واکنش پذیری روانشناختی در بیماران مبتلا به میگرن در کودکان و افراد سالم

Abstract The hypothesis that physiological responses of migraine patients are symptom-specific was evaluated in 29 children (age range 8–16 years) suffering from migraine and 10 healthy control children. The assessment included two major stress phases and a relaxation period. A standard laboratory stressor (a subtraction task) and parent–child conflict served as stressors. A total of six physiological parameters were measured: pulse amplitude at two extracranial (A. temporalis, A. supraorbitalis) and one peripheral (index finger) sites; finger temperature; heart rate; and skin-conductance level. There were no significant group differences in autonomic arousal. Moreover, extracranial and peripheral vasomotor activity was not different between groups, a finding which might be partially due to the considerable interindividual variability. The implications of the results are discussed taking into account that studying pediatric rather than adult migraine patients allows to minimize the potentially confounding impact of factors such as headache chronicity, medication, and additional nonmigraine headaches.

Introduction In contrast to earlier models, which typically postulated a primary vascular dysfunction [1]or neurological disturbance (e.g., ref. [2]) as the underlying cause of migraine, recent models of migraine pathophysiology posit that a migraine attack is the result of a complex interplay between neural, vascular and myofascial events (e.g., refs. [3]and [4]). It is assumed that neurovascular sensitivity to physical (e.g., exposure to light) and emotional stressors plays a central role in the proneness to recurrent migraine attacks 4 and 5. This claim is consistent with the clinical observation that adult and child migraine patients perceive “stress” as one of the most important migraine triggers (e.g., refs. 6, 7 and 8). Yet, it is less clear to what extent migraine patients exhibit a specific psychophysiological response pattern that would be indicative of the presumed neurovascular sensitivity (cf. ref. [9]). Migraine patients have, in general, although not consistently (e.g., ref. [10]), been shown to respond with a more pronounced vasodilation of the A. temporalis when exposed to aversive noise 11 and 12, stress imagery [13], and reaction time tasks 14 and 15, or when attempting to relax [16]. However, differences in stress-induced changes in heart rate, skin conductance, and finger temperature between migraine patients and control subjects have rarely been demonstrated 15 and 17. In contrast to the data on stress-induced reactivity, there is little evidence to support the hypothesis of significant baseline differences in sympathetic arousal and vasomotor functioning between migraine patients and control subjects (e.g., refs. 13, 14, 18, 19, 20 and 21). Only a small proportion of studies has specifically evaluated the recovery of psychophysiological responses. Whereas some studies have reported a delayed recovery of vasomotor activity 13, 14 and 22, other studies have failed to replicate these findings 18, 19 and 23. By contrast, there seems to be some consensus that indices of general arousal (i.e., heart rate or skin conductance) or indirect measures of peripheral vasoactivity (i.e., peripheral temperature) do not differ in their return to baseline between migraine and control subjects 13, 24 and 25. It cannot be ruled out that the inconsistent results might be a consequence of methodological and conceptual weaknesses (for a comprehensive review see ref. [26]). Assessing children with migraine not only minimizes the confounding effects of chronic headaches (HA) and long-term use of medication (especially vasoactive drugs), but the findings are also less likely to be influenced by the (largely unknown) effects of recurrent episodes of tension-type HA and the psychological sequelae of chronic pain (e.g., depression, anxiety). According to epidemiological and clinical observations, the prevalence of tension-type HA increases with age 8 and 27. Moreover, the level of depression, anxiety, and behavior problems of children suffering from migraine tends to be only slightly increased when compared with healthy controls 28, 29 and 30as opposed to the significant changes in psychological functioning typically found in adult migraine patients [31]. Despite its potential usefulness, psychophysiological reactivity in children with migraine has not yet been investigated systematically. The purpose of this study was to examine psychophysiological responses to stress in children with migraine in comparison to nonheadache controls with the main hypothesis that children suffering from migraine exhibit significantly more cranial vasodilation