Predicting the accuracy of facial affect recognition: The interaction of child maltreatment and intellectual functioning

https://doi.org/10.1016/j.jecp.2012.08.007Get rights and content

Abstract

Previous research demonstrates that both child maltreatment and intellectual performance contribute uniquely to the accurate identification of facial affect by children and adolescents. The purpose of this study was to extend this research by examining whether child maltreatment affects the accuracy of facial recognition differently at varying levels of intellectual functioning. A sample of maltreated (n = 50) and nonmaltreated (n = 56) adolescent females, 14 to 19 years of age, was recruited to participate in this study. Participants completed demographic and study-related questionnaires and interviews to control for potential psychological and psychiatric confounds such as symptoms of posttraumatic stress disorder, negative affect, and difficulties in emotion regulation. Participants also completed an experimental paradigm that recorded responses to facial affect displays starting in a neutral expression and changing into a full expression of one of six emotions: happiness, sadness, anger, disgust, fear, or surprise. Hierarchical multiple regression assessed the incremental advantage of evaluating the interaction between child maltreatment and intellectual functioning. Results indicated that the interaction term accounted for a significant amount of additional variance in the accurate identification of facial affect after controlling for relevant covariates and main effects. Specifically, maltreated females with lower levels of intellectual functioning were least accurate in identifying facial affect displays, whereas those with higher levels of intellectual functioning performed as well as nonmaltreated females. These results suggest that maltreatment and intellectual functioning interact to predict the recognition of facial affect, with potential long-term consequences for the interpersonal functioning of maltreated females.

Highlights

► Predict the accuracy of responses to facial affect displays. ► Sampled maltreated and non-maltreated adolescent females. ► Tested the interaction between child maltreatment and intellectual functioning. ► Found the interaction significantly predicted accuracy scores. ► Discuss implications for emotional and interpersonal development.

Introduction

Facial expressions provide a rich source of information that viewers can use to generate hypotheses about the current emotional state of another person, leading to contextually relevant behavioral responses that have important individual and interpersonal functions. For instance, certain discrete facial movements can inform the viewer that another person is experiencing sadness. Once these facial movements are detected and appropriately categorized as an expression of sadness, the viewer can then provide an empathic verbal response to achieve individual goals, such as identifying the conditions leading to this emotional state, or interpersonal goals, such as providing general emotional support to a friend or facilitating individual emotion regulation in a child. Such responses can also inform the person expressing the emotion on how to communicate sadness in a way that provides access to important individual and social contingencies, such as empathy. However, different expressions of affect communicate different interpersonal needs, and accurately recognizing a host of different affective expressions can increase the probability of discriminating which behavioral response to provide in a given context. Thus, accurately recognizing facial expressions of affect is a key developmental task for children and adolescents with significant implications for individual and interpersonal functioning.

Facial affect recognition is facilitated by several neurological processes. Discrete movements in facial muscles are detected visually (Ekman & Friesen, 1978) and are processed by both global and specific neural systems associated with affect (Sabatini et al., 2009), including the amygdala (Monk et al., 2003, Phillips et al., 2004), flexible fusiform area (van de Riet, Grezes, & de Gelder, 2009), inferior parietal cortex (Adolphs, Damasio, Tranel, & Damasio, 1996), and orbitofrontal and occipital cortices (Sabatinelli et al., 2011). Activation in these neural systems leads to further information processing in key frontal areas responsible for higher order processes and tasks, such as language and categorization (Sprengelmeyer, Rausch, Eysel, & Przuntek, 1998). Activating frontal cortices provides access to encoded labels for emotion previously paired with similarly expressed movements in facial muscles. This process allows the individual to put socially acceptable emotion labels to visually processed information in order to increase the probability of correctly categorizing the expressed facial display.

Prior learning also plays a critical role in how facial affect displays are classified into distinct emotion categories (Pollak & Kistler, 2002). Like many behaviors, the strength of accurately recognizing facial affect depends largely on the frequency and intensity of exposure to different expressions (Beale and Keil, 1995, Keyes, 2012, Pollak, 2003). Parents can play a particularly important role in shaping affect recognition by providing repeated occasions where specific emotion labels corresponding to current affective expressions are given to a child. In contrast to parents who do not provide such opportunities (Krause et al., 2003, Sullivan et al., 2010), parents who label and model affective expressions have children who are more competent emotionally (Denham et al., 1997, Fruzzetti and Shenk, 2008), setting them on a trajectory for improved emotional (Shipman et al., 2007), interpersonal (Eisenberg et al., 2001), and behavioral (Eisenberg et al., 2005) outcomes. As perceptual acuities, neurological systems, and learning experiences are continually refined and shaped across development, the ability to recognize emotions more accurately increases over time, providing important advantages for optimal development.

