Add To Cart

Section 14
Childhood Psychiatric Disorders Affecting the Grief Process

Question 14 | Answer Booklet | Table of Contents

Psychiatric Outcome
In this group of studies, the focus of assessment is on disorders thought likely to be associated with the loss of a major attachment figure, namely childhood depression, anxiety, and somatising disorders. Assessment has relied on interview data with either parents and/or children.

Dysphoria and depression. Childhood dysphoria is commonly reported in parentally bereaved children both by their parents, and the children themselves. This finding holds regardless of sample size, sample recruitment, and type of measure employed. Thus both small-scale studies that include referred children (Cheifetz, Stavrakakis, & Lester, 1989), and larger-scale community studies (Van Eederwegh et al., 1982, 1985) concur in reporting childhood dysphoria, although rates vary.

Using standardized psychiatric measures, Cheifetz and colleagues (1989) examined psychiatric disorder in 16 bereaved children (12 boys and 4 girls) aged 4 to 17 years whose parents had died in the preceding 2 years. They report that 69% of their sample met DSM-III criteria for dysthymia, the latter being defined as "a depressed mood or loss of interest or pleasure in almost all activities for a period of one year, which is associated with only mild impairment'' (p. 690).

Van Eederwegh et al. (1982, 1985) report on 105 children aged 2±17 years from 50 parentally bereaved families recruited from a community sample of young widows/widowers. Controls were 80 community children within the same age range. Parents were interviewed at 1 and 13 months post bereavement. High levels of child dysphoria (a combination of sadness and crying or irritability) were reported. Significantly more bereaved children (77%) showed dysphoria than did controls (34%). They were significantly more likely to be withdrawn (31% vs. 13%), experiencing sleep disturbance (19% vs. 6%) and decreased appetite (15% vs. 5%). Overall, significantly more bereaved children (14%) than controls (4%) were classed as mildly depressed i.e. displaying dysphoria plus three other symptoms such as loss of appetite, difficulty sleeping, and loss of interest in activities.

If we take the dysthmia described by Cheifetz et al. (1989) as analogous to the mild depression described by Van Eederwegh et al. (1982, 1985), then there is a wide difference in rates between the two studies (69% and 14%, respectively). This is likely to stem from differences between their samples and study designs. Cheifetz et al. include referred children, which is likely to lead to an overestimate of rates of disorder. Their higher rate for dysthymia may also be explained by the use of interview data from bereaved children themselves. Other studies find that when bereaved children are interviewed, they report more psychiatric symptoms than their parents do of them (Gersten, Beals, & Kallgren, 1991; R. A.Weller, Weller, Fristad, & Bowes, 1991). In contrast to Cheifetz et al., Van Eederwegh and colleagues used a community sample and excluded referred children. However, they only interviewed parents. Thus, although the study has a number of strengths, including its large sample size and use of community bereaved and comparison groups, the failure to interview the bereaved children may result in an underestimate of rates of dysthymia.

Given that the majority of parentally bereaved children will be in families where surviving parents are likely to show high levels of psychological morbidity for up to 1 year following the death of their partner (Dowdney et al., 1999), symptoms of mild depression and dysphoria in their children are unsurprising. The issue of whether bereaved children will exhibit a depressive disorder following parental death is more contentious. Here, sample characteristics, informant source, and measures employed influence reported outcome.

High rates of severe depression are reported by only two studies of bereaved children (Cheifetz et al., 1989; R. A. Weller et al., 1991). In the study by Cheifetz and colleagues, the authors' finding that five of the study children (31%) met DSM-III criteria for depression  cannot be generalized to other samples because, as outlined above, the sample includes children referred to psychiatric facilities.

The studies by Weller and colleagues (1991) (Fristad, Jedel, Weller, & Weller, 1993; Sanchez, Fristad, Weller, Weller, & Moye, 1994; Sood, Weller, Weller, Fristad, & Bowes, 1992) can be regarded as more robust. The authors examine depression, somatic complaints, psychosocial functioning, and anxiety in 38 prepubertal children from 26 families identified from obituary notices. Bereaved children previously referred for psychiatric help were excluded. Both parents and children were interviewed between 3 and 12 weeks after parental death using a standardized psychiatric interview.

In their first report (Weller et al., 1991), the authors note that on self-report, 26% of the 38 bereaved children met DSM-III criteria for a major depressive disorder, compared to only 8% according to parental interview. When either report was used, the investigators classified 37% of children as severely depressed, and 60% as dysphoric. Further, 23 (61%) children reported suicidal ideation, although none had attempted suicide. The authors also compare depression manifesting in these bereaved children with that shown in a sample of clinically depressed children. They conclude that although bereaved children frequently report suicidal ideation, they are significantly less likely than severely depressed children to attempt suicide. They are also less likely to report feelings of worthlessness and fatigue.

