The Bucharest Early Intervention Project (Romanian Orphans) – Charles Nelson, Nathan Fox, and Charles Zeanah

In the closing weeks of 1989, the violent overthrow and execution of Romanian dictator Nicolae Ceaușescu shattered the hermetic seal of one of the twentieth century’s most repressive totalitarian regimes. As Western journalists, humanitarian workers, and clinicians entered Romania, they uncovered an atrocity hidden within plain sight: a network of state-run pediatric warehouses known as leagăne (cradles), housing an estimated 170,000 abandoned, neglected, and systematically starved infants and young children. Confined to metal cribs in cavernous, silent wards, these children received minimal physical handling, virtually no cognitive or linguistic stimulation, and zero individualized emotional investment. The world was confronted with an unprecedented humanitarian disaster and an agonizing scientific puzzle regarding the limits of human resilience, the plasticity of the developing nervous system, and the biological consequences of severe psychosocial deprivation.

Recognizing the profound ethical and developmental stakes, a multidisciplinary triumvirate of American developmental scientists—neuroscientist Charles A. Nelson III of Harvard Medical School, developmental psychophysiologist Nathan A. Fox of the University of Maryland, and infant child psychiatrist Charles H. Zeanah Jr. of Tulane University School of Medicine—conceived the Bucharest Early Intervention Project (BEIP) in the year 2000. Designed at the invitation of Romanian pediatric leaders, the BEIP was structured as the first and only randomized controlled trial of foster care intervention for profoundly institutionalized children. By randomly assigning institutionalized infants either to remain in institutional “care-as-usual” or to enter high-quality, professionally supported foster homes created specifically by the research team, the BEIP sought to establish causal inferences regarding the impact of institutional rearing and the capacity of the brain to recover following environmental remediation.

Over more than two decades of rigorous longitudinal follow-up spanning infancy through early adulthood (ages 30 months, 42 months, 54 months, 8 years, 12 years, 16 years, and 21 years), the BEIP has fundamentally reshaped developmental neuroscience, pediatric medicine, clinical psychology, and international child welfare policy. The project’s findings demonstrated that early psychosocial neglect strikes at the structural core of the human brain, blunting electrophysiological activity, reducing gray and white matter volumes, and fracturing socio-emotional and cognitive trajectories. Crucially, the BEIP established the empirical reality of developmental “sensitive periods” in humans—demonstrating that remediating deprivation before the age of two yields striking cognitive, electrophysiological, and attachment recoveries, whereas chronic deprivation past this critical threshold leaves permanent neurobiological scars. The following comprehensive exploration documents the historical origins, methodological architecture, neurobiological discoveries, ethical complexities, and enduring global legacy of this landmark scientific endeavor.

1. Historical Context: Nicolae Ceaușescu’s Pronatalist Policies and Institutionalization

1.1 Decree 770 and Demographic Engineering

The institutional crisis that engulfed Romania was the direct consequence of deliberate, aggressive social engineering executed under the authoritarian regime of Nicolae Ceaușescu. Ascending to the leadership of the Romanian Communist Party in 1965, Ceaușescu harbored an obsession with transforming Romania from an agrarian society into an industrialized, heavily populated regional powerhouse. Convinced that economic modernization required a massive domestic workforce, Ceaușescu viewed the country’s declining fertility rate as an existential impediment to national grandeur. In October 1966, the regime promulgated Decree 770, an draconian pronatalist edict that outlawed abortion on demand and criminalized the importation, manufacture, and distribution of modern contraceptives.

To enforce Decree 770, the state deployed its pervasive security apparatus, the Securitate, directly into the reproductive lives of Romanian women. State-appointed gynecologists, colloquially and derisively termed the “menstrual police,” conducted mandatory monthly pelvic examinations on female factory workers and civil servants to detect early pregnancies and ensure they were carried to term. Unmarried individuals and married couples over the age of twenty-five who remained childless without a verified medical exemption were subjected to punitive “celibacy taxes,” which seized up to ten percent of their monthly wages. If a pregnancy miscarried, interrogators investigated the woman and her physician under suspicion of induced termination, driving thousands of desperate women toward catastrophic back-alley abortions that ultimately killed an estimated 10,000 women between 1966 and 1989.

The immediate demographic consequence was a short-term doubling of the birth rate in 1967–1968, generating a generation known as the decreței (children of the decree). However, this coerced population surge coincided with a disastrous domestic economic crisis during the late 1970s and 1980s, fueled by Ceaușescu’s decision to export Romania’s agricultural and industrial yield to liquidate its foreign sovereign debt. As basic food staples, heating fuel, and electricity were rigorously rationed, ordinary Romanian families faced profound poverty and systemic starvation. Unable to feed, clothe, or warm their children, parents were encouraged by state propaganda to surrender their offspring to public custody, under the Marxist-Leninist rationale that the socialist state was far better equipped to rear productive, collectivized citizens than the impoverished nuclear family.

1.2 The Leagăne Network and Structural Deprivation

To process the influx of surrendered children, the Romanian Ministry of Health established an expansive network of institutional facilities categorized strictly by the chronological age and developmental capacity of the pediatric population. Infants from birth through approximately three years of age were placed in leagăne (infant homes or “cradles”). If a child reached age three or four and exhibited typical physical and cognitive milestones, they were transferred to a casa de copii (children’s home) overseen by the Ministry of Education. If, however, the child exhibited physical stunting, cognitive delays, or neurological impairments—conditions often induced by the institutional environment itself—they were classified as “irrecuperable” (irecuperabili) and transferred to chronic psychiatric hospitals run by the Ministry of Labor, where conditions were notoriously fatal.

Within the leagăne, caregiving practices were systematically designed to minimize human attachment and maximize industrial efficiency. The physical architecture was characterized by vast, sterile rooms filled with rows of identical metal cribs with white-sheeted bars. Caregiver-to-child ratios were catastrophic, frequently ranging from 1:15 to 1:30 during daytime shifts and dropping to a single caregiver for up to 60 children during nighttime hours. Furthermore, institutional shifts rotated constantly, ensuring that an individual infant encountered dozens of different caregivers over a matter of months. Staff were explicitly instructed by institutional directors not to hold, rock, caress, or comfort crying infants, under the pseudoscientific justification that emotional affection would “spoil” the children, make them unmanageable, or transmit communicable diseases.

This organizational regime produced an environment of profound sensory, emotional, and cognitive starvation. Infants spent up to 20 to 22 hours a day lying supine in their cribs, staring at blank ceilings or uniform white walls, insulated from auditory variation, vocal exchange, and tactile comfort. Feeding was executed with propped bottles or through rigid, assembly-line feeding schedules where lukewarm porridge was shoveled into infants’ mouths without interpersonal gaze or communicative engagement. Diapers were changed mechanically in rapid succession without eye contact, verbalization, or smiles. In this vacuum of human reciprocity, normal developmental signaling broke down: infants rapidly ceased crying, having learned through behavioral extinction that distress signals elicited no maternal buffering, culminating in a state of quiet despair termed “functional institutionalism.”

1.3 Post-1989 Revelations and International Humanitarian Response

The execution of Nicolae and Elena Ceaușescu on Christmas Day 1989 suddenly exposed this vast institutional archipelago to the global public. Western news networks, led by landmark investigative broadcasts such as ABC’s 20/20, broadcast footage of thousands of emaciated, glassy-eyed toddlers rocking rhythmically in metal cages, smeared with feces, and exhibiting self-injurious behaviors such as head-banging and eye-gouging. The global audience reacted with horror, viewing what appeared to be an industrial-scale laboratory of human neglect. The revelations mobilized a massive, unregulated wave of Western humanitarian intervention, dominated by non-governmental organizations (NGOs), religious charities, and prospective adoptive parents from North America and Western Europe.

Between 1990 and the late 1990s, an estimated 30,000 Romanian children were adopted internationally. However, this early wave was plagued by systemic corruption, lack of oversight, and the emergence of a black-market infant trade where Romanian intermediaries brokered private, unregulated adoptions directly from impoverished biological parents and institutional directors. Prospective adoptive parents received virtually no preparation, screening, or post-adoption clinical support, and recipient governments implemented no unified tracking systems to record developmental outcomes, psychiatric morbidity, or placement dissolutions. Although early retrospective studies of adopted Romanian cohorts in the United Kingdom, Canada, and the United States—most notably the English and Romanian Adoptees (ERA) study led by Sir Michael Rutter—began documenting persistent patterns of quasi-autism, disinhibited social engagement, and cognitive deficits, these studies lacked pre-adoption baseline metrics and could not eliminate confounding genetic, prenatal, and selection biases.

By the late 1990s, the Romanian government, under severe diplomatic pressure from the European Union (which made modern child welfare reform an explicit prerequisite for accession), imposed moratoriums on international adoptions. Faced with tens of thousands of children still institutionalized and lacking an indigenous foster care system, Romanian pediatric leaders and governmental ministries recognized an acute need for definitive scientific data. They needed to ascertain whether domestic foster care could remediate the profound deficits observed in institutionalized infants, and whether the severe cognitive and neurobiological damage of early neglect was truly reversible. It was within this historic convergence of human crisis, political transition, and scientific urgency that the BEIP was born.

