Neuromotor Markers of Risk for Schizophrenia

Abstract

Objective

To evaluate the possibility of using neuromotor deviation (NMD) as an individual-orientated, early life marker for increased risk for schizophrenia.

Method

Previous findings based on systematic examination of NMD are summarised concerning schizophrenia patients and individuals at heightened genetic risk for schizophrenia.

Results

Significantly increased rates of NMD are found in patients with schizophrenia, both after and long before disease onset. Neuromotor deviation can be efficiently studied at all ages from infancy to adulthood, and is typically not in itself stigmatising. Neuromotor deviation may represent an unstable characteristic of individuals from infancy to middle-to-late childhood, possibly attaining greater stability thereafter. Neuromotor deviation may also be more stable over time in the high-risk offspring of parents with schizophrenia than in other children. Neuromotor deviation in childhood predicts subsequent serious mental disturbance in genetic high-risk children. The efficiency of NMD for predicting schizophrenia in unselected samples is unknown, as is the optimal age(s) for assessing NMD to predict schizophrenia in high-risk and normal-risk groups. Neuromotor deviation is also found in the well relatives of schizophrenia patients, individuals developing a broad range of mental and physical disorders, and mentally normal individuals in the general population.

Conclusions

Neuromotor deviation should definitely be included in batteries of methods for identifying individuals with notably heightened risk of schizophrenia. However, NMD is not pathognomonic for schizophrenia, and the investigation of NMD during childhood and adolescence must thus be used judiciously in order to avoid possible stigmatisation. Further research should be done concerning the specific forms and developmental timing of NMD as related to schizophrenia.

Neuromotor deviation (NMD) has received focal attention in the search for possible individual-orientated characteristics signalling heightened risk for schizophrenia. Researchers' interest in NMD has undoubtedly emanated from both early and consistent empirical observations of a strong association between NMD and schizophrenia. The increasing concentration on NMD, among many other early developmental factors, is likely due to the formulation of a neurodevelopmental theory of schizophrenia [1,2], and to attempts to delineate individual characteristics that would allow early identification of high-risk individuals who might benefit from intervention to prevent schizophrenia. As stated previously in this issue, the choice of markers for identifying individuals with a meaningfully increased risk of schizophrenia requires the availability of functional criteria and practicable methods, while avoiding stigmatisation of both true-positive and false-positive cases [3]. This paper thus explores central questions regarding (i) the sensitivity of NMD (i.e. to what extent does NMD characterise individuals who will later develop schizophrenia?); (ii) stability of NMD (i.e. at what ages could NMD be identified as an antecedent, and what form(s) would it have at the given age(s)?); (iii) methodological efficiency (i.e. whether NMD can be measured efficiently in practical contexts); and (iv) specificity (i.e. to what extent does NMD characterise only individuals with (pre)schizophrenia compared to other forms of (pre)psychopathology compared to normality?).

Neuromotor deviation represents inter alia motor coordination difficulties (e.g. clumsiness in walking), involuntary movements (e.g. tremor or twitching), mirror movements (e.g. in the contralateral hand when performing a task one-handed) and deviations in muscle power and tone. Neuromotor deviation items constitute the major portion of a more general adult neurological assessment, which also includes items representing cognitive, reflex and sensory domains [4]. Neuromotor deviation is also a prominent feature of childhood neurological assessment [5].

Neuromotor deviation is included in the study of schizophrenia because NMD has been observed to characterise patients with schizophrenia since the early conceptualisation of the disease [6], because schizophrenia is conceived of as a central nervous system (CNS) disease, and because assessment of NMD will provide an index of the individual's CNS functioning. The neurological abnormality associated with schizophrenia appears to represent a sign of an early and rather stable disease process [7], and is not generally a function of neuroleptic exposure [7]. A vast majority of the many studies of schizophrenia patients to date show no association at all between NMD and either previous or current neuroleptic medication [7].

Patient-Control Differences, Sensitivity and Predictability for Schizophrenia

Systematic studies conducted over more than three decades have consistently shown patients with schizophrenia to have significantly increased rates of neurological abnormality, and especially NMD, compared to both normal comparison subjects and patients with other psychiatric disorders including affective illness (for examples see Ismail et al. [4] and Torrey et al. [6]). For instance, in a recent study we conducted, we found schizophrenia patients to have significantly (p < 0.001) higher NMD scores (mean = 7.52, SD = 4.22) than normal comparison subjects (mean = 0.91, SD = 1.38), and this difference remained (p < 0.0001) when NMD that was potentially associated with neuroleptic exposure (i.e. involuntary movements), was removed from the scores [4].

