMeasured in the temporal region by 133Xe washout methodTestHead-up tilt	Blood flowNo difference between washout in MO and MA at any time. No side difference. No difference during attacks compared to interictallyHead-up tiltBlood flow significantly decreased by 27%, no difference between MA/MO, neither during attacks nor in headache-free periodDuring MA attacksDecrease in blood flow significantly higher contralateral (32%) compared to ipsilateral (16%)
	De Marinis et al.⁵⁴	1990	24 M#12 TTH#24 HC	InterictallyBlood flowAssessed by facial flushing assessed in seconds on a scale from 0 to 2, 0 = No flushing, 1 = Mild flushing, 2 = Severe flushing Induction of headache with intravenous infusion of histamine	Histamine infusionHistamine induced headache in all participants. Headache was preceded by a flushing which was more prominent and significantly longer lasting in M compared to the other groups
				Pretreatment with naloxone	Naloxone pretreatmentResulted in slight increase in pain intensity, but no changes in flushing intensity or duration
				Pretreatment with met-enkephalin analogue	Met-enkephalin pretreatmentNo difference in threshold or intensity In M reduction of intensity and duration of the facial flushing
	Zwetsloot et al.⁵⁵	1991	27 MO#4 MA#	Ictally and interictallyBlood flow velocityMeasured with Doppler in:CCAECAExtracranial part of the ICA	MABlood flow velocity lower during attacks compared headache-free periods, no differenceMOSignificantly lower blood flow velocity in CCA ictally compared to interictally. No side difference. No change in the other vessels. Unilateral headache showed no differences between painful and non-painful side
Facial skin blood flow	Caekebeke et al.⁵⁰	1992	47 M#	During unilateral attackSpontaneous migraine attacksTreated with placebo or 1, 2 or 3 mg subcutaneous sumatriptanBlood flow velocityMeasured with an EME TC2-64B transcranial pulsed Doppler system in:ECACCAICAMCAAll measured bilaterally immediately before and 30 min after injection of sumatriptan	PlaceboBlood flow velocity unchanged in CCA, ECA, ICA and nearly significantly decreased in the MCALow-dose sumatriptanBlood flow velocity unchanged in CCA and ECA, however, nearly significantly increased in MCA and significantly increased in ICAHigh-dose sumatriptanBlood flow velocity unchanged in CCA and ECA, but significantly increased in ICA and MCA
	Kara et al.⁵⁶	2003	30 M#31 HC	Ictally and interictallyPulsatility and blood flow velocityMeasured with a 7.5-MHz linear-array colour Doppler sonograph attached to a LOGIQ MD 400 ultrasound scanner in:OAPCACRAVA	IctalMean pulse and resistance was lower in all measured arteries ictally compared to interictally, no differenceInterictalMean pulsation and resistance in PCA and CRA was significantly higher in M bilaterally compared to controls Pulsation and resistance in left VA also significantly higher in M compared to controls, and when comparing M ictal to interictallyBlood flow velocityMean blood flow velocity in all arteries decreased during attack compared to interictally, no difference
	Avnon et al.⁵³	2004	21 MO#8 MA#	InterictallyArterial blood volumeMeasured with photoplethysmography in the forehead skin bilaterally and in the index fingerTestsparticipants were provoked diluted soapy eye drops and saline eye drops, respectively	Forehead vasodilatory responsePatients with left-sided pain had a significantly greater vasodilatory response to soapy eye drop compared to patients with right-sided pain. Response was present bilaterally, but significantly so ipsilateral to usual headache sideDigital blood volumeBlood volume in the index finger decreased immediately after soapy drop instillationNo difference in side of instillation and no difference between response in migraine patients with either right- or left-sided pain
	Park et al.⁵⁸	2009	40 MO#40 HC	AcupunctureNeedles applied to the hands corresponding to the areas for the CCA, the VA and the anterior and posterior circulation of the brain according to the KHT acupuncture principleBlood flow velocityFlow quantification (volume and velocity) performed with MRI (FLASH) both before and after applying the needles	Migraine patientsVolume and velocity of blood flow significantly decreased in the CCA but increased in vertebral arteries (not significant)Control groupVolume and velocity of blood flow decreased significantly in the CCA and only significantly for velocity in the vertebral arteriesChanges of patterns between the MO versus HC were significant on the right side
	Zaproudina et al.³⁸	2013-2	13 MO#25 HC (17/25 had a family history of migraine, 8/25 with no family history)	InterictallyBlood pulsation imagingDigital monochrome cameraRoom temperature is 22–°24CMeasured in the ‘facial area’	Pulse asymmetryNo difference in mean values in amplitude of blood pulsation between groups, however, asymmetry in pulsation was significantly higher in women with MO, and also women with a family history of migraine, compared HC without a family history of migraineBlood perfusionDistribution in MO different compared to HC. Maximum blood flow was reached at different times in the right versus the left cheekIn HC blood perfusion was evenly distributedBlood pulsationsMO and HC with a family history of migraine both had reduced amplitude of blood pulsation in the nose area compared to the cheeks
Facial skin blood flow	Ibrahimi et al.⁵⁹	2015	22 M# with MRM20 HC	Visit 1: Day 19–21 of cycleVisit 2: Day 1–2 of subsequent periodPost-menopausal: Two visits 7–10 days apartBlood flowDermal blood flow measured continoursly with a PeriScan PIM-3 system (Perimed) in the foreheadStimuliForehead skin exposed tonormal saline (electrical stimulation)0.06 mg/mL capsaicin6.0 mg/mL capsaicinProcedure repeated during the second visit	MRMForehead skin blood flow was slightly reduced in response to capsaicin exposure (0.06 mg/mL)Response was similar on days 1–2 compared to days 19–21 of the cycleHealthy volunteersForehead skin blood flow was higher in response to stimuli on days 1–2 of the cycle compared to day 19–21The response in blood flow increase to 0.06 mg/mL capsaicin was significantly higher compared to the response in MRM. Response to 6.0 mg/mL capsaicin was not significantly differentPostmenopausal womenResponse in forehead skin blood flow was similar during visits, and lower compared to both the other groups

