DOES THIS DATA APPLY TO INTERNATIONAL ADOPTEES?

Certainly, children are referred for adoption who were born with a head circumference more than two standard deviations below the mean and whose head remains in the microcephalic range. In these situations, the data above predict a higher risk of mental retardation and a reduction in IQ scores in direct proportion to the degree of microcephaly. However, this situation is not commonly encountered when reviewing medical records of institutionalized children. A personal series of 427 children referred for adoption from Russia showed that of those whose head circumference was available at birth and at the time of referral (n = 199), only 12% were in the microcephalic range at birth, and only 17% of these microcephalic children (2% of the total) remained so during their course of institutionalization.

A more commonly encountered situation is birth OFC within the normal range but extremely poor head growth during the first year of life. In institutionalized children from Romania, head circumference decreased in direct relationship to the length of orphanage confinement during early infancy (r = -0.54, p < 0.005 (21) (figure below, left). This early effect on head growth persisted into early childhood. Only 7% of infants < 10 months of age (n = 27) had head circumferences >= 2 SD below the mean vs. 41% of those >= 10 months of age (n = 37) (p < 0.005 by X2) (figure below, right).

The data of Rutter et al. (22) confirm the profound effect of institutionalization on head growth. Average head circumference on entry to the United Kingdom in his cohort of Romanian adoptees was in the microcephalic range for infants under (-2.1 SD) and over (-2.3 SD) six months of age.

Head Growth Pattern in Institutionalized Children

Segregating a lower risk group of children (n = 154) referred from Russia (normal head circumference at birth and no obvious indication of prenatal alcohol exposure), average head circumference decreased progressively during the first months and remained low throughout the first two years of institutionalization (figure below).

The percentage of children in this group whose head was within the microcephalic range (> 2 standard deviations below the mean) progressively increased during the first two years. Of those referred at 0-3 months of age, only 4% were microcephalic, in contrast to almost one-third of children referred at 12-24 months of age (figure below).

Among the studies cited above, the findings of Nelson (17) are most applicable, and argue that those children who are microcephalic at one year of age are at higher risk for mental retardation and lower IQ. However, perhaps most relevant to the situation we see in institutionalized children is the study by Avery et al. (23) of 100 children with severe illness in the first year of life. Twenty-eight percent of these children were microcephalic at one year of age-a figure quite comparable with the data shown above. Overall, 50% of these microcephalic children were developing normally or were only slightly delayed. However, within this microcephalic group, 12 had diagnoses implying possible injury to the central nervous system (e.g., birth trauma, seizures, multiple anomalies, microcephaly at birth, and bacterial and viral central nervous system infections). The incidence of moderate to severe mental retardation was significantly higher in this group of children compared to children whose primary diagnoses were usually not associated with mental retardation (75% vs. 31%). Three patterns of head growth were identified. Children whose head circumference followed closely at or just below two standard deviations below the mean were generally normal or mildly delayed. Children whose head circumferences were consistently less than two standard deviations below the mean were all moderately to severely retarded. Children whose head growth most closely approximated the growth patterns seen in institutionalized children were normal to mildly retarded in 33% of cases and moderately to severely retarded in 66% of cases.

Head Growth after Adoption

Catch-up head growth was documented in 85% of Eastern European orphans (n = 34) after arrival. Mean head circumference increased an average of 0.67 ± 0.82 SD from arrival (-1.07 ± 0.9 SD, mean age 13.2 ± 5.2 months, range 5.5-32 months) to follow-up (-0.40 ± 1 SD, mean age 26 ± 7 months, range 5.5-32 months) (p < 0.01 paired t test) (Aronson & Johnson, unpublished data).

The length of institutionalization appears to have a very strong effect on eventual head size. Benoit et al. (24) found that 13% of children institutionalized more than six months had a head circumference < 5th percentile an average of 12 months after arrival, while all children adopted at six months or less were within the normal range. Rutter et al. (22) found significant differences in mean head circumference at four years of age in children adopted at six months of age or more (-1.5 ± 1.0), children adopted prior to six months (-1.1 ± 1.0) and their control group of children adopted to the United Kingdom (-0.5 ± 0.8).

Effect of Early Brain Insults on Head Circumference

The pattern of brain growth seen in institutionalized children is consistent with the concept of an early brain insult; as such, the data on outcome of other conditions that cause early brain insults may be instructive. Poor brain growth in infancy is seen in a number of situations, including acquired intrauterine infections such as rubella (20) and radiation exposure (25). While the cause of poor brain growth within institutionalized settings is unknown, nutritional impairment leading to intrauterine or postnatal brain growth failure perhaps parallels the plight of institutionalized children most closely. An extremely large body of information is available on the effects of early childhood malnutrition including several excellent reviews of animal (26, 27) and human data (28). I particularly recommend the short summary on the effects of malnutrition on young children by Galler and Ross (29).