Neurological abnormality (including the specific NMD domain) demonstrated an impressive sensitivity to schizophrenia in these predefined groups. A level of neurological deviation (> 6 points), which was found in only 1% of the normal comparison subjects, described fully 80% of the patients with schizophrenia. The combination of both ‘hard’ and ‘soft’ neurological signs provided a better discrimination of patients compared to comparison subjects than did either hard signs or soft signs alone [4]. Of the two classes of neurological abnormality, hard signs were more discriminatory of patients than were soft signs [4].

Considerable scientific evidence indicates that NMD is a significant antecedent of schizophrenia. Follow-back studies have shown that neurological abnormality/NMD is significantly increased during childhood among individuals later developing schizophrenia. This is true among monozygotic twin pairs discordant for schizophrenia [8], representative school populations [9] and children already showing less serious forms of mental disturbance [10]. In this latter group, NMD had more value for predicting chronicity among preschizophrenic individuals than did any other single preschizophrenic characteristic [10]. In ingenious studies conducted by Walker and colleagues, blindly assessed home movie sequences from childhood yielded evidence that individuals who later develop schizophrenia had poorer fine and gross motor coordination before 5 years of age than did normal siblings or pre-affective individuals [11,12]. Most notably, these preschizophrenia NMD characteristics concerned deviations in hand posture during the first two years of life, but not thereafter [12]. The disappearance of significant markers during the course of development poses an important challenge to professionals who wish to use these markers to predict future schizophrenia.

Prospective investigations of offspring of parents with schizophrenia have consistently demonstrated significantly increased NMD in these high-risk groups, both in infancy [13,14] and middle childhood [5,15,16]. Neuromotor deviation in the high-risk cases earlier in life has been found to be significantly related to psychiatric and cognitive disorder at 10 years [13]; to schizophrenia-related affective flattening in adolescence [17]; to poor neurobehavioural functioning and poor psychiatric adjustment in adolescence [18]; to psychiatric treatment by later adolescence [19]; and to schizophrenia or schizotypal personality or schizophrenia spectrum disorders in adolescence [18] or adulthood [12]. Given the prospective nature of these studies, these findings represent impressive evidence for the importance of NMD in these selected groups. In contrast, NMD in aggressive, hyperactive or unselected populations in childhood may be a less effective predictor of adult psychopathology [20].

Stability of this Antecedent of Schizophrenia

Neuromotor deviation exhibits different specific forms at different ages. The most notable example of this is ‘primitive’ reflexes, whose absence is abnormal in neonates and presence is abnormal in adults. The stability of age-appropriate ‘abnormality’ should thus be considered rather than the stability of the specific behaviour or sign itself across different ages.

Individual subjects' stability on NMD over time is uncertain. A study of genetic high-risk offspring found no systematic relationship between the individuals' levels of neurological abnormality as newborns and in mid-childhood, in spite of increased frequencies of NMD in the offspring group at both these ages [5]. Neurologically abnormal neonates were not more often neurologically abnormal at 6 years (33%) than were other high-risk children (27%). Similar instability was reported by Marcus and colleagues; that is, 72% of the 22 offspring who demonstrated neurological abnormality in childhood or adolescence did not show it at both of these ages [21]. That research group's most recent report reiterates the finding of low stability of neurobehavioural deviation between infancy and adolescence [18]. In contrast, they found notably high stability of neurobehavioural deviation over the seven year period between school age (mean = 10.3 years) and adolescence (mean = 17.5 years) among the offspring of parents with schizophrenia [18]. Neuromotor deviation may well be more stable in individual subjects between late childhood (10 years) and adolescence than between infancy and late childhood or adolescence.

Furthermore, the stability of NMD over time may be greater in samples at high-risk for schizophrenia than in other groups [18]. In the study by Hans and colleagues, the offspring of normal controls and of parents with nonschizophrenic psychiatric diagnoses showed only moderate stability on neurobehavioural deviation from about 10–17 years and no stability at all from infancy to adolescence [18]. Furthermore, a study of children with attention deficit hyperactive disorder, conduct disorder, etc. found unsatisfactory retest stability for NMD after only two weeks [22], while a study of patients with manifest schizophrenia found that specific signs of NMD frequently were unstable at retest one or two days later [23].