M: migraine, unspecified; TTH: tension type headache; STA: superficial temporal artery; HC: healthy control; MA: migraine with aura; MO: migraine without aura; BPV: blood pulse volume; CCA: common carotid artery; ECA: external carotid artery; ICA: internal carotid artery; MCA: middle cerebral artery; OA: ophthalmic artery; PCA: posterior ciliary artery; CRA: central retinal artery; VA: vertebral artery; MRM: menstrually related migraine; IHS: International Headache Society.

^a Diagnosed according to IHS criteria.

^b Study depicted more than once in table.

Rosacea: Search results

The search on rosacea identified 49 publications; 11 of these were relevant for review, investigating temperature and/or blood flow in rosacea. One additional article was found after screening reference lists. Baseline skin temperatures were similar in patients with erythematotelangiectatic rosacea (ETR)^12,13 and papulopustular rosacea (PPR)¹² compared to controls. Blood flow in lesioned skin, however, was higher compared to the blood flow in the skin of healthy controls (HCs).^12,14

Discussion

Migraine

Facial temperature

Facial skin temperature may be used as a measure of facial skin perfusion, with changes in temperature being directly proportional to changes in perfusion.¹⁵ The assessment of facial skin temperature through infrared thermography was first introduced to study cerebrovascular disorders in 1965,¹⁶ and has since evolved into a very accurate method for obtaining absolute skin temperatures.¹⁵ Some studies included in this analysis were made with thermistors, temperature trainers, thermocouplers or older versions of infrared temperature equipment which leaves comparisons susceptible to differences in sensitivity in measurement variability.

A study of 161 healthy volunteers determined facial temperatures according to region from highest temperature (medial palpebral commissure) to the lowest (nasolabial area). Males generally had slightly higher facial skin temperatures than females.¹⁵

Our systematic search yielded 18 distinct studies investigating skin temperatures in migraine patients. The studies comprised 1990 patients with migraine (both migraine with aura (MA) and migraine without aura (MO)), 362 patients with other kinds of headache (tension type, tension-vascular, ‘mixed’ and cluster headache) and 281 HCs amalgamated. Patients were investigated both during and between headache attacks, and also during different physically and psychologically stressful situations. Higher temperatures ipsilateral to the usual side of headache were reported in the forehead^17,18 of interictal patients in two studies. In one of these studies, however, the pain side specificity was confined to only 3 out of 10 subjects.¹⁷ A tendency of slightly higher temperatures in the bilateral forehead and cheeks were reported in headache-free patients compared to during attacks, especially ipsilateral to the side of pain.¹⁹ Two studies comparing ictal and interictal observations found no difference in temperatures during or between attacks.^20,21

One study, comparing patients during spontaneous attacks to controls, reported no temperature difference in the orbits of the majority of patients, but found orbits to be cooler in 30% and warmer in 14% of the patients compared to the non-headache side,¹⁰ which corresponds to the previously reported vasoconstriction of the ophthalmic artery in some and vasodilation in others.¹⁰ Orbit temperatures have been found warmer during headache compared to controls, whereas they were colder during headache-free periods.¹¹ Different findings in the above-mentioned studies suggest a patient heterogeneity or variability in measurement capabilities between studies. In one study, no temperature difference was found, but after dividing migraine attacks into groups based on response to compression of the STA for pain relief, 32 attacks that were alleviated with compression were associated with significantly warmer foreheads on the headache side. Further compression for several minutes was associated with a decrease in forehead temperature.¹¹ The authors suggested that pain was associated with higher forehead blood flow and thus temperatures, and that a decrease in blood flow could relieve pain, but it was not reported how many patients these STA-sensitive attacks were spread across. Interestingly, thermal biofeedback in another study resulted in cooling of the temples and headache relief.²²

