Key Points
Question
What is the risk of experiencing adverse social and health outcomes in adulthood among children and adolescents placed in out-of-home care?
Findings
In this cohort study, risk of adverse social and health outcomes in adulthood were elevated 1.4- to 5-fold among children placed in out-of-home care compared with their siblings who had never been placed in out-of-home care. By comparing differentially exposed siblings, the study was able to account for shared genetic and environmental preplacement factors.
Meaning
Although it may be necessary to remove children from parents who expose them to severe maltreatment, neglect, or abuse, out-of-home care placement is associated with important outcomes that need careful review.
This cohort study estimates associations between being placed in out-of-home care in childhood and adolescence and subsequent risks of experiencing a wide range of social and health outcomes in adulthood following comprehensive adjustments for preplacement factors.
Abstract
Importance
Children who are placed in out-of-home care may have poorer outcomes in adulthood, on average, compared with their peers, but the direction and magnitude of these associations need clarification.
Objective
To estimate associations between being placed in out-of-home care in childhood and adolescence and subsequent risks of experiencing a wide range of social and health outcomes in adulthood following comprehensive adjustments for preplacement factors.
Design, Setting, and Participants
This cohort and cosibling study of all children born in Finland between 1986 and 2000 (N = 855 622) monitored each person from their 15th birthday either until the end of the study period (December 2018) or until they migrated, died, or experienced the outcome of interest. Cox and Poisson regression models were used to estimate associations with adjustment for measured confounders (from linked population registers) and unmeasured familial confounders (using sibling comparisons). Data were analyzed from October 2020 to August 2021.
Exposures
Placement in out-of-home care up to age 15 years.
Main Outcomes and Measures
Through national population, patient, prescription drug, cause of death, and crime registers, 16 specific outcomes were identified across the following categories: psychiatric disorders; low socioeconomic status; injuries and experiencing violence; and antisocial behaviors, suicidality, and premature mortality.
Results
A total of 30 127 individuals (3.4%) were identified who had been placed in out-of-home care for a median (interquartile range) period of 1.3 (0.2-5.1) years and 2 (1-3) placement episodes before age 15 years. Compared with their siblings, individuals who had been placed in out-of-home care were 1.4 to 5 times more likely to experience adverse outcomes in adulthood (adjusted hazard ratio [aHR] for those with a fall-related injury, 1.40; 95% CI, 1.25-1.57 and aHR for those with an unintentional poisoning injury, 4.79; 95% CI, 3.56-6.43, respectively). The highest relative risks were observed for those with violent crime arrests (aHR, 4.16; 95% CI, 3.74-4.62; cumulative incidence, 24.6% in individuals who had been placed in out-of-home care vs 5.1% in those who had not), substance misuse (aHR, 4.75; 95% CI, 4.25-5.30; cumulative incidence, 23.2% vs 4.6%), and unintentional poisoning injury (aHR 4.79; 95% CI, 3.56-6.43; cumulative incidence, 3.1% vs 0.6%). Additional adjustments for perinatal factors, childhood behavioral problems, and traumatic injuries, including experiencing violence, did not materially change the findings.
Conclusions and Relevance
Out-of-home care placement was associated with a wide range of adverse outcomes in adulthood, which persisted following adjustments for measured preplacement factors and unmeasured familial factors.
Introduction
Globally, it has been estimated that more than 3.2 million children reside in institutional settings on an annual basis.1 At least 660 000 children in the US2 are annually part of the foster care system with direct federal costs for child protective services reaching nearly $30 billion per year.3 Similarly, approximately 1 million children in European Union countries4 receive out-of-home care services. Despite efforts to compensate for early adverse home environments, systematic reviews5,6 of nonexperimental studies have consistently found that children placed in out-of-home care are far more likely than their peers to be unemployed, develop psychiatric disorders, engage in antisocial behaviors, and to die prematurely in adulthood. However, whether these associations reflect causal effects of the out-of-home placement or whether they can be explained by background risk factors that predate such placement has not been sufficiently accounted for in previous work.
Out-of-home placement aggregates in families,7 and many parental risk markers8,9 are at least moderately heritable, including low socioeconomic status,10,11,12 antisocial behaviors,13,14 and psychiatric disorders.15 Sibling comparison designs wherein risks of long-term outcomes are compared between biological full siblings who were differentially exposed to out-of-home care allow researchers to account for time-stable unmeasured familial confounders, including shared early-life environments and around half of cosegregating genes.16 The few existing sibling comparison studies on the effects of out-of-home care have reached mixed conclusions17,18 and require replication in considerably larger samples. Mixed findings have similarly been reported in studies using other designs.19,20,21,22 Large-scale studies examining the effect of placement across the entire period of childhood and adolescence as well as those examining differences between foster care and institutional care placement within sibships are lacking.
To examine associations between out-of-home care placement and health and social outcomes in adulthood while accounting for preexisting risk markers, we used a sibling control design in the entire population of Finland born between 1986 and 2000, which included more than 885 000 individuals and 622 000 siblings. We monitored these individuals for a mean (range) of 9 (9.0-10.3) years to investigate 16 objectively measured social and health outcomes while accounting for an extensive set of measured and unmeasured familial and individual-level confounders. We were also able to study the relative contributions of age at placement, care type, number of placement episodes, and duration of care with rigorous controls for preplacement confounders.
Methods
Our target study population included all individuals born in Finland between 1986 and 2000 who were alive and resided in Finland at age 15 years, which represented the baseline of the study. A total of 903 951 individuals were included in this population. All Finnish residents are assigned a personal identification number, which is used in different nationwide registers and provides accurate linkage.23 The Board of Statistical Ethics of Statistics Finland granted permission to use pseudonymized data. Finnish law does not require informed consent for register-based studies. The study was approved by the Ethics Board of Statistics Finland.
We identified children who were placed in out-of-home care from the Child Welfare Register maintained by the Finnish Institute for Health and Welfare, which includes data on all placement episodes since January 1991, including earlier episodes for those who were placed at any point after January 1990. The population registers maintained by Statistics Finland provided data on sociodemographic factors and migration dates between January 1987 and December 2018. Mortality dates and underlying and contributory causes based on International Classification of Diseases, Ninth Revision (ICD-9) and ICD-10 were derived from the Causes of Death Register, which is maintained by Statistics Finland and has excellent coverage (greater than 99%).24 Diagnoses of unintentional injuries, self-harm, and psychiatric and substance use disorders were retrieved from the Care Register for Health Care maintained by the Finnish Institute for Health and Welfare, which includes all inpatient care episodes (ICD-8, ICD-9, and ICD-10), and specialist outpatient visits (ICD-10). Similarly, we identified all reimbursed psychotropic drug purchases, classified according to the Anatomical Therapeutic Chemical system from the Prescription Drug Register (January 2001 to December 2018), maintained by the Social Insurance Institution of Finland. All individuals with criminal arrests from January 2001 to December 2017 were identified using police record registers maintained by Statistics Finland. Owing to the sensitive nature of the crime data, we were given access to annual arrest dates, which we subsequently set to December 31 for each year.
