Birth Family Contact from Childhood to Adulthood: Adjustment and Adoption Outcomes in Adopted Young Adults

Abstract

Experiences of contact between adopted persons and birth family members have implications for psychological adjustment of adopted persons. The current study utilizes four contact trajectory groups, spanning from middle childhood to young adulthood and encompassing three aspects of birth family contact, in predicting psychological adjustment and adoption-related outcomes in adopted young adults. Data come from a longitudinal study of adoptive families in which adopted persons were domestically adopted in infancy by same-race parents in the United States. Adopted young adults in the group characterized by sustained high levels of contact and satisfaction with contact over time (‘Extended Contact’) displayed lower levels of psychological distress and higher levels of psychological well-being than adopted persons in the group characterized by contact that increased over time but remained limited (‘Limited Contact’). Generally, adopted persons within the group characterized by consistent lack of contact (‘No Contact’) and the group characterized by contact that was initially present but ended (‘Stopped Contact’) did not differ in distress and well-being from those in the ‘Extended Contact’ group. No group differences were found on adoption dynamics and identity, however young adults in the ‘Extended Contact’ group generally reported more positive relationships with their birth mothers than those in the other groups. Findings are discussed in the context of heterogeneity in contact experiences and implications for policy and practice.

Most adoptions of children in mid-20^th century western countries involved the legal transfer of parental rights from birth parents to adoptive parents, at which time the identities of birth parents and adoptive parents were hidden from each other, and the child’s birth certificate was altered to indicate that the adoptive parents were the child’s parents of record. Since the 1970s, there has been movement toward sharing identifying information between adoptive and birth family members and permitting contact among them after placement. This trend has been pronounced in the United States, most strongly with domestic infant adoptions, to some degree in adoptions from the child welfare system, and less so with international adoptions (Grotevant, 2020). The trend has also gained traction in New South Wales (Australia), the United Kingdom, and New Zealand. Some European countries are also considering possibilities for contact. Although the movement toward more openness in adoption is global, changes are occurring unevenly across countries (see Grotevant, 2020, for review). How are such changes in contact arrangements related to outcomes for adopted persons, not just during childhood, but into adulthood?

Psychological adjustment of adopted persons has long been examined in the adoption literature. Meta-analyses indicate that adopted children are at increased risk for adjustment difficulties when compared to nonadopted children (Juffer & van IJzendoorn, 2005; Wierzbicki, 1993). These increased levels of adjustment difficulties have commonly been attributed to maltreatment prior to adoption, such as with the case of children who were previously institutionalized or removed from their birth homes by social services (Bos et al., 2011; Grotevant et al., 2006). However, there may also be an increased risk for adopted children who have not experienced significant pre-adoption maltreatment, such as some children domestically adopted in infancy through private agencies (Howard et al., 2004). Furthermore, post-adoption factors may be more predictive of adjustment challenges than pre-adoption adversity, as has been consistently found for many internationally and transracially adopted youth who experience racial discrimination post-adoption (Kim & Lee, 2020).

While much focus has been placed on childhood adjustment, examinations of adopted persons in adulthood are less common. Research suggests that differences in adjustment between adopted and non-adopted persons disappear over time (Palacios & Brodzinsky, 2010). However, a systematic review of studies of domestically adopted adults with varying levels of pre-adoption risk concluded that adopted adults were at higher risk for psychological difficulties than their nonadopted counterparts (Melero & Sanchez-Sandoval, 2017). Increased levels of psychological difficulties in adulthood have been attributed to post-adoption factors. For example, challenges in adoptive identity development predicted higher levels of internalizing difficulties in adults domestically adopted in infancy (Grotevant et al., 2017). Furthermore, consistent with findings on adopted youth, post-adoption experiences may be more predictive than pre-adoptive adversity for some adopted adults. This was the case for a sample of adopted adults previously institutionalized in Hong Kong, for which post-adoption family experiences appeared more predictive of adjustment than orphanage experiences (Rushton et al., 2013). Overall, the heterogeneity of experiences seen among adopted persons requires the study of multiple pre- and post-placement factors that may predict psychological adjustment. Within-group examinations emphasizing the heterogeneity of adoption experiences may prove more revealing than simple comparisons between adopted and nonadopted persons (Palacios & Brodzinsky, 2010).

Post-Placement Birth Family Contact

During the 1970s and 1980s, some critics of birth family contact hypothesized that connections with birth relatives would lead adopted children to feel confused as to who their “real parents” were (Kraft et al., 1985). Confusion would then have negative implications for psychological adjustment. A contrasting hypothesis is that birth family contact would have positive implications for adjustment, as contact allows an adopted person to access information about their own birth history that could be incorporated into their sense of self (Von Korff & Grotevant, 2011). Narrative theories of identity development emphasize the importance of creating coherent stories about one’s life in predicting positive adjustment in adulthood (McAdams, 2001; Waters & Fivush, 2015). Among adopted persons, information important for such narratives is often unknown but desired. For example, many adopted adolescents desire information about the circumstances of their placement (Wrobel & Dillon, 2009). Absence of such information may lead to formation of negative (and perhaps incorrect) conclusions about birth history (Dunbar & Grotevant, 2004). In adulthood, adopted persons from all adoptive backgrounds consistently report their desire for health history, in addition to the difficulties that arise from lack of such information (e.g., May et al., 2018). Adopted adults experience distress over what health challenges may be present in their genetics, shame when encountering healthcare professionals without such information to report, and a sense of loss in their understanding of their own histories (May et al., 2018). Birth family contact provides opportunities for all this information to be shared (e.g., Siegel, 2008; Von Korff & Grotevant, 2011), potentially alleviating distress and promoting well-being. Of note, adopted persons could gain negative information about birth history, birth family members, or health, contributing to negative feelings about the self and increased distress (Grotevant et al., 2017). Thus, in examining the relation between contact and adjustment, subjective experiences of the contact also need to be considered.

