1
Faculty of Social and Behavioural Sciences
Graduate School of Childhood Development and Education
The Relationship between Anxiety and Depressive Symptoms and
Parents’ Emotional Expression and Gaze during Triadic
Parent-Infant-Person Interactions
Research Master Child Development and Education Thesis II
Alyssa N. Kidd
Supervisors: Dr. E. Aktar and Dr. C. Colonessi
2
Abstract
Objective: The aim of this study was to evaluate the duration of parents’ emotional expression
and gaze during triadic interactions with their infant in the context of parent anxiety and/or depressive symptoms. Method: Typically developing children (N = 64) and parents (Nmothers = 50
and Nfathers = 14) were observed in a triadic parent-infant-person interaction during computerized
presentations of happy, sad, angry, fearful and neutral facial expressions. The duration of
parents’ emotional expressions (positive, negative and neutral), and gaze (to screen, to child and elsewhere), were observed. Parents’ anxiety and depressive symptoms were measured using the Depression Anxiety Stress Scales. Results: Parents’ anxiety symptoms did not predict parents’ emotional expressions or gaze. Higher levels of depressive symptoms were associated with increased neutral emotions during the fearful screen (p < .05). Parents were more positive (vs. neutral) and gazed less at the screen (vs. elsewhere) during the happy (vs. neutral) screen (p < .001). Mothers and fathers differed significantly in gaze during the angry screen (p < .01). Specifically, fathers gazed elsewhere more frequently. Conclusion: The findings provide preliminary support for the idea that sub-clinical levels of anxiety symptoms do not predict parents’ emotional expression or gaze. Specifically, anxiety symptoms may not impair parents’ ability to offer positive reinforcement in the face of ambiguous emotions during triadic
interactions. Parent depressive symptoms may have an impact upon parents’ emotions (but not gaze), in response to fearful expressions. Finally, fathers’ increased duration of gaze elsewhere during angry facial expressions may aim to divert the child’s attention elsewhere in response to ambiguous stimuli.
3
The Relationship between Anxiety and Depressive Symptoms and Parents’ Emotional Expression and Gaze during Triadic Parent-Infant-Person Interactions
Infants’ exposure to parents’ anxiety and/or depressive symptoms in early life has been linked to negative emotional, behavioural and psychological outcomes in infancy and beyond (Murray et al., 2011). During social referencing (SR), infants attend to their parents’ emotional signals coupled with parents’ gaze direction to link the emotional signals to ambiguous stimuli and to regulate their behaviour and emotional response accordingly (Feinman et al. 1992). Previous research indicates that parents with a lifetime or current diagnosis of social anxiety or with lifetime co-morbid social and other anxiety disorders express increased anxiety compared to parents without anxiety disorders during SR contexts (Aktar, Majdandžić, de Vente, & Bögels, 2013; Murray et al., 2008). Parents’ symptoms of depression may also interfere with parents’ emotional expressions during face-to-face interactions. Interactions of depressed mothers are often characterized by increased frequency of negative and neutral affect, as well as less frequent positive affect. This can have a subsequent negative impact upon infants’ emotion learning (Als et al., 1979; Cohn & Tronick, 1987). A recent study by Aktar et al (2015) explored the
associations between parents’ depression and anxiety symptoms and parents’ and infants’ emotional expression during face-to-face interactions in infants aged 3.5 to 5.5 months (N = 101). This study found that parents’ depression (but not anxiety) symptoms were linked to less positive and more neutral parent affect during face-to-face interactions. Additionally, infants of anxious parents expressed more positive and negative expressions, thus, suggesting a link between parents’ anxiety and infants’ expressions. However, the majority of previous studies have investigated the impact of parents’ anxiety and depressive symptoms upon parents’
4
emotional expression during dyadic face-to-face interactions, while the effect of parents’ anxiety and depressive symptoms during triadic parent-infant-person interactions remains to be
investigated.
Triadic parent-infant-person interactions offer a unique temporal environment to observe everyday triadic SR processes which incorporate parents’ emotional expressions and gaze. During triadic parent-infant-person interactions, infants utilize parents’ emotional signals and gaze as a reference to determine how to respond when confronted with unfamiliar people
(Feinman, 1982). Thus, during triadic parent-infant-person interactions, infants jointly rely upon parents’ emotional expression and the parent’s dynamic use of gaze. The infant’s simultaneous processing of emotional expressions and referential gaze direction forms the foundation of emotion-object associations (Feinman et al. 1992). For example, Moses, Baldwin, Rosicky, and Tidball (2001) investigated the impact of gaze direction during SR with infants at 12 months (N = 48) and 18 months (N = 32). Infants were presented with novel objects coupled with positive or negative emotional expressions together with gaze towards the object, or a non-observable gaze. A significant effect was observed for positive affect versus negative affect; infants approached the novel object more closely following positive affect (M = 2.2, SD = 1.0) compared to negative affect (M = 2.5, SD = .9). Interestingly, gaze direction had a significant effect upon 18-month old infants’ proximity to the novel object, however this result was not observed in 12-month old infants. These results suggest that 18-month old infants rely on adults’ gaze direction in SR situations to infer whether an emotional signal relates to the ambiguous stimulus. However, 12-month old infants did not demonstrate sensitivity to referential cues. However, the findings of Mumme and Fernald (2003) elicited conflicting results. This study investigated if emotional signals with gaze directed towards a novel object depicted on television
5
influenced 10-12 month old infants’ subsequent tendencies to interact with the object in real life. They concluded that infants at 10 months (N = 32) were not influenced by the emotional signals with gaze directed towards the object, however infants at 12 months (N = 32) did attend to gaze direction during SR situations and regulated their responses appropriately. However, most previous studies have investigated the impact of parents’ anxiety and depressive symptoms upon parents’ emotions and gaze during dyadic face-to-face interactions only, and did not consider the impact of parents’ anxiety and depressive symptoms upon parents’ emotions and gaze during triadic contexts.
