Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Feb 23.
Published in final edited form as: J Am Acad Child Adolesc Psychiatry. 2018 Sep 29;57(12):925–933.e3. doi: 10.1016/j.jaac.2018.05.028

A Developmental Twin Study of Emotion Recognition and Its Negative Affective Clinical Correlates

Lance M Rappaport 1, Dever M Carney 1, Brad Verhulst 1, Michael C Neale 1, James Blair 2, Melissa A Brotman 2, Daniel S Pine 2, Ellen Leibenluft 2, John M Hettema 3,*, Roxann Roberson-Nay 3,*
PMCID: PMC7036262  NIHMSID: NIHMS1558654  PMID: 30522738

Abstract

Objective.

Youth with psychiatric disorders distinguished by irritability, including depression and associated trait neuroticism, show deficits in the ability to recognize facial expressions of emotion, particularly happiness. However, the contribution of genetic and environmental factors to this ability remains unknown. The current study examined this trait in twins to assess the genetic and environmental influences on face-emotion recognition abilities and their association with irritability, neuroticism, and depression.

Method.

Child and adolescent twins (N = 957 from 496 families), aged 9 to 17 rated their irritability (on the Affective Reactivity Index), neuroticism (on the Junior Eysenck Personality Questionnaire), and depression (on the Short Mood and Feelings Questionnaire) and completed a face-emotion labeling task. Faces depicting anger, disgust, fear, happiness, sadness, and surprise were morphed with a neutral face yielding 10 levels of increasing emotional expressivity. Biometrical twin analyses evaluated contributions of genetic and environmental factors to the etiology of face-emotion recognition and its association with irritability, neuroticism, and depression.

Results.

Recognition of each emotion was heritable; common and specific sets of genetic factors influenced all emotions and individual emotions, respectively. Irritability, neuroticism, and depression are modestly, negatively correlated with emotion recognition, particularly the recognition of happiness. For irritability and neuroticism, this correlation appears largely due to genetic factors.

Conclusions.

This study maps genetic and environmental contributions to face-emotion recognition and its association with irritability, neuroticism, and depression. Findings implicate common genetic factors in deficits regarding the recognition of happiness associated with irritability and neuroticism in childhood and adolescence.

Keywords: irritability, face-emotion recognition, child, adolescent, twin, genetics

Introduction

Recognizing facial emotions is a fundamental human capability1. Emotions can be easily recognized by people from many different cultures. This may reflect the central role of face-emotion expression in social communication2,3 and the fact that face-emotion recognition is mediated by a core, evolutionarily conserved neural circuit4. Psychiatric problems often involve social communication deficits. For example, major depressive disorder5, bipolar disorder6,7, autism spectrum disorders8, alcohol dependence9, attention-deficit/hyperactivity disorder10, and disruptive mood dysregulation disorder (DMDD), in which irritability11 and elevated neuroticism are common.

Irritable mood, a common complaint amongst youth, is transdiagnostic11,12 and a core feature of several emotional disorders, including disruptive mood dysregulation disorder (DMDD)13,14. Youth with chronic, severe irritability (similar to the DMDD phenotype) exhibit deficits in face-emotion recognition12,15 and neural perturbations when processing emotional faces16,17, which suggests one aspect of diminished socioemotional functioning. Additionally, a recent meta-analysis implicates deficits in face-emotion recognition, specifically impaired recognition of happiness, in adult major depressive disorder5. This is consistent with existing cognitive theory18 and evidence of impaired recognition of happiness among adults who report high trait neuroticism19. As an initial investigation, the present study examined the role of common genetic contributions to explain theory and phenotypic evidence of face-emotion recognition deficits in individuals with elevated irritability, depression, and neuroticism severity.

However, potential pathways linking the etiologic underpinnings of the ability to recognize emotional expression in faces and childhood liability to irritability, neuroticism, and depression have not been investigated. Identifying the degree of overlap in shared genetic and environmental contributions to irritability, neuroticism, and depression and face-emotion recognition is necessary to locate diminished socioemotional functioning as a possible process in the development of irritability and internalizing disorders (e.g., depressive and anxiety disorders). To our knowledge, only one prior study has parsed genetic and environmental contributions to face-emotion recognition abilities20, and none consider the degree to which these contributions make shared or independent contributions to psychopathology. The current study addresses the need for such data.

Specifically, the primary aims of the current study are to estimate, in childhood and adolescence: i) the phenotypic correlation (i.e., within-subject association) of irritability, neuroticism, and depression with face-emotion recognition; ii) the contribution of genetic and environmental influences to face-emotion recognition; iii) the structure of these genetic and environmental influences across emotions; and iv) the degree to which the genetic and environmental contributions to face-emotion recognition are correlated with those for irritability, neuroticism, and depression. To achieve these aims, a large, genetically-informed epidemiological sample of child and adolescent twins, with known genetic concordance based on zygosity, self-reported on irritability, neuroticism, and depression and completed a face-emotion labeling task (FELT). Three standard multivariate twin models21, similar to those examined previously20, were compared to optimally model genetic and environmental factors in the etiology of face-emotion recognition. Irritability, neuroticism, and depression were then separately added to the best fitting model to examine the genetic and environmental contributions to their covariance with face-emotion recognition.

