- Email: [email protected]
Abnormal valence differentiation in alexithymia
Personality and Individual Differences 68 (2014) 102–106
Contents lists available at ScienceDirect
Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid
Abnormal valence differentiation in alexithymia Nancy S. Koven ⇑ Department of Psychology, Bates College, 4 Andrews Road, Lewiston, ME 04240, USA
a r t i c l e
i n f o
Article history: Received 24 December 2013 Received in revised form 4 April 2014 Accepted 9 April 2014
Keywords: Alexithymia Anhedonia Valence Arousal
a b s t r a c t As an at-risk personality trait, alexithymia includes difficulties attending to, distinguishing among, and verbalizing emotions. While these deficits are typically described as extending to both positive and negative emotions, some research suggests a valence-specific pattern such that alexithymic individuals are under-responsive to appetitive stimuli in particular. As this pattern suggests anhedonia, it is important to assess the degree to which alexithymia and anhedonia are confounded as contributors to appetitive hyporeactivity. A nonclinical sample of 96 adults rated word and picture stimuli on dimensions of valence and arousal. Participants were assessed for alexithymia, anhedonia, and mood disturbance with the Toronto Alexithymia Scale, the Chapman Revised Social and Physical Anhedonia Scales, and the Profile of Mood States, respectively. The subset of alexithymic individuals (n = 12) under-rated the valence of appetitive but not aversive stimuli relative to non-alexithymic peers (n = 13). Arousal ratings for all stimulus types were comparable across groups. Hierarchical regression in the full sample indicated that social anhedonia, which was confounded with alexithymia, contributed negligible variance to appetitive valence ratings. This study adds useful detail to the concept of diminished hedonic capacity in alexithymia by localizing the deficiency to poor affective differentiation rather than to hypoarousal. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Considering that alexithymia is characterized, in part, by difficulty identifying and describing emotions, it may be surprising that alexithymia is associated with depression in clinical and nonclinical samples (Lee & Guajardo, 2011; Ogrodniczuk, Sochting, Piper, & Joyce, 2012). Indeed, depression emerges as the most powerful individual predictor of alexithymia in the general population, prompting cautionary advice that depression be recognized as a potential confound in alexithymia research (Honkalampi, Hintikka, Tanskanen, Lehtonen, & Viinamäki, 2000). Despite the close relation, factor analysis positions them as distinct constructs (Hintikka, Honkalampi, Lehtonen, & Viinamäki, 2001). To explain the elevated rate of depression, it is possible that alexithymic, like depressed (Epp, Dobson, Dozois, & Frewen, 2012), individuals are vigilant for and/or overly reactive to aversive cues in the environment. However, research using emotional Stroop paradigms shows that alexithymic individuals have similar levels of attentional interference to general aversive stimuli compared to their non-alexithymic peers (Lundh & SimonssonSarnecki, 2002; Mueller, Alpers, & Reim, 2006; Pandey, 1995; Parker, Taylor, & Bagby, 1993), suggesting a normative degree of ⇑ Tel.: +1 207 786 6426; fax: +1 207 786 8338. E-mail address: [email protected] http://dx.doi.org/10.1016/j.paid.2014.04.007 0191-8869/Ó 2014 Elsevier Ltd. All rights reserved.
threat processing. Another possibility is that negative affect in alexithymia is secondary to deficient hedonic capacity (Dubey & Pandey, 2013; Prince & Berenbaum, 1993) such that depression results from a failure to attend to positive stimuli and/or process positive life events. Inattention to appetitive stimuli has previously been demonstrated in depressed individuals. For instance, relative to healthy peers, depressed subjects recall fewer pleasant words despite comparable recall of unpleasant words (Sloan, Strauss, & Wisner, 2001) and are less responsive to rewards (e.g., Pizzagalli, Iosifescu, Hallett, Ratner, & Fava, 2008). Such lack of reactivity to appetitive stimuli in depression is a more specific characteristic called anhedonia, and the suggestion of reduced appetitive responsivity in alexithymic individuals calls into question the construct boundaries between alexithymia and anhedonia. The little research to address discriminant validity among alexithymia and anhedonia measures presents mixed results. Loas, Fremaux, and Boyer (1997), for example, found no correlation in a nonclinical sample between a physical anhedonia scale and two facets of the Toronto Alexithymia Scale (TAS-20; Bagby, Parker, & Taylor, 1994), even after controlling for physical displeasure and depression. However, Deborde et al. (2006) found that TAS-20 scores were positively correlated with physical anhedonia scores and that the Bermond-Vorst Alexithymia Questionnaire was positively correlated with both physical and social anhedonia among healthy controls. Consistent with this, Prince
N.S. Koven / Personality and Individual Differences 68 (2014) 102–106
and Berenbaum (1993) found aspects of alexithymia to be associated with social anhedonia, even after controlling for depression and negative affect. It is unclear, however, to what extent anhedonia is confounded with alexithymia in predicting behavioral hyporeactivity to appetitive stimuli. The purpose of this study was to examine how subjective ratings of valence and arousal for appetitive, neutral, and aversive stimuli vary as a function of alexithymia, while controlling for anhedonia and mood disturbance. Given that few studies have provided evidence of reduced hedonic capacity in alexithymia, it is important to confirm the effect of selective hyporeactivity to appetitive cues and to determine if such an effect generalizes across verbal (word) and nonverbal (picture) stimulus types. A nonclinical sample of young adults was chosen to minimize the effects of confounding variables beyond those of interest. It was hypothesized that (1) high-alexithymic individuals would rate the valence and arousal of appetitive, but not aversive or neutral, verbal and nonverbal stimuli more weakly than would their non-alexithymic counterparts; (2) given that alexithymia and anhedonia are related yet distinct constructs in the general population, it was expected that an inverse relationship between alexithymia and reactivity to appetitive cues would be independent of contributions by anhedonia and/or mood disturbance. 2. Materials and methods 2.1. Participants The sample included 96 students (59.4% women) who received undergraduate course credit for participation. Participants ranged in age from 18 to 22 (M = 18.9, SD = 1.4) years, with 62% reporting their ethnicity as White, 21% Asian/Asian-American, 12% Latino, 4% Black, and 1% biracial. Individuals were excluded if they reported non-fluency in English, uncorrected vision, current or past diagnosed psychiatric/neurologic illness, or recent use of prescribed (e.g., methylphenidate) or illicit substances (e.g., cannabis) known to affect cognition or emotion. 2.2. Procedure After providing informed consent, participants completed selfreport measures in a pseudorandom order, including the TAS-20, the Profile of Mood States (McNair, Lorr, & Droppleman, 1992), and the Chapman Revised Social (Mishlove & Chapman, 1985) and Physical (Chapman, Chapman, & Raulin, 1976) Anhedonia Scales, and then evaluated affective stimuli. 2.3. Self-report measures Alexithymia status was determined with the TAS-20, a 20-item scale that assesses difficulty identifying (e.g., ‘‘I don’t know what’s going on inside me’’) and verbalizing feelings (e.g., ‘‘It is difficult for me to find the right words for my feelings’’) as well as externallyoriented thinking (e.g., ‘‘I prefer to analyze problems rather than just describe them’’) using a 5-point Likert scale. Total scores can range from 20 to 100. In keeping with earlier studies (Hesse & Floyd, 2011; Hesse et al., 2013; Joukamaa et al., 2007), participants with scores P60 (i.e., 60th percentile of the possible total score) were considered alexithymic. Full-scale Cronbach’s alpha was .80. The POMS was included in the protocol to assess and control for potential differences in negative affect between alexithymic and non-alexithymic participants. The POMS contains 65 adjectives, each rated on a 5-point Likert-type scale, that comprise six subscales: Tension/Anxiety (e.g., ‘‘nervous’’), Depression/Dejection (e.g., ‘‘gloomy’’), Anger/Hostility (e.g., ‘‘peeved’’), Vigor/Activity
103
(e.g., ‘‘alert’’), Fatigue/Inertia (e.g., ‘‘sluggish’’), and Confusion/ Bewilderment (e.g., ‘‘muddled’’). Total mood disturbance (TMD), calculated by subtracting the Vigor/Activity score from the sum of the other five scores, reflects overall negative affect at the time of testing and was the variable of interest in this study. A TMD zscore was calculated (M = 0, SD = 1) based on sex-related norms (McNair et al., 1992); higher z-scores reflect greater mood disturbance. Cronbach’s alpha for TMD was .85. The Chapman Revised Anhedonia Scales were included in order to track anhedonia levels among alexithymic and nonalexithymic participants. The 40-item Social Anhedonia scale and the 61-item Physical Anhedonia scale assess the ability to experience pleasure from human interaction (e.g., ‘‘Having close friends is not as important as many people say’’) and physical stimuli (e.g., ‘‘The beauty of sunsets is greatly overrated’’), respectively. In each case, items are presented in true/false format; higher scores reflect greater anhedonia. Cronbach’s alpha was 0.78 and 0.82, respectively.