The use of multiple indices of NMD or more global neurobehavioural abnormality may improve both individual stability and predictive or discriminatory power over that obtained using a single criterion. For example, Hans and coworkers' successful prediction of poor psychiatric adjustment in adolescence employed a two-factor ‘composite’ consisting of neuromotor and cognitive–attentional variables to represent ‘neurobehavioural functioning’ [18]. Those researchers have found that the use of such two-dimensional data provides a more efficient discriminatory tool for identifying poorly functioning individuals within groups at high-risk for schizophrenia. Furthermore, a longitudinal study showed better discrimination of high-risk compared to normal-risk subjects when based on multiple hits among the four measures of NMD (i.e. the neonatal and six-year neurological examinations and the locomotor and eye–hand coordination subscales in the Griffiths Developmental Scales) [5]. The ratio of schizophrenia high-risk to normal-risk subjects categorised as abnormal was 3.0 : 1.0 (85% high risk, 28% normal risk) when counted as deviation on only one or more of these four tests. This ratio increased to 13.5 : 1.0 (54% high risk, 4% normal risk) when the criterion required deviation on two or more tests. The latter deviation rates are also more congruent with the rates of future serious psychopathology that are expected in these high-risk and normal-risk groups, respectively.

Methodological Efficiency

A number of different examination systems exist for investigating NMD and other neurological abnormality in neonates, children and adolescents or adults. Unfortunately, the methodology used in different studies has been highly varied in both specific content and size or comprehensiveness, making comparisons across studies difficult [4]. To increase possible standardisation across future studies, we have recently provided an amalgamation of methods for adults [4].

The examinations we have proposed for the neonatal period [14], middle childhood [5] and adolescence and adulthood [4] each take approximately 30 min or less; can be performed in any (semi)quiet environment; require a minimum of equipment and removal of clothing; and can be efficiently conducted by assessors who have received adequate training. Physicians tend to receive this training routinely in the course of medical education, but other professionals clearly can be trained to conduct the examinations in a reliable manner. Regardless of professional background, all examiners should perform reliability testing. The examinations are typically not stigmatising for participants themselves, as the subtle deviations that are of interest to the researchers generally go unnoticed by the subjects.

The NMD parts of the childhood and adult schedules are especially operationally defined and behaviourally orientated to optimise reliable assessment. Brazelton's experience regarding neonatal assessment was that naïve examiners are able to reach acceptable reliability within only 10 training cases [24]. We have observed 91% perfect interexaminer agreement on the specific neonatal items we have suggested [14], and found intraclass correlation coefficients of +0.87 and +0.97 (p < 0.001) for interexaminer agreement on the adult assessment [4].

Specificity for Schizophrenia

As is the case for other early risk characteristics, such as minor physical anomalies and obstetric complications [3], NMD is clearly not pathognomonic for individuals with schizophrenia. Significantly increased rates of NMD are, for example, found in the normal relatives of patients with schizophrenia, including the well cotwins in monozygotic pairs discordant for schizophrenia [4,25]. In total, 38% of the normal siblings in a recent study had elevated neurological abnormality scores found in only 5% of the normal comparison subjects [4]. The level of neurological abnormality in the well sibling was significantly positively related to the level of neurological abnormality in the patient in the same family, suggesting a ‘familial component’ in NMD. This component need not be genetic in nature, as the neurological abnormality in the well siblings was significantly associated with a history of obstetric complications [26]. The rate of NMD observed in genetic high-risk cases greatly exceeds the rate of expected schizophrenia in these groups [5,18], which also suggests that NMD exists in many high-risk individuals without being associated with manifest schizophrenia.

Furthermore, increased NMD is also found in a range of childhood and adult disorders similar to the range of disorders associated with increased rates of minor physical anomalies (MPA): hyperactivity, impulsivity, learning disorders, aggressive conduct disorder, retarded cognitive development, anxiety and depression in children, as well as impulsivity and labile mood in adult psychosis and adolescent and adult affective disorders (see Ismail et al. [4] and Vitiello et al. [22]).

Conclusion

Increased rates of NMD clearly represent early life antecedents of both schizophrenia and many other mental disorders. Neuromotor deviation predicts serious subsequent mental disturbance including schizophrenia among children with genetic risk for schizophrenia. While NMD is seemingly closely related to the schizophrenia disease process, well relatives of schizophrenia patients also have increased rates of NMD. The exact efficacy of NMD as a predictor of schizophrenia in unselected samples is unknown. Neuromotor deviation can be studied efficiently in infancy, childhood, adolescence and adulthood, but it may be unstable over time, especially during the early developmental years. The most appropriate age(s) for its identification as an antecedent of schizophrenia thus remains unclear. Neuromotor deviations and minor physical anomalies appear to occur independently of one another in both patient samples [27] and risk groups (unpubl. data, McNeil). Neuromotor deviation and MPA should thus make independent contributions to identifying risk for psychopathology. The examination for NMD is not in itself stigmatising. Nevertheless, the identification of deviant neuromotor behaviour as indicating a ‘risk for schizophrenia’ or other serious mental disorders could potentially be emotionally problematic for the individuals and families in question, and such identification must be used with caution.