and peripheral vasoconstriction in response to stress. Measures of sympathetic arousal were included to determine the specificity of the physiological responses, especially given the lack of empirical data for children suffering from migraine. Vasomotor activity was recorded peripherally and from two cranial sites. In addition to the temporal artery, vasomotor responses were also measured from the A. supraorbitalis. This second cranial site was chosen because children with migraine often locate their HA above or behind the eye. Moreover, the inclusion of a second cranial recording site allowed the evaluation of the extent (“generality”) of abnormal cranial vasomotor activity in (pediatric) migraine 32 and 33. To increase the validity of the induced stress, parent–child conflict and a standard laboratory stressor were used. Both situations can be considered as (at least potentially) individually relevant stressors because achievement situations and emotional stress are among the most frequently endorsed migraine triggers in children

Results2 3.1. Measures of autonomic activity 3.1.1. Resting baseline values There were no significant differences between the two groups with respect to the resting levels of HR, SCL, and finger temperature. The respective means and standard deviations are presented in Table 1. 3.1.2. Stress reactivity As Fig. 1 illustrates, stress-induced changes in HR, SCL, and finger temperature did not differ between the children with migraine and the control children. Similarly, the interaction effects between patient status and stress phases did not reach significance for any of the three indices of autonomic activity. The stressors induced significantly different changes in HR, SCL, and finger temperature [HR: Wilk's lambda=0.3, F(6, 30)=11.3, p<0.001; SCL: Wilk's lambda=0.4, F(6, 30)=6.9, p<0.001; finger temperature: Wilk's lambda=0.16, F(6, 31)=28.0, p<0.001]. A priori contrasts revealed that (except for finger temperature in the subtraction task) the actual stressors (i.e., subtraction task and parent–child interaction) consistently induced a stronger physiological stress response than the anticipation of the stressor [subtraction—HR: t(37)=7.2, p<0.001; SCL: t(37)=6.1, p<0.001; parent–child interaction—HR: t(37)=2.03, p⩽0.05; SCL: t(37)=5.1, p<0.001; finger temperature: t(37)=10.6, p<0.001]. Moreover, the parent–child interaction elicited a stronger increase in SCL and only a slight increase in finger temperature in comparison to the subtraction task [SCL: t(37)=2.3, p<0.05; finger temperature: t(37)=−2.0, p<0.05]. By contrast, the subtraction task was associated with the most pronounced increase in HR [t(37)=4.2, p<0.001]. 3.1.3. Recovery phases There were no differences between the two groups with regard to return to baseline for HR, SCL, and finger temperature. HR and SCL changed significantly across the recovery phases [HR: Wilk's lambda=0.8, F(2, 36)=4.3, p<0.05; SCL: Wilk's lambda=0.7, F(2, 34)=6.5, p<0.01]. As suggested by the respective polynomial contrasts, SCL was highest and HR was lowest during the recovery phase following the parent–child interaction when compared with the other two recovery phases [quadratic trend: SCL: F(1, 35)=12.9, p⩽0.001; HR: F(1, 37)=6.3, p<0.05]. This pattern is consistent with the finding that SCL was highest during the parent–child interaction, which may have led to some carryover to the recovery phase. By contrast, the moderate increase in HR elicited by the interaction task may have facilitated the return to baseline. The lack of a group difference cannot be accounted for by an insufficient recovery process. As revealed by the comparison between the beginning and the end of the recovery phases, at least two of the autonomic measures showed a significant return to baseline. SCL decreased significantly from the beginning (minute 1) to the end (minute 5) of the recovery phases [F(1, 35)=29.5, p<0.001]. Moreover, finger temperature either showed a significant increase or remained unchanged, thus providing additional support for the presumed recovery process [main effect for minute: F(1, 36)=5.3, p<0.05); phase×minute: Wilk's lambda=0.5, F(2, 35)=16.9, p<0.001]. 3.2. Vasomotor activity: percent change from baseline 3.2.1. Stress reactivity Fig. 2 presents the changes in supraorbital, temporal, and digital pulse amplitude across the stressor phases. Statistical analyses did not yield evidence for a group effect for any of the recording sites. Similarly, there was no evidence that the two groups showed different extracranial vasomotor activity or digital BVP depending on the stressor. As illustrated in Fig. 2, the stressor phases led to significantly different changes in digital [Wilk's lambda=0.65, F(6, 27)=2.4, p⩽0.05] and temporal BVP [Wilk's lambda=0.6, F(6, 29)=3.4, p⩽0.01]. Unlike hypothesized, the significant phase effect could not be accounted for by a more pronounced change in digital and temporal BVP during the actual stress phases as compared with the anticipation phases. Moreover, there was no evidence for a significantly different digital vasomotor response associated with either the subtraction task or parent–child interaction. However, in comparison to the subtraction task, the parent–child interaction led to a significant vasodilation of the A. temporalis [t(37)=−2.9, p<0.01]. 