2.4. Affective judgment tasks Individuals were seated 24 inches in front of a 20-inch computer screen for two tasks run on Dell OptiPlex™ GX520 computers. For each task, participants were asked to evaluate the valence of the stimulus on a Likert scale of 1 (extremely negative) to 9 (extremely positive) followed by the intensity of the stimulus on a Likert-type scale of 1 (not intense at all) to 9 (strongest intensity) and to record their answers in a response booklet. Having stimulus presentation on the computer and the recording of responses on paper allowed participants to freely view each stimulus while deciding how to rank its valence and arousal. Tasks were untimed and self-paced. Half of the participants received the word task first, and the other half received the picture task first, enabling testing for order effects. Trials were presented in a pseudorandom order such that no more than two items from the same stimulus category would be contiguous. A total of 150 words were selected from the Affective Norms for English Words (ANEW; Bradley & Lang, 1999): 50 appetitive (e.g., ecstasy), 50 neutral (e.g., fabric), and 50 aversive words (e.g., killer). Words ranged from three to eight letters and were balanced across categories for word length, frequency of usage in the English language, and orthographic neighborhood (Table 1). Words were selected based on norms from the original ANEW validation study in which a mixed gender, nonclinical, adult sample rated words for valence and arousal with a similar 1 (low) to 9 (high) scale. Specifically, appetitive words were selected to be of high valence (close to 9) and high arousal (close to 9), aversive words to be of low valence (close to 1) and high arousal (close to 9), and neutral words to be of moderate valence (close to 5) and low arousal (close to 1). The average valence for the appetitive and aversive word sets was equally polarized (i.e., equidistant from the middle of the scale) and arousing. Words were presented individually in capital letters using Calibri 112-point black font atop a white background. Picture stimuli were drawn from the International Affective Picture System (IAPS; Lang, Bradley, & Cuthbert, 2005) and included 50 appetitive (e.g., laughing children), 50 neutral (e.g., household objects), and 50 aversive images (e.g., mutilated bodies). Pictures were chosen following the same principles as for the word set (Table 1) and guided by adult, nonclinical, mixed-gender norms from the original IAPS validation study. The appetitive and aversive image sets were equally polarized in valence and equally arousing. Appetitive, aversive, and neutral pictures were comparable in valence and arousal to appetitive, aversive, and neutral words, respectively.
104
N.S. Koven / Personality and Individual Differences 68 (2014) 102–106
Table 1 Characteristics of stimuli used in affective judgment tasks. Aversive stimuli
Valence Arousal Frequency Word length Orthographic neighborhood
Neutral stimuli
Appetitive stimuli
Words M ± SD
Pictures M ± SD
Words M ± SD
Pictures M ± SD
Words M ± SD
Pictures M ± SD
2.27 ± 0.46 6.53 ± 0.65 51.14 ± 72.78 5.52 ± 1.45 3.88 ± 5.12
2.19 ± 0.48 6.39 ± 0.65
5.15 ± 0.27 3.77 ± 0.36 50.12 ± 58.70 5.46 ± 1.42 3.74 ± 4.95
5.06 ± 0.29 3.80 ± 0.52
7.87 ± 0.44 6.46 ± 0.53 51.78 ± 66.75 5.60 ± 1.46 3.78 ± 4.32
7.74 ± 0.50 6.38 ± 0.76
Note: For word stimuli, values for valence, arousal, and frequency were derived from adult, mixed-sex norms associated with the ANEW list. Valence and arousal are measured on a 1–9 (low–high) scale. Frequency indicates how often a word is used in the English language, with higher numbers indicating greater frequency. Values for orthographic neighborhood, derived from the English Lexicon Project (Balota et al., 2007), refer to the number of additional words that can be derived from the target word by changing a single letter. For picture stimuli, valence and arousal values were derived from adult, mixed-sex norms associated with the IAPS using the same scaling structure.
2.5. Data analysis Analyses were computed using the Statistical Package for the Social Sciences (SPSS) software version 21.0 (Armonk, NY, USA). Normality was checked in the full sample with Kolmogorov–Smirnov tests and inspection of skew and kurtosis values. Participants with TAS-20 scores P60 (i.e., a threshold equating to 1.25 SD units above the sample mean) formed the alexithymic group (n = 12; 66.7% women). Participants with TAS-20 scores 635 (i.e., a value 1.25 SD units below the mean) formed the non-alexithymic group (n = 13; 69.2% women). Gender composition was comparable across groups, v2(1) with Yates’ correction = 0.09, p = ns. Affective ratings were analyzed with two multivariate analyses of variance (MANOVA) with alexithymia status as the grouping factor. The six valence and arousal variables were included as dependent variables in the first and second analysis, respectively, in order to determine if systematic bias existed in the evaluation of valence separate from any bias in the evaluation of arousal. Separating the valence from the arousal ratings across analyses also reduces multicollinearity, as valence and arousal ratings for the same stimulus category would be strongly correlated. One-way analyses of variance (ANOVA) were used to determine if TMD and anhedonia scores varied across groups. Given significant differences in social anhedonia, hierarchical regression in the full sample was used to analyze the relative contributions of alexithymia and social anhedonia as predictors of affective ratings. This method was chosen over analysis of covariance, as covariates should not be used to adjust for pre-existing differences between groups (Suckling, 2011). Using the results from the aforementioned MANOVAs as a guide for selection of dependent variables, the goal of the regression was to identify whether TAS-20 score contributed unique variance to affective ratings beyond that of the Social Anhedonia score. 3. Results 3.1. Descriptive statistics Table 2 provides descriptive statistics for all variables. Means and standard deviations for the TAS-20, Social Anhedonia, and Physical Anhedonia Scales were consistent with those from other young adult samples (Chmielewski, Fernandes, Yee, & Miller, 1995; Coffey, Berenbaum, & Kerns, 2003; Kwapil, Barrantes-Vidal, & Silvia, 2008). The TAS-20 and Physical Anhedonia scores were normally-distributed. The Social Anhedonia score was positively skewed but successfully normalized with logarithmic transformation. ANOVA showed that valence and arousal ratings for words and pictures did not vary as a function of presentation order, Wilks’ k = .78, F(12, 83) = 1.13, p = ns. Independent samples t-tests indi-
cated that the current sample rated appetitive words and pictures as less positive/less intense, aversive words as less intense, and aversive pictures as more negative/more intense relative to the original ANEW and IAPS normative cohorts; ratings of neutral stimuli differed across the cohorts as well (Table 2). Valence ratings for aversive words and pictures and arousal ratings for neutral words were positively skewed but subsequently corrected with logarithmic transformation. Valence ratings for neutral words were negatively skewed but corrected with a logarithmic transformation after subtracting a constant from each score to achieve a minimum score of one. All other ratings were normally-distributed. Transformed variables were used in lieu of untransformed variables for the remaining analyses. 3.2. Affective judgment as a function of alexithymia As seen in Table 2, ANOVA showed no group differences in TMD score, suggesting comparable degrees of negative affect at the time of testing. The alexithymic group had elevated Social but not Physical Anhedonia scores relative to the non-alexithymic group. The full model for the MANOVA with the six valence ratings as dependent variables was significant, Wilks’ k = .25, F(6, 18) = 8.18, p < .001, partial N2 = .75. Tests of between-subjects effects showed that the alexithymic group rated appetitive words, F(1, 23) = 12.08, p = .002, partial N2 = .37, and appetitive pictures, F(1, 23) = 5.65, p = .027, partial N2 = .21, less strongly than did the non-alexithymic group. The alexithymic group also rated neutral pictures as lower in valence than the non-alexithymic group, F(1, 23) = 6.69, p = .017, partial N2 = .24. Of these three effects, only the first one remained significant after Bonferroni correction (adjusted a = .008). There were no group differences in ratings of aversive pictures, aversive words, or neutral words. The full model for the second MANOVA indicated no group differences in arousal ratings, Wilks’ k = .76, F(6, 18) = .56, p = ns. The assumption of similar covariance matrix across groups was upheld for both the first, Box’s M = 44.89, F(23, 1592) = 1.46, p = ns, and second MANOVA, Box’s M = 40.80, F(23, 1592) = 1.33, p = ns. The assumption of similar error variances across groups for each dependent variable was upheld in both analyses (all p-values in Levene’s tests = ns). Given that the two groups differed in social anhedonia, a hierarchical regression was used to assess the contribution of alexithymia to appetitive word valence ratings in the full sample after controlling for social anhedonia. When added last to the regression mode, the TAS-20 score contributed significant incremental variance, DR2 = .054, DF(1, 93) = 4.23, p = .04. In the final model, the TAS-20 score, adjusted for its shared variance with the Social Anhedonia score, remained a significant negative predictor, beta = .26, p = .04, whereas the Social Anhedonia score, controlling for its shared variance with the TAS-20 score, was insignificant, beta = .05, p = ns.