Research on the effects of child maltreatment, including physical abuse, sexual abuse, and neglect, has highlighted the importance of how certain environmental events can disrupt key developmental processes, including the accurate recognition of facial affect (Pollak, 2008). Child maltreatment affects nearly 700,000 children each year in the United States (US Department of Health & Human Services, 2011) and is associated with a number of adverse developmental (Shields & Cicchetti, 2001), physical (Bentley & Widom, 2009), and psychological (Fergusson, Boden, & Horwood, 2008) health outcomes. There is a growing literature linking instances of child maltreatment to alterations in neurological structures, including the visual cortex (Tomoda, Navalta, Polcari, Sadato, & Teicher, 2009), orbitofrontal cortex (Hanson et al., 2010), amygdala (Maheu et al., 2010, Mehta et al., 2009), prefrontal cortex (Carrion et al., 2009), and cerebellar volumes (De Bellis & Kuchibhatla, 2006). This research suggests that child maltreatment may significantly affect brain–behavior functioning in the domains specifically needed for facial affect recognition (McCrory et al., 2010, Pollak et al., 2010). Child maltreatment is also linked to general and specific deficits in affect recognition when compared with nonmaltreated controls (Pears and Fisher, 2005, Pollak et al., 2001). Moreover, there are a number of learning experiences unique to maltreated children that can affect the accuracy of facial affect recognition. Neglected children have greater difficulty in discriminating between distinct affective expressions (Pollak, Cicchetti, Hornung, & Reed, 2000), which may be due to living in an environment that deprives these children of learning how to recognize and label different affective expressions. Physically abused children can more accurately detect displays of anger, even with less sensory information, when compared with nonmaltreated controls (Pollak & Sinha, 2002). This is believed to result from more frequent and intense exposure to anger displays in these homes (Pollak, Messner, Kistler, & Cohn, 2009) while serving to increase prediction of potential physical aggression in the future. Finally, maltreated children, including those who experienced sexual abuse, were no less accurate in recognizing facial affect but displayed significantly faster reaction times when compared with nonmaltreated controls (Masten et al., 2008). This suggests that maltreated children may become sensitized to the recognition of specific negative emotions that serve a potentially adaptive function in their current environment. Thus, child maltreatment may affect both the biological systems and learning processes central to the accurate recognition of facial affect.

Intellectual functioning plays a key role in accurately identifying expressions of affect (Anderson & Miller, 1998), most likely because it measures fluid and crystallized abilities that are shaped by both neurological development and prior learning experiences (Horn & Noll, 1997). Refinement of fluid and crystallized abilities corresponds with a developmental trend of improved affect recognition from childhood (Szekely et al., 2011), through adolescence (Gao & Maurer, 2009), and into adulthood (Horning, Cornwell, & Davis, in press). Thus, variations in the acquisition of fluid and crystallized abilities may have a differential impact on one’s ability to recognize affect in others. Deficits in intellectual functioning are consistently noted for maltreated children (Carrey et al., 1995, De Bellis et al., 2009), which can extend throughout the life course (Noll et al., 2010) and potentially limit one’s ability to recognize affect across developmental stages. However, much of the previous research on affect recognition in maltreated children views intellectual functioning as a nuisance variable requiring statistical or methodological control to detect the main effects of child maltreatment despite evidence that intellectual functioning plays an incremental role over neglectful parenting when modeling emotional knowledge (Sullivan et al., 2010). This approach prevents an opportunity to examine whether the relationship between child maltreatment and affect recognition varies across different levels of intellectual functioning. This could identify intellectual ability as a protective or risk factor for maltreated children, yielding important implications for prevention and intervention programs via direct targeting of fluid and crystallized abilities.