Methods and Materials

Participants

Child and adolescent participants (N = 1,279, 54.59 % female) were recruited as pairs of twins aged 9 to 20 from the mid-Atlantic area of the United States of America by the Mid Atlantic Twin Registry22. Detailed demographic information is provided in two published overviews of the child23 and adolescent samples24. The child sample exclusively recruited Caucasian twins; the adolescent sample primarily recruited Caucasian twins. This was done to enhance power for planned subsequent genomic research by reducing genetic heterogeneity due to population stratification25. All twins reported on irritability, neuroticism, and depression and completed the face-emotion labeling task (see below). To minimize the impact of potential confounding factors, participants were excluded from recruitment if they were intellectually disabled, had a prior diagnosis of an autism spectrum disorder, experienced a past or current psychotic episode, were currently using anxiolytic or antidepressant medication, or had been diagnosed with any medical condition that might have adversely impacted participants’ safety or ability to complete the study (e.g., unexplained seizures) including aspects of the study not described here23,24, such as hypersensitivity to a carbon dioxide challenge task26,27. To focus on face-emotion recognition during childhood and adolescence, the present study analyzed data on participants under 18 years of age (N = 957).

Rates of psychiatric diagnoses were assessed for all participants based on DSM-IV. Children were assessed by structured K-SADS assessment via parent interview28 by a study psychiatrist or clinical psychologist. Adolescents were assessed by participant-report of psychiatric symptoms on a modified version of the Composite International Diagnostic Interview – Short Form26,29. Diagnostic rates, based on DSM-IV, indicate that the sample is generally representative of children and adolescents in this age range. Lifetime prevalences for any depressive (2.2%) or anxiety disorder (27.4%) among children are consistent with rates reported elsewhere30,31; lifetime prevalences for major depressive disorder (14.8%), panic disorder (1.2%), and generalized anxiety disorder (2.6%) among adolescents are consistent with rates reported elsewhere32. Participants were recruited at two sites: Virginia Commonwealth University (VCU; Richmond, VA) and the National Institute of Mental Health (NIMH; Bethesda, MD), which included only child participants.

Procedure

Participants completed the FELT in the context of two larger studies on the heritability of potential endophenotypes of internalizing disorders in children or adolescents and young adults. A parent or legal guardian provided informed consent; youth provided assent before completing self-report questionnaires and laboratory-based tasks23,24.

To estimate test-retest reliability of the face-emotion recognition task, child participants attended a second session two weeks after the first. A planned missingness design minimized participant burden; 157 randomly-chosen participants completed the FELT task at follow-up. The protocol was approved by the Institutional Review Boards of VCU and the NIMH.

Materials

Twin Zygosity.

A parent or guardian answered questions regarding physical similarities of the twins (e.g., “Are your twin children often confused for one another?”), such as those described by Jackson and colleagues33. Prior research demonstrated that these questions assess twin zygosity with high agreement to blood34 and DNA assessments33 of zygosity including high agreement (κ = 0.95) with zygosity estimated from single nucleotide polymorphisms (SNPs) in the present sample24.

Clinical Correlates.

Participants reported irritability over the past six months on the Affective Reactivity Index (ARI), which assesses behavioral and emotional components of irritability and anger35. Participants completed neuroticism items from the Junior Eysenck Personality Questionnaire36 and the Short Mood and Feelings Questionnaire37 to assess neuroticism and recent depression symptoms, respectively. Prior research demonstrated reliability and validity of these measures in child and adolescent samples35–37 including the present study23,24.

Face-Emotion Labeling Task.

Participants completed a FELT in which they were presented with six fundamental emotions (i.e., happiness, anger, sadness, fear, disgust, and surprise) expressed by Caucasian adults (50% female) from the well-validated Pictures of Facial Affect Series38. The task was based on prior research on the clinical and neurobiological correlates of face-emotion processing39–41.

A photo portraying each emotion was morphed with a neutral face of the same person to produce 10 target images along a gradient from 10% to 100% emotional expression to yield 60 images (6 emotions × 10 morph steps). Each image was presented at 6 trials for a total of 360 trials. Trial order was randomized but consistent across participants. At each presentation, participants viewed a fixation cross for 250 ms before viewing the target image for 500 ms. Participants then pressed a button to endorse which emotion they saw from the six possible emotions. Task performance demonstrates high test-retest reliability among adult42 and child samples43. The task took approximately fifteen minutes to complete.

Data Analysis

Raw accuracy may reflect a participant’s tendency to endorse a specific emotion. To adjust for this association, participant scores were transformed as in Marsh et al.40 and Wagner44. A participant’s score was computed as the product of raw accuracy and differential accuracy (i.e., the proportion of trials in which an emotion is correctly selected). The result was adjusted for guessing by subtracting 1/6 and arcsine transformed to improve the normality of the distribution of scores. Similar to prior research using this task, each participant’s performance for each emotion was computed as the average score over all trials. Data from 94 individuals (9.82% of the sample) were removed due to suspected poor compliance with the task, (e.g. pressing the same button throughout the task without looking at the screen), or suspected difficulty attending to the task (e.g., self-reported headache). Research assistants who noted poor compliance were blind to other study data. In total, 189 monozygotic twin pairs, 271 dizygotic twin pairs, and 36 twin pairs of indeterminate zygosity completed the face-emotion labeling task.

Our main analyses proceeded in several steps. First, we verified the test-retest reliability of average performance on the FELT. Second, we examined the phenotypic correlations of face-emotion recognition with irritability, neuroticism, and depression. Third, we investigated the structure of genetic and environmental contributions to the recognition of all six emotions. Finally, the role of face-emotion recognition as an affective endophenotype was examined by estimating the degree of genetic and environmental correlations of face-emotion recognition with irritability, depression, and neuroticism. As participants are nested within families, test-retest reliability, phenotypic correlations, heritability, and genotypic correlations were estimated based on the biometrical model, which incorporates known nonindependence due to shared genetic and environmental influences that contribute to the correlation between mono- and dizygotic twins21.

