SMI860A

22/6/2000

14:11

Page 263

Stress Medicine Stress Med. 16: 263±268 (2000)

Pe r s o n a l i t y c h a n g e s i n d u c e d b y a u t o g e n i c t ra i n i n g p rac t i c e Mario A. FarneÁ1,*,{, MD and Noelia Jimenez-MunÄoz2, MD 1

Department of Psychology, University of Bologna, Italy Department of Psychiatry and Medical Psychology, University of Malaga, Spain

2

Summary

Although autogenic training is a relaxation method not regarded as `psychotherapeutic', it may induce psychotherapeutic personality changes. Study participants (N ˆ 60) had undertaken AT because they suffered from minor emotional problems caused by distress. All of them completed a series of inventories on two kinds of personality traits: a trait facilitating the stress response (anxiety) and traits moderating it (sense of personal control, barrier and Ego strength). The inventories were administered before the beginning of the course and after 8 months of constant practice at home. The results indicate a signi®cant decrease in emotional distress signs and anxiety and a signi®cant increase in the traits attenuating the stress response. Copyright # 2000 John Wiley & Sons, Ltd.

Key Words autogenic training; barrier; ego strength; personal control; psychotherapeutic effect; stress moderators Autogenic training (AT) is a psychophysiologically oriented approach generating a relaxation response which is the opposite of the stress response and has been de®ned as the result of homeostatic self-regulatory brain mechanisms.1,2 Autogenic therapy is based on a series of exercises carried out in a mental state called `passive concentration', i.e. a state which implies a casual attitude and functional passivity toward the intended outcome of the concentrative condition. In this state of consciousness, any goaldirected effort, active interest or apprehensiveness *Correspondence to: Dr Mario FarneÁ, Department of Psychology, University of Bologna, Viale Berti Pichat 5, 40127 Bologna, Italy. Tel: ‡39-051-351335. Fax: ‡39-051-243086. {E-mail: [email protected] Contract grant sponsor: Bologna University. Copyright # 2000 John Wiley & Sons, Ltd.

should be avoided. AT functions include the sensations of controlling one's own body and distress signs of danger.1,3 According to the criteria summarized by Smith and Sechrest,4 AT can be considered a `strong therapy', i.e. it has an a priori likelihood of obtaining its intended result. Other anti-stress processes similar to AT and not considered `psychotherapeutic' in nature (e.g. relaxation training and biofeedback) frequently yield personality outcomes such as an improved sense of self-ef®cacy and self-esteem, and what has been called a state of `cultivated low arousal'.5±7 Like these techniques, AT is not traditionally included in more standard psychotherapy approaches. However, anecdotal observations made by one of the authors (MAF) led to the hypothesis that AT does induce stable modi®cations of personality components. These results involve not only the components facilitating and/or following stress (e.g. anxiety), but also those moderating the negative consequences of stressors (e.g. a sense of personal control, PC, which could be compared to the sense of selfef®cacy). Received 17 May 1999 Accepted 15 November 1999

SMI860A

22/6/2000

14:11

Page 264

M. A. FarneÁ and N. Jimenez-MunÄoz Aim and Method The study was undertaken to investigate the in¯uence of AT practice on personality traits, i.e. enduring attributes. In particular, the hypotheses considered were that this kind of therapy contributes (a) to decrease the values of the characteristics which promote and/or typify the stress response, such as low mood states and anxiety, and (b) to increase the values of traits which moderate the stress response, such as a sense of personal control, Ego strength, and barrier. The use of the `barrier' (B) variable was suggested by Fisher's studies,8,9 leading to the description of a personality trait characterized by high levels of goal setting, a great need to persist in tasks and complete them, low suggestibility, self-indulgence and, most pertinent for the present study, an adequate capability to tolerate stress. Participants The participants comprised 77 self-referred patients (40 men and 37 women), aged between 20 and 61 years (mean age ˆ 38.65, SD ˆ 6.18). All suffered from minor psychological problems caused by distress. Procedure and materials The participants underwent a series of tests during the ®rst two interviews; thereafter, they began the AT course individually (standard exercises), according to the programme described by Schultz and Luthe.1 The same person (MAF) conducted the course, thus reducing the effects of therapist bias. The course lasted a mean period of 3 months, with one session a week with the therapist. After a further 8 months of practice at home, all the tests completed during the ®rst interviews were sent to the participants, together with a form inquiring into the frequency of their AT home sessions. A prepaid and preaddressed envelope was also enclosed to facilitate the return of the questionnaires. The criteria for inclusion in the study were that, during the 8-month period, the participants had practiced not less than four times a week. Seventeen of the participants were excluded, leaving 60 (30 men and 30 women) who completed the study. 264