105
N.S. Koven / Personality and Individual Differences 68 (2014) 102–106 Table 2 Descriptive statistics in the full sample and across groups. Variable
Full sample (n = 96) M ± SD
Alexithymic (n = 12) M ± SD
Non-alexithymic (n = 13) M ± SD
F
t
TAS-20 total Social Anhedonia Physical Anhedonia POMS TMD Appetitive word valence Appetitive word arousal Aversive word valence Aversive word arousal Neutral word valence Neutral word arousal Appetitive picture valence Appetitive picture arousal Aversive picture valence Aversive picture arousal Neutral picture valence Neutral picture arousal
47.1 ± 10.1 7.2 ± 5.9 11.2 ± 6.9 0.2 ± 0.8 6.9 ± 0.8 5.8 ± 1.2 2.3 ± 0.9 5.7 ± 1.6 5.2 ± 0.3 2.7 ± 1.6 6.8 ± 0.6 5.4 ± 1.3 1.8 ± 0.7 6.6 ± 1.7 5.3 ± 0.3 2.8 ± 1.3
64.1 ± 6.2 8.8 ± 4.8 13.0 ± 6.6 0.2 ± 0.7 6.3 ± 0.7 5.2 ± 1.5 2.8 ± 1.1 4.9 ± 1.4 5.2 ± 0.3 2.8 ± 2.0 6.4 ± 0.6 5.0 ± 1.5 1.8 ± 0.4 5.7 ± 2.0 5.2 ± 0.2 2.5 ± 1.4
33.7 ± 2.5 5.1 ± 3.7 10.7 ± 7.0 0 ± 0.7 7.2 ± 0.5 5.6 ± 1.1 2.2 ± 0.5 5.6 ± 1.4 5.4 ± 0.2 2.2 ± 1.3 6.9 ± 0.4 5.0 ± 1.1 1.8 ± 0.5 6.6 ± 1.4 5.5 ± 0.2 2.4 ± 1.1
267.42** 5.56* 1.22 1.11 12.08** 0.34 3.33 0.69 3.61 0.18 5.65* .03 .02 2.05 6.69* .20
18.83** 10.00** 0.55 9.96** 1.72 17.16** 17.27** 11.17** 7.26** 2.46* 7.72** 14.92**
Note: TAS-20 = Toronto Alexithymia Scale; POMS TMD = Profile of Mood States Total Mood Disturbance. Means and standard deviations reported here are from untransformed variables. The F statistic tested for group differences using transformed variables when appropriate. F-values for psychological variables were generated with ANOVA. Group differences in stimulus ratings were assessed with MANOVA, with the F-values drawn from tests of between-subjects effects. T-values are derived from independent samples t-tests, comparing stimulus ratings from the full sample to normed ratings from the ANEW and IAPS sets. * p < .05. ** p < .005.
4. Discussion This study confirmed that alexithymic, relative to non-alexithymic, individuals under-report the valence of appetitive stimuli. Although only the group difference for appetitive words survived Bonferroni correction, there was a similar but weaker valence difference for appetitive pictures, suggesting generalization across verbal and nonverbal content. However, the alexithymic group was more socially anhedonic than the non-alexithymic group, making it difficult to distinguish whether the above effects were secondary to alexithymia or social anhedonia. Hierarchical regression using the full sample determined that alexithymia contributed unique variance to appetitive valence ratings after controlling for social anhedonia. Taken together, it appears that alexithymia is characterized by diminished responsivity to appetitive stimuli, independent of anhedonia. One cannot conclude from these data, however, that alexithymic individuals have limited hedonic capacity, as the alexithymic group reported a level of arousal in response to appetitive stimuli comparable to that by the non-alexithymic group. The fact that the point of difference between groups was valence but not arousal suggests, instead, that the deficiency is one of affective differentiation. In other words, alexithymic individuals are sufficiently aroused by appetitive cues but struggle with the ability to ‘‘gauge the temperature’’ of this arousal to assign it a higher valence. That valence and arousal are separable is certainly not alien to personality theory or the psychopathology literature (Heller, Schmidtke, Nitschke, Koven, & Miller, 2002) and, indeed, their orthogonality is at the core of the valence-arousal model of emotion (Russell, 1980). Neuroimaging research also supports this distinction, with evidence suggesting that the evaluation of valence and arousal recruits different neural circuits in nonclinical samples (Gerber et al., 2008). Moreover, there is evidence (Nielen et al., 2009) of differential processing for appetitive (middle temporal and orbitofrontal cortex) and aversive (visual and lateral prefrontal regions) stimuli in healthy young adults as well as distinct brain regions for the conjoint processing of highly-arousing appetitive stimuli (occipital cortex, parahippocampal gyrus, and posterior cingulate) versus highly-arousing aversive stimuli (anterior insula).
It is difficult to ascertain the extent to which these findings apply to alexithymia, as most neuroimaging studies with alexithymic subjects overlook appetitive stimuli altogether and/or do not present stimuli in such a way as to separate contributions by valence and arousal to the resulting data. Of those with appropriate designs, Hesse et al. (2013) reported weaker activation in alexithymic participants in response to appetitive IAPS stimuli in several brain regions known for socioemotional perception. However, these authors did not assess anhedonia and did not include aversive stimuli, making it difficult to know if this pattern of hypoactivation is specific to the alexithymia construct and/or to appetitive stimuli. In a similar study, Heinzel et al. (2010) identified dorsal anterior cingulate activation in non-depressed alexithymic individuals in response to positive-valenced IAPS stimuli, which is noteworthy given the role of this brain region in purposeful inhibition of arousal (Beauregard, Lévesque, & Bourgouin, 2001). In application to the current data, this result raises the interesting possibility that the lower appetitive valence ratings by alexithymic individuals resulted from active suppression of arousal generated by those same stimuli, a premise consistent with the description of the ‘‘Type II’’ alexithymia subtype (Bagby et al., 2009); why this should be specific to appetitive stimuli, however, is an open question. As the present study was powered to find large effect sizes only, it is possible that alexithymic individuals exhibit mildly deficient responsivity to aversive stimuli, but, at the same time, the lack of a between-groups effect for aversive stimuli is consistent with other reports of normal threat processing in alexithymia (Lundh & Simonsson-Sarnecki, 2002; Mueller et al., 2006; Pandey, 1995; Parker et al., 1993). It is noteworthy that the present sample as a whole differed from the original ANEW and IAPS normative cohorts in their evaluation of most of the stimulus categories. While the ratings were such that the stimuli were judged to belong to the originallyintended category (e.g., aversive stimuli were still deemed aversive), the statistical differences may suggest a generational shift in the affective judgments of present-day young adults. Although the ANEW and IAPS norms were generated as recently as 1999 and 2005, respectively, the current results might signal a need for updated norms for these popular stimulus sets.