Much of what is known about birth family contact comes from studies of families who adopted domestically from private and public agencies. Among such families, limited evidence exists for a simple relation between birth family contact and behavioral difficulties in adopted persons, at least during childhood (Neil, 2009; Von Korff et al., 2006). Adopted persons who are less satisfied with their levels of birth family contact may feel they have less control over contact and less access to important information about themselves (Berge et al., 2006). Such dissatisfaction may then contribute to more adjustment difficulties. For example, among adoptees who were domestically adopted in infancy, higher family satisfaction with contact predicted lower levels of externalizing behavior problems in adolescence, and these positive implications continued into emerging adulthood (Grotevant et al., 2011). Thus, adopted persons’ subjective feelings about birth family contact may be more relevant to psychological adjustment than objective variables such as presence and frequency of contact.

For adoptive families of children domestically adopted in infancy, contact arrangements may change significantly throughout childhood, such as by becoming more open or discontinuing altogether (Grotevant & McRoy, 1998; Grotevant et al., 2004). As adopted children become more autonomous in adolescence, they may have increased responsibility in initiating and maintaining contact (Dunbar et al., 2006). In addition, the removal of legal barriers to adoption records in emerging adulthood allows adoptees opportunities to establish contact with birth relatives for the first time (Howard et al., 2010; Wrobel et al., 2013). Adopted persons’ satisfaction with contact may fluctuate with changing levels of contact, depending on the reasons for such changes and whether they are desired. For persons domestically adopted in infancy, satisfaction with contact generally relates to level of contact, such that adopted children and adults who have contact are more satisfied with their openness arrangements than those who do not (Mendenhall et al., 2004; Farr et al., 2014). Additionally, during adolescence, many who express dissatisfaction with contact appear more likely to want contact to increase than remain the same (Grotevant et al., 2007), suggesting that current satisfaction may also have implications for future levels of contact. For example, increased contact may result from intense curiosity about unknown information related to birth family members contributing to dissatisfaction (Grotevant et al, 2007; Wrobel et al., 2013). The implications of increases in contact for the psychological adjustment of adoptees are relatively unknown, as increased contact cannot guarantee the development of satisfying relationships or obtaining of desired information. Concerning decreases in contact, adolescents domestically adopted in infancy whose birth family contact had stopped were less satisfied with their contact arrangements than adolescents who had ongoing contact with birth parents that included in-person meetings (Grotevant et al., 2007). Such findings remained consistent through emerging adulthood (Farr et al., 2014). When adopted persons lose birth family contact in ways that are outside of their control, they may lose relationships that they valued and sources of important information about themselves. Examinations of the long-term implications of such losses are needed.

Trajectories of Birth Family Contact

Grotevant and colleagues (2019) used group-based trajectory modeling (GBTM; Nagin, 2005) to develop trajectories of birth family contact based on three contact variables informed by the literature: 1) Frequency of contact, defined as how often the adopted person had contact with their birth mother within the past year; 2) Satisfaction with contact, defined as how satisfied the adopted person currently was with contact with their birth mother; and 3) Participation in cross-network contact, defined as the sum of the number of persons in the adoptive family who had contact with any member of the adoptee’s birth family, and the number of persons in the adoptee’s birth family who had contact with any member of the adoptive family, within the past year. Trajectories reflected these variables across four waves, spanning from middle childhood to young adulthood. GBTM analyses identified four trajectory groups based upon the three contact variables (Figure 1), with each trajectory group representing an “average” contact experience across group members. In our presentation of the trajectory groups, we first describe the prototypical pattern of contact revealed by the quantitative data. We then describe themes that emerged when we previously randomly selected eight cases from each group and examined interview transcripts across the waves of data (Grotevant et al., 2019).

Figure 1.

Four multi-group trajectories of birth family contact based on frequency of contact with the birth mother, satisfaction with contact with the birth mother, and participation in cross-network contact (from Grotevant et al., 2019). BMO = Birth Mother. n = 190

Adopted persons in the ‘No Contact’ group (41.6% of the sample) typically had no contact with birth family members over time, associated with a linear decline in satisfaction with lack of contact. Some adopted persons expressed an interest in contact to fill gaps in their health history. Others expressed an openness or desire to have contact with birth family members, but with some degree of ambivalence about how much contact they would like. Adopted persons in the ‘Stopped Contact’ group (13.7%) typically had some contact with birth family members initially that decreased and ultimately stopped. Satisfaction also declined and leveled off at a moderate level. For some group members, contact appeared to fade over time for various reasons, including moves and busy lives. For others, it ended more abruptly due to conflicts between adoptive and birth family members. Adopted persons in the ‘Limited Contact’ group (26.3%) typically experienced increased frequency of contact with their birth mother and a level of cross-network contact that increased slightly but remained small. Satisfaction began high and decreased over time, rebounding slightly in young adulthood. Some group members emphasized how relationships with birth family members remained limited; some were satisfied with this situation, whereas others desired more contact. Others cited challenges in navigating increases in contact, such as managing differing expectations about level/intensity of desired contact. Lastly, adopted persons in the ‘Extended Contact’ group (18.4 %) typically reported high frequency of birth mother contact over time and a level of cross-network contact that began high and increased. Satisfaction began and remained high. Group members considered birth relatives as extended family members and noted continued interactions between adoptive and birth family members as “normal.” Also mentioned was the relational work needed to maintain contact throughout the years, such as the need for ongoing communication and consideration of others’ feelings.