2. Origins and Conceptual Framework of the Bucharest Early Intervention Project (BEIP)

2.1 The Core Investigative Triad: Nelson, Fox, and Zeanah

The scientific architecture of the BEIP was conceived and executed by three prominent American researchers, each bringing a complementary, world-class methodological specialization to developmental psychopathology. The first investigator, Charles A. Nelson III, was a cognitive neuroscientist specializing in infant brain development, visual perception, and the neural substrates of memory and facial processing. Nelson brought sophisticated electrophysiological technologies, including event-related potentials (ERPs) and high-density electroencephalography (EEG), alongside later structural magnetic resonance imaging (MRI), allowing the project to examine the invisible neuroanatomical and biophysical manifestations of deprivation beneath overt behavioral phenotypes.

The second member of the triad, Nathan A. Fox, was a pioneer in developmental psychophysiology and affective neuroscience. Fox’s career had centered on the physiological underpinnings of infant temperament, social-emotional processing, self-regulation, and autonomic nervous system activity. His expertise in quantitative EEG (qEEG)—specifically the functional interpretation of baseline spectral power, frontal alpha asymmetry, and resting-state cortical oscillations—provided the BEIP with a biological assay of cortical maturation and neural connectivity. Fox’s laboratory also contributed paradigms for assessing the autonomic nervous system via respiratory sinus arrhythmia (RSA) and stress-hormone reactivity via the hypothalamic-pituitary-adrenal (HPA) axis.

The third investigator, Charles H. Zeanah Jr., was a child and adolescent psychiatrist and an internationally recognized authority on infant mental health, attachment theory, and relational psychopathology. Zeanah had spent decades diagnosing and classifying severe disorders of attachment in maltreated and neglected children. His role was to oversee the psychiatric diagnostic framework, ensure the fidelity of clinical observational batteries (such as the Strange Situation Procedure), conceptualize the nuances of Reactive Attachment Disorder (RAD) and Disinhibited Social Engagement Disorder (DSED), and direct the clinical design, supervision, and therapeutic philosophy of the project-created foster care network.

2.2 Core Hypotheses on Sensitive Periods and Neuroplasticity

The BEIP was explicitly designed to test the limits of human developmental neuroplasticity and adjudicate between competing theoretical models of infant brain development. Central to the investigators’ framework was the biological concept of sensitive periods—genetically demarcated temporal windows during which the structural assembly and synaptic pruning of specific neural circuits are uniquely dependent upon expected environmental inputs. Drawing on David Hubel and Torsten Wiesel’s Nobel Prize-winning work on ocular dominance columns in the visual cortex, the BEIP team conceptualized institutional neglect as an unprecedented natural experiment in “experience-expectant” developmental deprivation.

The BEIP investigated three primary hypotheses:

• The Reversibility Hypothesis: Psychosocial deprivation causes profound multi-system developmental delays, but placement into an enriched, sensitive family environment can rescue these trajectories, stimulating catch-up growth across cognitive, linguistic, and socio-emotional domains.
• The Sensitive Period Hypothesis: The capacity for neural and psychological remediation is biologically constrained by chronological age at intervention. The team hypothesized that there is a critical threshold—speculatively identified in the literature between 18 and 30 months—beyond which the human brain loses its plasticity to fully recover typical architecture, leading to persistent deficits regardless of subsequent environmental enrichment.
• The Neurobiological Signatures Hypothesis: Institutional rearing leaves distinct, measurable biomarkers across cortical electrophysiology (reduced alpha power, elevated low-frequency theta power), neuroanatomy (volumetric reductions in gray and white matter), stress-response systems (blunted HPA-axis reactivity), and cellular aging (accelerated telomere attrition).

2.3 Project Inception and Romanian Institutional Cooperation

The implementation of the BEIP was initiated in the spring of 1999 when the investigators were approached by Dr. Cristian Tabacaru of the Institute for Maternal and Child Care (Institutul pentru Ocrotirea Mamei și Copilului – IOMC) in Bucharest and representatives from the Romanian State Secretariat for the Protection of Children. Romanian officials found themselves caught between competing recommendations: Western NGOs were aggressively promoting non-scientific deinstitutionalization policies, while traditionalist medical directors in Romania argued that institutional care was sanitary, safe, and developmentally adequate for abandoned children. The state urgently required empirical, local evidence to guide child welfare reform.

Nelson, Fox, and Zeanah spent over a year establishing deep collaborative ties with Romanian pediatricians, academic faculties at the University of Bucharest, and municipal child welfare directors. The research team conducted an initial screening and mapping of all six municipal leagăne operating within the Bucharest metropolitan region (Districts 1 through 6). These facilities housed hundreds of infants and toddlers who had been abandoned at birth or shortly thereafter. The team worked alongside bilingual Romanian psychologists, social workers, and pediatricians who were recruited and trained to conduct assessments in the native language, ensuring that the trial was not an extractive colonial expedition, but a cooperative venture designed to construct enduring pediatric research and social work capacity within post-communist Romania.

The collaborative agreement established that the BEIP would construct, finance, and operate a high-quality foster care system—a modality of substitute family care that did not exist in the municipality of Bucharest at that time. Crucially, the research protocol was integrated into the administrative workflow of the Bucharest child protection commissions. The data gathered would be systematically reported back to Romanian governmental bodies to provide objective, evidence-based recommendations for national legislative overhauls concerning child abandonment, foster licensing, and institutional closures.

3. Methodological Architecture: The Randomized Controlled Trial in Child Welfare

3.1 Sample Stratification and Cohort Characteristics

The methodological cornerstone of the BEIP was its design as a randomized controlled trial (RCT)—a design that had never previously been implemented to evaluate institutionalization versus family placement due to historical, political, and ethical barriers. The initial screening phase involved assessing 187 children residing across the six infant institutions in Bucharest. To ensure that observed deficits were attributable directly to institutional psychosocial neglect rather than preexisting congenital or neurological pathologies, the investigators instituted rigorous exclusionary screening criteria.

Children were systematically excluded if they presented with:

• Identifiable genetic syndromes, chromosomal aberrations (e.g., Down syndrome), or major dysmorphic abnormalities;
• Clinically documented microcephaly present at birth (as opposed to acquired microcephaly resulting from institutional deprivation);
• Overt physical manifestations of Fetal Alcohol Spectrum Disorder (FASD);
• Known focal neurological trauma, intracranial hemorrhage, or active pediatric seizure disorders;
• Age less than 6 months or greater than 31 months at the initial baseline evaluation.

Following this medical screening, exactly 136 institutionalized children (68 males, 68 females) met the study criteria and were enrolled in the core experimental cohort. The mean age of the institutionalized sample at baseline was 22 months (ranging from 6.8 to 31.8 months), with infants having spent an average of over 90% of their biological lives within institutional walls. Concurrently, the investigators recruited an age- and sex-matched normative community comparison group known as the Never-Institutionalized Group (NIG), comprising 72 children born in the same Bucharest hospitals and living with their biological families in typical socio-demographic neighborhoods. The NIG cohort served as a baseline control across all developmental, neurobiological, and psychiatric assessments.

3.2 Randomization Protocols: Institutionalized Rearing vs. Foster Care

Following comprehensive baseline multidisciplinary testing of all 136 children within the institutions, the cohort was randomized into two equal groups of 68 children:

• Care-as-Usual Group (CAUG, n=68): Children who remained in their current institutional placement, receiving standard care provided by state-run leagăne.
• Foster Care Group (FCG, n=68): Children who were systematically removed from the institutions and placed into the professional foster care network established, trained, and supervised directly by the BEIP team.

Randomization was conducted using a rigorous statistical allocation protocol to prevent selection bias. After baseline testing was completed, a biostatistician unaffiliated with the clinical testing teams placed randomized group assignments into sealed, opaque, sequentially numbered envelopes. To ensure baseline developmental equivalence across the two arms, the randomization protocol was stratified by sex and institutional site, with minimal balancing adjustments made for age and baseline developmental quotient (DQ) scores on the Bayley Scales of Infant Development.

Subsequent statistical verification confirmed that the CAUG and FCG arms were indistinguishable at baseline across every measured parameter. There were no statistically significant differences between the groups in chronological age, birth weight, gestational age, head circumference, height, weight, duration of institutionalization, developmental quotient, or baseline electrophysiological power. Consequently, any subsequent longitudinal divergences between the CAUG and FCG cohorts could be causally attributed to the foster care intervention rather than preexisting child-level characteristics or institutional history.

3.3 Multi-Informant and Multi-Modal Assessment Batteries

The BEIP deployed an exceptionally comprehensive longitudinal assessment battery spanning multiple levels of biological, cognitive, and interpersonal organization. Assessments were scheduled at precise developmental milestones: baseline (mean age 22 months), 30 months, 42 months, 54 months, 8 years, 12 years, 16 years, and 21 years of age. Testing sessions took place within dedicated, child-friendly laboratory suites established by the BEIP at the IOMC in Bucharest, equipped with standardized video recording technology, sound-dampened psychophysiological testing suites, and clinical interview rooms.

The assessment battery integrated three primary modalities:

• Psychometric and Cognitive Testing: The Bayley Scales of Infant Development (BSID-II) were utilized during infancy and toddlerhood (yielding Mental Development Index [MDI] and Psychomotor Development Index [PDI]). At older ages, the battery transitioned to the Wechsler Preschool and Primary Scale of Intelligence (WPPSI-R), the Wechsler Intelligence Scale for Children (WISC-IV), and computerized neuropsychological testing via the Cambridge Neuropsychological Test Automated Battery (CANTAB).
• Neurophysiological and Neuroimaging Paradigms: Electrophysiological metrics included high-density resting-state quantitative EEG (qEEG) to evaluate spectral power distribution across delta, theta, alpha, and beta frequency bands. Sensory and cognitive processing were evaluated using Event-Related Potentials (ERPs) recorded during visual face processing (P100, N170, and the negative central component [Nc]). At ages 8–10 and 16, structural MRI and Diffusion Tensor Imaging (DTI) were deployed to capture volumetric and white-matter microstructural integrity.
• Clinical, Attachment, and Psychiatric Interviews: Dyadic socio-emotional functioning was evaluated using the Ainsworth Strange Situation Procedure at 42 months to classify organized versus disorganized attachment. Diagnostic psychiatric interviews—specifically the Preschool Age Psychiatric Assessment (PAPA) and Child and Adolescent Psychiatric Assessment (CAPA)—were administered to foster mothers, institutional caregivers, and biological parents, complemented by teacher-report questionnaires and direct behavioral observations.