Hypothermia in the supraorbital area during headache was reported ipsilateral to the pain side,^11,23 and the intensity of pain was positively correlated to a decrease in forehead temperatures,²³ suggesting sympathetic constriction of forehead vessels in response to pain. Two cohort studies^24,25 investigated interictal patients by infrared thermography revealing a characteristic ‘cold patch’ (an area with a temperature that was ≥0.5°C colder than the surrounding skin area) in the forehead of 85.4% of MO, 89.1% MA and 62.5% of patients with vascular headaches (MA, MO and atypical migraines pooled).²⁵ The ‘cold patch’ was further investigated before and after preventive treatment with beta or calcium channel blockers.^26,27 One study found that the majority of patients experiencing clinical improvement from preventive treatment had smaller ‘cold patches’ compared to the group with no clinical improvement.²⁷ This finding contradicted an earlier study where some patients demonstrated a larger ‘cold patch’ despite clinical improvement from preventive treatment with beta or calcium channel blockers.²⁶ A limitation of these studies was that there was no placebo control and investigators were not blinded to clinical improvement. Furthermore, previous studies had shown that if the face was generally cool, the patch would grow larger, suggesting that the ‘cold patch’ may not be a sign of an area with poor circulation, but rather a sign of generally lower forehead temperatures in patients compared to controls, consistent with previous reports.^18,23

In two studies, temperatures in the hands of headache patients were reported to be lower compared to controls.^28,29 One of these studies reported that hand temperatures were inversely related to number of attacks and that patients had colder noses compared to controls,²⁹ and the other suggested that increasing the hand-to-forehead temperature ratio might provide headache relief.²⁸ Interestingly, the cold nose and hands were also present in some healthy volunteers, of which 50% presented with a family history of migraine.³⁰ Sublingual administration of glyceryl trinitrate induced headache in 5 out of 11 healthy volunteers, 4 of whom had a family history of migraine.³⁰ These volunteers had lower temperatures in the nasal area compared to the participants who did not report headache, suggesting the presence of an autonomic dysfunction in headache-prone subjects, possibly a sympathetic vasoconstriction or a dysfunction of the parasympathetic vasodilatory function.²⁹ This may also explain the previously reported association between migraine and Raynaud’s phenomenon.³¹

Psychological stress experiments (mental arithmetic, IQ test, exams) were applied in five separate studies. One study compared 37 MO to 37 HCs and found foreheads of MO to be warmer during experiments.³² In two studies, however, patients had cooler foreheads or slightly increased temperature compared to controls,^18,19 suggesting a dysfunction in the response to stress. Hyperoxia in headache patients would increase the temperature in the hand and forehead of 50% of the patients. Interestingly, there was a lower incidence (33%) of temperature increase in the patients who were currently on preventive treatment.³³ An attempt to use biofeedback in the control of STA pulsation would increase the forehead temperature significantly,³⁴ suggesting inherently low interictal temperatures in migraine patients. However, results should be cautiously inferred as they were not the focus of the study or part of the a priori hypotheses. A head-back tilt test induced ‘red and puffy faces’ in patients, most likely due to a vasoconstrictor reflex.¹⁸

Despite the vast amount of research in the facial skin temperature of migraine patients, results are equivocal, likely reflecting the use of different methods, the heterogeneity of patient populations and the fact that changes may be discrete and difficult to measure. There is a tendency towards lower temperatures in the nose and hands, and slightly higher temperatures in the forehead of headache patients compared to controls, both during and between attacks. Interestingly, the cold nose was also reported in some healthy volunteers, especially in those with a family history of headache. It is plausible to suggest a genetic disposition to an altered vascular response in these individuals, which may or may not result in headaches, depending on susceptibility. The altered temperature pattern in patients has been suggested to be the result of either sympathetic constriction of facial blood vessels or possibly under-responsiveness to vasodilation during stress.¹⁸

When comparing the studies, some limitations should be acknowledged. The majority of the studies were conducted before the introduction of the International Headache Classification³⁵ and investigated ‘vascular’ headaches which are not quite well-defined. Therefore, interpretation of reported data is difficult. In most studies, temperature measurements were focused on the frontotemporal area, the cheeks and/or the nose. However, some studies failed to differentiate between sides or the facial area in which the temperature was measured. This also complicates direct comparison, as intra-individual differences in the facial temperature have also been reported in HCs.¹⁵ Additionally, the method in the infrared thermography recordings was not clearly described in all studies, and there were discrepancies in the room temperature (range 20–25°C) during investigations, which is a common limitation in facial temperature studies.³⁶

Measurements of the STA

The STA is one of the terminal branches of the external carotid artery (ECA) which supplies the area near the orbit, along with the muscles and skin of the forehead via the anterior temporal, zygomatic and transverse facial arteries. A change in pulsation of the STA therefore entails a change in the blood flow supplying the orbit and forehead. Pulse amplitude of the STA was investigated in 12 separate studies comprising 520 migraine patients (both MO and MA), 57 patients with other types of headache and 108 HCs. Investigations were made both ictally and interictally in different stressful settings.

