Do frequency of visits with birth parents impact children’s mental health and parental stress in stable foster care settings

Abstract This article explores whether the number of visits by birth parents influence perceptions of attachment, children’s competence and mental health, and stress levels in foster parents. Foster parents acted as informants regarding 203 children living in kinship and non-kinship foster care. The children were young when placed in foster care, on average 2.3 years old (SD = 1.0) and had been living in the foster home for sometime at assessment, 5.4 years (SD = 3.0). Information were collected using validated instruments. The results showed that 47% of the children had monthly or more frequent visits with their mothers, whereas 21% of the fathers had visits this often. Visitations with birth parents did not significantly influence who was the main attachment figure or foster parental attachment relationships, the children’s psychosocial functioning or competence, or stress levels among the foster parents. These findings could indicate that social workers should emphasize the quality and short- and long-term consequences of visits for children when making decisions regarding the frequency of visits with birth parents. This could be done taking the child’s reactions and wishes into account, when evaluating the visit and the frequency of future visits. It is important that decisions concerning visits should be continuously revised in both the short- and the long term, since both wishes and practical aspects may change for all parties involved.

ABOUT THE AUTHORS Sturla Fossum works as an associate professor at UiT, the Arctic University of Norway (UiT). He has conducted studies of parent training for aggressive children, child welfare, and metaanalyses and reviews. Svein Arild Vis works as an associate professor at the same site and he has conducted research of child involvement in child welfare investigations, reviews of child welfare practices, and children living in foster care. Amy Holtan works as a professor at UiT. She has published articles in child welfare, foster care and in particular, on kinship foster care placement and co-authored reviews on kinship foster care. Holtan is project leader on "Outcome and experiences of kinship foster care placements" of which this paper is a part of. The project group is currently conducting a long-term follow-up of these children living in foster homes.

PUBLIC INTEREST STATEMENT
Foster care is a necessity for some children. When children grow up under such arrangements, birth parents have a legal right to meet with their child. In Norway, the extent of visitations is determined by courts. This is common. Knowledge of how visits influence foster children's mental health and foster parental stress are somewhat limited. This is relevant when courts determine the amount of parental visitations for foster children. Concerns raised by professionals and laypersons alike, that visits with birth parents negatively influence the children's mental health, attachment between the child and foster parents and foster parent stress influence these decisions. This study conducted in Norway suggested that negative effects are not true in all cases, at least when children were placed young and living in stable settings. Instead of being concerned with possible negative effects of visits, the most important consideration is the quality and not the frequency of visits. Still, some visits with birth parents constitute major stress for all parties, and the findings in this study is not an argument for more visits in these cases.