The Current Study

The current study examines the predictive effect of the birth family contact trajectories on adoptees’ psychological adjustment and adoption-related outcomes in adulthood. Previous examinations have been limited in their timeframe and ability to consider multiple aspects of contact simultaneously. The current study addresses these limitations through a person-centered approach: trajectories of birth family contact incorporating three features of contact (frequency of contact, satisfaction with contact, and participation in cross-network contact) across four time points (middle childhood, adolescence, emerging adulthood, and young adulthood) were used in examining the relation between contact and outcomes in a sample of young adults who were domestically adopted in infancy.

Higher satisfaction with contact predicts more positive psychological outcomes, and the limited findings pertaining to actual level of contact have favored contact over none (Grotevant et al., 2011; Von Korff et al., 2006). ‘Participation in cross-network contact’ has not been previously examined in the context of adjustment and adoption outcomes, despite its being a unique source of variability in contact experiences (Grotevant et al., 2019). Stable high contact with multiple birth family members and consistently high satisfaction would presumably reflect ongoing opportunities to gain information that could be incorporated into adoptees’ sense of self. Although contact may lead to some adopted adults gaining negative information about their histories that contribute to distress (e.g., perceived rejection by a birth parent), one would not expect those adoptees to continue having high levels of satisfying contact with birth family members across time. Therefore, it was hypothesized that adopted young adults in the ‘Extended Contact’ group, characterized by consistently high levels of contact, high satisfaction with contact, and large adoption kinship network, would have lower levels of psychological distress, higher levels of psychological well-being, and more positive adoption-related outcomes, than young adults in the other three groups. To test these hypotheses, pairwise comparisons between the ‘Extended Contact’ group and the other three groups were planned.

Method

Participants

Data came from 190 adoptive families (child gender: 47% female, 53% male) who participated in a four-wave longitudinal study of adoptive family dynamics in the United States. Families had varying levels of birth family contact and were recruited through 35 private adoption agencies across 23 states. Children were domestically placed before the age of one (mean age of placement = 4 weeks) into same-race (93.2% White, 2.1% Hispanic/Mexican American, 0.5% Black/African American, 4.2% unknown), heterosexual-parented families. No adoptions were special needs, and parents within each family were married at the time of recruitment.

The current study drew from four waves of data spanning childhood to young adulthood: Wave 1 (W1: 1986–1992; n = 190 families; child M age = 7.81 years, range = 4 to 12 years); Wave 2 (W2: 1996–2001; n = 177 families; adolescent M age = 15.73 years, range = 11 to 20 years); Wave 3 (W3: 2005–2008; n = 181 families; emerging adult M age = 24.95 years, range = 20 to 30 years); and Wave 4 (W4: 2012–2014; n = 112 young adults; young adult M age = 31.43 years, range = 26–38 years). Participants in these analyses included adoptive mothers at W1 (n = 190) and adopted persons at W2 (n = 156), W3 (n = 169), and W4 (n = 112). For full information on participants in the larger study, see Grotevant et al., 2013.

Procedure

At W1 and W2, data were collected during visits to adoptive families’ homes. Data collection lasted three to five hours and included interviews with the adoptive mother, adoptive father, and adopted person; administration of questionnaires; a couple’s interview (W1); and a family interaction task (W2). When home visits were not possible, interviews were conducted by phone and questionnaires were mailed. At W3, adopted emerging adults were interviewed and administered questionnaires online. At W4, adopted young adults completed questionnaires online. Procedures were approved by the Institutional Review Board at the University of Texas Austin for W1 (approval date: June 2, 1986), University of Minnesota for W2 and W3 (protocol number: 8909S2152), and University of Massachusetts Amherst for W4 (protocol number: 2011–1159). In addition, coding as part of the development of the trajectories was approved by the Institutional Review Board at Bethel University (protocol number: FA-21–19).

Measures

Psychological Adjustment

Psychological Distress.

Adopted young adults’ psychological distress at W4 was measured using the Brief Symptom Inventory (BSI; Derogatis, 1993), a 53-item measure with nine subscales: anxiety, depression, somatization, obsessive-compulsive, interpersonal sensitivity, phobic anxiety, paranoid ideation, and psychoticism. Each item is rated on a 5-point scale (0 = “Not at all”; 1 = “A little bit”; 2 = “Moderately”; 3 = “Quite a bit”; 4 = “Extremely”). Responses on each item are summed to create a global severity index (GSI) representing overall psychological distress. The GSI was utilized in the current study. The BSI has demonstrated excellent reliability and validity (Derogatis, 1993). Internal consistency for the GSI in the current sample was α = .98.

Psychological Well-Being.