From the target sample of 903 951 individuals (a flowchart of all subsamples can be found in eFigure 1 in the Supplement), we excluded 9585 individuals who were missing data on their biological parents and 8744 who were placed in out-home care for the first time after their 15th birthday. Our analytical sample therefore retained 98.0% of the target sample of 885 622. We monitored participants from their 15th birthday until the first of the following events: the individual experienced the outcome of interest, migrated, died, or was alive at the end of the follow-up period (December 2017 for the outcomes derived from patient and crime registers and December 2018 for the outcomes derived from the population and prescription drug registers).
We categorized individuals who had been placed in out-of-home care by child welfare services at least once before their 15th birthday as having been exposed to out-of-home care. There are 4 main types of out-of-home care settings in Finland: family foster care (kinship or nonrelative care), professional group homes, institutional care, and other or unclassified care. We assigned each child to the type of care setting to which they had been exposed for the longest period.
We examined a total of 16 outcomes that allowed us to comprehensively assess specific dimensions of health, social, and behavioral functioning across 4 categories. Psychiatric disorders included any inpatient episode or outpatient visit associated with any diagnosis of a severe mental illness (defined as schizophrenia-spectrum disorder and bipolar disorder);25 depression; anxiety; or personality disorder (ICD codes in eTable 1 in the Supplement; diagnostic validity in eAppendix 1 in the Supplement). Low socioeconomic status encompassed low education, which indicated that the individual had not achieved secondary school qualifications by the end of the study (not available for 53 730 individuals born in 2000); welfare benefits, which denoted whether the individual had received means-tested social assistance benefits; long-term unemployment, which indicated whether the individual had received unemployment benefits for a minimum of 6 months in a given year; and disability pension, which denoted whether the individual had received benefits because of work incapacity. Injuries and experiencing violence included any inpatient episode, outpatient visit, or death associated with any diagnosis for an unintentional motor vehicle–related injury, fall-related injury, unintentional poisoning injury, or assault. Antisocial behaviors, suicidality, and premature mortality included violent crime arrest (penal codes in eTable 2 in the Supplement),26 substance misuse (eg, a substance use disorder diagnosis, an arrest for a drug-related crime, or a prescription for medications used to treat substance use disorders; Anatomical Therapeutic Chemical codes in eTable 3 in the Supplement), suicidal behavior (a diagnostic code for suicide attempt or a completed suicide), and premature mortality, which indicated whether the individual died by the end of the study period (December 2018).
Data Analysis
We quantified the associations between being placed in out-of-home care and subsequent risk of long-term adverse outcomes by fitting Cox regression models with age as the primary time scale and adjusting for sex, birth year, and birth order. We treated the low educational attainment outcome as a binary measure by fitting a corresponding robust Poisson regression model. The associations were therefore either expressed as hazard ratios or rate ratios. We subsequently accounted for familial confounding by fitting stratified Cox regression models (or the analogous fixed-effects robust Poisson regression model for low educational attainment), where we allowed the baseline hazards to vary across clusters of biological full siblings using a fixed-effects estimator.16 This approach allowed us to compare outcome rates between 16 774 differentially exposed siblings in the same families (ie, where 1 sibling was placed in out-of-home care and cosiblings were not), a design that indirectly accounted for all time-invariant unmeasured familial confounders (eg, approximately half of the cosegregating genes and shared childhood environments). To additionally account for familial confounders that potentially varied between siblings, we further adjusted these models for the following covariates measured at offspring birth or age 1 year for those born in 1986 (definitions in eTable 4 in the Supplement): urbanicity, single-parent household, parental age, family income, parental long-term unemployment, welfare benefits, and disability pension. Complementary sensitivity analyses are described in eAppendix 2 in the Supplement.
In a subset of families where all children (n = 11 092) had been placed in out-of-home care, we tested for the following placement characteristics: care setting (eg, foster care vs institutional care), number of placement episodes, duration of care summed across all episodes (binary measure, age 5 to 15 years vs birth to age 4 years), and age at first placement (binary measure, age 12 to 15 years vs birth to age 11 years). We refitted the sibling models presented above using these factors as covariates in the model. We further accounted for measured preplacement factors by adjusting for specific traumatic injuries, neurodevelopmental and neurological disorders, and behavioral problems (eAppendix 2 in the Supplement). The estimates were determined to be statistically significant if their false discovery rate–corrected P values were less than .05.27
Results
Our nationwide sample included a total of 885 662 children, of whom 433 088 (48.9%) were female and 452 574 (51.1%) were male. The mean (SD) age at the end of follow-up was 25.1 (4.2) years. A total of 30 127 (3.4%) children had been placed in out-of-home care at least once before reaching age 15 years. These children had a median (interquartile range) of 2 (1-3) placements during the exposure period, which lasted for a median (interquartile range) of 1.3 (0.2-5.1) years accumulated across all placement episodes. Compared with their peers who had never been placed in out-of-home care, children who had been placed in out-of-home care experienced more psychosocial adversities and socioeconomic disadvantages (Table 1). We monitored individuals for a mean (range) of 9.4 (9.0-10.3) years across outcomes (Table 2). Children placed in out-of-home care had considerably higher unadjusted cumulative incidence rates of adverse outcomes compared with their peers.
Table 1. Sociodemographic Characteristics of the Study Population (N = 885 622).
