- Email: [email protected]
Ability to speak at the age of 1 year and alexithymia 30 years later
Journal of Psychosomatic Research 54 (2003) 491 – 495
Ability to speak at the age of 1 year and alexithymia 30 years later Pirkko Kokkonena,*, Juha Veijolaa, Juha T. Karvonena, Kristian La¨ksya, Jari Jokelainena, Marjo-Riitta Ja¨rvelinb,c, Matti Joukamaaa a Department of Psychiatry, Oulu University Hospital, Oulu, Finland Department of Public Health, Science and General Practice, University of Oulu, Oulu, Finland c Department of Epidemiology and Public Health, Imperial College School of Medicine, London, UK b
Received 30 October 2001; accepted 29 April 2002
Abstract Objective: We studied the association between speech development in the first year of life and alexithymia in young adulthood. Methods: The study forms a part of the Northern Finland 1966 Birth Cohort. The original material consisted of all liveborn children in the provinces of Lapland and Oulu in Finland with an expected delivery date during 1966. The comprehensive data collection began during the antenatal phase. In 1997, a 31-year follow-up study was made on a part of the initial sample. The 20-item version of the Toronto Alexithymia Scale (TAS-20) was given to 5983 subjects. Of them, 84% returned the questionnaire properly filled in. The ability to talk was classified according to whether the child spoke no words, one or two words, or three or more words at the age of 1 year. Statistical analyses on the association
between the ability to speak at the age of 1 year and alexithymia at the age of 31 years were performed, adjusted for birth weight, mother’s parity, place of residence and wantedness of pregnancy. Results: The mean of the total TAS score was lowest among early speakers and for both genders separately. The differences were statistically significant. A parallel significant difference was found among males on TAS Factors 2 and 3 and in case of females on TAS Factors 1 and 3. Conclusions: We found evidence for an association between speaking development in early childhood and later alexithymia. Our results support the theory that alexithymia may be a developmental process starting in early childhood and reinforcing itself in a social context. D 2003 Elsevier Science Inc. All rights reserved.
Keywords: Alexithymia; TAS-20; Cohort study; Prospective study; Speech development
Introduction Alexithymia, coined by Sifneos [1] in 1973, is a multifaceted construct including the following core features: difficulty in recognizing and verbalizing subjective feelings, a cognitive style characterized by absence of fantasy life and a tendency to recount the minute details of external events [2]. Although the clinical relevance of the alexithymia construct has been supported in a number of studies, there has been controversy and debate as to whether it is a stable personality trait, a transient state secondary to a stressful situation or a coping response to chronic illness [3]. It has even been postulated that alexithymia might just be a social or cultural phenomenon [4,5].
* Corresponding author. Department of Psychiatry, University of Oulu, Peltolantie 5, Oulu 90210, Finland. Fax: +358-8-333-167. E-mail address: [email protected] (P. Kokkonen).
Freyberger [6] conceptualized two kinds of alexithymia: primary alexithymia as a personality trait and secondary alexithymia as a state reaction resulting from stressful circumstances. Sifneos [7] attributes the term secondary alexithymia to alexithymic characteristics resulting from developmental arrests, massive psychological trauma, sociocultural factors or psychodynamic factors. Primary alexithymia is attributed to neurobiological deficits [7]. Sifneos [7] suggested that a disconnection between the limbic system and the verbal centers of the neocortex may result in primary alexithymia. Considering these neurobiological and neuropsychological theories on the origin of alexithymia, it would be interesting to investigate the possible associations between early speech development and later alexithymia. No longitudinal follow-up studies considering early developmental milestones and alexithymia have been conducted. The aim of this study was to examine the association between alexithymia and early speech development in childhood during the first year of life.
0022-3999/03/$ – see front matter D 2003 Elsevier Science Inc. All rights reserved. doi:10.1016/S0022-3999(02)00465-8
492
P. Kokkonen et al. / Journal of Psychosomatic Research 54 (2003) 491–495
Materials and methods Design This study forms a part of the Northern Finland 1966 Birth Cohort Project, which is a prospective, long-term, follow-up study. The original sample was collected from a geographically defined area of the two northernmost provinces of Finland. It consisted of an unselected, general population-based birth cohort of 12,058 live births in Northern Finland whose expected date on delivery fell in 1966, representing 96% of all such births [8]. The comprehensive data collection began during the antenatal phase. Subsequent to the baseline survey, a 1-year follow-up and a 14-year follow-up of the offspring have been performed. The detailed description of the study population and general design is presented elsewhere [8,9]. In 1997, a 31-year follow-up study was conducted through various questionnaires, interviews and clinical examinations. This follow-up study was composed of different subprojects and the alexithymia investigation formed a part of them. By using the Toronto Alexithymia Scale 20 (TAS-20) questionnaire, we had a possibility to assess the occurrence of alexithymia among these young adults. Using previously collected data, there was an opportunity to investigate connections between early childhood development and alexithymia in adulthood. The Ethics Committee of The Faculty of Medicine, University of Oulu, has approved this study. Subjects In 1997, 8417 of the original sample were living in the two northernmost provinces of Finland and in the capital city (Helsinki) area. As part of the 31-year follow-up study, an invitation to a field study was sent home to these cohort members. The field study consisted of a clinical examination, various questionnaires and interviews. Of the subjects contacted, 5983 participated in the field study and gave their written informed consent. A questionnaire including the TAS-20 was given to them to fill in at home and to return by mail. The TAS-20 questionnaire was returned properly by 84% (5028) of the subjects who had received it. The losses differed from the participants by being more commonly male, employed, married and less educated than the subjects participating [10]. Methods Of the different methods for measuring alexithymia, the TAS-20 is widely used and carefully validated. Its internal consistency, test – retest reliability, convergent, discriminant and concurrent validity have been demonstrated to be good [11,12]. The validity of the Finnish version of the TAS-20 has also been shown to be good [13]. The TAS-20 consists of 20 statements, each rated on a Likert-type scale from 1