Introduction
Every year, children all over the world are removed from their homes due to abuse, neglect, or other forms of maltreatment (Winokur, Holtan, & Batchelder, 2014). Although this is necessary at times, out-of-home care is one of the most disruptive, invasive and costly options are available to protect and care for children who cannot live with their birth families (James, Landsverk, & Slymen, 2004). In Norway, the municipal Child Protection Service (CPS) is responsible for investigating and when considered necessary bringing these cases to court. The CPS also provides social services for the children and foster families involved. Out-of-home placements such as kin and non-kin foster homes are widely used, and at the end of 2012, 9600 children were living in foster homes in Norway (Statistics Norway, 2012). Half of the children in foster homes in Norway were less than 12 years old (Backe-Hansen, Havik, & Grønningsaeter, 2013). The children in this sample lived in kinship foster homes implying that one of the foster parents were related to the child being for instance, a grandmother, aunt, or uncle or in regular foster homes, implying that the children were not related to the child.
Children with previous foster care placements are at greater risk of future emotional and behavioral problems compared to children and adolescents in general (Backe-Hansen, Madsen, Kristofersen, & Hvinden, 2014;Clausen & Kristoffersen, 2008), and children with a history of contact with child welfare authorities are at greater risk of adverse outcomes such as poorer education (Vinnerljung, Öman, & Gunnarson, 2005), teenage parenthood (Vinnerljung, Franzén, & Danielsson, 2007), attempted suicide and severe psychiatric morbidity (Vinnerljung, Hjern, & Lindblad, 2006). Stability in the foster home placement is therefore important for several reasons. The need for health services, including both in-and outpatient mental health services and paid claims in general health care setting in the United States of America (USA), is lower among children who experience greater stability in their placements (Rubin et al., 2004) and there are strong associations between stability in foster care placements and children's well-being as measured by the Child Behavior Checklist (CBCL-for information, see the Method section) (Rubin, O'Reilly, Luan, & Localio, 2007). Winokur et al. (2014) reported that children in kinship care experienced greater placement stability than children in nonkin foster homes, and outcomes were more favorable for children in kinship care with regard to behavioral problems, adaptive behaviors, psychiatric disorders, and well-being. Over a period of six years, foster carers requested a change in fostering arrangements in 42.2% of cases (Bernedo, Garcia-Martin, Salas, & Fuentes, 2015), which suggests that there is a need to learn more about the well-being of foster parents in order to prevent and avoid foster care breakdowns.
Maintaining contact between foster children and their birth parents is covered by the United Nations Convention on the Rights of the Child (UNCRC) (Unicef, 2012), and has been incorporated into Norwegian legislation (Haugli, 2007). In Norway, the number of visits with birth parents plays a central role in family care court hearings and the frequency of such visits varies. Norwegian Supreme Court judgments typically state that children and their birth parents should meet at least four to six times a year (Norges Offentlige Utredninger-Norwegian public research, [NOU], 2012). Children and their birth parents have a statutory right to meet unless otherwise is decided, according to paragraphs 42 and 43 of the UNCRC (Child Welfare Act, 1992). The overarching principle is that it is the best interest of the child to meet with birth parents, as is also mandated by the Norwegian Children's Act. Parental visits are regarded as an important means of maintaining the parent-child relationship (Haight, Kagle, & Black, 2003;Moyers, Farmer, & Lipscombe, 2006;Sen & Broadhurst, 2011) and parental visits are a strong predictor of family reunification prospects Davis, Landsverk, Newton, & Ganger, 1996).
Although contact with birth parents can be positive for children in foster care, visits may also be stressful for the children. In one study, about one-third of children in foster care felt stressed during visits with their birth parents, and contact was unproblematic for only a few of them (Neil, Beek, & Schofield, 2003). Furthermore, there are some concerns that visits with birth parents may negatively influence attachment between the children and their foster parents, i.e. the quality of the bond between the foster parents and the child (Browne & Moloney, 2002). Based on the experience of 17 social workers in 81 cases, the social workers reported of "Positive" reactions in the children in 50 cases (62%), in 43 cases (53%) the children had "Negative" reactions, and finally in 13 cases (16%), the reactions were "Improving" (Browne & Moloney, 2002). "Positive reactions" included that the children was looking forward to the visits, "Negative reactions" included an unwillingness to talk during the visits or the children refused to go to the visit in some occasions since sporadic visits had become more regular (Browne & Moloney, 2002). In a review, of 10 studies, which explored both the negative and positive consequences of foster child visits with their birth parents in both acute and long-term placements, the authors concluded that the evidence was not clear, and that many of the studies suffered from methodological weaknesses in designs, representativeness of the samples and statistics (Quinton, Rushton, Dance, & Mayes, 1997). In a dissertation from USA including 64 children and adolescents with age ranging from six to 18 years, contact with birth parents was not related to the child's mental health, neither internalizing nor externalizing problems (Rich, 2010).
This present study explored whether the number of visits with birth parents influences the functioning in various foster care arrangements, both kin and regular foster homes, using well-validated instruments of children's mental health, attachment, and level of stress among foster parents. At the start of the study, the children in our sample had been living in the foster home for at least a year. In particular, we were interested in exploring three topics: (i) Does the number of visits with birth parents influence children's competence and mental health?
(ii) Is attachment, that is who is the main attachment figure and the bond between child and foster parent, influenced by the number of visits with birth parents? (iii) Is the level of stress in foster parents influenced by the number of visits?