| Characteristic | No. (%) | ||
|---|---|---|---|
| Never placed in out-of-home care (n = 855 495) | Placed in out-of-home care at least once by age 15 y (n = 30 127) | ||
| Offspring demographic characteristics | |||
| Sex | |||
| Female | 417 912 (48.9) | 14 883 (49.4) | |
| Male | 437 583 (51.1) | 15 244 (50.6) | |
| Birth year | |||
| 1986-1990 | 291 941 (34.1) | 7706 (25.6) | |
| 1991-1995 | 299 217 (35.0) | 10 569 (35.1) | |
| 1996-2000 | 264 337 (30.9) | 11 852 (39.3) | |
| Birth order | |||
| First | 357 020 (41.7) | 12 320 (40.9) | |
| Second | 291 823 (34.1) | 8893 (29.5) | |
| Third | 133 158 (15.6) | 4949 (16.4) | |
| Fourth or higher | 73 494 (8.6) | 3965 (13.2) | |
| Urbanicity | |||
| Urban | 495 676 (57.9) | 21 485 (71.3) | |
| Semiurban | 134 043 (15.7) | 3519 (11.7) | |
| Rural | 225 776 (26.4) | 5123 (17.0) | |
| Parental sociodemographic factors measured at offspring birth | |||
| Immigrant background | 38 634 (4.5) | 2971 (9.9) | |
| Teenage mother at birth | 13 953 (1.6) | 2164 (7.2) | |
| Teenage father at birth | 3309 (0.4) | 645 (2.1) | |
| Single-parent household | 48 560 (5.7) | 6836 (22.7) | |
| Highest parental educational attainment | |||
| Primary | 59 368 (6.9) | 8515 (28.3) | |
| Secondary | 604 362 (70.6) | 19 884 (66.0) | |
| Tertiary | 191 765 (22.4) | 1728 (5.7) | |
| Family income in the bottom quintile | 159 904 (18.7) | 15 653 (52.0) | |
| Social assistance benefits | 107 903 (12.6) | 16 034 (53.2) | |
| Disability pension | 5560 (0.6) | 1461 (4.8) | |
| Lifetime parental history of psychiatric disorders | |||
| Severe mental illness | 38 899 (4.5) | 7143 (23.7) | |
| Depression | 124 219 (14.5) | 13 947 (46.3) | |
| Anxiety | 65 140 (7.6) | 7552 (25.1) | |
| Personality disorder | 31 059 (3.6) | 8221 (27.3) | |
| Lifetime parental history of injury and experiencing violence | |||
| Motor vehicle–related injury | 102 853 (12.0) | 6500 (21.6) | |
| Fall-related injury | 297 527 (34.8) | 14 976 (49.7) | |
| Unintentional poisoning injury | 13 327 (1.6) | 2174 (7.2) | |
| Experiencing violence | 24 742 (2.9) | 5445 (18.1) | |
| Lifetime parental history of antisocial behaviors and suicidality | |||
| Violent crime arrest | 98 537 (11.5) | 15 636 (51.9) | |
| Substance misuse | 81 466 (9.5) | 15 006 (49.8) | |
| Suicidal behavior | 39 869 (4.7) | 8102 (26.9) | |
| Offspring neurodevelopmental disorders and behavioral problems, birth to age 15 y | |||
| Attention-deficit/hyperactivity disorder | 12 297 (1.4) | 3088 (10.2) | |
| Intellectual disability | 4771 (0.6) | 654 (2.2) | |
| Autism spectrum disorder | 5570 (0.7) | 944 (3.1) | |
| Communication disorders | 20 100 (2.3) | 1741 (5.8) | |
| Learning disorders | 11 681 (1.4) | 1492 (5.0) | |
| Motor disorders | 8502 (1.0) | 804 (2.7) | |
| Other neurodevelopmental disorders | 2267 (0.3) | 354 (1.2) | |
| Any neurological condition | 31 738 (3.7) | 1887 (6.3) | |
| Conduct disorder or oppositional defiant disorder | 7253 (0.8) | 6138 (20.4) | |
| Offspring history of injury and experiencing violence, birth to age 15 y | |||
| Motor vehicle–related injury | 14 883 (1.7) | 902 (3.0) | |
| Fall-related injury | 64 723 (7.6) | 3302 (11.0) | |
| Unintentional poisoning injury | 5572 (0.7) | 1153 (3.8) | |
| Experiencing violence | 668 (0.1) | 278 (0.9) | |
Table 2. Person-Time at Risk, Number of Individuals, Cumulative Incidence Rate, and Incidence Rates per 1000 Person-Years Across Children Placed and Never Placed in Out-of-Home Care.
| Variable | Person-time at risk, y | Individuals, No. | Cumulative incidence rate, % (95% CI) | Incidence rate per 1000 person-years (95% CI) | |
|---|---|---|---|---|---|
| Total, No. | Mean (SD) per person | ||||
| Psychiatric disorder | |||||
| Severe mental illness | |||||
| All individuals | 8 334 509 | 9.4 (4.3) | 17 619 | 1.99 (1.96-2.02) | 2.11 (2.08-2.15) |
| Never placed | 8 087 744 | 9.5 (4.3) | 15 220 | 1.78 (1.75-1.81) | 1.88 (1.85-1.91) |
| Placed | 246 765 | 8.2 (4.2) | 2399 | 7.96 (7.66-8.27) | 9.72 (9.34-10.12) |
| Depression | |||||
| All individuals | 8 051 027 | 9.1 (4.4) | 68 592 | 7.75 (7.69-7.80) | 8.52 (8.46-8.58) |
| Never placed | 7 833 007 | 9.2 (4.4) | 61 571 | 7.20 (7.14-7.25) | 7.86 (7.80-7.92) |
| Placed | 218 020 | 7.2 (4.5) | 7021 | 23.30 (22.83-23.79) | 32.20 (31.45-32.97) |
| Anxiety | |||||
| All individuals | 8 193 183 | 9.3 (4.3) | 49 923 | 5.64 (5.59-5.69) | 6.09 (6.04-6.15) |
| Never placed | 7 956 497 | 9.3 (4.3) | 44 957 | 5.26 (5.21-5.30) | 5.65 (5.60-5.70) |
| Placed | 236 686 | 7.9 (4.4) | 4966 | 16.48 (16.07-16.91) | 20.98 (20.40-21.57) |
| Personality disorder | |||||
| All individuals | 8 369 072 | 9.4 (4.3) | 12 701 | 1.43 (1.41-1.46) | 1.52 (1.49-1.54) |
| Never placed | 8 118 380 | 9.5 (4.3) | 10 922 | 1.28 (1.25-1.30) | 1.35 (1.32-1.37) |
| Placed | 250 691 | 8.3 (4.1) | 1779 | 5.91 (5.64-6.18) | 7.10 (6.77-7.43) |
| Injuries and experiencing violence | |||||
| Motor vehicle–related injury | |||||
| All individuals | 8 177 305 | 9.2 (4.4) | 43 873 | 4.95 (4.91-5.00) | 5.37 (5.32-5.42) |
| Never placed | 7 929 272 | 9.3 (4.4) | 41 591 | 4.86 (4.82-4.91) | 5.25 (5.19-5.30) |
| Placed | 248 032 | 8.2 (4.3) | 2282 | 7.57 (7.28-7.88) | 9.20 (8.83-9.59) |
| Fall-related injury | |||||
| All individuals | 8 004 936 | 9.0 (4.4) | 80 455 | 9.08 (9.02-9.14) | 10.05 (9.98-10.12) |
| Never placed | 7 763 262 | 9.1 (4.4) | 76 663 | 8.96 (8.90-9.02) | 9.88 (9.81-9.95) |
| Placed | 241 674 | 8.0 (4.2) | 3792 | 12.59 (12.21-12.97) | 15.69 (15.20-16.20) |
| Unintentional poisoning injury | |||||
| All individuals | 8 396 751 | 9.5 (4.3) | 5619 | 0.63 (0.62-0.65) | 0.67 (0.65-0.69) |
| Never placed | 8 141 948 | 9.5 (4.3) | 4693 | 0.55 (0.53-0.56) | 0.58 (0.56-0.59) |