(strongly disagree) to 5 (strongly agree) for a maximum score of 100. The TAS-20 has a three-factor structure: Factor 1 assesses difficulty in identifying feelings; Factor 2 concerns itself with difficulty in describing feelings; and Factor 3 reflects externally oriented thinking. Based on the TAS-20, the total score and the scores of TAS Factors 1– 3 (hereafter TASF1, TASF2, TASF3) were calculated. According to the developers’ recommendation, subjects with a total TAS score 61 were considered alexithymic, and those with a score 51 were considered nonalexithymic [3]. In this study, subjects with total TAS scores from 52 to 60 were considered probably alexithymic. The first follow-up study considering the growth, development and health status of the children of this cohort was conducted at child welfare centers when the subjects were 1 year old. Data on the number of words spoken by the child at 12 months were obtained for 95% of the children [9]. The ability to talk was classified according to whether the child spoke no words (late speaker), one or two words (medium speaker), three or more words (early speaker). Statistical methods The mean score of the TAS-20 and its three factors was calculated in each speaking category and also stratified by gender. The statistical significance of differences between these means was evaluated using multivariate analysis of variance (MANOVA), with post hoc Dunnett’s test, controlled for the following variables: birth weight ( < 2500 g, 2500 g), mother’s parity in 1966 (1, 2 3, 4 5, 6), place of residence in 1966 (urban, rural) and wantedness of pregnancy (wanted/mistimed, unwanted). The variables were selected based on earlier studies in this cohort [8,9]. We also fit the proportional odds model [14] in order to analyse the effect of the speaking to the tricotomized TAS score (see above). The score test was used for assessing validity of the proportionality assumptions. The data analysis was generated using the SAS/STAT software, version 8, of the SAS system for windows [15,16].
Results The prevalence of alexithymia was lowest among subjects speaking three or more words at the age of 1 year (Table 1). The association was shown both in male subjects and in female subjects. Among males, the effect was statistically significant, and among female subjects, the same trend was seen. The score test for the proportional odds assumptions was fulfilled in all subjects (c2 = 7.0, df = 8, P = .536), and in males (c2 = 7.8, df = 8, P = .451) and females (c2 = 6.6, df = 8, P = .582). After adjusting for confounding factors, the statistical significance remained among male subjects and for the whole sample. Mean total scores of alexithymia were associated with speaking development both in males and females, e.g., the
P. Kokkonen et al. / Journal of Psychosomatic Research 54 (2003) 491–495
493
Table 1 Prevalence of alexithymia in adulthood and speaking ability at the age of 1 year Three or more words
One or two words
No words
n
%
n
%
n
%
Analysis c2a
P
c2b
P
All subjects TAS 61 52 < TAS < 60 TAS 51
159 400 1997
6.2 15.7 78.1
86 189 739
8.5 18.6 72.9
80 208 743
7.8 20.2 72.1
21.0
< .001
15.8
< .001
Men TAS 61 52 < TAS < 60 TAS 51
91 206 763
8.6 19.4 72.0
44 110 299
9.7 24.3 66.0
55 136 359
10.0 24.7 65.3
10.1
.038
7.6
.022
Women TAS 61 52 < TAS < 60 TAS 51
68 194 1234
4.6 13.0 82.5
42 79 440
7.5 14.1 78.4
25 72 384
5.2 15.6 79.8
8.8
.067
3.6
.164
TAS = Toronto Alexithymia Scale 20 total score. a Chi-square test, df = 4. b Wald chi-square test by proportional odds modelling after controlled potential confounding factors, df = 2.
mean of the total TAS score was lowest in early speakers in all subjects and in both genders separately (Table 2). The total TAS scores were significantly higher among medium and late speakers than among early speakers. No statistical significance was found on TASF1 among male subjects and on TASF2 among female subjects. Otherwise, the TAS factors were associated with speaking ability in the same way as the total TAS scores. After adjusting for confounding factors, TASF1 was still found to be nonsignificant among
male subjects ( P = .206) and TASF2 among female subjects ( P =.184). After ANOVA and Dunnett’s post hoc test, there was no statistical significance between late speakers and early speakers on TASF1 among all subjects, but medium speakers had higher scores than early speakers (ANOVA, P =.024; Dunnett’s test, P =.018). Among female subjects, medium speakers had higher TASF1 scores than early speakers (ANOVA, P =.013; Dunnett’s test, P =.011), but
Table 2 The TAS-20 scores and speaking ability at the age of 1 year Three or more words
One or two words
No words
Mean
S.D.
Mean
S.D.
Mean
S.D.
F
P
Fa
P
All TAS TASF1 TASF2 TASF3
43.6 12.8 11.3 19.6
10.3 4.8 4.1 4.6
45.2 13.3 11.9 20.1
10.4 4.8 4.3 4.6
45.7 13.2 11.9 20.6
10.3 4.8 4.2 4.7
17.56b 4.89 12.84 18.71
< .001 .008 < .001 < .001
14.82c 3.74 11.21 16.46
< .001 .024 < .001 < .001
Men TAS TASF1 TASF2 TASF3
46.0 12.4 12.3 21.3
10.2 4.8 4.1 4.3
47.6 12.8 13.0 21.8
9.9 4.6 4.3 4.3
47.8 12.9 12.8 22.1
10.1 4.8 4.3 4.4
6.18d 1.16 6.89 5.02
.002 .313 .001 .007
5.21e 0.77 6.34 4.45
.001 .463 .002 .012
Women TAS TASF1 TASF2 TASF3
41.9 13.0 10.5 18.3
10.1 4.8 3.9 4.4
43.3 13.7 11.0 18.7
10.4 4.9 4.1 4.3
43.2 13.5 10.8 18.9
10.1 4.8 3.9 4.5
6.29f 5.12 2.52 4.47
.002 .006 .080 .012
4.52g 4.32 1.57 3.04
.011 .013 .207 .048
a b c d e f g
Analysis of variance
Adjusted for birth weight, mother’s parity, place of residence and wantedness of pregnancy. df (2, 4505). df (2, 4499). df (2, 2019). df (2, 2013). df (2, 2483). df (2, 2477).
