Journal of Psychosomatic Research 76 (2014) 213–220

Contents lists available at ScienceDirect

Journal of Psychosomatic Research

A comparison of the emotional and behavioral problems of children of patients with cancer or a mental disorder and their association with parental quality of life Thomas Krattenmacher a,⁎, Franziska Kühne a, Susanne Halverscheid a, Silke Wiegand-Grefe a, Corinna Bergelt b, Georg Romer a, Birgit Möller a a b

University Medical Center Hamburg-Eppendorf, Department of Child Adolescent Psychiatry, Psychotherapy and Psychosomatics, Germany University Medical Center Hamburg-Eppendorf, Department of Medical Psychology, Germany

a r t i c l e

i n f o

Article history: Received 6 June 2013 Received in revised form 28 November 2013 Accepted 29 November 2013 Keywords: Child care Coping Parents Cancer Oncology Psychosocial aspects Quality of life

a b s t r a c t Objective: To compare the emotional and behavioral problems of children of patients suffering from cancer or a mental disorder and their association with parental quality of life. Methods: A total of 223 children from 136 families and their 160 parents were investigated from multiple perspectives in a cross-sectional study. The consistency of different adjustment reports between family members was examined. Through mixed models, the differences between parental HRQoL and the children's symptomatology were studied with regard to the type of parental illness. The prediction of children's adjustment through parental HRQoL was further examined. Additionally, gender and age of the children were considered. Results: Half of the children exhibited psychosocial problems. Gender and age differences were independent of the type of parental disease. In families with parental cancer, the reports of children's adjustment were more consistent between family members than in families where a parental mental disorder was present. We found differences in HRQoL between families with mentally ill parents and those with parental cancer patients. Specifically, the healthy partners of mentally ill parents showed worse HRQoL compared with healthy partners of cancer patients. Healthy parents' reduced HRQoL was associated with worse adjustment in their children, regardless of the type of parental illness, but this result was not found for ill parents. Conclusion: Family members confronted with parental cancer or mental disorders are more burdened compared with those from the “normal” population, independently of the type of disease. Our results indicate that the type of a parental disease has no direct effect on children's adjustment. However, there are disease-specific effects on parental HRQoL, which are associated with children's adjustment. © 2013 Elsevier Inc. All rights reserved.

Introduction Children of parents with severe chronic illnesses or mental disorders are at an increased risk of developing psychosocial problems [1–4] with prevalence rates up to 23% [5]. Comparisons of families with physically ill parents indicate only marginal effects of different diagnoses on child psychosocial adjustment [2]. Likely, in the case of parental somatic diseases, illness-related factors are not directly associated with children's adjustment, but they could indirectly affect the children's environment. A recent systematic review of studies on parental cancer [6] summarized associated factors of children's psychosocial adjustment and found that most studies reported only minor associations with illness-related factors but stronger associations with family or

⁎ Corresponding author at: University Medical Center Hamburg-Eppendorf, Department of Child and Adolescent Psychiatry, Psychotherapy and Psychosomatics, Martinistr, 52, D-20246 Hamburg, Germany. Tel.: +49 7410 57453; fax: +49 7410 55169. E-mail address: [email protected] (T. Krattenmacher). 0022-3999/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpsychores.2013.11.020

individual characteristics. Studies on mentally ill parents indicate that differences in children's adjustment based on different psychiatric diagnoses could also be less related to the specific mental disorder itself and more related to various environmental factors [7]. In contrast to parental somatic diseases, parental mental disorders could also have a heritable effect on children's adjustment. Compared with parental somatic diseases, this could augment children's vulnerability and increase their risk for developing psychosocial problems [8–10]. Direct comparisons of different illness types are important to identify their specific impact. Knowledge about common or different mechanisms of specific illness types enables the development of suitable psychosocial intervention programs for vulnerable populations. Nevertheless, studies directly comparing families with physically ill parents and families with a parental mental disorder are rare. Anthony (1970) [11] examined families with mentally ill parents and families with parents who suffer from tuberculosis. In this study, families with parental mental disorders had a lower socioeconomic status, which is associated with worse adjustment in dependent children [12,13]. Another study [14] compared families with depressed mothers, mothers with other

214

T. Krattenmacher et al. / Journal of Psychosomatic Research 76 (2014) 213–220

mental disorders and mothers with physical diseases. The authors concluded that the presence of a mental disorder in mothers was associated with worse adjustment in children compared to families with physically ill mothers. Additionally, several studies by the Hammen study group [15–18] also indicating that children of mentally ill parents show worse adjustment than those of physically ill parents. However, the conclusion that children of mentally ill parents are at a higher risk for maladjustment than those of physically ill parents should be interpreted with caution as the sample sizes in the previous studies were small (N b 20) [14–18]. Parent-related factors significantly predict children's adjustment [6,7,19], and parental well-being could be an especially important predictor of child maladjustment [20,21]. Health-related quality of life (HRQoL) is the “functional effect of a medical condition and/or its consequent therapy upon a patient. HRQoL is thus subjective and multidimensional, encompassing physical and occupational function, psychological state, social interaction and somatic sensation” [22]. HRQoL includes the physical, functional, social, and emotional wellbeing of an individual [23,24]. Mental disorders and severe chronic illnesses are associated with a decreased HRQoL [25,26], even in the long term [25,27]. Previous studies have found differences in HRQoL with respect to an individual's diagnosis [28–30]. Furthermore, not only the patient but also the entire family faces disruption and distress [11,31,32]. Partners of cancer patients are often more distressed than the patient [33–35] and report lower HRQoL than the average individual [36]. Partners can also be impaired by the mental illness of a spouse and show reduced quality of life compared with a reference norm [37]. In a general, population-based study, lower parental quality of life was associated with worse adjustment in children [38]. In a study on adolescent children of parents with chronic illnesses, lower HRQoL of both parents was associated with a greater number of emotional and behavioral problems in children [39]. Some studies on families with parental cancer found that lower parental HRQoL was associated with more emotional and behavioral problems in their dependent children [40,41], while other studies found no significant association of parental HRQoL with their children's emotional and behavioral problems [42–45]. Compared with healthy controls, parental pain and functional disability were associated with parent-rated child behavior problems [20]. HRQoL in families with mentally ill parents has not been directly examined, but previous studies have indirectly assessed the relationship of different parental qualities of life indicators, such as poor parental functioning, social disadvantage, marital discord, with children's adjustment [38] like poor parental functioning, social disadvantage, marital discord and family adversity [46]. At the child level, gender and age seem to moderate the association between a parental illness and children's adjustment, however, inconsistent results have been found for specific parental illness types [47–49]. Furthermore, it is important to consider who is rating the children's adjustment. Differences between mothers', fathers', and self-ratings of children's emotional and behavioral problems have been found in many contexts [50]. Previous studies on parental cancer found differences in the parental report of child adjustment based on health status and parental gender [6], as well as between parent- and self-rated adolescent adjustments [51]. Maternal depression could especially bias mothers' ratings of their children's adjustment; therefore, these ratings should be augmented by data from multiple informants [43,44]. Objectives The main objective was to study the differences between families with parental mental disorders and families with parental cancer regarding parental HRQoL and children's adjustment. Additionally, we investigated, whether age and gender effects of children's adjustment vary as a function of type of illness, and we included multiple