| Placed | 254 803 | 8.5 (4.2) | 926 | 3.07 (2.88-3.27) | 3.63 (3.40-3.88) |
| Experiencing violence | |||||
| All individuals | 8 366 590 | 9.4 (4.3) | 12 094 | 1.37 (1.34-1.39) | 1.45 (1.42-1.47) |
| Never placed | 8 113 670 | 9.5 (4.3) | 10 547 | 1.23 (1.21-1.26) | 1.30 (1.28-1.32) |
| Placed | 252 920 | 8.4 (4.1) | 1547 | 5.13 (4.89-5.39) | 6.12 (5.82-6.43) |
| Low socioeconomic status | |||||
| Low education | |||||
| All individuals | NA | NA | 96 401 | 11.59 (11.52-11.66) | NA |
| Never placed | NA | NA | 83 804 | 10.42 (10.36-10.49) | NA |
| Placed | NA | NA | 12 597 | 45.30 (44.72-45.89) | NA |
| Welfare benefits | |||||
| All individuals | 7 962 682 | 9.0 (4.4) | 239 763 | 27.07 (26.98-27.17) | 30.11 (29.99-30.23) |
| Never placed | 7 809 323 | 9.1 (4.4) | 216 421 | 25.30 (25.21-25.39) | 27.71 (27.60-27.83) |
| Placed | 153 358 | 5.1 (3.0) | 23 342 | 77.48 (77.00-77.95) | 152.21 (150.26-154.17) |
| Long-term unemployment | |||||
| All individuals | 8 382 883 | 9.5 (4.2) | 201 325 | 22.73 (22.65-22.82) | 24.02 (23.91-24.12) |
| Never placed | 8 151 165 | 9.5 (4.2) | 188 211 | 22.00 (21.91-22.09) | 23.09 (23.99-23.19) |
| Placed | 231 719 | 7.7 (3.6) | 13 114 | 43.53 (42.97-44.09) | 56.59 (55.63-57.57) |
| Disability pension | |||||
| All individuals | 9 165 601 | 10.3 (4.3) | 20 634 | 2.33 (2.30-2.36) | 2.25 (2.22-2.28) |
| Never placed | 8 888 410 | 10.4 (4.3) | 18 512 | 2.16 (2.13-2.19) | 2.08 (2.05-2.11) |
| Placed | 277 191 | 9.2 (4.2) | 2122 | 7.04 (6.76-7.34) | 7.66 (7.33-7.99) |
| Antisocial behaviors, suicidality, and premature mortality | |||||
| Violent crime arrest | |||||
| All individuals | 8 099 044 | 9.1 (4.3) | 50 676 | 5.72 (5.67-5.77) | 6.26 (6.20-6.31) |
| Never placed | 7 885 955 | 9.2 (4.3) | 43 252 | 5.06 (5.01-5.10) | 5.48 (5.43-5.54) |
| Placed | 213 089 | 7.1 (4.4) | 7424 | 24.64 (24.16-25.13) | 34.84 (34.05-35.64) |
| Substance misuse | |||||
| All individuals | 9 015 945 | 10.2 (4.4) | 46 040 | 5.20 (5.15-5.25) | 5.11 (5.06-5.15) |
| Never placed | 8 768 714 | 10.2 (4.3) | 39 063 | 4.57 (4.52-4.61) | 4.45 (4.41-4.50) |
| Placed | 247 230 | 8.2 (4.5) | 6977 | 23.16 (22.68-23.64) | 28.22 (27.56-28.89) |
| Suicidal behavior | |||||
| All individuals | 8 363 776 | 9.4 (4.3) | 13 457 | 1.52 (1.49-1.55) | 1.61 (1.58-1.64) |
| Never placed | 8 114 516 | 9.5 (4.3) | 11 311 | 1.32 (1.30-1.35) | 1.39 (1.37-1.42) |
| Placed | 249 260 | 8.3 (4.2) | 2146 | 7.12 (6.84-7.42) | 8.61 (8.25-8.98) |
| Premature mortality | |||||
| All individuals | 8 425 900 | 9.5 (4.3) | 3901 | 0.44 (0.43-0.45) | 0.46 (0.45-0.48) |
| Never placed | 8 166 370 | 9.5 (4.3) | 3494 | 0.41 (0.40-0.42) | 0.43 (0.41-0.44) |
| Placed | 259 530 | 8.6 (4.1) | 407 | 1.35 (1.22-1.49) | 1.57 (1.42-1.73) |
Abbreviation: NA, not applicable.
We initially found that the children placed in out-of-home care were between approximately 1.5 times (adjusted hazard ratio [aHR] for those with a fall-related injury, 1.53; 95% CI, 1.48-1.58) and 7 times (aHR for those receiving welfare benefits, 6.80; 95% CI, 6.71-6.89) more likely than their peers to meet criteria for adverse outcomes following adjustments for sex, birth year, and birth order (Figure). We subsequently adjusted for familial confounders by comparing full siblings who were differentially exposed to out-of-home care, which allowed us to account for all time-invariant unmeasured familial confounders. We additionally adjusted these models for several measured sociodemographic confounders, such as parental age or family income at birth, that varied between the siblings. In these analyses, we found that those who were placed in out-of-home care remained between 1.4 times (aHR for those with a fall-related injury, 1.40; 95% CI, 1.25-1.57) and 5 times (aHR for those with an unintentional poisoning injury, 4.79; 95% CI, 3.56-6.43) more likely than their siblings to meet criteria for the outcomes (Figure). The highest relative risks were observed for those with violent crime arrests (aHR, 4.16; 95% CI, 3.74-4.62; cumulative incidence, 24.6% in individuals who had been placed in out-of-home care vs 5.1% in those who had not), substance misuse (aHR, 4.75; 95% CI, 4.25-5.30; cumulative incidence, 23.2% vs 4.6%), and unintentional poisoning injury (aHR 4.79; 95% CI, 3.56-6.43; cumulative incidence, 3.1% vs 0.6%). In complementary sensitivity analyses, we found no evidence of systematic differences by sex (eFigure 2 in the Supplement). We obtained consistent results when we restricted the study group to cohorts born after 1990 for whom data on out-of-home placement covered all childhood years (eFigure 3 in the Supplement), used alternative outcome definitions (eFigure 4 in the Supplement), and measured out-of-home care placement up to age 21 years instead of age 15 years (eFigure 5 in the Supplement). Further adjustments for measured confounders and mediators, such as perinatal risks, neurodevelopmental disorders, problem behaviors, and injuries at birth to age 15 years, did not materially alter the presented findings (eFigure 6 in the Supplement). Although psychosocial and behavioral profiles varied between children who had been placed in out-of-home care and who were included in the sibling comparison models and those who were not (eTable 5 in the Supplement), we found that the crude associations were similar in the sibling and single-child family subsets (eFigure 7 in the Supplement), and we obtained commensurate results when we compared all children who had been placed in out-of-home care with their cousins who had not to account for unmeasured familial confounding in extended families (eFigure 8 in the Supplement).
