Depression in Childhood and Early Adolescence: Parental Expressed Emotion and Family Functioning

Special Article - Depression among Children and Adolescents

Ann Depress Anxiety. 2015; 2(7): 1070.

Depression in Childhood and Early Adolescence: Parental Expressed Emotion and Family Functioning

Tompson MC¹*, O Connor EE¹, Kemp GN¹, Langer DA¹ and Asarnow JR²

1Department of Psychological and Brain Sciences, Boston University, USA

2Department of Psychiatry and Biobehavioral Sciences, University of California Los Angeles, USA

*Corresponding author: Tompson MC, Department of Psychological and Brain Sciences, Boston University, 648 Beacon Street, 4th Floor, Boston, USA

Received: September 21, 2015; Accepted: December 04, 2015; Published: December 10, 2015

Abstract

Across development depression is associated with impairments in interpersonal and family functioning. In turn, these impairments may predict a more negative depression course and outcome. This study examined family functioning and parental Expressed Emotion (EE) among depressed youth during middle childhood and early adolescence and their relationship to demographic and clinical factors. Data were drawn from pretreatment evaluations of 132 depressed youth ages 7-14 and their families enrolled in a randomized clinical trial comparing family to individual treatment for youth depressive disorders. Families completed semi-structured diagnostic interviews, self-report measures of family functioning, and the Five Minute Speech Sample EE measure. High parental EE was more common in one-parent, as opposed to two-parent families, and early adolescent youth were more likely than pre-adolescent youth to have high critical EE parents. Severity and chronicity of child depression, child comorbidity, functional impairment, and maternal depressive symptoms were not associated with parental EE. Parental high EE overall and critical EE in particular were associated with reports of higher conflict and lower cohesion by both parents and children when compared to low parental EE. Similar patterns of associations were evident for youth across pre-adolescent and early adolescent developmental periods. Single parent status may be an indicator of greater family stress; and higher levels of critical EE may reflect the higher levels of parent-child conflict characteristic of the transition from late childhood to early adolescence. Among youth with depression parental EE appears to reflect potentially important impairments in family functioning.

Keywords: Depression; Youth; Family; Family functioning; Expressed emotion

Introduction

Depression is relatively rare prior to adolescence with 1-3% of children suffering from Major Depressive Disorder (MDD) and slightly higher rates of children meeting criteria for dysthymic (DD) and minor depressive disorders (for review, see) [1]. Prevalence increases in early adolescence, and by age 18 approximately 20 percent of youth in the U.S. are likely to have suffered from a depressive episode [2]. Early onset depression is characterized by high levels of chronicity and severity [3], relapse and social impairment [4-6], increased risk for subsequent manic episodes [7], drug/alcohol abuse, suicide [8], and recurrence in adulthood [9]. By understanding factors contributing to the development and maintenance of depression in childhood, increasingly effective treatments can be developed and preventive efforts undertaken.

Given the centrality of the family in the developmental context of middle and late childhood, understanding the impact of family functioning on youth depression may be particularly important during this developmental period [10]. Using a variety of methods, a number of family characteristics have been examined in relation to youth depression. In studies utilizing self-report measures, depressed children and their parents report high levels of stress and negative life events [11,12], low family cohesion [13,14], high levels of coercion, control [15], conflict [13,16], and parental rejection/ hostility [17]. Some of these studies suggest that difficulties in family functioning may precede symptoms and increase risk for symptom exacerbation [18]. However, the relationship between depression and family functioning is likely bidirectional. Depressive symptoms may contribute to lower parent-child relationship quality [19] and higher parent–child conflict [20], fueling family stress and contributing to problems with family functioning.

Parental Expressed Emotion (EE) - an index of the degree to which parents express critical (CRIT) and/or Emotionally - Over Involved (EOI) attitudes during an individual interview or speech sample-has also been examined among youth with depression. EE is thought to be an index of the relational quality of family members and may contribute to more negative clinical outcomes through increasing stress on patients [21,22]. High EE ratings suggest not merely a greater number of negative descriptive statements but greater negative emotional reaction on the part of the parent. High family EE predicts poorer outcomes in a variety of psychiatric disorders in adults, particularly among individuals with mood disorders [21]. Although statistically significant relationships between EE ratings and clinical characteristics (such as symptom severity) are not consistently found by researchers using cross-sectional analyses, EE ratings may be important in prospectively predicting relapse and clinical course [21,22].

Among depressed youth, high parental CRIT has been found to prospectively predict a) relapse and/or non-recovery in 7-14 year olds in the year following inpatient hospitalization [23], b) lower social functioning and higher persistence of depression among outpatient youth with depression [24], and c) more depressive episodes among depressed youth and youth at risk due to family history [25]. Overall, these early data suggest that high parental EE, particularly CRIT, may be associated with a more negative course among youth with depressive disorders. Although rates of high parental overall EE and CRIT may be higher among youth with depression compared to normal controls and youth with other forms of psychopathology [26,27], not all youth with depression have high EE or high CRIT parents. By identifying factors associated with risk, we can more specifically target both acute treatment and prevention of recurrence. In a study of 7-14 year old depressed inpatients, the likelihood of high parental EE, particularly CRIT, was found to be associated with chronicity (greater than one year) of illness and comorbidity with disruptive behavior disorders [26]. In a study of outpatient depressed youth, overall functional impairment was found to be a general risk factor for high parental EE [27].