perspectives of children's adjustment. Lastly, the relationship between parental HRQoL and children's adjustment was studied. Hypotheses Based on previous findings, the main hypotheses were: H1. (a) In both samples, we expect worse adjustment of children compared with a reference norm. (b) We expect worse adjustment in children of families with parental mental disorders compared with children of parents who have cancer. (c) We expect less consistent ratings in families with parental mental disorders. H2. (a) In both samples, we expect lower parental HRQoL compared with the reference norms. (b) With regard to patients' HRQoL, we expect lower mental HRQoL (MCS) in patients with a mental disorder and lower physical HRQoL (PCS) in cancer patients. (c) We expect no significant differences in healthy partners' HRQoL. Furthermore we tested the following additional hypothesis: H3. We expect that the HRQoL of both parents is associated with children's adjustment, with worse HRQoL in parents being related to worse adjustment in children. Methods Participants Our study was part of two studies evaluating interventions for children of parents with cancer and parents with mental disorders. Data were collected between October 2009 and February 2011 in the Department of Child and Adolescent Psychiatry, Psychotherapy and Psychosomatics, University Medical Center Hamburg-Eppendorf, Germany. Patients who were parents, their partners, and their children (0–18 years old) were recruited in standard oncological or psychiatric care. They were offered specialized intervention programs and invited to participate in the studies. Study participation was not necessary to receive the respective specialized intervention. We obtained ethical approval, and each participant gave informed consent after having received oral and written information about the studies. Family members were instructed to complete the questionnaires independently and to not consult with other family members. We included data from families with at least 1 child between 0 and 18 years of age. Children older than 18 years were excluded. Families with parents who had both mental and physical illnesses were also excluded. Parents and adolescents were

Table 1 Psychiatric diagnoses of mentally ill (N = 69) cancer diagnoses of somatically ill parents (N = 67) Classification of DSM-IV diagnoses

N

%

Alcoholism Paranoid schizophrenia Depressive disorders Phobic disorders Eating disorders Personality disorders

4 13 20 11 5 16

5.8 18.8 29.1 15.9 7.2 23.2

Classification of primary tumor

N

%

Breast cancer Gynecological cancer Lung cancer Colorectal cancer Hematological cancer Skin cancer Gastro-intestinal cancer Other

21 4 7 4 7 2 9 13

31.3 6.0 10.4 6.0 10.4 3.0 13.4 19.4

T. Krattenmacher et al. / Journal of Psychosomatic Research 76 (2014) 213–220

required to be fluent in German. Cancer diagnoses were made by oncologists. Mental diagnoses (Table 1) were made by psychotherapists/ psychiatrists and validated with the Structured Clinical Interview for DSM-IV (SKID) [52]. Finally, a sample of 223 children (N = 123) and adolescents (N = 100) from 136 families, as well as their respective 160 parents, were eligible. Missing values Missing data are unavoidable in clinical research [53]. In our dataset, 33% of all participating subjects had at least one missing value, and there were significantly (p b 0.001) more missing values in families with parental mental disorders. There were no missing values for illness or sociodemographic variables. Missing data were assumed to be dependent on other observed variables (e.g., sociodemographic and illness variables), and, therefore, they were not missing completely at random (MCAR) [54]. Deleting cases or replacing missing values with mean values to address the missing data could produce crucial sample biases [54–57]. In this study, a multiple imputation model was used to account for missing data [58]. This approach is widely accepted as a state-of-theart method [55,57,59] for reducing bias and improving the validity of the analyses [53]. The multiple imputations were based on our analysis models and included all independent and dependent variables. Five iterations, with a maximum of 100 cases and 30 parameters, were conducted. Categorical variables were estimated by logistic regression models, and interval-scaled variables were estimated by linear regression models. Sample characteristics are presented in Table 2.

215

informants. An ICC below 0.40 indicates poor agreement, between 0.40 and 0.75 indicates moderate to good agreement, and above 0.75 indicates excellent agreement [60]. For group comparisons and predictive analyses, a mixed linear model (MLM) approach was used to examine dependent data [61]. This method is appropriate for data with nested sources of variability and involves the nesting of lower level units (parents and children) within a higher level unit (families). This technique allows for the simultaneous examination of the effects of both higher and lower level variables on individual-level outcomes and accounts for the nonindependence of observations within groups. We used a random intercept with fixed slope model in all mixed-model analyses. The family was included as a random effect. Fixed effects were dependent on the focus of the analyses as described below, with illness group as a fixed effect in group analyses and HRQoL components of ill and healthy parents as fixed effects in predictive models of child adjustment. Sample differences (sociodemographic status, time since diagnosis) were included as fixed effect covariates in all analyses. The significance level was set at p b 0.05. Holm–Bonferroni corrections were applied to correct for multiple testing [62]. First, we used MLMs to assess group differences in parental HRQoL between patients with cancer or a mental disorder and their partners. Second, we used MLMs to calculate differences in children's adjustment between families with parental cancer and those with parental mental disorders. Finally, we conducted regression analyses using MLMs to estimate the effect of parental HRQoL on children's adjustment. Additionally, we analyzed the effect of parental status (single vs. partnership) on child adjustment.