Three multivariate models were fitted to face-emotion recognition data: a common pathway model (CPM); an independent pathway model (IPM); and a “saturated” model, based on a Cholesky decomposition of the variance-covariance matrix. The latter allows for six sets of three factors that account for different structures of additive genetic (A), shared (between twins) environmental (C), and nonshared environmental (E) factors to influence the recognition of each emotion. The CPM and IPM are nested within the Cholesky decomposition to evaluate whether separate A, C, and E factors directly influence the measures (i.e., the independent pathways model), or covariance in the etiology of performance between emotions is best summarized by a single, latent common factor (i.e., CPM). Performance on the face emotion recognition task was standardized prior to analyses. Analyses of face-emotion recognition were repeated adjusting for the impact of age and sex on mean performance; results regarding the optimal model, structure of the model (see Figure 1), and estimated heritability were unchanged. An additional analysis examined potential differences in the contribution of genetic and environmental factors by participant sex. While female participants demonstrated higher face-emotion recognition across all six emotions (see Supplemental Table 1), multivariate sex limitation models45 demonstrated equivalent genetic and environmental contributions to face-emotion recognition across sex (see Supplemental Table 2).

Figure 1.

Figure 1.

Open in a new tab

Best Fitting Model for Genetic and Environmental Contributions to Face-emotion Recognition

Note. A = additive genetic factor; C = shared environmental factor; E = nonshared-environmental factor; Factor numbers correspond to Supplemental Table 2 under “Cholesky Factor Loading Estimates,” specifically factors 1–6; factors with estimated loadings below 0.3 are omitted for clarity. Loadings are standardized. Variances for latent variables, not depicted, are constrained to 1.

Once a suitable, best-fitting model for the etiology of face-emotion recognition was identified, irritability, neuroticism, and depression were separately added to examine shared etiology with face-emotion recognition. Preliminary linear mixed effects models, which adjust for nesting of twins within family, estimated the influence of age and sex on irritability, neuroticism, depression, and face-emotion recognition (see Supplemental Table 1). Older age was associated with reduced irritability, neuroticism, and depression but improved face-emotion recognition across all emotions. The present research may clarify mechanisms underlying age-related changes in irritability, neuroticism, depression, and face-emotion recognition. For example, age-related changes in irritability and face-emotion recognition may co-occur due to common developmental pathways and due to shared genetic risk. Age and sex also accounted for a substantial proportion of variance in face-emotion recognition (conditional R2 range: 0.33 – 0.43; see Supplemental Table 1). Therefore, adjusting for age and sex may statistically underestimate the correlation of face-emotion recognition with irritability, neuroticism, and depression.

Parameters from each structural equation model were iteratively fixed to zero to evaluate the statistical significance of additive genetic (A) and shared environmental (C) factors to face-emotion recognition, irritability, neuroticism, and depression and the covariance between them. Profile-based confidence intervals were computed using maximum likelihood estimation within OpenMx. Analyses were conducted in R version 3.2.346 using the following packages: psych47, car48, plyr49, ggplot250, nlme51, umx52, and OpenMx53.

Results

Descriptive Statistics and Face-Emotion Recognition Deficits

Table 1 lists descriptive data on the FELT outcomes. Test-retest reliability for performance on the FELT is high for each emotion and similar across emotions as evidenced by overlapping confidence intervals for reliability estimates. Performance in recognizing each emotion is correlated across emotions. However, as also shown in Table 1, mean performance for each emotion differs significantly, reflecting greater accuracy to identify certain emotions (e.g. happiness) over others (e.g. disgust).

Table 1.

Test-Retest Reliability, Descriptive Statistics, and Phenotypic Correlations Among Recognition of Each Emotion

Emotion Mean (SD) Test-Retest
Reliability:
r (95% CI)		 Correlation Matrix
Anger Happiness Sadness Fear Surprise Disgust
Anger 0.03 (0.07)a,b 0.69 (0.52, 0.80) 1
Happiness 0.38 (0.17)a,b 0.69 (0.51, 0.80) 0.55 1
Sadness 0.08 (0.11)a,b 0.76 (0.63, 0.83) 0.51 0.62 1
Fear 0.06 (0.13)a,b 0.64 (0.43, 0.77) 0.52 0.59 0.61 1
Surprise 0.19 (0.14)a,b 0.67 (0.50, 0.79) 0.50 0.58 0.53 0.64 1
Disgust −0.04 (0.09)a,b 0.75 (0.61, 0.85) 0.50 0.37 0.38 0.33 0.31 1
Irritability 2.96 (2.66)b -- 0.01
(−0.06, 0.08)		 −0.11 **
(−0.18, −0.04) 		−0.11 **
(−0.17, −0.04) 		−0.12 **
(−0.18, −0.05) 		−0.05
(−0.12, 0.02)		 −0.03
(−0.1, 0.04)
Neuroticism 7.04 (4.57)b -- −0.05
(−0.13, 0.02)		 −0.16 ****
(−0.23, −0.09) 		−0.14 ****
(−0.21, −0.07) 		−0.14 ***
(−0.21, −0.06) 		−0.08 *
(−0.15, −0.01) 		−0.09 **
(−0.17, −0.02)
Depression 5.28 (4.24)b -- 0.05
(−0.02, 0.12)		 −0.08 *
(−0.15, −0.01) 		−0.07
(−0.14, 0)		 −0.02
(−0.1, 0.05)		 −0.02
(−0.09, 0.05)		 0
(−0.07, 0.07)
Open in a new tab

Note.