4. Ethical Deliberations, Institutional Review, and the Equipoise Doctrine

4.1 Clinical Equipoise and the Justification for Randomization

The decision to utilize a randomized controlled trial design within an active human rights crisis generated intense, polarizing debate within bioethics, developmental psychology, and pediatric medicine. Critics, including several prominent Western bioethicists, argued that the harmful nature of institutionalization had already been definitively demonstrated by mid-twentieth-century pioneers such as John Bowlby, René Spitz, and Anna Freud. From this perspective, assigning vulnerable infants to remain in environments recognized as developmentally toxic violated the fundamental Hippocratic tenet of primum non nocere (first, do no harm) and fell short of genuine clinical equipoise.

The BEIP investigators defended the RCT design through a rigorously articulated doctrine of “local clinical equipoise.” They emphasized that while institutionalization was widely viewed as suboptimal, the scientific literature at that time suffered from severe methodological flaws, including retrospective designs, unmeasured pre-adoption confounding, and selection biases. More critically, the project existed within a specific geopolitical reality: the Romanian state did not possess a foster care system in Bucharest, and municipal authorities were actively questioning whether institutional care was truly harmful, with many pediatricians arguing that communal infant homes were medically superior to impoverished rural families. The investigators lacked the legal authority, logistical resources, and mandate to place all institutionalized children into foster care; they had the resources to construct a pilot network capable of serving only a fraction of the population.

Under these constraints, the investigators argued that randomization represented the only scientifically rigorous and morally justifiable method to allocate scarce therapeutic resources without researcher bias. If children were selected for foster care based on clinician preference, the healthiest, most charismatic, or most intellectually intact children would inevitably have been chosen, fatally compromising the scientific validity of the findings. Crucially, Nelson, Fox, and Zeanah contended that only the gold-standard evidence of a randomized controlled trial would carry sufficient empirical weight to dismantle the entrenched political and medical bureaucracy preserving the Romanian institutional system.

4.2 Institutional Review Board (IRB) Oversight and Dual Governance

To navigate these unprecedented bioethical challenges, the BEIP established a dual-governance institutional review apparatus that subjected the study to continuous, multi-institutional ethical scrutiny. Because the primary investigators held academic appointments at three major American universities, the study protocol was independently evaluated, approved, and annually re-reviewed by the Institutional Review Boards (IRBs) of:

• Harvard University / Boston Children’s Hospital;
• The University of Maryland, College Park;
• Tulane University Health Sciences Center.

Recognizing that Western IRB approval alone was insufficient and potentially vulnerable to accusations of academic imperialism, the team mandated that the project operate under direct Romanian domestic oversight. A local Ethical Advisory Commission was established in Bucharest, composed of Romanian pediatricians, academic philosophers, child psychologists, legal experts, and child welfare advocates. This body met regularly to inspect field operations, review testing procedures, and monitor the psychological and physical well-being of the enrolled children. The trial strictly adhered to the ethical provisions of the World Medical Association Declaration of Helsinki and the guidelines outlined in the United States Belmont Report, particularly concerning the protection of vulnerable populations incapable of providing informed consent.

Because the enrolled institutionalized infants had been legally surrendered or abandoned, parental consent was legally absent. Under Romanian law, municipal Child Protective Services Commissions served as the legal guardians of the institutionalized children. Formal written informed consent for participation was obtained directly from these governmental child welfare authorities, as well as from the institutional directors of each leagăn. When children were placed into foster homes, ongoing informed assent and parental consent were obtained from the foster parents. Biological parents who retained partial parental rights were consulted whenever legally traceable.

4.3 Non-Interference Mandates and Ongoing Ethical Dilemmas

A central ethical commitment embedded in the BEIP protocol was the policy of strict non-interference. The investigators recognized that it would be profoundly unethical to actively prevent an institutionalized child from accessing positive placement opportunities outside the trial. Consequently, the research team established an absolute rule: if a child assigned to the Care-as-Usual Group (CAUG) became eligible for domestic adoption, reunification with biological family members, or placement into government-sponsored foster care outside the project, the BEIP investigators did not intervene, object, or impede the process in any manner.

As the trial progressed over its 20-year span, this non-interference mandate resulted in significant “cross-over” between experimental arms. Substantial numbers of children originally randomized to CAUG were eventually adopted domestically, reunited with their biological parents, or placed into a burgeoning Romanian state foster care network that expanded throughout the mid-2000s. While these cross-overs introduced complex methodological challenges that required sophisticated “intent-to-treat” (ITT) and “as-treated” statistical modeling, it preserved the moral integrity of the project by ensuring that no child was ever held in institutional custody for the purpose of research observations.

Despite these safeguards, the BEIP remains a centerpiece of modern bioethical discourse. Scholars such as Joseph Millum, Alan Wertheimer, and Ezekiel Emanuel have analyzed the project as a paradigmatic case study in translational ethics, examining questions of post-trial responsibility: What obligations did the researchers owe to the foster parents and children once the initial NIH funding expired? The investigators successfully addressed this by transferring the ongoing financial and clinical management of the foster network to a local Romanian non-governmental organization (Fundatia SERA Romania), ensuring that no foster child was summarily returned to an institution when the primary grant ended. Nonetheless, the existential question of conducting randomized clinical trials on severely traumatized pediatric cohorts continues to serve as an indispensable reference point in the ethics of global human rights research.

5. Designing and Implementing High-Quality Foster Care Intervention

5.1 Creating an Unprecedented Foster Care Network

When the BEIP commenced in 2000, foster care was virtually non-existent within the municipality of Bucharest. The Romanian child protection system relied almost exclusively on large-scale institutional residential centers. To provide a valid, therapeutic alternative to institutionalization, the BEIP team was forced to build an entire foster care infrastructure from the ground up. This required establishing a dedicated clinical non-governmental organization, securing municipal licensing agreements, and engineering an unprecedented recruitment and training apparatus.

The recruitment process was launched across the greater Bucharest metropolitan region through print advertisements, local media broadcasts, and community outreach. Over 100 prospective families applied and were subjected to an exhaustive, multi-stage clinical screening protocol conducted by the BEIP’s Romanian clinical team. Screening criteria were specifically formulated to identify emotionally resilient, warm, and highly committed caregivers. The evaluation included comprehensive in-home visits, criminal background checks, assessments of marital stability and family cohesion, evaluations of prospective caregivers’ own attachment histories, and pediatric health clearances. Ultimately, 56 high-quality foster families were selected, licensed, and trained to care for the 68 children randomized to the foster care arm (several families took sibling pairs or sequential placements).

A distinctive feature of the BEIP foster care network was its structural and financial independence from the state system. Foster parents were hired as professional, salaried employees of the project. They received monthly stipends that exceeded typical Romanian civil-service wages, ensuring financial stability and preventing economic coercion. Crucially, the BEIP team supplied all necessary material resources directly to the homes, including modern cribs, age-appropriate toys, clothing, formula, specialized pediatric medical supplies, and developmental play materials. This eliminated the structural poverty that had historically driven Romanian families to institutionalize their biological children, isolating high-quality caregiving as the primary experimental variable.

5.2 Theoretical Framework of the Foster Care Intervention

The theoretical architecture of the foster care intervention was rooted in attachment theory, infant mental health principles, and the clinical science of developmental relational repair. Zeanah and his team drew heavily upon the Attachment and Biobehavioral Catch-up (ABC) framework developed by Mary Dozier, adapting its core tenets to the post-totalitarian Romanian context. The central clinical thesis was that overcoming profound institutional deprivation required far more than physical safety, hygiene, and nutrition; it necessitated the formation of an individualized, enduring, and psychologically committed attachment bond between the child and a dedicated primary caregiver.

The training curriculum emphasized four core clinical pillars:

• Sensitive and Contingent Responsiveness: Foster mothers were trained to decode subtle, ambiguous, or disorganized behavioral cues in toddlers who had learned to suppress distress. Caregivers were taught to respond immediately, predictably, and warmly to crying, vocalization, and proximity-seeking, establishing a reliable contingency between infant signaling and parental soothing.
• Active Psychological Investment and “Claiming”: Unlike traditional institutional paradigms that discouraged emotional bonding, the BEIP actively encouraged foster parents to “claim” the children as their own. Foster parents were coached to view the child not as a temporary medical ward, but as an authentic member of their family, cultivating unconditional commitment that could withstand developmental regressions.
• Behavioral Co-Regulation and Sensory Soothing: Institutionalized toddlers frequently presented with severe emotional dysregulation, autonomic hyper-reactivity, and bizarre motor stereotypies (e.g., violent rocking, head-banging). Foster parents were trained to avoid punitive discipline, utilizing physical proximity, gentle rocking, vocal soothing, and deep pressure to regulate the child’s overactivated nervous system.
• Linguistic and Cognitive Enrichment: Caregivers were trained to immerse children in rich, contingent linguistic environments. They were taught to narrate daily activities, maintain sustained mutual eye contact, engage in joint-attention play routines, and read illustrated books, thereby remediating the profound communicative deprivation of the leagăne.

