Psychometric properties of the Toronto Alexithymia Scale (TAS-20) in Brazil

Abstract

Internationally, the most widely used self-report instrument to assess alexithymia (a clinical condition characterized by serious deficits in the cognitive processing of emotions) is the Toronto Alexithymia Scale (TAS-20). Because this measure has been poorly investigated in Brazil and emerging research suggests that TAS-20 scores might vary from one culture to another, we sought to: (1) investigate the psychometric properties of a Brazilian TAS-20; and (2) examine the degree to which Brazilian non-clinical TAS-20 scores differ from the scores of non-clinical samples from other cultures. A sample of 850 non-clinical Brazilian adults were administered a number of questionnaires and performance-based measures via online data collection. Data analyses inspected internal consistency and factor structure of the TAS-20, and tested the association of TAS-20 scores to emotional functioning and psychopathology. In line with previous international research, the Brazilian TAS-20 showed acceptable to adequate psychometric properties. Furthermore, TAS-20 scores associated negatively with empathy and emotional perception, and positively with emotion dysregulation and personality traits like dependency, abrupt changes in mood, and avoidance of criticism. Also noteworthy, our non-clinical Brazilian TAS-20 scores were very similar to those observed in other previously published non-clinical TAS-20 scores from seven non-Brazilian cultures.

Keywords

alexithymia Brazil emotion TAS-20 validity

Alexithymia (literally “no word for feeling”) was defined and initially studied as a condition of psychosomatic patients who had great difficulty in fantasizing and in decoding or expressing emotions (Taylor, 1987). The current definition of alexithymia includes: (1) difficulty recognizing emotional states; (2) a low ability to fantasize; (3) focusing interests on external events rather than internal or introspective thinking; and (4) concrete verbalizations, with reduced or non-symbolic thinking (Bagby, Taylor, & Parker, 1994; Taylor & Bagby, 2004).

Over the years, a large number of studies have been conducted and the construct of alexithymia has been investigated in relation to clinical conditions such as eating disorders (Brewer, Cook, Cardi, Treasure, & Bird, 2015; Kessler, Schwarze, Filipic, Traue, & von Wietersheim, 2006), psychosomatic gastrointestinal disorders (Gerhards, 1998), inflammatory bowel disease (Porcelli & Meyer, 2002), dependent and other personality disorders (Loas & Cormier, 2009; Nicolò et al., 2011), alcohol dependence and motivation for drinking (Maciel & Yoshida, 2006; Stewart, Zvolensky, & Eifert, 2002), anorexia nervosa (Torres et al., 2011), and constriction of feelings (Akimoto, Fukunishi, Baba, Mori, & Wai, 2002), among others.

In addition, being a relevant construct of social interaction, the relationship of alexithymia to interpersonal abilities like emotional perception has been investigated. In fact, a large number of authors have reported that people with a lower ability to perceive emotional expression in faces (Brewer et al., 2015; Cook, Brewer, Shah, & Bird, 2013; Grynberg et al., 2012; Jongen et al., 2014; Parker, Taylor, & Bagby, 1993; Prkachin, Casey, & Prkachin, 2009; van der Velde et al., 2015) and voices (Heaton et al., 2012) tend to display higher levels of alexithymia. Other studies found little or no association between alexithymia and emotional perception (Gerhards, 1998; Kessler et al., 2006; Miguel et al., 2010; Zuanazzi, Ricci, & Miguel, 2015). Furthermore, some authors have suggested that alexithymia might be associated with poor empathy skills (e.g., Guttman & Laporte, 2002; Jonason & Krause, 2013; Moriguchi et al., 2007; Silani et al., 2008; Swart, Kortekaas, & Aleman, 2009), though additional research on this topic would be beneficial.

Regarding the assessment of alexithymia, the most frequently used instrument is the Toronto Alexithymia Scale (TAS-20), a self-report instrument composed of 20 items (Bagby, Parker, & Taylor, 1994). The 20 items are spread across three subscales: (1) Difficulty Identifying Feelings (DIF); (2) Difficulty Describing Feelings (DDF); and (3) Externally Oriented Thinking (EOT). Cross-cultural studies on the TAS-20 have been conducted in many different countries, such as Finland (Säkkinen, Kaltiala-Heino, Ranta, Haataja, & Joukamaa, 2007), France (Loas, Parker, Otmani, Verrier, & Fremaux, 1997), Germany (Bach, Bach, de Zwaan, Serim, & Böhmer, 1996), Greece (Tsaousis et al., 2010), Italy (Bressi et al., 1996; Craparo, Faraci, & Gori, 2015), Iran (Besharat, 2007, 2008, 2015; Pourhosein, Rostami, Besharat, & Mirzamani, 2006), Japan (Akimoto et al., 2002), Peru (Loiselle & Cossette, 2001), and Spain (Fernández-Jiménez et al., 2013). In addition, an internet version of the TAS-20 has been studied recently, showing psychometric properties similar to the original paper-and-pencil version (Bagby, Ayearst, Morariu, Watters, & Taylor, 2014). In general, all these studies concur to indicate that while the DIF and DDF factors possess relatively sound psychometric properties, EOT often produces low reliability indexes (Kooiman, Spinhoven, & Trijsburg, 2002). Interestingly, however, in a relatively recent, well-designed study by Prkachin et al. (2009), EOT correlated with perception of facial expressions of emotion more strongly than did DIF and DDF.

Despite the large number of international studies on the assessment of alexithymia, the TAS-20 has been poorly investigated in Brazil. In fact, only two articles (i.e., Balbinotti, Wiethaeuper, Barbosa, Ferreira, & Gonzalez, 2008; and Wiethaeuper, Balbinotti, Pelisoli, & Barbosa, 2005) have examined a Brazilian-Portuguese version of the TAS-20, and neither of them have been published in English-language journals. Moreover, both of those articles focused on the factor structure of the TAS-20 but did not investigate its convergent or divergent validity, and only one of them reported Cronbach’s alphas. Thus, additional research on the psychometric properties of the Brazilian TAS-20 is needed.

