Relationship status and quality of the partner relationship in parents of long‐term childhood cancer survivors: The Swiss Childhood Cancer Survivor Study‐Parents

The intensive and long‐lasting experience of childhood cancer is a tremendous stressor for the parental relationship. We aimed to (1) compare civil status and partner relationship of parents of long‐term childhood cancer survivors with population‐based comparisons, (2) identify cancer‐related characteristics associated with not being married, and (3) evaluate the quality of the partner relationship.


| BACKGROUND
"Childhood cancer is family cancer! Our family, our partner relationship did not function well for many years. The partner relationship remained affected until today!" (father of renal tumour survivor, 36 years after diagnosis). This comment in our questionnaire survey including parents of long-term childhood cancer survivors in Switzerland highlights that the intensive and long-lasting experience of childhood cancer affects the psychosocial functioning of all family members. [1][2][3] Managing the child's disease alongside other everyday activities is challenging for the parents. After diagnosis, parents are confronted with the potential fatality of the disease and conflicting caregiving, emotional, and practical demands they were not prepared for. 4 A recent review concluded that although most parent dyads adapt well to the crisis of having a child with cancer, findings related to conflicts in the parental relationship are mixed. 5 Even years after successful treatment, parents may experience uncertainties related to cancer relapse or late effects and continue to play an active role in the child's long-term care. 6,7 A majority of parents indicated that they felt well prepared for the child's cancer treatment; however, fewer reported feeling prepared for the life after cancer. 8 This points towards a persistent impact on parents' psychosocial functioning including the partner relationship even years after the child's recovery. However, the long-term impact of childhood cancer on the partner relationship from the perspectives of both parents remains largely unknown. 5 In this population-based questionnaire survey, we aimed to (1)

| Parents of childhood cancer survivors
This study is part of the Swiss Childhood Cancer Survivor Study (SCCSS) 9 and includes a sample of parents of long-term childhood cancer survivors (SCCSS-Parents) derived from the Swiss Childhood Cancer Registry (SCCR). 10,11 The SCCR centrally registers children and adolescents aged <20 years, who were diagnosed with leukaemia, lymphoma, central nervous system (CNS) tumour, malignant solid tumour, or Langerhans cell histiocytosis in Switzerland. 10

| Quality of the partner relationship
The quality of the partner relationship was evaluated using the relationship-specific attachment scale for adults by Asendorpf et al, which showed satisfactory psychometric properties in German settings. 12 The instrument consists of 14 items assessing the dimensions security (secure-fearful; 6 items) and dependency (dependent-independent; 8 items). Participants indicated their level of agreement with each item using Likert scales (1 = not at all to 5 = completely). If necessary, items were reverse-coded with higher scores indicating higher security and higher dependency. In our sample, internal consistency was satisfactory for the dimensions security (Cronbach α = 0.73) and dependency (α = 0.72). 13

| Cancer-related characteristics
The following cancer-related characteristics were extracted from the SCCR: cancer diagnosis, treatment, age at diagnosis (<5, 5-10, >10 years), time since diagnosis (<20, 20-30, >30 years), and relapse (yes, no). Cancer diagnoses were classified according to the International Classification of Childhood Cancer-Third Edition (ICCC-3) 16 and categorized into leukaemia, lymphoma (including Langerhans cell histiocytosis), CNS tumour, and solid tumour (neuroblastoma, retinoblastoma, renal tumour, hepatic tumour, bone tumour, soft tissue sarcoma, and germ cell tumour). Treatment was categorized hierarchically into surgery only, chemotherapy (may have had surgery), radiotherapy (may have had surgery and/or chemotherapy), and stem cell transplantation.

| Statistical analysis
All analyses were performed using Stata version 14.2 (StataCorp LP, College Station, TX). As a persons' relationship status may be associated with sociocultural background, we weighted comparison parents on age, sex, and migration background according to the distribution in parents of survivors. 15,17 Sociodemographic characteristics of comparison parents before weighting are presented in Table SS1. Subsequent analyses were based on weighted comparisons. We used descriptive statistics to compare sociodemographic characteristics, civil status, and partner relationship between parents of survivors and comparison parents. We determined associations between not being married and cancer-related characteristics using logistic regression models adjusted for parental age. A multilevel modelling approach with random intercepts was used to account for family clustering. To evaluate the quality of the partner relationship, we used means and 95% confidence intervals to describe the dimensions security and dependency. We   (Table SS2; all P > 0.05). Among 1255 population-based comparisons who participated in our survey (response rate: 23.6%), 471 parents (272 mothers, 199 fathers) were included in the analysis ( Figure SS2).
The mean age of parents of survivors and weighted comparison parents was 62.3 years (SD = 6.9) and 61.7 years (SD = 7.7), respectively. Parents of survivors were more likely to have ≥2 children (P < 0.001) than comparisons (Table 1). Mean time since diagnosis was 24.0 years (SD = 6.9), and survivors were on average aged 6.9 years (SD = 4.5) at diagnosis.