Participants
Foster parents of children ranging in age from four to 13 years old, living in court-ordered kin and non-kin foster homes in 2000, were recruited for participation. In the kinship group, 238 foster homes in 104 municipalities were asked to participate. These represented 98% of all registered kinship foster care placements in Norway at the time. About half of these homes were enrolled in the study (n = 124). In the comparison group, 90 of 192 non-kin foster homes agreed to participate. Foster parents were interviewed and completed the questionnaires.
Originally, 246 children were included in the study. Thirteen were excluded due to missing data and 30 cases were eliminated since the children were siblings. Siblings living in the same foster home were excluded in order to avoid dependency in the data due to all measures being reported by the foster parents. The total number of cases used for analysis was 203. The average age of the children when placed in care was 2.3 years (SD = 1.0). At recruitment, the children had been living in the same foster home for an average of 5.4 years (SD = 3.0). Some of the children, in all 90 (44.8%), had been in one or more foster homes prior to their current placement. Rather more than half of the children (56.1%) were boys, and the mean age of the total sample was 9.2 (SD = 2.8). The mean age of the foster mothers was 45.8 (SD = 7.8) and foster fathers 47.1 (SD = 8.5) years. A total of 182 (84.3%) of the foster homes consisted of two parents.

Demographic information and placement characteristics
This questionnaire was designed for topics concerning: (1) children's history of care (e.g. age at first removal, number of previous placements and duration in care); (2) children's family contact (e.g. visiting arrangements with birth parents and siblings, location of foster and birth homes); and (3) caregiver characteristics (e.g. age, marital status, education, income, health, family relationship of child and caregiver) including information of who the foster parents considered were the children's main attachment figure (Holtan, 2002).

Child mental health and competence
The foster parents completed the Child Behavior Checklist (CBCL). The competence scales include Activities, which is the number of and quality of activities in sport, non-sport hobbies, and job chores, Social competence refers to the number and quality of relationships and organizational involvements, and school, which comprises academic performances and grade repetition. The total competence scale serves as a global index summing these scales. The problem section of the CBCL consists of 118 items (0-2 scale) that refer to emotional and behavioral problems during the past six months (Achenbach, 1991).
In this study, the total problem score as well as the internalizing and externalizing syndrome scores were converted to t-scores based on the US normative sample, using the ASEBA software package.

Foster parental stress
Foster parents completed the Parent Stress Index (PSI), which consists of 101 items rated on a 1-5-point Likert scale that measures parents' perceived total stress (Abidin, 1995). In this article, we also discuss whether attachment between foster parents and children is influenced by visits, using the attachment section of PSI, which consists of seven items. Higher scores on this section could suggest two sources of dysfunction, that is a parental lack of closeness to the child or parent's real or perceived inability to understand or observe the child's need. A score of 18 or higher equals the 95th percentile of the attachment section. PSI do not have Norwegian norms and we use scores from the US normative sample to determine clinical cut-off. The PSI manual was employed to determine how to deal with missing data and to determine the clinical level of stress, i.e. a score equal to or above 250 corresponding to the 80th percentile. A total score of 222 equals the 50th percentile.

Procedures
Participants were included in the study and responded to the questionnaires independently of time gone since the last visits of the birth parent. The study was approved by the regional committees for medical and health research ethics in Norway, and was conducted in line with the Helsinki Declaration of ethical principles for medical research involving human subjects published by the World Medical Association (2008).

