- Psychophysiological assessment: Profile of stress related to pain. This is a computerized battery made up of 20 situations supposedly related to pain, and which the subject is required to personalize and briefly imagine. During the test, and by means of a computerized system of registration and processing of psychophysiological responses (J&J I-330) connected to a PC, we recorded the following physiological responses: frontal electromyographical activity (EMG1) and electromyographical activity in the muscles of the trapezius (EMG2); respiratory rate (RR); heart rate (HR); blood pulse volume (BPV); peripheral temperature (TEMP); and skin conductance (CON).

- Expectancies about the psychophysiological assessment: questionnaire on expectancies related to this test, based on the scale developed by Borkovec and Nau (1972). Subjects were required to respond on a scale of 0-10 to the following items: REALISM (degree of realism in imagining the situations), EQUIPMENT (rating of the use of computer equipment in the diagnosis), SUITABILITY (suitability of the test for their problem), LOGIC (rating of the logic of the test), CONTRIBUTION (contribution of the test to the solution of the problem), KNOWLEDGE (contribution of the test to knowledge of the problem), SOLUTION (current confidence in solution of the problem), RECOMMENDATION (intention to recommend the test to relatives or friends with similar problems), ACCESSIBILITY (rating of the accessibility of the test), UNIVERSITY (opinion on whether the tests should be carried out in the university).

- Locus of control related to health: measured by means of form A of the "Multidimensional Health Locus of Control" (MHLC) (Wallston and cols., 1978). Measures were obtained of its three dimensions: INTERNALITY, POWER of OTHERS and LUCK.

- Degree of suggestibility: measured by means of the FALLING-BACKWARDS test of hypnotic suggestibility, first item of forms A and B of the Stanford scale (Weitzenhoffer and Hilgard, 1959). This test was applied three times to each subject, so that the score could range from 0 to 3, depending on the number of times the subject reacted to the suggestions of falling.

- Type A Behaviour Pattern (TABP): measured using the Jenkins Activity Scale (JAS) (Jenkins and cols., 1979), form C, Spanish version by Fernández-Abascal (1992). We obtained, in addition to the TYPE-A measure itself, those of each of its three components: IMPATIENCE, WORK and COMPETITIVENESS.

- Depression: measured with Beck’s Depression Inventory (BDI) (Beck and cols., 1961), Spanish version by Conde and Useros (1975).

- ASSOCIATIONS in the baseline: percentage of occasions on which subjects associated concrete situations with the onset of a headache episode. This measure was obtained from the self-records made during the baseline month.

- Percentage of IMPROVEMENT: degree of reduction of the headache index during each post-treatment period with respect to the baseline. Headache index (frequency x duration x intensity) was obtained from the daily pain self-records.

1) The greater the AGE of subjects, the greater the improvement with the Passive treatment. This prediction contrasts with numerous studies in which it is reported that young subjects tend to respond to psychological headache treatments better than older subjects (e.g., Werder and cols., 1981; Chapman, 1986). This contradiction may be explained by the peculiarities of the Passive approach that most differentiate it from conventional treatments. For example, while in normal psychological treatments (those that could be considered within the framework of the Active approach) it is necessary to take precautions with older subjects in order to compensate for certain deficits (Arena and cols., 1988 and 1991), in this work the very characteristics of the Passive treatment appeared to make such precautions unnecessary. In sum, it seems reasonable to state that the Passive treatment, on not requiring the learning or use of any type of ability, is a suitable approach for older headache patients.

2) The lower the EDUCATIONAL level, the better the response to the Passive treatment. Level of education, though a classic variable in the study of therapeutic processes (Shoham-Salomon and Hannah, 1991), does not tend to be useful in the prediction of the effectiveness of psychological treatment of headache (Werder and cols., 1981). The negative relationship between educational level and effectiveness of the Passive treatment obtained in this study may be the result of secondary suggestibility (Eysenck and Furneaux, 1945), a variable positively related to ingenuousness and credulity (Eysenck, 1989). Nevertheless, the finding of this relationship does not appear to be consistent with other results, as we shall see on discussing the next variable.