One study measured the diameter of the STA in 25 patients during and between migraine attacks. There was no side difference in diameter between attacks, but the STA on the headache side was larger during attacks than on the non-headache side. Further analyses revealed that the headache side was unchanged during attacks and the non-headache side was constricted. This was based on baseline measurements from another day leaving the post hoc analyses vulnerable to day-to-day variations.³⁷

Pulsation of the STA was found to be increased ipsilateral to usual side of headache both during attacks¹⁰ and interictally.³⁸ In line with this, exposing patients to psychological stress (mental arithmetic),^18,19,39 light exercise,^40,41 painful stimulation with ice-water on non-dominant hand⁴² or stimulation of the ophthalmic division of the trigeminal nerve with a soapy eye drop⁴³ would increase pulsation in the STA ipsilateral to usual side of headache. This suggests a more readily response to stressful stimulations. Interestingly, one study measured STA pulsation in a large population during a longer period of time and found that asymmetry in pulsation was present both during and between headaches and that variability became more pronounced in the days preceding an onset of headache.⁴⁴

The effect of biofeedback sessions on headache was tested in two separate studies. One study found that regular use of a relaxation technique reduced headache frequency, increased pulsation in the finger and decreased frontotemporal pulsation.⁴⁵ Another study found unchanged STA pulsation after biofeedback in migraine patients but reported a positive effect on the headache.²⁰

Collectively, data suggested asymmetrical reactivity in the STA in headache patients with changes in pulsation amplitude related to headache phase. Furthermore, long-term relaxation exercises were reported to have a possible beneficial effect on pulsation asymmetry, also resulting in a decrease in headache pain. A reason for this may be the resemblance of these exercises to meditation, resulting in decreased stress levels and thus less stress-related headaches. This is in line with clinical reports from migraine patients experiencing worsening of their migraine during stressful periods. Interestingly, despite the difference in STA pulsation in migraine patients compared to controls, it seems that cardiac responses are similar to those of HCs¹⁸ and the changes in STA pulsation may thus be seen as local changes in the vessels of the face and scalp as opposed to a general vasomotor instability in migraine patients.

The method for recording pulse amplitude is a well-recognized and sensitive method.⁴³ A limitation of pulsation measurement is, however, that the amplitude of STA can vary with position and pressure of the transducer, and even small movements can obscure results. In some of the studies, investigators tried to avoid this by placing the Korotkoff sounds microphone in a pressure-adjustable band.^10,18,19

Compression of STA/CCA during migraine attacks

Numerous methods exist for testing vascular distension and compression in the cranial region of migraine patients. The purpose of compression studies is to test whether flow in the arteries modulates migraine pain. The STA originates from the CCA, and alleviation of migraine pain from compression of either STA or the CCA, has been investigated in a total of 243 migraine patients (both MO and MA) and 73 patients with other headaches.^10,11,46,47

One study found that compression of the CCA could alleviate pain in 100% of investigated attacks and that some attacks could even be temporarily abolished.⁴⁶ Compression of the STA or the occipital artery, however, would only result in relief in the anterior or the posterior half of the head, respectively. A more recent study found compression of the STA during 202 attacks in 186 patients to be more effective in alleviating head pain than compression of the CCA.¹¹ Circular compression of the head was tested in one study,⁴⁷ resulting in alleviation of pain in 42% of patients, whereas 28% felt a worsening in pain and 24% experienced no effect.

Collectively, compression may result in headache alleviation in up to 33% of patients, with compression of the STA being the most effective. However, inconsistency of results and large inter- and intra-individual differences should be acknowledged as some patients may experience relief by compression of STA in one attack, and relief by compression of CCA, or no relief at all, in another attack. Most studies applied manual compression which may have varied in the intensity of the pressure across studies and studies lacked a reliable placebo/sham treatment option. Results should therefore be cautiously interpreted.

Facial skin blood flow

Provocation experiments in migraine patients using infusion of vasoactive neuropeptides such as pituitary adenylate cyclase-activating polypeptide suggested an association between flushing and immediate headache in up to 100% of patients.⁴⁸

Fifteen studies comprising a total of 542 migraine patients, 182 HCs and 12 with tension-type headache investigated facial skin blood flow both during and between attacks. An increased facial skin blood flow was found ipsilateral to headache,^9,49 which was normalized after treatment with ergotamine tartrate.⁹ A subcutaneous injection of sumatriptan (3 mg) during spontaneous attacks did not have any effect on blood flow velocity in the ECA.⁵⁰ Teeth clenching or chewing resulted in an increased temporal muscle blood flow which was about doubled interictally compared to ictally.^51,52 This indicates that blood flow is already increased during attacks. Provocation with soapy eye drops in interictal patients would increase vasodilation in the forehead, especially ipsilateral to usual headache side, which was particularly prominent in patients who usually had migraines on the left side.⁵³