Furthermore, emerging research suggests that cultural values might contribute to shaping one of the core features of alexithymia. Indeed, Dere et al. (2013) reported that TAS-20 EOT scores differed when considering Euro-Canadian versus Chinese-Canadian students, and that modernization and Euro-American values negatively associated with EOT scores in a sample of Chinese outpatients. However, the empirical data on this topic is still very scarce.

To extend the available literature in these important but under-investigated areas, the current study aimed at: (i) testing the psychometric properties of a Brazilian-Portuguese version of the TAS-20; (ii) investigating the association of alexithymia to emotion regulation, emotional perception, and empathy; and (iii) comparing Brazilian non-clinical TAS-20 scores against other cultures’ non-clinical TAS-20 scores. To do so, we recruited via web-based procedures a relatively large sample of Brazilian adult participants, and administered them a number of self-report, as well as performance-based, psychological measures.

Methods

Participants

This study used a sample of 850 adult participants, with ages ranging from 18 to 71 (M = 30.97, SD = 10.80), with 98.6% of participants aged under 60. The majority of the sample were women (83.2%), with men being about one sixth (16.8%); and education level was in its majority college education (52.5% were graduates or enrolled in graduate courses and 18.0% were post-graduate).

Procedure

Data collection followed standard ethical procedures for web-based psychological research. Potential participants were informed about the possibility to volunteer for our study through a post on social media (Facebook). By clicking on the invitation, prospective participants were conducted to another website where an informed consent form, containing the information about the research, was displayed. Consent was provided by accepting to create a username to log into the web page, which gave access to the instruments, all administered in Brazilian Portuguese. There was no financial compensation for participants, although results for the instruments were displayed at the end, with a brief explanation of each construct. Data collection took four months.

Measures

Below is the list of measures available for each participant to respond to. In order to avoid careless responding, participants had the option to choose which measures they wanted to fill in and were informed about the possibility to discontinue participation at any time. Therefore, not all participants filled in all instruments.

The Toronto Alexithymia Scale (TAS-20; Bagby, Parker, et al., 1994)

As noted above, the TAS-20 is the most-used instrument for assessing alexithymia. It is comprised of 20 items assessing three dimensions of alexithymia, i.e., Difficulty Identifying Feelings (DIF), Difficulty Describing Feelings (DDF), and Externally Oriented Thinking (EOT). Several international studies concur that while DIF and DDF possess good psychometric properties, the reliability of EOT is doubtful (for a review, see Kooiman et al., 2002). The current study used the Brazilian TAS-20 developed by Wiethaeuper et al. (2005).

Difficulties in Emotion Regulation Scale (DERS; Gratz & Roemer, 2004)

This instrument is a 36-item questionnaire that assesses six factors of emotional dysregulation: Non-Acceptance of Emotional Responses (Acceptance); Difficulties in Engaging in Goal-Oriented Behaviors (Goals); Difficulties in Control Impulses (Impulses); Lack of Emotional Knowledge (Knowledge); Restricted Access to Emotional Regulation Strategies (Strategies); and Lack of Emotional Clarity (Clarity). Respondents must classify the frequency with which each sentence applies to them, on a five-point Likert scale varying from 1 (almost never 0–10%) to 5 (almost always 91–100%). Gratz and Roemer (2004) reported good psychometric properties of this instrument in their original study, with alphas ranging from 0.77 to 0.93, besides other good indicators of construct validity and test-retest stability. Several international studies also support the strong validity and reliability of the DERS (e.g., Cho & Hong, 2013; Coutinho, Ribeiro, Ferreirinha, & Dias, 2010; Giromini, Velotti, de Campora, Bonalume, & Zavattini, 2012; Ruganci & Gençöz, 2010). The current study used the Brazilian-Portuguese version of the scale developed by Miguel et al. (2017), which produced, in our sample, Cronbach’s alphas ranging from 0.89 to 0.94.

Questionnaire for Cognitive and Affective Empathy (QCAE; Reniers, Corcoran, Drake, Shryane, & Völlm, 2011)

The QCAE is an instrument comprised of 31 items that assesses cognitive and affective empathy. A four-point Likert scale ranging from strongly disagree to strongly agree is provided to measure the agreement of the respondents with each of the affirmatives. Reniers et al. (2011) reported acceptable psychometric properties of the scale in the original study. This study used the Brazilian-Portuguese version of the QCAE, which was developed by Miguel et al. (2017) via translation/back-translation procedures and in consultation with Dr. Reniers, the author of the original instrument. In our sample, alphas for the QCAE total score of affective and cognitive empathy were respectively 0.68 and 0.87. For the subscales, the alpha ranged from 0.34 to 0.87. The alpha for the total score was 0.82.

Clinical Dimensional Personality Inventory (IDCP-2; Carvalho & Primi, 2015)

This test assesses characteristics associated with personality disorders using a Likert scale of four points to measure the agreement of the respondent with each affirmative. The scale can be used to assess 12 factors of the personality, besides a small questionnaire about the psychiatric and psychological history and suicide concerns. In preview studies, the IDCP-2 showed good psychometric properties (Abela, Carvalho, Cho, & Yazigi, 2015; Carvalho & Primi, 2015). In our sample, the alphas for the subscales of IDCP-2 were 0.91 for Dependency, 0.89 for Aggressiveness, 0.86 for Mood instability, 0.86 for Eccentricity, 0.80 for Attention seeking, 0.87 for Distrust, 0.86 for Grandiosity, 0.88 for Isolation, 0.89 for Criticism avoidance, 0.90 for Self-sacrifice, 0.84 for Conscientiousness, and 0.89 for Impulsiveness.

Primary Emotion Perception Test (PEP; Miguel & Primi, 2014)

This instrument is a computer-based measure that assesses one of the abilities of Emotional Intelligence, which is the capacity to perceive basic emotions (Mayer, Roberts, & Barsade, 2008). The task involves watching 35 videos of people expressing basic emotions (joy, love, fear, sadness, surprise, disgust, anger, and curiosity), with the purpose to identify the emotion being shown. The PEP has shown satisfactory psychometric properties in previous Brazilian studies, with a test-retest correlation of 0.86 in seven months and Rasch reliability of 0.68. The Rasch reliability for our sample was 0.63.