Figure. Associations Between Any Out-of-Home Placement Episode (Birth to Age 15 Years) and Poor Functioning in Adulthood.
All models were adjusted for sex, birth year, and birth order. The crude models were fitted to the entire sample (N = 885 622). The adjusted models refer to within-family estimates comparing differentially exposed siblings (eg, families in which at least 1 sibling was placed and any of their cosiblings were not; n = 16 774). The latter models were further adjusted for parental age at birth, urbanicity, single-parent household, family income, parental welfare benefits, and disability pension.
A total of 11 092 children who were placed in out-of-home care had siblings who had all been placed in out-of-home care. Within this sample, we found that children who were primarily placed in institutional care settings had, compared with their siblings who were placed in foster care settings, a 1.3- to 2.3-fold increase in risk of being diagnosed with a severe mental illness or depression, having low educational attainment, receiving welfare benefits, and being arrested for violent crime (aHRs or adjusted rate ratio range, 1.29-2.33) (Table 3). We also found that each additional placement episode was associated with a 7% to 18% increased rate of being diagnosed with a severe mental illness or depression, having a fall-related injury, receiving welfare benefits, and engaging in antisocial or suicidal behaviors (aHRs or adjusted rate ratio, 1.07-1.18) (Table 3), whereas the duration of placement was not associated with any of the outcomes. Age 12 to 15 years at first placement was associated with a 2-fold increased risk of a subsequent motor vehicle–related injury (aHR, 1.98; 95% CI, 1.21-3.23) but none of the other outcomes. We found similar results when we restricted these analyses to those born after 1990 to account for left-truncation bias (eTable 6 in the Supplement).
Table 3. Out-of-Home Care Placement Characteristics on Poor Adulthood Functioning Among Siblings in Families Where All Children Had Been Exposed to at Least 1 Placement Episode (n = 11 092).
| Characteristic | aHR/aRR (95% CI)a | |||
|---|---|---|---|---|
| Institutional care vs foster care | No. of episodes | Long-term placement | Age at first placement (12-15 y vs 0-11 y) | |
| Psychiatric disorders | ||||
| Severe mental illness | 2.33 (1.38-3.94)b | 1.09 (1.02-1.16)b | 1.58 (1.00-2.48) | 1.51 (0.98-2.34) |
| Depression | 1.54 (1.14-2.09)b | 1.07 (1.03-1.12)b | 0.84 (0.62-1.13) | 0.92 (0.70-1.21) |
| Anxiety | 1.50 (1.04-2.16) | 1.02 (0.97-1.08) | 0.94 (0.66-1.34) | 1.12 (0.80-1.59) |
| Personality disorder | 1.96 (0.96-3.99) | 1.10 (1.01-1.20) | 1.10 (0.60-2.02) | 1.16 (0.65-2.06) |
| Injuries and violence | ||||
| Motor vehicle–related injury | 1.41 (0.87-2.30) | 0.99 (0.91-1.07) | 0.82 (0.48-1.40) | 1.98 (1.21-3.23)b |
| Fall-related injury | 1.03 (0.69-1.53) | 1.12 (1.05-1.20)b | 0.80 (0.55-1.18) | 1.12 (0.77-1.62) |
| Unintentional poisoning injury | 1.62 (0.65-4.06) | 1.09 (0.96-1.24) | 1.44 (0.51-4.10) | 0.80 (0.31-2.09) |
| Experiencing violence | 1.35 (0.71-2.57) | 1.07 (0.99-1.16) | 1.10 (0.60-2.01) | 1.07 (0.60-1.91) |
| Socioeconomic status | ||||
| Low education | 1.29 (1.08-1.54)b | 1.02 (1.00-1.05) | 0.96 (0.81-1.14) | 1.03 (0.88-1.21) |
| Welfare benefits | 1.94 (1.37-2.75)b | 1.09 (1.03-1.15)b | 0.89 (0.63-1.27) | 1.17 (0.87-1.56) |
| Long-term unemployment | 1.21 (0.93-1.58) | 1.03 (1.00-1.07) | 0.93 (0.72-1.21) | 0.90 (0.71-1.14) |
| Disability pension | 1.35 (0.79-2.33) | 1.06 (0.97-1.15) | 1.51 (0.88-2.58) | 1.04 (0.60-1.80) |
| Antisocial behaviors and suicidality | ||||
| Violent crime arrest | 2.23 (1.59-3.13)b | 1.18 (1.11-1.24)b | 1.05 (0.77-1.44) | 1.27 (0.94-1.72) |
| Substance misuse | 1.44 (1.05-1.99) | 1.13 (1.07-1.19)b | 1.13 (0.84-1.54) | 1.09 (0.81-1.46) |
| Suicidal behavior | 1.34 (0.78-2.31) | 1.11 (1.03-1.19)b | 1.32 (0.77-2.27) | 1.21 (0.73-2.02) |
Abbreviations: aHR, adjusted hazard ratio; aRR, adjusted rate ratio.
The estimates refer to within-family estimates comparing siblings who had been placed in care but were differentially exposed to placement characteristics. The estimates were further adjusted for sex, birth year, birth order, parental age at birth, urbanicity, single-parent household, family income, parental welfare benefits, and disability pension in addition to the following preplacement confounders: attention-deficit/hyperactivity disorder; autism spectrum disorder; intellectual disability; communication disorder; learning disorder; motor disorder; other neurodevelopmental disorders; any neurological disorder, conduct disorder, or oppositional defiant disorder; motor vehicle–related injury, fall-related injury, unintentional poisoning injury, and experiencing violence. The estimates for premature mortality as outcome lacked sufficient statistical power to be presented.
Estimates with false discovery rate–corrected P values below .05.
Discussion
In this nationwide cohort study more than 885 000 Finnish children, we examined the associations between being in out-of-home care at any time from birth until age 15 with a range of adverse health and social outcomes in adulthood. We monitored individuals for an average of nearly a decade from their 15th birthdays and examined differentially exposed full siblings to account for various sources of confounding. We report 3 principal findings.