Data analyses Instruments All analyses were performed using SPSS 20. We performed independent t-tests and chi-square tests to examine differences in sociodemographic characteristics between both samples. To analyze the consistency among ratings of family members, intra-class coefficients (ICCs) were calculated to assess pair-wise (absolute) agreement between

Parental HRQoL was measured with the SF-8 version of the wellestablished Medical Outcomes Health Survey [63]. The SF-8 consists of eight items assessed on five-point Likert scales, with higher values indicating a better HRQoL. The Physical Component Summary (PCS), which

Table 2 Sociodemographic characteristics Mentally ill parents Mean (SD) Family characteristics (N = 136) Number of families Socioeconomic status

Number of children in family

Parental characteristics (N = 160) Ill parents Healthy parents Parental status Time since diagnosis

Children and adolescents' characteristics (N Children (N = 123) (0–10 years) Adolescents (N = 100) (11–18 years)

= 223) Female Age Female Age

N (%)

Mean (SD)

69 (100) 35 (50.7) 24 (34.8) 10 (14.5) 33 (28.7) 58 (50.4) 24 (20.8)

Low Middle High 1 2 ≥3 Female Age Male Age Single-parent Both parents Months b1 year 1–3 years N3 years

Parental cancer

67 (100) 8 (11.9) 25 (37.3) 34 (50.7) 34 (31.5) 54 (50.0) 20 (18.5)

55 (79.7) 39.4 (8.1)

45 (67.2) 42.9 (6.7)

26 (72.2) 42.7 (7.1)

34 (66.7) 44.8 (7.4)

31 (44.9) 38 (55.1) 111.6 (98.2)

22 (32.8) 45 (67.2)

p

b.001

n.s.

n.s. n.s. n.s. n.s. n.s.

29.8 (40.3) 8 (11.6) 13 (18.8) 48 (69.6)

40 (65.6) 5.3 (2.5)

31 (46.3) 21 (31.3) 15 (22.4)

b.001

32 (51.6)

n.s. n.s. n.s. n.s.

6.4 (2.8) 31 (57.4)

15.6 (2.7)

N (%)

20 (43.5) 14.6 (2.5)

Note: Means, standard deviations (SD), frequencies and percentage of the imputed sample. Characteristics of the imputed and original samples are equally distributed. Significant differences between both samples are calculated by t-tests or chi-tests [2] and are indicated in the right column.

216

T. Krattenmacher et al. / Journal of Psychosomatic Research 76 (2014) 213–220

measures physical functioning, the role of physical and bodily pain, and global health, as well as the Mental Component Summary (MCS), which measures social functioning and the role of emotional and mental health, were used for our analyses. German norms exist for these summaries [64], and good psychometric properties have been reported for the SF-8, with a Cronbach's α of 0.70 or higher for each item [63] and a retest reliability ranging from 0.61 to 0.70 [63]. Convergent correlations have been reported to range from 0.70 to 0.88 [63]. Children's and adolescents' adjustment was measured using the Child Behavior Checklist (CBCL) [65] to assess the emotional and behavioral functioning of the children from the parents' perspective. Adolescents (11–18 years) also completed the Youth Self-Report (YSR), a self-report version of the CBCL [66]. The YSR consists of 102 items, and the CBCL consists of 120 items, with higher values indicating more problems. We used the total problem scale, as well as the internalizing and externalizing subscales, of the CBCL/YSR. The internalizing subscale represents emotional problem syndromes, and the externalizing subscale represents behavioral problem syndromes. The total problem scale is a sum score of internalizing, externalizing, cognitive, social, and attention problems. Normative German data on the CBCL and YSR are available. Raw scores were transformed into T-scores based on normative data to assess possible differences between age and gender, as well as between informants, beyond expected differences in the general population. Cut-off points for T-scores exist for clinically elevated problems. The CBCL is widely used and has good psychometric properties [58,59]. The validity of the CBCL/YSR has been supported in a large number of studies [65,66], with Cronbach's α values ranging from 0.78 to 0.97 [58,59]. Results Children's emotional and behavioral problems Parents with a mental disorder reported sub-clinically elevated scores for emotional and behavioral problems in 11.1% of their children and clinically relevant scores in 40.9% of their children. Healthy partners of mentally ill parents reported sub-clinical problems in 8% of their children and clinical problems in 38% of their children. A total of 32.4% of the adolescents self-reported clinical problems, and 17.6% reported sub-clinical problems. In the case of parental cancer, ill parents reported that 14.9% of their children showed sub-clinically relevant symptoms, and 24.3% of their children showed clinically relevant symptoms. Healthy partners of parental cancer patients reported that 11.1% of their children showed sub-clinical and 27.8% clinical problems. A total of 30.6% and 19.4% of adolescents self-reported clinical and subclinical problems, respectively. Differences in children's adjustment with respect to type of illness In accordance with our hypothesis (H1a), children of both, parental cancer patients and mentally ill parents, showed significantly higher emotional and behavioral problem rates from parental and self-perspectives when compared with normal reference samples (p b 0.001). Parents with mental disorders rated their children as having significantly (p b 0.001) more clinically relevant symptoms compared with parents with cancer (H1b). In contrast, differences between samples were not significant when adjusting for sample differences (socioeconomic status, time since diagnosis, and parental status) and intra-correlations within families. Furthermore, in accordance with our hypothesis (H1c), the consistency (ICC) of internalization and externalization reports of family members in families with parental cancer was significantly higher than that in families with parental mental disorders. For an overview of intra-class correlation coefficients see Table 3.