Battery of Reasoning Tests (BPR-5; Primi & Almeida, 2000)

This is a Brazilian instrument battery that measures general reasoning abilities. In this research, we used two of the five subscales of this battery: Verbal Reasoning (VR) and Abstract Reasoning (AR). The BPR-5 is composed of 25 items where the individual has to identify for the verbal and abstract reasoning the relationship between words and abstract geometric figures, respectively. In our sample, internal consistency was 0.80 for AR and 0.66 for VR.

Results

TAS-20 scores

Table 1 presents the descriptive statistics for the entire sample, separated by gender for the three factors, and the total score of the TAS-20. We analyzed the differences between women and men’s scores, and found only one significant difference for EOT, with men scoring higher. However, the effect size of the difference was small (Cohen, 1988), which indicates that there are no notable distinctions between women and men regarding their scores in TAS-20. Due to the difference between the number of women (n = 707) and men (n = 143), we also ran a Mann-Whitney U test, a non-parametric comparison between groups. The results of these additional analyses, however, led to the same conclusion, confirming that the Brazilian men and women included in this research scored similarly on the TAS-20.

Table 1.

Descriptive statistics, gender differences and correlations with age for the TAS-20.

	Women (n = 707)		Men (n = 143)		Total (N = 850)		Gender differences			Correlation with age
	M	SD	M	SD	M	SD	t (848)	p	d	r	p
DIF	17.11	6.84	16.97	6.96	17.09	6.86	0.22	0.828	0.02	–0.26	<0.001
DDF	13.71	4.63	13.77	4.9	13.72	4.68	0.14	0.888	0.01	–0.27	<0.001
EOT	18.12	4.95	17.23	4.65	17.97	4.91	1.97	0.049	0.18	–0.01	0.841
Total	48.93	12.89	47.97	13.25	48.77	12.95	0.81	0.418	0.07	–0.24	<0.001

TAS-20 scores showed medium-sized significant correlations with age, in the expected direction (i.e., older participants displayed less alexithymia than younger participants, as emotion regulation typically improves with age), except for EOT, whose correlation value approached zero.

Internal consistency

We calculated Cronbach’s alphas and item-total correlations to examine the internal consistency of the TAS-20, presented in Table 2. Cronbach’s alphas ranged from 0.51 to 0.83. Apart from the EOT subscale, all indices were satisfactory. Our results are very similar to those found in the scale development studies (Bagby, Parker, et al., 1994), the internet version (Bagby et al., 2014), and a smaller Brazilian sample (Wiethaeuper et al., 2005).

Table 2.

Internal consistency analyses for TAS-20.

	No. of items	Cronbach’s alpha	Range of item-total correlations
DIF	7	0.83	0.80–0.84
DDF	5	0.68	0.59–0.67
EOT	8	0.51	0.42–0.51
Total	20	0.82	0.80–0.83

Factor structure

We conducted confirmatory factor analyses (CFAs) to analyze the proposed factor structure for TAS-20. The resulting model is presented in Figure 1. The root mean square error of approximation (RMSEA) was 0.056 (90% confidence interval of 0.051–0.061), the non-normed fit index (NNFI) was 0.95, the comparative fit index (CFI) was 0.95, and the standardized RMR was 0.048. According to commonly accepted benchmarks—i.e., RMSEA ≈ 0.05 for a close fit (Browne & Cudeck, 1993); NNFI and CFI ≥ 0.90 for a good fit (Bentler & Bonett, 1980); and standardized RMR < 0.08 for a good fit (Hu & Bentler, 1999)—our data adequately fit the model proposed by Bagby, Parker, et al. (1994) .

Figure 1.

Factor structure of the TAS-20.

The relationship of TAS-20 to emotion regulation, empathy, and emotional perception

Because emotion regulation typically improves with age (and in our sample the TAS-20 correlated with age), to test the relationship of alexithymia to emotion regulation and empathy we performed a series of partial correlations aimed at controlling for the effects of age (Table 3). The results of all these analyses were statistically significant, with few exceptions. Among the three TAS-20 factors, DIF produced the strongest correlations with difficulties regulating emotions, while EOT produced the strongest (negative) correlations with empathy. While most correlations with empathy were negative, there were some positive correlations. They were for QCAE’s Emotion Contagion, and between the subscale DDF and Affective Empathy. Taken together, these results indicate that alexithymia correlates positively with difficulties in regulating emotions and negatively with cognitive empathy, though DDF positively correlates with emotional contagion and affective empathy.

Table 3.

Partial correlations of TAS-20 to DERS and QCAE, controlled by age.

	DIF	DDF	EOT	Total
DERS (n = 722)	.45***	.28***	.18***	.41***
Nonacceptance	.34***	.19***	.12***	.30***
Goals	.46***	.22***	.19***	.40***
Impulse	.24***	.28***	.41***	.39***
Awareness	.54***	.29***	.23***	.48***
Strategies	.77***	.56***	.37***	.75***
Clarity	.66***	.41***	.33***	.63***
Total	.45***	.28***	.18***	.41***
QCAE (n = 847)
Perspective Taking	–.19***	–.20***	–.26***	–.27***
Online Simulation	–.27***	–.20***	–.34***	–.35***
Emotion Contagion	.26***	.14***	.08*	.22***
Proximal Responsivity	.01	–.08*	–.17***	–.09**
Peripheral Responsivity	–.05	–.08*	–.27***	–.16***
Cognitive Empathy	–.28***	–.24***	–.37***	–.38***
Affective Empathy	.12***	.00	–.15***	.01
Total	–.17***	–.19***	–.37***	–.30***

*p < .05; **p < .01; ***p < .001.

Deficits in emotion regulation are typical of several psychopathological conditions. Thus, to further investigate the link between alexithymia and emotion regulation, we also inspected the partial correlations (controlling for age) between the TAS-20 and IDCP-2 (Table 4). Some of the most prominent results were that alexithymia correlated to dependency on other people, mood instability, criticism avoidance, and self-sacrifice. The EOT subscale showed the lowest correlations with IDCP-2, with almost half of them non-significant. The pathological traits of need for attention, suspiciousness, grandiosity, excessive conscientiousness, and impulsiveness showed lower magnitude correlations with alexithymia, albeit some were still significant. DIF correlated with most of the psychiatric questions (dummy variables), albeit with a small effect size. Total alexithymia correlated only with suicide concern questions.