106
N.S. Koven / Personality and Individual Differences 68 (2014) 102–106
Although this study adds useful detail to the concept of diminished hedonic capacity in alexithymia, several limitations of the study must be acknowledged. First, even though stimuli were carefully chosen based on established norms, low arousal stimuli were not included and, therefore, it is unknown whether subjective ratings of appetitive valence by alexithymic individuals change with decreased intensity and if any change would differ from patterns found in non-alexithymic individuals. Second, although selection of a healthy sample was purposeful, the restricted age range limits the generalizability of these results to older adult cohorts. Moreover, while this study suggests that anhedonia is non-contributory to appetitive valence ratings in alexithymia, the reliance on a cross-sectional design with self-report methods curtails any conclusions concerning the underlying mechanisms of or the etiological reasons for abnormal valence differentiation. Future studies that combine neuroimaging, assessment of possible confounds, and balanced emotion stimuli in larger, community-based samples will be well positioned to characterize the physiological and cognitive interface of this affective style as well as to assess the generalizability of effects across potential alexithymia subtypes. References Bagby, R. M., Parker, J. D. A., & Taylor, G. L. (1994). The twenty-item Toronto Alexithymia Scale - 1: item selection and cross-validation of the factor structure. Journal of Psychosomatic Research, 38, 23–32. http://dx.doi.org/ 10.1016/0022-3999(94)90005-1. Bagby, R., Quilty, L. C., Taylor, G. J., Grabe, H. J., Luminet, O., Verissimo, R., et al. (2009). Are there subtypes of alexithymia? Personality and Individual Differences, 47(5), 413–418. http://dx.doi.org/10.1016/j.paid.2009.04.012. Balota, D. A., Yap, M. J., Cortese, M. J., Hutchison, K. A., Kessler, B., Loftis, B., et al. (2007). The English lexicon project. Behavior Research Methods, 39(3), 445–459. http://dx.doi.org/10.3758/BF03193014. Beauregard, M., Lévesque, J., & Bourgouin, P. (2001). Neural correlates of conscious self-regulation of emotion. The Journal of Neuroscience, 21(18), 6993–7000. Bradley, M. M., & Lang, P. J. (1999). Affective norms for English words (ANEW): Stimuli, instruction manual, and affective ratings. Gainesville, FL, USA: The Center for Research in Psychophysiology, University of Florida. Chapman, L. J., Chapman, J. P., & Raulin, M. L. (1976). Scales for physical and social anhedonia. Journal of Abnormal Psychology, 85(4), 374–382. http://dx.doi.org/ 10.1037/0021-843X.85.4.374. Chmielewski, P. M., Fernandes, L. O. L., Yee, C. M., & Miller, G. A. (1995). Ethnicity and gender in scales of psychosis proneness and mood disorders. Journal of Abnormal Psychology, 104(3), 464–470. http://dx.doi.org/10.1037/0021843X.104.3.464. Coffey, E., Berenbaum, H., & Kerns, J. G. (2003). The dimensions of emotional intelligence, alexithymia, and mood awareness: Associations with personality and performance on an emotional Stroop task. Cognition and Emotion, 17(4), 671–679. http://dx.doi.org/10.1080/02699930302304. Deborde, A. S., Berthoz, S., Godart, N., Perdereau, F., Corcos, M., & Jeammet, P. (2006). Étude des relations entre alexithymie et anhédonie chez des femmes présentant des troubles du comportement alimentaire et chez des témoins. L’Encéphale: Revue De Psychiatrie Clinique Biologique Et Thérapeutique, 32(1), 83–91. doi:10.1016/S0013-7006(06)76140-1. Dubey, A., & Pandey, R. (2013). Mental health problems in alexithymia: Role of positive and negative emotional experiences. Journal of Projective Psychology and Mental Health, 20(2), 128–136. Epp, A. M., Dobson, K. S., Dozois, D. J. A., & Frewen, P. A. (2012). A systematic metaanalysis of the Stroop task in depression. Clinical Psychology Review, 32(4), 316–328. http://dx.doi.org/10.1016/j.cpr.2012.02.005. Gerber, A. J., Posner, J., Gorman, D., Colibazzi, T., Yu, S., Wang, Z., et al. (2008). An affective circumplex model of neural systems subserving valence, arousal, and cognitive overlay during the appraisal of emotional faces. Neuropsychologia, 46(8), 2129–2139. http://dx.doi.org/10.1016/j.neuropsychologia.2008.02.032. Heinzel, A., Schäfer, R., Müller, H., Schieffer, A., Ingenhag, A., Northoff, G., et al. (2010). Differential modulation of valence and arousal in high-alexithymic and low-alexithymic individuals. NeuroReport, 21(15), 998–1002.
Heller, W., Schmidtke, J. I., Nitschke, J. B., Koven, N. S., & Miller, G. A. (2002). States, traits, and symptoms: Investigating the neural correlates of emotion, personality, and psychopathology. In D. Cervone, & W. Mischel (Eds.), (pp. 106–126). New York, NY, USA: Guilford Press. Hesse, C., & Floyd, K. (2011). The impact of alexithymia on initial interactions. Personal Relationships, 18(3), 453–470. http://dx.doi.org/10.1111/j.14756811.2010.01311.x. Hesse, C., Floyd, K., Rauscher, E. A., Frye-Cox, N., Hegarty, J. P., & Peng, H. (2013). Alexithymia and impairment of decoding positive affect: An fMRI study. Journal of Communication, 63(4), 786–806. http://dx.doi.org/10.1111/jcom.12039. Hintikka, J., Honkalampi, K., Lehtonen, J., & Viinamäki, H. (2001). Are alexithymia and depression distinct or overlapping constructs? A study in a general population. Comprehensive Psychiatry, 42(3), 234–239. http://dx.doi.org/ 10.1053/comp.2001.23147. Honkalampi, K., Hintikka, J., Tanskanen, A., Lehtonen, J., & Viinamäki, H. (2000). Depression is strongly associated with alexithymia in the general population. Journal of Psychosomatic Research, 48(1), 99–104. http://dx.doi.org/10.1016/ S0022-3999(99)00083-5. Joukamaa, M., Taanila, A., Miettunen, J., Karvonen, J. T., Koskinen, M., & Veijola, J. (2007). Epidemiology of alexithymia among adolescents. Journal of Psychosomatic Research, 63(4), 373–376. http://dx.doi.org/10.1016/ j.jpsychores.2007.01.018. Kwapil, T. R., Barrantes-Vidal, N., & Silvia, P. J. (2008). The dimensional structure of the Wisconsin Schizotypy Scales: Factor identification and construct validity. Schizophrenia Bulletin, 34(3), 444–457. http://dx.doi.org/10.1093/schbul/ sbm098. Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2005). International affective picture system (IAPS): Affective ratings of pictures and instruction manual. Gainesville, FL, USA: University of Florida. Lee, S. A., & Guajardo, N. R. (2011). Affect intensity and alexithymia differentially influence the relationship between neuroticism and depressive symptomatology among college students. Personality and Individual Differences, 50(5), 646–650. http://dx.doi.org/10.1016/j.paid.2010.11.036. Loas, G., Fremaux, D., & Boyer, P. (1997). Anhedonia and alexithymia: Distinct or overlapping constructs. Perceptual and Motor Skills, 84(2), 415–425. http:// dx.doi.org/10.2466/pms.1997.84.2.415. Lundh, L., & Simonsson-Sarnecki, M. (2002). Alexithymia and cognitive bias for emotional information. Personality and Individual Differences, 32(6), 1063–1075. http://dx.doi.org/10.1016/S0191-8869(01)00110-6. McNair, D. M., Lorr, M., & Droppleman, L. F. (1992). Profile of mood states manual. San Diego, CA, USA: Educational and Industrial Testing Service. Mishlove, M., & Chapman, L. J. (1985). Social anhedonia in the prediction of psychosis proneness. Journal of Abnormal Psychology, 94(3), 384–396. http:// dx.doi.org/10.1037/0021-843X.94.3.384. Mueller, J., Alpers, G. W., & Reim, N. (2006). Dissociation of rated emotional valence and Stroop interference in observer-rated alexithymia. Journal of Psychosomatic Research, 61(2), 261–269. http://dx.doi.org/10.1016/j.jpsychores.2006.02.017. Nielen, M. M., Heslenfeld, D. J., Heinen, K., Van Strien, J. W., Witter, M. P., Jonker, C., et al. (2009). Distinct brain systems underlie the processing of valence and arousal of affective pictures. Brain and Cognition, 71(3), 387–396. Ogrodniczuk, J. S., Sochting, I., Piper, W. E., & Joyce, A. S. (2012). A naturalistic study of alexithymia among psychiatric outpatients treated in an integrated group therapy program. Psychology and Psychotherapy: Theory, Research, and Practice, 85(3), 278–291. http://dx.doi.org/10.1111/j.2044-8341.2011.02032.x. Pandey, R. (1995). Stroop interference effect of emotion-arousing words in alexithymia. Journal of the Indian Academy of Applied Psychology, 21(1), 21–28. Parker, J. D. A., Taylor, G., & Bagby, M. (1993). Alexithymia and the processing of emotional stimuli: An experimental study. New Trends in Experimental and Clinical Psychiatry, 9(1–2), 9–14. Pizzagalli, D. A., Iosifescu, D., Hallett, L. A., Ratner, K. G., & Fava, M. (2008). Reduced hedonic capacity in major depressive disorder: Evidence from a probabilistic reward task. Journal of Psychiatric Research, 43(1), 76–87. http://dx.doi.org/ 10.1016/j.jpsychires.2008.03.001. Prince, J. D., & Berenbaum, H. (1993). Alexithymia and hedonic capacity. Journal of Research in Personality, 27(1), 15–22. http://dx.doi.org/10.1006/jrpe.1993.1002. Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39(6), 1161–1178. Sloan, D. M., Strauss, M. E., & Wisner, K. L. (2001). Diminished response to pleasant stimuli by depressed women. Journal of Abnormal Psychology, 110(3), 488–493. http://dx.doi.org/10.1037/0021-843X.110.3.488. Suckling, J. (2011). Correlated covariates in ANCOVA cannot adjust for pre-existing differences between groups. Schizophrenia Research, 126(1–3), 310–311. http:// dx.doi.org/10.1016/j.schres.2010.08.034.