494
P. Kokkonen et al. / Journal of Psychosomatic Research 54 (2003) 491–495
late speakers and early speakers did not differ from each other. On TASF2, early and medium speakers had higher scores than early speakers among all subjects (ANOVA, P < .001; Dunnett’s test, later vs. early, P < .001, and medium vs. early, P < .001). Among male subjects, both later and medium speakers differed from early speakers on TASF2 (ANOVA, P =.002; Dunnett’s test, later vs. early, P =.019, and medium vs. early, P =.003). On TASF3, early and medium speakers had higher scores among all subjects (ANOVA, P < .001; Dunnett’s test, later vs. early, P < .001, and medium vs. early, P =.032). Among male and female subjects, early speakers had also lowest scores on TASF3 (ANOVA, P =.012 among males, P =.048 among females; Dunnett’s test, later vs. early, P =.006 in males and P =.037 in females). No statistical significance was found between early and medium speakers.
Discussion We found evidence for an association between alexithymia in adulthood and timing of speech development. The early speakers, e.g., children who could speak three or more words at the age of 1 year, had the lowest mean score of the total TAS among all subjects and among both genders separately. The TAS factors were associated with speaking ability in the same way as the total TAS scores, but no statistical significance was found on TASF1 among male subjects and on TASF2 among female subjects. After adjusting for confounding factors, the abovementioned associations still remained statistically significant. Among male subjects, prevalence of later alexithymia was lowest in early speakers. Among female subjects, there was a similar trend but it fell below conventional statistical significance. TASF3 was associated with speaking development in all subjects and in both genders separately. There were gender differences considering the association with speech development on TASF1 and TASF2. Lane et al. [17] have speculated that for certain subjects, self-reported ratings on TASF1 and TASF2 may be less accurate than those on the TASF3. They suggested that high alexithymic subjects rate themselves unreliably because of their own lack of awareness of the deficits [17]. Suslow et al. [18] have proposed that TASF2, ‘‘Difficulties Describing Feelings,’’ does not measure primarily an impairment in symbolizing emotions. There are only few studies considering antecedents of alexithymia [19 – 26]. Diminished expressiveness in childhood [21,22], mother’s poor care [19,20], maternal alexithymia and disturbed family functioning [23] have all been shown to be associated with alexithymia. Kooiman et al. [25] could not confirm the relation of alexithymia to disturbances in the early parent –child relationship, while Berenbaum [24] found an association between childhood abuse and alexithymia in adulthood. All these studies have been cross-sectional and consequently vulnerable to recall
bias. The strength of our study was the prospective design: all developmental data were collected at the age of 1 year and measuring of alexithymia using the TAS-20 was conducted at the age of 31 years. In the present study, we found an association between later speech development and alexithymia in adulthood. In a study with nonclinical sample, Parker et al. [27] observed a slight correlation between alexithymia, on one hand, and intelligence score and mother’s education, on the other. Therefore, we also analysed separately the relationship between speaking development and mother’s education. No association was found. In the 1960s, women were not very highly educated in Finland. In the Northern Finland 1966 Birth Cohort, 86% of the mothers had less than 8 years of education [8]. This may be one reason why no significant association was found. Lamberty and Holt [28] suggested that an individual’s difficulty with verbalization of emotional material might be related to specific developmental verbal deficits, independent of other proposed etiological factors. The deficits underlying alexithymia have been attributed, at least in part, to an arrest in affect development during early childhood [3,29]. As Kauhanen [30] has stated, alexithymia may be ‘‘an accumulative process starting in early childhood and developing and reinforcing itself in a social context.’’ However, etiology of alexithymia probably involves multiple factors including constitutionally inherited variations in brain organization and deficiencies in the early family and social environment [3]. There are some limitations in our study. Firstly, all the subjects were of the same age. A positive association between alexithymia and age appears likely [31]. However, the sample was large and consisted of a population of normal young adults. To our knowledge, there exists no prospective longitudinal population-based study considering associations between alexithymia and development in early life. Secondly, because of our large cohort, the effect sizes are extremely small, which ought to be considered when interpreting the results. Thirdly, regarding speech development, the examination at 1 year of age may be too early to detect an effect. There is no pathology or real speech delay at the age of 1 year. However, early developmental differences in speaking ability were apparent in children who later developed alexithymic features. The TAS-20 was the measure of alexithymia. To date, the TAS-20 is the most commonly used and best validated measure of alexithymia [17]. Recently, however, self-rating measures like the TAS-20 have been seriously criticised [17,18]. As we already noted, Lane et al. [17] have proposed that an inherent difficulty may exist in the use of self-reports to measure something (i.e., alexithymia) that some of the respondents do not comprehend. It is possible that accurate responses in the TAS-20 may require the absence of extreme degrees of the trait that it measures [17]. So it may be possible that the TAS-20 cannot detect the most severe cases of alexithymia [32].
P. Kokkonen et al. / Journal of Psychosomatic Research 54 (2003) 491–495
The focus of our study was the developmental course of alexithymia. The concept of alexithymia has much to offer heuristically [33]. Well-conducted, longitudinal research from childhood to adulthood has been lacking. Developmental information might improve our understanding of alexithymia. In fact, alexithymia is conceptualized as a dimensional construct rather than a categorical phenomenon [3].
Acknowledgments This work was supported by grants from the Signe and Ane Gyllenberg Foundation and the Academy of Finland.