Table 4.

Partial correlations of TAS-20 to pathological personality traits, controlled by age.

	DIF	DDF	EOT	Total
IDCP (n = 227)
Dependency	0.45***	0.29***	0.19**	0.41***
Aggressiveness	0.29***	0.28***	0.13*	0.30***
Mood Instability	0.50***	0.31***	0.16*	0.44***
Eccentricity	0.35***	0.41***	0.09	0.37***
Need for Attention	0.24***	0.07	–0.06	0.14*
Suspiciousness	0.28***	0.29***	0.10	0.29***
Grandiosity	0.12	0.13*	0.01	0.11
Isolation	0.26***	0.37***	0.24***	0.36***
Criticism Avoidance	0.42***	0.36***	0.25***	0.45***
Self-Sacrifice	0.46***	0.36***	0.17*	0.44***
Conscientiousness	0.18**	0.28***	0.10	0.24***
Impulsiveness	0.23***	0.25***	0.08	0.24***
Individual Questions^a
Psychiatric Treatment?	0.13*	–0.01	–0.07	0.04
Psychological Treatment?	0.13*	0.00	0.06	0.09
Psychiatric Medication?	0.11	0.00	–0.02	0.06
Suicide Attempts?	0.18**	0.14*	0.00	0.15*
Suicide Thoughts?	0.24***	0.09	–0.01	0.16*
Psychiatric Diagnosis?	0.16*	0.03	–0.01	0.09

^aPoint bi-serial correlations (0 = no, 1 = yes); *p < 0.05; **p < 0.01; ***p < 0.001.

Next, we investigated the relationship between alexithymia and emotional perception by testing partial correlations (controlling for age) between the TAS-20 and PEP. Interestingly, among the three TAS-20 scales, EOT produced the strongest negative correlation with the ability to identify emotions in other people’s faces (Table 5). The effect size of this association, however, was still relatively small, i.e., small to medium.

Table 5.

Partial correlations of TAS-20 to cognitive tasks, controlled by age.

	DIF	DDF	EOT	Total
Emotional Perception (n = 846)	–0.09**	–0.08*	–0.19***	–0.15***
Verbal Reasoning (n = 162)	–0.10	–0.05	–0.11	–0.11
Abstract Reasoning (n = 327)	–0.10	0.02	–0.15**	–0.11

*p < 0.05; **p < 0.01; ***p < 0.001.

Additionally, as a means to investigate divergent validity, we also ran correlation analyses between TAS-20 and two intelligence tasks. The results of these analyses are presented in Table 5. There was only one statistically significant correlation, which was low. The results give support to the differentiation of alexithymia from abstract and verbal reasoning.

Comparison between Brazilian versus other cultures’ scores on TAS-20

To examine the extent to which our Brazilian TAS-20 scores differed from those obtained in other cultures, we retrieved from the literature all cross-cultural validation studies of the TAS-20 that were conducted with non-clinical samples, and for which the mean and standard deviation values of DIF, DDF, EOT, and Total scores had been reported (only articles published in English-language journals were considered, to this aim). Accordingly, we inspected TAS-20 data from Greek (Tsaousis et al., 2010), Italian (Bressi et al. 1996), American-English and Peruvian-Spanish (Loiselle & Cossette, 2001), Farsi (Besharat, 2007), and Chinese (Zhu et al., 2007) TAS-20 versions.

To estimate TAS-20 descriptive statistics for all available samples (N = 3414), we added the data from the international samples described above to our Brazilian sample. Specifically, for each TAS-20 score we calculated a weighted average, i.e., the mean of the combined sample was calculated as the sum of the products of each sample's mean value times the size of that specific sample, divided by the N of the combined sample. A similar approach was then used also to estimate the dispersion statistics, i.e., variance and standard deviation, of this combined sample. As a result, the mean score of the entire combined sample was 48.1 (SD = 11.0) for the Total score, 16.2 (SD = 5.4) for TAS-20 DIF, 13.0 (SD = 4.0) for TAS-20 DDF, and 18.4 (SD = 4.7) for TAS-20 EOT.

Based on these values, we finally transformed each sample’s raw scores into T scores (i.e., scores with a mean of 50 and SD of 10), and plotted the results in Figure 2, so as to examine the extent to which our Brazilian scores differed from those obtained in other, cross-cultural validation studies involving non-clinical participants. Interestingly, all samples produced T scores within one standard deviation from the grand mean of the entire sample (i.e., between 40 and 60). The Brazilian sample was close to the grand mean values for all TAS-20 scores (i.e., close to T = 50); the highest scores were obtained with the Peruvian version; and the lowest scores were obtained with the Greek and Farsi versions.

Figure 2.

Distribution of TAS-20 T scores in multiple cross-cultural validation studies.

Discussion

Much empirical evidence indicates that alexithymia, broadly defined as a deficit in the cognitive processing of emotions, associates with difficulties in emotion regulation and psychopathology. Additionally, some authors have suggested that alexithymic personality traits may also associate with deficits in interpersonal abilities such as empathy and emotional perception, though the empirical research findings are somewhat mixed in this regard. To contribute to the study of alexithymia and investigate its relationship to emotion regulation, empathy, and emotional perception, the current study inspected a relatively large sample of Brazilian adult individuals, and tested the correlations between the TAS-20 and a number of self-report and performance-based psychological measures. Besides this, since the literature on the validity and reliability of the TAS-20 in Brazil is scarce, our research also aimed at examining the psychometric properties of a Brazilian-Portuguese version of the TAS-20, and at inspecting whether a relatively large non-clinical Brazilian sample would score differently on the TAS-20 when compared to other non-clinical samples from different cultures.

