Is Early Experience Destiny? Review of Research on Long-Term Outcomes following International Adoption with Special Reference to the British Chinese Adoption Study
. All adopted between 1970 and 1977 and brought up by families in southern Sweden. Two-thirds were females. This group was formed of the oldest participants from previous study of 152 adolescents.
88% born in Asia (mostly India, Thailand, and Sri Lanka), and the remainder mostly in Latin America. All except one had spent some time in orphanage care prior to adoption.
Mean age: 28 years; range: 25–34 years.
Measures included Adult Attachment Interview (AAI, semistructured), mainly about relationships with adoptive parents during childhood. Transcripts coded to adult attachment classifications: autonomous, dismissive, or preoccupied.
Compared to AAI results of a “norm group” of nonclinical mothers from a meta-analysis.
Adopted group’s AAI scores were not found to differ significantly from nonclinical comparison group. Revisiting country of origin and ethnic self-identity were not related to attachment classification, but later age at adoption predicted greater likelihood of dismissive or preoccupied attachment classification.
= 6,065. Born between 1963 and 1973. Two-thirds females. Age at immigration available for 90% of the sample, of which 54% <2 yrs; 26% 2-3 yrs; and 20% >4 yrs. Identified via National Swedish Registers. Main areas of origin were Far East Asia (54%) and South Asia (18%).
No specific information on this cohort, general discussion of possible preadoption adversity.
Range: 29–39 years.
Suicide attempt and suicide death data taken from national hospital discharge and cause of death registers.
IC adopted group had higher suicide attempt (RR 4.5) and suicide death (RR 3.6) rates than gen. pop. and sibling samples. IC adopted women’s risk compared to other female groups was elevated to a greater extent than men’s. Domestically adopted adults also had higher risks than gen. pop. but less than the IC adopted.
Tieman et al., 2005 (a), 2006 (b), and 2008 (c) [25–27]
(a) = 1,484. (b), (c) = 1,417. Adopted into Holland 1972–1975 at mean age of 28 months. 55% females. Accessed through central records of adoption (sample (a) = 72.5% of original sample at earlier sweep). Attrition rate higher for men with higher scores (therefore more problems) on Child Behaviour Checklist.
Approx. 47% experienced neglect and 12% preadoption abuse. Rates based on data reported by adopters, using predefined scales of level of adversity; only information that parents reported as being “certain” of was included.
Range: 22–32 years at most recent time of data collection.
(a) Composite International Diagnostic Interview (standardised psychiatric interview that generates DSM-IV diagnoses). Some items from National Institute of Mental Health Diagnostic I/v Schedule. (b) Social functioning assessed via standardised semistructured 115-item interview. (c) Measures included: questions about searching for birth parents; DSM-IV diagnoses (see above); Child Behaviour Checklist; family functioning.
(a) 695 age-matched general population sample. (b) Subsample of 713 adults from the above gen. pop. sample, on whom additional data was available. (c) Within-group only.
(a) Adopted group at higher risk than control group to meet criteria for anxiety disorder (1.52 times), substance abuse or dependence (2.05 times), and, for men only, mood disorder (3.76 times). No significant difference for disruptive disorder, which differs from findings from earlier follow-up with same cohort. (b) Adopted group less likely than nonadopted to be living with a partner or have had a relationship lasting more than a year. But adopted people without partners functioned better than their nonadopted single counterparts. Authors conclude that groups were similar in social contacts. (c) 32% had searched for birth parents. Greater likelihood to search was associated with country of origin, older age at placement, greater interest in searching during adolescence, more problematic behaviour, and psychiatric diagnoses. Greater interest in (but not levels of) searching found for females.
van den Berg et al., 2008 [28] (same sample as Tieman et al. studies)
= 1,475
As above.
As above.
Adopted adults’ self-reported ratings of internalising and externalising problems on standardised scales. Parented-reported Young Adult Behaviour Checklist.
Within-group only: compared biologically related siblings, nonrelated siblings, and singles.
Only study to explore genetic and environmental influences on adult outcomes. Both data sources (self- and parent-report) indicated that genetic influences were greater for internalising problems and environmental influences were greater for externalising problems. This reversed findings from the same sample during adolescence.
van der Vegt et al., 2009 (a), 2009 (b), and 2010 (c) [29–31] (same sample as Tieman et al. studies)
(a) = 1,364. (b) = 623. (c) = 429. (b) and (c) both subsamples from (a). Sample (a) = 64% of baseline sample. As above, selective attrition at follow-up was noted; nonparticipants had higher levels of mental health problems in childhood and adolescence.
As above.
