In most but not all studies (see Walker & Emory, 1985, for review), during infancy neurological signs or neuromotor dysfunctions are found more frequently in children at risk for schizophrenia than in controls. In these studies neuromotor anomalies were assessed by observation during a pediatric neurological examination or by performance on standardized tests of infant development (e.g., the Bayley). Neurological signs and neuromotor dysfunctions are not specific to infants at risk for schizophrenia and are not rare events in the general pediatric population. Neurological abnormalities in neonates typically tend to improve. In contrast, it appears that these abnormalities in children at risk for schizophrenia persist, and may worsen over time. Infants with neurological or neuromotor abnormalities are the high-risk infants most likely to develop schizophrenic disorders in adolescence and early adulthood (Fish, 1987; Marcus et al., 1987; Parnas, 1982). Neurologic dysregulation in infancy predicts the development of schizophrenia spectrum disorders (Fish, 1984). Impaired performance on tasks with extensive motor demands during middle childhood also predicts the presence of schizophrenia spectrum disorders during adolescence (Hans et al., 1999).

Disturbances in early social development are found more frequently in children at risk for schizophrenia than in controls. Depending on the study, these disturbances are manifested in difficult temperaments, apathy or withdrawal, being inhibited, less spontaneous, and imitative, reduced social contact with mothers, and absence of fear of strangers. The absence of fear of strangers during infancy could be an indication that the child does not differentiate between familiar adults to whom the child is attached (for example, the parents) and others. This absence of the fear of strangers may reflect inadequately developed attachment. These disturbances are not specific to children at risk for schizophrenia. They are also associated with broad risk factors such as socioeconomic status, general maternal distress, early trauma or neglect, and poor quality of parenting. Although there are scant data on how well these disturbances in infant social development predict the development of schizophrenia, many of these findings are related to the development of social competencies (Watt & Saiz, 1991).

During early childhood (2 to 4 years of age), children at risk for schizophrenia are more likely than controls to show poor fine and gross motor coordination and low reactivity. Although poor fine and gross motor coordination was found in a sample of children that was different from the samples of infants at risk for schizophrenia who showed a variety of neurological signs and neuromotor dysfunction, these data suggest that the dysfunctions observed in infancy are persistent.

In early childhood there is an increased occurrence of internalizing symptoms (depression and anxiety), angry and hostile dispositions, and schizoid behavior (emotionally flat, socially withdrawn, passive and distractible) in children at risk for schizophrenia. Again, these characteristics are not specific to children at risk for schizophrenia and there is no evidence that these characteristics are strongly predictive of the later development of schizophrenia.

Neuromotor impairments, including gross motor skills (Marcus et al., 1993), are found more frequently in children at risk for schizophrenia than in controls during middle childhood and early adolescence (4 to 14 years of age). One of the most robust cross-sectional findings during middle childhood and early adolescence is the presence of neurocognitive impairments, especially on measures with high attention demands.

A subgroup of children at risk for schizophrenia showed impairments on some of the same tasks for which patients with schizophrenia show impairments (Asarnow, 1988). The neurocognitive tasks for which children at risk for schizophrenia show impairments include measures of sustained attention (various continuous performance tests) and secondary memory (e.g., memory for stories). For example, children at risk for schizophrenia, as well as acutely disturbed and partially remitted schizophrenia patients, performed poorly on a partial-report-span-of-apprehension task (Asarnow, 1983) in the high attention/processing demand condition. The span of apprehension measures the rate of early visual information processing (Fig. 7.6).