| Civil status, partner relationship, and determinants of marriage
Parents of survivors were less likely to be divorced/separated (Figure 1; 9.0% vs 17.5%, P < 0.001) and more likely to be married (83.4% vs 75.1%, P = 0.002) or in a partner relationship (89.9% vs 85.0%, P = 0.010) than comparison parents. There were no associations between not being married and cancer-related characteristics while adjusting for parental age and family clustering (all P > 0.05; Table SS3).  (Table 3; all P > 0.05). Both fathers of survivors and comparison fathers indicated a higher security (P = 0.037; P < 0.001) and a higher dependency (P < 0.001; P = 0.005) than mothers of survivors and comparison mothers, respectively. In the analysis restricted to parent dyads of survivors, where both the child's mother and father responded, fathers reported a higher dependency (P = 0.032), but similar security (P = 0.556) as mothers. We observed similar results in sensitivity analyses using imputed missing scale scores.

| DISCUSSION
This study highlights that parenting a child with cancer does not adversely affect parents' civil status or partner relationship in the long term. More than 20 years after the child's diagnosis, parental marriage was not related to the child's cancer diagnosis or treatment. However, parents of survivors reported a higher perceived dependency within the partner relationship than comparison parents.
The devastating experience of having a child with cancer is an enormous challenge for the parental relationship. 2,18,19 Yet, in line with our findings, a large registry-based study from Norway found no evidence for increased parental divorce rates. 20 The authors further concluded that the divorce risk was not associated with cancer diagnosis, prognosis, or time since diagnosis. 20 Similarly, parental marriage was not related to cancer-related characteristics in our study, suggesting that the burden going along with the child's cancer diagnosis may not necessarily be reflected in parents' marital status. A persons' marital status may also be influenced by the respective sociocultural background 21 and other partnership arrangements such as cohabitation without being married (which is nowadays increasingly established in Switzerland). 22 However, even after standardizing for sociodemographic characteristics, our study revealed that parents of survivors were more often in a partner relationship than comparison parents. Similarly, a Danish study concluded that childhood cancer did not affect parental separation rates. 23 Collectively, these findings suggest that many parent dyads adapt well to the crisis of having a child with cancer. Indeed, a recent review concluded that childhood cancer does not necessarily affect parent's functioning in terms of emotional closeness, support, and marital satisfaction. 5 Yet, difficulties in communication, sexuality, or role changes have been observed shortly after diagnosis. 5,[24][25][26] In our study, parents of survivors reported similar security and higher dependency within the partner relationship as comparison parents many years after the child's diagnosis. One explanation may be that parents of survivors feel more comfortable with this dependency or more often allow it to happen as they manage the child's disease in a collective effort that lasts long into survivorship. Additionally, parents of survivors reported having more children than comparison parents. This might contribute to an increased perceived dependency within the partner relationship.
Previous studies indicated that having a child with cancer resulted in increased mutual commitment and strengthened parental bonds. 19,20,[26][27][28] However, the impact of the cancer diagnosis on the parental relationship also largely depends on the couple's precancer functioning, 2 an aspect we could not account for in our study.   We further showed that both fathers of survivors and comparison fathers reported a higher dependency within the relationship than mothers. This pattern was confirmed among parent dyads of survivors, where we directly compared the perceptions of the survivors' mother and father. This is in contrast to a review that observed only few gender differences in perceived marital functioning among parents of children with cancer. 29 One explanation may be the rather traditional division of parental tasks in Switzerland. We previously showed that fathers of survivors were more often engaged in work and in charge of guaranteeing the household's financial stability than mothers. 15,17 Mothers typically adopts the role of the child's primary caregiver thereby contributing to fathers' perceived dependency, particularly if the child is suffering from a severe disease. Alternatively, fathers may feel more comfortable reporting dependency due to their role as the families' breadwinner. Such diverging roles or responsibilities 27 may have contributed to different perceptions of the relationship between mothers and fathers of children with cancer.

| Study limitations
A limitation of our study is the response rate of 44% among parents of survivors, which is lower than the median participation rate of 65% reported in a recent systematic review on studies including childhood cancer survivors. 30  This enabled a detailed evaluation of the parental relationship from both, the mothers' and fathers' perspective.

| Clinical implications
Understanding the impact of childhood cancer on the parental relationship is crucial to develop appropriate family support strategies along the child's cancer trajectory. More than 20 years after the child's cancer diagnosis, the partner relationship of Swiss parents of survivors was not adversely affected. The observed increased dependency within the relationship suggests that in Switzerland, the child's disease is managed in a collective effort by the parents. Two aspects may contribute to these encouraging findings: First, Swiss paediatric oncology settings provide comprehensive family-centred care during as well as after the acute treatment phase in the context of long-term followup care; second, specifically trained psychooncologists have been part of the care team for many years in order to address potential psychosocial consequences of the child's disease. Based on our findings, the development or implementation of additional interventions focusing on the parental relationship (eg, marital counselling) may be in imbalance in regard to costs and benefits in Switzerland. However, parents of children with cancer facing problems in the partner relationship should be identified early in the cancer trajectory and offered additional guidance in order to optimize parent, child, and family outcomes in the long term. [33][34][35][36]  Abbreviation: CI, confidence interval. a Analysis based on complete cases. Participants indicated their level of agreement with each item using Likert scales (1 = not at all to 5 = completely). Higher mean scores indicate higher security and higher dependency, respectively. b P value from adjusted Wald tests comparing mothers of survivors with fathers of survivors and comparison mothers with comparison fathers. c P value from adjusted Wald tests comparing mothers of survivors with comparison mothers and fathers of survivors with comparison fathers.