Brain growth is clearly affected in severe protein-energy malnutrition (2). In studies conducted in young Jamaican children between 6 and 24 months, head circumference averaged only 91% of expected at the time the child was admitted to the hospital for treatment (30). Head growth failure persisted during follow up despite improved nutrition. While weight for height after recovery compared favorably with the control group within one month, 36 months later mean head circumference was still only 94% of expected (approximately 1.8 SD below the mean). Stoch and Smythe, studying 21 children in South Africa, documented severe, persistent impairment in head growth secondary to early malnutrition (31). Initial OFCs in the malnourished group were > 2 SD below the mean. After 15 years of adequate nutrition, the mean head circumference of this group continued to be more than two standard deviations below the mean (32).

During the acute period of malnutrition, children are apathetic, demonstrate delays in all developmental scales (especially language), and demonstrate abnormal crying patterns and altered mother-infant interactions. During recovery from malnutrition, children have reversal of apathy and improved motor and exploratory skills, but continue to have delays in language and mental development and a reduced developmental quotient. Long-term sequelae through adolescence include decreased IQ scores, delayed cognitive development, impaired sensory integration, impaired school performance, a fourfold increase in attention deficit disorder compared to the control group (60% vs. 15%) and low self-esteem. While head growth failure in institutionalized children may not be due exclusively to malnutrition, these data on the short- and long-term cognitive and behavioral effects of early malnutrition are startlingly similar to behavioral and cognitive findings in post-institutionalized adoptees (29).

The Affect of Home Placement on Children with Early Brain Insult

The above data imply that children with early brain insult are irreversibly affected and may be profoundly impaired on follow up. Indeed, the prognosis for children who continue to experience ongoing deprivation is not good (29,33,34). However, the positive effects of an enriched environment have been repeatedly demonstrated in children experiencing adverse conditions during early life. In the case of malnourished children, those who participated in a home-visiting program of psychosocial stimulation showed a marked advantage over the nonintervention group that persisted for at least six years (33,34). Children placed in single foster homes after an episode of early malnutrition had significantly higher IQ scores than similarly malnourished children subjected to multiple foster placements (35). Improvement in outcome within the enriched environments of adoptive homes or single placement foster homes has been demonstrated in malnourished children from Vietnam (36), Korea (37,38), Chile (39) and Peru (40), in intrauterine cocaine-exposed children in Canada (41), and in children adopted into higher socioeconomic groups in the United Kingdom (42).

Adoption of normal children prior to one year of life also appears to confer an advantage in terms of IQ at four years, though not at seven years of age (43). Scarr and Weinberg (44) showed that IQ scores in young adopted children (mean age of 7 years) were similar to their unrelated siblings; however, by adolescence, they were similar to their biological parents and siblings. Other studies utilizing adopted children have clearly demonstrated that under normal circumstances, while early experience may confer a temporary advantage, ultimately IQ is strongly determined by genetic factors (45).

In summary, there is no better review of the effect of a child's environment on cognitive development than that authored by Michael Rutter (46). He concluded that:

• Environmental effects on IQ are relatively modest within the normal range of environments, but that the effects in markedly disadvantageous circumstances are very substantial.
• Cognitive development is influenced by direct effects on cognition and by indirect effects through alterations in self-concept, aspirations, attitudes to learning and styles of interaction with other people.

Is There a Correlation Between OFC and Outcome in International Adoptees?

The study of Romanian children by Rutter et al. (22) is the only investigation to date that attempted to correlate outcome with head circumference. The authors demonstrated a statistically significant correlation between head circumference and the Denver Quotient at the time of entry into the adoptive home. However, they failed to demonstrate a clear relationship between head circumference at arrival or at 4 years with cognitive outcome at four years of age (Denver Quotient or McCarthy scores). McCarthy scores at four years of age did not differ between children who were normocephalic and microcephalic at entry into their adoptive homes [104.0 ± 16.7 (SD) vs. 99.6 ± 20.3, respectively]. This finding provides additional support for the benefits of the enriched environment of an adoptive home for at-risk children.

Is Bigger Better?

Several of the studies referenced above examined the outcome of children who had large heads. Nelson and Deutschberger (17) documented that the 1% of children who had the largest head circumferences at one year of age [> 49 cm in girls (+2.5 SD) and > 50 cm in boys (+2.1 SD)] had the highest mean IQ at four years of the groups examined, and a higher proportion of these children had IQs of 120 or greater. Using a cohort of children enrolled in the Collaborative, Fisch et al. expanded the observations of Nelson et al. by confirming that children with superior intelligence (IQ >= 120) at seven years of age had significantly higher mean head circumferences at one, four and seven years of age than children with average and low intelligence (47). In the St. Louis school study, 50% of children with the largest heads (>= 55 cm) had IQs of 120 or above and none were below 90 (19).

Large heads are also associated with hydrocephalus and rare neurologic conditions, grouped under the descriptive term megalencephaly, which are associated with poor neurologic outcome. Perhaps this is why Nelson and Deutschberger (17) also found a higher number of children within the lowest IQ groups in their group with the largest head circumference and Desch et al. (14) found minor differences in IQ and mathematical achievement tests.

Does This Data Apply to International Adoptees? | Summary