Additionally, we analyzed age and gender effects. Overall, ill parents reported more internalization problems in girls than in boys (p = 0.015). With regard to age differences, this result was also found for adolescents (p = 0.021), but not for latency-aged children (p = 0.687), which was also consistent with our hypothesis. Gender and age differences were not significant regarding type of parental illness. Differences in parental HRQoL based on type of illness To test hypothesis H2a, we compared families with parental cancer and families with parental mental disorders with a German norm population [64], and found that both illness groups exhibited significantly (p b 0.001) lower HRQoL than the norms (Table 4). The hypothesis (H2b) that patients' HRQoL differed significantly between families with parental mental disorders and families with parental cancer was examined while controlling for sample differences. We found that parents with cancer reported significantly better mental HRQoL (p = 0.014) than parents with a mental disorder, which was in accordance with our hypothesis. Contrary to hypothesis H2b, no significant differences were found in the physical component scale reported by ill parents (p = 0.240). Healthy partners' HRQoL did not significantly vary based on the type of illness, which was also contrary to our hypothesis (H2c). While controlling for sample differences, we found that partners of parental cancer patients reported better mental HRQoL than partners of mentally ill parents (p b 0.001), but lower physical HRQoL than healthy partners of mentally ill parents (p b 0.001). Parental HRQoL and children's adjustment Our hypothesis (H3) that lower parental HRQoL predicts more emotional and behavioral problems in children was tested while controlling for sample differences (Table 5). We found that the mental and physical component summary of the healthy parent predicted the children's internalizing and total problems, rated from both parental perspectives (p b 0.05). According to ill parents' HRQoL, neither the mental nor the physical component summary significantly predicted the children's adjustment. No significant relationship between parental HRQoL and adolescents' self-rated adjustment was found. Of the included covariates, the type of illness and time since diagnosis did not significantly predict children's adjustment, but higher socioeconomic status was significantly associated with higher numbers of problems reported by ill and healthy parents (p b 0.05). However, we did not find any sample differences in parental status. We included parental status post-hoc in our analyses and found that single parent status was associated with more internalizing child problems from the ill and healthy parents' perspectives (p b 0.05).

Discussion The main objective of this study was to compare families with parental mental disorders and families with parental cancer with regard to children's adjustment and parental HRQoL. We investigated, whether children's adjustment and parental HRQoL varied as a function of illness type. Additionally, we examined the relationship between parental HRQoL and children's adjustment with regard to the type of parental illness. Children's emotional and behavioral problems Frequency analyses showed that 38.9 to 52.0% of the children in families with parental cancer or mental disorders displayed emotional and behavioral problems and were at higher risk for developing these problems than children from normal populations. With regard to parental mental disorders, up to 40.8% of the children displayed clinically significant symptoms, while in parental cancer families, a maximum of 27.9% of children showed clinically relevant symptoms. This result indicates a higher vulnerability of children to mental disorders which

Table 3 Consistency among raters of children's and adolescents' psychosocial adjustment

Functioning of

Reports of

Children Adolescents

Ill and healthy parents Ill and healthy parents Ill parents and adolescents Healthy parents and adolescents

Families with parental psychiatric disorder

Families with parental cancer

Internalization

Externalization

Internalization

Externalization

.309 .295 .358 .125

.408 .316 .555 .073

.474 .506 .490 .396

.572 .416 .540 .365

Note: Intraclass coefficients (ICCs) with one-way random intra-class analyses for families with parental psychiatric disorder and families with parental cancer.

T. Krattenmacher et al. / Journal of Psychosomatic Research 76 (2014) 213–220

217

Table 4 Ill and healthy parents' health-related quality of life Norm population

III Parent Healthy Parent

Mental component Physical component Mental component Physical component

Parental mental disorder

Parental cancer

MLM

Mean

SD

N

Mean

SD

N

Mean

SD

p

53.25 50.30 53.25 50.30

7.82 8.39 7.82 8.39

115 115 115 115

36.82 42.85 36.33 48.74

5.18 7.52 3.84 7.73

108 108 108 108

38.39 39.34 50.05 42.35

11.04 10.75 5.75 10.73

.014 .240 .000 .000

Norm Norm Norm Norm

N N N N

cancer N mental mental = cancer cancer N mental mental N cancer

Note: Means and standard deviations (SD) of the mental and physical components of ill and healthy parent's health-related quality of life (SF-8) in families with parental mental disorder or parental cancer compared with a German norm population. Significant differences tested by a mixed linear model (MLM) with socioeconomic status, partnership, and month since diagnosis as covariates.

is in line with previous studies [14–18] and might be due to the heritability of metal disorders [8–10]. Alternatively, the result could also be attributed to environmental factors. Previous studies have found that low socioeconomic status and single parental status significantly predicted the development of mental disorders in children [12,13]. We included socioeconomic status and single parental status in our analyses and also found significant relationships, so these environmental factors could also explain the higher rates of clinically relevant symptoms in the population of parents with mental disorders, which indicates that family functioning could be a critical variable [67–69]. Furthermore, when including those environmental variables, we did not find significant differences between children's adjustment due to type of parental disease. This result is in contrast to previous studies with smaller sample sizes, which did not account for intervening variables [14–18]. Ratings of children's adjustment varied between family members [50]. To identify absolute agreement between family members, we calculated intra-class coefficients and found that families with parental mental disorders reported less consistent information about children's adjustment than families of parental cancer patients. Previous studies have shown that parents with depression overestimate their children's/adolescents' symptomatology [70,71], while parents with bipolar disorder or cancer [72] underestimate their children's symptoms [15]. A number of studies [6,51,73] have shown that depressive cancer patients rated worse child adjustment than non-depressive cancer patients or their healthy partners. In line with our results, the fact that families with mentally ill parents reported less consistent results could lead to the conclusion that (additional) parental mental health problems reduce parental sensitivity for observing the mental state of their children. As we did not collect self-ratings of younger children, this result should be interpreted with caution. Additionally we investigated children's age and gender. Previous studies on parental cancer found that adolescent daughters show more maladjustment compared with sons [6,73–75] due to specific

interactional patterns. The high rates of breast cancer patients in some samples and worse adjustment of adolescent girls might be related to cancer-specific relationships, such as girls' anxiety about breast cancer heritability [73] and body identification [76–79]. In accordance with previous results, we found that ill parents reported more internalization problems in adolescent girls than boys. It has been hypothesized in other studies that this result could be an illness-specific effect, but we found no difference with regard to the type of disease. Our results indicate that gender and age differences are independent of the specific illness and likely a function of different developmental processes [80,81], as found in normal populations [82]. Differences in parental HRQoL with regard to type of illness Both, families with parental cancer and families with parental mental disorders reported lower parental HRQoL than norms. Previous studies have found differences in HRQoL depending on the presenting diagnosis [28–30]. We found that patients with cancer reported better mental HRQoL than patients with a mental disorder. This result could be attributed to specific disease impairments, as mentally ill parents could be more impaired with respect to social, cognitive, and emotional processes [83,84]. In contrast, no significant differences were found in the physical HRQoL reported by ill parents, indicating that parents with mental disorders are as physically impaired as parents with cancer. This result should be further investigated to exclude sample selection bias. Partners of cancer and mentally ill patients also displayed impaired HRQoL [33–37]. We expected that the HRQoL of healthy partners would not significantly vary between types of illness, as impairments are not a result of the specific disease, but rather are a result of general caregiver requirements [31]. In contrast, we found that healthy partners of cancer patients reported better mental HRQoL, but lower physical HRQoL, than partners of mentally ill parents. Contrary to our results, a previous study found that spouses of cancer patients reported worse