Another possible risk factor for high EE may be parental psychopathology, although the studies are mixed in this regard. Extensive literature documents the negative impact of maternal depression on child psychopathology, as well as maternal interactional behavior and parenting [28], and it is possible that maternal depression could be linked to higher levels of parental EE [29]. Although parental psychopathology has been found to be unrelated to parental EE among parents of adult bipolar patients [30], in a study of preadolescent youth, mothers with a history of depression were found to have twice the rate of high overall EE and high CRIT specifically than were mothers without a history of depression [31]. Similarly, in a study of youth with behavior problems, current maternal depressive symptoms were associated with higher maternal CRIT [29].

Age and gender may predict depressive symptoms and disorders, and careful attention must be paid to the possible role of both factors in moderating the relationship between family factors and youth depression. During the developmental shift from late childhood to early adolescence, normative reductions in positive emotions, increases in negative emotions [32] and intensification in parentchild negativity and conflict [33] are evident. Complicating these developmental trends, gender differences in depression emerge between 12 and 14 years of age when rates of depression for girls rise more steeply than those for boys [1]. These important changes have potentially strong implications for family interactions and relationships.

Although EE is thought to be a measure of the family environment, few studies have examined its association with family members’ own perceptions of their family’s functioning. These few do suggest that high EE relatives are perceived of as more critical than low EE relatives among adults with schizophrenia [34] and depression [21]. However, the relationship between observed EE and family members’ self-reported measures of family functioning has rarely been assessed. Understanding the association between parental EE and self-report indices of family functioning in depressed youth would assist in providing information on the most appropriate targets for interventions aimed at enhancing parent-child relationships and support. Given the emphasis of EE measures on assessing relative criticism, we would anticipate that high parental EE, particularly CRIT, would be associated with perceptions of greater conflict and less family cohesion among family members.

Current study

In this study we sought to understand the association between parental EE, child clinical characteristics, maternal depressive symptoms, and family functioning among pre- and early-adolescent youth with depressive disorders. First, we examined the association between parental EE and demographic and clinical characteristics. We hypothesized that, compared to low parental EE, high parental EE, particularly CRIT, would be associated with chronicity, functional impairment, externalizing comorbidity and maternal depressive symptoms. Second, we examined differences between children (ages 7-11) and early adolescence (ages 12-14) and between boys and girls. We hypothesized that the early adolescent group would have higher rates of high parental EE and more negative family functioning than the childhood group; however, we did not anticipate gender differences in parental EE. Third, we examined the association between parental EE and indices of family functioning. We hypothesized that, compared to low parental EE, high parental EE, particularly CRIT, would be associated with more negative family functioning, including higher conflict and lower cohesion, as rated by mothers, fathers and youth.

Participants

Participants were drawn from 134 families with a child between ages 7 and 14 who were enrolled in a two-site randomized clinical trial [35] comparing two treatment models for youth depression. Of the 134 families, 2 families did not have EE data; these were due to audio- recording difficulties or refusal of audio-recording altogether. Thus, the total sample for the current study comprised 132 families who had at least one parent with complete EE measures at the baseline evaluation.

Children were eligible if they met the following inclusion criteria: (a) DSM-IV-TR [36] diagnosis of current MDD, DD, Double Depression (MDD + DD), or Depression Not Otherwise Specified, (b) ages 7-14; (c) living with at least one parent or parental figure willing to participate in the evaluation and treatment sessions; and (d) parents and youth able and willing to provide informed consent (assent).Exclusion criteria included (a) thought or other disturbance that would interfere with the ability to benefit from the intervention and participate in treatment or assessments (e.g., psychotic disorder, pervasive developmental disorder, Tourette’s syndrome, severe Obsessive Compulsive Disorder, active substance abuse/dependence, mental retardation), (b) severe conduct disorders that threatened the stability of the home environment (e.g. youth with recent arrests and/ or juvenile justice or children’s protective service involvement) due to the potential impact on treatment implementation; and (c) youth or primary caregivers did not speak English.

Procedures

Parents and children reviewed and signed informed consent (assent) documents and were then interviewed and completed self-report measures separately. All procedures were approved by Institutional Review Boards at Boston University and UCLA.

