Foster care intervention protects cognitive control skills in young adults exposed to institutional rearing
Open accessThe DOI matches its Crossref record; 46 of the 46 references it cites were found in a registry or an index; 29 of the 40 sources its claims cite were checked for support.
Abstract
Abstract Prior work from the Bucharest Early Intervention Project (BEIP) suggests that institutional care impairs cognitive control, whereas foster care may help preserve these skills. However, it remains unclear whether these effects persist into adulthood. We examined EEG data from a flanker task administered at BEIP’s 22-year follow-up. Participants ( N = 127; Age: M = 23.5 years, SD = 1.0, range = 21.4–25.6; 61% Romanian ethnicity) were initially enrolled during childhood while in institutional care in Bucharest, Romania and included individuals randomized to remain in institutional care (CAUG; n = 42; 50% female) or transition to foster care (FCG; n = 46; 54% female), along with never-institutionalized controls (NIG; n = 39; 67% female). At the 22-year follow-up, young adults in the FCG showed event-related potential (ERP) markers of performance monitoring comparable to the NIG but different from the CAUG ( η2 p = 0.06). In contrast, the CAUG showed convergent evidence of disrupted cognitive control, including poorer behavioral performance and altered neural responses across ERP and time–frequency measures relative to the NIG. These neural disruptions were associated with lower IQ and elevated mental health symptoms. Overall, results highlight enduring effects of early caregiving on cognitive control in adulthood.
Overview
CoChat Summary
TL;DR
At the 22-year follow-up of the Bucharest Early Intervention Project, young adults randomized to foster care as infants showed event-related potential markers of performance monitoring more similar to never-institutionalized controls than to those who remained in institutional care. However, error-related theta power showed persistent effects of institutionalization for both ever-institutionalized groups, and neural measures were linked to IQ and mental health symptoms.
Key findings
- In intent-to-treat analyses, the care-as-usual group (CAUG) had larger (more negative) ERP amplitudes than the foster care group (FCG), \(F(1,84.27)=5.14, p=.026, \eta^2_p=0.06\); CAUG mean = -0.41 µV vs. FCG mean = 0.42 µV. In complete-group analyses, CAUG had significantly larger (more negative) amplitudes than never-institutionalized controls (NIG; CAUG \(M=-0.40\) µV vs. NIG \(M=1.08\) µV, \(p=.001, p\)-adj \(=.004, d=-0.85\)), while FCG vs. NIG was not significant.
- Behavioral performance was poorer in CAUG than NIG: NIG outperformed CAUG on accuracy (\(t(162)=2.99, p=.003, p\)-adj \(=.005, d=0.47\)), and NIG responded faster than both FCG and CAUG on overall reaction time (NIG \(M=381.59\) ms vs. FCG \(M=416.46\) ms, \(p=.005, p\)-adj \(=.015, d=-0.45\); vs. CAUG \(M=412.07\) ms, \(p=.016, p\)-adj \(=.024, d=-0.40\)).
- Error-related theta power was greater in NIG than both CAUG (\(p=.005, p\)-adj \(=.015, d=-0.66\)) and FCG (\(p=.008, p\)-adj \(=.021, d=-0.83\)), with no significant CAUG–FCG difference, indicating an effect of institutionalization exposure rather than a foster care intervention effect on theta.
- Neural measures were associated with cognitive and psychosocial functioning: higher IQ correlated with smaller (less negative) ERN amplitude (\(r=.34, p=.001\)), smaller CRN amplitude (\(r=.23, p=.014\)), and greater error theta power (\(r=.30, p=.004\)); greater ASR total problems correlated with larger (more negative) ERN amplitude (\(r=-.23, p=.031\)) and lower error theta power (\(r=-.33, p=.002\)). Exploratory mediation found ERN amplitude mediated the ITT effect on IQ (ACME = 2.11, 95% CI [0.18, 4.75], \(p=.024\); 31.5% mediated).
Methods
- Data came from the Bucharest Early Intervention Project (BEIP), a randomized controlled trial in which institutionalized infants in Bucharest, Romania were randomized to remain in institutional care (CAUG) or be placed in high-quality foster care (FCG), alongside a never-institutionalized group (NIG). The 22-year follow-up analyzed \(N=127\) (CAUG \(n=42\), FCG \(n=46\), NIG \(n=39\); mean age approximately 23.5 years) using an intent-to-treat approach.