The consideration of infant exposure to parents’ anxiety symptoms during triadic parent-infant-person interactions is valuable because it seems that observational learning of parental behavior via SR may contribute to children’s learning of anxiety from parents (Fisak & Grills-Taquechel, 2007; Murray, Creswell, & Cooper, 2009). Indeed, children whose parents have anxiety disorders are more likely to subsequently develop anxiety disorders themselves
(Weissman et al. 1984; Turner et al. 1987). Furthermore, higher levels of previously expressed anxiety during SR tasks by parents with lifetime co-morbid social and other anxiety disorders predicted higher levels of anxiety/avoidance in children at 30 months, suggesting that infants’ exposure to parents’ anxiety during SR in infancy contributes to the intergenerational
transmission of anxiety (Aktar, Majdandžić, Vente, & Bögels, 2014).
The consideration of infant exposure to parents’ depressive symptoms during triadic parent-infant-person interactions is also valuable as it is suggested that infants of clinically depressed parents experience an atypical emotional environment during parent-infant-person interactions. This atypical environment is often characterized by disproportionately high exposure to negative or neutral facial expressions. A recent discussion suggested that the
6
characteristic flat affect of depressed parents might compromise parents’ ability to display appropriate emotional signals to their infant during SR situations (Peláez, Virues-Ortega, Field, Amir-Kiaei, and Schnerch, 2013). This may, in turn, impact the infant’s attention allocation to novelty during SR situations. Within dyadic contexts, infants of depressed mothers were less responsive to facial stimuli and did not appear to perceive sad expressions as novel, compared with infants of non-depressed parents (Field, Diego, & Hernanez-Reif, 2009). Specifically, a study of 3 and 6 month old infants of depressed and non-depressed mothers (Field, Pickens, Fox, Gonzalez & Nawrocki, 1998) found that infants of depressed mothers attended to sad faces for less time in comparison to infants of non-depressed mothers. These findings support the suggestion that negative expressions of emotion are perceived as less novel by infants of
depressed parents, most likely due to frequent exposure to sad facial expressions. Furthermore, it has also been suggested that infant exposure to parents’ depressive symptoms may effect infant emotional expression. An early study investigated infant exploration and parents’ emotional expression in both dyadic and triadic contexts (Hart, Field, Del Valle, & Peláez-Nogueras, 1998). This study found that daughters of depressed parents expressed less positive and more negative affect than daughters of non-depressed parents. However, the same results were not elicited for boys of depressed parents.
Previous SR research has indicated that the age where infants’ SR behaviour is observed to be the most salient is between the ages of 10 to 14 months (Emde, 1992). However, infants are thought to rely on adults’ gaze direction to infer whether an emotional signal relates to the
ambiguous stimulus from the age of 12 months (Moses, Baldwin, Rosicky, and Tidball, 2001; Mumme and Fernald, 2003). It is thought that infants younger than 12 months are not influenced by gaze direction during SR situations in order to regulate their responses (Mumme & Fernald,
7
2003). 12-month-old infants were therefore chosen for the purpose of this study in order to attempt replication of the above findings in a triadic parent-infant-person SR context.
Interestingly, the role of fathers in SR has rarely been considered. This is despite the hypothesis that fathers’ expressed anxiety in SR situations is as influential as mothers’ at 12 months (Aktar et al., 2013). Moreover, it has been suggested in recent anxiety development models that the father’s role becomes more influential as children grow older, stemming from their role in the development of anxiety which is related to their greater evolutionary experience in exploring the external world (Bögels & Perotti, 2011; Bögels & Phares, 2008). Therefore, the inclusion of fathers as well as mothers is thought to be valuable in the continued study of SR.
Previous research has investigated the effect of infant gender upon dyadic parent-infant interactions, yet has yielded inconsistent results. Several studies (e.g., Cohn & Tronick, 1987; Kaye & Fogel, 1980) have found no gender effects upon affect expression within mother-infant dyads. Similarly, in a study of postpartum depressed and non-depressed mothers alongside their 2-month old children, it was found that the bidirectional effects of mother-infant behaviour were unrelated to infant sex (Cohn, Campbell, Matias & Hopkins, 1990). On the other hand, a study by Haviland (1977) found that female participants were less responsive to negative expressions in daughters in comparison to in sons.
To summarize, the evidence suggests that parents’ anxiety and/or depressive symptoms during dyadic SR contexts can have a significant relationship with parents’ emotional expression and gaze, and a subsequent impact upon infants’ emotion learning. However, the relationship between parents’ anxiety and/or depressive symptoms and parents’ emotional expression and gaze during triadic parent-infant-person interactions has yet to be explored. Thus, this study will offer innovative insights into the effect of parents’ anxiety and depressive symptoms upon
8
parents’ emotional expression and gaze in the context of triadic parent-infant-person interactions. This is scientifically relevant as to our knowledge, parents’ anxiety and depression symptoms have not yet been investigated in such triadic SR situations (Aktar et al, 2015). Furthermore, the consideration of parents from the general population, as opposed to a clinical population, offers insights into the impact of exposure to positive and negative emotional expressions upon infants during SR interactions (Cicchetti, 2006). This study aims to investigate the relationship between parents’ anxiety and depressive symptoms upon parents’ emotional expression and gaze during a triadic parent-infant-person context.