Statistics
When two or more siblings were included in the study, in order to ensure data independence, only the sibling with the first birthday during the calendar year was included in the analyses. A score equal to or above the borderline on the CBCL total score (t-score ≥ 60) was used to determine the number of children within the clinical range. In evaluating stress levels in foster parents, a score equal to or above 250 on the PSI was set to identify a clinical level of stress. Regression analyses were used to explore whether the time since the last visit with birth parents influenced CBCL scores (i.e. total competence and total problems) and foster parents' PSI scores. Visits with birth parents were categorized as follows: frequent (high) referred to monthly or more frequent visits; less frequent (low) indicated visits less often than once a month; and no visitation (none) referred to none at all. F-tests and χ 2 tests were used to test for differences on continuous and categorical variables. Welch's t-tests were used as a corrective for variables with unequal variances.

Amount of visitations
Slightly less than half of the children (47.8%) had monthly or more frequent visit with their mothers, while fewer fathers (21.6%) visited as frequently. Birth mothers were not allowed to take the children out unsupervised in 53 (29.0%) cases, while the same held true for fathers in 165 (81.5%) cases. Relatively few mothers (7.4%) had no visits at all with their child, compared to a fairly large proportion of the fathers (41.2%). The reasons for no visit taking place varied. In only one case (0.5%), had the court ruled against visiting with the mother, while three mothers (1.4%) had died and seven others (3.2%) did not show up for their scheduled visits. The corresponding figures for fathers were 10 (4.6%), 19 (8.8%), and 5 (2.3%), respectively. Table 1 presents data on the number of visits of both mothers and fathers.
As far as associations between the number of days since previous visit and the mental health of the children are concerned, none of the CBCL variables were significant. For birth mothers, neither the CBCL total competence nor CBCL total problems scores were significant (b = 0.01, t-value = 1.83 and b = −0.03, t-value = −0.59, respectively), nor were the associations for visits with fathers and CBCL (b = −0.01, t-value = −0.88 and b = 0.00, t-value 0.05). The level of foster parent stress measured by PSI total stress were not significantly related to visits with birth mothers (b = −0.08, t-value = −0.99) or birth fathers (b = −0.06, t-value = −0.54).

Associations between number of visits and child and foster care characteristics
With regard to child characteristics, more girls than boys had frequent visits with their fathers. A larger proportion of the children who were in contact with their birth mothers less than once a month identified their foster parents as their main attachment figures. In other respects, the variables concerning child characteristics were insignificant. With regard to foster family characteristics, the distance in kilometers from the birth mother's dwelling was significantly shorter in those children with more frequent visits than in those with less frequent visits (see Table 2). In other respects, none of the variables differed significantly with regard to visits with parents. This is true for the variables of attachment, both on PSI attachment and perceptions of the foster parents being the prime attachment figure, foster parental levels of stress or child characteristics being a girl, child age, age moving from birth parents, and time spent in the present foster home. Table 2 presents data on the associations between high vs. low frequency of visits with birth parents and the characteristics of both the children and their foster families.

Table 1. Frequency of visitations with birth parents
Note: For mothers n = 203 and for fathers n = 202. A total of 20 (15.3%) foster parents reported stress levels ≥250 on the PSI but this was not related to visits with birth mothers: 10 (15.4%) from the sample with monthly or more frequent visits, six (10.9%) in the group with visits less often than once a month, and four (36.4%) in the group with no visits. This difference is not significant χ 2 (2) = 4.6, ns. For visitations with birth fathers, the distribution was five (15.6%), seven (13.2%), and 10 (16.9%), respectively, producing in a non-significant χ 2 (2) = 0.3.

Variance in frequency of visitations and mental health problems among the children
Analysis of the difference in frequency of visits with both birth mothers and fathers (visits equal to or more frequent than once a month vs. less frequent than once a month) and children's mental health or competency scales on the CBCL did not indicate any relationship between children's mental health and frequency of visitations. This was true for all eight sub-scales, the two sub-domain scores, the total score, the three competencies scales, and the total competence score as reported by the foster parents'. Table 3 presents detailed information on frequency of visitations with mothers and fathers and detailed information regarding the CBCL scores.