- Participants completed a flanker task with EEG recording; analyses focused on incongruent trials, ERP components (ERN/CRN at FCz), and time-frequency theta power (4–8 Hz at FCz). Accuracy and reaction time were analyzed with repeated-measures ANOVAs; EEG measures used linear mixed-effects models with group and response type as fixed effects, plus correlations and exploratory mediation models involving WAIS-IV Full Scale IQ and Adult Self-Report scores.
Limitations
- The paper states that although retention at the 22-year assessment was high (77.9% CAUG and 79.4% FCG), any attrition introduces possible selection bias; they used an analytic approach retaining nearly all participants with any flanker EEG data. The paper also states that prior BEIP flanker reports varied in preprocessing pipelines, EEG scoring, and statistical approaches, so it remains unclear whether disparate findings across waves reflect those factors versus developmental change, and harmonized longitudinal analyses are needed.
- The paper notes that the adaptive flanker task, which sped up or slowed down based on overall accuracy, limits conclusions based on task behavior alone. Evident from the text: the study focuses on the 22-year assessment only, so it cannot directly test within-person trajectories of error monitoring across development.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
14 Valid · 1 Not valid · 4 Not checked · 74 Own results and interpretations
Some cited sources support this claim and some don’t — read each source’s verdict below.
In the paper
“Previous work consistently finds differences in ERN amplitude between children and adolescents exposed to early institutional care and those who have never been institutionalized (Debnath et al., 2023; Loman et al., 2013; McDermott et al., 2013; TrollerRenfree et al., 2016).”
What the sources say
- Disputedcites Debnath et al., 2023 sourceExperimental — accuracy not yet measured
The Introduction explicitly says previous studies of ERN in children exposed to institutional care produced mixed findings, including studies finding no amplitude difference. The Results report a smaller ERN in the prolonged-care group than in never-institutionalized adolescents, but no significant difference between the foster-care group and never-institutionalized adolescents. Thus the paper does not support the claim that differences are consistently found. (LLM, conf=0.88)
Checked by a reasoning model — may be wrong.
- Partly supportedcites Loman et al., 2013 sourceExperimental — accuracy not yet measured
The Results (Flanker Event-Related Potentials; page not provided) report a smaller ERN amplitude in post-institutionalized children than in non-adopted children. But the Results (Go/No-go Event-Related Potentials; page not provided) report no group difference in ERN amplitude, so this paper does not support the claim that such differences are consistently found across tasks. (LLM, conf=0.88)
- Disputedcites McDermott et al., 2013 sourceExperimental — accuracy not yet measured
The Results (Event-related potentials, ERN) report no significant difference in ERN peak amplitude among the institutional-care, foster-care, and never-institutionalized groups (p = 0.275). The Discussion likewise states that group differences were not found on the ERN. Although the Introduction describes differences reported in earlier studies, this paper's own finding runs contrary to the claim that the cited work consistently finds such differences. (LLM, conf=0.90)
Checked by a reasoning model — may be wrong.
boostedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Disputed
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Findings both from BEIP and other independent samples have revealed long-term error monitoring differences in previously institutionalized children that persist at least through adolescence (Buzzell et al., 2020; Debnath et al., 2023; Frenkel et al., 2020; Loman et al., 2013; McDermott et al., 2013; Troller-Renfree et al., 2016).”
What the sources say
- Supportedcites Buzzell et al., 2020 source
The Results section (‘Effects of early experience on 16-year mediofrontal theta’; page not provided) reports lower error-related mediofrontal theta at age 16 in the BEIP care-as-usual group than in both the foster-care and never-institutionalized groups. This supports the BEIP finding of an error-monitoring difference persisting into adolescence; the other cited studies, not this paper, are offered for the independent-samples portion of the sentence. (LLM, conf=0.91)
- Supportedcites Debnath et al., 2023 source
The Participants section describes the BEIP follow-up at age 16, and the Conclusions state that institutional care can disrupt error monitoring through adolescence. The study distinguishes adolescents who continued receiving care as usual from those placed in foster care, whose error monitoring was comparable to never-institutionalized peers. This paper supports the BEIP portion of the jointly cited sentence; it does not itself study an independent sample. (LLM, conf=0.90)
- Partly supportedcites Troller-Renfree et al., 2016 sourceExperimental — accuracy not yet measured