Clinical experience indicates that a wish for death in bereaved children is closer to a wish to be reunited with the dead parent, rather than a desire for an end to life itself. Sadness and despair may, therefore, differ in both process and form between depressed and bereaved children. However, as the matter has yet to be adequately researched, this represents a clinical hypothesis requiring further investigation.

Although the Weller et al. study (1991) has a number of methodological strengths, two design flaws should be taken into account when considering apparent rates of depression. First, the bereaved children were identified via obituary notices or funeral directors. This strategy may result in a loss of up 30% of potential informants (Gersten et al., 1991), although whether this reflects a systematic bias is unknown. Second, the depressed comparison group, although matched on age and sex, were children admitted to inpatient facilities due to the  severity of their depressive illness. Although this allows the authors to pose interesting questions concerning the presentation of depression in bereaved and nonbereaved children, the findings cannot be generalized to less extreme samples. Indeed, community-based studies concur in reporting lower rates of severe depression (Gersten et al., 1991; Van Eederwegh et al., 1982, 1985). The Van Eederwegh studies report that over the 13-month period following parental death, only 6 of the 105 children studied (5%) were reported by their parents to be severely depressed, compared to 3 (4%) of community controls. As outlined previously, only parental data were available for this study.

Gersten and colleagues (1991; West, Sandler, Pillow, Baca, & Gersten, 1991) also found few severely depressed children in their study of 92 children aged 8 to 15 years chosen at random from families identified via death records. Parental death had occurred 3 to 24 months previously. The bereaved children were compared with a control group matched on sex, approximate age, and geographical location, and children in both groups were interviewed using the Child Assessment Schedule (CAS, Hodges, Kline, Stern, Cytryn, & McKnew, 1982). Only 9±8% of the bereaved children met DSM-III criteria for depressive disorder, although this was a significantly higher proportion than in the control group (1±3%). The bereaved children also had significantly higher mean depression scores. As with the R. A. Weller et al. study (1991), it was child data rather than parental report that distinguished the two groups, although in the Gersten et al. (1991) study, parental report was by questionnaire (the Child Behavior Checklist) and child report by interview (the CAS). Given the representative nature of the bereaved sample used by Gersten et al., the low rates of depressive disorder in this group appear reasonably reliable. However, a major limitation in the selection of the control group (who were identified and recruited from the telephone directory), means that statistical comparisons between the bereaved and control children are unsound.

We can conclude that only a small minority of bereaved children are likely to be at risk for a depressive disorder following the loss of a parent. The extent to which depression will be higher amongst bereaved children than in a demographically matched control group remains unknown.

Anxiety and somatizing disorders. There is little evidence of generalized anxiety or somatising disorders in bereaved children, although specific anxieties concerning separation and death are common (Sanchez et al., 1994; Silverman & Worden, 1992a). Both parents and the children themselves give accounts of increased child  anxiety, the content of which is largely confined to specific concerns about the safety of the surviving parent or other family members, and to anxiety about separation from  their remaining attachment figure. When somatising symptoms increase they most commonly take the form of headaches or stomachaches and, less usually, bedwetting (see Silverman &Worden, 1992a; Sood et al., 1992; Van Eederwegh et al., 1982, 1985). As with other symptoms of distress, bereaved children report more somatic complaints than their parents do of them.

Comorbid disorders. There is little to suggest that bereaved children experience comorbid disorders, except where a particularly traumatic death such as the murder or suicide of a parent (or sibling) occurs. In such cases, depressive disorder and post-traumatic stress disorder (PTSD) can co-occur, leading to greater difficulties in dealing with the death (see, e.g., Black & Harris- Hendriks, 1992). Further details concerning children's responses to traumatic death are presented in the section  on Moderating Variables.

Although comorbid disorders are unusual, several studies have found that psychiatric disorders such as depression or dysphoria can be accompanied by a number of other symptoms of disturbance. Van Eederwegh and colleagues (1982) report that significantly more bereaved than control children had temper tantrums (35% vs. 20% respectively). Cheifetz et al. (1989) note that depression scores in bereaved children were significantly associated with reports of behavioral disturbance, hyperactivity, anxiety, and withdrawal. Sanchez et al. (1994) reported that within their bereaved group, those with a major depressive disorder had significantly higher mean anxiety scores than the remainder.

Nonspecific Disturbance
Interviews with bereaved children and adolescents reveal marked affective responses to parental death including crying, sadness, anger, guilt, and despair (Kranzler et al., 1990; Silverman, Nickman, & Worden, 1992; Silverman & Worden, 1992a; R. A. Weller et al.,  1991). As outlined above, studies reporting specific psychiatric disorder in bereaved children also find concurrent symptoms of emotional and behavioral disturbance. Studies that focus on assessing such nonspecific emotional and behavioral disturbance note that it can manifest in a variety of ways.