3) The greater the suggestibility (FALLING-BACKWARDS), the better the response to the Passive treatment. Suggestibility, as measured with the FALLING-BACKWARDS test, would respond to characteristics of primary suggestibility (Eysenck and Furneaux, 1945), a variable strongly correlated with hypnotizability (Eysenck, 1989). The results of this study support the positive relationship observed between primary suggestibility (hypnotizability) and improvement after the psychological treatment of chronic pain (Spinhoven, 1988). Nevertheless, it should be pointed out that the relationship between primary suggestibility and the improvement found with a placebo treatment (an effect of no little importance in the case of Passive treatment) is not usually significant (Evans, 1989). Also, the fact that primary suggestibility and low educational level act jointly to predict effectiveness of the Passive treatment appears to be incongruent with previous results, since a lack of relationship between primary and secondary suggestibility is usually reported (Eysenck, 1989). In view of the above, it would seem contradictory to consider secondary suggestibility as a mediator between low educational level and improvement.

4) Low level of muscular tension in the trapezia (EMG2) as a predictor of improvement with Passive treatment. Although the relationship between muscular tension and headache is quite complex, the literature reviewed provides no data that can help to illustrate this result.

5) The lower the score in IMPATIENCE, the greater the improvement after Passive treatment. This relationship would appear to indicate the greater effectiveness of the Passive treatment in people that are not particularly irritable, and who do not tend to behave in an impatient manner or put pressure on others (Fernández-Abascal, 1994). Although some studies point to a positive relationship between headache and global TABP score (Rappaport and cols., 1988), there are no data with respect to the relationship between headache and each of the components of the JAS, nor on the relationship between these components and improvement subsequent to intervention. Our own results, though referring to only one component of the pattern, are in contrast to those reported in the review by Chapman (1986), who suggests that the presence of high scores in TABP is a good predictor of the effectiveness of treatments such as biofeedback or relaxation. Taking all of this into account, we can speculate that the involvement of the variable IMPATIENCE in the improvement obtained with the Passive treatment may be attributable to those people that, on not presenting characteristics typical of impatient people, do not "need" an active intervention that modulates their behaviour style.

6) The greater the number of ASSOCIATIONS in the baseline, the greater the improvement. This relationship is somewhat paradoxical. Capacity to observe relationships between pain and triggering situations would appear to provide a good prediction for the success of an approach such as the Active one, which is based on the modification of such relationships. Consequently, its involvement in the success of the Passive phase is not coherent with the logic of this type of treatment, which appears to lie more in the measurement of non-specific cognitive variables than in the subject’s ability for self-observation. An aspect that could be related to the intervention of self-observation ability in the improvement achieved in the Passive phase may be precisely a paradoxical effect, such as that postulated by some authors (Arntz and Schmidt, 1989; Vallejo and Comeche, 1994): the Passive treatment, on focusing the responsibility for control on an external source, would turn out to be beneficial for the patients with the greatest capacity for self-observation.

1 & 2) The lower the EDUCATIONAL level and level of muscular tension in the trapezia (EMG2), the better the response to Active treatment in PT2. These two variables already formed part of the regression equation obtained for predicting improvement in PT1. Their involvement, though difficult to explain (as already pointed out), appears to constitute a continuation of the effect found in the first treatment phase.

3) Considering the initial test of psychophysiological assessment as not very SUITABLE for the subjects’ problem is a good predictor of improvement in the Active phase of PT2. This relationship is congruent with the characteristics of the Active phase. It appears to denote an attitude of rejection of machines, and consequently a better response to the Active treatment received in this second phase. That is, subjects that considered the initial computerized test as fairly unsuitable improved with an Active treatment, perhaps because they found it radically different from the test of which they disapproved.