The research questions are:
a) Do parents’ anxiety symptoms explain differences in parents’ emotional expression and gaze during triadic parent-infant-person interactions in response to happy, sad, angry and fearful (vs. neutral) faces?
b) Do parents’ depressive symptoms explain differences in parents’ emotional expression and gaze during triadic parent-infant-person interactions in response to happy, sad, angry and fearful (vs. neutral) faces?
c) Is parent gender associated with differences in parents’ emotional expression and gaze during triadic interactions?
d) Is infant gender associated with differences in parents’ emotional expression and gaze during triadic interactions?
9
The following results have been hypothesized:
a) Based upon studies revealing increased anxiety expression in parents with anxiety during SR contexts, we predicted that higher levels of parent anxiety symptoms would be associated with longer duration of negative emotion and increased gaze towards the infant during triadic interactions.
b) Based upon studies revealing more neutral/negative and less positive affect in depressed parents, we predicted that higher levels of parent depressive symptoms would be
associated with longer duration of neutral emotion and less gaze towards the infant during triadic interactions.
c) As the role of fathers in SR has rarely been considered, we wished to explore possible associations between parent gender and parent emotion and gaze during triadic
interactions.
d) Based upon inconsistent results regarding the effect of infant gender upon dyadic parent-infant interactions, we wish to explore possible associations between parent-infant gender and parents’ emotion and gaze during triadic interactions.
Method
Participants
The sample consisted of typically developing infants aged 12 months (N = 64), along with infants’ parents (Nmothers = 50 and Nfathers = 14). Table 1 presents the descriptive statistics of
participating infants and parents. The sample included 25 boys (39.1%) and 39 girls who ranged in age from 10.7 months to 13 months (M = 12.08 months, SD = .60). Parents ranged in age from
10
26 years to 62 years (M = 34.4, SD = 5.01). The majority of parents were of Dutch nationality (84.4%), however the sample also included parents from Brazil, Canada, Croatia, France, Germany, Italy, Poland, Romania, Slovakia, Surinam, United Kingdom and Venezuela. An additional three infants were tested but removed from the analysis due to fussiness or equipment failure. Parents were not required to have a clinical diagnosis of anxiety and/or depression to participate in the study. Instead, parents’ self-reported levels of anxiety and/or depressive symptoms were viewed on a continuum for the purpose of this study. Families were part of a larger sample recruited via invitation letters sent by the municipality to families who recently became parents. The study was approved by the ethics committee at the University of
Amsterdam. All participating parents received information detailing the purpose and procedure of the study and informed consent was obtained.
Parents’ education level varied from one (primary/ elementary education) to eight (university) with parents having a mean education of 7.43 (SD = 1.09). Parents’ gross monthly income (excluding their partner’s income) varied from one (<500 euro) to seven (>5000 euro) with a mean of 5.15 (SD = 1.98) Parent’s anxiety symptoms varied from 0 to 11 (M = 0.95, SD = 1.56) and parent’s depressive symptoms varied from 0 to 10 (M = 1.82, SD = 2.28).
Materials and Procedure
Stimuli
The stimuli were computerized video presentations of four unfamiliar models exhibiting neutral, happy, sad, fearful and angry facial expressions. The videos were taken from the
11
Doosje, 2011). Previous research suggests that such pre-recorded stimuli allow careful manipulation of the information presented to participants (Mumme & Fernald, 2003). Thus, video stimuli act as a reliable medium for well-controlled experiments with infants. Each experiment consisted of two blocks of four trials. Each block included two male and two female models of North-European descent between the ages of 18 and 25. The same four models
appeared again in the second block in a randomly generated order. Each trial began with a visual attention getter displayed at the centre of the screen. Attention getters consisted of moving images accompanied by a sound. At the beginning of each block, an experimenter monitored infants’ attention from another room, and repeated the attention getters when required. Following the attention getter, the experiment continued with the presentation of a neutral expression and progressed in a randomly generated order of (non-neutral) emotions. For the purpose of this study, only the first block of four trials was included in the analysis.
Procedure
Each infant and parent (mother or father) pair was observed in a triadic parent-infant-person context during a computer task where unfamiliar neutral, happy, sad, fearful and angry facial expressions were presented. Infants were secured in a car seat mounted on a table that was situated 60cm away from the computer screen. The parent sat on a chair adjacent to the infant and facing the computer screen during the task. Parents were asked to attend to the computer screen and not to intervene with their infant unless the infant sought attention or required
soothing. In these cases, parents were asked to interact with their infant as they normally would, but without using any toys. Figure 1 demonstrates the set-up of the experiment.
12
All observations were recorded using two cameras, one camera recording the face and upper body of the infant and parent, and one camera recording the facial expression on the screen. The experimenter was not present during the interaction. The state events of parents’ emotional expression and gaze were coded, frame-by-frame, during the first four blocks of each triadic interaction using the software Observer XT 11.0 (Noldus, 2009). Outcome variables included the duration (in seconds) of parent emotions. Parents’ emotions were coded as positive, negative or neutral and considered mutually exclusive (parents may only give one emotional expression at a time). Positive emotions (joy or happiness) refer to closed and open smiles identified by raising corners of the lips, constriction of the eyes, raising of the cheeks and opening of the mouth (Ekman & Friesen, 1978; Messinger, Fogel, & Dickson, 2001). Negative emotions (sad, fearful or angry) refer to frowns, sad and cry faces identified by lowering the brows, constriction of the eyes and opening of the mouth (Bolzani Dinehart, Messinger, Acosta, Cassel, Ambadar, & Cohn, 2005). The individual characteristics of the parent were taken into consideration while coding the emotions. Neutral emotions refer to moments when the parent did not show a positive or negative emotion. Additionally, the duration (in seconds) and direction of parent gaze (towards infant, towards screen or elsewhere) were observed.