As expected, all TAS-20 scales positively correlated with the DERS. When considering the total DERS score, the effect sizes of these correlations were large (r ≈ 0.50) for DIF, medium (r ≈ 0.30) for DDF, and small to medium (r ≈ 0.20) for EOT (for benchmarks for interpreting r values, see Cohen, 1988). Thus, identifying feelings may be particularly important to regulate emotions, while DDF and EOT possibly play a minor role in this. This finding is not too surprising: focusing on one’s internal world (EOT) and/or being able to describe these internal experiences (DDF) is important for, but does not necessarily lead to a clear identification or understanding of, one’s own mental states and feelings (DIF). Conversely, emotion regulation largely depends on the ability to attend to, acknowledge, and be clear about the emotions one is experiencing (Gratz & Roemer, 2004).

Alexithymia also correlated with empathy, as measured by the QCAE. However, while the effect size of this negative correlation was medium to large (r ≈ –0.40) for the QCAE subscale Cognitive Empathy, it approached zero for the QCAE subscale Affective Empathy, and even became positive for the QCAE Emotion Contagion subscale (r ≈ 0.20). Thus, the presence of alexithymic traits seems to interfere with the ability to cognitively process and understand the emotions experienced by others, but not with the ability to implicitly and perhaps automatically share and feel those same emotions. Put differently, alexithymia seems to allow for an automatic emotional contagion and embodied simulation of others’ emotions, even in the absence of a clear understanding of what the meanings and implications of these shared experiences, feelings, and emotions might be. To some extent, this finding is in line with recent research suggesting that affective empathy is less adaptive and more associated with psychopathology than is cognitive empathy (e.g., Giromini et al., 2016; Michaels et al., 2014).

Some empirical data support the hypotheses that higher alexithymia is correlated with a lower ability to recognize emotions in other people (Brewer et al., 2015; Cook et al., 2013; Grynberg et al., 2012; Jongen et al., 2014; Parker et al., 1993; Prkachin et al., 2009; van der Velde et al., 2015). However, some articles report partial or no association between these two constructs (Gerhards, 1998; Kessler et al., 2006; Miguel et al., 2010; Zuanazzi et al., 2015), and our study also showed a weak association between emotional perception and alexithymia. In an attempt to better understand these seemingly contradictory findings, future studies should test whether the association between these constructs would be mediated by other psychological aspects, like pathological personality traits. In fact, a recent study has shown that alexithymia could be a mediator of the relationship between personality and mental health (Atari & Yaghoubirad, 2016). As such, personality could also mediate the extent to which people with alexithymic traits do or do not impact emotional recognition.

An intriguing finding of our study is that among the three TAS-20 scales, EOT produced the strongest negative correlations both with empathy (r ≈ –0.40 with the total QCAE, versus r ≈ –0.20 for DIF and DDF) and with emotional perception (r ≈ –0.20 with the PEP, versus r > –0.10 for DIF and DDF). As noted in the Introduction, also in Prkachin et al.’s (2009) study, EOT correlated more strongly than DIF and DDF with perception of facial expressions of emotion. As a possible explanation for their findings, Prkachin et al. (2009) suggested that when a person focuses on external events rather than internal or introspective thinking, they likely pay no or little attention to their emotions while observing another individual’s emotional face. Thus, because emotional perception presumably involves people processing visual information while retrieving memories of their own emotional states associated with the observed emotional stimuli, people with high scores on EOT might perform poorly on emotion perception tasks because they do not encode their own emotional states while observing emotionally relevant information displayed by others. This hypothesis, which is consistent with the “shared manifold of intersubjectivity” theory (Gallese, 2003), seems to be supported also by our data, given that EOT correlated, more strongly than DIF and DDF, with both the QCAE and PEP.

We also found that alexithymia, as measured by the TAS-20, is related to several pathological personality traits, confirming previous findings in the literature (Loas & Cormier, 2009; Nicolò et al., 2011). Not all IDCP-2 traits correlated significantly with the TAS-20 and its subfactors, and some of the significant correlations magnitudes were low. However, the correlations that displayed the highest magnitudes were coherent with the description of alexithymia, suggesting that it is related to dependency on other people’s opinions and guidance, frequent and abrupt changes in mood, and putting oneself in situations to avoid feeling distressed from being criticized and to receive positive feedback. As shown in previous studies (Glenn & Klonsky, 2009; McLaughlin, Mennin, & Farach, 2007; Miguel et al., 2016), pathological or maladaptive personality traits are associated with emotional dysregulation. These results point to a lower ability to regulate emotions, thus using behavioral strategies to put oneself in situations to avoid emotional distress.

In general, our findings attest to the reliability of the TAS-20 scores, except for EOT. On the one hand, international studies show similar findings for that subscale, which may indicate a further need for improving its items. The number of items per scale might also account for the different alpha values from one scale to another. On the other hand, however, the idea that Cronbach’s alphas should be higher than 0.70 has recently been criticized by statisticians, as high internal consistency often reflects redundancy across items, rather than reliability (e.g., John and Soto, 2007; Sijtsma, 2009). Given that the debate on the reliability of EOT is far from being settled, in the meantime we recommend that individual interpretations made with EOT be conducted carefully.

Lastly, our study also allowed us to compare the TAS-20 scores produced by a number of non-clinical samples from seven different cultures. Differently from recent research supporting the possibility of a cultural interference on alexithymia (Dere et al., 2013), the differences in this investigation were rather small. Indeed, the TAS-20 scores from the American, Chinese, Greek, Italian, Farsi, and Peruvian versions did not notably differ from each other or from our Brazilian data—in fact, each sample’s scores fell within one standard deviation of the grand mean of the entire sample combining all data. With this respect, two considerations can be made. First, our findings do not support the hypothesis that culture plays a key role in the manifestation or development of alexithymia traits. Second, the TAS-20 scores produced by our non-clinical Brazilian sample very closely resembled those produced by the other non-clinical samples under investigation, which provides additional support to the hypothesis that the TAS-20 should perform well in Brazil too.