The Discussion reports error-monitoring differences in the BEIP sample at age 12, supporting persistence from early institutionalization into early adolescence. This paper does not establish persistence through later adolescence; its Introduction describes findings from another sample at ages 8–11 as mixed. (LLM, conf=0.82)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“In the BEIP, adolescents in the CAUG had lower mid-frontal theta power compared to those in the FCG at 16 years (Buzzell et al., 2020; Debnath et al., 2023; Wade et al., 2023)”
What the sources say
- Partly supportedcites Buzzell et al., 2020 sourceExperimental — accuracy not yet measured
The Results section, “Effects of early experience on 16-year mediofrontal theta” (page not provided), reports lower theta power in the CAUG than the FCG on error trials. It found no group difference on correct trials and no overall main effect of group, so the unqualified claim of lower theta power omits an important condition. (LLM, conf=0.88)
- Supportedcites Debnath et al., 2023 source
The Introduction identifies this as a BEIP follow-up at age 16. The Results (Theta power) report significantly lower theta power at the mid-frontal FCz electrode in the CAUG than in the FCG (p = .01). (LLM, conf=0.94)
- Partly supportedcites Wade et al., 2023 sourceExperimental — accuracy not yet measured
The Results (Step 2; page unavailable) report that mediofrontal theta power at age 16 declined with increasing dependent stressful life events in the care-as-usual group but remained high in the foster-care group. They also state that group alone was not significantly associated with theta power, so the paper does not support an unqualified claim that the care-as-usual group had lower power. (LLM, conf=0.88)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Nevertheless, across multiple ages and even in the absence of intervention group differences, ERN amplitude has been associated with psychosocial functioning in the BEIP at age 8, 12, and 16 years (Debnath et al., 2023; McDermott et al., 2013; Troller-Renfree et al., 2016), albeit sometimes in different directions, likely due to differences in analytic approach”
What the sources say
- Supportedcites Debnath et al., 2023 source
The Introduction reports associations between ERN amplitude and socio-emotional or externalizing problems at ages 8 and 12, and notes that the intervention groups did not differ in ERN amplitude at age 8. The Results (Mental health) report an association at age 16 in the care-as-usual group, though not in the foster-care or community groups. (LLM, conf=0.90)
- Supportedcites McDermott et al., 2013 source
The Participants section identifies this as a BEIP assessment at age 8. The Event-related potentials results report no significant group difference in ERN amplitude, while the Response monitoring results and Discussion link ERN amplitude to externalizing-ADHD behavior and academic competence within the foster-care group. The other ages in the jointly cited sentence need not be established by this source alone. (LLM, conf=0.90)
- Supportedcites Troller-Renfree et al., 2016 source
This source supports the age-12 portion of the multi-citation claim; the other cited papers are offered for the other ages. In the Results (Mental health), ΔERN moderated the association between time in institutional care and externalizing/ADHD behaviors, but not internalizing behaviors. The Results (ERP analyses) report only a marginal foster-care versus care-as-usual difference in ΔERN (p=.063), consistent with the absence of a significant intervention-group difference. (LLM, conf=0.90)
hedgedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“By age 16 years, intervention effects were more clearly evident such that the CAUG showed significantly smaller (less negative) Δ ERNs than the FCG (Debnath et al., 2023).”
What the sources say
- Supportedcites Debnath et al., 2023 source
The Introduction identifies this as the 16-year follow-up and says the group difference was absent at age 8 but emerged at age 12. The Results (ERN) report that the FCG had a significantly larger, more negative ΔERN than the CAUG (p = .032), equivalent to the CAUG having a smaller, less negative ΔERN. (LLM, conf=0.94)
boostedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
29 more claims
Create a free account to see all 93 claims and every source verdict.
- It may state two separate claims as one. 39
- Probably not a substantial claim. 22
- Not yet classified. 1
Statistics that don’t add up
Each reported p-value recomputed from its test statistic and degrees of freedom.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“t (164) = 0.33, p = .725”
Reported p = .725 · recomputed p = .742
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“t (164) = 0.34, p = .707”
Reported p = .707 · recomputed p = .734
Statistics recomputed
Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.
- Doesn’t add upt(164) = 0.33reported p = .725 · recomputed p = .742
“t (164) = 0.33, p = .725”
- Doesn’t add upt(164) = 0.34reported p = .707 · recomputed p = .734
“t (164) = 0.34, p = .707”
- ConsistentF(1, 81) = 149.6reported p < .001 · recomputed p < .001
“F (1,81) = 149.60, p < .001”
- ConsistentF(1, 81) = 3.74reported p = .057 · recomputed p = .057
“F (1, 81) = 3.74, p = .057”
- ConsistentF(1, 81) = 1.87reported p = .175 · recomputed p = .175
“F (1, 81) = 1.87, p = .175”
49 more recomputed tests
Create a free account to see every test the Stats check recomputed.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Flanker task
“if a participant ' s cumulative accuracy was greater than 60%, the subsequent trial ' s stimulus presentation was shortened by 50 ms.”