The Ryff Scales of Psychological Well-being (Ryff, 1989) medium form was used to assess psychological well-being at W4. The measure contains 54 items separated into six, 9-item scales. The current study utilized three scales: autonomy, environmental mastery, and positive relations with others. Scales were chosen for relevance to dimensions described in influential models of adult development, mainly the dimensions of life skills and healthy relationships (Scales et al., 2016; Wright et al., 2023). Participants responded to each item using a 6-point Likert scale ranging from 1 = “strongly disagree” to 6 = “strongly agree”. Responses on items within each scale were summed to create total scores. Internal consistencies in the current sample ranged from α = .80 to α = .89.

Adoption Outcomes

Adoptive Identity.

The Adoptive Identity Scale (AIS; Lee, 2007) was used to measure adoptive identity at W4. The AIS includes 14 items rated using a 5-point Likert scale ranging from 1 = “strongly disagree” to 5 = “strongly agree”. AIS items capture multiple aspects of adoptive identity, including how much respondents identify with other adoptees/the adoption community (e.g., “I have a lot of pride in adoptees and their accomplishments”), general feelings about being adopted (e.g., “I am happy that I am adopted”) and adoption’s connection with participants’ sense of self (e.g., “Being adopted is an important reflection of who I am”). Responses on items were summed to create a total score; internal consistency was α = .84.

Adoption Experience.

The Adoption Dynamics Questionnaire (ADQ; Benson et al., 1994) was used to measure feelings and experiences pertaining to adoption at W4. Two subscales of the ADQ were employed: Positive Affect about own adoption (PA; 20 items), and Negative Experiences with own adoption (NE; 7 items). Items are rated on a 5-point Likert scale from 1 = “not true/strongly disagree/never” to 5 = “always true/strongly agree/always”. Example items include “I like the fact that I’m adopted” (PA) and “I get teased about being adopted” (NE). Mean scores were created for each scale. Internal consistencies for the PA and NE subscales were α = .91 and α = .59, respectively.

Feelings about Birth Mother.

Three measures assessed adopted young adults’ feelings towards their birth mothers at W4. Measures were created through exploratory factor analysis on adoption-related items completed by adopted young adults. Items were based on themes seen in interviews with the adopted persons at earlier waves of the study. Items are rated on a 5-point Likert scale from 1 = “strongly disagree” to 5 = “strongly agree”. The first measure captures adoptee’s desirability of contact with their birth mother and consisted of four items, including “I have good feelings about my birth mother” and “Having contact with my birth mother is desirable”. The second measure captures adoptee’s satisfaction with aspects of the relationship with their birth mother and consisted of three items, including “I am satisfied with the amount of control I have over the contact with my birth mother” and “I am angry about my lack of relationship with my birth mother” (reverse coded). The third measure captured adoptee’s feelings about being placed for adoption by their birth mother and consisted of two items: “I am grateful to my birth mother for placing me for adoption” and “I have negative feelings about the fact that my birth mother placed me for adoption” (reverse coded). Mean scores were created for each of the three scales. Internal consistencies ranged from α = .75 to α = .88.

Trajectory Variables

The three contact variables used in the creation of the four contact trajectory groups were coded from semi-structured interviews with members of the adoptive family at W1–W3 and from responses on questionnaires completed by adopted young adults at W4. At W1, variables were coded from interviews with the adoptive mother, as adopted children were considered too young to be accurate reporters. At W2 and W3, variables were coded from interviews with the adopted person. Frequency of contact was coded as follows: 0 = no current contact, 1 = rarely (less than once per year), 2 = occasionally (once or twice per year), 3 = often (3–11 times a year, or once a year for an extended visit), and 4 = frequently (once a month or more, or more than one extended visit a year). “Extended visits”, which were those that lasted multiple days, were considered in the frequency variable because for some adoptees, less frequent but longer visits were more logical than multiple shorter visits (e.g., adoptee lived far away from the birth family). Satisfaction with contact was coded as follows: 0 = very dissatisfied, 1 = dissatisfied, 2 = neutral (or approximately balanced between dissatisfied and satisfied), 3 = satisfied, and 4 = very satisfied. For the W1 interviews with the adoptive mothers, coders focused on adoptive mothers’ reported observations of their child’s satisfaction with the contact. For participation in cross-network contact, members of the adoption kinship network (e.g., adopted person, adoptive mother, adoptive father, birth mother, birth father, birth sibling, etc.) were coded as “yes” or “no”, with yes indicating that the member participated in contact. The number of “yes” responses were then summed. Participation in cross-network contact at W4 was not included in the trajectories, as directly comparable data to create this variable were not available. For more information on the contact variables and development of trajectories, see Grotevant, et al., (2019).

Data Analyses

To examine whether trajectory group membership significantly predicted outcomes at W4, the study utilized fixed-effect modeling through the MIXED procedure in SPSS 25. The MIXED procedure accounts for missing data through maximum likelihood estimation (ML/FIML estimation; Schafer & Graham, 2002), which was preferred as FIML estimation had been utilized to address missing data during the development of the contact trajectories (Grotevant, 2019). The procedure also produces omnibus F-tests for models, which were used to account for multiple analyses. Specifically, significance on the omnibus F-tests suggested confidence in significant pairwise differences on an outcome measure. Significance of pairwise differences were examined using an adjusted p-value of .0167 to account for the three pairwise comparisons made per model. Outcome variables were initially screened for normality and determined to be acceptable.