References [1] Sifneos PE. The prevalence of alexithymic characteristics in psychosomatic patients. Psychother Psychosom 1973;22:255 – 62. [2] Taylor GJ. Alexithymia: concept, measurement, and implications for treatment. Am J Psychiatry 1984;141:725 – 32. [3] Taylor GJ, Bagby RM, Parker JDA, editors. Disorders of affect regulation: alexithymia in medical and psychiatric illness. Cambridge: Cambridge Univ. Press, 1997. [4] Borens R, Grosse-Schulte E, Jaensch W, Kortemme KH. Is ‘alexithymia’ but a social phenomenon? Psychother Psychosom 1977;28: 193 – 8. [5] Kirmayer LJ. Languages of suffering and healing: alexithymia as a social and cultural process. Transcult Psychiatr Res Rev 1987;24: 119 – 36. [6] Freyberger H. Supportive psychotherapeutic techniques in primary and secondary alexithymia. Psychother Psychosom 1977;28:337 – 43. [7] Sifneos PE. Alexithymia and its relationship to hemispheric specialization, affect, and creativity. Psychiatr Clin North Am 1988;11: 287 – 92. [8] Rantakallio P. A longitudinal study of the Northern Finland birth cohort 1966. Paediatr Perinat Epidemiol 1988;2:59 – 88. [9] Rantakallio P, von Wendt L, Ma¨kinen H. Influence of social background on psychomotor development in the first year of life and its correlation with later intellectual capacity: a prospective cohort study. Early Hum Dev 1985;11:141 – 8. [10] Kokkonen P, Karvonen JT, Veijola J, La¨ksy K, Jokelainen J, Ja¨rvelin MR, Joukamaa M. Prevalence and sociodemographic correlates of alexithymia in a population sample of young adults. Comp Psychiatry 2001;42:471 – 6. [11] Bagby RM, Parker JDA, Taylor GJ. The twenty-item Toronto Alexithymia Scale: I. Item selection and cross-validation of the factor structure. J Psychosom Res 1994;38:23 – 32. [12] Bagby RM, Taylor GJ, Parker JDA. The twenty-item Toronto Alexithymia Scale: II. Convergent, discriminant, and concurrent validity. J Psychosom Res 1994;38:33 – 40.
495
[13] Joukamaa M, Miettunen J, Kokkonen P, Koskinen M, Julkunen J, Kauhanen J, Jokelainen J, Veijola J, La¨ksy K, Ja¨rvelin MR. The Finnish version of the 20-item Toronto Alexithymia Scale. Nord J Psychiatry 2001;2:123 – 7. [14] McCullah P. Regression models for ordinal data. J R Stat Soc, B 1980;42:109 – 42. [15] SAS Institute. SAS/STAT user’s guide, Version 8. Gary (NC): SAS Institute, 1999. [16] Allison PD. Logistic regression using the SAS system: theory and application. Gary (NC): SAS Institute, 1999. [17] Lane RD, Sechrest L, Riedel R. Sociodemographic correlates of alexithymia. Comp Psychiatry 1998;39:377 – 85. [18] Suslow T, Donges US, Kersting A, Arolt V. 20-Item Toronto Alexithymia Scale: do difficulties describing feelings assess proneness to shame instead of difficulties symbolizing emotions? Scand J Psychol 2000;41:329 – 34. [19] Fukunishi I, Kawamura T, Ago Y, Sei H, Morita Y, Rahe RH. Mothers’ low care in the development of alexithymia: a preliminary study in Japanese college students’. Psychol Rep 1997;80:143 – 6. [20] Fukunishi I, Sei H, Morita Y, Rahe RH. Sympathetic activity in alexithymics with mother’s low care. J Psychosom Res 1999;46:579 – 89. [21] Kench S, Irwin HJ. Alexithymia and childhood family environment. J Clin Psychol 2000;56:737 – 45. [22] Berenbaum H, James T. Correlates and retrospectively reported antecedents of alexithymia. Psychosom Med 1994;56:353 – 9. [23] Lumley MA, Mader C, Gramzow J, Papineau KTI. Family factors related to alexithymia characteristics. Psychosom Med 1996;58: 211 – 6. [24] Berenbaum H. Childhood abuse, alexithymia and personality disorder. J Psychosom Res 1996;41:585 – 95. [25] Kooiman CG, Spinhoven P, Trijsburg RW, Rooijmans HG. Perceived parental attitude, alexithymia and defence style in psychiatric outpatients. Psychother Psychosom 1998;67:81 – 7. [26] King JL. Family environment and alexithymia in clients and no-clients. Psychother Res 2000;10:78 – 86. [27] Parker JDA, Taylor GJ, Bagby RM. The alexithymia construct: relationship with sociodemographic variables and intelligence. Comp Psychiatry 1989;30:434 – 41. [28] Lamberty GJ, Holt CS. Evidence for a verbal deficit in alexithymia. J Neuropsychiatry Clin Neurosci 1995;7:320 – 4. [29] Lane RD, Scwartz GE. Levels of emotional awareness: a cognitive – developmental theory and its application to psychopathology. Am J Psychiatry 1987;144:133 – 43. [30] Kauhanen J. Dealing with emotions and health: a population study of alexithymia in middle-aged men. In: Kuopio University Publications. D, Medical sciences, vol. 25. Kuopio: Kuopio University Printing Office, 1993. ¨ a¨rela¨ E, Toikka T, Kauhanen J. Preva[31] Salminen J, Saarija¨rvi S, A lence of alexithymia and its associations with sociodemographic variables in the general population of Finland. J Psychosom Res 1999; 46:75 – 82. [32] Lane RD, Lee S, Reidel R, Weldon V, Kaszniak A, Schwartz GE. Impaired verbal and nonverbal recognition in alexithymia. Psychosom Med 1996;58:203 – 10. [33] Lesser IM. A review of the alexithymia concept. Psychosom Med 1981;43:531 – 43.