Three trained observers coded the emotional expression and gaze of parents. Observers were blind to parents’ reported levels of anxiety and depressive symptoms in order to minimise bias. For reliability purposes, 20% of the total observations were coded by all three observers. The inter-observer reliability for parents’ emotional expression was Cohen’s kappa = 0.85 and the inter-observer reliability for parents’ gaze was Cohen’s kappa = 0.91.
13
Instruments
Depression Anxiety Stress Scales (DASS; Lovibond, & Lovibond, 1995). The DASS was
completed by each participating parent. The DASS is a 42-item self-report questionnaire, which evaluates the three related negative emotional states of depression, anxiety and stress.
Respondents rate the extent to which they have experienced each symptom in the past week on a 4-point severity/ frequency scale ranging from 0 (did not apply to me at all) to 3 (applied to me very much or most of the time). The depression scale assesses dysphoria, hopelessness,
devaluation of life, self-deprecation, lack of interest/involvement, anhedonia, and inertia. An example item of the DASS depression subscale is “I felt that I had lost interest in just about everything”. The anxiety scale assesses autonomic arousal, skeletal muscle effects, situational anxiety, and subjective experience of anxious affect. An example item of the anxiety subscale is “I felt scared without any good reason”, and an example item of the stress subscale is “I found that I was very irritable”. The DASS has been found to have good psychometric properties. In the current sample, the reliability of the DASS (Cronbach’s alpha) was 0.85 for depressive symptoms and 0.73 for anxiety symptoms.
Statistical Analyses
Observational data were available for all participants (N = 64). The first block of four trials was included in the analysis. Ten seconds of observational data was extracted for each of the 20 facial expressions. This included a five second duration during the apex of each
expression on the screen, as well as a five second duration after the decline of the apex of the expression. For four parents (6.25%), the total number of observations exceeded 20 facial
14
expressions on the screen. Additionally, for six parents (9.38%), the total number of observations did not reach 20 facial expressions. Thus, total number of observed screens was included as a covariate in the multilevel models in order to control for differences in the number of
observations for each parent. For mothers, two DASS depression symptom reports (4.2%) and two anxiety symptom reports (4.2%) were missing and for fathers, two depression reports (14.3%) and one anxiety report (7.14%) were missing.
Multilevel regression models were used to examine differences in the duration of parents’ emotional expression and gaze. Multilevel analysis was used to account for the nested data structure stemming from repeated observations. The structure of the multilevel model with the duration of parent emotion as the outcome included the repeated observations of parent emotion (positive, and negative vs. neutral) nested within repeated observations of the screen (positive, sad, angry, and fearful vs. neutral). The structure of the multilevel model with the duration of parent gaze as the outcome included the repeated observations of parent gaze (to screen and to child vs. elsewhere) nested within the repeated observations of the screen (positive, sad, angry and fearful vs. neutral). All continuous outcome and predictor variables were transformed into normal standard scores (M = 0, SD = 1). The significance of effects was evaluated at p ≤ .05.
Inspection of the distributions revealed sufficient normality (<2) for all variables, except for the following variables: parent gaze elsewhere during happy screen, parent gaze to video and to child during sad screen, parent gaze to video and child during angry screen, parent gaze to video and to child during fearful screen, parent negative expression during sad screen, parent anxiety, parent age and parent education. Several outliers with Z-scores larger than (-) 3.29 were identified within these variables (Tabachnick & Fidell, 2001). Outliers (> 2.5 SD)
15
were replaced by values 0.1 unit beyond the next most extreme non-outlying value in the distributions.
Parent emotional expression and gaze were analyzed within separate fixed-effects multilevel models using the maximum likelihood as the estimation method. Data were analyzed with IBM SPSS Statistics Version 21. The duration of parent’s emotion was first analyzed with a main effects model including the main effects of screen (happy, sad, fearful and angry vs.
neutral) and parent emotion (positive and negative vs. neutral), parent age, parent gender, child gender, parent depression symptoms and parent anxiety symptoms. Theoretically relevant
interactions were added one-by-one to the model and significant interaction terms were retained.
First, the interactions between screen and parent emotion/gaze were included in the model, to investigate if parents’ emotions/gaze differed as a function of the emotion on the screen. Next, the interactions between parent gender and screen, followed by parent gender and emotion/gaze were added in order to investigate whether the link between the duration of parent emotion/gaze and the screen or parent emotion/gaze differed between mothers and fathers. This was followed by a test of the same interactions, but with child gender included. This allowed the investigation of the link between the duration of parents’ emotion/gaze and the screen or parent emotion/gaze as a function of child gender (girls v. boys). Finally, to test differences in the duration of parent emotion/gaze as a function of parents’ anxiety and depressive symptoms, the two-way interactions between parent anxiety and depressive symptoms with screen, followed by emotion/gaze were included.