3.2.2. Recovery phases There was no evidence for a difference between the two groups with regard to the return to baseline of extracranial and digital vasomotor activity. For both extracranial sites the data revealed a significant vasoconstriction toward the end of the recovery phases, which reflects a return to baseline after the stress-induced vasodilation [A. supraorbitalis: F(1, 36)=8.8; p<0.01; temporal artery: F(1, 36)=10.4, p<0.01]. The A. temporalis showed the highest degree of poststressor vasodilation subsequent to the parent–child interaction, which is consistent with the finding that the parent–child interaction led to the most noticeable stress-related changes in psychophysiological activity [main effect-phase: Wilk's lambda=0.8, F(2, 35)=4.4, p<0.05; quadratic trend:F(1, 36)=4.3, p<0.05]. Overall, the analyses did not reveal any evidence for a differential recovery as a function of diagnostic status. 3.2.3. Interindividual variability As was observed in previous studies 12 and 45, inspection of the data suggests considerable interindividual variability in vasomotor response. This response variability was explored by classifying the subjects depending on the magnitude and direction of the percent change from baseline of the digital and temporal BVP elicited by the subtraction and the parent–child interaction task. Supraorbital vasomotor responses were not considered because they had failed to demonstrate stressor-induced response variability. In an attempt to represent the full range of the observed vasomotor responses, the following categories were defined for each direction of change of BVP from baseline: minimal (<10% change); moderate (⩾10% and <50% change), large (⩾50% and <100% change); and extreme (⩾100% change). Although a formal statistical comparison was not possible because the expected cell frequency was less than 5 for more than 50% of the cells, the data did not readily suggest a different response pattern between the two groups. In fact, for both temporal and digital BVP, and for both stressor phases, the proportion of subjects displaying minimal, moderate, large, and extreme vasodilation or vasoconstriction was essentially equal in both groups. It is noteworthy that neither of the two stressors consistently elicited vasoconstriction or vasodilation across subjects. 3.3. Subjective ratings of discomfort A repeated measurement MANOVA, which included the ratings for both the baseline, recovery, and stress phases, did not reveal any overall differences in subjective discomfort between the two groups. The phases, however, differed with regard to the induced discomfort [Wilk's lambda=0.28, F(6, 31)=13.1, p<0.001]. The perceived distress decreased significantly over the course of the assessment, irrespective of the type of phase, that is, stressor phase or recovery phase [F(1, 36)=22.6, p<0.001]. Nonetheless, the stressor phases were associated with a slight increase in discomfort in comparison to the recovery phases [F(1, 36)=7.6, p<0.01). Thus, the subjective ratings of discomfort gave evidence for a gradual decline throughout the assessment, but suggest that the stressor phases led to a transient increase in subjective discomfort. 3.4. Parent–child interaction The anger ratings for the discussed issues, as indicated on the Issues Checklist, did not differ between the groups (patients: Full-size image (<1 K); controls: Full-size image (<1 K)). A MANOVA did not reveal significant differences between patients and controls with respect to the negative and positive interaction behavior of the parent and the child. Similarly, the parent–child interactions were not characterized by a different degree of problem resolution (patients: Full-size image (<1 K); controls: Full-size image (<1 K)), putting each other down (patients: Full-size image (<1 K); controls: Full-size image (<1 K)), friendliness (patients: Full-size image (<1 K); controls: Full-size image (<1 K)), and effectiveness at conflict resolution (patients: Full-size image (<1 K); controls: Full-size image (<1 K)). To summarize, both groups perceived the parent–child interaction as equally anger-provoking and showed a similar amount of negative interaction behavior.