Table 5 Relationship between parents' health-related quality of life and children's and adolescents' functioning Self

MC of ill parent PC of ill parent MC of healthy parent PC of healthy parent Type of illness Socioeconomic status Time since diagnosis Parental status

Ill parent

Healthy parent

Internalizing

Externalizing

Overall

Internalizing

Externalizing

Overall

Internalizing

Externalizing

Overall

−1.41 −1.06 −3.01 −2.15 0.14 −2.23 −0.34 −1.56

−1.28 −1.04 −2.06 −1.02 0.11 −2.15 −0.36 −1.61

−1.39 −1.08 −2.42 −1.14 0.18 −2.24 −0.65 −1.43

−2.01 −1.06 −3.00⁎ −2.22⁎

−1.02 −1.03 −2.05 −1.76 0.09 −2.01 −0.34 0.75

−1.00 −1.08 −2.67⁎ −2.77⁎

−1.16 −1.03 −3.16⁎ −2.54⁎

−1.11 −1.00 −2.55⁎ −2.65⁎

0.13 −2.23⁎ 0.27 −1.03

0.15 −2.19 −0.45 −2.07⁎

−1.02 −1.01 −1.01 −1.88 0.12 −2.01 −0.33 −0.65

0.17 −2.07 0.21 −2.08⁎

0.11 −2.22⁎ −0.23 −1.40

Note: This table lists the predicted adjustment based on mixed linear models of children and adolescents using the HRQoL mental component (MC) and the physical component (PC) of the ill and the healthy parent, respectively. Estimated coefficients are calculated by controlling for the socioeconomic status (low vs. middle vs. high), time since diagnosis (months) and parental status (single vs. partnership). Type of illness (psychiatric vs. cancer) was also included as a fixed effect. Parent's HRQoL was measured using the SF-8. Internalizing, externalizing and overall problems were assessed by the Child Behavior Checklist (CBCL) and the Youth Self Report (YSR). ⁎ p b .05.

218

T. Krattenmacher et al. / Journal of Psychosomatic Research 76 (2014) 213–220

mental HRQoL, but not physical HRQoL, compared with a normal population [36]. Families with parental cancer are generally characterized by young breast cancer patients with a lower mortality rate [6]. Therefore, the healthy partner could be impaired by the physical limitations of the cancer patient (fatigue and bodily weakness) that disrupt daily family routines instead of worrying about the health of the ill partner. In a previous study, reduced physical and mental HRQoL was found in partners of mentally ill patients [85]. This result was not replicated in our sample. Parental HRQoL and children's adjustment Both the mental and physical HRQoL components of the healthy parent are associated with children's emotional and behavioral problems; otherwise, ill parents' HRQoL could not predict children's adjustment. Healthy partners of cancer patients are often more distressed than the patients [33–35]. Contrary to previous studies [16,43], healthy parents' HRQoL seems to be more important than ill parents' HRQoL with regard to children's adjustment when a parent is ill [39], and this could reflect the disruption of daily family routines, which could additionally burden dependent children. Based on this result, we analyzed whether parental status predicted child symptomatology, as has been observed in previous studies [86,87]. We found that single parenting predicted internalizing problems, even when we controlled for socioeconomic status [88]. Therefore, we concluded that when HRQoL is not reduced, the healthy parent could be a protective factor, likely balancing the adverse effects of the ill parent [21,89]. Furthermore, no significant relationship between the type of disease and the children's adjustment was found, indicating that HRQoL could be an important mediator of parental illness on child functioning, regardless of the specific illness. Methodological limitations This study included a large sample size with reduced bias resulting from missing values due to multiple imputations. We accounted for dependences within the data by using mixed linear models. Furthermore, multiple perspectives (ill/healthy parent, child) were used to account for reporting differences of child adjustment. Despite these strengths, this study has some limitations. First, as in other studies [6,51], there was little information on families with ill fathers. Second, we could not draw any conclusions regarding unadjusted sample differences such as family functioning. Third, we cannot generalize the results of this study to other somatic diseases, as other diseases could have different impacts on children's symptomatology [2]. In this study, participants with both a mental disorder and a physical illness were excluded. Therefore, we cannot draw any conclusions about the additive effects of illnesses (e.g., cancer patients with major depression). Cancer patients could be additionally affected by mental disorders [90] with an increased risk of maladjustment in children [6]. Fourth, the effect of specific mental diagnoses could vary significantly and patients with mental disorders often have comorbid mental disorders. We could show here a general effect of mental disorders on children's adjustment but not of comorbidity or differential effects of specific mental diagnoses. Higher sample sizes are needed to examine effects of different diagnoses. Further studies should provide interrater-reliability of diagnoses. Finally, we are unable to determine whether parental HRQoL is causally related to children's adjustment because of the cross-sectional design. Conclusion Children show worse adjustment compared with a normal population independently of the type of the parental illness. Age and gender differences in children's adjustment are independent of the type of parental illness. The accordance of ratings between family members on children's adjustment is less consistent in families with parental mental disorder.