*

p ≤ 0.05,

**

p ≤ 0.01,

***

p ≤ 0.001,

****

p ≤ 0.0001;

a

All means are significantly different from one another; an overall ANOVA and equivalence of each pair of means were tested within the biometrical model. Significance was evaluated using the chi-square difference test, which evaluates worsened model fit due to constraining the mean of all emotions (the overall ANOVA; p = 3.34e−255) or each pair of two emotions to be equal (p ≤ 2.16e−8).

b

Mean face-emotion recognition differs as a function of participant age and sex; female participants generally demonstrated better recognition as well as higher neuroticism and depression, but similar irritability severity (see Supplemental Table 1).

The correlation of irritability, neuroticism, and depression with each emotion was estimated within the biometrical model to account for nonindependence between twins (see Data Analysis, above). Elevated depression, neuroticism, and irritability are associated with reduced recognition of facial expression of happiness (see Table 1), which appears to be linear upon investigation of bivariate scatterplots (see Supplemental Figure 1). Additionally, participants who reported elevated irritability and neuroticism demonstrate reduced recognition of sadness and fear. Elevated neuroticism is, more broadly, also associated with reduced recognition of surprise and disgust.

As noted in previous literature20, a common factor may usefully describe overall face-emotion recognition ability. The phenotypic correlation of irritability, neuroticism, and depression with overall face-emotion recognition ability was estimated within a common factor reflecting all six emotions, which is identical to the CPM (see Supplemental Table 3, Common Factor Model, for factor loadings). Irritability, r = −0.10, p = 0.007, and neuroticism, r = −0.13, p = 0.0007, though not depression, r = −0.05, p = 0.23, were modestly, negatively correlated with overall face-emotion recognition.

Etiology of Face-emotion Recognition

Evaluation of the three multivariate twin models for each of the six emotions differentiated potential etiological contributors to emotion-recognition ability. The comparison of model fits for the Cholesky vs. the CPM and IPM indicated that the Cholesky model fit the data better than either of the other two models (Table 2). This suggests that several additive genetic and environmental factors explain etiology more comprehensively than does a single common factor. Specifically, the parameter estimates for the three models demonstrated a high level of consistency between the factor loadings for the CPM, the additive genetic factor loadings for the IPM, and the first latent factor for the Cholesky decomposition (see Figure 1, Supplemental Table 3). The CPM and IPM, however, did not account for additional genetic components for emotion recognition, particularly for happiness and disgust, nor the residual genetic correlation between fear and surprise. Moreover, the factor loading results for both the CPM and the IPM diverged substantially from the results for the best fitting model regarding shared and unique environmental influences.

Table 2.

Nested Model Fit for Multivariate Models

model np −2LL df AIC ΔLL Δdf p
“Saturated” Cholesky 69 12411.09 5156 2099.09 -- -- --
CPM 33 12525.79 5193 2139.79 114.70 37 7.01 e−10
IPM 42 12476.10 5183 2110.10 65.01 27 5.60 e−05
Open in a new tab

Notes. IPM = independent pathway model; CPM = common pathway model; np = number of parameters; −2LL = −2 log likelihood; df = degrees of freedom; AIC = Akaike Information Criteria; ΔLL = log likelihood difference, Δdf = degree of freedom difference.

Genetic and environmental contributors to face-emotion recognition were ascertained from the best-fitting model based on the Cholesky decomposition. Strong evidence arose for familial aggregation of emotion recognition ability (i.e., cannot drop both A and C sources of familial aggregation, Table 3), with additional evidence suggesting that similarity among twins’ performance predominantly reflected genetic factors (Table 4).

Table 3.

Nested Models Fit for Variance Components of Best-Fitting Multivariate Model

comparison np −2LL df AIC ΔLL Δdf p
ACE 69 14103.12 5870 2363.12 -- -- --
CE 48 12455.76 5177 2101.76 44.67 21 0.002
AE 48 12424.80 5177 2070.80 13.71 21 0.88
E 27 12627.50 5198 2231.50 216.41 42 2.47 e−25
Open in a new tab

Notes. np = number of parameters; −2LL = −2 log likelihood; df = degrees of freedom; AIC = Akaike Information Criteria; ΔLL = log likelihood difference, Δdf = degree of freedom difference.

Table 4.

Variance Components for Recognition of Each Emotion

Emotion Additive Genetic Factors
(A)		 Shared Environmental Factors
(C)		 Nonshared Environmental Factors
(E)
Anger 0.40 (0.23, 0.53) 0.08 (0, 0.20) 0.53 (0.43, 0.63)
Happiness 0.57 (0.35, 0.67) 0.03 (0, 0.21) 0.40 (0.32, 0.49)
Sadness 0.34 (0.18, 0.49) 0.10 (0, 0.26) 0.56 (0.47, 0.65)
Fear 0.44 (0.24, 0.60) 0.12 (0, 0.29) 0.43 (0.35, 0.53)
Surprise 0.49 (0.30, 0.59) 0.01 (0, 0.16) 0.50 (0.41, 0.60)
Disgust 0.41 (0.16, 0.54) 0.06 (0, 0.25) 0.53 (0.44, 0.65)
Open in a new tab

For specific emotions, consistent evidence implicated genetic factors with similarly strong influences among emotions as indicated by overlapping confidence intervals (Table 4). Significance testing in nested-model comparisons generated little evidence of shared environmental factors influences (Tables 3 and 4), although these comparisons may be underpowered for tests with larger degrees of freedom54.