First, we found that being placed in out-of-home care was associated with all examined outcomes, including important morbidities, such as common psychiatric disorders, suicidality, and injury, as well as premature mortality. In addition, risks of behavioral outcomes, such as experiencing violence and antisocial behaviors, were increased. We confirmed and replicated these findings by examining siblings who had never been placed in out-of-home care as comparators, which allowed us to account for an aggregate of all familial confounders that were shared between the siblings (eg, early family environments and genetic risks). These analyses indicated that out-of-home placement was associated with a 1.4- to 5-fold increased risk of the outcomes with the largest effect sizes being observed for violent crime arrests and poisoning injuries. In terms of absolute risks, those placed in out-of-home care had cumulative incidence rates of 24.6% for violent crime arrests and 3.1% for poisoning injuries compared with 5.1% and 0.6% among those who had never been placed in out-of-home care. Various sensitivity analyses ruled out common sources of bias and suggested that the main findings were consistent over varying economic periods.
Two previous studies, based on a total of 1384 siblings and examining within-family associations between long-term foster care placement and social and health outcomes in adulthood, found that those who were placed either did not differ from their siblings or had worse outcomes.17,18 However, the limited statistical power in these studies likely explains the lack of associations. Investigations using other natural experimental approaches, such as child welfare policy reforms19 and rotationally assigned child welfare investigators,20,28,29 have reported mixed results. An advantage of our study compared with these is that we were able to study the placement trajectories of the individuals in our sample throughout their entire childhood and adolescence to obtain more precise estimates of the associations with a large number of health and social outcomes in adulthood.
Second, children who were primarily placed in institutional care were around twice as likely as their siblings who were placed in foster care settings to experience 5 of the tested long-term outcomes. Although consistent with the literature,30 our study was also able to replicate the findings with careful adjustments for preplacement behavioral problems and traumatic injuries in addition to duration of care, placement instability, age at first placement, and familial confounders. The increased rates of adverse outcomes observed in children placed in institutionalized care cannot therefore be solely attributed to their preplacement behavioral problems. Instead, our findings suggest that environmental factors within the institutions may contribute to the risk increases, which could include peer influences, child-to-caregiver ratio, staff turnover, and training.31,32 The specific etiological relevance of these factors needs clarification in large-scale quasiexperimental studies to inform the development of effective interventions. Although we did not have access to data on the reasons for institutional placement, a US-based study33 found that nearly one-third of children who had been placed in restrictive care were placed there for reasons unrelated to their behavior. Our findings therefore emphasize that foster care placement should be prioritized and the quality of care improved.
Each additional placement episode was associated with an increased risk of many of the examined outcomes, whereas the duration of placement was not. To our knowledge, this is the first study that has demonstrated that the associations between placement instability and long-term outcomes remain following adjustments for shared unmeasured familial confounding. This finding is potentially important as children placed in out-of-home care had experienced a median of 2 different placements before reaching age 15 years. Reducing the risks of placement instability should therefore be considered, and could include the following strategies: comprehensive assessments of the needs of the children in care (eg, symptoms of posttraumatic stress and behavioral problems) to inform early interventions, improved strategies to match children with placement settings that meet the needs of the child at the outset, and improved training for caregivers to reduce the risks of conflicts.34
The strengths of our study included the use of Finnish national registers that allowed us to study 16 objectively measured outcomes in adulthood among more than 885 000 individuals, of whom approximately 30 000 had been placed in out-of-home care. To our knowledge, this is the first study to adjust for unmeasured familial confounding by adopting the full-sibling comparison design in a sample with negligible selection bias. Importantly, we were able to estimate the relative contributions of placement characteristics, including care settings, the number of placement episodes, duration of care, and age at first placement.
Limitations
This study has limitations. First, we cannot rule out the impact of residual genetic confounding as biological full siblings only share, on average, half of their cosegregating genes.16 However, even if we were to assume that such confounders would have attenuated the reported sibling estimates by half, the magnitude of the associations would remain large. Second, despite our large sample size, we were underpowered to examine premature mortality as an outcome in some of the analyses. Third, although we carefully attempted to account for behavioral problems by adjusting for clinical diagnoses of conduct and oppositional defiant disorders in addition to neurodevelopmental disorders, we were unable account for less severe problems that did not meet clinical criteria for such conditions. Given the negligible attenuation of the reported associations following adjustments for severe behavioral problems, it seems unlikely that less severe measures would have stronger explanatory power. Although clinical studies are warranted to explore this possibility, we note that alternative approaches to measuring behavioral traits, such as self-reporting, commonly result in selection bias and measurement error, the latter of which further inflates in sibling comparison designs and typically causes artificial reductions of the associations.35
The generalizability of our findings is an important consideration. The cumulative exposure to out-of-home care placement up to age 18 years is similar between Finland (5.8%)36 and the US (5.9%),37 but lower estimates have been reported in other high-income countries.38 Although direct cross-country comparisons remain challenging owing to large differences in selection factors into out-of-home care and the specific services offered in care,39 the weight of the evidence suggests that being placed in out-of-home care is associated with poorer long-term outcomes in high-income countries.5,6,40 Large-scale and genetically informative replication efforts in other countries are nevertheless warranted to investigate whether the magnitude of these associations vary between countries.
Conclusions
In this nationwide Finnish cohort study, we found that children who were placed in out-of-home care were more likely than their siblings who had never been placed to meet criteria for multiple adverse health and social outcomes in adulthood, even after adjustments for preplacement behavioral problems, traumatic incidents, and family background. Institutionalization and placement instability further contributed to the risks of many of the outcomes. While children who have been placed in out-of-home care have, on average, considerably worse outcomes in adulthood than their peers and siblings, it is important to bear in mind that many children lead better lives as a result of having been placed in out-of-home care. Our findings therefore suggest that out-of-home care placement should remain a last resort intervention, and efforts should be directed toward to improving the quality of care and reducing institutional placement and placement instability.