The developmental traumatology model (De Bellis et al., 1999) specifies that child maltreatment affects developing biological mechanisms that alter the performance and structure of neurological systems regulating developmental and behavioral outcomes. Extending this model to the domain of affect recognition, maltreatment status and intellectual functioning are regarded as independent variables, whereas accurately detecting and discriminating between different emotions is a key developmental outcome. An important area of research in developmental traumatology is to assess the interactions between independent variables, such as child maltreatment and intellectual functioning, to understand the impact of these variables on child development more completely. Directly examining the interaction between child maltreatment and intellectual functioning can identify whether the effects of child maltreatment on affect recognition are constant across varying levels of intellectual ability.

The primary research hypothesis for the current study stated that child maltreatment and intellectual functioning will interact to predict the accuracy of facial affect recognition. Given the extant literature showing the unique effects of both child maltreatment and intellectual functioning as main effect variables, it was specified that the interaction term will explain a significant amount of additional variance in the accurate recognition of facial affect when compared with a prior model estimating child maltreatment and intellectual functioning as single indicator variables. To test this hypothesis, adolescents experiencing multiple forms of substantiated child maltreatment, sexual abuse, physical abuse, and neglect, were recruited along with a nonmaltreated comparison group. An assessment of overall accuracy in response to the facial expressions of six primary emotions—happiness, sadness, fear, disgust, surprise, and anger—was undertaken. Common psychological and psychiatric outcomes of maltreatment, such as symptoms of posttraumatic stress disorder (PTSD), difficulties in emotion regulation, and negative affect, were assessed for use as potential covariates in statistical modeling because this has been noted as a significant limitation of research on affect recognition with maltreated samples (Hart and Rubia, 2012, Saigh et al., 2006).

Section snippets

Participants

Adolescent females (N = 106), 14 to 19 years of age, participated in this study. An exclusively female sample was recruited because females are more likely to be abused overall and are more likely to experience sexual abuse specifically (Sedlak et al., 2010), an underrepresented subpopulation in the existing literature. Moreover, older adolescents were recruited given results suggesting that the effect of trauma, such as child maltreatment, may be mitigated during later stages of development (

Results

Preliminary data analysis was conducted to identify potential covariates for subsequent statistical modeling. Covariates were identified based on demographic and study-related variables that differed significantly between maltreated and nonmaltreated groups or that had a significant relationship with the primary outcome, that is, overall accuracy in detecting facial affect displays. Chi-square and multivariate analysis of variance (MANOVA) tests assessed differences between maltreatment and

Discussion and conclusions

Results from the current study support general findings in the existing research literature on facial affect recognition while contributing substantive new findings. First, maltreatment status was significantly related to the accuracy of affect recognition, most notably in response to expressions of fear. This finding supports previous research identifying deficits in the accurate recognition of facial affect, especially fear, in maltreated samples (Pollak et al., 2000). Moreover, the

Acknowledgements

This project was supported by an Institutional Clinical and Translational Science Award (NIH/NCRR Grant #: 1UL1RR026314), the National Institute on Child Health and Human Development (Noll: R01HD052533), and the National Institute of Diabetes and Digestive and Kidney Diseases (Powers: T32DK063929).

References (71)

  • C.L. Masten et al.

    Recognition of facial emotions among maltreated children with high rates of post-traumatic stress disorder

    Child Abuse & Neglect

    (2008)
  • M.L. Phillips et al.

    Differential neural responses to overt and covert presentations of facial expressions of fear and disgust

    NeuroImage

    (2004)
  • S.D. Pollak et al.

    Development of perceptual expertise in emotion recognition

    Cognition

    (2009)
  • D. Sabatinelli et al.

    Emotional perception: Meta-analyses of face and natural scene processing

    NeuroImage

    (2011)
  • E. Sabatini et al.

    Brain structures activated by overt and covert emotional visual stimuli

    Brain Research Bulletin

    (2009)
  • A. Tomoda et al.

    Childhood sexual abuse is associated with reduced gray matter volume in visual cortex of young women

    Biological Psychiatry

    (2009)
  • E.W. Twamley et al.

    Neuropsychological function in college students with and without posttraumatic stress disorder

    Psychiatry Research

    (2004)
  • R. Adolphs et al.

    Cortical systems for the recognition of emotion in facial expressions

    Journal of Neuroscience

    (1996)
  • American Psychiatric Association

    Diagnostic and statisical manual of mental disorders-IV – Text revision (DSM-IV – TR)

    (2000)
  • R.T. Ammerman et al.

    An open trial of in-home CBT for depressed mothers in home visitation

    Maternal and Child Health Journal

    (2011)
  • M. Anderson et al.