1) The greater the REALISM on imagining the situations described in the initial psychophysiological assessment, the better the response to the Active treatment. This relationship would appear, in principle, more congruent with the characteristics of the Passive treatment than with those of the Active phase. Indeed, the involvement of imagination as a personality trait has frequently been positively correlated with hypnotic responsiveness (Kirsch, 1990). However, the fact that in this group of subjects there is observed a negative and significant correlation between the variables REALISM and FALLING-BACKWARDS would indicate that subjects who imagined the situations with the greatest degree of realism were in turn the least suggestible, so that it does not seem reasonable to interpret the prediction of the variable REALISM in the sense of hypnotizability. It would appear more coherent to deduce that the greater degree of REALISM is a reflection of a voluntary process, of wishing to fulfil the therapeutic prescription, since this behavioural style is ideally productive for an approach like the Active one, in which the permanent and active collaboration of the subject is necessary throughout the therapeutic process.

2) Low score in the externality factor POWER of OTHERS as a good predictor of improvement with the Active treatment. This relationship is coherent with the demands of this approach. Given that subjects do not focus their own expectancies of cure on factors external to themselves, it appears logical that they tend to become more involved in their own therapeutic process, and consequently obtain better results with an Active approach. If we consider this result in relation to the literature on the subject, it is easy to find data on the positive intervention of internality (Hudzinski and Levenson, 1985; Härkäpää, 1991), but not on the negative relationship between externality and improvement. However, discrepant results also appear, such as the lack of utility of locus of control as a predictor of headache severity (Jones and Page, 1986), or as a predictor of the efficacy of real or placebo feedback (Díaz and Vallejo, 1987), and, contrary to expectations (Gale and Funch, 1984), the beneficial effects of POWER of OTHERS in the effectiveness of Behavioural Therapy. In sum, although the relationship between POWER of OTHERS and improvement has not so far found direct support in the literature, the involvement of this variable in the effectiveness of Active treatment appears to be coherent with the peculiarities of that treatment.

3) The higher the level of expectancies of KNOWLEDGE of the problem through the Psychophysiological Assessment, the greater the improvement with the Active treatment. This relationship appears to be congruent, since the content of the questions asked during the assessment covered a set of aspects potentially related to pain (stress-generating situations, typical thoughts in these patients, activities that relieve pain, etc.) –aspects usually considered as related to the onset, worsening or relief of headache (Drummond, 1985).

1) The greater the IMPROVEMENT in PT1, the greater the improvement after the Passive phase in PT2. This relationship appears to indicate that the completion of the entire sequence produces an optimization of the results obtained in the first part of that sequence. In this case it would indicate, specifically, that the improvement experienced after the Active treatment phase is consolidated and strengthened with the carrying out of the Passive phase.

2) A high level of frontal muscular tension (EMG1) during the Psychophysiological Assessment as a predictor of the effectiveness of the Passive treatment in the PT2. This result is congruent with the findings of Cram (1980), who found that tension headache patients with high levels of EMG in the initial assessment (moderately stressful cognitive task) presented a greater improvement in a treatment programme with biofeedback and relaxation. However, this result, in turn, contradicts the findings of Blanchard and cols. (1983), who reported less improvement after biofeedback training in subjects with higher basal levels of EMG.

3) The greater the level of EMG2 tension, the greater the improvement after the Passive treatment in PT2. This relationship also appeared in the prediction of the efficacy of the two types of treatment in the Group I subjects; as already pointed out, there are no data in the literature reviewed that could help us illustrate the relationship between these variables.

4) High score in the variable RECOMMENDATION as a good predictor of improvement after the Passive treatment in PT2. This relationship appears to be coherent with that postulated by Kirsch’s team (Kirsch, 1985; Kirsch and Council, 1989) on the positive relationship between expectancies and pain reduction, though it is essential to qualify this assertion: first, Kirsch speaks of correlation, while in our case the relationship between the two variables is found by means of a multiple regression analysis; second, Kirsch’s hypothesis is established with regard to the relationship between direct expectancies of pain reduction and actual reduction. The intention to RECOMMEND a treatment would appear to be closer to an indirect measure of such expectancy, even more so considering that the questionnaire included items dealing directly with the expectancies, and which were not found to be related to improvement.