eAppendix 1. The diagnostic validity of the psychiatric disorders
eAppendix 2. Complementary sensitivity analyses
eTable 1. ICD codes
eTable 2. Penal codes
eTable 3. ATC codes
eTable 4. Measured confounders
eTable 5. Descriptive statistics for placed children either without siblings or with siblings stratified across those with siblings who were all placed and those who had at least one sibling who was never placed
eTable 6. Out-of-home care placement characteristics on poor adulthood functioning among siblings born 1991-2000 in families where all children had been exposed to at least one placement episode (n = 6755)
eFigure 1. Flowchart of the samples
eFigure 2. Associations between any out-of-home placement episode (ages 0-15 years) and poor functioning in adulthood stratified by sex
eFigure 3. Associations between any out-of-home placement episode (ages 0-15 years) and poor functioning in adulthood stratified across cohorts born 1986-2000 and 1991-2000
eFigure 4. Associations between any out-of-home placement episode (ages 0-15 years) and alternative outcome definitions of psychiatric morbidity and substance misuse
eFigure 5. Associations between any out-of-home placement episode (ages 0-15 years vs. 0-21 years) and subsequent poor functioning
eFigure 6. Associations between any out-of-home placement episode (ages 0-15 years) and poor functioning in adulthood, adjusted for additional measured confounders and mediators
eFigure 7. Associations between any out-of-home placement episode (ages 0-15 years) and poor functioning in adulthood, stratified across the sibling and single-child family subsets
eFigure 8. Associations between any out-of-home placement episode (ages 0-15 years) and poor functioning in adulthood using cousins as comparators to account for unmeasured familial confounding in the extended family
eReferences
References
- 1.Desmond C, Watt K, Saha A, Huang J, Lu C. Prevalence and number of children living in institutional care: global, regional, and country estimates. Lancet Child Adolesc Health. 2020;4(5):370-377. doi: 10.1016/S2352-4642(20)30022-5 [DOI] [PubMed] [Google Scholar]
- 2.U.S. Department of Health and Human Services . The AFCARS Report. Published June 23, 2020. Accessed December 27, 2020. https://www.acf.hhs.gov/sites/default/files/documents/cb/afcarsreport27.pdf
- 3.Rosinsky K, Connelly D; National Children’s Advocacy Center. Child welfare financing SFY 2014: a survey of federal, state, and local expenditures. December 30, 2020. https://www.childtrends.org/wp-content/uploads/2016/10/2016-53ChildWelfareFinancingSFY2014-1.pdf
- 4.Better Care Network . Children in Alternative Care: National Surveys 2nd Edition. Accessed December 27, 2020. https://bettercarenetwork.org/library/social-welfare-systems/data-and-monitoring-tools/children-in-alternative-care-national-surveys-2nd-edition
- 5.Gypen L, Vanderfaeillie J, De Maeyer S, Belenger L, Van Holen F. Outcomes of children who grew up in foster care: systematic-review. Child Youth Serv Rev. 2017;76:74-83. doi: 10.1016/j.childyouth.2017.02.035 [DOI] [Google Scholar]
- 6.Kääriälä A, Hiilamo H. Children in out-of-home care as young adults: a systematic review of outcomes in the Nordic countries. Child Youth Serv Rev. 2017;79:107-114. doi: 10.1016/j.childyouth.2017.05.030 [DOI] [Google Scholar]
- 7.Wall-Wieler E, Almquist Y, Liu C, Vinnerljung B, Hjern A. Intergenerational transmission of out-of-home care in Sweden: a population-based cohort study. Child Abuse Negl. 2018;83:42-51. doi: 10.1016/j.chiabu.2018.07.007 [DOI] [PubMed] [Google Scholar]
- 8.Simkiss DE, Stallard N, Thorogood M. A systematic literature review of the risk factors associated with children entering public care. Child Care Health Dev. 2013;39(5):628-642. doi: 10.1111/cch.12010 [DOI] [PubMed] [Google Scholar]
- 9.Meinhofer A, Angleró-Díaz Y. Trends in foster care entry among children removed from their homes because of parental drug use, 2000 to 2017. JAMA Pediatr. 2019;173(9):881-883. doi: 10.1001/jamapediatrics.2019.1738 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Sariaslan A, Fazel S, D’Onofrio BM, et al. Schizophrenia and subsequent neighborhood deprivation: revisiting the social drift hypothesis using population, twin and molecular genetic data. Transl Psychiatry. 2016;6:e796. doi: 10.1038/tp.2016.62 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hyytinen A, Ilmakunnas P, Johansson E, Toivanen O. Heritability of lifetime earnings. J Econ Inequal. 2019;17(3):319-335. doi: 10.1007/s10888-019-09413-x [DOI] [Google Scholar]
- 12.Hill WD, Davies NM, Ritchie SJ, et al. Genome-wide analysis identifies molecular systems and 149 genetic loci associated with income. Nat Commun. 2019;10(1):5741. doi: 10.1038/s41467-019-13585-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Burt SA. Are there meaningful etiological differences within antisocial behavior? results of a meta-analysis. Clin Psychol Rev. 2009;29(2):163-178. doi: 10.1016/j.cpr.2008.12.004 [DOI] [PubMed] [Google Scholar]
- 14.Sariaslan A, Larsson H, Fazel S. Genetic and environmental determinants of violence risk in psychotic disorders: a multivariate quantitative genetic study of 1.8 million Swedish twins and siblings. Mol Psychiatry. 2016;21(9):1251-1256. doi: 10.1038/mp.2015.184 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Pettersson E, Lichtenstein P, Larsson H, et al. ; Attention Deficit/Hyperactivity Disorder Working Group of the iPSYCH-Broad-PGC Consortium; Autism Spectrum Disorder Working Group of the iPSYCH-Broad-PGC Consortium; Bipolar Disorder Working Group of the PGC; Eating Disorder Working Group of the PGC; Major Depressive Disorder Working Group of the PGC; Obsessive Compulsive Disorders and Tourette Syndrome Working Group of the PGC; Schizophrenia CLOZUK; Substance Use Disorder Working Group of the PGC . Genetic influences on eight psychiatric disorders based on family data of 4 408 646 full and half-siblings, and genetic data of 333 748 cases and controls. Psychol Med. 2019;49(7):1166-1173. doi: 10.1017/S0033291718002039 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.D’Onofrio BM, Lahey BB, Turkheimer E, Lichtenstein P. Critical need for family-based, quasi-experimental designs in integrating genetic and social science research. Am J Public Health. 2013;103 Suppl(Suppl 1):S46-55. doi: 10.2105/AJPH.2013.301252 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Brännström L, Vinnerljung B, Hjern A. Outcomes in adulthood after long-term foster care: a sibling approach. Child Maltreat. 2020;25(4):383-392. doi: 10.1177/1077559519898755 [DOI] [PubMed] [Google Scholar]
- 18.Vinnerljung B. Fosterbarn Som Vuxna (Foster Children as Adults). Arkiv; 1996. [Google Scholar]