5.3 Clinical Social Work Support and Ongoing Psychoeducation

Recognizing that placing traumatized, developmentally delayed infants into novice foster homes carried an exceptionally high risk of placement disruption and parental burnout, the BEIP established an intensive, continuous clinical support network. The project recruited and trained a cadre of master’s-level Romanian clinical social workers and psychologists who operated under the direct weekly supervision of Zeanah and clinical child psychologists from the United States. This clinical team functioned as an emotional scaffold for the foster parents throughout the lifecycle of the project.

Supportive mechanisms included:

• Bi-Weekly Home Visits: Social workers conducted home visits every two weeks during the initial placement phases, tapering to monthly visits as placements stabilized. These sessions involved observing caregiver-child interactions, troubleshooting behavioral challenges, and delivering real-time psychoeducation regarding institutional trauma, attachment insecurity, and developmental delays.
• 24/7 Crisis Response Infrastructure: Foster parents were provided direct, 24-hour cellular access to the BEIP clinical team to resolve acute behavioral crises, medical emergencies, or parental exhaustion, completely eliminating the impulse to return difficult children to state institutions.
• Reflective Functioning and Burnout Mitigation: Social workers actively targeted the foster parents’ “reflective functioning”—their capacity to understand the internal mental and emotional states of both themselves and the child. When a child exhibited withdrawal, aggression, or indiscriminate friendliness, social workers helped parents interpret these behaviors as manifestations of early trauma rather than personal rejection.
• Peer Support Networks: The project organized monthly parent support groups where BEIP foster mothers convened to share experiences, normalize common challenges, celebrate developmental gains, and build a cohesive community of practice.

The efficacy of this intensive clinical infrastructure was demonstrated by its placement stability. Over the first decade of the project, placement disruptions within the BEIP foster network were exceptionally rare—under 5%—a stability metric that stands in stark contrast to foster care systems in the United States and Western Europe, where placement breakdown rates frequently exceed 30% to 50%.

6. Baseline Deprivation Profiles: Growth, Motor, and Cognitive Deficits

6.1 Anthropometric and Physical Growth Suppression

The baseline physical profiles of the 136 institutionalized infants assessed by the BEIP team revealed catastrophic, multi-system somatic growth suppression. When initial anthropometric measurements were standardized against World Health Organization (WHO) and international pediatric growth charts, the institutionalized cohort exhibited profound physical stunting, pervasive bodily wasting, and acquired microcephaly. At a mean chronological age of 22 months, the average height and weight of institutionalized infants were hovering between 1.5 and 2.5 standard deviations below the normative mean, with over a third of the cohort meeting the clinical definition of failure to thrive.

Most alarmingly, head circumference—a reliable pediatric proxy for total intracranial volume and brain growth—was markedly depressed across the institutionalized cohort, with baseline scores averaging nearly two standard deviations below typical community controls. The investigators confirmed that this microcephaly was not congenital: hospital birth records demonstrated that the children had been born with normal head circumferences and birth weights. The suppression of cranial and somatic growth was directly proportional to the length of time spent in the institution: for every month an infant remained in the leagăn, their height, weight, and head circumference fell further behind international developmental trajectories.

This somatic collapse represented a classic manifestation of psychosocial dwarfism (or psychosocial short stature). This condition is characterized by profound endocrine disruption resulting directly from chronic emotional and sensory neglect. In the absence of warm, contingent maternal care, the child’s body exists in a state of chronic autonomic stress, altering the hypothalamic-pituitary axis. This disruption suppresses the normal pulsatile release of growth hormone (GH) from the anterior pituitary and downregulates hepatic production of Insulin-Like Growth Factor 1 (IGF-1). Despite receiving caloric intake that was theoretically adequate for survival, the infants’ bodies were biologically incapable of utilizing nutrients for somatic assembly and cellular proliferation.

6.2 Profound Cognitive and Motor Impairments

Baseline psychometric testing utilizing the Bayley Scales of Infant Development (BSID-II) demonstrated devastating cognitive and psychomotor deficits among the institutionalized children compared to the community-dwelling Never-Institutionalized Group (NIG). The BSID-II assesses two primary developmental domains: the Mental Development Index (MDI), capturing sensorimotor integration, memory, problem-solving, and early language; and the Psychomotor Development Index (PDI), assessing gross and fine motor coordination and control.

The baseline data revealed a stark cognitive gulf:

• Mental Development Index (MDI): While the community control cohort (NIG) scored within the normal range with a mean MDI of 100.4 ± 11.2, the institutionalized cohort exhibited a catastrophic mean MDI of 73.4 ± 12.8. Over 65% of the institutionalized infants scored at or below an MDI of 69, meeting the clinical threshold for significant developmental disability.
• Psychomotor Development Index (PDI): Similar suppression was observed in motor systems. The NIG control group demonstrated typical motor capacity with a mean PDI of 101.6 ± 12.7, whereas the institutionalized children registered a mean PDI of 73.2 ± 13.9.

Qualitative behavioral observations during these testing sessions documented an utter absence of symbolic, pretend, or exploratory play. Institutionalized toddlers presented with severe gross motor delays; many children aged 18 to 24 months were incapable of independent sitting, crawling, or pulling to a stand, having spent their lives confined to flat, narrow crib mattresses without physical space or environmental incentive to mobilize. Fine motor dexterity was similarly paralyzed, characterized by immature palmar grasps, absent pincer grips, and pronounced difficulties in sensorimotor coordination. Expressive language was virtually nonexistent across the institutionalized cohort: infants did not babble, engage in reciprocal vocal turn-taking, or utilize communicative gestures such as pointing or waving, reflecting an environment completely devoid of reciprocal linguistic input.

6.3 Atypical Sensory Processing and Self-Soothing Stereotypies

One of the most striking and distressing clinical features documented at baseline was the near-universal presence of complex, atypical motor stereotypies and profound sensory processing pathologies. Within the institutional wards, children had evolved bizarre, homeostatic auto-regulatory behaviors to compensate for the profound absence of external vestibular, kinesthetic, and sensory stimulation. Over 60% of the institutionalized cohort exhibited severe, repetitive stereotypic movements, including continuous rhythmic body rocking, rapid head-rolling from side to side in their cribs, and chronic hand-flapping.

In more severe presentations, these auto-stimulatory patterns deteriorated into violent, self-injurious behaviors:

• Rhythmic, repetitive head-banging against the hard wooden or metal bars of their cribs;
• Compulsive eye-poking, corneal gouging, and persistent hair-pulling;
• Severe biting of fingers, wrists, and arms, frequently producing open cutaneous lesions and localized infections.

Electrophysiological and observational assessments revealed profound anomalies in sensory modulation. Institutionalized infants lived in a state of sensory paradox: they were simultaneously hyper-reactive to sudden tactile or auditory transitions, yet profoundly hypo-reactive to basic sensory thresholds. For example, a sudden acoustic stimulus (such as a door closing or an unfamiliar voice) would trigger severe autonomic panic or catastrophic behavioral freezing. Conversely, infants demonstrated marked tactile defensiveness, shuddering or withdrawing violently when touched, brushed, or embraced by an examiner, having never experienced affective touch as a source of safety. Autonomic nervous system monitoring via baseline heart rate and vagal tone revealed blunted, disorganized physiological regulatory capacities, signaling a shattered internal homeostatic architecture.

7. Neurobiological Alterations: EEG, ERP, and Brain Structural Trajectories

7.1 Quantitative EEG Anomalies and ‘Hypoactivation’

Under the direction of Nathan Fox, the BEIP integrated high-density quantitative electroencephalography (qEEG) to evaluate the functional neurophysiological maturation of the cerebral cortex. The team collected resting-state EEG data across 12 scalp locations, analyzing absolute and relative power across distinct spectral bands: delta (1–4 Hz), theta (4–8 Hz), alpha (8–12 Hz), and beta (12–30 Hz). In typically developing children, the developmental trajectory is characterized by a progressive decrease in low-frequency power (delta and theta) and a concurrent increase in high-frequency power (alpha and beta), reflecting synaptogenesis, myelination, and the functional maturation of corticocortical networks.

The baseline and longitudinal EEG findings demonstrated a stark, profound neurobiological signature of institutional deprivation termed the neural hypoactivation model:

• Elevated Low-Frequency Power (Theta): Institutionalized children (CAUG) exhibited massively elevated power in the slow-wave theta band across all scalp regions compared to community controls (NIG).
• Suppressed High-Frequency Power (Alpha): Concurrently, institutionalized children displayed profound reductions in high-frequency alpha power. Because alpha rhythms reflect active, alert cortical processing and functional thalamocortical coordination, this suppression pointed to an under-aroused, neurophysiologically quiescent cerebrum.

This electrophysiological profile—dramatically elevated theta power combined with suppressed alpha power—mirrors the neural architecture observed in severe developmental learning disabilities, microcephaly, and traumatic brain injury. It suggested that institutional sensory deprivation induces a state of electrical “starvation,” severely delaying the normal developmental shift toward high-frequency oscillatory dynamics. However, the longitudinal tracking of this cohort revealed a remarkable, intervention-driven neuroplastic recovery: children placed into foster care prior to 24 months of age exhibited a significant longitudinal increase in high-frequency alpha power, eventually becoming indistinguishable from typically developing community controls by age 8. In stark contrast, children who remained institutionalized (CAUG) continued to exhibit persistent, blunted alpha activity well into adolescence.