Ability to speak at the age of 1 year and alexithymia 30 years later Pirkko Kokkonena,*, Juha Veijolaa, Juha T. Karvonena, Kristian La¨ksya, Jari Jokelainena, Marjo-Riitta Ja¨rvelinb,c, Matti Joukamaaa a Department of Psychiatry, Oulu University Hospital, Oulu, Finland Department of Public Health, Science and General Practice, University of Oulu, Oulu, Finland c Department of Epidemiology and Public Health, Imperial College School of Medicine, London, UK b
Received 30 October 2001; accepted 29 April 2002
Abstract Objective: We studied the association between speech development in the first year of life and alexithymia in young adulthood. Methods: The study forms a part of the Northern Finland 1966 Birth Cohort. The original material consisted of all liveborn children in the provinces of Lapland and Oulu in Finland with an expected delivery date during 1966. The comprehensive data collection began during the antenatal phase. In 1997, a 31-year follow-up study was made on a part of the initial sample. The 20-item version of the Toronto Alexithymia Scale (TAS-20) was given to 5983 subjects. Of them, 84% returned the questionnaire properly filled in. The ability to talk was classified according to whether the child spoke no words, one or two words, or three or more words at the age of 1 year. Statistical analyses on the association
between the ability to speak at the age of 1 year and alexithymia at the age of 31 years were performed, adjusted for birth weight, mother’s parity, place of residence and wantedness of pregnancy. Results: The mean of the total TAS score was lowest among early speakers and for both genders separately. The differences were statistically significant. A parallel significant difference was found among males on TAS Factors 2 and 3 and in case of females on TAS Factors 1 and 3. Conclusions: We found evidence for an association between speaking development in early childhood and later alexithymia. Our results support the theory that alexithymia may be a developmental process starting in early childhood and reinforcing itself in a social context. D 2003 Elsevier Science Inc. All rights reserved.
Keywords: Alexithymia; TAS-20; Cohort study; Prospective study; Speech development
Introduction Alexithymia, coined by Sifneos [1] in 1973, is a multifaceted construct including the following core features: difficulty in recognizing and verbalizing subjective feelings, a cognitive style characterized by absence of fantasy life and a tendency to recount the minute details of external events [2]. Although the clinical relevance of the alexithymia construct has been supported in a number of studies, there has been controversy and debate as to whether it is a stable personality trait, a transient state secondary to a stressful situation or a coping response to chronic illness [3]. It has even been postulated that alexithymia might just be a social or cultural phenomenon [4,5].
* Corresponding author. Department of Psychiatry, University of Oulu, Peltolantie 5, Oulu 90210, Finland. Fax: +358-8-333-167. E-mail address: [email protected] (P. Kokkonen).
Freyberger [6] conceptualized two kinds of alexithymia: primary alexithymia as a personality trait and secondary alexithymia as a state reaction resulting from stressful circumstances. Sifneos [7] attributes the term secondary alexithymia to alexithymic characteristics resulting from developmental arrests, massive psychological trauma, sociocultural factors or psychodynamic factors. Primary alexithymia is attributed to neurobiological deficits [7]. Sifneos [7] suggested that a disconnection between the limbic system and the verbal centers of the neocortex may result in primary alexithymia. Considering these neurobiological and neuropsychological theories on the origin of alexithymia, it would be interesting to investigate the possible associations between early speech development and later alexithymia. No longitudinal follow-up studies considering early developmental milestones and alexithymia have been conducted. The aim of this study was to examine the association between alexithymia and early speech development in childhood during the first year of life.
0022-3999/03/$ – see front matter D 2003 Elsevier Science Inc. All rights reserved. doi:10.1016/S0022-3999(02)00465-8
492
P. Kokkonen et al. / Journal of Psychosomatic Research 54 (2003) 491–495
Materials and methods Design This study forms a part of the Northern Finland 1966 Birth Cohort Project, which is a prospective, long-term, follow-up study. The original sample was collected from a geographically defined area of the two northernmost provinces of Finland. It consisted of an unselected, general population-based birth cohort of 12,058 live births in Northern Finland whose expected date on delivery fell in 1966, representing 96% of all such births [8]. The comprehensive data collection began during the antenatal phase. Subsequent to the baseline survey, a 1-year follow-up and a 14-year follow-up of the offspring have been performed. The detailed description of the study population and general design is presented elsewhere [8,9]. In 1997, a 31-year follow-up study was conducted through various questionnaires, interviews and clinical examinations. This follow-up study was composed of different subprojects and the alexithymia investigation formed a part of them. By using the Toronto Alexithymia Scale 20 (TAS-20) questionnaire, we had a possibility to assess the occurrence of alexithymia among these young adults. Using previously collected data, there was an opportunity to investigate connections between early childhood development and alexithymia in adulthood. The Ethics Committee of The Faculty of Medicine, University of Oulu, has approved this study. Subjects In 1997, 8417 of the original sample were living in the two northernmost provinces of Finland and in the capital city (Helsinki) area. As part of the 31-year follow-up study, an invitation to a field study was sent home to these cohort members. The field study consisted of a clinical examination, various questionnaires and interviews. Of the subjects contacted, 5983 participated in the field study and gave their written informed consent. A questionnaire including the TAS-20 was given to them to fill in at home and to return by mail. The TAS-20 questionnaire was returned properly by 84% (5028) of the subjects who had received it. The losses differed from the participants by being more commonly male, employed, married and less educated than the subjects participating [10]. Methods Of the different methods for measuring alexithymia, the TAS-20 is widely used and carefully validated. Its internal consistency, test – retest reliability, convergent, discriminant and concurrent validity have been demonstrated to be good [11,12]. The validity of the Finnish version of the TAS-20 has also been shown to be good [13]. The TAS-20 consists of 20 statements, each rated on a Likert-type scale from 1
