Original Article
Medication overuse reinstates conditioned pain modulation in women with migraine
Cephalalgia 0(0) 1–11 ! International Headache Society 2017 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0333102417727545 journals.sagepub.com/home/cep
Nathalie Guy1,2,3, Daniel Voisin4,5, Aure´lien Mulliez2, Pierre Clavelou1,2,3 and Radhouane Dallel1,2,3 Abstract Background: This study investigated the effects of medication overuse and withdrawal on modulation of pain processing in women with migraine. Temporal summation of laser-evoked thermal pain was used to measure the effects of conditioned pain modulation. Methods: 36 female participants (12 healthy volunteers, 12 with episodic migraine and 12 with medication overuse headache) were included in a two session protocol. Medication overuse headache subjects were also tested three weeks after medication overuse headache withdrawal. Mechanical and laser-evoked thermal pain thresholds were measured on the back of the non-dominant hand where, later, temporal summation of laser-evoked thermal pain to repetitive thermal stimuli was elicited for 30 min, at an intensity producing moderate pain. Between the 10th and 20th minutes, the contralateral foot was immersed into a water bath at a not painful (30 C) or painfully cold (8 C; conditioned pain modulation) temperature. Results: Episodic migraine, medication overuse headache and medication overuse headache withdrawal were associated with an increase in extracephalic temporal summation of laser-evoked thermal pain as compared to healthy volunteer subjects, while there was no alteration of laser-evoked thermal and mechanical extracephalic pain thresholds in these subjects. Conditioned pain modulation was highly efficient in temporal summation of laser-evoked thermal pain in healthy volunteer subjects, with a solid post-effect (reduction of pain). Conditioned pain modulation was still present, but reduced, in episodic migraine. By contrast, conditioned pain modulation was normal in medication overuse headache and strongly reduced in medication overuse headache withdrawal. Furthermore, in medication overuse headache withdrawal, the post-effect was no longer a decrease, but a facilitation of pain. Conclusions: These data show that a decrease in conditioned pain modulation does not underlie medication overuse headache in women. On the contrary, medication overuse reinstated conditioned pain modulation in female migraine patients. They also identify different phenotypes of pain modulation in migraine patients. Registration number N 2008-A00471-54. Keywords Migraine, medication overuse headache, conditioned pain modulation, pain summation Date received: 4 May 2017; revised: 15 June 2017; accepted: 4 July 2017
Introduction Migraine is a common and disabling neurological disorder that manifests as a recurrent, episodic, often severe, headache. Among episodic migraineurs, a subset develops chronic migraine, a very invalidating condition characterized by headaches for more than 15 days a month. The most prominent risk factors that can promote chronic migraine include the frequency of attacks, psychiatric co-morbidities such as anxiety, depression, addictive behavior, and medication overuse (1). Indeed, the regular overuse for more than three months of one or more of the
1
Universite´ Clermont Auvergne, Clermont-Ferrand, France CHU Clermont-Ferrand, Clermont-Ferrand, France 3 Inserm U1107, Neuro-Dol, Clermont-Ferrand, France 4 Neurocentre Magendie, Inserm U1215, Bordeaux, France 5 Universite´ de Bordeaux, Bordeaux, France 2
Corresponding author: Radhouane Dallel, Inserm/UCA U1107, Neuro-Dol: Trigeminal pain and Migraine, Faculte´ de Chirurgie Dentaire, 2 rue de Braga, 63100 Clermont-Ferrand, France. Email: [email protected]
2 drugs normally used to relieve acute and/or symptomatic headache can lead to medication overuse headache (MOH). According to cross-sectional studies, chronic migraine and MOH are associated in about 65% of patients (2). The pathogenesis of medication overuse headache is unclear. It is still not known whether chronic medication use increases sensitivity to pain perception and thus leads to chronicity, or whether migraine progression on its own facilitates medication abuse. However, one important issue to understand the mechanisms underlying such progression, from episodic to chronic migraine associated with medication overuse, is to determine whether and how endogenous pain modulation systems are altered in medication overuse headache vs. episodic migraine. One of the major endogenous pain inhibitory systems is conditioned pain modulation (CPM; formerly diffuse noxious inhibitory controls (DNIC)), a powerful general endogenous analgesic mechanism that can completely inhibit incoming nociceptor signals at the primary synapse (3). CPM is believed to play an important role in the development and exacerbation of chronic pain, because dysfunction of CPM is associated with a shift in balance between pain facilitation and pain inhibition. In many (but not all) patients with central sensitization, CPM is less efficacious (4). Migraine has been linked to an impairment of endogenous pain inhibitory systems, including subtle deficits in CPM (5). MOH could also result from a deficit in CPM (6,7). Chronic treatment with oral opioids is associated with altered CPM in patients with chronic pain (8). Accordingly, CPM was found to be reduced in MOH patients, while the situation improved after medication withdrawal (9). By contrast, a failure of CPM upon the R2 component of the blink-reflex was reported in another population of chronic migraineurs who had discontinued drug overuse for at least two months (10). There are thus possible conflicting results for the effect of medication overuse and withdrawal on modulation of pain processing in migraine patients. Here, we have sought to determine whether medication overuse and withdrawal could interfere with central modulation of pain processing in migraine and, more specifically, whether a decrease in CPM could play a part in medication overuse headache. To address this issue, we evaluated the effect of CPM on extracephalic pain perception in healthy volunteers (HV), episodic migraineurs (EM), and MOH patients, before and three weeks after withdrawal of medications. Since migraine is mostly a women’s disease and there are sex differences in CPM (11,12), we only included women in our study.