Sustained flushing during histamine-induced headache could be partly antagonized by pretreatment with met-enkephalin analogue.⁵⁴

Ictal blood flow was increased in four studies^9,49,51,52 compared to the interictal period. In one of these studies, blood flow was increased ipsilateral to headache,⁵¹ two studies reported decreased blood flow velocity during attacks,^55,56 and one study found no difference between ictal and interictal blood flow.⁵⁷ Acupuncture has been found to decrease blood flow velocity in the CCA in migraine patients (no sham control),⁵⁸ and a decrease in blood flow in response to local capsaicin was also found in patients with menstrually related migraine compared to both HCs and postmenopausal women.⁵⁹ Blood flow was only measured interictally and remained unchanged during the whole menstrual cycle.

Taken together, these studies suggested increased blood flow in the large arteries of the face during attacks. This is consistent with findings of an increase in ictal forehead temperature due to vasodilation and supports the hypothesis of temperature being proportionate to blood flow. Using facial thermography, it is possible to investigate the relationship between blood flow velocity in the feeding arteries and microvascular changes in the skin after sumatriptan. It should be noticed that studies on facial skin blood flow included small populations and some conducted before the International Headache Society classification, making it difficult to compare groups across studies. Blood flow is a function of arterial cross-section and intravascular blood flow velocity, and when only one of these variables is measured, it is not possible to conclude whether changes in, for example, blood flow are a result of changes in velocity, arterial dimensions or both. Further studies are needed to investigate blood flow in a standardized manner in larger populations.

Rosacea

The systematic search on rosacea uncovered 12 studies. Rosacea flushing could be induced by ingestion of alcohol^60
–62 and hot beverages,⁶¹ whereas cold beverages (cold coffee, fizzy water and caffeine tablets) failed to induce flushing. The flushing induced by alcohol and hot beverages was accompanied by an increase in temperature which reached a maximum faster for hot beverages⁶¹ than for alcohol.⁶⁰ Further, alcohol-induced flushing was shown to be partly inhibited by clonidine, methysergide and indomethacin.⁶² Exposure to heat^13,63,64 or cold⁶⁴ could also induce facial flushing which was accompanied by a temperature increase that was more rapid in rosacea patients, but otherwise similar to that of controls.¹³ No differences in skin temperatures were found in skin affected by erythematotelangiectatic rosacea (ETR) or papulopustular rosacea (PPR) compared to HC.¹²

In one study, baseline skin blood flow was statistically equivalent for rosacea patients and controls. Iontophoresis of acetylcholine induced a similar vasodilatory response, however, the surrounding skin area of the iontophoresis reacted with an increase in blood flow in patients, but not in controls.⁶⁵ In addition, patients with severe rosacea had a slightly higher blood flow compared to controls.⁶⁶ In these patients, physical and psychological stress induced a similar blood flow increase in patients and controls, however, blood flow in patients with severe rosacea remained slightly higher.⁶⁶ Skin blood flow assessed by laser Doppler flowmetry was found to be increased 3- to 4-fold in rosacea patients compared to controls in one study.¹⁴ In patients with PPR, blood flow in lesioned skin was higher in comparison to non-lesioned skin.¹² Treatment of rosacea skin with intense pulsed light resulted in an average decrease in blood flow by 30%⁶⁷ Angiogenesis, increased microvessel density with oedema and morphological changes were present more often in lesioned skin compared to non-lesioned skin.⁶⁸ Interestingly, an age-related decline in cutaneous vasoconstriction has been reported in healthy volunteers,⁶⁹ which might contribute to rosacea symptoms in the middle-aged.

Collectively, the studies on rosacea suggest altered blood flow and temperature in patients compared to controls, particularly in lesioned skin. Rosacea flushing can be induced by the ingestion of alcohol and hot beverages, prolonged exposure of the skin to cold or heat and physical and psychological stress. The observation that intake of hot beverages induces a more rapid flushing compared to use of alcohol⁶¹ suggests that the initiation pathway by the two different triggers is somewhat different, and that intraoral heating in itself may initiate flushing, and not the coffee, as caffeine tablets failed to induce flushing.

Interestingly, in one study on migraine patients, intraoral (maxillary) chilling had a beneficial effect on nausea and pain intensity.⁷⁰ To our knowledge, this has not been systematically tested as a treatment in rosacea patients.

It seems there is an augmented sympathetic nerve activity found in forehead skin in rosacea patients during stressful tasks which may be a sign of autonomic dysregulation in rosacea. Similar findings of autonomic dysregulation have been reported in migraine.⁶ Given that migraine and rosacea are highly associated in epidemiological studies,⁵ it is important to investigate shared molecular mechanisms to understand a significant overlap between the two disorders.