Copyright # 2000 John Wiley & Sons, Ltd.

Six tests were used, detailed as follows. (a) The Pro®le of Mood States (POMS) Inventory,10,11 which measures ®ve mood factors (anxiety, depression, hostility, fatigue, and confusion± bewilderment) by means of 49 adjectives. A Total Mood Disturbance (TMD) score was obtained by summing the scores of the ®ve factors. The test±retest reliability coef®cient of the ®ve scales ranged between 0.65 and 0.74.10 (b) The State±Trait Anxiety Inventory (STAI),12 of which we used only the 20 items assessing trait anxiety, i.e. a stable personality characteristic (test±retest reliability coef®cient ˆ 0.79). (c) The Personal Control (PC) Scale measures the con®dence in mastering the course of one's existence: a coping resource which moderates the stress response.13 The 11 items of the scale were derived from Mirels'14 abbreviated form of Rotter's Internal versus External Control Scale.15 The respondents had to choose between statements like `In my case getting what I want has little or nothing to do with luck' and `Many times we might just as well decide what to do by ¯ipping a coin'. In order to mislead them as to the aim of the scale, four statements similar to, but not the same as, those of Rotter's were added. Mirels does not report the reliability values of the scale. Tests (d), (e) and ( f) were developed by variously recombining part of the 567 items from the second edition of the Minnesota Multiphasic Personality Inventory (MMPI-2).16 (d) The Anxiety Scale (ANX) is a 23-item test which measures anxiety as a trait.16 It covers a wide variety of anxious behaviours. In particular, the respondent with high ANX scores has a high trait anxiety level and may be considered a `worrier', s/he is predisposed to experience great emotional discomfort in stressful situations (test± retest reliability coef®cient ˆ 0.88). (e) The Ego Strength (Es) Scale is a 52-item test originally developed to predict the response of neurotic patients to individual psychotherapy.16 Now there is a tendency to consider high Es-scale scorers as better adjusted psychologically, with a secure sense of reality, and more able to cope with problems and stressors in their life situations (test±retest reliability coef®cient ˆ 0.72). ( f) The Barrier (B) Scale is a 30-item test developed by FarneÁ et al.17 to measure the `barrier' personality trait.8,9 The validity of the scale was tested by having 104 university students complete it and Stress Med. 16: 263±268 (2000)

SMI860A

22/6/2000

14:11

Page 265

Personality changes induced by AT practice respond to a series of inkblot reproductions, which are a useful means to assess the B characteristic.9 The results of the two tests were signi®cantly correlated (r ˆ 0.74, p 5 0.0118). The test±retest reliability coef®cient for a sample of 215 university students (121 women and 94 men) and for a retest interval of 4 months was 0.68. The personal characteristics assessed by these tests were of the kind either facilitating and/or typifying the stress response (tests (a), (b) and (d)) or moderating it (tests (c), (e) and ( f)). Results Of the 17 excluded subjects, ®ve no longer practiced AT, while the others exercised no more than three times a week; the total number

of subjects excluded represents 20 per cent of the initial cohort. Statistical analysis was therefore conducted on the 60 remaining subjects. Male and female scores on self-report measures are presented in Table I. Signi®cant differences were found on TMD-1, where women reported a higher result than men (t ˆ ÿ2.86, p 5 0.006). Differences were also found on B-1 and Es-2 where men, in contrast, reported a higher result than women (t ˆ 2.34, p 5 0.023 and t ˆ 2.22, p 5 0.030, respectively). In these three cases of discrepancies, we analyzed the results obtained by men and women separately. Differences between the self-report measures obtained before learning AT and after 8 months of practice are shown in Table II. The t-test for paired samples was used: in each comparison, the values led to a rejection of the hypothesis of equality with signi®cance levels lower than 0.001.