16
Results
Preliminary Analyses
The associations between the continuous predictors (parents’ age, education, income, depressive and anxiety symptoms) and parent emotion/gaze were first examined via zero-order correlations. There were significant correlations between parents’ age and income (r = .376, p = .007). Parents’ depression and anxiety levels were also significantly correlated (r = .340, p = .008). Among the correlations between the continuous predictors and the outcome variables, parents’ education was significantly correlated with parents’ neutral emotion during the sad screen (r = .322, p = .014), and with neutral emotion during the fearful screen (r = .443, p = .000). There was also a significant correlation between parent depressive symptoms and gaze elsewhere during the fearful screen (r = .456, p = .033). Parent anxiety symptoms were not significantly correlated with parent emotion and gaze.
Main Analyses
Table 3 displays the results of the multilevel analyses with the duration of parents’ emotional expressions and gaze as the outcome. First level units were the repeated observations of parent emotions (positive and negative vs. neutral) and gaze (to screen, to child vs. elsewhere), second level units were the happy, sad, fearful, angry (vs. neutral) screens.
17
Duration of parents’ emotional expressions
Initially, a main-effects model for parent emotion was fit (-2log = 1251.253, AIC = 1271.253, BIC = 1315.694). This model explained 27% of the variance for the duration of parent emotional expression. The aforementioned interaction terms were then added one-by-one. Among theoretically relevant interactions, the interaction between the happy screen and parent positive emotion was significant (β = .89, SE = .15, p = .000) and improved the model fit (-2log = 1197.336, AIC = 1233.336, BIC = 1313.331). This interaction revealed that parents were significantly more positive (vs. neutral) during the happy screen. The addition of the interaction between parent depressive symptoms and screen improved the model fit (-2log = 1121.563, AIC = 1173.563, BIC = 1287.314), and was significant between depressive symptoms and the fearful screen (β = .89, SE = .04, p = .010). Higher levels of depressive symptoms were associated with increased neutral emotion during the fearful screen. The final model for the duration of parent emotional expression, eliminating the non-significant effects, is presented in table 2a. The final model explained 31% of the variance for the duration of parent emotional expression.
Duration of parent gaze
First, main-effects only model for parent gaze was fit (-2log = 1718.481, AIC = 1738.481,
BIC = 1784.534). The main-effects model explained 23% of the variance for parent gaze. The
interaction terms were then added step-by-step. Among theoretically relevant interactions, the interaction between screen and parent gaze was significant (β = -.93, SE = .25, p = .000), and significantly improved the model fit (-2log = 1685.300, AIC = 1721.300, BIC = 1804.195). This revealed that parents were less likely to gaze at the screen (vs. elsewhere) during the presentation
18
of happy (vs. neutral) expressions on the screen. Second, the inclusion of the interaction between parent gender and screen significantly improved the model fit (-2log = 1679.244, AIC =
1725.244, BIC = 1831.166) and was significant between parent gender and angry screen (β = .47, SE = .20, p = .016). This suggests that mothers and fathers differed significantly in their duration of gaze during the angry screen. Specifically, fathers had an increased duration of gaze elsewhere during the angry screen, compared to mothers who gazed more often at the child. Following the addition of the interaction between parent depressive symptoms and screen, a trend towards significance was identified in the interaction between depressive symptoms and the fearful screen (p = .071), however this term did not reach significance. The model significantly improved (-2log = 1576.085, AIC = 1636.085, BIC = 1771.968), however inclusion of this interaction did not significantly improve the model fit of subsequent models or yield additional significant interaction terms. Thus, this interaction was removed from the model. The final, parsimonious model for the duration of parent gaze, eliminating the non-significant effects, is presented in table 2b. The final model explained 27% of the variance for the duration of parent emotion.
Discussion
In this study, parents’ emotional expressions and gaze were observed during
computerized presentations of happy, sad, angry, fearful and neutral facial expressions. Parents’ self-reported anxiety and depressive symptoms were measured using the DASS. The key
findings can be summarized as follows; (a) Parents were significantly more positive (vs. neutral) and were less likely to gaze at the screen (vs. elsewhere) during the happy (vs. neutral)
19
expressions. (b) Mothers and fathers expressed similar levels of positive and negative (vs. neutral) emotions. However, mothers and fathers differed significantly in the duration of their gaze during the angry screen. Specifically, fathers gazed elsewhere for a longer duration,
compared to mothers who gazed at the child for a longer duration. (c) Parents’ emotions or gaze did not differ significantly as a function of child gender. (d) Higher levels of depressive
symptoms in mothers and fathers were associated with more neutral emotions during the fearful screen. However, depressive symptoms did not predict parents’ gaze. (e) Parents’ anxiety symptoms did not predict parents’ emotional expression or gaze.
The results lend partial support to the identified hypotheses. The first hypothesis predicted that higher levels of parent anxiety symptoms would be associated with increased negative emotional expression and gaze towards the infant. The findings suggest that parent anxiety symptoms did not predict parents’ emotional expression or gaze. Previous research suggests that children’s learning of anxiety may be partly attributed to observational learning of parental anxiety (Fisak & Grills-Taquechel, 2007; Murray, Creswell, & Cooper, 2009), as infants of anxious parents are more likely to be provided with expressions of parental anxiety (Muris, Steerneman, Merckelbach, & Meesters, 1996). The current findings suggest that parents experiencing more anxiety symptoms do not differ from parents without anxiety symptoms in their emotional expressions or gaze during triadic interactions. This may be attributed to the fact that participating parents did not have clinical diagnoses of anxiety. Thus, these findings provide preliminary support for the idea that expressed anxiety during triadic interactions does not differ between parents experiencing higher levels of anxiety symptoms, in comparison to parents who do not experience anxiety symptoms. Although it is known that parental anxiety may impair parents’ ability to positively reinforce their child’s reaction to novelty (Fisak & Grills-Taquechel,
20
2007; Murray et al., 2009), it may be suggested that sub-clinical levels of anxiety symptoms may not impair parents’ ability to offer positive reinforcement in the face of novelty during triadic interactions. It is important to acknowledge that the current findings may be attributed to differences in the current triadic context in comparison to previous SR contexts. It may be suggested that strangers’ facial expressions in the current study may not be considered as ambiguous as alternative ambiguous stimuli such as novel objects or real-life strangers.
