Systematic review
What are the consequences of shared physical custody for children? A systematic review
Systematic review
|Published
We were asked to summarize studies that could say something about the consequences (effect) of shared custody for children.
Download
Key message
In recent years, Norway has experienced a large increase in the use of shared custody for children among parents living separately. The same trend is seen in many other western countries. The Norwegian Directorate for Children, Youth and Family Affairs (Bufdir) has asked the Norwegian Institute of Public Health to conduct a systematic review of studies that look at consequences of shared custody for children.
We conducted a systematic review of studies that examine the consequences of different living arrangements for children. We included five studies of children aged 0-6 years. Four of the studies were conducted in the USA and one in Australia. They were published between 1999 and 2016. The studies examined outcomes related to attachment, psychological health and relationships between parents and children.
Because the studies were too heterogeneous, we could not perform meta-analyses. The studies found quite heterogeneous results. However, there appears to be a pattern in the results, related to the age of the children: for children over three years who have (more) overnights with the other parent, there seem to be more positive results, while for children younger than three years who have (more) overnights with the other parent, the results are more mixed and/or negative. However, our assessment shows that we have very low confidence in all of the results. This is because the studies are observational, most have very few participants, and several of the studies have high risk of bias.
We were asked to summarize studies that could say something about the consequences (effect) of shared custody for children. Our assessment is that it is very uncertain what we can conclude about such effects.
Summary
Background
In recent years, Norway has experienced a large increase in the use of shared physical custody for children among parents living in separate households. In 2012, 26% of the children lived in shared custody, compared with 10% in 2004 and 8% in 2002. The same trend is seen in many other western countries. How does living in two homes, compared to living most of the time in one home, affect the children? In particular, how does it affect the youngest children? These questions are associated with uncertainty and disagreement. The Norwegian Directorate for Children, Youth and Family Affairs (Bufdir) has asked the Norwegian Institute of Public Health to conduct a systematic review of studies that examine consequences of shared custody for children.
Aim
The aim of this review was to identify, assess and synthesise research about the consequences of shared physical custody on child outcomes.
Method
We conducted a systematic review in accordance with the division for the health services’ handbook. We performed a literature search in relevant databases in August 2016 and a search for gray literature in March 2017. Systematic reviews of high or moderate methodological quality and/or longitudinal, controlled studies were eligible. To be included, the primary studies had to include analysis of longitudinal data and have a relevant control group. Outcomes were attachment, psychological health and parent-child relationship. To assess eligibility, two researchers first screened all titles and abstracts and subsequently relevant articles in full text. After reviewing all relevant studies, we found that none of the studies met the criteria for study design. After discussions with the commissioner, we decided to change the criteria for study design for the studies that investigated outcomes in children aged 0-6 years. This was because the need for knowledge about what research can and cannot tell us about how shared physical custody affect children in this age group is particularly important. For this age group, we chose to include studies that had a relevant control group. Two researchers critically appraised the included studies for risk of bias. We synthesised the results narratively and in tables, but the studies were too heterogenius to conduct meta-analyses. For each finding we considered the certainty of the evidence by using the GRADE approach (Grading of Recommendations Assessment, Development and Evaluation).
Results
We read 4475 abstracts and 61 articles in full text. We included five unique studies. The studies compared outcomes for children between 0-6 years who had different custody/overnight arrangement. Four of the studies were conducted in the USA and one was conducted in Australia. They were published between 1999 and 2016, and were either cross-sectional or longitudinal. The longitudinal studies presented only relevant analyses of cross-sectional data. The studies examined outcomes such as attachment, psychological health, and parent-child relationship. Overall, they presented results from five different comparisons, i.e., of different custody/overnight arrangements.
Three studies compared children who had overnight stays with their father with children who had no overnight stays with their father. For children 0-3 years, none of the studies found any differences between the groups for the outcomes secure- and insecure attachment. For the outcome disorganized attachment, more of the children who stayed overnight had a disorganized attachment style (RR [Risk Ratio]: 1.54, CI [Confidence Interval]: 1.04-2.26). For the outcome psychological health, measured with the CBCL (Child Behavior Checklist), there were no differences between the groups. For the outcome relationship to father, there was a positive relationship for the children staying with their father, for both children under 1 year (r: 0.356, p <0.05) and for 2 year olds (r: 0.500, p <0, 01). For children 4-6 years, one study measured psychological health with CBCL, finding positive results for the children who had overnight stays.