7.2 Event-Related Potentials and Sensory Face Processing

Led by Charles Nelson, the BEIP utilized event-related potentials (ERPs) to examine the fine-grained temporal dynamics of neural information processing, specifically focusing on visual face processing and emotional recognition. Because faces represent the primary vehicle for early social communication and infant-caregiver attachment, the team designed experimental paradigms where children were presented with visual displays of unfamiliar female faces versus their primary caregiver’s face, displaying varied emotional expressions (happy, fearful, sad, and neutral).

The investigations focused on three key ERP components:

• The P100 Wave: An early, obligate sensory component reflecting primary visual cortex activation;
• The N170 Wave: A structurally sensitive component generated within the fusiform face area (FFA) and superior temporal sulcus, reflecting the structural encoding of human faces;
• The Negative Central (Nc) Component: A mid-latency frontal component reflecting endogenous attentional allocation and the salience of emotional stimuli.

The findings demonstrated severe, persistent disruptions in facial processing networks among institutionalized children. At baseline and early follow-ups, CAUG children exhibited significantly smaller, blunted P100 and N170 amplitudes and delayed latencies compared to community controls. Most critically, while typically developing children (NIG) exhibited robust neural differentiation on the Nc component—showing a significantly larger negative amplitude to their primary caregiver’s face compared to a stranger’s face—institutionalized children demonstrated a total failure of neural discrimination. Their brains responded with identical, blunted Nc waveforms regardless of whether they were viewing their primary institutional nurse or a complete stranger.

Remediation of these face-processing circuits was critically dependent upon the timing of foster placement. Children placed into foster care before 20 to 24 months of age exhibited significant neural catch-up: by age 42 months, their ERP waveforms began demonstrating distinct morphological differentiation between familiar caregivers and strangers, along with enhanced Nc amplitudes to emotional faces. Conversely, children placed into foster homes after this developmental cut-off demonstrated persistent deficits in facial encoding and emotional discrimination at age 8 and 12, providing empirical confirmation of a sensitive period for the functional organization of human social-visual circuitry.

7.3 Structural MRI Alterations: Gray and White Matter Reductions

As the BEIP cohort reached late childhood and adolescence (ages 8 to 10 and 16 years), the research team transported participants to advanced neuroimaging facilities to acquire high-resolution structural magnetic resonance imaging (sMRI) and diffusion tensor imaging (DTI). These scans yielded the first definitive three-dimensional quantitative reconstructions of brain anatomy in children subjected to randomized early psychosocial deprivation.

The volumetric findings were profound and somber:

• Cortical Gray Matter Reductions: Children with a history of institutional rearing exhibited dramatic, global reductions in total cortical gray matter volume compared to never-institutionalized controls. Gray matter loss was most pronounced in the prefrontal cortex, temporal lobes, and parietal association cortices. Strikingly, the foster care intervention had no significant remediating effect on cortical gray matter volume. Regardless of whether a child was placed into foster care at 10 months or remained in the institution, gray matter volume remained significantly diminished into adolescence, suggesting that early psychosocial deprivation inflicts a permanent constraint on cortical cellular assembly, dendritic arborization, and synaptogenesis.
• White Matter Volume Mitigation: In dramatic contrast to gray matter, total white matter volume exhibited remarkable neuroplastic recovery directly driven by the foster care intervention. While CAUG children exhibited severe, persistent reductions in white matter volume, children placed into high-quality foster care displayed significant catch-up growth in total white matter volume, rendering them statistically comparable to never-institutionalized controls by age 8.
• Microstructural White Matter Tractography (DTI): Diffusion tensor imaging revealed that institutionalization compromised the microstructural integrity and fractional anisotropy (FA) of major associative white-matter tracts. Most notably, the uncinate fasciculus (connecting the amygdala and limbic structures to the orbitofrontal cortex) and the superior longitudinal fasciculus (subserving fronto-parietal executive and linguistic networks) exhibited profound microstructural disorganization, accounting for the persistent executive dysregulation and affective instability documented in this cohort.

8. Cognitive and Intellectual Trajectories: The Critical Period Hypothesis

8.1 Longitudinal IQ Testing and the 24-Month Threshold

The long-term tracking of full-scale intelligence across the BEIP cohort represents one of the most cited contributions to modern developmental science. The research team evaluated intellectual performance utilizing age-appropriate, standardized intelligence scales—specifically the Wechsler Preschool and Primary Scale of Intelligence (WPPSI-R) at 42 and 54 months, and the Wechsler Intelligence Scale for Children (WISC-IV) at ages 8 and 16 years. These longitudinal batteries allowed the team to construct dynamic developmental trajectories across the CAUG, FCG, and NIG cohorts.

The data revealed an enduring, statistically robust intellectual divergence across the study arms:

• Never-Institutionalized Group (NIG): Displayed stable, normative intellectual performance across all developmental waves, maintaining an average Full-Scale IQ ranging between 100 and 109.
• Care-as-Usual Group (CAUG): Exhibited profound, chronic cognitive suppression. At age 42 months, the mean IQ of the institutionalized cohort was 73. At age 8, it remained flat at 73, and by age 16, the mean Full-Scale IQ stood at 73.3, with more than half of the chronically institutionalized children functioning in the clinical range of intellectual disability (IQ < 70).
• Foster Care Group (FCG): Demonstrated statistically significant, sustained intellectual recovery, achieving an average Full-Scale IQ of approximately 85 to 90 at age 8 and maintaining this cognitive elevation through age 16.

The most momentous discovery emerging from this cognitive data was the empirical verification of a definitive sensitive period for intellectual recovery bounded at 24 months of age. When the investigators stratified the Foster Care Group by the exact chronological age at which the child was removed from the institution, a striking step-function appeared: children placed into foster care before 24 months of age achieved mean IQ scores of 89 to 94 by age 8, functioning well within typical limits. In sharp contrast, children placed into foster care after 24 months of age exhibited mean IQ scores hovering around 75 to 79—scores that were not statistically distinguishable from the chronically institutionalized CAUG cohort. This 24-month threshold demonstrated that human intellectual potential is acutely sensitive to early experience, and that the window for remediating severe cognitive neglect narrows dramatically after the second year of life.

8.2 Executive Functioning and Working Memory Deficits

Beyond global psychometric intelligence, the BEIP administered rigorous neuropsychological batteries to evaluate the functional integrity of higher-order executive functioning networks. Utilizing computerized tasks from the Cambridge Neuropsychological Test Automated Battery (CANTAB) alongside behavioral paradigms, the researchers evaluated spatial working memory, cognitive flexibility, attentional set-shifting, and inhibitory motor control (via Go/No-Go and Flanker tasks) at ages 8, 12, and 16.

The findings demonstrated that executive functions are disproportionately vulnerable to early institutional rearing. Children with a history of institutionalization exhibited severe impairments in spatial working memory, committing significantly more errors on the CANTAB Spatial Working Memory (SWM) task and demonstrating an inability to utilize organized search strategies. On inhibitory control paradigms, institutionalized children displayed profound deficits, exhibiting significantly elevated commission errors (failing to inhibit a motor response on “No-Go” trials) and erratic response latency times, reflecting severe dysregulation of the prefrontal-striatal circuits that govern behavioral inhibition.

Critically, unlike global IQ, executive functioning domains demonstrated very limited remediation following foster care placement. Even children placed into high-quality foster homes prior to 24 months continued to display persistent, measurable deficits in working memory capacity and attentional set-shifting at age 16 compared to never-institutionalized controls. This suggested that the prolonged, complex developmental timeline of prefrontal cortex maturation—which continues through adolescence—renders these circuits vulnerable to early insults that cannot be completely overcome by subsequent environmental enrichment. Longitudinal mediation modeling confirmed that these early executive deficits directly mediated the persistent academic underachievement, educational disruptions, and occupational difficulties experienced by institutionalized children as they transitioned into young adulthood.

8.3 Language Acquisition and Communicative Competence

Language development in the BEIP cohort presented a fascinating contrast between extraordinary biological resilience and persistent communicative vulnerability. At baseline, institutionalized toddlers exhibited an almost complete absence of expressive language and profound receptive delays, having lived in an acoustic environment devoid of child-directed speech (“motherese”), contingent conversational turn-taking, or shared reading routines.

Following assignment to the Foster Care Group, children experienced dramatic, rapid catch-up growth in basic structural language mechanics. Testing at 30, 42, and 54 months via the Reynell Developmental Language Scales demonstrated that children placed into foster care—particularly those placed prior to 24 months—rapidly acquired vocabulary, mastered morphological markers, and assembled syntactic sentence structures at rates that closed the gap with typically developing peers by age 8. This rapid recovery demonstrated the biological robustness of “domain-specific” linguistic modules, which appear equipped with powerful innate learning mechanisms capable of accelerating once linguistic exposure is provided.

However, when the BEIP evaluated higher-order pragmatic language, socio-communicative competence, and discourse coherence at ages 12 and 16, enduring deficits were unmasked. Adolescents with a history of institutional rearing—including many who had been placed into foster care—struggled with narrative coherence, contextual pragmatic comprehension (e.g., interpreting irony, metaphor, and non-literal speech), and the reciprocal social use of language. While the structural syntax of language proved remarkably plastic and recoverable, the sociopragmatic application of language—which relies heavily on the integration of linguistic modules with prefrontal executive control and Theory of Mind networks—remained compromised by early psychosocial deprivation.