Contact trajectories were developed utilizing data from all 190 adoptive families across the four waves of the study, whereas outcomes were measured at W4. The sample at W4 included 112 adopted persons out of the 190 whose families were recruited at W1. Proportions of available data on W4 outcome variables ranged from 54% to 58% out of the 190 families. When using FIML estimation for large amounts of missing data (e.g., 50 percent or greater), bias in parameter estimates does not appear to occur in a consistent direction; however, standard errors may be inflated (Dong & Peng, 2013; Schlomer et al., 2010). In such situations, auxiliary variables related to missingness are typically included in models to support the assumption that data are missing at random (MAR; Dong & Peng, 2013; Nicholson et al., 2017). Furthermore, inclusion of auxiliary variables significantly related to variables with missing data can improve FIML estimation’s accuracy in estimating missing data (Dong & Peng, 2013). Given the proportion of missing data in the current study, preliminary analyses involved determination of auxiliary variables to include in the models. For the selection process, several variables were tested to see if they related to missingness at W4. Focus was placed on important demographic variables as well as variables expected to be significantly related to the W4 outcomes.

In examining relations to missingness, comparisons were made between adopted persons from W1 families who participated at W4 and those who did not. Women were more likely to participate than men at W4; χ2(1) = 6.984, p = .008, r=.19 (W4 gender distribution: 55% female, 45% male). No differences were found for age of the adopted person at W1, level of birth family contact at W1, or education levels of participants’ adoptive parents at W1. Multiple indicators of psychological adjustment spanning W1 through W3 were also examined, and no indicators were found to consistently relate to both missingness and W4 outcome variables. Lastly, three measures of the ADQ at W3 were examined: positive affect with adoption, negative experiences with adoption, and preoccupation with adoption. Those who participated at W4 reported significantly lower levels of negative experiences with adoption at W3 than those who did not; t(79.52) = 2.12, p = .037, r = −.19 (W4 participants: M = 10.51, SD = 3.10; Non-W4 participants: M = 11.94, SD = 4.40). Level of negative experiences at W3 was significantly related with the following W4 outcome variables: Positive Relations with Others, Adoptive Identity, Positive Affect with own Adoption, Negative Experiences with own Adoption, and Feelings about being Placed for Adoption (r’s = .30–.64).

In choosing auxiliary variables, focus was placed on balancing the need to include variables that would maximize the efficiency of FIML estimations, with the desire to produce interpretable models. Gender was selected due to being the only variable related to missingness that had data for all 190 cases. In addition, negative experiences at W3 was selected due to being related to missingness as well as multiple outcomes of interest. Both variables were included as covariates in all models. Lastly, adopted young adults at W4 ranged widely in age (26 years to 38 years) and likely differed in ways related to their life stages (e.g., occupation level, financial stability, independence from adoptive family). Such factors could theoretically capture meaningful variability in the adjustment and adoption outcomes (e.g., Niemeyer et al., 2019). Thus, adopted young adults’ age and highest level of education at W4 were also included as covariates.

Results

Correlations among study variables as well as overall means and standard deviations are presented in Table 1. Results from fixed-effects models are presented in Table 2.

Table 1.

Bivariate correlations for covariates, auxiliary variables, and W4 outcome variables.

Variables	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1. Age (W4)	-
2. Education (W4)	.00	-
3. Gender	−.06	.13	-
4. Negative Exp (W3)	−.35**	−.10	.03	-
5. BSI GSI	.11	−.12	.14	.17	-
6. Autonomy	.04	.06	−.15	−.08	−.45**	-
7. Env. Mastery	−.09	.19	.03	−.06	−.67**	.42**	-
8. Positive Relations	.09	.12	.01	−.30**	−.53**	.50**	.62**	-
9. AIS	−.01	.14	.18	−.31**	−.18	.24*	.26**	.40**	-
10. Positive Affect	.14	−.00	−.06	−.47**	−.29**	.11	.20*	.29**	.48**	-
11. Negative Exp. (W4)	−.14	−.12	.07	.64**	.27**	−.19	−.13	−.30**	−.42**	−.51**	-
12. Desirability	.02	−.14	−.04	−.05	−.13	.06	.08	.15	.21*	.18	−.07	-
13. Satisfaction	−.11	.12	−.12	−.03	−.14	.04	.25*	.18	.05	.19	−.09	.23*	-
14. Being Placed	.03	.03	−.17	−.33**	−.39**	.13	.31**	.39**	.31**	.63**	−.38**	.26**	.35**	-
Mean	31.43	7.14	-	1.57	51.02	40.00	40.09	44.19	43.69	4.26	1.52	3.80	3.84	4.36
Standard Deviation	2.05	1.64	-	.52	14.25	7.44	8.75	7.76	5.93	.72	.42	.97	.99	.77

Note. Total sample size at W4 was n = 112. Gender was available for all 190 adopted persons. Negative Experiences at W3 was available for 155 adopted emerging adults. Negative Exp = Negative Experiences; BSI GSI = Brief Symptom Inventory Global Severity Index; Env Mastery = Environmental Mastery; AIS = Adoptive Identity Scale.

^**p < .01;

^*p < .05

Table 2.

Adjustment and adoption outcomes by trajectory group membership.