The second hypothesis was that higher levels of parent depressive symptoms would be associated with longer duration of neutral emotions and less gaze towards the infant. This hypothesis was only supported during the presentation of fearful expressions on the screen. That is, higher levels of depressive symptoms were associated with a longer duration of neutral emotions during the fearful screen only. Longer duration of neutral emotions for parents with more depressive symptoms is in alignment with the suggestions of Cohn, Campbell, Matias, & Hopkins (1990) who acknowledged that flat affect characterizes the interactions of most mothers experiencing symptoms of depression (specifically post-partum depression). Thus, parents’ depressive symptoms may have a salient impact upon parents’ SR during triadic interactions, particularly in response to fearful expressions. However, in this study, depressive symptoms did not predict parents’ gaze. Taken together, these findings suggest that depressive symptoms may impact upon parents’ emotional expressions, but not gaze, in response to fearful stimuli during triadic interactions.
The third hypothesis explored associations between parent gender and parent emotion and gaze. It was found that overall; parent emotional expression did not differ significantly between mothers and fathers. These findings offer a preliminary insight into the relatively unknown role of fathers during SR (Aktar et al., 2013). This study found that fathers gazed
21
elsewhere for a longer duration during the angry screen, compared to mothers who gazed at the child for a longer duration. This may suggest that fathers aim to divert the child’s attention elsewhere when faced with ambiguous stimuli, specifically angry facial expressions. It may be proposed that this response by fathers may aim to protect their child from adverse reactions when faced with ambiguous stimuli. This holds significant importance, as recent models exploring the development of anxiety in infants (Bögels & Perotti, 2011; Bögels & Phares, 2008), inferred that fathers have a central role in the development of anxiety based upon the evolutionary role of fathers in their exploration of the external world.
The final hypothesis investigated the associations between infant gender and parent emotional expression and gaze. It was found that parents’ duration of emotions and gaze did not differ between daughters vs. sons. This finding is consistent with previous studies of face-to-face interactions, which also failed to find significant infant gender effects upon parents’ emotions and gaze (e.g., Cohn & Tronick, 1987; Kaye & Fogel, 1980). This finding suggests that parents may not adjust their emotional expressions and gaze during triadic interactions according to the sex of their child. However, these findings are in contrast to the findings of Cohn, Campbell, Matias, & Hopkins (1990), who found that depressed mothers of boys, although not of girls, were less positive during face-to-face interactions.
Limitations
Due to the cross-sectional nature of the data, causal inferences cannot be assumed
between parent’s anxiety and depressive symptoms and parent’s observed emotional expressions and gaze. Further longitudinal studies investigating parent’s emotion and gaze in parents with reported anxiety and/or depressive symptoms during triadic interactions would lead to an
22
increased ability to make causal inferences. Additionally, it is important to acknowledge that parent’s anxiety and depressive symptoms were self-reported, and were not indicative of parents with clinical diagnoses of anxiety and/or depression. Therefore, the findings of this study may not be generalized to the wider population of parents with clinical levels of anxiety and/or depression. The total number of participating fathers in this study was small (N = 14). Thus, there may be limited power to identify effects for fathers within this study. Finally, as 84.4% of the current sample were Dutch in nationality, further research with a more representative sample would be required to clarify the generalizability of these results.
Implications
In conclusion, the current study found that parents were significantly more positive (vs. neutral) and were less likely to gaze at the screen (vs. elsewhere) in response to happy (vs. neutral) expressions on the screen. The findings additionally suggest that fathers had a
significantly increased duration of gaze elsewhere during the angry screen compared to mothers. This study also provided evidence of an association between higher levels of depressive
symptoms and more neutral emotions in response to fearful facial expressions on a screen. These results offer further understanding into parents’ emotional expression and gaze during everyday triadic interactions with their infant. The findings illustrate the presence of parents’ positive social referencing cues in response to happy (vs. neutral) facial expressions. For parents experiencing depressive symptoms, the results indicate that future intervention should place an emphasis upon parents’ awareness of the potential impact of depressive symptoms upon their social referencing cues during triadic interactions with their infant. Finally, the current findings regarding infant gender effects upon parent emotional expression and gaze are in contrast to
23
some previous findings. This suggests that further research into the effects of infant gender upon parent emotional expression and gaze during triadic interactions would be beneficial.
24
References
Aktar, E., & Bögels, S. M. (In preparation). Exposure to parents’ negative emotions as a developmental pathway in the family aggregation of depression and anxiety in the first year of life. Clinical Child and Family Psychology Review.
Aktar, E., Majdandžić, M., de Vente, W., & Bögels, S. M. (2013). The interplay between
expressed parental anxiety and infant behavioral inhibition predicts infant avoidance in a social referencing paradigm. Journal of Child Psychology and Psychiatry, 54, 144-156.