5) The lower the CHRONICITY level, the greater the improvement with the Passive treatment in PT2. Although the chronicity variable is a classic one in this type of study, it does not tend to be a good predictor of treatment effectiveness (Werder and cols., 1981). Nevertheless, the result of our work seems logical from theoretical positions such as that of Bakal and Kaganov (1977), in their psychobiological model based on chronicity. In this model, severity of pain is made to depend on degree of chronicity, and although the relationship between severity and chronicity does not necessarily involve improvement after treatment, it appears logical that the treatment of less chronic, and therefore less severe, headaches is more accessible. Some authors (Nicholson and Blanchard, 1993), reporting recent results in this line, refer to a strong negative correlation between chronicity (in years) and improvement after treatment. Bearing in mind the caution that should be exercised on comparing these data from correlational analyses with data from multiple regression analysis, they would nevertheless appear to concur with the findings of the present study.

6) The lower the BDI score, the greater the improvement with the Passive treatment in PT2. Score in the BDI has been considered a good predictor of the effectiveness of behavioural treatment of headache in some research (e.g., Jacob and cols., 1983). These studies coincide in pointing out that the majority of patients that showed improvement had scores of less than 8. Despite the fact that in our study the majority of subjects in Group II scored above this cut-off point (mean 16.5; range 4-29), the relationship found is nevertheless congruent with the findings of these authors.

7) The lower the level of INTERNALITY, the greater the improvement with the Passive treatment in PT2. This relationship would appear to be congruent with the logic of this type of treatment. That is, subjects who do not perceive that their pain problem depends on their own actions improve more with the Passive treatment since, clearly, it is coherent with their beliefs.

8) A low percentage of ASSOCIATIONS as a good predictor of improvement after Passive treatment in PT2. This relationship contrasts with the involvement of this same variable when the Passive phase was received in first position, that is, in Group I. The obvious differences involved in receiving the Passive treatment in an isolated way or subsequent to the Active treatment may explain this distinction. As pointed out earlier, the involvement of a high percentage of associations in improvement after the Passive phase can most plausibly be interpreted by relating, on the one hand, the fact of not having to use any control strategy in this type of treatment with, on the other, the benefits of such behaviour for those subjects most preoccupied with their self-observation. Nevertheless, when the Passive treatment is applied at the end of the sequence, that is, after the Active treatment, the prediction changes its sign, as in this case the patients that benefit most are those that initially presented low levels of association. It would appear, then, that these subjects benefit most from the Passive treatment only when they have first received the Active treatment, in which they have evidently learned to associate certain situations with the onset or worsening of their pain.

1. That the peculiarities of some of the variables that predict the improvement achieved with the Passive treatment would indicate, by and large, the implication of cognitive aspects such as suggestibility, or person characteristics such as more advanced age or low level of education, which may reinforce the involvement of these cognitive aspects. Also, variables such as low impatience characterize patients that are less "needful" of an Active treatment to remedy their problems of perception of the environment or of their relationship with others. Finally, the intervention of the percentage of associations during the baseline can be explained as a paradoxical effect of Passive treatment in hypervigilant subjects.

2. That the three variables involved in the prediction of the effectiveness of Active treatment are, in sum, variables coherent with the logic of that approach, since they combine different beneficial aspects for the active participation of patients in their treatment process.

3. That the association established by the patient, through the self-record, between the pain episodes and concrete events, is especially relevant. This relevance derives from the fact that this same variable forms part of the two regression equations that predict the efficacy of the Passive treatment and, moreover, has a different sign in each equation. That is, and without forgetting the other set of variables involved in each case in the prediction, there exists the possibility of assigning subjects to one or the other treatment programme according to their differential characteristics in self-observation ability. Thus, an element as useful and common in the behavioural treatment of headache as the observation of situations functionally related to pain may aid our decision on whether to begin with the Active or Passive phase of treatment.