Contents lists available at ScienceDirect
Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid
Abnormal valence differentiation in alexithymia Nancy S. Koven ⇑ Department of Psychology, Bates College, 4 Andrews Road, Lewiston, ME 04240, USA
a r t i c l e
i n f o
Article history: Received 24 December 2013 Received in revised form 4 April 2014 Accepted 9 April 2014
Keywords: Alexithymia Anhedonia Valence Arousal
a b s t r a c t As an at-risk personality trait, alexithymia includes difficulties attending to, distinguishing among, and verbalizing emotions. While these deficits are typically described as extending to both positive and negative emotions, some research suggests a valence-specific pattern such that alexithymic individuals are under-responsive to appetitive stimuli in particular. As this pattern suggests anhedonia, it is important to assess the degree to which alexithymia and anhedonia are confounded as contributors to appetitive hyporeactivity. A nonclinical sample of 96 adults rated word and picture stimuli on dimensions of valence and arousal. Participants were assessed for alexithymia, anhedonia, and mood disturbance with the Toronto Alexithymia Scale, the Chapman Revised Social and Physical Anhedonia Scales, and the Profile of Mood States, respectively. The subset of alexithymic individuals (n = 12) under-rated the valence of appetitive but not aversive stimuli relative to non-alexithymic peers (n = 13). Arousal ratings for all stimulus types were comparable across groups. Hierarchical regression in the full sample indicated that social anhedonia, which was confounded with alexithymia, contributed negligible variance to appetitive valence ratings. This study adds useful detail to the concept of diminished hedonic capacity in alexithymia by localizing the deficiency to poor affective differentiation rather than to hypoarousal. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Considering that alexithymia is characterized, in part, by difficulty identifying and describing emotions, it may be surprising that alexithymia is associated with depression in clinical and nonclinical samples (Lee & Guajardo, 2011; Ogrodniczuk, Sochting, Piper, & Joyce, 2012). Indeed, depression emerges as the most powerful individual predictor of alexithymia in the general population, prompting cautionary advice that depression be recognized as a potential confound in alexithymia research (Honkalampi, Hintikka, Tanskanen, Lehtonen, & Viinamäki, 2000). Despite the close relation, factor analysis positions them as distinct constructs (Hintikka, Honkalampi, Lehtonen, & Viinamäki, 2001). To explain the elevated rate of depression, it is possible that alexithymic, like depressed (Epp, Dobson, Dozois, & Frewen, 2012), individuals are vigilant for and/or overly reactive to aversive cues in the environment. However, research using emotional Stroop paradigms shows that alexithymic individuals have similar levels of attentional interference to general aversive stimuli compared to their non-alexithymic peers (Lundh & SimonssonSarnecki, 2002; Mueller, Alpers, & Reim, 2006; Pandey, 1995; Parker, Taylor, & Bagby, 1993), suggesting a normative degree of ⇑ Tel.: +1 207 786 6426; fax: +1 207 786 8338. E-mail address: [email protected] http://dx.doi.org/10.1016/j.paid.2014.04.007 0191-8869/Ó 2014 Elsevier Ltd. All rights reserved.
threat processing. Another possibility is that negative affect in alexithymia is secondary to deficient hedonic capacity (Dubey & Pandey, 2013; Prince & Berenbaum, 1993) such that depression results from a failure to attend to positive stimuli and/or process positive life events. Inattention to appetitive stimuli has previously been demonstrated in depressed individuals. For instance, relative to healthy peers, depressed subjects recall fewer pleasant words despite comparable recall of unpleasant words (Sloan, Strauss, & Wisner, 2001) and are less responsive to rewards (e.g., Pizzagalli, Iosifescu, Hallett, Ratner, & Fava, 2008). Such lack of reactivity to appetitive stimuli in depression is a more specific characteristic called anhedonia, and the suggestion of reduced appetitive responsivity in alexithymic individuals calls into question the construct boundaries between alexithymia and anhedonia. The little research to address discriminant validity among alexithymia and anhedonia measures presents mixed results. Loas, Fremaux, and Boyer (1997), for example, found no correlation in a nonclinical sample between a physical anhedonia scale and two facets of the Toronto Alexithymia Scale (TAS-20; Bagby, Parker, & Taylor, 1994), even after controlling for physical displeasure and depression. However, Deborde et al. (2006) found that TAS-20 scores were positively correlated with physical anhedonia scores and that the Bermond-Vorst Alexithymia Questionnaire was positively correlated with both physical and social anhedonia among healthy controls. Consistent with this, Prince
N.S. Koven / Personality and Individual Differences 68 (2014) 102–106
and Berenbaum (1993) found aspects of alexithymia to be associated with social anhedonia, even after controlling for depression and negative affect. It is unclear, however, to what extent anhedonia is confounded with alexithymia in predicting behavioral hyporeactivity to appetitive stimuli. The purpose of this study was to examine how subjective ratings of valence and arousal for appetitive, neutral, and aversive stimuli vary as a function of alexithymia, while controlling for anhedonia and mood disturbance. Given that few studies have provided evidence of reduced hedonic capacity in alexithymia, it is important to confirm the effect of selective hyporeactivity to appetitive cues and to determine if such an effect generalizes across verbal (word) and nonverbal (picture) stimulus types. A nonclinical sample of young adults was chosen to minimize the effects of confounding variables beyond those of interest. It was hypothesized that (1) high-alexithymic individuals would rate the valence and arousal of appetitive, but not aversive or neutral, verbal and nonverbal stimuli more weakly than would their non-alexithymic counterparts; (2) given that alexithymia and anhedonia are related yet distinct constructs in the general population, it was expected that an inverse relationship between alexithymia and reactivity to appetitive cues would be independent of contributions by anhedonia and/or mood disturbance. 2. Materials and methods 2.1. Participants The sample included 96 students (59.4% women) who received undergraduate course credit for participation. Participants ranged in age from 18 to 22 (M = 18.9, SD = 1.4) years, with 62% reporting their ethnicity as White, 21% Asian/Asian-American, 12% Latino, 4% Black, and 1% biracial. Individuals were excluded if they reported non-fluency in English, uncorrected vision, current or past diagnosed psychiatric/neurologic illness, or recent use of prescribed (e.g., methylphenidate) or illicit substances (e.g., cannabis) known to affect cognition or emotion. 2.2. Procedure After providing informed consent, participants completed selfreport measures in a pseudorandom order, including the TAS-20, the Profile of Mood States (McNair, Lorr, & Droppleman, 1992), and the Chapman Revised Social (Mishlove & Chapman, 1985) and Physical (Chapman, Chapman, & Raulin, 1976) Anhedonia Scales, and then evaluated affective stimuli. 2.3. Self-report measures Alexithymia status was determined with the TAS-20, a 20-item scale that assesses difficulty identifying (e.g., ‘‘I don’t know what’s going on inside me’’) and verbalizing feelings (e.g., ‘‘It is difficult for me to find the right words for my feelings’’) as well as externallyoriented thinking (e.g., ‘‘I prefer to analyze problems rather than just describe them’’) using a 5-point Likert scale. Total scores can range from 20 to 100. In keeping with earlier studies (Hesse & Floyd, 2011; Hesse et al., 2013; Joukamaa et al., 2007), participants with scores P60 (i.e., 60th percentile of the possible total score) were considered alexithymic. Full-scale Cronbach’s alpha was .80. The POMS was included in the protocol to assess and control for potential differences in negative affect between alexithymic and non-alexithymic participants. The POMS contains 65 adjectives, each rated on a 5-point Likert-type scale, that comprise six subscales: Tension/Anxiety (e.g., ‘‘nervous’’), Depression/Dejection (e.g., ‘‘gloomy’’), Anger/Hostility (e.g., ‘‘peeved’’), Vigor/Activity
103
(e.g., ‘‘alert’’), Fatigue/Inertia (e.g., ‘‘sluggish’’), and Confusion/ Bewilderment (e.g., ‘‘muddled’’). Total mood disturbance (TMD), calculated by subtracting the Vigor/Activity score from the sum of the other five scores, reflects overall negative affect at the time of testing and was the variable of interest in this study. A TMD zscore was calculated (M = 0, SD = 1) based on sex-related norms (McNair et al., 1992); higher z-scores reflect greater mood disturbance. Cronbach’s alpha for TMD was .85. The Chapman Revised Anhedonia Scales were included in order to track anhedonia levels among alexithymic and nonalexithymic participants. The 40-item Social Anhedonia scale and the 61-item Physical Anhedonia scale assess the ability to experience pleasure from human interaction (e.g., ‘‘Having close friends is not as important as many people say’’) and physical stimuli (e.g., ‘‘The beauty of sunsets is greatly overrated’’), respectively. In each case, items are presented in true/false format; higher scores reflect greater anhedonia. Cronbach’s alpha was 0.78 and 0.82, respectively.