Child diagnosis and symptom evaluation

Categorical DSM-IV-TR diagnoses were made based on information derived from the Schedule for Affective Disorders and Schizophrenia for School-Aged Children-Present and Lifetime Versions (K-SADS-PL) [37] administered to the parent about the child and to the child about him/herself. Diagnoses evaluated included depressive disorders (MDD, DD, Double Depression, DDNOS), anxiety disorders, elimination disorders, disruptive behavior disorders, tic disorders and substance use disorders. As in past studies [23], chronicity was categorized into chronic when depressive disorder criteria had been met for over one year and nonchronic when depressive disorder criteria had been met for less than one year. Estimates of inter-rater agreement based on 60 cases independently rated by two diagnosticians indicated excellent reliability for MDD (kappa = 0.95) and any depression diagnosis (kappa = 0.91). Diagnosis for comorbid disorders also demonstrated adequate reliability, including disruptive behavior disorders (kappa = 0.76), anxiety disorders (kappa = 0.77) and elimination disorders (kappa = 1.00). Depression severity and overall functional impairment were assessed dimensionally. Depression severity over the past two weeks was measured using the 17-item, clinician-rated Children’s Depression Rating Scale-Revised (CDRS-R) [38], and inter-rater reliability was excellent for total scores (n = 57; ICC = 0.94).Overall functional impairment was evaluated using the Children’s Global Adjustment Scale (C-GAS) [39] and was based on all available information from K-SADS-PL and CDRS-R. Inter-rater agreement was adequate for C-GAS (n = 57, ICC = 0.77).

Parental expressed emotion

The Five Minute Speech Sample [40] was used to assess the parents’ EE toward the child. It was administered individually to each parent, who was prompted to talk for five minutes without interruptions, describing “what kind of person [child] is and how you get along together.”Speech samples were audio-recorded, transcribed, and scored. Parents were rated high in EE by scoring high on CRIT, EOI, or both. Scoring criteria for a CRIT rating are: (1) a negative initial statement, (2) an overall indication of a negative child-parent relationship, or (3) one or more critical comments. Scoring criteria for a high EOI rating are: (1) evidence of self-sacrificing/overprotective behavior; (2) an emotional display, such as crying; or (3) at least two of the following: (a) excessive focus on the child’s early life, (b) one or more statements of idealized love or willingness to do anything for the child, and (c) five or more positive remarks. Those parents who do not display behaviors characterized by either CRIT or EOI categories are considered low EE. Reliability of Five Minute Speech Sample ratings as assessed on 39 co-rated samples(30%) was adequate for overall EE (kappa = 0.74), CRIT (kappa = 0.79) and EOI (kappa = 0.66).

Measures

Self-reports of family functioning

Two measures of family functioning were used. First, the 20- item Conflict Behavior Questionnaire (CBQ) [41] was used to assess the youth’s perspective on levels of conflict and negative communication within the family. Internal consistency was excellent on the CBQ (Cronbach’s alpha = 0.91). Second, two subscales of the Family Environment Scale [42] - 9 items each - were used to assess participants’ perspectives of the family environment. The family cohesion scale assesses commitment to family members, help-giving behavior, and supportiveness (e.g., “in our family members really help and support one another,” “we really get along well with each other”). The family conflict scale assesses conflicts, anger, and disagreements (e.g., “we fight a lot in our family,” “sometimes family members get so angry they throw things”). Internal consistency was adequate for cohesion (Cronbach’s alphas = 0.78 parents; 0.67 children) and for conflict (Cronbach’s alphas = 0.78 parents; 0.71 children).

Maternal depressive symptoms

The total score (range 0-63) on the Beck Depression Inventory (BDI-II) [43] was used to assess current maternal depressive symptoms. Internal consistency in this sample was high (Cronbach’s alpha = 0.99).

Results

Sample demographic and clinical characteristics

(Table 1) presents descriptive information on participants overall and by high versus low parental CRIT. Children ranged in age from 7 to 14 years (M = 10.63; SD = 2.08); 56% were girls and 44% were boys; a range of ethnic groups participated. The majority of families were two-parent (n = 79; 60%) of which 65 had two biological parents, two had adoptive parents, and 12 had a biological parent plus a stepparent. There were 51 (38.5%) one-parent families, and 2 (1.5%) with other family structures (one with grandparent and one shifting regularly between parental households). Approximately 18% of families reported annual income of less than $30,000; 37% reported annual income between $30,000 and $75,000; and 45% reported annual income exceeding $75,000. At some time in their lives, 53 (40%)received public assistance (e.g., Food stamps, WIC, Medicaid). Although parental education ranged from less than high school completion to graduate degree attainment, both mothers and fathers typically had around 2 years of college coursework. Across the sample 49 (37%) children had at least one anxiety disorders with many showing multiple anxiety diagnoses, including separation anxiety disorder (SAD; n = 16), simple phobias (n = 10), social phobias (n = 19), generalized anxiety disorder (n = 29), and post-traumatic stress disorder (n = 2). There were also a number of children with elimination disorders (n = 11; 8%) and ADHD (n = 40; 30%). There were no eating or substance use disorders in this sample.

Of the full sample, 34% (42 of 122) of mothers and 25% of fathers (11 of 44) were rated high EE. Of these, 28% of mothers (34 of 122) and 9% of fathers (4 of 44) were rated high CRIT; 12% of mothers (15 of 122) and 16% of fathers (7 of 44) were rated high EOI. CRIT and EOI were largely independent as only 7 mothers and no fathers were rated high on both CRIT and EOI; in only 6 families were both parents rated high EE.