Table I Ð Differences between men and women on self-report measures (two-tail t-test for paired samples). Men Mean/SD

Women Mean/SD

B-1

19.633 3.439

16.933 5.298

B-2

22.733 4.409

Es-1

t

df

p

2.34

58

50.023

20.867 5.387

1.47

58

50.147

39.533 8.835

35.400 8.063

1.89

58

50.063

Es-2

47.567 5.782

42.900 9.963

2.22

58

50.030

PC-1

5.200 2.140

4.667

1.01

58

50.316

PC-2

8.000 2.133

7.000 1.965

1.89

58

50.064

ANX-1

22.600 9.250

23.633 7.586

ÿ0.47

58

50.638

ANX-2

15.100 9.095

15.600 7.454

ÿ0.23

58

50.817

STAI-1

48.000 10.052

51.767 9.825

ÿ1.47

58

50.148

STAI-2

37.933 7.853

42.333 9.490

ÿ1.72

58

50.090

TMD-1

77.967 14.922

94.667 28.292

ÿ2.86

58

50.006

TMD-2

59.267 22.659

54.351 18.933

0.89

56

50.376

Note: B ˆ Barrier; Es ˆ Ego strength; PC ˆ Personal Control; ANX ˆ Anxiety Score; STAI ˆ State±Trait Anxiety Inventory; TMD ˆ Total Mood Disturbance.

Copyright # 2000 John Wiley & Sons, Ltd.

Stress Med. 16: 263±268 (2000)

265

SMI860A

22/6/2000

14:11

Page 266

M. A. FarneÁ and N. Jimenez-MunÄoz Table II Ð Differences between the self-report measures obtained before the learning of AT and after an 8-month practice (two-tail t-test for paired samples). Before AT Mean/SD B

m

After 8 months Mean/SD

t

df

p

4.69

29

50.001

3.76

29

50.001

6.70

29

50.001

4.97

29

50.001

19.633 3.439 16.933 5.298

22.733 4.409 20.866 5.387

39.533 8.835 35.400 8.063

47.566 5.782 42.900 9.963

4.933 2.041

7.500 2.095

12.28

59

50.001

ANX

23.117 8.403

15.350 8.248

ÿ8.41

59

50.001

STAI

49.883 10.036

40.133 10.050

7.00

59

50.001

77.966 14.922 98.214 25.744

59.266 22.659 54.357 18.933

ÿ3.75

27

50.001

ÿ6.66

27

50.001

w Es

m

w PC

TMD m w

Note: m ˆ men; w ˆ women; B ˆ Barrier; Es ˆ Ego strength; PC ˆ Personal Control; ANX ˆ Anxiety Score; STAI ˆ State± Trait Anxiety Inventory; TMD ˆ Total Mood Disturbance.

The measures obtained for B and Es were correlated in both phases of the study (Pearson r ˆ 0.3752, p ˆ 0.003, and r ˆ 0.3785, p ˆ 0.003).

Discussion The decrease in the personality traits facilitating the stress response or generated by it (as assessed by POMS and STAI) indicates that AT practice alone is suf®cient to keep these characteristics at normal levels. This result is explained by the fact that the symptoms were a mere reaction to a distress condition, without a neurotic basis. AT alone is not suf®cient in cases of neurosis, but has to be combined with other therapies.1 The POMS results are practically the same as those obtained in another study at the end of the 3-month AT course.19 The increased values of the personality traits attenuating the stress response seem to be due to a psychotherapeutic effect of AT. An 11-month practice period is somewhat short and 266

Copyright # 2000 John Wiley & Sons, Ltd.