2.4. Affective judgment tasks Individuals were seated 24 inches in front of a 20-inch computer screen for two tasks run on Dell OptiPlex™ GX520 computers. For each task, participants were asked to evaluate the valence of the stimulus on a Likert scale of 1 (extremely negative) to 9 (extremely positive) followed by the intensity of the stimulus on a Likert-type scale of 1 (not intense at all) to 9 (strongest intensity) and to record their answers in a response booklet. Having stimulus presentation on the computer and the recording of responses on paper allowed participants to freely view each stimulus while deciding how to rank its valence and arousal. Tasks were untimed and self-paced. Half of the participants received the word task first, and the other half received the picture task first, enabling testing for order effects. Trials were presented in a pseudorandom order such that no more than two items from the same stimulus category would be contiguous. A total of 150 words were selected from the Affective Norms for English Words (ANEW; Bradley & Lang, 1999): 50 appetitive (e.g., ecstasy), 50 neutral (e.g., fabric), and 50 aversive words (e.g., killer). Words ranged from three to eight letters and were balanced across categories for word length, frequency of usage in the English language, and orthographic neighborhood (Table 1). Words were selected based on norms from the original ANEW validation study in which a mixed gender, nonclinical, adult sample rated words for valence and arousal with a similar 1 (low) to 9 (high) scale. Specifically, appetitive words were selected to be of high valence (close to 9) and high arousal (close to 9), aversive words to be of low valence (close to 1) and high arousal (close to 9), and neutral words to be of moderate valence (close to 5) and low arousal (close to 1). The average valence for the appetitive and aversive word sets was equally polarized (i.e., equidistant from the middle of the scale) and arousing. Words were presented individually in capital letters using Calibri 112-point black font atop a white background. Picture stimuli were drawn from the International Affective Picture System (IAPS; Lang, Bradley, & Cuthbert, 2005) and included 50 appetitive (e.g., laughing children), 50 neutral (e.g., household objects), and 50 aversive images (e.g., mutilated bodies). Pictures were chosen following the same principles as for the word set (Table 1) and guided by adult, nonclinical, mixed-gender norms from the original IAPS validation study. The appetitive and aversive image sets were equally polarized in valence and equally arousing. Appetitive, aversive, and neutral pictures were comparable in valence and arousal to appetitive, aversive, and neutral words, respectively.
104
N.S. Koven / Personality and Individual Differences 68 (2014) 102–106
Table 1 Characteristics of stimuli used in affective judgment tasks. Aversive stimuli
Valence Arousal Frequency Word length Orthographic neighborhood
Neutral stimuli
Appetitive stimuli
Words M ± SD
Pictures M ± SD
Words M ± SD
Pictures M ± SD
Words M ± SD
Pictures M ± SD
2.27 ± 0.46 6.53 ± 0.65 51.14 ± 72.78 5.52 ± 1.45 3.88 ± 5.12
2.19 ± 0.48 6.39 ± 0.65
5.15 ± 0.27 3.77 ± 0.36 50.12 ± 58.70 5.46 ± 1.42 3.74 ± 4.95
5.06 ± 0.29 3.80 ± 0.52
7.87 ± 0.44 6.46 ± 0.53 51.78 ± 66.75 5.60 ± 1.46 3.78 ± 4.32
7.74 ± 0.50 6.38 ± 0.76
Note: For word stimuli, values for valence, arousal, and frequency were derived from adult, mixed-sex norms associated with the ANEW list. Valence and arousal are measured on a 1–9 (low–high) scale. Frequency indicates how often a word is used in the English language, with higher numbers indicating greater frequency. Values for orthographic neighborhood, derived from the English Lexicon Project (Balota et al., 2007), refer to the number of additional words that can be derived from the target word by changing a single letter. For picture stimuli, valence and arousal values were derived from adult, mixed-sex norms associated with the IAPS using the same scaling structure.
2.5. Data analysis Analyses were computed using the Statistical Package for the Social Sciences (SPSS) software version 21.0 (Armonk, NY, USA). Normality was checked in the full sample with Kolmogorov–Smirnov tests and inspection of skew and kurtosis values. Participants with TAS-20 scores P60 (i.e., a threshold equating to 1.25 SD units above the sample mean) formed the alexithymic group (n = 12; 66.7% women). Participants with TAS-20 scores 635 (i.e., a value 1.25 SD units below the mean) formed the non-alexithymic group (n = 13; 69.2% women). Gender composition was comparable across groups, v2(1) with Yates’ correction = 0.09, p = ns. Affective ratings were analyzed with two multivariate analyses of variance (MANOVA) with alexithymia status as the grouping factor. The six valence and arousal variables were included as dependent variables in the first and second analysis, respectively, in order to determine if systematic bias existed in the evaluation of valence separate from any bias in the evaluation of arousal. Separating the valence from the arousal ratings across analyses also reduces multicollinearity, as valence and arousal ratings for the same stimulus category would be strongly correlated. One-way analyses of variance (ANOVA) were used to determine if TMD and anhedonia scores varied across groups. Given significant differences in social anhedonia, hierarchical regression in the full sample was used to analyze the relative contributions of alexithymia and social anhedonia as predictors of affective ratings. This method was chosen over analysis of covariance, as covariates should not be used to adjust for pre-existing differences between groups (Suckling, 2011). Using the results from the aforementioned MANOVAs as a guide for selection of dependent variables, the goal of the regression was to identify whether TAS-20 score contributed unique variance to affective ratings beyond that of the Social Anhedonia score. 3. Results 3.1. Descriptive statistics Table 2 provides descriptive statistics for all variables. Means and standard deviations for the TAS-20, Social Anhedonia, and Physical Anhedonia Scales were consistent with those from other young adult samples (Chmielewski, Fernandes, Yee, & Miller, 1995; Coffey, Berenbaum, & Kerns, 2003; Kwapil, Barrantes-Vidal, & Silvia, 2008). The TAS-20 and Physical Anhedonia scores were normally-distributed. The Social Anhedonia score was positively skewed but successfully normalized with logarithmic transformation. ANOVA showed that valence and arousal ratings for words and pictures did not vary as a function of presentation order, Wilks’ k = .78, F(12, 83) = 1.13, p = ns. Independent samples t-tests indi-
cated that the current sample rated appetitive words and pictures as less positive/less intense, aversive words as less intense, and aversive pictures as more negative/more intense relative to the original ANEW and IAPS normative cohorts; ratings of neutral stimuli differed across the cohorts as well (Table 2). Valence ratings for aversive words and pictures and arousal ratings for neutral words were positively skewed but subsequently corrected with logarithmic transformation. Valence ratings for neutral words were negatively skewed but corrected with a logarithmic transformation after subtracting a constant from each score to achieve a minimum score of one. All other ratings were normally-distributed. Transformed variables were used in lieu of untransformed variables for the remaining analyses. 3.2. Affective judgment as a function of alexithymia As seen in Table 2, ANOVA showed no group differences in TMD score, suggesting comparable degrees of negative affect at the time of testing. The alexithymic group had elevated Social but not Physical Anhedonia scores relative to the non-alexithymic group. The full model for the MANOVA with the six valence ratings as dependent variables was significant, Wilks’ k = .25, F(6, 18) = 8.18, p < .001, partial N2 = .75. Tests of between-subjects effects showed that the alexithymic group rated appetitive words, F(1, 23) = 12.08, p = .002, partial N2 = .37, and appetitive pictures, F(1, 23) = 5.65, p = .027, partial N2 = .21, less strongly than did the non-alexithymic group. The alexithymic group also rated neutral pictures as lower in valence than the non-alexithymic group, F(1, 23) = 6.69, p = .017, partial N2 = .24. Of these three effects, only the first one remained significant after Bonferroni correction (adjusted a = .008). There were no group differences in ratings of aversive pictures, aversive words, or neutral words. The full model for the second MANOVA indicated no group differences in arousal ratings, Wilks’ k = .76, F(6, 18) = .56, p = ns. The assumption of similar covariance matrix across groups was upheld for both the first, Box’s M = 44.89, F(23, 1592) = 1.46, p = ns, and second MANOVA, Box’s M = 40.80, F(23, 1592) = 1.33, p = ns. The assumption of similar error variances across groups for each dependent variable was upheld in both analyses (all p-values in Levene’s tests = ns). Given that the two groups differed in social anhedonia, a hierarchical regression was used to assess the contribution of alexithymia to appetitive word valence ratings in the full sample after controlling for social anhedonia. When added last to the regression mode, the TAS-20 score contributed significant incremental variance, DR2 = .054, DF(1, 93) = 4.23, p = .04. In the final model, the TAS-20 score, adjusted for its shared variance with the Social Anhedonia score, remained a significant negative predictor, beta = .26, p = .04, whereas the Social Anhedonia score, controlling for its shared variance with the TAS-20 score, was insignificant, beta = .05, p = ns.