Shared Etiology of Face-emotion Recognition with Irritability, Neuroticism, and Depression

The phenotypic correlation of irritability, neuroticism, and depression with recognition of happiness appears largely due to common genetic factors. Genetic contributions to irritability, λ (factor loading) = −0.54, 95% CI [−0.66, −0.02], p = 0.04, and neuroticism, λ = −0.58, 95% CI [−0.72, −0.11], p = 0.01, loaded significantly onto the residual genetic factor for the recognition of happiness.A Notably, the loading of the genetic contribution to depression onto the factor for the recognition of happiness was similar but did not quite reach statistical significance, λ = −0.58, p = 0.09. There is no evidence of additional associations of genetic contributions to irritability, neuroticism, or depression with other genetic factors underlying face-emotion recognition, such as A1, A3, and A4 (see Figure 1). These correlations could be constrained to zero without appreciable loss of model fit for irritability, χ2 (5) = 2.50, p = 0.78, neuroticism, χ2 (5) = 2.07, p = 0.84, and depression, χ2 (5) = 3.15, p = 0.68.

Discussion

The present study used a genetically-informed epidemiological sample of child and adolescent twins to examine the contribution of genetic and environmental factors to face-emotion recognition and its association with irritability, neuroticism, and depression. After demonstrating strong test-retest reliability for the FELT task, the study demonstrated the heritability of both a common emotion recognition ability and specific abilities to recognize happiness, disgust, and fear/surprise. The present study replicated and extended prior findings linking childhood irritability12,55 and adult neuroticism19 and depression5 with impaired face-emotion recognition, specifically impaired recognition of happiness. Finally, the present study provides the first exploration of the contribution of common genetic factors to these face-emotion recognition deficits.

The study replicates the only prior twin study of emotion recognition, which found a large additive genetic contribution (i.e., 75%)20. This prior study attributed the ability to recognize each emotion to a single latent common factor, whereas the present, considerably larger study found that additional additive genetic factors for specific emotions augment this common factor18. The current study then extends prior research by leveraging the structure of genetic factors to clarify the contribution of shared genetic factors to the phenotypic association of negative affect with impaired face-emotion recognition.

Impaired face-emotion recognition has been implicated in many psychiatric conditions in children, adolescents, and adults including major depressive disorder5, bipolar disorder6,7, autism spectrum disorders8, alcohol dependence9, and attention-deficit/hyperactivity disorder10, several of which are characterized by irritability11 and neuroticism19. As hypothesized18, elevated depression symptoms were associated with a specific deficit in recognizing happiness. Additionally, irritability and neuroticism showed deficits in recognizing a broader range of emotions. Along with similar genetic associations of irritability, neuroticism, and depression with the recognition of happiness, this suggests that face-emotion recognition deficits may clarify points of convergence and divergence in the socioemotional functioning of multiple psychiatric disorders. Given evidence of common genetic influences to emotion recognition, irritability, and neuroticism, further research is needed to clarify the role of common genetic influences to the association of face-emotion recognition with the range of psychiatric conditions in children, adolescents, and adults. With further research, evidence of common genetic influences to face-emotion recognition deficits within multiple disorders might clarify a socioemotional pathway in the development of multiple comorbid psychiatric disorders56.

To clarify the role of face-emotion recognition as a socioemotional developmental pathway, the present data begin to satisfy the proposed criteria57 to suggest that face-emotion recognition, specifically the recognition of happiness, may present a useful psychiatric endophenotype58. The present data demonstrate that face-emotion recognition deficits, specifically regarding happiness, co-segregate with irritability, neuroticism, and depression within families partially due to mutual genetic influences. Additionally, the present community sample satisfies the requirement that the endophenotype be ‘state independent,’ i.e., manifest outside of active psychiatric episodes. While the sample approximates developmentally expected psychopathology rates23,26, results demonstrate the contribution of genetic and environmental factors to the association of face-emotion recognition with irritability, neuroticism, and depression in a community sample. Finally, the present study replicates existing evidence of impaired emotion recognition in children with elevated irritability12, neuroticism19, and related conditions14.

Several limitations should be addressed in future research. First, data were collected at a single time point. Developmental mechanisms are presumed to originate prior to the associated disorder. Longitudinal research is needed to characterize the development of face-emotion recognition and associated clinical phenotypes. For example, recent statistical developments estimate causal relationships within longitudinal research of mono- and dizygotic twins59. Second, although the genetic and environmental contributions to face-emotion recognition were similar between male and female participants (see Supplemental Table 3), sex effects in the genetic and environmental contributions to emotion recognition warrant further investigation. Lau and colleagues examine sex effects20 though both their sample and ours are underpowered to robustly estimate whether sex moderates genetic contributions60. Third, since participants are predominantly from one ancestral group, the present study optimizes statistical power for genomic analysis. However, further research is needed to generalize results to youth from other backgrounds. Finally, the present analyses are consistent with a small contribution of shared, familial environment to face-emotion recognition (Table 4). Though this is not statistically significant, careful interpretation is warranted as the chi-square approach is biased downwards in multivariate models54.

Despite the above limitations, these data demonstrate the importance of face-emotion recognition to socioemotional functioning in childhood and adolescence. Given deficits associated with irritability, neuroticism, and depression, the present study implicates face-emotion recognition in multiple psychiatric conditions. This supports incorporating emotion recognition and the amelioration of related deficits into existing treatment for childhood irritability, depression, and related psychopathologies. The present study encourages further study into the potentially broad clinical impact of existing interventions that address emotion recognition and regulation in childhood and adolescence61. Finally, considering improved emotion recognition during adult pharmacotherapy62, further research might examine face-emotion recognition as mechanism-of-action for pharmacological and cognitive-behavioral treatment of child and adolescent conditions.

Conclusion.