Kaffman and Elizur, in their studies of Israeli kibbutz children, report that parental death led to serious psychological sequelae in these children (Elizur & Kaffman, 1982, 1983; Kaffman & Elizur, 1979). For example, in their initial study, Kaffman and Elizur used parental and teacher data to report on 24 Israeli kibbutz children aged 1±10 years from 14 families where a father had been killed in war 1±6 months previously. Forty-five percent of the children aged over 2 were judged by the investigators to be showing severe and widespread disturbance sufficient to be handicapping in their daily lives and to require referral. These findings are still frequently cited in the research and clinical literature (e.g., Baker et al., 1992; Black & Urbanowicz, 1987; Raphael, 1996; Sanchez et al., 1994; Schneiderman, Winders, Tallett, & Feldman,  1996). Yet, severe methodological limitations render those conclusions questionable. These include the absence of standardized measures and a control group, and the use of retrospective parental accounts of child functioning prior to the death. Further, confining the study to paternal death in war leads to a difficulty in disentangling the effects of war and parental death.

The finding that parental death can lead to a wide range of nonspecific child disturbance does, however, receive support from the more methodologically rigorous work of Kranzler et al. (1990) and Dowdney et al. (1999), who study children whose parents died from a variety of causes. Both employ standardized and validated questionnaire measures (CBCL, Teacher Report Form; Achenbach & Edelbrock, 1980, 1983), both use teacher and parental report, and both report higher levels of psychological disturbance in bereaved children in comparison  with controls.

Kranzler et al. (1990) confine their study to 26 bereaved children aged 3±6 years referred from a variety of sources  following media publicity of the study. Inclusion criteria were that the child lived with both parents prior to the death, or had daily contact with the deceased, and that the death had occurred within the previous 6 months. Both parents and teachers reported significantly higher overall problem scores in the bereaved children than in same-aged controls. The parents of the bereaved children reported significantly more emotional problems in their children than did comparison parents. Further, by parental report, 40% of the bereaved children had problem scores comparable to those found in clinical populations. The same was true for only 10%of controls.

In a study employing a prospective case control design, Dowdney et al. (1999) assess children aged 2±18 years from 45 families between 3±12 months following parental death. The sample was identified from death records within a circumscribed geographical area and includes only children living with both parents at the time of death. Those not living with the surviving parent, those in institutions or fostered, and those where one parent had murdered the other, are excluded. One child was chosen at random for detailed assessment from each of the 45 participating families. Parents report that 28% of the total sample of bereaved children had overall problem scores of clinical severity. When comparing the bereaved children of school age with randomly chosen classroom controls matched on age and sex, teachers rate them as displaying significantly more emotional and behavioral problems. They also report the bereaved children to be significantly more anxious and depressed, and as showing more attention and thinking difficulties. Teachers rated 10 of the bereaved children as having overall problem scores of clinical severity, compared to none of the controls. A significant level of agreement is found between parents and teachers on the severity of the problems shown, with a 78% pairwise level of agreement on whether bereaved children have problems of clinical severity.

Although both studies concur in finding significant differences between bereaved children and controls on the basis of parental and teacher reports, the overall rates of disturbance differ considerably. The parents in the Dowdney et al. (1999) study rate 28% of bereaved children as having overall problem scores of clinical significance, whereas the parents in the Kranzler et al. (1990) study report 40% of their children functioning at this level. The difference probably stems from the samples employed. Rates of disturbance are likely to be inflated in the Kranzler et al. study as the bereaved children derive from a variety of sources and include some referred because of their distress. In contrast, Dowdney and  colleagues use a representative general population sample. Further support for parental ratings in the latter study is provided by the highly significant agreement between teacher and parent pairs on whether children's scores lay within the clinical range.
- Dowdney, L. (2000). Annotation: Childhood Bereavement Following Parental Death. Journal of Child Psychology and Psychiatry, 41(7), 819-830. doi:10.1111/1469-7610.00670

Update
Prolonged Grief Disorder in ICD-11
and DSM-5-TR: Challenges and Controversies

- Eisma M. C. (2023). Prolonged grief disorder in ICD-11 and DSM-5-TR: Challenges and controversies. The Australian and New Zealand journal of psychiatry, 57(7), 944–951.

Peer-Reviewed Journal Article References:
Donohue, M. R., & Tully, E. C. (2019). Reparative prosocial behaviors alleviate children’s guilt. Developmental Psychology, 55(10), 2102–2113.

Kiser, L. J., Miller, A. B., Mooney, M. A., Vivrette, R., & Davis, S. R. (2020). Integrating parents with trauma histories into child trauma treatment: Establishing core components. Practice Innovations, 5(1), 65–80.

Marshall, A. D., Roettger, M. E., Mattern, A. C., Feinberg, M. E., & Jones, D. E. (2018). Trauma exposure and aggression toward partners and children: Contextual influences of fear and anger. Journal of Family Psychology, 32(6), 710–721. 

QUESTION 14
What was found in a study performed by Weller and colleagues related to the comparison of the manifestation of depression in bereaved children and a sample of clinically depressed children? To select and enter your answer go to Answer Booklet.

Answer Booklet
Section 15
Table of Contents
Top