Methods This is an experimental study conducted in the Neurology Department of Clermont-Ferrand University Hospital (France).
Cephalalgia 0(0)
Standard protocol approvals, registrations, and subject consents The study was performed after approval was obtained from the regional Ethics Committee ‘‘CCP Sud-Est’’ in 2012. Standard written informed consent was obtained from all participants, according to the 1964 declaration of Helsinki. The study was registered as N 2008-A00471-54.
Participants In a population of female subjects aged 18–65 years having an appointment for migraine at the Neurology Department of Clermont-Ferrand University Hospital, two groups of subjects were successively recruited: Episodic migraineurs without aura (EM), and subjects suffering from medication overuse headache (MOH). After the first phone contact for appointment, a three-month diary was sent to patients and they were asked to indicate headache episodes and treatment intake on a daily basis. A third group of age-matched volunteer females was later recruited from a database of healthy people. Episodic migraineurs met the diagnosis criteria of the International Headache Society (IHS), had two to five attacks per month without aura, and had not used preventive medications for at least two months. Subjects suffering from MOH met the diagnosis criteria of the revised International Classification of headache Disorders that was available at the time of the study (second edition, ICHD-II (13)) and had migraine as the primary headache. MOH subjects with tension-type headache were not considered. Only women were included in the study, since migraine is mostly a women’s pathology and CPMs display sexspecific features (11,12,14). All subjects underwent a standardized interview, as well as a clinical neurological examination, and all were free of any other neurological dysfunction or chronic pain disease. All experiments were performed during the menstrual cycle follicular phase. Oral contraception was not an exclusion criterion. In patients with no menstrual cycle as a consequence of hormonal treatment, the treatment was maintained during experiments. In all groups, exclusion criteria were as follows: Treatment with tricyclic antidepressants or serotonin reuptake inhibitors, alcohol or drug addiction, pregnancy. All experiments in EM and MOH subjects were performed at least 24 hours after any acute migraine attack. If an attack happened within the 24 hours after a test, patients were asked to notify the investigator and the experimental results were discarded. In the MOH group, CPM experiments were performed again three weeks after treatment withdrawal (MOHw).
3
Guy et al.
Questionnaires Standard demographic information, duration of migraine, prophylactic and attack headache medications used before were collected on the day of inclusion. At the same time, all subjects also completed a series of selfreport questionnaires including the Beck Depression Inventory short form (BDI-SF), the State trait inventory anxiety (STAI YA) form, the Pain Catastrophizing Scale (PCS), the Headache Impact Test (HIT-6) and the Medication Dependence Questionnaire in Headache patients (MDQ-H).