Future perspectives and conclusions

In the reviewed articles, an altered temperature pattern was found in migraine patients who generally showed lower temperature in the forehead, nose and hands. STA was shown to be more susceptible to both physiological and psychological changes in migraine patients compared to controls, and compression of either STA or CCA during attacks may lead to transient pain relief. Blood flow was also found to be asymmetrical in migraine patients, which is consistent with the asymmetrical findings in temperature, indicating an altered blood flow in the face of migraine patients. Collectively, the findings suggest that migraine patients may exhibit dysfunction of the autonomic functions of the extracranial circulation. It is, however, not quite clear whether this is due to sympathetic or parasympathetic dysregulation. It has been suggested that both systems are active, but that the greatest variability may lie in the sympathetic nervous system which seems downregulated between attacks and, maybe due to hypersensitivity, upregulated during attacks.¹ The pathophysiology of rosacea is not better understood, but, as in migraine, a genetic component is believed to play a role along with a dysregulation of both immune, vascular and nervous systems.⁷ Further studies are needed to investigate pathophysiology in both diseases and may lead towards acknowledging, understanding and possibly treating a somewhat overlooked symptom in migraine.

Clinical implications

Extracranial vasomotor reactivity in the face is altered in migraine and rosacea.

Altered vasomotor reactivity may reflect dysfunction of the autonomic nervous system.

Better understanding of links between migraine and rosacea may encourage future development of novel therapeutic options in both diseases.

Footnotes

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Dr Wienholtz reports no disclosures relevant to the manuscript. Dr Christensen reports grant from Lundbeck Foundation (R249-2017-1608). Dr Egeberg has received research funding from Pfizer, Eli Lilly, the Danish National Psoriasis Foundation and the Kgl Hofbundtmager Aage Bang Foundation, and honoraria as consultant and/or speaker from Almirall, Leo Pharma, Samsung Bioepis Co., Ltd, Pfizer, Eli Lilly, Novartis, Galderma and Janssen Pharmaceuticals. Dr Thyssen has received a speaker fee from Galderma on two occasions. Dr Ashina is principal investigator for the following clinical trials: Amgen 20120178 (Phase 2), 20120295 (Phase 2), 20130255 (Open label extension), 20120297 (Phase 3), 20150308 (Phase 2), ElectroCore GM-11 gamma-Core-R, TEVATV48125-CNS-30068 (Phase 3), Novartis CAMG334A2301(Phase 3) and Alder PROMISE-2. Dr Ashina has no ownership interest and does not own stocks of any pharmaceutical company. Dr. Ashina reports personal fees from Alder BioPharmaceuticals, Allergan, Amgen, Alder, Eli Lilly, Novartis and Teva. Dr Ashina serves as an associated editor of Cephalalgia and a coeditor of the Journal of Headache and Pain.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded by grants from the Novo Nordisk Foundation (NNF17OC0029698), the Augustinus Foundation and the Lundbeck Foundation (R155-2014-171).

References

Miglis

. Migraine and autonomic dysfunction: Which is the horse and which is the jockey? Curr Pain Headache Rep 2018; 22: 19.

Harvey

. Neuropsychiatry: migraine-a vascular neurosis. Cal West Med 1929; 30: 54.

Roberts

Tanner

. Vascular response in migraine. Br Med J 1964; 1: 195–196.

Möllendorff

. Ueber hemikranie. Arch Pathol 1867; 41: 385–395.

Egeberg

Ashina

Gaist

. Prevalence and risk of migraine in patients with rosacea: a population-based cohort study. J Am Acad Dermatol 2017; 76: 1–5.

Christensen

Andersen

Wienholtz

. The relationship between migraine and rosacea: systematic review and meta-analysis. Cephalalgia 2017; 38: 1387–1398. DOI: 033310241773177.

Steinhoff

Schauber

Leyden

. New insights into rosacea pathophysiology: a review of recent findings. J Am Acad Dermatol 2013; 69: 15–26.

Petersen

Lassen

Birk

. BIBN4096BS antagonizes human alfa-calcitonin gene related peptide-induced headache and extracerebral artery dilatation. Clin Pharmacol Ther 2005; 77: 202–213.

Elkind

Friedman

. Cutaneous blood flow in vascular headaches of the migraine type. Neurology 1964; 14: 24–30.

10.

Drummond

Lance

. Extracranial vascular changes and the source of pain in migraine headache. Ann Neurol 1983; 13: 32–37.

11.

Drummond

Lance

. Facial temperature in migraine, tension-vascular and tension headache. Cephalalgia 1984; 4: 149–158.

12.

Guzman-Sanchez

Ishiuji

Patel

. Enhanced skin blood flow and sensitivity to noxious heat stimuli in papulopustular rosacea. J Am Acad Dermatol 2007; 57: 800–805.