Several limitations can be considered in this study. First, the majority of participants were women, which may restrict generalizability of the results for the population. Even though the magnitude of difference between genders was small, a study with a larger sample of men could investigate whether gender differences display a larger effect size. Second, the age range was very broad (from 18 to 71 years old). However, most of the participants were young or relatively young adults, so that participants older than 60 (1.4%) might be considered to be outlier data. We tested whether excluding them from the analysis would impact the results and found no support in this sense. Nevertheless, future studies might attempt to replicate our study by testing older adults. Third, all instruments were collected with online administration. According to Bagby et al. (2014), the results of their study “comparing Internet and paper-and-pencil administrations of the TAS-20 indicate that the two methods are comparable and can be used interchangeably” (p. 20). However, it is still possible that a number of participants were not in a neutral environment or did not put enough effort into answering some of the tests and/or items. In addition, the participants may represent a part of the population with a better economic status and resources to be able to use the internet, while less privileged people may not be represented. Fourth, the instruments that displayed higher correlation magnitudes are self-report, just like the TAS-20, and may be influenced by social desirability and other biases. Further studies could address this issue by using more performance-based or behavioral measures, or comparison between criteria groups. Despite these limitations, our study has the merit of providing some insight into the relationship between alexithymia, emotional functioning, and mental health. It is the first study of the TAS-20 in a large Brazilian sample, providing some validity evidence for the scale.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Maíra Stivaleti Colombarolli, a psychologist, is currently a Master’s student on the Psychology Post Graduation Program in the Faculty of Philosophy, Science and Letters of Ribeirão Preto, at the University of São Paulo, Brazil, and a member of the Center of Research on Psychodiagnostics in the Department of Psychology at USP. Her research includes transcultural adaptation and validation of psychological measures for Brazilian culture, especially those related to emotion and cognitive processes.

Ana Carolina Zuanazzi, MS, is a doctoral student in Psychology at São Francisco University in Brazil. Her research and work focus on assessment of emotional intelligence and personality, mainly with projective tests.

Fabiano Koich Miguel, PhD, is a professor of psychology in the Department of Psychology and Psychoanalysis at the State University of Londrina in Brazil. His research and work focus on assessment of emotional intelligence and personality, and the development of computerized tests. Dr. Miguel was visiting professor at the University of Turin in 2015–2016.

Luciano Giromini, PhD, is a lecturer in the Department of Psychology, and core faculty member for the PhD program in Psychological, Anthropological and Educational Sciences of the University of Turin, Italy. He has taught psychological assessment and psychometrics at two Italian universities and one Californian university, and is currently coordinator for the Evidence-Based Psychological Assessment research team at the University of Turin. He is a member of the R-PAS Research and Development group, and one of the authors of the Inventory of Problems (IOP-29), a newly developed tool to detect non-credible cognitive or psychiatric complaints. His published works focus on assessment, malingering, and cross-cultural adaptations of psychological assessment tools.

References

Abela

R. K.

Carvalho

L. F.

Cho

S. J. M.

Yazigi

(2015) Validity evidences for the Dimensional Clinical Personality Inventory in outpatient psychiatric sample. Paidéia (Ribeirão Preto) 25(61): 221–228.

Akimoto

Fukunishi

Baba

Mori

M. M.

Wai

(2002) Alexithymia and sociocultural factors in a Japanese sample: A study with the Rorschach. Psychological Reports 90(1): 205–211.

Atari

Yaghoubirad

(2016) The big five personality dimensions and mental health: The mediating role of alexithymia. Asian Journal of Psychiatry 24: 59–64.

Bach

de Zwaan

Serim

Böhmer

(1996) Validation of the German version of the 20-item Toronto Alexithymia Scale in normal persons and psychiatric patients. Psychotherapie Psychosomatik Medizinische Psychologie 46(1): 23–28.

Bagby

R. M.

Ayearst

L. E.

Morariu

R. A.

Watters

Taylor

G. J.

(2014) The internet administration version of the 20-item Toronto Alexithymia Scale. Psychological Assessment 26(1): 16–22.

Bagby

R. M.

Parker

J. D. A.

Taylor

G. J.

(1994) The twenty-item Toronto Alexithymia scale—I. Item selection and cross-validation of the factor structure. Journal of Psychosomatic Research 38(1): 23–32. . http://doi.org/10.1016/0022-3999(94) 90005-1.

Bagby

R. M.

Taylor

G. J.

Parker

J. D. A.

(1994) The twenty-item Toronto Alexithymia scale—II. Convergent, discriminant, and concurrent validity. Journal of Psychosomatic Research 38(1): 33–40.

Balbinotti

M. A. A.

Wiethaeuper

Barbosa

M. L. L.

Ferreira

A. O.

Gonzalez

R. H.

(2008) Qualidades métricas da escala Toronto de Alexitimia com estudantes de educação física. Coleção Pesquisa em Educação Física 7(3): 345–350.

Besharat

M. A.

(2007) Reliability and factorial validity of a Farsi version of the 20-item Toronto Alexithymia Scale with a sample of Iranian students. Psychological Reports 101(1): 209–220.

10.

Besharat

M. A.

(2008) Assessing reliability and validity of the Farsi version of the Toronto Alexithymia Scale in a sample of substance-using patients. Psychological Reports 102(1): 259–270.

11.

Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness-of-fit in the analysis of covariance structures. Psychological Bulletin, 88, 588–600.

12.

Besharat

M. A.

(2015) Psychometric characteristics of Persian version of the Toronto alexithymia scale-20 in clinical and non-clinical samples. Iranian Journal of Medical Sciences 33(1): 1–6.

13.

Bressi

Taylor

Parker

Bressi

Brambilla

Aguglia

Invernizzi

(1996) Cross validation of the factor structure of the 20-item Toronto Alexithymia Scale: An Italian multicenter study. Journal of Psychosomatic Research 41(6): 551–559.

14.

Brewer

Cook

Cardi

Treasure

Bird

(2015) Emotion recognition deficits in eating disorders are explained by co-occurring alexithymia. Royal Society Open Science 2(1): 140382–140382.

15.

Browne M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K.A. Bollen & J. S. Long (ed.), Testing Structural Equation Models, pp. 136–162. Beverly Hills, CA: Sage.

16.

Carvalho

L. F.

Primi

(2015) Development and internal structure investigation of the Dimensional Clinical Personality Inventory. Psicologia: Reflexão E Crítica 28(2): 322–330.

17.

Cho, Y., & Hong, S. (2013). The new factor structure of the Korean version of the Difficulties in Emotion Regulation Scale (K-DERS) incorporating method factor. Measurement and Evaluation in Counseling and Development, 46(3), 192–201. doi: 10.1177/0748175613484033.