the interviews with the specialist were limited to the introductory 3 months, and were devoted to teaching the exercises. However, it is possible to conclude that deep processes are already set in motion during this period. The nature of these processes is a matter of conjecture. We have seen that AT can be regarded as a strong treatment, but it specializes in distress signs of danger. In the case of the personality changes observed here, we hypothesize that the effective variable associated with AT is the training commitment. This conjecture is substantiated by the fact that, thanks to AT practice, the subject realizes that s/he can obtain psychosomatic improvements and, consequently, this may improve his or her self-appraisal of competence. In the words of Linden:2 Learning to relax via AT not only has desirable acute effects, but tends to generalize insofar as patients typically learn to perceive themselves as being at the control of their stress responses; this Stress Med. 16: 263±268 (2000)

SMI860A

22/6/2000

14:11

Page 267

Personality changes induced by AT practice in turn has a positive impact on the way they perceive potential stress triggers and how they respond to them ( p. 224; our italics). Langen,3 in turn, concludes that: the subject, thanks to AT practice, can regulate blood circulation and other body functions usually considered unintentional, autonomous . . . The control of these functions is a means to reach major objectives: to in¯uence one's attitudes toward oneself and life and, moreover, to work upon one's personality ( p. 42; our italics). More generally, Bandura5 maintains that the enhancement of perceived ef®cacy is the decisive mechanism common to different but effective treatment modalities. For instance, he summarizes the results of his studies on `participant modeling', a psychological treatment developed from learning principles, thus: `Successful performance provides the most in¯uential source of ef®cacy information because it is based on authentic experiences of mastery' 5[ p. 640]. He claims that perceived self-ef®cacy in turn in¯uences behaviour, thought patterns and emotional reactions; it also `determines how much effort individuals will expend, and how long they will persevere in the face of obstacles and discouraging experiences' 5[ p. 640]. In other words, perceived self-ef®cacy fosters initiation and persistence of coping and the factors affecting coping resources. This assumption is supported by an experiment carried out by Bandura et al.7 on patients during different phase of participant modeling treatment and, consequently, with different degrees of perceived self-ef®cacy. The conclusion was that the weaker the sense of controllability and self-ef®cacy, the greater the catecholamine secretion as an index of the acute stress response. In the present study, improved PC is a direct outcome of the processes mentioned here.2,3,5,7 The same conclusion might be drawn for the other two personality variables analyzed (Es and B), since the de®nitions of these traits coincide with Linden's,2 Langen's3 and Bandura's5 conclusions on the control of the stress response and on the sense of self-ef®cacy. A previous study Copyright # 2000 John Wiley & Sons, Ltd.

carried out with 95 university students showed that the measures obtained from the PC, Es and B tests are correlated.20 These ®ndings and considerations partly explain why a control group was not deemed necessary in this study. In addition, weighing the ethical problems involved in withholding of treatment to patients, we are convinced that the control groups can be represented by the `normal' (i.e., without any apparent disturbance) subjects used in evaluating the test±retest reliability of the measures we adopted. Our ®ndings contrast with studies on compliance in medical practice, where treatment nonadherence generally ranges from 30 to 60 per cent.21 In addition, a study on compliance during the AT course2 concluded that 29 per cent of patients do not follow the instructions. In contrast, the fact that only 20 per cent of our subjects no longer exercised regularly 8 months after the end of the AT course indicates a good treatment adherence. There are two main reasons for this. First, the subjects had volunteered for the AT course, and this increases the probability of lasting behaviour. Second, during the 3 months of the course there was a one-to-one therapeutic relationship that may have enhanced the patient's adherence. The positiveness of this relationship is also substantiated by the fact that, after the ®rst interviews and before the beginning of the AT course, there was a signi®cant decrease in the distress symptoms.19 This is evidence of the phenomenon that a positive doctor±patient relationship in itself has a curative effect.22 The study expands the notion of AT from a relaxation and anti-stress technique to a means of modifying personality traits. In fact, the resulting changes are more typical of a psychotherapeutic process. In our case, the process has two advantages: (a) deep-seated changes in different (albeit converging) aspects of personality in a short period of time, which are not always feasible in short-term psychotherapy, and (b) these results are self-generated (i.e. autogenic), the intervention of the therapist being limited to the ®rst 3 months of training. Acknowledgements

This research was supported by funds from the Associazione Industriali della Provincia di Stress Med. 16: 263±268 (2000)

267

SMI860A

22/6/2000

14:11

Page 268

M. A. FarneÁ and N. Jimenez-MunÄoz Bologna Ð Con®ndustria. Dr N. Jimenez-MunÄoz was the recipient of an Erasmus grant for 1997±98 at Bologna University.