9. Attachment Disruptions and Socio-Emotional Maladaptation

9.1 The Strange Situation Procedure and Attachment Disorganization

The evaluation of infant-caregiver attachment was directed by Charles Zeanah, utilizing the gold-standard developmental paradigm: the Ainsworth Strange Situation Procedure (SSP). Conducted at 42 months of age, the SSP is a structured 21-minute laboratory simulation comprising eight sequential episodes of brief separations and reunions with the primary caregiver in the presence of an unfamiliar adult. Infant behaviors upon reunion are coded to classify attachment into four recognized typologies: Secure (Group B), Insecure-Avoidant (Group A), Insecure-Resistant/Ambivalent (Group C), and Disorganized/Disoriented (Group D).

The baseline and 42-month attachment profiles revealed a catastrophic collapse of organized attachment relationships within the institutions:

• Care-as-Usual Group (CAUG): Over 65% of institutionalized children exhibited Disorganized Attachment (Group D), characterized by contradictory, fearful, or disoriented behaviors upon the caregiver’s return (e.g., freezing in place, dropping to the floor, approaching while actively averting gaze). Most strikingly, another 13% exhibited behaviors that were completely unclassifiable under standard Ainsworth protocols, presenting a bizarre clinical picture of total relational vacancy and lack of behavioral strategy. Only a minuscule fraction (3%) exhibited signs of secure attachment to their favorite institutional nurse.
• Never-Institutionalized Group (NIG): Exhibited normative attachment distributions, with approximately 65% classified as Secure (Group B) and low rates of disorganization.
• Foster Care Group (FCG): The foster care intervention effected an extraordinary, rapid reorganization of attachment. By 42 months, nearly 50% of the children placed into foster care had formed fully Secure attachments (Group B) to their foster mothers.

The critical variable governing attachment remediation was the timing of the intervention. Children placed into foster homes before 18 to 24 months of age were four times more likely to develop secure, organized attachments than children placed after 24 months. For infants placed early, the warm, sensitive responsiveness of the BEIP foster mothers successfully repaired their disrupted internal working models, enabling them to utilize the caregiver as a safe haven for soothing and a secure base for environmental exploration. For children remaining in institutions past their second birthday, the biological window for organizing a first attachment bond appeared to close, leaving them at high risk for permanent relational pathology.

9.2 Indiscriminate Friendliness and Disinhibited Social Engagement Disorder (DSED)

One of the most persistent, confounding clinical phenotypes documented across the BEIP institutionalized cohort was the phenomenon of “indiscriminate friendliness,” codified in the DSM-5 as Disinhibited Social Engagement Disorder (DSED). In typically developing infants, the emergence of attachment at 7 to 9 months is accompanied by normative “stranger wariness” and separation anxiety—evolutionary adaptations designed to keep the mobile infant tethered to protective adults. Institutionalized children, however, exhibited a total absence of reticence toward unfamiliar adults.

To evaluate DSED quantitatively, the BEIP developed the Stranger-at-the-Door paradigm and observational rating scales assessing three core behavioral features:

• Absence of Checking Back: The child willingly wanders away into unfamiliar environments with complete strangers without looking back to locate their primary caregiver;
• Violation of Physical Boundaries: Approaching unfamiliar adults to climb into their laps, hug them, fondle their faces, or hold their hands within minutes of first meeting;
• Overly Familiar Verbal and Social Demeanor: Engaging in excessively intimate, non-normative verbal exchanges with strangers without age-appropriate hesitation.

The longitudinal data demonstrated that DSED operates under a radically different developmental and biological etiology than classic attachment insecurity. While placement into high-quality foster care rapidly remediated attachment insecurity and eradicated Reactive Attachment Disorder (RAD)—a condition where a child is emotionally withdrawn and fails to seek comfort—foster care had minimal impact on reducing DSED. Children placed into foster care before 24 months continued to display high levels of disinhibited, boundary-less social behaviors at ages 4, 8, 12, and 16 years. Neuroimaging analyses linked persistent DSED to microstructural alterations in amygdala volume and fronto-limbic functional connectivity rather than attachment failure per se, proving that social disinhibition is an enduring, neurobiologically mediated deficit in inhibitory social control rather than a manifestation of relationship quality.

9.3 Theory of Mind and Empathy Development

As the BEIP cohort traversed middle childhood and entered early adolescence, the researchers administered structured laboratory paradigms to measure the development of Theory of Mind (ToM)—the cognitive capacity to attribute mental states (beliefs, desires, intentions, emotions) to oneself and others, and to recognize that others possess mental states distinct from one’s own. Testing utilized classic first-order and second-order false-belief tasks (e.g., the Sally-Anne paradigm, unexpected contents tasks) at ages 54 months and 8 years, followed by the “Reading the Mind in the Eyes” task at ages 12 and 16.

The findings demonstrated severe, protracted delays in mentalizing capacities among institutionalized children:

• False-Belief Performance: At age 54 months and age 8, CAUG children exhibited severe impairments in false-belief understanding, consistently failing to predict a protagonist’s behavior based on a mistaken belief. They operated under a rigid, reality-bound cognitive framework, demonstrating an inability to decouple another person’s internal cognitive representation from objective physical reality.
• Foster Care Catch-Up: Children randomized to the Foster Care Group exhibited significant, measurable improvements in Theory of Mind performance compared to institutionalized peers. The daily experience of living within a communicative family environment—where mental states were regularly discussed, negotiated, and validated—stimulated cognitive perspective-taking mechanisms.

Despite these compensatory gains, long-term testing at age 16 revealed that subtle socio-emotional vulnerabilities persisted. On advanced tests of affective Theory of Mind, adolescents with a history of institutionalization exhibited reduced accuracy in decoding complex emotional states (such as skepticism, anticipation, or guilt) from subtle cues in the eye region of human faces. These neurocognitive mentalizing deficits manifested in the real world as significant difficulties in navigating adolescent peer dynamics, increased susceptibility to peer victimization, elevated rates of social isolation, and compromised capacities for resolving relational conflict.

10. Psychiatric Morbidity and Psychopathology Across Development

10.1 Externalizing Behaviors and ADHD Manifestations

Psychiatric diagnostic evaluations were administered longitudinally across the BEIP cohort using the Preschool Age Psychiatric Assessment (PAPA) at age 54 months and the Child and Adolescent Psychiatric Assessment (CAPA) at ages 8, 12, and 16 years. These comprehensive clinical interviews, conducted independently with caregivers and teachers, evaluated the entire spectrum of DSM-IV and DSM-5 psychiatric morbidity.

The most pervasive and structurally intractable psychiatric diagnosis that emerged across the institutionalized cohort was Attention-Deficit/Hyperactivity Disorder (ADHD). At age 8 and age 12, over 40% of children with a history of institutional care met full diagnostic criteria for ADHD—a prevalence rate more than four to five times higher than that observed in typical community populations. Intriguingly, the clinical presentation of institutionalization-induced ADHD differed markedly from classic, genetically mediated ADHD:

• Symptom Profile: The institutional cohort presented predominantly with the Inattentive Subtype, exhibiting catastrophic deficits in sustained attention, distractibility, and disorganization, with less prominent hyperactive-impulsive motor agitation.
• Resistance to Environmental Remediation: Placement into high-quality foster care exerted virtually no therapeutic effect on ADHD rates. Children randomized to the FCG exhibited rates of ADHD that were virtually identical to their peers who remained in the institutions (CAUG), regardless of the age at which they were placed into foster homes.

This persistent vulnerability, which researchers termed “institutional ADHD,” is mechanistically rooted in structural prefrontal cortical insults. As documented in the BEIP neuroimaging studies, early sensory and social neglect permanently blunts the development of cortical gray matter, diminishes white-matter microstructural integrity within the fronto-striatal loops, and alters the dopaminergic and noradrenergic circuitry that underpins executive attentional control. Because these structures undergo rapid, irreplaceable organization during the first two years of life, subsequent environmental enrichment cannot fully remediate the underlying neural architecture, leaving a permanent psychiatric vulnerability.

10.2 Internalizing Disorders: Anxiety and Depression in Adolescence

While externalizing disorders (ADHD, oppositional defiant disorder) dominated the early childhood psychiatric profile, the onset of puberty and adolescence unmasked a surge in internalizing psychopathology, including Major Depressive Disorder (MDD), Generalized Anxiety Disorder (GAD), and panic disorders. The longitudinal evaluations at ages 12 and 16 revealed a striking, sex-differentiated divergence in emotional vulnerability:

Cohort Group	Female Internalizing Rate (Age 12-16)	Male Internalizing Rate (Age 12-16)	Primary Diagnoses Observed
Care-as-Usual (CAUG)	45% – 52%	18% – 22%	Major Depression, Social Anxiety, GAD
Foster Care Group (FCG)	20% – 24%	15% – 19%	Mild Generalized Anxiety, Specific Phobias
Never-Institutionalized (NIG)	10% – 12%	8% – 10%	Transient Anxiety, Adjustment Reactions

As illustrated in the data above, institutionalized adolescent females represented an exceptionally vulnerable subgroup: over half of CAUG females manifested a clinical internalizing disorder by age 16. In dramatic contrast to the findings regarding ADHD, the foster care intervention served as an exceptionally potent, protective buffer against adolescent internalizing disorders. Females placed into high-quality foster care experienced a greater than 50% reduction in rates of depression and generalized anxiety compared to their chronically institutionalized female peers.