Pain threshold measurements Extracephalic mechanical and thermal pain thresholds were assessed in all subjects, while they sat quietly in a room maintained at a constant temperature (23 C). Mechanical stimuli were applied on the back of the non-dominant hand using an electronic algometer (electronic von Frey, Bioseb, France). The strain gauge was connected to a plastic sterile cone (Eppendorf, Hamburg, Germany), the tip of which was applied perpendicularly to the studied skin area. The punctate pressure was gradually increased with a constant slope under visual control of the pressure value up to the detection of the mechanical pain threshold. The threshold was defined as the lowest pressure that produced a sensation of pain. The results of three separate consecutive measurements at different points in the testing area were averaged to establish the mechanical pain threshold value. Thermal pain was evoked by laser stimulation of the back of the non-dominant hand using a YAP laser stimulator (fiber–optic guidance, diameter 5 mm, duration 2 ms). The stimulator, with optic-fiber guidance, was placed on a skin area of about 6 cm2 at the dorsal aspect of the hand. Cutaneous heat stimuli were applied in incremental intensities in steps of 0.25 J, starting at a minimum of 0.75 J and reaching a maximum of 2.00 J. The threshold was defined as the lowest intensity that produced a sensation of pain.
CPM measurements CPM was assessed on the temporal summation of pain evoked by thermal stimuli. Blocks of four thermal test stimuli (Nd:YAP laser stimulator, diameter 5 mm; duration 2 ms) were delivered at 0.2 Hz on the back of the non-dominant hand. First, the lowest laser energy able to produce a moderate pain (3–6 on a 0–10 VAS scale) was determined (10 min before CPM assessment) by applying one block of four thermal test stimuli (spaced 5 sec apart) at an intensity just below the pain threshold. Thereafter, the intensity of the stimulation was adjusted according to the subject’s response, and
further blocks were repeated every 40 sec for 30 min, at an intensity producing a moderate pain (3–6 on a 0–10 VAS scale). After each block of stimulation, subjects were asked to score pain on a 0–10 VAS scale. The laser beam was moved (1 cm) between each block to avoid skin lesions. To produce CPM, the contralateral foot was immersed into a water bath (a conditioning stimulus) at a noxious cold (8 C) temperature, between the 10th and 20th minute of the experiment. In control experiments, performed on a different day in the same subjects, the water bath temperature was neutral (30 C). In all cases, water was constantly re-circulated to prevent laminar warming around the immersed foot. All subjects were asked to rate the cold pain on a 0–10 VAS scale immediately at the end of immersion. The order of CPM vs. control experiments was assigned randomly. The responses for each block in each condition were plotted over time. The first 15 responses were used to study the unconditioned temporal summation of pain, the next 15 to study CPM, and the 15 final responses to study the post-effects of CPM.
Treatment withdrawal management One week after the initial CPM experiments, MOH subjects were instructed to abruptly stop headache medication intake and immediately hospitalized for five days. During this period, headache was managed using ice packs applied on the forehead, and when necessary, sedative medication (hydroxyzine, 25 mg per day). At discharge from hospital, patients were asked to continue withdrawal at home, until they could wake up without headache. Afterwards, they were allowed to take a triptan at the beginning of the next attack. Preventive therapy was begun three weeks after treatment withdrawal, just after the last CPM experiment was performed.
Data analysis Statistics were computed with STATA V12 (Stata Corp, College Station, TX, USA). Results were expressed as mean standard error of the mean (SEM) or median (interquartile range) and as frequencies (percentage). Groups were compared using the Chi square test (or Fisher’s exact test when appropriate) for categorical data, followed by the Marascuillo procedure when signification was reached comparing groups, and using Student t-test (for comparing two groups) or ANOVA (for comparing three or more groups), or the Mann-Whitney/Kruskal-Wallis test (for comparing two or more groups), depending on the data distribution for continuous data. Those tests were followed by a Dunn’s post-hoc test when significance was reached between
4
Cephalalgia 0(0)
groups. Relationships between the number of headache days and BDI, STAI, PCS and MDQ-H scores were assessed by Spearman’s correlation coefficient. The evolution of the number of headache days and tablet intake from baseline to three weeks after withdrawal was analyzed using the Wilcoxon matched-pairs signed rank test. Pain perception was analyzed using linear mixed models with random subject intercept. Groups, time, and foot water immersion were tested as fixed effects. Those methods were developed for temporal subdivision (control, water immersion and post-water immersion periods) in order to test: a) the session effect, b) the effect of the repeated painful stimulation, and c) the water immersion effect and the post-water immersion effect. Those analyses were repeated intra and inter groups. All tests were two-sided and a p-value < 5% was considered statistically significant.