This indicates that mental disorders reduce parental sensitivity for observing the emotional well-being of their children. Additionally, we found evidence that parental HRQoL is a key variable, independent of the parents' diagnosis to predict children's adjustment. Thus, when screening for higher risk families, health professionals should use parental HRQoL. Additionally, parental HRQoL should be supported in specialized intervention programs. Future studies should generally include different ratings of children's psychosocial adjustment, especially in samples with parental mental disorders, and should analyze mediators of the different rates of emotional and behavioral problems (e.g., family functioning and coping) according to the type of the parental illness. This is important to improve intervention programs for families affected by parental illness according to the specific needs. Conflict of interest There is no conflict of interest. Acknowledgments This study is part of the German multi-site research project entitled “Psychosocial Services for Children of Parents with Cancer,” which is supported by the German Cancer Aid (Deutsche Krebshilfe, grant # 108303). In this multisite project, the following institutions and principal investigators are collaborating: • Dept. of Child and Adolescent Psychiatry and Psychotherapy, Hamburg-Eppendorf University Medical Center (Prof. Georg Romer); • Institute of Medical Psychology, Hamburg-Eppendorf University Medical Center (Prof. Uwe-Koch-Gromus); • Dept. of Child and Adolescent Psychiatry, Psychosomatics & Psychotherapy, Charité University Medical Center, Berlin (Prof. Ulrike Lehmkuhl); • Dept. of Child and Adolescent Psychiatry, Psychotherapy & Psychosomatics, University Medical Center, Leipzig (Prof. Kai v. Klitzing); • Institute of Medical Psychology, University Medical Center, Leipzig (Prof. Elmar Brähler); • Dept. of Psychosomatic and General Clinical Medicine; University Medical Center, Heidelberg (Prof. Wolfgang Herzog); • Dept. of Child and Adolescent Psychiatry, University Medical Center, Heidelberg (Prof. Franz Resch); • Dept. of Child and Adolescent Psychiatry, Psychotherapy and Psychosomatics, Otto-von-Guericke University, Magdeburg (Prof. Hans-Henning Flechtner).

References [1] Barkmann C, Romer G, Watson M, Schulte-Markwort M. Parental physical illness as a risk for psychosocial maladjustment in children and adolescents: epidemiological findings from a national survey in Germany. Psychosomatics 2007;48:476–81. [2] Sieh DS, Meijer AM, Oort FJ, Visser-Meily JM, Van der Leij DA. Problem behavior in children of chronically ill parents: a meta-analysis. Clin Child Fam Psychol Rev 2010;13:384–97. [3] Ramchandani P, Psychogiou L. Paternal psychiatric disorders and children's psychosocial development. Lancet 2009;374:646–53. [4] Rutter M. Children of sick parents: an environmental and psychiatric study. London: Oxford University Press; 1966. [5] Maybery DJ, Reupert AE, Patrick K, Goodyear M, Crase L. Prevalence of parental mental illness in Australian families. Psychiatrist 2009;33:22–6. [6] Krattenmacher T, Kuhne F, Ernst J, Bergelt C, Romer G, Moller B. Parental cancer: factors associated with children's psychosocial adjustment-a systematic review. J Psychosom Res 2012;72:344–56. [7] Reupert A, Maybery D. Families affected by parental mental illness: a multiperspective account of issues and interventions. Am J Orthopsychiatry 2007;77:362–9. [8] Hyman SE. Grouping diagnoses of mental disorders by their common risk factors. Am J Psychiatry 2011;168:1–3. [9] Kendler KS, Kessler RC, Walters EE, MacLean C, Neale MC, Heath AC, et al. Stressful life events, genetic liability, and onset of an episode of major depression in women. Am J Psychiatry 1995;152:833–42. [10] Tsuang MT, Bar JL, Stone WS, Faraone SV. Gene-environment interactions in mental disorders. World Psychiatry 2004;3:73–83. [11] Anthony FJ. The impact of mental and physical illness on family life. Am J Psychiatry 1970;127:138–46.