Arena, J.G.; Hannah, S.L.; Bruno, G.M. and Meador, K.J. (1991) Electromyographic biofeedback training for tension headache in the elderly: a prospective study. Biofeedback and Self-Regulation, 16: 379-390.

Arena, J.G.; Hightower, N.E,. and Chong, G.C. (1988) Relaxation therapy for tension headache in the elderly: A prospective study. Psychology and Aging, 1: 96-98.

Arntz, A. and Schmidt, A.J.M. (1989) Perceived control and the experience of pain. In: A. Steptoe and A. Appels (Eds.) Stress, Personal Control and Health. Brussels: Wiley.

Bakal, D.A. and Kaganov, J.A. (1977) Muscle contraction and migraine headache: Psychophysiologic comparison. Headache, 17: 208-215.

Beck, A.T.; Ward, D.H.; Mendelson, M.; Mock, J. and Erbaugh, J. (1961) An inventory for measuring depression. Archives of General Psychiatry, 4: 53-63.

Blanchard, E.B.; Andrasik, F.; Arena, J.G.; Neff, D.F.; Saunders, N,.L.; Jurish, S.E.; Teders, S.J. and Rodichok, L.D. (1983) Psychophysiological responses as predictors of response to behavioral treatment of chronic headache. Behaviour Therapy, 14: 357-374.

Borkovec, T.D. and Nau, S.D. (1972) Credibility of analogue therapy rationales. Journal of Behaviour Therapy and Experimental Psychiatry, 3: 257-260.

Comeche, M.I.; Díaz, M.I. and Vallejo, M.A. (1997) Tratamiento psicológico de las cefaleas: potenciación de los efectos terapéuticos mediante un acercamiento activo y pasivo. Análisis y Modificación de Conducta, 23: 527-548.

Conde, V. and Useros, E. (1975) Adaptación castellana de la escala de evaluación conductual para la depresión de Beck. Revista de Psiquiatría y Psicología Médica de Europa y América, 12: 217-236.

Cram, J.R. (1980) EMG biofeedback and the treatment of tension headaches: A systematic analysis of treatment components. Behaviour Therapy, 11: 699-710.

Chapman, S.L. (1986) A review and clinical perspective on the use of EMG and thermal biofeedback for chronic headaches. Pain, 27: 1-43.

Díaz, A. and Vallejo M.A. (1987) Influencia del placebo en el tratamiento de la cefalea. Estudios de Psicología. Monográfico: El Efecto Placebo, 31: 53-68.

Drummond, P.D. (1985) Predisposing, precipitating and relieving factors in different categories of headache. Headache, 25: 16-22.

Evans, F.J. (1989) The independence of suggestibility, placebo response, and hypnotizability. In: V.A. Gheorghiu, P. Netter, H.J. Eysenck and R. Rosenthal (Eds.). Suggestion and suggestibility, Berlin: Springer-Verlag.

Eysenck, H.J. (1989) Personality, Primary and Secondary Suggestibility and Hypnosis. In: V.A. Gheorghiu, P. Netter, H.J. Eysenck and R. Rosenthal (Eds.). Suggestion and suggestibility, Berlin: Springer-Verlag.

Eysenck, H.J. and Furneaux, W.D. (1945) Primary and secondary suggestibility: an experimental and statistical study. Journal of Experimental Psychology, 35: 485-503.

Fernández-Abascal, E.G. (1992) El Inventario de Actividad de Jenkins, Madrid: Tea.

Fernández-Abascal, E.G. (1994) Intervención comportamental en los trastornos cardiovasculares, Madrid: Fundación Universidad-Empresa.

Flor, H. and Turk, D.C. (1989) Psychophysiology of chronic pain: do chronic pain patients exhibit symptom-specific psychophysiological responses?, Psychological Bulletin, 2: 215-259.

Gale, E.N. and Funch, D.P. (1984) Factors Associated with successful outcome from behavioral therapy for chronic temporomandibular joint (TMJ) pain. Journal of Psychosomatic Research, 28: 441-448.