Aktar, E., Majdandžić, M., Vente, W., & Bögels, S. M. (2014). Parental social anxiety disorder prospectively predicts toddlers' fear/avoidance in a social referencing paradigm. Journal of Child Psychology and Psychiatry, 55, 77-87.
Aktar, E., Mandell, D., de Vente, W., Majdandzic, M., Raijmakers, M.E.J., & Bögels, S. M. (under review). Infants’ temperament and mothers’, and fathers’ depression predict infants’ attention to objects paired with emotional faces. Journal of Abnormal Child
Psychology.
Als, H., Tronick, E., & Brazelton, T. B. (1979). Analysis of face-to-face interaction in infant-adult dyads. In M. E. Lamb, S. J. Suomi, & G. R. Stephenson (Eds), Social
25
interaction analysis: Methodological Issues, (pp. 33-77). Madison: University of
Wisconsin.
Bögels, S.M., & Perotti, E.C. (2011). Does father know best? A formal model of the paternal influence on childhood social anxiety. Journal of Child and Family Studies, 20, 171-181.
Bögels, S.M., & Phares, V. (2008). Fathers’ role in the etiology, prevention and treatment of child anxiety: A review and a new model. Clinical Psychology Review, 28, 539-558.
Bögels, S.M., & Van Melick, M. (2004). The relationship between child reports, parent self-report, and partner report of perceived parental rearing behaviours and anxiety in children and parents. Personality and Individual Differences, 37, 1583.
Bolzani Dinehart, L. H. B., Messinger, D. S., Acosta, S. I., Cassel, T., Ambadar, Z. and Cohn, J. (2005), Adult perceptions of positive and negative infant emotional expressions. Infancy,
8, 279–303.
Cohn, J. F., Campbell, S.B., Matias, R., & Hopkins, J. (1990). Face-to-face interactions of postpartum depressed and nondepressed mother-infant pairs at 2 months. Developmental
26
Cohn, J. F., & Tronick, E. Z. (1987). Mother-infant face-to-face interaction: The sequence of dyadic states at 3, 6, and 9 months. Developmental Psychology, 23, 68-77.
Dawson, G., Ashman, S.B., Panagiotides, H., Hessel, D., Self, J., Yamada, E., & Embry, L. (2003). Child Development, 74, 1158-1175.
De Haan, M., Belsky, J., Reid, V., Volein, A., & Johnson, M. H. (2004). Maternal personality and infants’ neural and visual responsivity to facial expressions of emotion. Journal of
Child Psychology and Psychiatry, 45, 1209.
Ekman, P., & Friesen, W.V. (1978). Facial action coding system. Palo Alto: Consulting Psychologist Press.
Emde, R.N. (1992). Social referencing research: Uncertainty, self and search for meaning. In Feinman, S. (Ed.), Social referencing and the social construction of reality in infancy (pp 79-95). New York: Plenum Press.
Feinman, S., Roberts, D., Hsieh, K. F., Sawyer, D., & Swanson, D. (1992). A critical review of social referencing in infancy. In S. Feinman (Ed.), Social referencing and the social
27
Field, T., Diego, M., & Hernanez-Reif, M. (2009). Depressed mothers’ infants are less responsive to faces and voices. Infant Behaviour and Development, 32, 239-244.
Field, T., Pickens, J., Fox, N. A., Gonzalez, J., & Nawrocki, T. (1998). Facial expression and EEG responses to happy and sad faces/ voices by 3-month-old infants of depressed mothers. British Journal of Developmental Psychology, 16, 485.
Fisak, B.J.R., & Grills-Taquechel, E.A. (2007). Parental modeling, reinforcement, and information transfer: Risk factors in the development of child anxiety? Clinical Child and
Family Psychology, 10, 213–231.
Hart, S., Field, T., Del Valle, C., & Peláez-Nogueras, M. (1998). Depressed mothers’ interactions with their one year old infants. Infant behaviour and Development, 21, 519-525.
Haviland, J. M. (1977). Sex-related pragmatics in infants’ nonverbal communication.
Communication in Infancy and Early Childhood, 27, 80-84.
Kaye, K., & Fogel, A. (1980). The temporal structure of face-to-face communication between mothers and infants. Developmental Psychology, 16, 454-464.
28
Langner, O., Dotsch, R., Bijlstra, G., Wigboldus, D.H.J., Hawk, S.T., & Van Knippenberg, A. (2010). Presentation and validation of the radboud faces database. Cognition & Emotion,
24, 1377.
Lovibond, S.H. & Lovibond, P.F. (1995). Manual for the Depression Anxiety Stress Scales.
(2nd. Ed.) Sydney: Psychology Foundation. ISBN 7334-1423-0.
Messinger, D.S., Fogel, A., & Dickson, K. L. (2001). All smiles are positive, but some smiles are more positive than others. Developmental Psychology, 37, 642-653.
Moses, L. J., Baldwin, D. A., Rosicky, J. G., & Tidball, G. (2001). Evidence for referential understanding in the emotions domain at twelve and eighteen months. Child
Development, 72, 718.
Mumme, D. L., & Fernald, A. (2003). The infant as onlooker: Learning from emotional reactions observed in a television scenario. Child Development, 74, 221.
Muris, P., Steerneman, P., Merckelbach, H., & Meesters, C. (1996). The role of parental fearfulness and modeling in children’s fear. Behavior Research Therapy, 34, 265-268.
29
Murray, L., Arteche, A., Fearon, P., Halligan, S., Goodyer, I., & Cooper, P. (2011). Maternal postnatal depression and the development of depression in offspring up to 16 years of age. Journal of the American Academy of Child & Adolescent Psychiatry, 50, 460-470.