13.

Metzler-Wilson

Toma

Sammons

. Augmented supraorbital skin sympathetic nerve activity responses to symptom trigger events in rosacea patients. J Neurophysiol 2015; 114: 1530–1537.

14.

Sibenge

Gawkrodger

. Rosacea: a study of clinical patterns, blood flow, and the role of demodex folliculorum. J Am Acad Dermatol 1992; 26: 590–593.

15.

Haddad

Brioschi

Baladi

. A new evaluation of heat distribution on facial skin surface by infrared thermography. Dentomaxillofac Radiol 2016; 45: 1–10.

16.

Wood

. Thermography in the diagnosis of cerebrovascular disease. Radiology 1965; 85: 270–283.

17.

Lance

Anthony

Somerville

. Facial thermography in cerebral vascular insufficiency and migraine. Proc Aust Assoc Neurol 1973; 9: 31–39.

18.

Drummond

. Extracranial and cardiovascular reactivity migrainous subjects. J Psychosom Res 1982; 26: 317–331.

19.

Drummond

. Vascular responses in headache-prone subjects during stress. Biol Psychol 1985; 21: 11–25.

20.

McArthur

Cohen

. Measures of forehead and finger temperature, frontalis EMG, heart rate and finger pulse amplitude during and between migraine headaches. Headache J Head Face Pain 1980; 20: 134–136.

21.

Drummond

. Disturbances in ocular sympathetic function and facial blood flow in unilateral migraine headache. J Neurol Neurosurg Psychiatry 1990; 53: 121–125.

22.

Gamble

Elder

. Multimodal biofeedback in the treatment of migraine. Biofeedback Self Regul 1983; 8: 383–392.

23.

Mathew

Alvarez

. The usefulness of thermography in headache. Prog Migraine Res F Clofford Rose, FRSP 1984, Pittman, London 1984; 232–45.

24.

Swerdlow

Dieter

. The validity of the vascular “cold patch” in the diagnosis of chronic headache. Headache 1986; 26: 22–26.

25.

Ford

. Thermography in the diagnosis of headache. Semin Neurol 1997; 17: 343–349.

26.

Swerdlow

Dieter

. The vascular ‘cold patch’ is not a prognostic index for headache. Headache 1989; 29: 562–568.

27.

Dalla Volta

Anzola

DiMonda

. The disappearance of the ‘cold patch’ in recovered migraine patients: thermographic findings. Headache 1991; 31: 305–309.

28.

Sargent

Green

Walthers

. The use of autogenic feedback training in a pilot study of migraine and tension headaches. Headache 1972; 12: 120–124.

29.

Zaproudina

Lipponen

Karjalainen

. Acral coldness in migraineurs. Auton Neurosci 2014; 180: 70–73.

30.

Zaproudina

Närhi

Lipponen

. Nitroglycerin-induced changes in facial skin temperature: ‘cold nose’ as a predictor of headache? Clin Physiol Funct Imaging 2013; 33: 409–417.

31.

Cavestro

Ferrero

. Migraine in systemic autoimmune diseases. Endocr Metab Immune Disord Drug Targets 2018; 18: 124–134.

32.

Passchier

Goudswaard

Orlebeke

. Abnormal extracranial vasomotor response in migraine sufferers to real-life stress. J Psychosom Res 1993; 37: 405–414.

33.

Swerdlow

Dieter

. The thermographically observed effects of hyperoxia on vascular headache patients and non-headache individuals. Headache 1987; 27: 533–539.

34.

Falkenstein

Hoormann

. Psychophysiological correlates of vasomotor self control mediated by biofeedback. J Psychophysiol 1987; 4: 383–392.

35.

Olesen

Bendtsen

Dodick

. Headache classification committee of the International Headache Society (IHS) the international classification of headache disorders, 3rd edition. Cephalalgia 2018; 38: 1–211.

36.

Zaproudina

Airaksinen

Närhi

. Are the infrared thermography findings skin temperature-dependent? A study on neck pain patients. Ski Res Technol 2012; 19: 537–545.

37.

Iversen

Nielsen

Olesen

. Arterial responses during migraine headache. Lancet 1990; 6: 837–839.

38.

Zaproudina

Teplov

Nippolainen

. Asynchronicity of facial blood perfusion in migraine. PLoS One 2013; 8: e80189.

39.

Drummond

. Effects of body heating and mental arithmetic on facial sweating and blood flow in unilateral migraine headache. Psychophysiology 1991; 28: 172–177.

40.

Drummond

Lance

. Extracranial vascular reactivity in migraine and tension headache. Cephalalgia 1981; 1: 149–155.

41.

Drummond

. Extracranial vascular changes during headache, exercise and stress. J Psychosom Res 1984; 28: 133–138.

42.

Drummond

Granston

. Facilitation of extracranial vasodilatation to limb pain in migraine sufferers. Neurology 2003; 61: 60–63.