T. Krattenmacher et al. / Journal of Psychosomatic Research 76 (2014) 213–220 [12] Bøe T, Øverland S, Lundervold A, Hysing M. Socioeconomic status and children's mental health: results from the Bergen Child Study. Soc Psychiatry Psychiatr Epidemiol 2012;47:1557–66. [13] Velez CN, Johnson J, Cohen P. A longitudinal analysis of selected risk factors for childhood psychopathology. J Am Acad Child Adolesc Psychiatry 1989;28:861–4. [14] Lee C, Gotlib I. Maternal depression and child adjustment: a longitudinal analysis. J Abnorm Psychol 1989;98:78–85. [15] Anderson CA, Hammen CL. Psychosocial outcomes of children of unipolar depressed, bipolar, medically ill, and normal women: a longitudinal study. J Consult Clin Psychol 1993;61:448–54. [16] Conrad M, Hammen C. Protective and resource factors in high and low-risk children: a comparison of children with unipolar, bipolar, medically ill, and normal mothers. Dev Psychopathol 1993;5:593–607. [17] Hammen C, Burge D, Burney E, Adrian C. Longitudinal study of diagnoses in children of women with unipolar and bipolar affective disorder. Arch Gen Psychiatry 1990;47:1112–7. [18] Hammen C, Gordon D, Burge D, Adrian C, Jaenicke C, Hiroto D. Maternal affective disorders, illness, and stress: risk for children's psychopathology. Am J Psychiatry 1987;144:736–41. [19] Manning C, Gregoire A. Effects of parental mental illness on children. Psychiatry 2008;8:7–9. [20] Chun DY, Turner JA, Romano JM. Children of chronic pain patients: risk factors for maladjustment. Pain 1993;52:311–7. [21] Romer G, Barkmann C, Schulte-Markwort M, Thomalla G, Riedesser P. Children of somatically ill parents: a methodological review. Clin Child Psychol Psychiatry 2002;7:17–38. [22] (ISOQOL) ISfQoLR. What is health-related quality of life?; 2012 [http://www.isoqol. org/about-isoqol/what-is-health-related-quality-of-life-research]. [23] Fayers P, Hays R, editors. Assessing quality of life in clinical trials. 2nd ed. Oxford: Oxford University Press; 2005. [24] Guyatt G, Feeny D, Patrick D. Measuring health-related quality of life. Ann Intern Med 1993;118:622–8. [25] Arndt V, Merx H, Stegmaier C, Ziegler H, Brenner H. Quality of life in patients with colorectal cancer 1 year after diagnosis compared with the general population: a population-based study. J Clin Oncol 2004;22:4829–36. [26] Rapaport MH, Clary C, Fayyad R, Endicott J. Quality-of-life impairment in depressive and anxiety disorders. Am J Psychiatry 2005;162:1171–8. [27] Boini S, Briancon S, Guillemin F, Galan P, Hercberg S. Impact of cancer occurrence on health-related quality of life: a longitudinal pre-post assessment. Health Qual Life Outcomes 2004;2:4. [28] Stewart AL, Greenfield S, Hays RD, Wells K, Rogers WH, Berry SD, et al. Functional status and well-being of patients with chronic conditions. results from the Medical Outcomes Study. J Am Med Assoc 1989;262:907–13. [29] Wells KB, Stewart A, Hays RD, Burnam MA, Rogers W, Daniels M, et al. The functioning and well-being of depressed patients. Results from the Medical Outcomes Study. J Am Med Assoc 1989;262:914–9. [30] Lochner C, Mogotsi M, du Toit PL, Kaminer D, Niehaus DJ, Stein DJ. Quality of life in anxiety disorders: a comparison of obsessive–compulsive disorder, social anxiety disorder, and panic disorder. Psychopathology 2003;36:255–62. [31] Baronet AM. Factors associated with caregiver burden in mental illness: a critical review of the research literature. Clin Psychol Rev 1999;19:819–41. [32] Pitceathly C, Maguire P. The psychological impact of cancer on patients' partners and other key relatives: a review. Eur J Cancer 2003;39:1517–24. [33] Axelsson B, Sjoden PO. Quality of life of cancer patients and their spouses in palliative home care. Palliat Med 1998;12:29–39. [34] Eton DT, Lepore SJ, Helgeson VS. Psychological distress in spouses of men treated for early-stage prostate carcinoma. Cancer 2005;103:2412–8. [35] Kornblith AB, Herr HW, Ofman US, Scher HI, Holland JC. Quality of life of patients with prostate cancer and their spouses. The value of a data base in clinical care. Cancer 1994;73:2791–802. [36] Bergelt C, Koch U, Petersen C. Quality of life in partners of patients with cancer. Qual Life Res 2008;17:653–63. [37] Angermeyer MC, Kilian R, Wilms HU, Wittmund B. Quality of life of spouses of mentally ill people. Int J Soc Psychiatry 2006;52:278–85. [38] Giannakopoulos G, Dimitrakaki C, Pedeli X, Kolaitis G, Rotsika V, Ravens-Sieberer U, et al. Adolescents' wellbeing and functioning: relationships with parents' subjective general physical and mental health. Health Qual Life Outcomes 2009;7:100. [39] Rodrigue JR, Houck CD. Parental health and adolescent behavioral adjustment. Child Health Care 2001;30:79–91. [40] Visser A, Huizinga GA, Hoekstra HJ, van der Graaf WT, Hoekstra-Weebers JE. Parental cancer: characteristics of parents as predictors for child functioning. Cancer 2006;106:1178–87. [41] Watson M, St James-Roberts I, Ashley S, Tilney C, Brougham B, Edwards L, et al. Factors associated with emotional and behavioural problems among school age children of breast cancer patients. Br J Cancer 2006;94:43–50. [42] Edwards L, Watson M, St James-Roberts I, Ashley S, Tilney C, Brougham B, et al. Adolescent's stress responses and psychological functioning when a parent has early breast cancer. Psychooncology 2008;17:1039–47. [43] Gazendam-Donofrio SM, Hoekstra HJ, van der Graaf WT, Pras E, Visser A, Huizinga GA, et al. Quality of life of parents with children living at home: when one parent has cancer. Support Care Cancer 2008;16:133–41. [44] Lindqvist B, Schmitt F, Santalahti P, Romer G, Piha J. Factors associated with the mental health of adolescents when a parent has cancer. Scand J Psychol 2007;48:345–51.