				Trajectory Group M(SD) [Min, Max]
Outcome Variable	F-value	Pseudo R2	No (n = 79)	Stopped (n = 26)	Limited (n = 50)	Extended (n = 35)
BSI GSI	4.52**	.26	49.00(14.67) [30, 80]	50.75(14.13) [30, 80]	56.78(13.35)* [35, 80]	47.75(13.60) [30, 80]
Autonomy	2.91*	.12	40.55(7.14) [27, 54]	38.19(9.19) [18, 53]	37.82(6.02)* [23, 48]	42.92(7.42) [30, 54]
Env. Mastery	2.30	.22	40.43(9.11) [23, 52]	37.88(10.43) [17, 50]	38.68(8.27) [21, 52]	42.71(7.24) [26, 54]
Positive Relations	6.21**	.30	45.19(8.01) [26, 54]	41.75(7.72)* [29, 53]	41.61(7.65)* [26, 54]	47.33(6.36) [35, 54]
Adoptive Identity Scale	.24	.20	42.75(6.32) [27, 53]	44.44(5.23) [35, 56]	44.23(6.27) [30, 55]	43.87(5.55) [32, 55]
Positive Affect	.67	.20	87.89(10.26) [64, 100]	83.67(16.20) [51, 100]	84.00(16.69) [36, 100]	83.96(15.86) [29, 100]
Negative Experiences	1.36	.50	10.73(2.34) [7, 15]	9.52(2.20) [7, 15]	10.94(2.92) [7, 19]	10.92(3.97) [7, 21]
Desirability	9.64**	.26	13.03(3.43)* [6, 19]	15.18(3.63)* [7, 20]	15.59(3.55)* [10, 20]	18.08(3.17) [5, 20]
Satisfaction	5.23**	.15	10.36(3.13)* [3, 15]	11.71(2.66) [8, 15]	11.21(3.08)* [4, 15]	13.50(1.72) [10, 15]
Being Placed	2.04	.17	8.35(1.53) [6, 10]	9.12(1.54) [4, 10]	8.55(1.74) [4, 10]	9.25(1.11) [6, 10]

Note. Group sizes are based on all 190 cases whereas means and standard deviations are based on available data. An asterisk on a group mean/standard deviation indicates significant difference from the Extended Contact group. All models include gender, W3 negative experiences, W4 age, and W4 education as covariates. Reported F-values are for the effect of trajectory group membership on the outcome variable. Pseudo R² values represent comparisons between intercept-only models and models with trajectory group and covariates/auxiliary variables as predictors. BSI GSI = Brief Symptom Inventory Global Severity Index; Env. Mastery = Environmental Mastery.

For F-values: **p < .01, *p < .05;

For pairwise comparisons: *p < .0167.

Psychological Adjustment

A T-score ≥ 63 on the GSI of the BSI indicates significant levels of psychological distress (Derogatis, 1993). In the overall sample, the mean T-score was 51.02 (SD = 14.25). No trajectory group had an average T-score that reached clinically significant levels, with the highest average T-score seen in the ‘Limited Contact’ group (M= 56.78, SD = 13.35). Results of fixed-effect modeling indicated an overall significant effect of trajectory group membership on psychological distress, F(3, 94) = 4.52, p = .005. Pairwise comparisons were made between the ‘Extended Contact’ group and all other groups, and the comparison between the ‘Extended Contact’ group and the ‘Limited Contact’ group was significant (t(94) = 2.99, p = .004). Young adults in the ‘Extended Contact’ group had lower levels of psychological distress than those in the ‘Limited Contact’ group. Comparisons between the ‘Extended Contact’ group and the ‘No Contact’ and ‘Stopped Contact’ groups were not significant.

Concerning psychological well-being, as measured by the Ryff Scales, there was a significant overall effect of trajectory group membership on autonomy, F(3, 95) = 2.91, p = .038. Comparisons between the ‘Extended Contact’ group and remaining groups indicated higher levels of autonomy in the ‘Extended Contact’ group compared to the ‘Limited Contact’ group (t(95) = −2.62, p = .010). In addition, there was a significant overall effect of trajectory group membership on positive relations with others, F(3, 95) = 6.21, p = .001. Participants in the ‘Extended Contact’ group had significantly higher levels of positive relations to others than those in the ‘Limited Contact’ (t(95) = −3.53, p = .001) and ‘Stopped Contact’ groups (t(95) = −2.92, p = .004). There was no significant effect of trajectory group membership on environmental mastery, F(3, 95) = 2.30, p = .082.

Adoption Outcomes

There were no significant effects of trajectory group membership on the AIS [F(3, 98) = .24, p = .871] or either the PA [F(3, 101) = .51, p = .673] or NE [F(3, 101) = 1.36, p = .258] subscales of the ADQ. Trajectory group membership significantly predicted desirability of contact with the birth mother, F(3, 98) = 9.64, p < .001. Young adults in the ‘Extended Contact’ group expressed significantly higher levels of desirability than those in the ‘Limited Contact’ (t(98) = −2.61, p = .011), ‘Stopped Contact’ (t(98) = −2.65, p = .009), and ‘No Contact’ (t(98) = −5.34, p < .001) groups. Similarly, trajectory group membership significantly predicted young adults’ satisfaction with their relationship with their birth mother, F(3, 98) = 5.23, p = .002. Participants in the ‘Extended Contact’ group expressed significantly higher levels of satisfaction than those in the ‘Limited Contact’ (t(98) = −2.80, p = .006) and ‘No Contact’ (t(98) = −3.82, p < .001) groups. Trajectory group membership did not significantly predict young adult’s feelings about being placed for adoption, F(3, 98) = 2.04, p = .113.