(strongly disagree) to 5 (strongly agree) for a maximum score of 100. The TAS-20 has a three-factor structure: Factor 1 assesses difficulty in identifying feelings; Factor 2 concerns itself with difficulty in describing feelings; and Factor 3 reflects externally oriented thinking. Based on the TAS-20, the total score and the scores of TAS Factors 1– 3 (hereafter TASF1, TASF2, TASF3) were calculated. According to the developers’ recommendation, subjects with a total TAS score 61 were considered alexithymic, and those with a score 51 were considered nonalexithymic [3]. In this study, subjects with total TAS scores from 52 to 60 were considered probably alexithymic. The first follow-up study considering the growth, development and health status of the children of this cohort was conducted at child welfare centers when the subjects were 1 year old. Data on the number of words spoken by the child at 12 months were obtained for 95% of the children [9]. The ability to talk was classified according to whether the child spoke no words (late speaker), one or two words (medium speaker), three or more words (early speaker). Statistical methods The mean score of the TAS-20 and its three factors was calculated in each speaking category and also stratified by gender. The statistical significance of differences between these means was evaluated using multivariate analysis of variance (MANOVA), with post hoc Dunnett’s test, controlled for the following variables: birth weight ( < 2500 g, 2500 g), mother’s parity in 1966 (1, 2 3, 4 5, 6), place of residence in 1966 (urban, rural) and wantedness of pregnancy (wanted/mistimed, unwanted). The variables were selected based on earlier studies in this cohort [8,9]. We also fit the proportional odds model [14] in order to analyse the effect of the speaking to the tricotomized TAS score (see above). The score test was used for assessing validity of the proportionality assumptions. The data analysis was generated using the SAS/STAT software, version 8, of the SAS system for windows [15,16].
Results The prevalence of alexithymia was lowest among subjects speaking three or more words at the age of 1 year (Table 1). The association was shown both in male subjects and in female subjects. Among males, the effect was statistically significant, and among female subjects, the same trend was seen. The score test for the proportional odds assumptions was fulfilled in all subjects (c2 = 7.0, df = 8, P = .536), and in males (c2 = 7.8, df = 8, P = .451) and females (c2 = 6.6, df = 8, P = .582). After adjusting for confounding factors, the statistical significance remained among male subjects and for the whole sample. Mean total scores of alexithymia were associated with speaking development both in males and females, e.g., the
P. Kokkonen et al. / Journal of Psychosomatic Research 54 (2003) 491–495
493
Table 1 Prevalence of alexithymia in adulthood and speaking ability at the age of 1 year Three or more words
One or two words
No words
n
%
n
%
n
%
Analysis c2a
P
c2b
P
All subjects TAS 61 52 < TAS < 60 TAS 51
159 400 1997
6.2 15.7 78.1
86 189 739
8.5 18.6 72.9
80 208 743
7.8 20.2 72.1
21.0
< .001
15.8
< .001
Men TAS 61 52 < TAS < 60 TAS 51
91 206 763
8.6 19.4 72.0
44 110 299
9.7 24.3 66.0
55 136 359
10.0 24.7 65.3
10.1
.038
7.6
.022
Women TAS 61 52 < TAS < 60 TAS 51
68 194 1234
4.6 13.0 82.5
42 79 440
7.5 14.1 78.4
25 72 384
5.2 15.6 79.8
8.8
.067
3.6
.164
TAS = Toronto Alexithymia Scale 20 total score. a Chi-square test, df = 4. b Wald chi-square test by proportional odds modelling after controlled potential confounding factors, df = 2.
mean of the total TAS score was lowest in early speakers in all subjects and in both genders separately (Table 2). The total TAS scores were significantly higher among medium and late speakers than among early speakers. No statistical significance was found on TASF1 among male subjects and on TASF2 among female subjects. Otherwise, the TAS factors were associated with speaking ability in the same way as the total TAS scores. After adjusting for confounding factors, TASF1 was still found to be nonsignificant among
male subjects ( P = .206) and TASF2 among female subjects ( P =.184). After ANOVA and Dunnett’s post hoc test, there was no statistical significance between late speakers and early speakers on TASF1 among all subjects, but medium speakers had higher scores than early speakers (ANOVA, P =.024; Dunnett’s test, P =.018). Among female subjects, medium speakers had higher TASF1 scores than early speakers (ANOVA, P =.013; Dunnett’s test, P =.011), but
Table 2 The TAS-20 scores and speaking ability at the age of 1 year Three or more words
One or two words
No words
Mean
S.D.
Mean
S.D.
Mean
S.D.
F
P
Fa
P
All TAS TASF1 TASF2 TASF3
43.6 12.8 11.3 19.6
10.3 4.8 4.1 4.6
45.2 13.3 11.9 20.1
10.4 4.8 4.3 4.6
45.7 13.2 11.9 20.6
10.3 4.8 4.2 4.7
17.56b 4.89 12.84 18.71
< .001 .008 < .001 < .001
14.82c 3.74 11.21 16.46
< .001 .024 < .001 < .001
Men TAS TASF1 TASF2 TASF3
46.0 12.4 12.3 21.3
10.2 4.8 4.1 4.3
47.6 12.8 13.0 21.8
9.9 4.6 4.3 4.3
47.8 12.9 12.8 22.1
10.1 4.8 4.3 4.4
6.18d 1.16 6.89 5.02
.002 .313 .001 .007
5.21e 0.77 6.34 4.45
.001 .463 .002 .012
Women TAS TASF1 TASF2 TASF3
41.9 13.0 10.5 18.3
10.1 4.8 3.9 4.4
43.3 13.7 11.0 18.7
10.4 4.9 4.1 4.3
43.2 13.5 10.8 18.9
10.1 4.8 3.9 4.5
6.29f 5.12 2.52 4.47
.002 .006 .080 .012
4.52g 4.32 1.57 3.04
.011 .013 .207 .048
a b c d e f g
Analysis of variance
Adjusted for birth weight, mother’s parity, place of residence and wantedness of pregnancy. df (2, 4505). df (2, 4499). df (2, 2019). df (2, 2013). df (2, 2483). df (2, 2477).