Two studies compared children who had frequent overnight stays versus some overnight stays. For children 0-1 year, there was no difference between the groups regarding psychosomatic health. For the outcomes related to emotional regulation, children with frequent overnight stays fared worse (B: -0.37, CI: -0.74-0, p = 0.05). For the outcome attachment, there were more children with insecure attachment in the group with frequent overnights (OR: 5.62, missing CI). For children older than 2-3 years, the outcomes related to emotional regulation showed that children with frequent overnights fared worse than children with some overnight stays (B: -0.32, CI: 0.01-0.64 p = 0,05, B: -2.92, CI: -4.97-0.87, p = 0.01). For the outcomes attachment and prosocial behavior, there were no differences between the groups. For children aged 4-5 years, there were no differences between the groups.
Two studies compared frequent overnight stays versus no overnight stays. For children 0-1 year, there were no differences between the groups on the outcomes related to psychosomatic health and emotional regulation. For children aged 2-3 years, there were inconsistent results: For the outcomes related to emotional regulation, there was a difference between the groups when measured with the BITSEA Problems scale and The Emotional Functioning scale. When the outcome was measured with the Persistence scale, children with frequent stays fared worse than those who had no overnights (B: 0.42, KI: 0.13-0.72, p <0.01). For the outcome attachment, there were no differences between the groups. For children aged 4-5 years, there were no differences between the groups for the outcomes psychosomatic health, emotional regulation, and attachment. For the outcome prosocial behavior, children with frequent overnights scored higher on behavior than those who did not stay overnight (B: -1,47, p<0.05).
One study compared (more) overnight stays with the father versus less/no overnights. For children 0-1 year, for outcomes related to relationship to father, there was a positive relationship between (more) overnights compared to less/no overnights for four out of five outcomes. For example, for the outcome 'Experience of quality in relation to father 0-16 years', the result was r = .295 (p <0.05). For outcomes related to relationship to mother, there was no association between (more) overnights compared to less/no overnights for four out of five outcomes. For children aged 2 years, there were mostly positive relationships between (more) overnights and outcomes related to how the grown up children rated their relationship to their parents.
Confidence in the results
Our assessment of each finding show that we have very low confidence in all of the results. This is because, among other things, all included studies are observational, with cross-sectional data. Many of the studies are small and had a high risk of bias, due to, for example, how outcomes were measured.
Discussion
Overall, the results of the studies are quite heterogeneous. The results are based on cross-sectional data from five studies. The results, which are based on very low certainty of evidence, should therefore be interpreted with caution. Nevertheless, some patterns in the results related to the age of the children can be observed: For children 4-6 years, the studies found either positive relationships between shared custody/overnights and outcomes for children – this applies to the outcomes psychological health, prosocial behavior and somatic disorders – or they found no differences between the groups. For children 0-3 years, the studies found mixed results. Several of the studies found negative relationships between overnights and outcomes in the children – this applies to the outcomes attachment and emotional regulation. At the same time, for other outcomes, there were no differences between the groups. Some studies also found positive relationships between overnights and outcomes related to the children – for example outcomes related to parent-child relationship.
Conclusion
We were asked to summarize studies that could answer questions about the consequences (effect) of shared custody or overnight stays, for children. Our assessment is that it is very uncertain what we can conclude about this. For the five studies we included, we assessed that we have very low confidence in the results. This is due to the study design, size, outcome and measurement methods of the studies, and means that we are uncertain whether the results are true.
Knowledge gaps and research needs
One of the goals of this report was to investigate whether there are studies with a longitudinal design, control group and follow-up measurement of the same outcome. We found no such studies. To build a more solid empirical evidence base on questions about the consequences (effects) of shared custody for children, longitudinal studies are necessary. We believe it would be useful to examine questions related to correlations – such as shared custody, conflict between parents and child outcomes. Furthermore, we believe there is a need for more qualitative studies that investigate how children feel about living in shared custody, and we believe that a summary of such studies would be valuable.