18.

Cohen

(1988) Statistical power analysis for the behavioral sciences, 2nd ed. Hillsdale, NJ: Erlbaum.

19.

Cook

Brewer

Shah

Bird

(2013) Alexithymia, not autism, predicts poor recognition of emotional facial expressions. Psychological Science 24(5): 723–732.

20.

Coutinho, J., Ribeiro, E., Ferreirinha, R., & Dias, P. (2010). The Portuguese version of the Difficulties in Emotion Regulation Scale and its relationship with psychopathological symptoms. Revista de Psiquiatria Clínica, 37(4), 145–151.

21.

Craparo

Faraci

Gori

(2015) Psychometric properties of the 20-item Toronto Alexithymia Scale in a group of Italian younger adolescents. Psychiatry Investigation 12(4): 500–2.

22.

Dere, J., Sun, J., Zhao, Y., Persson, T. J., Zhu, X., Yao, S., … Ryder, A. G. (2013). Beyond “somatization” and “psychologization”: symptom-level variation in depressed Han Chinese and Euro-Canadian outpatients. Frontiers in Psychology, 4, 377. doi:10.3389/fpsyg.2013.00377.

23.

Fernández-Jiménez

Pérez-San-Gregorio

M. Á.

Taylor

G. J.

Bagby

R. M.

Ayearst

L. E.

Izquierdo

(2013) Psychometric properties of a revised Spanish 20-item Toronto Alexithymia Scale adaptation in multiple sclerosis patients. International Journal of Clinical and Health Psychology 13(3): 226–234.

24.

Gallese

(2003) The roots of empathy: The shared manifold hypothesis and the neural basis of intersubjectivity. Psychopathology 36(4): 171–180.

25.

Gerhards

(1998) Decoding of facial expressed emotions in patients with psychosomatic gastrointestinal disorders. Behavior Modification 22: 184–191.

26.

Giromini, L., Velotti, P., de Campora, G., Bonalume, L., & Zavattini, G. C. (2012). Cultural adaptation of the Difficulties in Emotion Regulation Scale: Reliability and validity of an Italian version. Journal of Clinical Psychology, 68, 989–1007.

27.

Giromini, L., de Campora, G., Brusadelli, E., D'Onofrio, E., Zennaro, A., Zavattini, G. C., & Lang, M. (2016). Validity and reliability of the Interpersonal Competence Questionnaire: Empirical evidence from an Italian study. Journal of Psychopathology and Behavioral Assessment, 38, 113–123.

28.

Glenn

Klonsky

E. D.

(2009) Emotion dysregulation as a core feature of borderline personality disorder. Journal of Personality Disorders 23: 20–28.

29.

Gratz

K. L.

Roemer

(2004) Multidimensional assessment of emotion regulation and dysregulation: Development, factor structure, and initial validation of the Difficulties in Emotion Regulation Scale. Journal of Psychopathology and Behavioral Assessment 26(1): 41–54.

30.

Grynberg

Chang

Corneille

Maurage

Vermeulen

Berthoz

Luminet

(2012) Alexithymia and the processing of emotional facial expressions (EFEs): Systematic review, unanswered questions and further perspectives. PLoS ONE 7(8): e42429.

31.

Guttman

H. A.

Laporte

(2002) Alexithymia, empathy, and psychological symptoms in a family context. Comprehensive Psychiatry 43(6): 448–455.

32.

Heaton

Reichenbacher

Sauter

Allen

Scott

Hill

(2012) Measuring the effects of alexithymia on perception of emotional vocalizations in autistic spectrum disorder and typical development. Psychological Medicine 42(11): 2453–2459.

33.

Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55.

34.

John

O. P.

Soto

C. J.

(2007) The importance of being valid: Reliability and the process of construct validation. In: Robins

R. W.

Fraley

R. C.

Krueger

R. F.

(eds) Handbook of research methods in personality psychology, New York: Guilford, pp. 461–494.

35.

Jonason

P. K.

Krause

(2013) The emotional deficits associated with the dark triad traits: Cognitive empathy, affective empathy, and alexithymia. Personality and Individual Differences 55(5): 532–537.

36.

Jongen

Axmacher

Kremers

N. A. W.

Hoffmann

Limbrecht-Ecklundt

Traue

H. C.

Kessler

(2014) An investigation of facial emotion recognition impairments in alexithymia and its neural correlates. Behavioural Brain Research 271: 129–139.

37.

Kessler

Schwarze

Filipic

Traue

H. C.

von Wietersheim

(2006) Alexithymia and facial emotion recognition in patients with eating disorders. International Journal of Eating Disorders 39(3): 245–251.

38.

Kooiman

C. G.

Spinhoven

Trijsburg

R. W.

(2002) The assessment of alexithymia: A critical review of the literature and a psychometric study of the Toronto Alexithymia Scale-20. Journal of Psychosomatic Research 53(6): 1083–1090.

39.

Loas

Cormier

(2009) Personnalité dépendante et alexithymie. Annales Médico-Psychologiques, Revue Psychiatrique 167(2): 110–114.

40.

Loas

Parker

J. D. A.

Otmani

Verrier

Fremaux

(1997) Confirmatory factor analysis of the French translation of the 20-item Toronto Alexithymia Scale. Perceptual and Motor Skills 85(3): 1018–1018.

41.

Loiselle

C. G.

Cossette

(2001) Cross-cultural validation of the Toronto Alexithymia Scale (TAS-20) in U.S. and Peruvian populations. Transcultural Psychiatry 38(3): 348–362.

42.

Maciel

M. J. N.

Yoshida

E. M. P.

(2006) Avaliação de alexitimia, neuroticismo e depressão em dependentes de álcool. Avaliação Psicológica 5(1): 43–54.

43.

Mayer

J. D.

Roberts

R. D.

Barsade

S. G.

(2008) Human abilities: Emotional intelligence. Annual Review of Psychology 59: 507–536.

44.

McLaughlin

K. A.

Mennin

D. S.

Farach

F. J.

(2007) The contributory role of worry in emotion generation and dysregulation in generalized anxiety disorder. Behaviour Research and Therapy 45: 1735–1752.

45.