105
N.S. Koven / Personality and Individual Differences 68 (2014) 102–106 Table 2 Descriptive statistics in the full sample and across groups. Variable
Full sample (n = 96) M ± SD
Alexithymic (n = 12) M ± SD
Non-alexithymic (n = 13) M ± SD
F
t
TAS-20 total Social Anhedonia Physical Anhedonia POMS TMD Appetitive word valence Appetitive word arousal Aversive word valence Aversive word arousal Neutral word valence Neutral word arousal Appetitive picture valence Appetitive picture arousal Aversive picture valence Aversive picture arousal Neutral picture valence Neutral picture arousal
47.1 ± 10.1 7.2 ± 5.9 11.2 ± 6.9 0.2 ± 0.8 6.9 ± 0.8 5.8 ± 1.2 2.3 ± 0.9 5.7 ± 1.6 5.2 ± 0.3 2.7 ± 1.6 6.8 ± 0.6 5.4 ± 1.3 1.8 ± 0.7 6.6 ± 1.7 5.3 ± 0.3 2.8 ± 1.3
64.1 ± 6.2 8.8 ± 4.8 13.0 ± 6.6 0.2 ± 0.7 6.3 ± 0.7 5.2 ± 1.5 2.8 ± 1.1 4.9 ± 1.4 5.2 ± 0.3 2.8 ± 2.0 6.4 ± 0.6 5.0 ± 1.5 1.8 ± 0.4 5.7 ± 2.0 5.2 ± 0.2 2.5 ± 1.4
33.7 ± 2.5 5.1 ± 3.7 10.7 ± 7.0 0 ± 0.7 7.2 ± 0.5 5.6 ± 1.1 2.2 ± 0.5 5.6 ± 1.4 5.4 ± 0.2 2.2 ± 1.3 6.9 ± 0.4 5.0 ± 1.1 1.8 ± 0.5 6.6 ± 1.4 5.5 ± 0.2 2.4 ± 1.1
267.42** 5.56* 1.22 1.11 12.08** 0.34 3.33 0.69 3.61 0.18 5.65* .03 .02 2.05 6.69* .20
18.83** 10.00** 0.55 9.96** 1.72 17.16** 17.27** 11.17** 7.26** 2.46* 7.72** 14.92**
Note: TAS-20 = Toronto Alexithymia Scale; POMS TMD = Profile of Mood States Total Mood Disturbance. Means and standard deviations reported here are from untransformed variables. The F statistic tested for group differences using transformed variables when appropriate. F-values for psychological variables were generated with ANOVA. Group differences in stimulus ratings were assessed with MANOVA, with the F-values drawn from tests of between-subjects effects. T-values are derived from independent samples t-tests, comparing stimulus ratings from the full sample to normed ratings from the ANEW and IAPS sets. * p < .05. ** p < .005.
4. Discussion This study confirmed that alexithymic, relative to non-alexithymic, individuals under-report the valence of appetitive stimuli. Although only the group difference for appetitive words survived Bonferroni correction, there was a similar but weaker valence difference for appetitive pictures, suggesting generalization across verbal and nonverbal content. However, the alexithymic group was more socially anhedonic than the non-alexithymic group, making it difficult to distinguish whether the above effects were secondary to alexithymia or social anhedonia. Hierarchical regression using the full sample determined that alexithymia contributed unique variance to appetitive valence ratings after controlling for social anhedonia. Taken together, it appears that alexithymia is characterized by diminished responsivity to appetitive stimuli, independent of anhedonia. One cannot conclude from these data, however, that alexithymic individuals have limited hedonic capacity, as the alexithymic group reported a level of arousal in response to appetitive stimuli comparable to that by the non-alexithymic group. The fact that the point of difference between groups was valence but not arousal suggests, instead, that the deficiency is one of affective differentiation. In other words, alexithymic individuals are sufficiently aroused by appetitive cues but struggle with the ability to ‘‘gauge the temperature’’ of this arousal to assign it a higher valence. That valence and arousal are separable is certainly not alien to personality theory or the psychopathology literature (Heller, Schmidtke, Nitschke, Koven, & Miller, 2002) and, indeed, their orthogonality is at the core of the valence-arousal model of emotion (Russell, 1980). Neuroimaging research also supports this distinction, with evidence suggesting that the evaluation of valence and arousal recruits different neural circuits in nonclinical samples (Gerber et al., 2008). Moreover, there is evidence (Nielen et al., 2009) of differential processing for appetitive (middle temporal and orbitofrontal cortex) and aversive (visual and lateral prefrontal regions) stimuli in healthy young adults as well as distinct brain regions for the conjoint processing of highly-arousing appetitive stimuli (occipital cortex, parahippocampal gyrus, and posterior cingulate) versus highly-arousing aversive stimuli (anterior insula).
It is difficult to ascertain the extent to which these findings apply to alexithymia, as most neuroimaging studies with alexithymic subjects overlook appetitive stimuli altogether and/or do not present stimuli in such a way as to separate contributions by valence and arousal to the resulting data. Of those with appropriate designs, Hesse et al. (2013) reported weaker activation in alexithymic participants in response to appetitive IAPS stimuli in several brain regions known for socioemotional perception. However, these authors did not assess anhedonia and did not include aversive stimuli, making it difficult to know if this pattern of hypoactivation is specific to the alexithymia construct and/or to appetitive stimuli. In a similar study, Heinzel et al. (2010) identified dorsal anterior cingulate activation in non-depressed alexithymic individuals in response to positive-valenced IAPS stimuli, which is noteworthy given the role of this brain region in purposeful inhibition of arousal (Beauregard, Lévesque, & Bourgouin, 2001). In application to the current data, this result raises the interesting possibility that the lower appetitive valence ratings by alexithymic individuals resulted from active suppression of arousal generated by those same stimuli, a premise consistent with the description of the ‘‘Type II’’ alexithymia subtype (Bagby et al., 2009); why this should be specific to appetitive stimuli, however, is an open question. As the present study was powered to find large effect sizes only, it is possible that alexithymic individuals exhibit mildly deficient responsivity to aversive stimuli, but, at the same time, the lack of a between-groups effect for aversive stimuli is consistent with other reports of normal threat processing in alexithymia (Lundh & Simonsson-Sarnecki, 2002; Mueller et al., 2006; Pandey, 1995; Parker et al., 1993). It is noteworthy that the present sample as a whole differed from the original ANEW and IAPS normative cohorts in their evaluation of most of the stimulus categories. While the ratings were such that the stimuli were judged to belong to the originallyintended category (e.g., aversive stimuli were still deemed aversive), the statistical differences may suggest a generational shift in the affective judgments of present-day young adults. Although the ANEW and IAPS norms were generated as recently as 1999 and 2005, respectively, the current results might signal a need for updated norms for these popular stimulus sets.