Results Study population Thirty-six female subjects were included in the present study, out of 42 who were screened. Six subjects were excluded (four in the HV group, two in the MOH group) because their pain tests were not sufficiently painful (
Medication overuse reinstates conditioned pain modulation in women with migraine
Cephalalgia 0(0) 1–11 ! International Headache Society 2017 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0333102417727545 journals.sagepub.com/home/cep
Nathalie Guy1,2,3, Daniel Voisin4,5, Aure´lien Mulliez2, Pierre Clavelou1,2,3 and Radhouane Dallel1,2,3 Abstract Background: This study investigated the effects of medication overuse and withdrawal on modulation of pain processing in women with migraine. Temporal summation of laser-evoked thermal pain was used to measure the effects of conditioned pain modulation. Methods: 36 female participants (12 healthy volunteers, 12 with episodic migraine and 12 with medication overuse headache) were included in a two session protocol. Medication overuse headache subjects were also tested three weeks after medication overuse headache withdrawal. Mechanical and laser-evoked thermal pain thresholds were measured on the back of the non-dominant hand where, later, temporal summation of laser-evoked thermal pain to repetitive thermal stimuli was elicited for 30 min, at an intensity producing moderate pain. Between the 10th and 20th minutes, the contralateral foot was immersed into a water bath at a not painful (30 C) or painfully cold (8 C; conditioned pain modulation) temperature. Results: Episodic migraine, medication overuse headache and medication overuse headache withdrawal were associated with an increase in extracephalic temporal summation of laser-evoked thermal pain as compared to healthy volunteer subjects, while there was no alteration of laser-evoked thermal and mechanical extracephalic pain thresholds in these subjects. Conditioned pain modulation was highly efficient in temporal summation of laser-evoked thermal pain in healthy volunteer subjects, with a solid post-effect (reduction of pain). Conditioned pain modulation was still present, but reduced, in episodic migraine. By contrast, conditioned pain modulation was normal in medication overuse headache and strongly reduced in medication overuse headache withdrawal. Furthermore, in medication overuse headache withdrawal, the post-effect was no longer a decrease, but a facilitation of pain. Conclusions: These data show that a decrease in conditioned pain modulation does not underlie medication overuse headache in women. On the contrary, medication overuse reinstated conditioned pain modulation in female migraine patients. They also identify different phenotypes of pain modulation in migraine patients. Registration number N 2008-A00471-54. Keywords Migraine, medication overuse headache, conditioned pain modulation, pain summation Date received: 4 May 2017; revised: 15 June 2017; accepted: 4 July 2017
Introduction Migraine is a common and disabling neurological disorder that manifests as a recurrent, episodic, often severe, headache. Among episodic migraineurs, a subset develops chronic migraine, a very invalidating condition characterized by headaches for more than 15 days a month. The most prominent risk factors that can promote chronic migraine include the frequency of attacks, psychiatric co-morbidities such as anxiety, depression, addictive behavior, and medication overuse (1). Indeed, the regular overuse for more than three months of one or more of the
1
Universite´ Clermont Auvergne, Clermont-Ferrand, France CHU Clermont-Ferrand, Clermont-Ferrand, France 3 Inserm U1107, Neuro-Dol, Clermont-Ferrand, France 4 Neurocentre Magendie, Inserm U1215, Bordeaux, France 5 Universite´ de Bordeaux, Bordeaux, France 2
Corresponding author: Radhouane Dallel, Inserm/UCA U1107, Neuro-Dol: Trigeminal pain and Migraine, Faculte´ de Chirurgie Dentaire, 2 rue de Braga, 63100 Clermont-Ferrand, France. Email: [email protected]
2 drugs normally used to relieve acute and/or symptomatic headache can lead to medication overuse headache (MOH). According to cross-sectional studies, chronic migraine and MOH are associated in about 65% of patients (2). The pathogenesis of medication overuse headache is unclear. It is still not known whether chronic medication use increases sensitivity to pain perception and thus leads to chronicity, or whether migraine progression on its own facilitates medication abuse. However, one important issue to understand the mechanisms underlying such progression, from episodic to chronic migraine associated with medication overuse, is to determine whether and how endogenous pain modulation systems are altered in medication overuse headache vs. episodic migraine. One of the major endogenous pain inhibitory systems is conditioned pain modulation (CPM; formerly diffuse noxious inhibitory controls (DNIC)), a powerful general endogenous analgesic mechanism that can completely inhibit incoming nociceptor signals at the primary synapse (3). CPM is believed to play an important role in the development and exacerbation of chronic pain, because dysfunction of CPM is associated with a shift in balance between pain facilitation and pain inhibition. In many (but not all) patients with central sensitization, CPM is less efficacious (4). Migraine has been linked to an impairment of endogenous pain inhibitory systems, including subtle deficits in CPM (5). MOH could also result from a deficit in CPM (6,7). Chronic treatment with oral opioids is associated with altered CPM in patients with chronic pain (8). Accordingly, CPM was found to be reduced in MOH patients, while the situation improved after medication withdrawal (9). By contrast, a failure of CPM upon the R2 component of the blink-reflex was reported in another population of chronic migraineurs who had discontinued drug overuse for at least two months (10). There are thus possible conflicting results for the effect of medication overuse and withdrawal on modulation of pain processing in migraine patients. Here, we have sought to determine whether medication overuse and withdrawal could interfere with central modulation of pain processing in migraine and, more specifically, whether a decrease in CPM could play a part in medication overuse headache. To address this issue, we evaluated the effect of CPM on extracephalic pain perception in healthy volunteers (HV), episodic migraineurs (EM), and MOH patients, before and three weeks after withdrawal of medications. Since migraine is mostly a women’s disease and there are sex differences in CPM (11,12), we only included women in our study.