Discussion

This study examined whether patterns of contact with birth family members from childhood to adulthood predicted psychological adjustment and adoption-related outcomes in adopted young adults. Results were consistent with hypotheses that young adults in the ‘Extended Contact’ group would display more positive psychological adjustment than those in other trajectory groups. Group differences in adjustment were most often seen between the ‘Extended Contact’ group and the ‘Limited Contact’ group. One difference was found with the ‘Stopped Contact’ group, and no differences were found with the ‘No Contact’ group. Average T-scores for the four groups on the GSI were all in the nonclinical range, suggesting that although some adoptees reported more positive adjustment than others, no contact experiences were specifically tied with clinically significant psychological distress. Young adult adoptees appeared to navigate their experiences with contact in resilient ways that resulted in levels of adjustment comparable to other adults in the community (Derogatis, 1993).

Within the ‘Extended Contact’ group, young adults had experienced contact with birth relatives for most of their lives. Consistent with emotional distance regulation theory (Grotevant, 2009), their interviews had noted the relational “work” required of such contact: maintaining communication with multiple members of the birth family, being flexible in the face of conflicting needs, and dealing adaptively with occasional challenges (Grotevant et al., 2019). If they had not been successful in maintaining contact, they would likely have been in the ‘Stopped Contact’ or ‘Limited Contact’ groups. Thus, participants in the ‘Extended Contact’ group had successfully navigated the challenges of open adoption.

The current study found no evidence that adoptees’ longitudinal contact experiences were predictive of adoptive identity or general adoption experiences. Regarding adoptive identity, it was initially hypothesized that consistently high contact and satisfaction with contact would indicate opportunities for the construction of coherent adoptive identity narratives. However, the AIS used in the current study included many ‘social identity’ aspects of adoption (e.g., solidarity with other adoptees) and general feelings about being adopted, as opposed to primarily focusing on the creation of internally consistent adoption stories. Concerning the ADQ, the PA scale speaks to general feelings about being adopted and feelings about adoptive parents, whereas the NE scale pertains mostly to social interactions and stigma around adoption. That no differences were found may in fact be unsurprising, as such constructs may be less relevant to birth family relationships and the benefits that they bring (e.g., connections/information for sense of self) or are less salient in young adulthood than in earlier developmental stages. Lastly, the ‘Extended Group’ differentiated itself through participants reporting more positive feelings and relational experiences with their birth mothers. Such findings were consistent with the ‘Extended Group’ being characterized by ongoing relationships with the birth mother/birth relatives and consistently high degree of satisfaction with contact.

Pathways to Outcomes

Although adoption is a positive intervention in the lives of most children (Palacios et al 2019), it is fundamentally grounded in loss – loss of one’s birth family, loss of aspects of one’s identity, and in some cases, loss of one’s culture and language (Leon, 2002). Adoptive parents vary in their ability to navigate such relational complexities over the life course, and adopted children vary in how they respond to their distinctive family situations (Pinderhughes & Brodzinsky, 2019). Practitioners began advocating for contact between birth and adoptive family members decades ago, arguing that contact could mitigate some of the negative effects of loss (Pannor & Baran, 1984). The movement toward openness has been predicated on the notion that secrecy is potentially harmful to adopted persons and members of the broader adoptive kinship network, that authentic relationships can be forged between adopted persons and their birth relatives, and that these connections can provide adoptees with access to knowledge about their history and connections to other people who care about them (Grotevant, 2020).

Adoptees in the ‘Extended Contact’ group experienced contact with their birth mother/relatives over many years and found those openness arrangements to be consistently satisfying. Participating in such relationships means that adoptees’ questions can be answered so that they can develop reality-based narratives about their lives that support well-being and buffer against distress. For relationships in the broader adoptive kinship network to be sustained, it also means that birth relatives and adoptive parents must be able to engage with each other constructively, further supporting the well-being of the adopted person over time (Neil et al., 2015). In explaining lack of evidence of differences involving the ‘No Contact’ and ‘Stopped Contact’ groups, young adults in those groups may not have experienced as significant challenges around loss in sense of self, otherwise they may have sought out more contact and been categorized in a different group. Within narrative identity literature, narrative coherence appears more important for well-being when the narrative is more central to identity (Waters & Fivush, 2015). Thus, for some participants in those groups, the development of a coherent birth history narrative (as well as the contact that would aid such development) may be less essential in the construction of their sense of self. Alternatively, some participants may have been satisfied developing coherent narratives just from information provided to them by their adoptive family (e.g., narratives about being chosen by adoptive parents). Thus, they may have felt less need to seek out information from birth relatives to develop their self-concepts. Lastly, it is possible that participants in the ‘No Contact’ group had more positive adjustment prior to the start of the study and felt less need to seek out contact throughout their lives. However, existence of contact in the current study did not equate to adopted persons feeling the need to initiate contact, as many families agreed to birth family contact upon placement of the child. Regarding group differences that were identified, interviews with the adoptees revealed that some participants in the ‘Limited Contact’ groups experienced challenges in maintaining/negotiating contact (Grotevant et al., 2019). Such difficulties may have interfered with the establishment of relationships that are sources of highly sought-after information and that are as satisfying as those experienced by members of the ‘Extended Contact’ group. Furthermore, adopted persons in those groups could have gained negative information about birth family members and birth histories, which would be consistent with the decrease in satisfaction seen in those groups.

Human development unfolds over time in the context of dynamic and reciprocal relationships (Fiese et al., 2019). Thus, the development of positive adjustment also serves as a resource for the young person, further contributing to ongoing positive relationships and well-being. Although theory and temporal precedence suggest that the predominant direction of effect is from openness arrangements to adjustment, it is important to consider bidirectional effects. Young adults with significant and sustained adjustment challenges could have been less able to create and sustain the relationships characteristic of the ‘Extended Contact’ group, and thus may have experienced earlier termination of contact (‘Stopped Contact’ group) or limitations in relationships with birth family members (‘Limited Contact’ group).