Deficits in recognizing others’ facial expression of emotion have been identified in multiple psychiatric conditions, particularly those characterized by negative affect. The present study provides a critical first step to evaluate the contribution of genetic and environmental influences to face-emotion recognition deficits. Moreover, the present study demonstrates that common genetic influences might explain specific deficits recognizing happiness among children and adolescents who report elevated irritability and neuroticism. Finally, to examine whether face-emotion recognition may present a candidate psychiatric endophenotype57, the present study demonstrates that face-emotion recognition deficits are, within a representative community sample, associated phenotypically with irritability, neuroticism, and depression; heritable; and co-segregate with irritability and neuroticism based on correlated genetic factors.

Supplementary Material

Supplemental Tables
Supplemental Figure S1

Supplemental Figure 1. Bivariate Scatterplots of Irritability, Neuroticism, and Depression with Recognition of Happiness

Note. Lines represent an empirical generalized additive model, shaded areas indicate a 95% confidence interval.

Acknowledgments

This study was supported by the National Institute of Mental Health (R01MH098055 to JMH, R01MH101518 to RR, NIMH-IRP-ziamh002781 to DSP, and T32MH020030) and by UL1TR000058 from the NCRR. The authors do not have any financial interests that might influence this research.

Footnotes

A

Residual genetic variance for irritability, neuroticism, and depression is sufficiently small that genetic correlations of each with the genetic factor specific to the recognition of happiness approach −1. For this reason, the standardized loading is presented to index each association distinct from the amount of residual variance, which obscures the genetic correlation.

The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. The material contained in this article has not been published elsewhere nor is it under consideration elsewhere. Two prior papers, which described the child and adolescent samples that comprise the present study, respectively, are cited in the text. Additionally, one prior paper (Cecilione et al., 2017) reported on the test-retest reliability data, which are described here using an alternative data analytic approach. This is described in the text. Preliminary research findings were presented at the 2017 annual meeting of the Anxiety and Depression Association of America (ADAA) in San Francisco, CA.