106
N.S. Koven / Personality and Individual Differences 68 (2014) 102–106
Although this study adds useful detail to the concept of diminished hedonic capacity in alexithymia, several limitations of the study must be acknowledged. First, even though stimuli were carefully chosen based on established norms, low arousal stimuli were not included and, therefore, it is unknown whether subjective ratings of appetitive valence by alexithymic individuals change with decreased intensity and if any change would differ from patterns found in non-alexithymic individuals. Second, although selection of a healthy sample was purposeful, the restricted age range limits the generalizability of these results to older adult cohorts. Moreover, while this study suggests that anhedonia is non-contributory to appetitive valence ratings in alexithymia, the reliance on a cross-sectional design with self-report methods curtails any conclusions concerning the underlying mechanisms of or the etiological reasons for abnormal valence differentiation. Future studies that combine neuroimaging, assessment of possible confounds, and balanced emotion stimuli in larger, community-based samples will be well positioned to characterize the physiological and cognitive interface of this affective style as well as to assess the generalizability of effects across potential alexithymia subtypes. References Bagby, R. M., Parker, J. D. A., & Taylor, G. L. (1994). The twenty-item Toronto Alexithymia Scale - 1: item selection and cross-validation of the factor structure. Journal of Psychosomatic Research, 38, 23–32. http://dx.doi.org/ 10.1016/0022-3999(94)90005-1. Bagby, R., Quilty, L. C., Taylor, G. J., Grabe, H. J., Luminet, O., Verissimo, R., et al. (2009). Are there subtypes of alexithymia? Personality and Individual Differences, 47(5), 413–418. http://dx.doi.org/10.1016/j.paid.2009.04.012. Balota, D. A., Yap, M. J., Cortese, M. J., Hutchison, K. A., Kessler, B., Loftis, B., et al. (2007). The English lexicon project. Behavior Research Methods, 39(3), 445–459. http://dx.doi.org/10.3758/BF03193014. Beauregard, M., Lévesque, J., & Bourgouin, P. (2001). Neural correlates of conscious self-regulation of emotion. The Journal of Neuroscience, 21(18), 6993–7000. Bradley, M. M., & Lang, P. J. (1999). Affective norms for English words (ANEW): Stimuli, instruction manual, and affective ratings. Gainesville, FL, USA: The Center for Research in Psychophysiology, University of Florida. Chapman, L. J., Chapman, J. P., & Raulin, M. L. (1976). Scales for physical and social anhedonia. Journal of Abnormal Psychology, 85(4), 374–382. http://dx.doi.org/ 10.1037/0021-843X.85.4.374. Chmielewski, P. M., Fernandes, L. O. L., Yee, C. M., & Miller, G. A. (1995). Ethnicity and gender in scales of psychosis proneness and mood disorders. Journal of Abnormal Psychology, 104(3), 464–470. http://dx.doi.org/10.1037/0021843X.104.3.464. Coffey, E., Berenbaum, H., & Kerns, J. G. (2003). The dimensions of emotional intelligence, alexithymia, and mood awareness: Associations with personality and performance on an emotional Stroop task. Cognition and Emotion, 17(4), 671–679. http://dx.doi.org/10.1080/02699930302304. Deborde, A. S., Berthoz, S., Godart, N., Perdereau, F., Corcos, M., & Jeammet, P. (2006). Étude des relations entre alexithymie et anhédonie chez des femmes présentant des troubles du comportement alimentaire et chez des témoins. L’Encéphale: Revue De Psychiatrie Clinique Biologique Et Thérapeutique, 32(1), 83–91. doi:10.1016/S0013-7006(06)76140-1. Dubey, A., & Pandey, R. (2013). Mental health problems in alexithymia: Role of positive and negative emotional experiences. Journal of Projective Psychology and Mental Health, 20(2), 128–136. Epp, A. M., Dobson, K. S., Dozois, D. J. A., & Frewen, P. A. (2012). A systematic metaanalysis of the Stroop task in depression. Clinical Psychology Review, 32(4), 316–328. http://dx.doi.org/10.1016/j.cpr.2012.02.005. Gerber, A. J., Posner, J., Gorman, D., Colibazzi, T., Yu, S., Wang, Z., et al. (2008). An affective circumplex model of neural systems subserving valence, arousal, and cognitive overlay during the appraisal of emotional faces. Neuropsychologia, 46(8), 2129–2139. http://dx.doi.org/10.1016/j.neuropsychologia.2008.02.032. Heinzel, A., Schäfer, R., Müller, H., Schieffer, A., Ingenhag, A., Northoff, G., et al. (2010). Differential modulation of valence and arousal in high-alexithymic and low-alexithymic individuals. NeuroReport, 21(15), 998–1002.
Heller, W., Schmidtke, J. I., Nitschke, J. B., Koven, N. S., & Miller, G. A. (2002). States, traits, and symptoms: Investigating the neural correlates of emotion, personality, and psychopathology. In D. Cervone, & W. Mischel (Eds.), (pp. 106–126). New York, NY, USA: Guilford Press. Hesse, C., & Floyd, K. (2011). The impact of alexithymia on initial interactions. Personal Relationships, 18(3), 453–470. http://dx.doi.org/10.1111/j.14756811.2010.01311.x. Hesse, C., Floyd, K., Rauscher, E. A., Frye-Cox, N., Hegarty, J. P., & Peng, H. (2013). Alexithymia and impairment of decoding positive affect: An fMRI study. Journal of Communication, 63(4), 786–806. http://dx.doi.org/10.1111/jcom.12039. Hintikka, J., Honkalampi, K., Lehtonen, J., & Viinamäki, H. (2001). Are alexithymia and depression distinct or overlapping constructs? A study in a general population. Comprehensive Psychiatry, 42(3), 234–239. http://dx.doi.org/ 10.1053/comp.2001.23147. Honkalampi, K., Hintikka, J., Tanskanen, A., Lehtonen, J., & Viinamäki, H. (2000). Depression is strongly associated with alexithymia in the general population. Journal of Psychosomatic Research, 48(1), 99–104. http://dx.doi.org/10.1016/ S0022-3999(99)00083-5. Joukamaa, M., Taanila, A., Miettunen, J., Karvonen, J. T., Koskinen, M., & Veijola, J. (2007). Epidemiology of alexithymia among adolescents. Journal of Psychosomatic Research, 63(4), 373–376. http://dx.doi.org/10.1016/ j.jpsychores.2007.01.018. Kwapil, T. R., Barrantes-Vidal, N., & Silvia, P. J. (2008). The dimensional structure of the Wisconsin Schizotypy Scales: Factor identification and construct validity. Schizophrenia Bulletin, 34(3), 444–457. http://dx.doi.org/10.1093/schbul/ sbm098. Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2005). International affective picture system (IAPS): Affective ratings of pictures and instruction manual. Gainesville, FL, USA: University of Florida. Lee, S. A., & Guajardo, N. R. (2011). Affect intensity and alexithymia differentially influence the relationship between neuroticism and depressive symptomatology among college students. Personality and Individual Differences, 50(5), 646–650. http://dx.doi.org/10.1016/j.paid.2010.11.036. Loas, G., Fremaux, D., & Boyer, P. (1997). Anhedonia and alexithymia: Distinct or overlapping constructs. Perceptual and Motor Skills, 84(2), 415–425. http:// dx.doi.org/10.2466/pms.1997.84.2.415. Lundh, L., & Simonsson-Sarnecki, M. (2002). Alexithymia and cognitive bias for emotional information. Personality and Individual Differences, 32(6), 1063–1075. http://dx.doi.org/10.1016/S0191-8869(01)00110-6. McNair, D. M., Lorr, M., & Droppleman, L. F. (1992). Profile of mood states manual. San Diego, CA, USA: Educational and Industrial Testing Service. Mishlove, M., & Chapman, L. J. (1985). Social anhedonia in the prediction of psychosis proneness. Journal of Abnormal Psychology, 94(3), 384–396. http:// dx.doi.org/10.1037/0021-843X.94.3.384. Mueller, J., Alpers, G. W., & Reim, N. (2006). Dissociation of rated emotional valence and Stroop interference in observer-rated alexithymia. Journal of Psychosomatic Research, 61(2), 261–269. http://dx.doi.org/10.1016/j.jpsychores.2006.02.017. Nielen, M. M., Heslenfeld, D. J., Heinen, K., Van Strien, J. W., Witter, M. P., Jonker, C., et al. (2009). Distinct brain systems underlie the processing of valence and arousal of affective pictures. Brain and Cognition, 71(3), 387–396. Ogrodniczuk, J. S., Sochting, I., Piper, W. E., & Joyce, A. S. (2012). A naturalistic study of alexithymia among psychiatric outpatients treated in an integrated group therapy program. Psychology and Psychotherapy: Theory, Research, and Practice, 85(3), 278–291. http://dx.doi.org/10.1111/j.2044-8341.2011.02032.x. Pandey, R. (1995). Stroop interference effect of emotion-arousing words in alexithymia. Journal of the Indian Academy of Applied Psychology, 21(1), 21–28. Parker, J. D. A., Taylor, G., & Bagby, M. (1993). Alexithymia and the processing of emotional stimuli: An experimental study. New Trends in Experimental and Clinical Psychiatry, 9(1–2), 9–14. Pizzagalli, D. A., Iosifescu, D., Hallett, L. A., Ratner, K. G., & Fava, M. (2008). Reduced hedonic capacity in major depressive disorder: Evidence from a probabilistic reward task. Journal of Psychiatric Research, 43(1), 76–87. http://dx.doi.org/ 10.1016/j.jpsychires.2008.03.001. Prince, J. D., & Berenbaum, H. (1993). Alexithymia and hedonic capacity. Journal of Research in Personality, 27(1), 15–22. http://dx.doi.org/10.1006/jrpe.1993.1002. Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39(6), 1161–1178. Sloan, D. M., Strauss, M. E., & Wisner, K. L. (2001). Diminished response to pleasant stimuli by depressed women. Journal of Abnormal Psychology, 110(3), 488–493. http://dx.doi.org/10.1037/0021-843X.110.3.488. Suckling, J. (2011). Correlated covariates in ANCOVA cannot adjust for pre-existing differences between groups. Schizophrenia Research, 126(1–3), 310–311. http:// dx.doi.org/10.1016/j.schres.2010.08.034.