Table 2. Associations between frequency of visits with birth parents and child and foster care characteristics
Note: High ≥ monthly, Low < monthly. None = no contact; PSI = Parent Stress Index. *p < 0.05. **p < 0.001. A total of 61 (27.9%) children scored in the borderline/clinical range on the CBCL total score, i.e. t-score ≥ 60. When we examine the children in the clinical range in relation to visits with their mothers, 28 (13.8%) had monthly or more frequent visits, 26 (12.8%) had visits less than once a month, and seven children (3.4%) had no visitation with their mothers. This difference is not significant χ 2 (2) = 2.2. With regard to visits with fathers, seven (3.5%) children who scored in the clinical range on the CBCL had frequent visits, while 27 (13.4%) had fewer visits than once a month and a further 27 (13.4%) did not have any visits with their birth father. This distribution is not significant χ 2 (2) = 4.9.

Discussion
The purpose of this study was to explore differences between young children living in the same foster care setting for at least one year and visits with birth parents. Birth parent visits were either monthly and more frequent or less frequent than this and outcome variables were the children's mental health and competencies, who was the main attachment figure to the child and foster parental levels of attachment to the child and finally total level of foster parent stress. The 203 children Table 3

Birth mother
Externalizing 8.2 (7.2) 10.5 (9.7) 13.7 (12.6) 3.10 8.1 (7.2) 9.6 (8.4) 10.7 (9.6) 1.22 were on average 2.3 years when they were placed for the first time in foster care, and their living situation had been stable since moving to their current foster home. The placement alternatives included both kin and non-kin foster homes. Ninety-seven (47.8%) of the birth mothers had monthly or more frequent visits, while 43 (21.6%) of the birth fathers had visits that often. An overwhelming majority of the children (81-93%) experienced a foster parent being the main attachment figure according to the foster parents. Our most striking finding was the similarities between children who had high as opposed to low numbers or no visits with their birth parents. This was the case for the children's psychosocial functioning and competence, attachment-including who was the main attachment figure and the level of foster parental rating of attachment on the PSI, and foster parents' levels of stress. Nor were there significant differences relative to the number of visits with birth parents, between children scoring in clinical levels (27.9% of the children) or not on the CBCL and foster parents levels of stress reported by the results on the PSI. These results indicate scores above (15.3% of the foster parents) or below clinical cut-off on the children's psychosocial functioning and the foster parents' levels of stress. A similar finding that visits did not seem to influence children's psychosocial functioning was also reported in a study of 64 slightly older foster children conducted in USA (Rich, 2010). In our sample the girls had more frequent contact with their fathers than the boys, but a large proportion of girls had no contact at all with their fathers. This distribution was significant, but the reasons for this distribution is not known. With regard to contact with birth mothers, there were no differences between boys and girls.
It is important to take into consideration that the children were very young when first placed outside their birth homes, and that they had been living in stable foster relationships for sometime when data were collected. These factors are probably relevant to the finding that most children did not display clinical levels of mental health problems. In this sample, less than 1/3 of the children scored above the 60th percentile on CBCL total score. Placement instability increases the risk of more mental health problems (Newton, Litrownik, & Landsverk, 2000;Rubin et al., 2004), and in children with elevated levels of behavioral problems, the risk of placement breakdown increases (Oosterman, Schuengel, Wim Slot, Bullens, & Doreleijers, 2007). We found that psychosocial functioning was not related to the number of visits with birth parents. Nevertheless, externalizing problems, that is aggression, opposition, delinquency and more, were more evident among the children who had little or no contact with their birth mothers than among those with more frequent visits, but none of these variables were significant when adjusted for the unequal variance in the data.
Attachment is regarded as important because it reflects the relationship between the child and the caregiver early in the child's development as well as being associated with the child's social functioning later in life (Dozier, Stoval, Albus, & Bates, 2001). In a literature review, Haight et al. (2003) strongly argued for consideration of the quality of attachment when deciding on visits with birth parents. In view of this, the finding that children's attachment to foster parents was not influenced by the number of visits with birth parents is of importance. This was true for both the foster parents' own reports on whether or not they served as the main attachment figure, and for the analysis of attachment quality between foster parent and child on the PSI. Furthermore, the frequency of visits did not appear to influence the level of stress in the foster parents. In all, 84.7% of the foster parents reported non-clinical levels of stress. This suggests that the distribution of stress in foster parents is similar to that in the general population. Interestingly, and possibly related, Fuentes, Salas, Bernedo, and Carcia-Martin (2015) reported that slightly more than 90% of foster parent reported no burdens in relation to the role of being a foster parent, but those reporting of worries stated concerns regarding the children's future.