Methods This is an experimental study conducted in the Neurology Department of Clermont-Ferrand University Hospital (France).
Cephalalgia 0(0)
Standard protocol approvals, registrations, and subject consents The study was performed after approval was obtained from the regional Ethics Committee ‘‘CCP Sud-Est’’ in 2012. Standard written informed consent was obtained from all participants, according to the 1964 declaration of Helsinki. The study was registered as N 2008-A00471-54.
Participants In a population of female subjects aged 18–65 years having an appointment for migraine at the Neurology Department of Clermont-Ferrand University Hospital, two groups of subjects were successively recruited: Episodic migraineurs without aura (EM), and subjects suffering from medication overuse headache (MOH). After the first phone contact for appointment, a three-month diary was sent to patients and they were asked to indicate headache episodes and treatment intake on a daily basis. A third group of age-matched volunteer females was later recruited from a database of healthy people. Episodic migraineurs met the diagnosis criteria of the International Headache Society (IHS), had two to five attacks per month without aura, and had not used preventive medications for at least two months. Subjects suffering from MOH met the diagnosis criteria of the revised International Classification of headache Disorders that was available at the time of the study (second edition, ICHD-II (13)) and had migraine as the primary headache. MOH subjects with tension-type headache were not considered. Only women were included in the study, since migraine is mostly a women’s pathology and CPMs display sexspecific features (11,12,14). All subjects underwent a standardized interview, as well as a clinical neurological examination, and all were free of any other neurological dysfunction or chronic pain disease. All experiments were performed during the menstrual cycle follicular phase. Oral contraception was not an exclusion criterion. In patients with no menstrual cycle as a consequence of hormonal treatment, the treatment was maintained during experiments. In all groups, exclusion criteria were as follows: Treatment with tricyclic antidepressants or serotonin reuptake inhibitors, alcohol or drug addiction, pregnancy. All experiments in EM and MOH subjects were performed at least 24 hours after any acute migraine attack. If an attack happened within the 24 hours after a test, patients were asked to notify the investigator and the experimental results were discarded. In the MOH group, CPM experiments were performed again three weeks after treatment withdrawal (MOHw).
3
Guy et al.
Questionnaires Standard demographic information, duration of migraine, prophylactic and attack headache medications used before were collected on the day of inclusion. At the same time, all subjects also completed a series of selfreport questionnaires including the Beck Depression Inventory short form (BDI-SF), the State trait inventory anxiety (STAI YA) form, the Pain Catastrophizing Scale (PCS), the Headache Impact Test (HIT-6) and the Medication Dependence Questionnaire in Headache patients (MDQ-H).