- 19.Warburton WP, Warburton RN, Sweetman A, Hertzman C. The impact of placing adolescent males into foster care on education, income assistance, and convictions. Can d’économique. 2014;47(1):35-69. doi: 10.1111/caje.12064 [DOI] [Google Scholar]
- 20.Doyle JJ Jr, Aizer A. Economics of child protection: maltreatment, foster care, and intimate partner violence. Annu Rev Econom. 2018;10(1):87-108. doi: 10.1146/annurev-economics-080217-053237 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bald A, Chyn E, Hastings J, Machelett M. The causal impact of removing children from abusive and neglectful homes. National Bureau of Economic Research. 2019. doi: 10.3386/w25419 [DOI] [Google Scholar]
- 22.Gross M, Baron EJ. Temporary stays and persistent gains: the causal effects of foster care. SSRN Electron J. 2020. doi: 10.2139/ssrn.3576640 [DOI]
- 23.Sund R, Gissler M, Hakulinen T, Rosén M. Use of Health Registers. In: Ahrens W, Pigeot I, eds. Handbook of Epidemiology. Springer; 2014. doi: 10.1007/978-0-387-09834-0_5 [DOI] [Google Scholar]
- 24.Official Statistics of Finland. Quality Description: Causes of death 2017. Accessed May 11, 2020. https://www.stat.fi/til/ksyyt/2017/ksyyt_2017_2018-12-17_laa_001_en.html
- 25.Ronaldson A, Elton L, Jayakumar S, Jieman A, Halvorsrud K, Bhui K. Severe mental illness and health service utilisation for nonpsychiatric medical disorders: a systematic review and meta-analysis. PLoS Med. 2020;17(9):e1003284. doi: 10.1371/journal.pmed.1003284 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Sariaslan A, Mikkonen J, Aaltonen M, Hiilamo H, Martikainen P, Fazel S. No causal associations between childhood family income and subsequent psychiatric disorders, substance misuse and violent crime arrests: a nationwide Finnish study of >650 000 individuals and their siblings. Int J Epidemiol. 2021;dyab099. doi: 10.1093/ije/dyab099 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc (B). 1995;57(1):289-300. doi: 10.1111/j.2517-6161.1995.tb02031.x [DOI] [Google Scholar]
- 28.Doyle JJ. Causal effects of foster care: an instrumental-variables approach. Child Youth Serv Rev. 2013;35(7):1143-1151. doi: 10.1016/j.childyouth.2011.03.014 [DOI] [Google Scholar]
- 29.Fallesen P. Time well spent: the duration of foster care and early adult labor market, educational, and health outcomes. J Adolesc. 2013;36(6):1003-1011. doi: 10.1016/j.adolescence.2013.08.010 [DOI] [PubMed] [Google Scholar]
- 30.Li D, Chng GS, Chu CM. Comparing long-term placement outcomes of residential and family foster care: a meta-analysis. Trauma Violence Abuse. 2019;20(5):653-664. doi: 10.1177/1524838017726427 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Robst J, Armstrong M, Dollard N. Peer contagion among youth in Medicaid-funded residential mental health treatment programs. Child Youth Serv Rev. 2011;33(5):651-655. doi: 10.1016/j.childyouth.2010.11.007 [DOI] [Google Scholar]
- 32.Barone L, Dellagiulia A, Lionetti F. When the primary caregiver is missing: investigating proximal and distal variables involved in institutionalised children’s adjustment. Child Abus Rev. 2016;25(6):454-468. doi: 10.1002/car.2365 [DOI] [Google Scholar]
- 33.James S, Landsverk J, Leslie LK, Slymen DJ, Zhang J. Entry into restrictive care settings: placements of last resort? Fam Soc. 2008;89(3):348-359. doi: 10.1606/1044-3894.3760 [DOI] [Google Scholar]
- 34.Konijn C, Admiraal S, Baart J, et al. Foster care placement instability: a meta-analytic review. Child Youth Serv Rev. 2019;96:483-499. doi: 10.1016/j.childyouth.2018.12.002 [DOI] [Google Scholar]
- 35.Frisell T, Öberg S, Kuja-Halkola R, Sjölander A. Sibling comparison designs: bias from non-shared confounders and measurement error. Epidemiology. 2012;23(5):713-720. doi: 10.1097/EDE.0b013e31825fa230 [DOI] [PubMed] [Google Scholar]
- 36.Forsell M, Heino T, Kuoppala T. Kodin ulkopuolelle sijoitettujen lasten sijoitusten kestot 2016 (Placement durations among children in out-of-home care in 2016). SVTT Statistic Report. Accessed February 20, 2021. http://urn.fi/URN:NBN:fi-fe2018042318219
- 37.Wildeman C, Emanuel N. Cumulative risks of foster care placement by age 18 for US children, 2000-2011. PLoS One. 2014;9(3):e92785. doi: 10.1371/journal.pone.0092785 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Ubbesen MB, Gilbert R, Thoburn J. Cumulative incidence of entry into out-of-home care: changes over time in Denmark and England. Child Abuse Negl. 2015;42:63-71. doi: 10.1016/j.chiabu.2014.10.006 [DOI] [PubMed] [Google Scholar]
- 39.Gilbert N. A comparative study of child welfare systems: abstract orientations and concrete results. Child Youth Serv Rev. 2012;34(3):532-536. doi: 10.1016/j.childyouth.2011.10.014 [DOI] [Google Scholar]
- 40.Chaiyachati BH, Wood JN, Mitra N, Chaiyachati KH. All-cause mortality among children in the US foster care system, 2003-2016. JAMA Pediatr. 2020;174(9):896-898. doi: 10.1001/jamapediatrics.2020.0715 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eAppendix 1. The diagnostic validity of the psychiatric disorders
eAppendix 2. Complementary sensitivity analyses
eTable 1. ICD codes
eTable 2. Penal codes
eTable 3. ATC codes
eTable 4. Measured confounders
eTable 5. Descriptive statistics for placed children either without siblings or with siblings stratified across those with siblings who were all placed and those who had at least one sibling who was never placed
eTable 6. Out-of-home care placement characteristics on poor adulthood functioning among siblings born 1991-2000 in families where all children had been exposed to at least one placement episode (n = 6755)
eFigure 1. Flowchart of the samples
eFigure 2. Associations between any out-of-home placement episode (ages 0-15 years) and poor functioning in adulthood stratified by sex
eFigure 3. Associations between any out-of-home placement episode (ages 0-15 years) and poor functioning in adulthood stratified across cohorts born 1986-2000 and 1991-2000
eFigure 4. Associations between any out-of-home placement episode (ages 0-15 years) and alternative outcome definitions of psychiatric morbidity and substance misuse
eFigure 5. Associations between any out-of-home placement episode (ages 0-15 years vs. 0-21 years) and subsequent poor functioning
eFigure 6. Associations between any out-of-home placement episode (ages 0-15 years) and poor functioning in adulthood, adjusted for additional measured confounders and mediators
eFigure 7. Associations between any out-of-home placement episode (ages 0-15 years) and poor functioning in adulthood, stratified across the sibling and single-child family subsets
eFigure 8. Associations between any out-of-home placement episode (ages 0-15 years) and poor functioning in adulthood using cousins as comparators to account for unmeasured familial confounding in the extended family
eReferences