This protective buffering operates through both psychological and neurobiological mechanisms. Psychologically, the presence of an emotionally available, sensitive foster parent provides the adolescent with an external regulatory figure, buffering against the catastrophic feelings of worthlessness, hopelessness, and existential abandonment that characterize chronically institutionalized teenagers. Biologically, sensitive caregiving shields the developing adolescent brain from the neurotoxic cascades of chronic stress, downregulating amygdala hyper-reactivity and preserving the integrity of fronto-limbic regulatory circuits.

10.3 Stress Physiology: HPA Axis and Autonomic Dysregulation

To unravel the physiological mechanisms linking early institutional neglect to multi-system psychopathology, the BEIP conducted comprehensive assays of the body’s primary stress-response architectures: the Hypothalamic-Pituitary-Adrenal (HPA) axis and the Autonomic Nervous System (ANS). HPA axis functioning was evaluated via salivary cortisol collected across the circadian cycle to measure diurnal baseline rhythms, as well as in response to acute laboratory stress challenges (e.g., the Trier Social Stress Test for Children [TSST-C]).

The findings demonstrated severe, systemic physiological dysregulation:

• Blunted Diurnal Cortisol Rhythms: Chronically institutionalized children exhibited a profound flattening of the typical diurnal cortisol curve. Rather than displaying the typical steep surge upon waking (the Cortisol Awakening Response [CAR]) followed by a progressive decline toward a nocturnal nadir, institutionalized children presented with chronically flat, low morning cortisol levels, reflecting homeostatic down-regulation or “burnout” of the adrenal cortex following sustained, unmitigated toxic stress.
• Blunted Stress Reactivity: During the TSST-C challenge, typically developing children (NIG) mounted a robust, adaptive cortisol surge that gradually returned to baseline. Institutionalized children (CAUG) exhibited a completely blunted, unresponsive hormonal profile, mounting virtually no cortisol response to acute social threat.
• Autonomic Dysregulation (RSA): Assessment of high-frequency Heart Rate Variability (Respiratory Sinus Arrhythmia [RSA]), which indexes parasympathetic cardiac regulation via the vagus nerve, showed marked suppression in institutionalized children. Lower resting RSA and atypical RSA suppression during challenge signaled a compromised capacity for physiological self-soothing and social engagement.

At the deepest biological level, the BEIP investigated cellular and epigenetic biomarkers of early adversity. In a landmark 2013 study led by Stacy Drury and the BEIP team, blood and saliva samples collected at ages 8 and 16 were analyzed for telomere length—the protective nucleoprotein caps at the ends of eukaryotic chromosomes that serve as an objective index of biological aging and cumulative cellular stress. The results were striking: children with a history of institutional care exhibited significantly accelerated telomere attrition compared to never-institutionalized controls. Crucially, the extent of telomere shortening was directly correlated with the cumulative duration of institutional care: the more months a child spent in a leagăn, the shorter their telomeres were in adolescence. Concurrently, epigenetic investigations revealed extensive, altered DNA methylation patterns across critical candidate genes regulating neurodevelopment and glucocorticoid signaling, including the serotonin transporter gene (SLC6A4) and the glucocorticoid receptor gene (NR3C1), demonstrating that institutional deprivation alters the chemical software governing human gene expression.

11. Adolescent and Early Adulthood Outcomes: The Long-Term Follow-Ups

11.1 The Age 12, 16, and 21 Assessment Waves

Longitudinal retention across pediatric clinical cohorts represents one of the most formidable logistical challenges in developmental science. Over its two decades of operation, the BEIP achieved participant retention rates exceeding 75% to 80% across its major assessment waves at ages 12, 16, and 21. This continuity was maintained through the relentless dedication of the local Romanian clinical and research infrastructure, which tracked participants across geographical migrations, domestic adoptions, and the challenging transition out of secondary education.

The age 16 and 21 assessment waves marked a critical developmental transition: the evaluated cohort moved from pediatric dependence to biological maturity, independent living, vocational employment, or continued state institutional adult care. To evaluate real-world functional capacity, the team administered the Vineland Adaptive Behavior Scales alongside comprehensive evaluations of academic completion, occupational integration, independent living skills, and financial autonomy.

The long-term functional trajectories revealed marked divergences across the randomized arms:

• The Care-as-Usual Group: Individuals who remained in institutional care throughout childhood faced devastating developmental transitions. A substantial proportion aged out of pediatric institutions only to be immediately transferred to adult municipal psychiatric warehouses, homeless shelters, or rural subsistence living. Rates of independent employment were minimal, with the vast majority lacking basic life-management skills, financial literacy, or independent housing capability.
• The Foster Care Group: Youths placed into the BEIP foster care network demonstrated significantly higher rates of high school graduation, vocational training acquisition, and independent community living. Crucially, their adaptive behavior scores were substantially elevated compared to CAUG peers, driven by the enduring scaffolding of their foster families, who continued to provide financial, emotional, and practical housing support well past the legal age of majority.

11.2 Structural and Functional Brain Wiring in Late Adolescence

During the age 16 neuroimaging wave, the BEIP deployed advanced connectomics and resting-state functional magnetic resonance imaging (rs-fMRI) to interrogate the whole-brain functional architecture of the participants. These studies moved beyond isolated regional volumes to map the topological efficiency and functional connectivity of the human connectome—the complex network of structural and functional connections linking distributed brain regions.

The connectomic analyses revealed persistent topological abnormalities in adolescents with a history of institutionalization:

• Fronto-Limbic Dysconnectivity: CAUG adolescents exhibited pronounced functional hypoconnectivity between the ventromedial prefrontal cortex (vmPFC) and the amygdala. In typically developing adolescents, mature vmPFC-amygdala connectivity enables the top-down cognitive suppression of negative emotional arousal; its functional disruption in institutionalized youths accounted directly for their persistent emotional lability, explosive dysregulation, and susceptibility to affective disorders.
• Default Mode Network (DMN) Alterations: Significant functional alterations were documented within the Default Mode Network, which subserves self-referential thought, autobiographical memory, and social cognition. Institutionalized youths displayed fragmented intra-network cohesion within the posterior cingulate cortex (PCC) and medial prefrontal nodes.
• Compensatory Neural Plasticity: Remarkably, fMRI paradigms revealed that FCG adolescents who exhibited successful cognitive and behavioral rehabilitation recruited alternative, compensatory neural circuits during working memory and executive tasks. When performing cognitive challenges, remediated foster youths recruited broader bilateral fronto-parietal networks compared to the more focal, lateralized activation seen in never-institutionalized controls, demonstrating that behavioral recovery can occur via atypical, compensatory neurodevelopmental pathways.

11.3 Substance Use, Risky Behaviors, and Lawful Socialization

The transition through late adolescence and early adulthood brings an escalation in risk-taking behaviors, impulsivity, substance exploration, and potential conflict with legal authorities. At the age 16 and 21 follow-up waves, the BEIP systematically quantified delinquency, criminal offenses, substance use disorders (alcohol, cannabis, illicit narcotics), and risky sexual behaviors utilizing confidential self-report inventories, computerized risk-taking tasks (such as the Balloon Analogue Risk Task [BART]), and official municipal criminal justice records.

The data demonstrated that the high-quality foster care intervention exerted an enduring, protective effect against severe antisocial trajectories:

• Delinquency and Police Contacts: Chronically institutionalized youths (CAUG) exhibited significantly elevated rates of police contact, property crime, physical altercations, and municipal juvenile adjudications compared to community controls. In contrast, youths in the Foster Care Group exhibited rates of legal conflicts and antisocial infractions that were substantially lower than the CAUG cohort, tracking much closer to normative community baselines.
• Substance Dependence: While rates of adolescent alcohol experimentation were culturally widespread across Romania, severe substance dependence, binge drinking, and chemical abuse were significantly concentrated within the chronically institutionalized cohort. The absence of parental monitoring, coupled with early prefrontal executive dysfunction and peer socialization within institutional subcultures, drove CAUG youths toward maladaptive chemical self-medication.
• The Stabilizing Power of Family Ties: Longitudinal modeling revealed that the single most powerful factor protecting foster youth from antisocial and substance-dependent pathways was placement stability and emotional commitment. Foster adolescents who remained in stable, continuous foster homes throughout their childhood and adolescence demonstrated the lowest rates of behavioral delinquency, underscoring that human relational scaffolding serves as an indispensable external regulator of youthful biological impulsivity.

12. Global Policy Transformations and the Paradigm Shift in Child Welfare

12.1 Revolutionizing the Romanian Child Protection System

The primary translational objective of the BEIP was to generate definitive, empirical data to compel structural child welfare reform within Romania. As the empirical findings of the trial began emerging in the early and mid-2000s, Nelson, Fox, and Zeanah took the deliberate step of presenting their scientific data directly to the Romanian Prime Minister, the Ministry of Health, the National Authority for Child Protection, and the Romanian Parliament, long before theoretical academic publication in Western journals.

The impact of this translational advocacy was seismic:

• Passage of Romanian Law 272/2004: In 2004, armed with the BEIP’s empirical data demonstrating the profound cognitive, neurobiological, and attachment damage wrought by infant institutionalization—alongside the landmark 24-month sensitive period cutoff—the Romanian Parliament enacted Law 272/2004. This historic legislation formally made it illegal to place any child under the age of two into an institutional facility, unless the child presented with profound, life-threatening physical disabilities. In 2014, following subsequent BEIP follow-ups, this institutional ban was extended to children under the age of three, and subsequently raised to age seven.
• Nationwide Expansion of Foster Care: Prior to the BEIP, municipal foster care in Bucharest was nonexistent. The operational success, clinical training models, and social work protocols pioneered by the BEIP provided the direct administrative and clinical blueprint utilized by the Romanian government and European Union accession funds to train, license, and deploy tens of thousands of professional foster parents (asistenți maternali profesioniști) across all 41 administrative counties of Romania.
• Systematic Decommissioning of the Leagăne: The empirical evidence that institutions actively suppress brain growth dismantled the traditional medicalized defense of communal infant care. Between 2000 and 2020, the Romanian government systematically closed hundreds of large-scale infant and children’s institutions, reducing the population of institutionalized children from an estimated 170,000 in 1990 to fewer than 7,000 by the early 2020s, with institutional care permanently replaced by domestic foster networks, kinship placements, and family-preservation services.