(a) Mean age: 26.3 years; range: 22–32 years; SD: 1.4. (b) and (c) both subsamples from (a).
(a) Measures included interviews generating information on DSM-IV codes of mental disorders. Recorded these diagnoses individually plus created an “any disorder” variable. Plus parent reports on early abuse, neglect, and multiple preadoption placements. (b) Parent reports on preadoption adversity (as above). Cortisol levels collected 4 times per day via saliva samples. (c) As above, plus Child Behaviour Checklist.
Within-group only.
(a) Multiple early adversities associated with increased risk of adulthood anxiety (OR = 2.22; 95% CI 1.11–4.45); mood disorders (OR = 2.20; 95% CI 1.00–4.86); or substance abuse or dependence (OR = 3.81; 95% CI 1.62–8.98). After controlling for childhood onset of mental health problems, differences remained. Level of de novo onset suggests that consequences of early adversity can appear many years later. (b) Early neglect and abuse both predicted altered cortisol levels compared to nonneglected/nonabused groups. (c) Severe early maltreatment found to modify the relationship between anxiety disorders and cortisol secretion, but not mood disorders and cortisol secretion.
. Korean-born adults adopted into white US families. 18% men, 82% women.
Majority of time prior to adoption spent in orphanages (35%), foster families (39%), birth family (13%), unknown (11%), or other (2%).
Mean age: 31 years. All over 18 years.
Online questionnaire, incl. Family of Origin Scale, Multigroup Ethnic Identity Measure, Rosenberg Self-Esteem, and Satisfaction with Life. Also questions about (1) changes in self-identification and (2) support services.
156 white American-born adopted adults, from same total sample as Korean-born group. Mean age: 44 years (13 years older than Korean-born group).
Adoption is an increasingly significant aspect of identity across lifespan for both groups; “race”/ethnicity is increasingly important for Korean group, peaking in adulthood. Feeling more comfortable with adoptive identity was associated with higher life satisfaction for White group. In contrast, feeling more comfortable with adoptive identity was associated with female gender, higher life satisfaction and higher self-esteem for Korean-born group.
. Children adopted in infancy from Greece into Netherlands before 1970. 30 men and 23 women (of 60 randomly selected from 121 traced; total potential sample was 400).
Authors characterise this as a group who did not suffer severe deprivation: relatively limited number of caregivers in relatively limited number of caregivers in orphanage and attempts at consistency in care.
Mean age: 29 years; range: 25–36 years.
General questionnaire about adult life circumstances and adoption, plus mental health (Symptom Check-List 90), well-being (Satisfaction with Life scale), and self-esteem (Rosenberg).
Normative data on Dutch-born young adults, taken from different sources for each measure.
Only significant difference from comparisons on mental health, well-being, and self-esteem was higher rate of depression for adopted men in comparison to nonadopted men. Within-group: those who searched for birth parents reported more difficulties in mental health, well-being, and self-esteem. Further analysis identified that negative appraisal of adoption was a stronger predictor than search status for mental health outcomes and well-being.
. Hong Kong ex-orphanage girls adopted into the UK in the 1960s. Duration of orphanage care: mean age: 20 months (range: 5–83 months; SD: 13.5). Age at placement for adoption: mean age: 23 months (range: 8–83 months; SD: 14) Participants were from original groups of 100 women traced individually nearly 50 years later and recruited to the study.
Relatively well-run orphanages with adequate physical care and nutrition but overcrowding and lack of opportunity for selective attachment to adult carers.
Mean age: 48 years; range: 42–53 years; SD: 2.4.
Orphanage/adoption records. Questionnaires covering mental health (General Health Questionnaire and Malaise Inventory), self-esteem (Rosenberg), life satisfaction, personality profiles, community connectedness, partnerships, and adoptive family relationships. Interviews on life history and current circumstances (usually 2-3 hours).
(a) Within-group only. (b) 5,115 age- and sex-matched general population sample (National Child Development Study) and a domestically adopted subsample () drawn from the same dataset.
(a) Ethnic and social identification not found to predict psychological adjustment. (b) Outcomes were commensurate with the comparison groups in terms of mental and physical health measures. Serious psychiatric and social difficulties were largely absent. Timing and extent of exposure to orphanage care did not influence outcome, but self-reported poorer quality adoptive experience and a negative view of their adoption were significantly associated with poorer mental health outcomes (difference in means = 0.76, 95% CI 1.33–0.19, ; difference in means = 1.2, 95% CI 0.68–1.73, , resp.).
Age of participants at point of data reported in this paper. IC = intercountry, OR = odds ratio, RR = risk ratio, gen pop = general population, and I/v = interview.