Michaels

T. M.

Horan

W. P.

Ginger

E. J.

Martinovich

Pinkham

A. E.

Smith

M. J.

(2014) Cognitive empathy contributes to poor social functioning in schizophrenia: Evidence from a new self-report measure of cognitive and affective empathy. Psychiatry Research 220(3): 803–810.

46.

Miguel

F. K.

Bueno

J. M. H.

Noronha

A. P. P.

Couto

Primi

Muniz

(2010) Alexitimia e inteligência emocional: Estudo correlacional. Psicologia: Teoria e Prática 12(3): 52–65.

47.

Miguel, F. K., Giromini, L., Colombarolli, M. S., Zuanazzi, A. C., & Zennaro, A. (2017). A Brazilian investigation of the 36- and 16-item Difficulties in Emotion Regulation Scales. Journal of Clinical Psychology, 73(9), 1146–1159. https://doi.org/10.1002/jclp.22404.

48.

Miguel

F. K.

Primi

(2014) Estudo psicométrico do Teste Informatizado de Percepção de Emoções Primárias. Avaliação Psicológica 13(1): 1–9.

49.

Moriguchi, Y., Decety, J., Ohnishi, T., Maeda, M., Mori, T., Nemoto, K., … Komaki, G. (2007). Empathy and judging other's pain: An fMRI study of alexithymia. Cerebral Cortex, 17(9), 2223–2234. https://doi.org/10.1093/cercor/bhl13.

50.

Nicolò

Semerari

Lysaker

P. H.

Dimaggio

Conti

D’Angerio

Carcione

(2011) Alexithymia in personality disorders: Correlations with symptoms and interpersonal functioning. Psychiatry Research 190(1): 37–42.

51.

Parker

J. D.

Taylor

G. J.

Bagby

R. M.

(1993) Alexithymia and the recognition of facial expressions of emotion. Psychotherapy and Psychosomatics 3–4: 197–202.

52.

Porcelli

Meyer

G. J.

(2002) Construct validity of Rorschach variables for alexithymia. Psychosomatics 43(5): 360–369.

53.

Pourhosein

Rostami

Besharat

M. A.

Mirzamani

(2006) Assessing reliability and validity of Farsi version of the Toronto Alexithymia Scale-20 in a sample of opioid substance use disordered patients. Iranian Journal of Psychiatry 1(4): 133–139.

54.

Primi

Almeida

L. S.

(2000) Bateria de Provas de Raciocínio (BPR-5): Manual técnico, São Paulo: Casa do Psicólogo.

55.

Prkachin

G. C.

Casey

Prkachin

K. M.

(2009) Alexithymia and perception of facial expressions of emotion. Personality and Individual Differences 46(4): 412–417.

56.

Reniers

R. L. E. P.

Corcoran

Drake

Shryane

N. M.

Völlm

B. A.

(2011) The QCAE: A Questionnaire of Cognitive and Affective Empathy. Journal of Personality Assessment 93(1): 84–95.

57.

Ruganci, R. N., & Gençöz, T. (2010). Psychometric properties of a Turkish version of the difficulties in emotion regulation scale. Journal of Clinical Psychology, 66(4), 442–455.

58.

Säkkinen

Kaltiala-Heino

Ranta

Haataja

Joukamaa

(2007) Psychometric properties of the 20-item Toronto Alexithymia Scale and prevalence of alexithymia in a Finnish adolescent population. Psychosomatics 48(2): 154–161.

59.

Sijtsma

(2009) On the use, the misuse, and the very limited usefulness of Cronbach’s alpha. Psychometrika 74: 107–120.

60.

Silani

Bird

Brindley

Singer

Frith

(2008) Levels of emotional awareness and autism: An fMRI study. Social Neuroscience 3(2): 97–112.

61.

Stewart

S. H.

Zvolensky

M. J.

Eifert

G. H.

(2002) The relations of anxiety sensitivity, experiential avoidance, and alexithymic coping to young adults’ motivations for drinking. Behavior Modification 26: 274–296.

62.

Swart

Kortekaas

Aleman

(2009) Dealing with feelings: Characterization of trait alexithymia on emotion regulation strategies and cognitive-emotional processing. PLoS ONE 4(6): e5751.

63.

Taylor

G. J.

(1987) Alexithymia: History validation of the concept. Transcultural Psychiatry 24: 85–95.

64.

Taylor

G. J.

Bagby

R. M.

(2004) New trends in alexithymia research. Psychotherapy and Psychosomatics 73(2): 68–77.

65.

Torres

Guerra

M. P.

Lencastre

Vieira

Roma-Torres

Brandão

(2011) Prevalência da alexitimia na anorexia nervosa e sua associação com variáveis clínicas e sociodemográficas. Jornal Brasileiro de Psiquiatria 60(3): 182–189.

66.

Tsaousis

Taylor

Quilty

Georgiades

Stavrogiannopoulos

Bagby

R. M.

(2010) Validation of a Greek adaptation of the 20-item Toronto Alexithymia Scale. Comprehensive Psychiatry 51(4): 443–448.

67.

van der Velde

Gromann

P. M.

Swart

Wiersma

de Haan

Bruggeman

Aleman

(2015) Alexithymia influences brain activation during emotion perception but not regulation. Social Cognitive and Affective Neuroscience 10(2): 285–293.

68.

Wiethaeuper

Balbinotti

M. A. A.

Pelisoli

Barbosa

M. L. L.

(2005) Estudos da consistência interna e fatorial confirmatório da Escala Toronto de Alexitimia-20 (ETA-20). Interamerican Journal of Psychology 39(2): 221–230.

69.

Zhu, X., Yi, J., Yao, S., Ryder, A. G., Taylor, G. J., & Bagby, R. M. (2007). Cross-cultural validation of a Chinese translation of the 20-item Toronto Alexithymia Scale. Comprehensive Psychiatry, 48(5), 489–496. http://doi.org/10.1016/j.comppsych.2007.04.007.

70.

Zuanazzi

A. C.

Ricci

D. S.

Miguel

F. K.

(2015) Avaliação da alexitimia e percepção emocional: Comparação entre autorrelato e desempenho. Temas em Psicologia 23(4): 831–842.