§ Flanker task
“If cumulative accuracy was greater than 60%, the subsequent trial ' s stimulus presentation was lengthened by 50 ms.”
Citations
Cites
Delorme, A., & Makeig, S. (2004) · Journal of Neuroscience Methods · doi:10.1016/j.jneumeth.2003.10.009 · Cited by 24,371
Oostenveld, R., Fries, P., Maris, E., & Schoffelen, J.-M. (2011) · Computational Intelligence and Neuroscience · doi:10.1155/2011/156869 · Cited by 10,323
mediation:R Package for Causal Mediation Analysis
Tingley, D., Yamamoto, T., Hirose, K., Keele, L., & Imai, K. (2014) · Journal of Statistical Software · doi:10.18637/jss.v059.i05 · Cited by 4,155
Frontal theta as a mechanism for cognitive control
Cavanagh, J. F., & Frank, M. J. (2014) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2014.04.012 · Cited by 2,726
Wechsler Adult Intelligence Scale--Fourth Edition
Wechsler, D. (2012) · PsycTESTS Dataset · doi:10.1037/t15169-000 · Cited by 2,232
Falkenstein, M., Hohnsbein, J., Hoormann, J., & Blanke, L. (1991) · Electroencephalography and Clinical Neurophysiology · doi:10.1016/0013-4694(91)90062-9 · Cited by 1,706
The Timing of Action-Monitoring Processes in the Anterior Cingulate Cortex
Veen, V. V., & Carter, C. S. (2002) · Journal of Cognitive Neuroscience · doi:10.1162/08989290260045837 · Cited by 1,298
ADJUST: An automatic EEG artifact detector based on the joint use of spatial and temporal features
Mognon, A., Jovicich, J., Bruzzone, L., & Buiatti, M. (2011) · Psychophysiology · doi:10.1111/j.1469-8986.2010.01061.x · Cited by 1,290
Localization of a Neural System for Error Detection and Compensation
Dehaene, S., Posner, M. I., & Tucker, D. M. (1994) · Psychological Science · doi:10.1111/j.1467-9280.1994.tb00630.x · Cited by 1,255
FASTER: Fully Automated Statistical Thresholding for EEG artifact Rejection
Nolan, H., Whelan, R., & Reilly, R. B. (2010) · Journal of Neuroscience Methods · doi:10.1016/j.jneumeth.2010.07.015 · Cited by 1,077
Research works printed but not linked: 1
- Achenbach, 2003, Manual for the ASEBA adult forms and profiles
Metadata
- Authors
- Valadez, E. A., Jopling, E., Macera, E., Wade, M., Zeanah, C. H., Nelson, C. A., & Fox, N. A.
- Affiliations
- University of Southern California (US)
- Dalhousie University (CA)
- University of Toronto (CA)
- Tulane University (US)
- Boston Children's Hospital (US)
- University of Maryland, College Park (US)
- Corresponding author
- Valadez, Emilio A.
- Venue
- Development and Psychopathology
- Publisher
- Cambridge University Press (CUP)
- Published
- 2026-09-23
- Pages
- 1-10
- ISSN
- 0954-5794 (ISSN-L)
- 1469-2198
- OpenAlex
- W7214090914
- Cited by
- 0
- References
- 40
- FWCI
- 0.00
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Locations
- Development and Psychopathology (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Funders
- John D. and Catherine T. MacArthur Foundation
- Binder Family Foundation (R01MH091363)
- Jacobs Foundation
- National Institutes of Health
- National Institute of Mental Health (R01MH091363, K23MH130751)
- Topics
- Social Sciences › Sociology and Political Science › Child Welfare and Adoption
- Social Sciences › Psychology › Clinical Psychology › Child Abuse and Trauma
- Social Sciences › Education › Early Childhood Education and Development
- Keywords
- foster care, institutional upbringing, cognitive control, performance monitoring, Bucharest Early Intervention Project, event-related potentials, EEG, flanker task, early caregiving, young adults, IQ, psychiatric symptoms
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-09-29, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-09-29 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Foster care intervention protects cognitive control skills in young adults exposed to institutional rearing. https://app.cochat.ai/papers/10.1017/s0954579426101928
- The paper
- Valadez, E. A., Jopling, E., Macera, E., Wade, M., Zeanah, C. H., Nelson, C. A., & Fox, N. A. (2026). Foster care intervention protects cognitive control skills in young adults exposed to institutional rearing. Development and Psychopathology. https://doi.org/10.1017/s0954579426101928
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.