There are some data on the predictive validity of the neurocognitive impairments identified during middle childhood and early adolescence. In the New York High-Risk project, the presence of impairments on a number of attentional tasks (an Attentional Deviance Index) given in middle childhood predicted 58% of the subjects who developed schizophrenia-related psychoses by mid-adulthood (Erlenmeyer-Kimling et al., 2000). Attentional impairments in mid-dle childhood were also associated with anhedonia (Freedman, Rock, Roberts, Cornblatt, & Erlenmeyer-Kimling, 1998) in adolescents prior to the onset of schizophrenia and social deficits during early adulthood (Cornblatt, Lenzenweger, Dworkin, & Erlenmeyer-Kimling, 1992; Freedman et al., 1998). Neuromotor dysfunction during childhood (assessed by the Lincoln-Oseretsky Motor Development Scale) identified 75% of the high-risk children who developed schizophrenia-related psychoses during adulthood (Erlenmeyer-Kimling et al., 2000). A verbal short-term memory factor that included a childhood digit span task and a complex attention task predicted 83% of the New York high-risk children who developed schizophrenia-related psychoses during adulthood and showed high specificity to those psychoses (Erlenmeyer-Kimling et al., 2000). If replicated, these findings would suggest that the combination of genetic risk (being the child of a parent who has schizophrenia) and neurocognitive impairments during middle childhood might identify individuals with a greatly increased risk for developing schizophrenia. The sensitivity (correctly predicting the onset of schizophrenia-related psychoses) was higher for the verbal memory (83%) and motor skills (75%) factors than for the attentional factor (58%). Conversely, the false-positive rate (incorrectly predicting that a child

would develop schizophrenia) was lower for the Attentional Deviance Index (18%) than for the memory factor (28%) and motor factor (27%).

The short-term follow-up in the Jerusalem High-Risk study provides an important link between the attentional impairments frequently observed in children at risk for schizophrenia during adolescence and the motor impairments found during infancy and early childhood. The children who showed impaired neuromotor performance during childhood were the subjects most likely to show impairments on a variety of measures of attention and information processing during early adolescence (Hans et al., 1999).

During middle childhood, children at risk for schizophrenia receive an increased frequency of a variety of psychiatric diagnoses, including an attention deficit disorder (ADD)-like syndrome. Poor affective control, including emotional instability and impulsivity, as well as aggression and disruptive behaviors are found more frequently in children at risk for schizophrenia than in controls. Early precursors of thought disorder may be reflected in the presence of cognitive slippage. Poor peer relations are one of the more frequently found behavioral characteristics during middle childhood and early adolescence. None of these symptoms is specific to children at risk for schizophrenia. For example, poor affective control is found in children who subsequently develop an affective disorder.

Table 7.1 shows a consistency across studies of children at risk for schizophrenia, birth cohort studies, and retrospective studies in the presence of motor and language problems during infancy. This consistency is particularly impressive given the considerable variation across studies in the ways in which motor functioning and language were assessed.

During early childhood (2 to 4 years of age), neuromotor problems are observed in all three types of studies. In birth cohort studies and retrospective studies, impaired language is noted. In high-risk studies children at risk for schizophrenia are noted as being depressed, anxious, angry, and schizoid, whereas in birth cohort studies they are noted as preferring solitary play.

During middle childhood (4 to 14 years of age) there is a persistence of neurologic impairments reflected in poor motor functioning in high-risk, birth cohort, and follow-back studies. High-risk studies, unlike birth cohort studies and retrospective studies, included laboratory measures of attention information processing. On these tasks, children at risk for schizophrenia showed attentional impairment under conditions of high processing demands. This may be related to the poor academic achievement observed in birth cohort studies and retrospective studies during middle childhood as well as the frequent diagnosis of ADHD. In contrast to the persistence of neuromotor problems, language problems tend to diminish over time, so by middle childhood they are rarely noted across the three classes of studies. In adolescents who develop a schizophrenic disorder, language functions are relatively preserved compared to visual-spatial and motor functioning (Asarnow, Tanguay, Bott, & Freeman, 1987).

The results of this brief review suggest a developmental pathway from precursors first identified in infancy to the development of schizophrenia-related psychoses in late adolescence and early adulthood. Neurologic signs or neuromotor dysfunctions are present in infancy and persist through early and middle childhood and early adolescence. Neuromotor dysfunction in early childhood predicts the presence of attentional impairments under high processing demands during early adolescence. Neuromotor dysfunctions and attentional impairments during adolescence predict the development of schizophrenia-related psychoses. Because the characterization of key points in this developmental sequence is based on only one or two studies, clearly this model needs to be tested in future research.

The developmental pathway sketched here has potentially interesting implications for our understanding of the neurobiology of schizophrenia. What brain systems are involved in the control of simple motor functions and attention? The developmental link between early neuromotor dysfunction and later attentional impairments may implicate cortical-striatal pathways that support both motor functions and attentional control mechanisms. Striatal dysfunction results in impaired sequential motor performance and chunking of action sequences.