219

[45] Sigal JJ, Perry JC, Robbins JM, Gagne MA, Nassif E. Maternal preoccupation and parenting as predictors of emotional and behavioral problems in children of women with breast cancer. J Clin Oncol 2003;21:1155–60. [46] Fergusson DM, Horwood LJ, Lynskey MT. Maternal depressive symptoms and depressive symptoms in adolescents. J Child Psychol Psychiatry 1995;36:1161–78. [47] Cummings EM, Davies PT. Maternal depression and child development. J Child Psychol Psychiatry 1994;35:73–112. [48] Davies PT, Windle M. Gender-specific pathways between maternal depressive symptoms, family discord, and adolescent adjustment. Dev Psychol 1997;33:657–68. [49] Gross D, Conrad B, Fogg L, Willis L, Garvey C. A longitudinal study of maternal depression and preschool children's mental health. Nurs Res 1995;44:96–101. [50] Achenbach TM, McConaughy SH, Howell CT. Child/adolescent behavioral and emotional problems: implications of cross-informant correlations for situational specificity. Psychol Bull 1987;101:213–32. [51] Osborn T. The psychosocial impact of parental cancer on children and adolescents: a systematic review. Psychooncology 2007;16:101–26. [52] Wittchen H-U, Zaudig M, Fydrich T. Strukturiertes Klinisches Interview für DSM-IV. Göttingen: Hogrefe; 1997. [53] Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ 2009;338:b2393. [54] Donders AR, van der Heijden GJ, Stijnen T, Moons KG. Review: a gentle introduction to imputation of missing values. J Clin Epidemiol 2006;59:1087–91. [55] Enders CK. Applied missing data analysis. New York, NY: Guilford Press; 2010 [US]. [56] Moons KG, Donders RA, Stijnen T, Harrell FE, Moons KGM, Donders RART, et al. Using the outcome for imputation of missing predictor values was preferred. J Clin Epidemiol 2006;59:1092–101. [57] Schafer JL, Graham JW, Schafer JL, Graham JW. Missing data: our view of the state of the art. Psychol Methods 2002;7:147–77. [58] Rubin DB, Little RJA. Statistical analysis with missing data. 2nd ed. New York: Wiley; 2002. [59] Allison PD. Missing data thousand oaks. Sage Publications; 2001. [60] Novella JL, Jochum C, Jolly D, Morrone I, Ankri J, Bureau F, et al. Agreement between patients' and proxies' reports of quality of life in Alzheimer's disease. Qual Life Res 2001;10:443–52. [61] Raudenbush SW, Bryk AS. Hierarchical linear models: applications and data analysis methods. 2nd ed. Oaks (California): Sage Publications; 2001. [62] Holm S. A simple sequentially rejective multiple test procedure. Scand J Stat 1979;6:65–70. [63] Ware J, Kosinski M, Dewey J, Gandek B. How to score and interpret single-item health status measures: a manual for users of the SF-8 health survey. Lincoln, RI: QualityMetric, Inc.; 2001. [64] Beierlein V, Morfeld M, Bergelt C, Bullinger M, Brähler E. Measuring health-related quality of life with the SF-8: German norms from a representative self-administered survey. Diagnostika 2012;58:145–53. [65] Achenbach T. Manual for the child behavior checklist and 1991 profiles. Burlington: University of Vermont, Department of Psychiatry; 1991. [66] Achenbach T. Manual for the youth self-report and 1991 profile. Burlington: University of Vermont, Department of Psychiatry; 1991. [67] Gazendam-Donofrio SM, Hoekstra HJ, van der Graaf WT, van de Wiel HB, Visser A, Huizinga GA, et al. Family functioning and adolescents' emotional and behavioral problems: when a parent has cancer. Ann Oncol 2007;18:1951–6. [68] Jeske J, Bullinger M, Wiegand-Grefe S. Do attachment patterns of parents with a mental illness have an impact upon how they view the quality of life of their children? Vulnerable Child Youth Stud 2011;6:39–50. [69] Schmitt F, Piha J, Helenius H, Baldus C, Kienbacher C, Steck B, et al. Multinational study of cancer patients and their children: factors associated with family functioning. J Clin Oncol 2008;26:5877–83. [70] Ordway MR. Depressed mothers as informants on child behavior: methodological issues. Res Nurs Health 2011;34:520–32. [71] Richters JE. Depressed mothers as informants about their children: a critical review of the evidence for distortion. Psychol Bull 1992;112:485–99. [72] Welch AS, Wadsworth ME, Compas BE. Adjustment of children and adolescents to parental cancer. Parents' and children's perspectives. Cancer 1996;77: 1409–18. [73] Grabiak B, Bender C, Puskar K. The impact of parental cancer on the adolescent: an analysis of the literature. Psychooncology 2007;16:127–37. [74] Compas BE, Worsham NL, Epping-Jordan JE, Grant KE, Mireault G, Howell DC, et al. When mom or dad has cancer: markers of psychological distress in cancer patients, spouses, and children. Health Psychol 1994;13:507–15. [75] Quinn-Beers J. Attachment needs of adolescent daughters of women with cancer. J Psychosoc Oncol 2001;19:35–48. [76] Spira M, Kenemore EJ. Adolescent daughters of mothers with breast cancer: impact and implications. Clin Soc Work J 2000;28:183–95. [77] Wellisch DK, Gritz ER, Schain W, Wang HJ, Siau J. Psychological functioning of daughters of breast cancer patients. Part I: daughters and comparison subjects. Psychosomatics 1991;32:324–36. [78] Wellisch DK, Gritz ER, Schain W, Wang HJ, Siau J. Psychological functioning of daughters of breast cancer patients. Part II: characterizing the distressed daughter of the breast cancer patient. Psychosomatics 1992;33:171–9. [79] Wellisch DK, Schains W, Gritz ER, Wang HJ. Psychological functioning of daughters of breast cancer patients. Part III: experiences and perceptions of daughters related to mother's breast cancer. Psychooncology 1996;5:271–81. [80] Krattenmacher T, Kühne F, Führer D, Beierlein V, Brähler E, Resch F, et al. Coping skills and mental health status in adolescents when a parent has cancer: a multicenter and multi-perspective study. J Psychosom Res 2013;4:252–9.

220

T. Krattenmacher et al. / Journal of Psychosomatic Research 76 (2014) 213–220

[81] Zahn-Waxler C, Shirtcliff EA, Marceau K. Disorders of childhood and adolescence: gender and psychopathology. Annu Rev Clin Psychol 2008;4: 275–303. [82] Bongers I, Koot HM, Ende J, Verhulst F. The normative development of child and adolescent problem behavior. J Abnorm Psychol 2003;112:179–92. [83] Ginieri-Coccossis M, Triantafillou E, Tomaras V, Liappas I, Christodoulou G, Papadimitriou G. Quality of life in mentally ill, physically ill and healthy individuals: the validation of the Greek version of the World Health Organization Quality of Life (WHOQOL-100) questionnaire. Ann Gen Psychiatry 2009;8:23. [84] Guan B, Cohen P, Deng Y, Chen H. Relative impact of Axis I mental disorders on quality of life among adults in the community. J Affect Disord 2011;131:293–8. [85] Heru AM, Ryan CE, Vlastos K. Quality of life and family functioning in caregivers of relatives with mood disorders. Psychiatr Rehabil J 2004;28: 67–71.

[86] Amato PR, Keith B. Parental divorce and the well-being of children: a meta-analysis. Psychol Bull 1991;110:26–46. [87] Cherlin AJ, Furstenberg FF, Chase-Lansdale L, Kiernan KE, Robins PK, Morrison DR, et al. Longitudinal studies of effects of divorce on children in Great Britain and the United States. Science 1991;252:1386–9. [88] Gelles RJ. Child abuse and violence in single-parent families: parent absence and economic deprivation. Am J Orthopsychiatry 1989;59:492–501. [89] Krattenmacher T, Kuhne F, Fuhrer D, Ernst J, Brahler E, Herzog W, et al. Parental cancer—parents ways of coping, family functioning, and psychosocial adjustment of dependent children. Prax Kinderpsychol Kinderpsychiatr 2012;61:447–62. [90] Harter M, Reuter K, Aschenbrenner A, Schretzmann B, Marschner N, Hasenburg A, et al. Psychiatric disorders and associated factors in cancer: results of an interview study with patients in inpatient, rehabilitation and outpatient treatment. Eur J Cancer 2001;37:1385–93.