Murray, L., Cooper, P., Creswell, C., Schofield, E., & Sack, C. (2007). The effects of maternal social phobia on mother-infant interactions and infant social responsiveness. Journal of
Child Psychology and Psychiatry, 48, 45-52.
Murray, L., Creswell, C., & Cooper, P. J. (2009). The development of anxiety disorders in childhood: an integrative review. Psychological Medicine, 39, 1413-1423.
Murray, L., De Rosnay, M., Pearson, J., Bergeron, C., Schofield, E., Royal-Lawson, M., & Cooper, P. J. (2008). Intergenerational transmission of social anxiety: the role of social referencing processes in infancy. Child Development, 79, 1049.
Peláez, M., Virues-Ortega, J., Field, T.M., Amir-Kiaei, Y., and Schnerch, G. (2013). Social referencing in infants of mothers with symptoms of depression. Infant Behaviour and
Development, 36, 548-556.
30
Sorce, F. F., Emde, R. N., Campos, J. J. & Klinnert, M. D. (1985). Maternal emotional signalling: its effect on the visual cliff behavior of 1-year-olds. Developmental
Psychology. 21, 195–200.
Tabachnik, B.G. & Fidell, L.S. (2001). Using Multivariate Statistics. New York: Harper & Row.
Turner, S. M., Beidel, D. C., & Costello, A. (1987). Psychopathology in the offspring of anxiety disorders patients. Journal of Consulting and Clinical Psychology, 55, 229-235.
Van der Schalk, J., Hawk, S. T., Fischer, A. H., & Doosje, B. J. (2011). Moving faces, looking places: The Amsterdam Dynamic Facial Expressions Set (ADFES), Emotion, 11, 907-920. DOI: 10.1037/a0023853.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: the PANAS scales. Journal of Personality
and Social Psychology, 54, 1063–1070.
Weissman, M.M., Leckman, J.F., Merikangas, K.R., Gammon, G.D., & Prusoff, B.A. (1984). Depression and anxiety disorders in parents and children. Archives of General
31
32
Table 1. Sociodemographic Characteristics (Means, standard deviations and range)
Mothers Fathers Girls Boys
(n = 50) (n = 14) (n = 39) (n = 25) ______________________________________________________________________________ Age 33.56 (4.50, 27-43) 37.75 (9.26, 26-62) 12.07 (.60) 12.09 (.61) Dutch Origin 33 (66%) 9 (64.3%) Parent Educationa 7.43 (1.08, 3-8) 7.33 (1.15, 4-8) Monthly Incomeb 4.53 (1.26, 2-7) 4.45 (1.69, 1-7) Anxiety 1.04 (1.73, 0-9) 0.62 (0.65, 0-2) Depression 1.67 (2.13, 0-8) 2.42 (2.84, 0-10)
Notes. Parent age measured in years and child age measured in months.
a
Measured with an 8-point scale from 1 (primary/elementary education) to 8 (university). b
Excluding partner’s income. Measured with a 7-point scale from 1 (<500 euro/month) to 7 (>5000 euro/month).
33
Table 2. Multilevel regressions of the percentage duration of parent emotion (2a) and percentage
duration of parent gaze (2b).
2a 2b
β SE P β SE P
Intercept 2.07 .36 < .001 Intercept 0.46 .36 .202
Parent gender 0.05 .25 .843 Parent gender -0.17 .23 .475
Child gender -0.08 .20 .706 Child gender -0.02 .15 .921
Total Screens -0.29 .07 < .001 Total Screens -0.21 .07 .002
Depression -0.07 .05 .193 Depression -0.02 .04 .546 Anxiety -0.03 .08 .680 Anxiety 0.02 .06 .687 Happy -0.71 .12 < .001 Happy 0.22 .26 .392 Sad -0.20 .12 .091 Sad -0.22 .25 .392 Angry -0.37 .12 < .050 Angry -0.19 .27 .473 Fearful -0.43 .12 < .001 Fearful -0.04 .25 .869
Positive -0.83 .11 < .001 Gaze video 1.32 .21 < .001
Negative -1.09 .29 < .001 Gaze child 0.06 .21 .781
Happy * positive 0.89 .15 < .001 Happy * gaze screen -1.02 .28 < .001 Sad * positive 0.09 .16 .572 Sad * gaze screen -0.26 .28 .357 Angry * positive 0.14 .16 .375 Angry * gaze screen -0.33 .29 .259 Fearful * positive -0.05 .15 .763 Fearful * gaze screen -0.44 .27 .107 Happy * negative 0.29 .41 .482 Happy * gaze child -0.07 .28 .817 Sad * negative 0.23 .32 .469 Sad * gaze child 0.11 .28 .702 Angry * negative 0.21 .32 .514 Angry * gaze child 0.23 .29 .417 Fearful * negative 0.24 .32 .455 Fearful * gaze child 0.19 .27 .493 Happy * depression 0.04 .03 .194 Happy * depression 0.36 .21 .089
34
Sad * depression -0.01 .03 .722 Sad * depression 0.33 .21 .122 Angry * depression 0.02 .03 .500 Angry * depression 0.59 .21 .006 Fearful * depression 0.09 .03 < .010 Fearful * depression 0.29 .21 .170
Notes: Positive = positive (vs. neutral) parent emotion, negative = negative (vs. neutral) parent
emotion.
Happy, sad, angry, fearful = happy screen, sad screen, angry screen and fearful screen (vs. neutral screen).