12.2 Deinstitutionalization Movements and Global Human Rights Policy

The conceptual and empirical reverberations of the BEIP extended far beyond the borders of Romania, catalyzing a worldwide paradigm shift in international development, pediatric public health, and human rights law. Prior to the publication of the BEIP’s landmark papers—most notably their 2007 report in Science—large-scale orphanages were widely accepted across developing nations and Eastern Europe as an acceptable, economically efficient modality for housing abandoned, poor, or displaced children.

The BEIP permanently demolished this consensus, providing unassailable, gold-standard proof that institutionalization is an inherent form of severe developmental maltreatment. Major international organizations mobilized the BEIP’s neurobiological and developmental data to drive global deinstitutionalization agendas:

• The United Nations and UNICEF: The BEIP served as the primary scientific foundation for the drafting and global adoption of the United Nations Guidelines for the Alternative Care of Children in 2009. The UN formally recognized institutional care not merely as suboptimal, but as a direct violation of the UN Convention on the Rights of the Child (UNCRC), calling for the progressive global phase-out of all pediatric institutions.
• The World Health Organization (WHO): The WHO incorporated the BEIP’s electrophysiological, neuroimaging, and sensitive-period findings into its global mental health and early childhood development frameworks, declaring that institutional care inflicts active harm on brain architecture.
• The Shift in International Philanthropy: The dissemination of the BEIP’s findings triggered a massive reallocation of Western philanthropic capital. Major international donors, faith-based charities, and NGOs ceased financing the construction and maintenance of orphanages in sub-Saharan Africa, Southeast Asia, and Latin America. Instead, global child welfare funding shifted dramatically toward “orphan prevention”—investing resources directly into biological family strengthening, cash transfers, maternal mental health, and community-based foster care networks.

12.3 Enduring Scientific Legacy of Nelson, Fox, and Zeanah’s Work

Scientifically, the Bucharest Early Intervention Project stands alongside classic, monumental investigations such as the Framingham Heart Study, the Dunedin Multidisciplinary Health and Development Study, and the Perry Preschool Project as one of the most consequential clinical investigations in the history of behavioral science. Nelson, Fox, and Zeanah accomplished what developmental scientists had long theorized but had never causally proven: demonstrating the biological embodiment of early experience in human beings.

The enduring scientific legacies of the BEIP include:

• Causal Proof of Sensitive Periods in Humans: The project definitively established that human neurodevelopment is bounded by temporal windows of experience-expectant plasticity. By demonstrating that cognitive, electrophysiological, and attachment recoveries are highly contingent upon remediation before 24 months of age, the BEIP proved that early intervention is not merely politically desirable, but a strict neurobiological imperative.
• Methodological Blueprint for Translational Neuroscience: The project established a gold-standard methodological paradigm for integrating multi-level biological assays—spanning genetics, telomere biology, neuroendocrinology, electrophysiology, structural/functional neuroimaging, and psychiatric observation—within a single, longitudinal, ethically sound framework operating in a low-resource global setting.
• The Biological Primacy of Relational Care: Ultimately, the deepest scientific and philosophical truth established by the BEIP is that the human brain does not, and cannot, assemble its architecture in social isolation. The human infant is fundamentally an “experience-expectant” organism whose developing cortex requires continuous, contingent, and emotionally committed human interactions—conversations, reciprocal gazes, comforting embraces, and sensitive co-regulation—to fuel synaptic proliferation, guide myelination, and stabilize neural circuitry. In documenting the devastating biological consequences of withholding that care, Charles Nelson, Nathan Fox, and Charles Zeanah provided irrefutable proof that love, relationship, and human connection are biological requirements for the human species.

---

References

• Almas, A. N., Degnan, K. A., Nelson, C. A., Zeanah, C. H., & Fox, N. A. (2012). IQ at age 8 in children institutionalized in infancy: The role of caregiver responsiveness. Development and Psychopathology, 24(4), 1269–1279. https://doi.org/10.1017/S095457941200070X
• Bick, J., Zhu, T., Stamoulis, C., Fox, N. A., Zeanah, C., & Nelson, C. A. (2015). Effect of early institutionalization and foster care on long-term white matter development: A randomized clinical trial. JAMA Pediatrics, 169(3), 211–219. https://doi.org/10.1001/jamapediatrics.2014.3212
• Drury, S. S., Theall, K., Gleason, M. M., Smyke, A. T., De Vivo, I., Wong, J. Y., Fox, N. A., Zeanah, C. H., & Nelson, C. A. (2012). Telomere length and early severe social deprivation: Linking early adversity to cellular aging. Molecular Psychiatry, 17(7), 719–727. https://doi.org/10.1038/mp.2011.53
• Fox, N. A., Nelson, C. A., & Zeanah, C. H. (2017). The Bucharest Early Intervention Project: A review of methods and findings. Social Science & Medicine, 183, 148–156. https://doi.org/10.1016/j.socscimed.2017.04.053
• Ghera, M. M., Marshall, P. J., Fox, N. A., Zeanah, C. H., Smyke, A. T., & Nelson, C. A. (2009). The effects of early institutionalization on the development of emotion processing: An ERP study. Development and Psychopathology, 21(3), 769–784. https://doi.org/10.1017/S0954579409000424
• Marshall, P. J., Fox, N. A., & BEIP Core Group. (2004). A comparison of the electroencephalogram between institutionalized and community children in Romania. Journal of Cognitive Neuroscience, 16(8), 1327–1338. https://doi.org/10.1162/0898929042304723
• McLaughlin, K. A., Fox, N. A., Zeanah, C. H., Sheridan, M. A., Marshall, P., & Nelson, C. A. (2010). Delayed maturation in event-related potentials to emotional faces in institutionalized children: The role of early social deprivation. Journal of Child Psychology and Psychiatry, 51(3), 329–337. https://doi.org/10.1111/j.1469-7610.2009.02157.x
• McLaughlin, K. A., Sheridan, M. A., Winter, W., Fox, N. A., Zeanah, C. H., & Nelson, C. A. (2014). Widespread reductions in cortical thickness following severe early-life deprivation: A neurodevelopmental pathway to attention-deficit/hyperactivity disorder. Biological Psychiatry, 76(8), 629–638. https://doi.org/10.1016/j.biopsych.2013.08.016
• Nelson, C. A., Fox, N. A., & Zeanah, C. H. (2014). Romania’s Abandoned Children: Deprivation, Brain Development, and the Search for Improvement. Harvard University Press. https://www.hup.harvard.edu/books/9780674724709
• Nelson, C. A., Zeanah, C. H., Fox, N. A., Marshall, P. J., Smyke, A. T., & Guthrie, D. (2007). Cognitive recovery in socially deprived young children: The Bucharest Early Intervention Project. Science, 318(5858), 1937–1940. https://doi.org/10.1126/science.1143921
• Sheridan, M. A., Fox, N. A., Zeanah, C. H., McLaughlin, K. A., & Nelson, C. A. (2012). Variation in neural development as a result of exposure to extreme institutionalization in infancy. Proceedings of the National Academy of Sciences, 109(32), 12927–12932. https://doi.org/10.1073/pnas.1200041109
• Smyke, A. T., Dumitrescu, A., & Zeanah, C. H. (2002). Attachment disturbances in young children. I: The continuum of caretaking casualty. Journal of the American Academy of Child & Adolescent Psychiatry, 41(8), 972–982. https://doi.org/10.1097/00004583-200208000-00016
• Smyke, A. T., Zeanah, C. H., Fox, N. A., Nelson, C. A., & Guthrie, D. (2010). Placement in foster care enhances quality of attachment among young institutionalized children. Child Development, 81(1), 212–223. https://doi.org/10.1111/j.1467-8624.2009.01391.x
• Vanderwert, R. E., Marshall, P. J., Nelson, C. A., Zeanah, C. H., & Fox, N. A. (2010). Timing of intervention affects brain electrical activity in children exposed to severe psychosocial neglect. PLOS ONE, 5(7), e11415. https://doi.org/10.1371/journal.pone.0011415
• Wade, M., Fox, N. A., Zeanah, C. H., & Nelson, C. A. (2018). Effect of early psychosocial neglect on academic achievement in late adolescence: A randomized clinical trial of foster care. Development and Psychopathology, 30(5), 1845–1856. https://doi.org/10.1017/S0954579418000854
• Zeanah, C. H., Egger, H. L., Smyke, A. T., Nelson, C. A., & Fox, N. A. (2009). Institutional rearing and psychiatric disorders in Romanian preschool children. American Journal of Psychiatry, 166(7), 777–785. https://doi.org/10.1176/appi.ajp.2009.08091438
• Zeanah, C. H., Nelson, C. A., Fox, N. A., Smyke, A. T., Marshall, P., Parker, S. W., & Koga, S. (2003). Designing research on institutionalized children: The Bucharest Early Intervention Project. Development and Psychopathology, 15(4), 885–907. https://doi.org/10.1017/s0954579403000452