Pain threshold measurements Extracephalic mechanical and thermal pain thresholds were assessed in all subjects, while they sat quietly in a room maintained at a constant temperature (23 C). Mechanical stimuli were applied on the back of the non-dominant hand using an electronic algometer (electronic von Frey, Bioseb, France). The strain gauge was connected to a plastic sterile cone (Eppendorf, Hamburg, Germany), the tip of which was applied perpendicularly to the studied skin area. The punctate pressure was gradually increased with a constant slope under visual control of the pressure value up to the detection of the mechanical pain threshold. The threshold was defined as the lowest pressure that produced a sensation of pain. The results of three separate consecutive measurements at different points in the testing area were averaged to establish the mechanical pain threshold value. Thermal pain was evoked by laser stimulation of the back of the non-dominant hand using a YAP laser stimulator (fiber–optic guidance, diameter 5 mm, duration 2 ms). The stimulator, with optic-fiber guidance, was placed on a skin area of about 6 cm2 at the dorsal aspect of the hand. Cutaneous heat stimuli were applied in incremental intensities in steps of 0.25 J, starting at a minimum of 0.75 J and reaching a maximum of 2.00 J. The threshold was defined as the lowest intensity that produced a sensation of pain.
CPM measurements CPM was assessed on the temporal summation of pain evoked by thermal stimuli. Blocks of four thermal test stimuli (Nd:YAP laser stimulator, diameter 5 mm; duration 2 ms) were delivered at 0.2 Hz on the back of the non-dominant hand. First, the lowest laser energy able to produce a moderate pain (3–6 on a 0–10 VAS scale) was determined (10 min before CPM assessment) by applying one block of four thermal test stimuli (spaced 5 sec apart) at an intensity just below the pain threshold. Thereafter, the intensity of the stimulation was adjusted according to the subject’s response, and
further blocks were repeated every 40 sec for 30 min, at an intensity producing a moderate pain (3–6 on a 0–10 VAS scale). After each block of stimulation, subjects were asked to score pain on a 0–10 VAS scale. The laser beam was moved (1 cm) between each block to avoid skin lesions. To produce CPM, the contralateral foot was immersed into a water bath (a conditioning stimulus) at a noxious cold (8 C) temperature, between the 10th and 20th minute of the experiment. In control experiments, performed on a different day in the same subjects, the water bath temperature was neutral (30 C). In all cases, water was constantly re-circulated to prevent laminar warming around the immersed foot. All subjects were asked to rate the cold pain on a 0–10 VAS scale immediately at the end of immersion. The order of CPM vs. control experiments was assigned randomly. The responses for each block in each condition were plotted over time. The first 15 responses were used to study the unconditioned temporal summation of pain, the next 15 to study CPM, and the 15 final responses to study the post-effects of CPM.
Treatment withdrawal management One week after the initial CPM experiments, MOH subjects were instructed to abruptly stop headache medication intake and immediately hospitalized for five days. During this period, headache was managed using ice packs applied on the forehead, and when necessary, sedative medication (hydroxyzine, 25 mg per day). At discharge from hospital, patients were asked to continue withdrawal at home, until they could wake up without headache. Afterwards, they were allowed to take a triptan at the beginning of the next attack. Preventive therapy was begun three weeks after treatment withdrawal, just after the last CPM experiment was performed.
Data analysis Statistics were computed with STATA V12 (Stata Corp, College Station, TX, USA). Results were expressed as mean standard error of the mean (SEM) or median (interquartile range) and as frequencies (percentage). Groups were compared using the Chi square test (or Fisher’s exact test when appropriate) for categorical data, followed by the Marascuillo procedure when signification was reached comparing groups, and using Student t-test (for comparing two groups) or ANOVA (for comparing three or more groups), or the Mann-Whitney/Kruskal-Wallis test (for comparing two or more groups), depending on the data distribution for continuous data. Those tests were followed by a Dunn’s post-hoc test when significance was reached between
4
Cephalalgia 0(0)
groups. Relationships between the number of headache days and BDI, STAI, PCS and MDQ-H scores were assessed by Spearman’s correlation coefficient. The evolution of the number of headache days and tablet intake from baseline to three weeks after withdrawal was analyzed using the Wilcoxon matched-pairs signed rank test. Pain perception was analyzed using linear mixed models with random subject intercept. Groups, time, and foot water immersion were tested as fixed effects. Those methods were developed for temporal subdivision (control, water immersion and post-water immersion periods) in order to test: a) the session effect, b) the effect of the repeated painful stimulation, and c) the water immersion effect and the post-water immersion effect. Those analyses were repeated intra and inter groups. All tests were two-sided and a p-value < 5% was considered statistically significant.
Results Study population Thirty-six female subjects were included in the present study, out of 42 who were screened. Six subjects were excluded (four in the HV group, two in the MOH group) because their pain tests were not sufficiently painful (