494
P. Kokkonen et al. / Journal of Psychosomatic Research 54 (2003) 491–495
late speakers and early speakers did not differ from each other. On TASF2, early and medium speakers had higher scores than early speakers among all subjects (ANOVA, P < .001; Dunnett’s test, later vs. early, P < .001, and medium vs. early, P < .001). Among male subjects, both later and medium speakers differed from early speakers on TASF2 (ANOVA, P =.002; Dunnett’s test, later vs. early, P =.019, and medium vs. early, P =.003). On TASF3, early and medium speakers had higher scores among all subjects (ANOVA, P < .001; Dunnett’s test, later vs. early, P < .001, and medium vs. early, P =.032). Among male and female subjects, early speakers had also lowest scores on TASF3 (ANOVA, P =.012 among males, P =.048 among females; Dunnett’s test, later vs. early, P =.006 in males and P =.037 in females). No statistical significance was found between early and medium speakers.
Discussion We found evidence for an association between alexithymia in adulthood and timing of speech development. The early speakers, e.g., children who could speak three or more words at the age of 1 year, had the lowest mean score of the total TAS among all subjects and among both genders separately. The TAS factors were associated with speaking ability in the same way as the total TAS scores, but no statistical significance was found on TASF1 among male subjects and on TASF2 among female subjects. After adjusting for confounding factors, the abovementioned associations still remained statistically significant. Among male subjects, prevalence of later alexithymia was lowest in early speakers. Among female subjects, there was a similar trend but it fell below conventional statistical significance. TASF3 was associated with speaking development in all subjects and in both genders separately. There were gender differences considering the association with speech development on TASF1 and TASF2. Lane et al. [17] have speculated that for certain subjects, self-reported ratings on TASF1 and TASF2 may be less accurate than those on the TASF3. They suggested that high alexithymic subjects rate themselves unreliably because of their own lack of awareness of the deficits [17]. Suslow et al. [18] have proposed that TASF2, ‘‘Difficulties Describing Feelings,’’ does not measure primarily an impairment in symbolizing emotions. There are only few studies considering antecedents of alexithymia [19 – 26]. Diminished expressiveness in childhood [21,22], mother’s poor care [19,20], maternal alexithymia and disturbed family functioning [23] have all been shown to be associated with alexithymia. Kooiman et al. [25] could not confirm the relation of alexithymia to disturbances in the early parent –child relationship, while Berenbaum [24] found an association between childhood abuse and alexithymia in adulthood. All these studies have been cross-sectional and consequently vulnerable to recall
bias. The strength of our study was the prospective design: all developmental data were collected at the age of 1 year and measuring of alexithymia using the TAS-20 was conducted at the age of 31 years. In the present study, we found an association between later speech development and alexithymia in adulthood. In a study with nonclinical sample, Parker et al. [27] observed a slight correlation between alexithymia, on one hand, and intelligence score and mother’s education, on the other. Therefore, we also analysed separately the relationship between speaking development and mother’s education. No association was found. In the 1960s, women were not very highly educated in Finland. In the Northern Finland 1966 Birth Cohort, 86% of the mothers had less than 8 years of education [8]. This may be one reason why no significant association was found. Lamberty and Holt [28] suggested that an individual’s difficulty with verbalization of emotional material might be related to specific developmental verbal deficits, independent of other proposed etiological factors. The deficits underlying alexithymia have been attributed, at least in part, to an arrest in affect development during early childhood [3,29]. As Kauhanen [30] has stated, alexithymia may be ‘‘an accumulative process starting in early childhood and developing and reinforcing itself in a social context.’’ However, etiology of alexithymia probably involves multiple factors including constitutionally inherited variations in brain organization and deficiencies in the early family and social environment [3]. There are some limitations in our study. Firstly, all the subjects were of the same age. A positive association between alexithymia and age appears likely [31]. However, the sample was large and consisted of a population of normal young adults. To our knowledge, there exists no prospective longitudinal population-based study considering associations between alexithymia and development in early life. Secondly, because of our large cohort, the effect sizes are extremely small, which ought to be considered when interpreting the results. Thirdly, regarding speech development, the examination at 1 year of age may be too early to detect an effect. There is no pathology or real speech delay at the age of 1 year. However, early developmental differences in speaking ability were apparent in children who later developed alexithymic features. The TAS-20 was the measure of alexithymia. To date, the TAS-20 is the most commonly used and best validated measure of alexithymia [17]. Recently, however, self-rating measures like the TAS-20 have been seriously criticised [17,18]. As we already noted, Lane et al. [17] have proposed that an inherent difficulty may exist in the use of self-reports to measure something (i.e., alexithymia) that some of the respondents do not comprehend. It is possible that accurate responses in the TAS-20 may require the absence of extreme degrees of the trait that it measures [17]. So it may be possible that the TAS-20 cannot detect the most severe cases of alexithymia [32].
P. Kokkonen et al. / Journal of Psychosomatic Research 54 (2003) 491–495
The focus of our study was the developmental course of alexithymia. The concept of alexithymia has much to offer heuristically [33]. Well-conducted, longitudinal research from childhood to adulthood has been lacking. Developmental information might improve our understanding of alexithymia. In fact, alexithymia is conceptualized as a dimensional construct rather than a categorical phenomenon [3].
Acknowledgments This work was supported by grants from the Signe and Ane Gyllenberg Foundation and the Academy of Finland.