    Modularity, mental retardation, and speed of processing

    Developmental Science

    (1998)
  • T. Bentley et al.

    A 30-year follow-up of the effects of child abuse and neglect on obesity in adulthood

    Obesity

    (2009)
  • J.F. Cohn et al.

    Automated face analysis by feature point tracking has high concurrent validity with manual FACS coding

    Psychophysiology

    (1999)
  • J.R. Crawford et al.

    The Positive and Negative Affect Schedule (PANAS): Construct validity, measurement properties, and normative data in a large non-clinical sample

    British Journal of Clinical Psychology

    (2004)
  • J. Davidson et al.

    Validity and reliability of the DSM-III criteria for posttraumatic stress disorder: Experience with a structured interview

    Journal of Nervous and Mental Disease

    (1989)
  • M.D. De Bellis et al.

    Neuropsychological findings in childhood neglect and their relationships to pediatric PTSD

    Journal of the International Neuropsychological Society

    (2009)
  • S.A. Denham et al.

    Parental contributions to preschoolers’ emotional competence. Direct and indirect effects

    Motivation & Emotion

    (1997)
  • N. Eisenberg et al.

    Mother’s emotional expressivity and children’s behavior problems and social competence. Mediation through children’s regulation

    Developmental Psychology

    (2001)
  • N. Eisenberg et al.

    Relations among positive parenting, children’s effortful control, and externalizing problems: A three-wave longitudinal study

    Child Development

    (2005)
  • P. Ekman et al.

    Facial action coding system: A technique for the measurement of facial movement

    (1978)
  • M.J. Farah et al.

    What is “special” about face perception?

    Psychological Review

    (1998)
  • A.E. Fruzzetti et al.

    Fostering validating responses in families

    Social Work in Mental Health

    (2008)
  • O. Golan et al.

    Systemizing empathy: Teaching adults with Asperger syndrome or high-functioning autism to recognize complex emotions using interactive multimedia

    Development and Psychopathology

    (2006)
  • K.L. Gratz et al.

    Multidimensional assessment of emotion regulation and dysregulation: Development, factor structure, and initial validation of the difficulties in emotion regulation scale

    Journal of Psychopathology & Behavioral Assessment

    (2004)
  • M.J. Guralnick

    Early intervention for children with intellectual disabilities: Current knowledge and future prospects

    Journal of Applied Research in Intellectual Disabilities

    (2005)
  • Cited by (21)

    • A systematic review of measures of theory of mind for children

      2023, Developmental Review
      Citation Excerpt :

      In the preschool years (ages 3 to 6 years), children rapidly develop their basic ToM functioning. In their third year, children can match emotion words to photographs of faces (the developmental component “emotion distinction”) (Assed et al., 2020; Masten et al., 2008; Miller, 2006; Shenk et al., 2013). Children also begin to develop awareness of a sense of self and recognize that they are dissociated from others.

    • Language disorders in victims of domestic violence in children's homes

      2018, Child Abuse and Neglect
      Citation Excerpt :

      Neuroscience has enriched this field of study as it has demonstrated how child abuse affects the brain structure and its functioning (Mesa & Moya, 2011; Carrion et al., 2008; Wilson, Hansen, & Li, 2011). Such effects have been tightly related to alterations in the cognitive functions (Bücker et al., 2012; Carrion, Wong, & Kletter, 2013; Kale & Urbano, 2013; Shenk, Putnam, & Noll, 2013). Among the cognitive function effects, language, which is considered as the mediator of behavior and other cognitive functions (Escoriza, 2003), turns out to be the most affected one.

    • Electrocortical processing of social signals of threat in combat-related post-traumatic stress disorder

      2013, Biological Psychology
      Citation Excerpt :

      Based on prior work, we thought it possible that we would observe either increased amplitudes to threatening stimuli, reflecting hypervigilance (Karl et al., 2006; Shin et al., 2005) or decreased amplitudes, possibly reflecting avoidance or numbing symptoms (Felmingham et al., 2003; Felmingham, Bryant, Kendall, & Gordon, 2002). Behaviorally, we expected to observe reduced accuracy for individuals with PTSD – especially for trials containing threatening expressions (Pollak, Cicchetti, Hornung, & Reed, 2000; Shenk, Putnam, & Noll, 2012). Participant demographics, symptomatology and comorbid diagnoses are presented in Table 1.

    View all citing articles on Scopus
    View full text