References

  • 1.McClure EB. A meta-analytic review of sex differences in facial expression processing and their development in infants, children, and adolescents. Psychological Bulletin. 2000;126:424–453. [DOI] [PubMed] [Google Scholar]
  • 2.Izard CE. Innate and universal facial expressions: evidence from developmental and cross-cultural research. Psychological Bulletin. 1994;115:288–299. [DOI] [PubMed] [Google Scholar]
  • 3.Nesse RM. Evolutionary explanations of emotions. Human nature. 1990;1(3):261–289. [DOI] [PubMed] [Google Scholar]
  • 4.Haxby JV, Hoffman EA, Gobbini MI. Human neural systems for face recognition and social communication. Biological psychiatry. 2002;51(1):59–67. [DOI] [PubMed] [Google Scholar]
  • 5.Dalili MN, Penton-Voak IS, Harmer CJ, Munafò MR. Meta-analysis of emotion recognition deficits in major depressive disorder. Psychological Medicine. 2015;45(06):1135–1144. doi: 10.1017/S0033291714002591 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Brotman MA, Guyer AE, Lawson ES, et al. Facial emotion labeling deficits in children and adolescents at risk for bipolar disorder. American Journal of Psychiatry. 2008;165:385–389. [DOI] [PubMed] [Google Scholar]
  • 7.Vierck E, Porter RJ, Joyce PR. Facial recognition deficits as a potential endophenotype in bipolar disorder. Psychiatry Research. 2015;230(1):102–107. doi: 10.1016/j.psychres.2015.08.033 [DOI] [PubMed] [Google Scholar]
  • 8.Lozier LM, Vanmeter JW, Marsh AA. Impairments in facial affect recognition associated with autism spectrum disorders: A meta-analysis. Development and Psychopathology. 2014;26(4pt1):933–945. doi: 10.1017/S0954579414000479 [DOI] [PubMed] [Google Scholar]
  • 9.Quaglino V, De Wever E, Maurage P. Relations between cognitive abilities, drinking characteristics, and emotional recognition in alcohol dependence: A preliminary exploration. Alcoholism: Clinical and Experimental Research. 2015;39(10):2032–2038. doi: 10.1111/acer.12841 [DOI] [PubMed] [Google Scholar]
  • 10.Shaw P, Stringaris A, Nigg J, Leibenluft E. Emotion dysregulation in attention deficit hyperactivity disorder. American Journal of Psychiatry. 2014;171(3):276–293. doi: 10.1176/appi.ajp.2013.13070966 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Vidal-Ribas P, Brotman MA, Valdivieso I, Leibenluft E, Stringaris A. The status of irritability in psychiatry: a conceptual and quantitative review. Journal of the American Academy of Child & Adolescent Psychiatry. 2016;55(7):556–570. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Leibenluft E Severe mood dysregulation, irritability, and the diagnostic boundaries of bipolar disorder in youths. American Journal of Psychiatry. 2011;168(2):129–142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Stringaris A, Cohen P, Pine DS, Leibenluft E. Adult outcomes of youth irritability: a 20-year prospective community-based study. American Journal of Psychiatry. 2009;166(9):1048–1054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Copeland WE, Shanahan L, Egger H, Angold A, Costello EJ. Adult diagnostic and functional outcomes of DSM-5 disruptive mood dysregulation disorder. American journal of psychiatry. 2014;171(6):668–674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Rich BA, Grimley ME, Schmajuk M, Blair KS, Blair RJR, Leibenluft E. Face emotion labeling deficits in children with bipolar disorder and severe mood dysregulation. Development and Psychopathology. 2008;20(02):529–546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Brotman MA, Rich BA, Guyer AE, et al. Amygdala activation during emotion processing of neutral faces in childen with severe mood dysregulation versus ADHD or bipolar disorder. American Journal of Psychiatry. 2010;167:61–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Thomas LA, Brotman MA, Bones BL, et al. Neural circuitry of masked emotional face processing in youth with bipolar disorder, severe mood dysregulation, and healthy volunteers. Developmental Cognitive Neuroscience. 2014;8:110–120. doi: 10.1016/j.dcn.2013.09.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Bourke C, Douglas K, Porter R. Processing of facial emotion expression in major depression: a review. Australian and New Zealand Journal of Psychiatry. 2010;44(8):681–696. [DOI] [PubMed] [Google Scholar]
  • 19.Chan SWY, Goodwin GM, Harmer CJ. Highly neurotic never-depressed students have negative biases in information processing. Psychological Medicine. 2007;37(09):1281. doi: 10.1017/S0033291707000669 [DOI] [PubMed] [Google Scholar]
  • 20.Lau JYF, Burt M, Leibenluft E, et al. Individual Differences in Children’s Facial Expression Recognition Ability: The Role of Nature and Nurture. Developmental Neuropsychology. 2009;34(1):37–51. doi: 10.1080/87565640802564424 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Neale MC, Cardon L. Methodology for Genetic Studies of Twins and Families. Vol 67 Springer Science & Business Media; 1992. [Google Scholar]
  • 22.Lilley EC, Silberg JL. The mid-Atlantic twin registry, revisited. Twin Research and Human Genetics. 2013;16:424–428. [DOI] [PubMed] [Google Scholar]
  • 23.Carney DM, Moroney E, Machlin L, et al. The twin study of negative valence emotional constructs. Twin Research and Human Genetics. 2016;19(05):456–464. doi: 10.1017/thg.2016.59 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Cecilione JL, Rappaport LM, Hahn SE, et al. Genetic and Environmental Contributions of Negative Valence Systems to Internalizing Pathways. Twin Research and Human Genetics. 2018;21(01):12–23. doi: 10.1017/thg.2017.72 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Price AL, Zaitlen NA, Reich D, Patterson N. New approaches to population stratification in genome-wide association studies. Nature Reviews Genetics. 2010;11(7):459–463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Rappaport LM, Sheerin C, Savage JE, Hettema JM, Roberson-Nay R. Clinical correlates of latent classes of CO2 hypersensitivity over time in adolescents and young adults. Behaviour Research and Therapy. 2017;93:95–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Rappaport LM, Sheerin C, Carney DM, et al. Clinical Correlates of Carbon Dioxide Hypersensitivity in Children. Journal of the American Academy of Child & Adolescent Psychiatry. September 2017. doi: 10.1016/j.jaac.2017.09.423 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Kaufman J, Birmaher B, Brent D, et al. Schedule for affective disorders and schizophrenia for school-age children-present and lifetime version (K-SADS-PL): initial reliability and validity data. Journal of the American Academy of Child & Adolescent Psychiatry. 1997;36:980–988. [DOI] [PubMed] [Google Scholar]
  • 29.Wittchen H-U. Reliability and validity studies of the WHO-composite international diagnostic interview (CIDI): A critical review. Journal of Psychiatric Research. 1994;28:57–84. [DOI] [PubMed] [Google Scholar]
  • 30.Fergusson DM, Horwood J, Lynskey MT. Prevalence and comorbidity of DSM-III-R diagnoses in a birth cohort of 15 year olds. Journal of the American Academy of Child & Adolescent Psychiatry. 1993;32:1127–1134. [DOI] [PubMed] [Google Scholar]