43.

Avnon

Nitzan

Sprecher

. Different patterns of parasympathetic activation in uni- and bilateral migraineurs. Brain 2003; 126: 1660–1670.

44.

Tunis

Wolff

. Long-term observations of the reactivity of the cranial arteries in subjects with vascular headache of the migraine type. AMA Arch NeurPsych 1953; 70: 551–557.

45.

Dalessio

Kunzel

Sternbach

. Conditioned adaptation-relaxation reflex in migraine therapy. Jama 1979; 242: 2102–2104.

46.

Graham

Wolff

. Mechanism of migraine headache and action of ergotamine tartrate. Arch NeurPsych 1938; 39: 737–763.

47.

Blau

Dexter

. The site of pain origin during migraine attacks. Cephalalgia 1981; 1: 143–147.

48.

Amin

Hougaard

Schytz

. Investigation of the pathophysiological mechanisms of migraine attacks induced by pituitary adenylate cyclase-activating polypeptide-38. Brain 2014; 137: 779–794.

49.

Sakai

Meyer

. Regional cerebral hemodynamics during migraine and cluster headaches measured by 133Xe inhalation. Headache 1978; 18: 122–132.

50.

Caekebeke

JVF

Ferrari

Zwetsloot

. Antimigraine drug sumatriptan increases blood flow velocity in large cerebral arteries during migraine attacks. Neurology 1992; 42: 1522–1526.

51.

Jensen

Olesen

. Common migraine and temporal muscle blood flow. Acta Neurol Scand Suppl 1984; 69: 254–255.

52.

Jensen

Olesen

. Temporal muscle blood flow in common migraine. Acta Neurol Scand 1985; 72: 561–570.

53.

Avnon

Nitzan

Sprecher

. Autonomic asymmetry in migraine: Augmented parasympathetic activation in left unilateral migraineurs. Brain 2004; 127: 2099–2108.

54.

De Marinis

Feliciani

Janiri

. Increased reactivity to a met-enkephalin analogue in the control of autonomic responses in migraine patients. Clin Neuropharmacol 1990; 13: 507–521.

55.

Zwetsloot

Caekebeke

Jansen

. Blood flow velocity changes in migraine attacks-a transcranial Doppler study. Cephalalgia 1991; 11: 103–107.

56.

Kara

Erdemoĝlu

Karadeniz

. Color Doppler sonography of orbital and vertebral arteries in migraineurs without aura. J Clin Ultrasound 2003; 31: 308–314.

57.

Jensen

. Subcutaneous blood flow in the temporal region of migraine patients. Acta Neurol Scand 1987; 75: 310–318.

58.

Park

Kim

Baek

. Effect of acupuncture on blood flow velocity and volume in common carotid and vertebral arteries in migraine patients. Med Acupunct 2009; 21: 47–54.

59.

Ibrahimi

van Oosterhout

WPJ

van Dorp

. Reduced trigeminovascular cyclicity in patients with menstrually related migraine. Neurology 2015; 84: 125–131.

60.

Parodi

Guarrera

Rebora

. Dermatological flushing in rosacea: an experimental approach. Arch Dermatol Res 1980; 273: 269–273.

61.

Wilkin

. Oral thermal-induced flushing in erythematotelangiectatic rosacea. J Invest Dermatol 1981; 76: 15–18.

62.

Guarrera

Parodi

Cipriani

. Flushing in rosacea: a possible mechanism. Arch Dermatol Res 1982; 272: 311–316.

63.

Brinnel

Friedel

Caputa

. Rosacea: Disturbed defense against brain overheating. Arch Dermatol Res 1989; 281: 66–72.

64.

Haxthausen

. Changes in skin vessels from protracted action of climatic factors and their significance in various skin diseases. Br J Dermatol 1930; 42: 105–125.

65.

Drummond

. Endothelial and axon reflex vasodilatation to acetylcholine in rosacea-affected skin. Arch Dermatol Res 2012; 304: 133–137.

66.

Drummond

. Blushing in rosacea sufferers. J Psychosom Res 2012; 72: 153–158.

67.

Mark

Sparacio

Voigt

. Objective and quantitative improvement of rosacea-associated erythema after intense pulsed light treatment. Dermatologic Surg 2003; 29: 600–604.

68.

Aroni

Tsagroni

Kavantzas

. A study of the pathogenesis of rosacea: How angiogenesis and mast cells may participate in a complex multifactorial process. Arch Dermatol Res 2008; 300: 125–131.

69.

Holowatz

Thompson-Torgerson

Kenney

. Aging and the control of human skin blood flow. Front Biosci (Landmark Ed) 2010; 15: 718–739.

70.

Friedman

Peterson

Behar

. Intraoral chilling versus oral sumatriptan for acute migraine. Hear Dis 2001; 3: 357–361.

71.

Moher

Liberati

Tetzlaff

. The PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009; 6(7): e1000097.