References 1. Schultz JH, Luthe W. Autogenic Therapy, vol. 1. Grune and Stratton: New York, 1969. 2. Linden W. The autogenic training of J. H. Schultz. In Principles and Practice of Stress Management (2nd edn), Lehrer PM, Woolfolk RL (eds). The Guilford Press: New York, 1993; 205±229. È bungshaft fuÈr das Autogene Training 3. Langen D. U [Manual for the Autogenic Training Exercises] (19th edn). Georg Thieme Verlag: Stuttgart, 1980. 4. Smith B, Sechrest L. Treatment of aptitude  treatment interaction. In Methodological Issues and Strategies in Clinical Research, Kazdin AE (ed.). American Psychological Association: Washington, DC, 1992; 557±584. 5. Bandura A. Perceived self-ef®cacy: An explanatory mechanism of behavioral change. In Psychology (3rd edn), Lindzey G, Thompson RF, Spring B (eds). Worth Publishers: New York, 1988; 640±641. 6. Everly GS. A Clinical Guide to the Treatment of the Human Stress Response. Plenum Press: New York, 1992. 7. Bandura A, Taylor CB, Williams SL, Mefford IN, Barchas JD. Catecholamine secretion as a function of perceived coping self-ef®cacy. J. Consult. Clin. Psychol. 1985; 53: 406±414. 8. Fisher S. Development and Structure of the Body Image. Lawrence Erlbaum: Hillsdale, NJ, 1986. 9. Fisher S, Cleveland SE. Body Image and Personality. Dover: New York, 1968.

268

Copyright # 2000 John Wiley & Sons, Ltd.

10. McNair DM, Lorr M, Droppleman LF. Manual for the Pro®le of Mood States (POMS). Educational and Industrial Testing Service: San Diego, CA, 1981. 11. FarneÁ MA, Sebellico A, Gnugnoli D. POMS: Adattamento Italiano. Organizzazioni Speciali: Firenze, 1991. 12. Spielberger C, Gorsuch R, Lushene R. The STAI Manual. Consulting Psychologists Press: Palo Alto, CA, 1970. 13. Lazarus RS, Folkman S. Stress, Appraisal and Coping. Springer Publishing Co.: New York, 1984. 14. Mirels HL. Dimensions of internal versus external control. J. Consult. Clin. Psychol. 1970; 34: 226±228. 15. Rotter JB. Generalized expectancies for internal versus external control of reinforcement. Psychol. Monograph 1966; 80(609). 16. Graham JR. MMPI-2: Assessing Personality and Psychopathology (2nd edn). Oxford University Press: New York, 1993. 17. FarneÁ MA, Sebellico A, Gnugnoli D. An objective measure of the Barrier personality trait: results of an inquiry. Boll. Soc. Ital. Biol. Sper. 1986; 42: 1165±1169. 18. FarneÁ MA, Cutajar R, Sebellico A. The Barrier inventory: a validity study. Boll. Soc. Ital. Biol. Sper. 1990; 46: 387±390. 19. FarneÁ MA, Gnugnoli D. Autogenic training immediate effects on emotional distress symptoms. Stress Med. 2000; 16: 259±261. 20. FarneÁ MA, Sebellico A, Gnugnoli D, Corallo A. Personality variables as moderators between hassles and subjective indications of distress. Stress Med. 1992; 8: 161±165. 21. Meichenbaum D, Turk DC. Facilitating Treatment Adherence. Plenum Press: New York, 1987. 22. Balint M. The Doctor, his Patient and the Illness. Pittman Medical Publ.: London, 1957.

Stress Med. 16: 263±268 (2000)