  • 31.Costello EJ, Egger HL, Angold A. The developmental epidemiology of anxiety disorders: Phenomenology, prevalence, and comorbidity. Child and Adolescent Psychiatric Clinics of North America. 2005;14(4):631–648. doi: 10.1016/j.chc.2005.06.003 [DOI] [PubMed] [Google Scholar]
  • 32.Merikangas KR, He J, Burstein M, et al. Lifetime Prevalence of Mental Disorders in U.S. Adolescents: Results from the National Comorbidity Survey Replication–Adolescent Supplement (NCS-A). Journal of the American Academy of Child & Adolescent Psychiatry. 2010;49(10):980–989. doi: 10.1016/j.jaac.2010.05.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Jackson RW, Snieder H, Davis H, Treiber FA. Determination of twin zygosity: a comparison of DNA with various questionnaire indices. Twin Research. 2001;4(01):12–18. [DOI] [PubMed] [Google Scholar]
  • 34.Kasriel J, Eaves L. The zygosity of twins: Further evidence on the agreement between diagnosis by blood groups and written questionnaires. Journal of Biosocial Science. 1976;8:263–266. [DOI] [PubMed] [Google Scholar]
  • 35.Stringaris A, Goodman R, Ferdinando S, et al. The Affective Reactivity Index: a concise irritability scale for clinical and research settings. Journal of Child Psychology and Psychiatry. 2012;53(11):1109–1117. doi: 10.1111/j.1469-7610.2012.02561.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Eysenck SBG. A new scale for personality measurements in children. British Journal of Educational Psychology. 1965;35:362–367. [DOI] [PubMed] [Google Scholar]
  • 37.Angold A, Costello EJ, Messer SC, Pickles A, Winder F, Silver D. Development of a short questionnaire for use in epidemiological studies of depression in children and adolescents. International Journal of Methods in Psychiatric Research. 1995;5:237–249. [Google Scholar]
  • 38.Ekman P, Friesen W. Pictures of Facial Affect. Palo Alto: Consulting Psychologists; 1976. [Google Scholar]
  • 39.Blair RJR, Colledge E, Murray L, Mitchell DGV. A selective impairment in the processing of sad and fearful expressions in children with psychopathic tendencies. Journal of Abnormal Child Psychology. 2001;29(6):491–498. [DOI] [PubMed] [Google Scholar]
  • 40.Marsh AA, Yu HH, Pine DS, Blair RJR. Oxytocin improves specific recognition of positive facial expressions. Psychopharmacology. 2010;209(3):225–232. doi: 10.1007/s00213-010-1780-4 [DOI] [PubMed] [Google Scholar]
  • 41.Young AW, Rowland D, Calder AJ, Etcoff NL, Seth A, Perrett DI. Facial expression megamix: Tests of dimensional and category accounts of emotion recognition. Cognition. 1997;63:271–313. [DOI] [PubMed] [Google Scholar]
  • 42.Adams T, Pounder Z, Preston S, et al. Test-retest reliability and task order effects of emotional cognitive tests in healthy subjects. Cognition and Emotion. 2015:1–13. [DOI] [PubMed] [Google Scholar]
  • 43.Cecilione JL, Rappaport LM, Verhulst B, et al. Test-retest reliability of the facial expression labeling task. Psychological Assessment. 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Wagner HL. On measuring performance in category judgment studies of nonverbal behavior. Journal of Nonverbal Behavior. 1993;17(1):3–28. [Google Scholar]
  • 45.Neale MC, Røysamb E, Jacobson K. Multivariate Genetic Analysis of Sex Limitation and G × E Interaction . Twin Research and Human Genetics. 2006;9(04):481–489. doi: 10.1375/twin.9.4.481 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria; 2015. https://www.R-project.org/. [Google Scholar]
  • 47.Revelle W Psych: Procedures for Personality and Psychological Research. Evanston, Illinois, USA: Northwestern University; 2015. http://CRAN.R-project.org/package=psych. [Google Scholar]
  • 48.Fox J, Weisberg S. An {R} Companion to Applied Regression. Thousand Oaks, CA: Sage; 2011. http://socserv.socsci.mcmaster.ca/jfox/Books/Companion. [Google Scholar]
  • 49.Wickham H The split-apply-combine strategy for data analysis. Journal of Statistical Software. 2011;40:1–29. [Google Scholar]
  • 50.Wickham H Ggplot2: Elegant Graphics for Data Analysis. New York: Springer-Verlag; 2009. [Google Scholar]
  • 51.Pinheiro J, Bates D, DebRoy S, Sarkar D, R Core Team. Nlme: Linear and Nonlinear Mixed Effects Models.; 2016. http://CRAN.R-project.org/package=nlme. [Google Scholar]
  • 52.Bates TC, Neale M, Maes H. umx: Twin and Path-based Structural Equation Modeling in OpenMx.:20. [DOI] [PubMed] [Google Scholar]
  • 53.Neale MC, Hunter MD, Pritikin JN, et al. OpenMx 2.0: Extended structural equation and statistical modeling. Psychometrika. 2016;81:535–549. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Wu H, Neale MC. On the Likelihood Ratio Tests in Bivariate ACDE Models. Psychometrika 2013;78(3):441–463. doi: 10.1007/s11336-012-9304-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Brotman MA, Kircanski K, Leibenluft E. Irritability in children and adolescents. Annual Review of Clinical Psychology. 2017;13:317–341. [DOI] [PubMed] [Google Scholar]
  • 56.Costello EJ, Mustillo S, Erkanli A, Keeler G, Angold A. Prevalence and development of psychiatric disorders in childhood and adolescence. Archives of General Psychiatry. 2003;60:837–844. [DOI] [PubMed] [Google Scholar]
  • 57.Gottesman II, Gould TD. The endophenotype concept in psychiatry: Etymology and strategic intentions. American Journal of Psychiatry. 2003;160(4):636–645. [DOI] [PubMed] [Google Scholar]
  • 58.Insel TR, Cuthbert BN. Endophenotypes: Bridging genomic complexity and disorder heterogeneity. Biological Psychiatry. 2009;66(11):988–989. doi: 10.1016/j.biopsych.2009.10.008 [DOI] [PubMed] [Google Scholar]
  • 59.Duffy DL, Martin NG. Inferring the direction of causation in cross-sectional twin data: Theoretical and empirical considerations. Genetic epidemiology. 1994;11(6):483–502. [DOI] [PubMed] [Google Scholar]
  • 60.Verhulst B A Power Calculator for the Classical Twin Design. Behavior Genetics. 2017;47:255–261. doi: 10.1007/s10519-016-9828-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Bauminger N The Facilitation of Social-Emotional Understanding and Social Interaction in High-Functioning Children with Autism: Intervention Outcomes. Journal of Autism and Developmental Disorders. 2002;32:283–298. [DOI] [PubMed] [Google Scholar]
  • 62.Tranter R, Bell D, Gutting P, Harmer C, Healy D, Anderson IM. The effect of serotonergic and noradrenergic antidepressants on face emotion processing in depressed patients. Journal of Affective Disorders. 2009;118(1–3):87–93. doi: 10.1016/j.jad.2009.01.028 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Tables
NIHMS1558654-supplement-Supplemental_Tables.docx (27.1KB, docx)
Supplemental Figure S1

Supplemental Figure 1. Bivariate Scatterplots of Irritability, Neuroticism, and Depression with Recognition of Happiness

Note. Lines represent an empirical generalized additive model, shaded areas indicate a 95% confidence interval.

NIHMS1558654-supplement-Supplemental_Figure_S1.pdf (31.4KB, pdf)

RESOURCES