Recommendations for practice
We would like to suggest some possible recommendations for child welfare agencies based on the results in this study. It is pertinent to point out that the findings in this study primarily maybe generalized to cases where children were young at placement and lived in stable foster home arrangements. Our findings suggest that child welfare workers should consider the quality and safety of interaction between children and parents during visits. When interaction between child and parents during visits are considered to be sufficiently safe and visitation rights are thereby granted according to standards set by courts, there is no particular reason to assume that the frequency of visitations has a significant impact upon children's mental health or adjustment to the foster home. One implication of this is that there should be some scope for adapting visiting arrangements toward children's wishes and to take the children's perspectives and wishes into account. This is strongly argued for in a study of 21 adolescents growing up in kinship foster care advising not to push for more or less contact with birth parents or the extended family, but instead listen more carefully to their perspectives (Kiraly & Humphreys, 2013). This is possible for younger children too. In 151 cases in Norway with 10-year-old children, 55.1% of the foster children wished for more visits with their mothers, 40% wished for more visits with the fathers, while the rest did not want to change their present arrangements with the birth parents (Vis & Fossum, 2013). In cases with younger children, the child's reactions could be monitored by observing smiles, discomfort and gesticulations. For instance, if a child expresses joy and pleasure of seeing his or her mother, this could be an indication. This is not to state that children's opinions automatically should be taken into account, but their perspectives should be appreciated at least.
Practicalities and arrangements of visits influence all parties involved and need to be taken into account. In our sample, there was a significant difference in distances from the birth home and frequency of visits with birth mothers. Those birth mothers living closer to the foster home visited the child more often than mothers living further away. As young children grow up it is important to ensure that family networks and social support for adolescents are established and can be maintained when they leave foster care (Hedin, 2014;Singer, Berzin, & Hokanson, 2013). This further highlight the importance of taking children's perspective into account when frequency of visitations are determined since many foster parents report being concerned for their foster child's social functioning when he or she turns 18 .

Limitations
This study was not primarily designed to measure the influence of visitations with biological parents on child mental health, attachment, and foster parent stress. As a consequence, cases were not selected to represent the distribution of visitation frequency found in the population. This limit the generalizability of the findings. Future studies should be designed with this in mind. Furthermore, the research was not randomized or conducted in a fashion that ruled out the possibility of unknown confounders that could potentially bias the results. We assume that foster home availability, various safety issues, and children's special needs were important factors for placement that also impacted the frequency of visitations. Consequently, it is important to exercise caution in interpreting the study results.

Conclusions
For young children living in stable foster care settings, the frequencies of visits with birth parents do not seem to negatively influence the children or the foster parents. That is, neither children's mental health, children's level of competencies, who is the perceived main attachment figure or foster parents' attachment to the child, nor foster parents' level of stress, are significantly associated with visits with birth parents. One possible implication of the findings is that children's developmental needs should not be the only consideration when frequency of visitations are determined. The findings indicate that some latitude should be given to adjust frequency of visits according to the children's own views. It is vital to consider the value of such visits both in short-and long-term perspectives for the children when deciding the frequency and duration of the visits.

Funding
This work was supported by the UiT The artic university of Norway.