Strengths, Limitations, and Future Directions

This study is unique in taking an over 20-year longitudinal approach involving multiple aspects of contact in extending research on the implications of birth family contact experiences. It is rooted in a larger study of contact that has been ongoing for the longest duration with the largest sample (190 families). Furthermore, it is based on a specific set of adoptive families (domestic, infant adoptions with low-risk histories), maximizing depth and minimizing heterogeneity (Jager et al., 2017). However, the study is limited in its generalizability to other types of adoptive families, especially those who have adopted through the child welfare system or internationally (see Grotevant, 2020, for further discussion). For example, it is possible that differences in identity on the AIS and general adoptive experiences on the ADQ may have been found had the current sample included transracially adopted adults as opposed to only adults from same-race (primarily White) placements. Transracially adopted persons, including individuals adopted internationally from Asian, Latin American, or African countries, navigate creating ethnic-racial identities within families that are ethnically and racially different from themselves. Furthermore, they more often encounter racial/ethnic discrimination than their White, adopted peers (Kim & Lee, 2020). Given the centrality of such factors to many transracially adopted individuals’ lived experiences (Kim & Lee, 2020), it would be important to determine how birth family contact relates to adoptive identity and adoption experiences among transracially adopted adults.

The study was limited in its statistical power due to trajectory group sizes and moderately high degrees of missingness in the outcome variables. Although auxiliary variables were added to address the MAR assumption, the degree of missingness still potentially led to inflated standard errors and reduced ability to detect significant differences (Schlomer et al., 2010). Nevertheless, a consistent and theoretically coherent pattern of results was identified. Lastly, the internal consistency of the Negative Experiences variable at W4 was low, limiting the study’s ability to accurately capture that construct as an outcome. Regarding future research, it would be important to include the experiences of birth family members in examining the implications of contact, as birth family members and adoptive family members have been found to sometimes differ in their perceptions of contact (e.g., reasons for contact ending; Dunbar et al., 2006).

Implications for Practice and Policy

In recent years, senior project personnel have been invited to present our work (e.g., Grotevant, 2017; Grotevant, 2019; Wrobel, 2018) to practitioners and policymakers in the UK, Spain, Germany, Canada, Australia, Norway, New Zealand, the Netherlands, and across the US. In each setting, audience members spoke about the impact that the research is already having on their practice with adoptive families and/or on policy discussions within their respective countries. Results of this study will aid professionals who interact with adoptive family members over time. For social workers preparing families for adoption, our results emphasize that openness can change dynamically over time as a function of the members involved, their continually changing contexts, and their abilities to create and maintain relationships that do not have clear societal definitions. Professionals such as adoption agency staff, health professionals, and educators will have the knowledge that presence of birth family contact itself does not appear harmful to adopted persons, and that sustained and rewarding contact experiences may be protective into young adulthood. It will be the role of practitioners to determine the conditions that contribute to creating and maintaining effective relationships within adoption kinship networks.

Although the findings of this study were developed in the context of same-race domestic infant adoptions, they have broader relevance. For example, most U.S. adoptions currently come from the child welfare system; those children have typically been maltreated in their families of origin, after which they experienced one or more placements in foster care. Nevertheless, many children who have had such experiences long for contact with family members, including siblings, parents, and extended family (Affronti et al., 2015). Birth family contact in adoptions from care is becoming increasingly more common in the US (Howard & Smith, 2003) as well as in countries such as the UK (Neil et al., 2015) and Australia (del Pozo de Bolger et al., 2018).

Child welfare practitioners are currently discussing how to assure children’s safety and well-being while permitting contact with family members. One mechanism is the Post Adoption Contact Agreement, in which adoptive and birth family members agree in writing on the principles that will guide their contact with the child (Allisan, 2019). The Australian state of New South Wales stipulates that in adoptions from foster care, arrangements permitting direct contact with birth relatives are required. The parties to the adoption must develop a post-adoption contact plan, which is agreed upon by all parties and approved by the court prior to the adoption (del Pozo de Bolger et al., 2018). Research validating the usefulness of such agreements is needed. Neil (2019) has identified three overarching issues that are particularly important for child welfare professionals to consider when working with children being adopted from care: a) making individualized decisions rather than creating “one size fits all” policies; b) considering the quality of the contact (rather than simply its presence) in deciding how it should move forward; and c) offering support when it is needed.

Results of this study also have valuable implications for international adoptees. The Hague Conference on Private International Law (1993) specifies that information about the identity of the child’s birth parents and their medical histories be preserved and available to the child or their representative. Reports of falsified or missing identity or medical information (Smolin, 2010) have motivated many internationally adopted young adults to search for birth relatives, despite having to do it across national, cultural, and language barriers (Baden et al., 2013; Grotevant, 2020).

Despite differences in particular adoption circumstances, the current research lays the foundation from which professionals working in different contexts (e.g., child welfare adoptions, international adoptions) can shape culturally appropriate practices that ensure child safety while simultaneously promoting child well-being and maintaining important family connections. Results of this study support the argument that trends towards openness can be beneficial for adopted persons, and approaches focusing on a variety of aspects of contact should be considered in developing effective practice and policy.