References [1] Sifneos PE. The prevalence of alexithymic characteristics in psychosomatic patients. Psychother Psychosom 1973;22:255 – 62. [2] Taylor GJ. Alexithymia: concept, measurement, and implications for treatment. Am J Psychiatry 1984;141:725 – 32. [3] Taylor GJ, Bagby RM, Parker JDA, editors. Disorders of affect regulation: alexithymia in medical and psychiatric illness. Cambridge: Cambridge Univ. Press, 1997. [4] Borens R, Grosse-Schulte E, Jaensch W, Kortemme KH. Is ‘alexithymia’ but a social phenomenon? Psychother Psychosom 1977;28: 193 – 8. [5] Kirmayer LJ. Languages of suffering and healing: alexithymia as a social and cultural process. Transcult Psychiatr Res Rev 1987;24: 119 – 36. [6] Freyberger H. Supportive psychotherapeutic techniques in primary and secondary alexithymia. Psychother Psychosom 1977;28:337 – 43. [7] Sifneos PE. Alexithymia and its relationship to hemispheric specialization, affect, and creativity. Psychiatr Clin North Am 1988;11: 287 – 92. [8] Rantakallio P. A longitudinal study of the Northern Finland birth cohort 1966. Paediatr Perinat Epidemiol 1988;2:59 – 88. [9] Rantakallio P, von Wendt L, Ma¨kinen H. Influence of social background on psychomotor development in the first year of life and its correlation with later intellectual capacity: a prospective cohort study. Early Hum Dev 1985;11:141 – 8. [10] Kokkonen P, Karvonen JT, Veijola J, La¨ksy K, Jokelainen J, Ja¨rvelin MR, Joukamaa M. Prevalence and sociodemographic correlates of alexithymia in a population sample of young adults. Comp Psychiatry 2001;42:471 – 6. [11] Bagby RM, Parker JDA, Taylor GJ. The twenty-item Toronto Alexithymia Scale: I. Item selection and cross-validation of the factor structure. J Psychosom Res 1994;38:23 – 32. [12] Bagby RM, Taylor GJ, Parker JDA. The twenty-item Toronto Alexithymia Scale: II. Convergent, discriminant, and concurrent validity. J Psychosom Res 1994;38:33 – 40.
495
[13] Joukamaa M, Miettunen J, Kokkonen P, Koskinen M, Julkunen J, Kauhanen J, Jokelainen J, Veijola J, La¨ksy K, Ja¨rvelin MR. The Finnish version of the 20-item Toronto Alexithymia Scale. Nord J Psychiatry 2001;2:123 – 7. [14] McCullah P. Regression models for ordinal data. J R Stat Soc, B 1980;42:109 – 42. [15] SAS Institute. SAS/STAT user’s guide, Version 8. Gary (NC): SAS Institute, 1999. [16] Allison PD. Logistic regression using the SAS system: theory and application. Gary (NC): SAS Institute, 1999. [17] Lane RD, Sechrest L, Riedel R. Sociodemographic correlates of alexithymia. Comp Psychiatry 1998;39:377 – 85. [18] Suslow T, Donges US, Kersting A, Arolt V. 20-Item Toronto Alexithymia Scale: do difficulties describing feelings assess proneness to shame instead of difficulties symbolizing emotions? Scand J Psychol 2000;41:329 – 34. [19] Fukunishi I, Kawamura T, Ago Y, Sei H, Morita Y, Rahe RH. Mothers’ low care in the development of alexithymia: a preliminary study in Japanese college students’. Psychol Rep 1997;80:143 – 6. [20] Fukunishi I, Sei H, Morita Y, Rahe RH. Sympathetic activity in alexithymics with mother’s low care. J Psychosom Res 1999;46:579 – 89. [21] Kench S, Irwin HJ. Alexithymia and childhood family environment. J Clin Psychol 2000;56:737 – 45. [22] Berenbaum H, James T. Correlates and retrospectively reported antecedents of alexithymia. Psychosom Med 1994;56:353 – 9. [23] Lumley MA, Mader C, Gramzow J, Papineau KTI. Family factors related to alexithymia characteristics. Psychosom Med 1996;58: 211 – 6. [24] Berenbaum H. Childhood abuse, alexithymia and personality disorder. J Psychosom Res 1996;41:585 – 95. [25] Kooiman CG, Spinhoven P, Trijsburg RW, Rooijmans HG. Perceived parental attitude, alexithymia and defence style in psychiatric outpatients. Psychother Psychosom 1998;67:81 – 7. [26] King JL. Family environment and alexithymia in clients and no-clients. Psychother Res 2000;10:78 – 86. [27] Parker JDA, Taylor GJ, Bagby RM. The alexithymia construct: relationship with sociodemographic variables and intelligence. Comp Psychiatry 1989;30:434 – 41. [28] Lamberty GJ, Holt CS. Evidence for a verbal deficit in alexithymia. J Neuropsychiatry Clin Neurosci 1995;7:320 – 4. [29] Lane RD, Scwartz GE. Levels of emotional awareness: a cognitive – developmental theory and its application to psychopathology. Am J Psychiatry 1987;144:133 – 43. [30] Kauhanen J. Dealing with emotions and health: a population study of alexithymia in middle-aged men. In: Kuopio University Publications. D, Medical sciences, vol. 25. Kuopio: Kuopio University Printing Office, 1993. ¨ a¨rela¨ E, Toikka T, Kauhanen J. Preva[31] Salminen J, Saarija¨rvi S, A lence of alexithymia and its associations with sociodemographic variables in the general population of Finland. J Psychosom Res 1999; 46:75 – 82. [32] Lane RD, Lee S, Reidel R, Weldon V, Kaszniak A, Schwartz GE. Impaired verbal and nonverbal recognition in alexithymia. Psychosom Med 1996;58:203 – 10. [33] Lesser IM. A review of the alexithymia concept. Psychosom Med 1981;43:531 – 43.