Härkäpää, K. (1991) Relationships of Psychological Distress and Health Locus of Control Beliefs with the Use of Cognitive and Behavioral Coping Strategies in Low Back Pain Patients. The Clinical Journal of Pain, 7: 275-282.

Hudzinski, L.G. and Levenson, H. (1985) Biofeedback behavioral treatment of headache with locus of control pain analysis: A 20-month retrospective study. Headache, 25: 380-386.

Jacob, R.G.; Turner, S.N.; Szekely, B.C. and Eidelman, B.H. (1983) Predicting outcome of relaxation therapy in headaches: The role of depression. Behaviour Therapy, 14: 457-465.

Jenkins, C.D.; Zyzanski, S.J. and Rosenman, R.H. (1979) Jenkins Activity Survey, New York: Psychological Co.

Jensen, M.P.; Turner, J.A. and Romano, J.M. (1991) Self-efficacy and outcome expectancies: relationship to chronic pain coping strategies and adjustment. Pain, 44: 263-269.

Jones, C. and Page, S. (1986) Locus of Control, Assertiveness an Anxiety as Personality Variables in Stress-Related Headaches. Headache, 26: 369-374.

Kirsch, I. (1985) Response expectancy as a determinant of experience and behaviour. American Psychologist, 40: 1189-1202.

Kirsch, I. (1990) Changing Expectations. A key to effective psychotherapy, Pacific Grove: Brooks/Cole P.C.

Kirsch, I. and Council, J.R. (1989) Response expectancy as a determinant of hypnotic behaviour. In: N.P. Spanos and J.F. Chaves (Eds). Hypnosis: The cognitive-behavioral perspective, Buffalo: Prometheus.

Malone, M.D. and Strube, M.J. (1988) Metaanalysis of non-medical treatments for chronic pain, Pain, 34: 231-244.

Nicholson, N.L. and Blanchard, E.B. (1993) A controlled evaluation of behavioral treatment of chronic headache in the elderly. Behaviour Therapy, 24: 395-408.

Rappaport, N.B.; McAnulty, D.P. and Brantley, P.J. (1988) Exploration of the Type A behaviour pattern in chronic headache sufferers. Journal of Consulting and Clinical Psychology, 56: 621-623.

Rudy, R.E.; Kerns, R.D. and Turk, D.C. (1988) Chronic pain and depression: Toward a cognitive-behavioral mediation model, Pain, 35: 129-140.

Shoham-Salomon, V. and Hannah, M.T. (1991) Client-Treatment Interaction in the Study of Differential Change Processes. Journal of Consulting and Clinical Psychology, 59: 217-225.

Spinhoven, P. (1988) Similarities and dissimilarities in hypnotic and non-hypnotic procedures for headache control. American Journal of Clinical Hypnosis, 30: 183-194.

Turk, D.C. (1990) Customizing treatment for chronic pain patients: Who, what and why?. Clinical Journal of Pain, 6: 255-270.

Turk, D.C. and Rudy, T.E. (1992) Cognitive factors and persistent pain: a glimpse into Pandora’s box. Cognitive Therapy and Research, 16: 99-122.

Vallejo, M.A. and Comeche, M.I. (1994) Evaluación y Tratamiento Psicológico del dolor crónico, Madrid: Fundación Universidad-Empresa.

Wallston, K.A.; Wallston, B.S. and DeVellis, R. (1978) Development of the multidimensional health locus of control (MHLC) scales. Health Education Monographs, 6: 160-170.

Weitzenhoffer, A.M. and Hilgard, E.R. (1959) Stanford Hypnotic Susceptibility Scale, Forms A and B, Palo Alto: Consulting Psychologists Press.

Werder, D.S.; Sargent, J.D. and Coyne, L. (1981) MMPI profiles of headache patients using self-regulation to control headache activity. Paper presented at the American Association of Biofeedback Clinicians Meeting, Kansas City.