Motor Issues in the Clinical High Risk Phase of Psychosis

A reliable body of research suggests that reducing the duration of untreated psychosis among those in their first episode results in demonstrable gains in clinical and functional outcomes. Along a similar line, growing evidence suggests that intervention in the Clinical High-Risk (CHR)/Ultra High-Risk phase of illness may be associated with positive outcomes including symptom improvement, delay of onset, and potential prevention of diagnosable psychosis. Early intervention in this phase of illness, however, remains controversial given that conversion rates from a typically defined risk state to diagnosable psychosis remain well below a clinically useful threshold (~36% over 3 years), and conversion itself may not be an optimal endpoint of which to predict. Nonetheless, as the risk state represents a clinically meaningful condition regardless of conversion or outcome (due to distress and impairment associated with risk symptoms), interventions that are low risk (eg, cognitive behavior therapy), low stigma, developmentally tailored, and personalized with respect to clinical targets and functional goals provide a foundation for early intervention. At present, however, treatment efforts are nonspecific, lacking precision with respect to client need (eg, psychosis vs other mental/behavioral health outcomes; see Peralta and Cuesta, this issue).¹ Improved predictive accuracy for the array of possible clinical outcomes is needed to enhance prevention/intervention effectiveness.

Self-report of psychological processes through clinical interviews such as the Structured Interview for Psychosis Risk Syndromes (SIPS) and the Comprehensive Assessment of At-Risk Mental States remain the standard bearers for prediction. A variety of additional predictors have supplemented gold standard interviews with varying implications for sensitivity and specificity. In the largest predictive model of its kind, using a North American sample from the NAPLS consortium, Cannon et al’s² risk calculator noted decline in social functioning, lower memory and verbal learning performance, slower processing speed, and younger age of risk symptom onset as significant contributors to risk over and above clinical predictors (unusual thought content and suspiciousness) for conversion among those deemed at risk through the SIPS. Other work with smaller samples have supported the added predictive value of nonpsychological/nonself-report input towards prediction, including biological measures such as in vivo imaging, psychophysiology, and blood markers.

Sharing a long history with premorbid and prodromal psychosis research,¹ motor issues may serve as an additional neurodevelopmental biological marker useful in psychosis prediction. Movement issues have been a part of the phenomenological description of schizophrenia as early as Kraepelin, and have been considered relevant signs of illness for as long as the disorder has been characterized.³ More recently, the field has accumulated a strong body of related retrospective and prospective empirical research providing evidence for motor abnormalities preceding the onset of psychosis in the absence of obvious confounds such as medication side effects.^4–6 Collectively, these bodies of literature have led many to the conclusion that motor abnormalities are a hallmark feature of psychotic illness.¹

The assessment of motor abnormalities holds an array of advantages relative to some other potential indicators of risk. Movement assessments can be performed at very low costs compared to technologies such as MRI. Additionally, with proper training and depending on the task, coding of motor abnormalities can be extremely reliable. Similar to other behaviors that reflect biological processes, much of the motor functioning relevant for psychosis is automatic and involuntary, thus bypassing any subjective perceptions or biases that might interfere with assessment of information.

Innovative motor-related research is enhancing our understanding of the underlying pathophysiology of the psychosis risk phase of illness from a number of angles (also see Mittal et al and Garvey and Cuthbert, this issue). Studies of dyskinesias, “motor agency” (degree of sense of self as the agent of a collision between an object and a pendulum using a force transducer), motor-dependent procedural learning (eg, rotor tasks), motor speed, fine motor skills (eg, arm steadiness, aiming, velocity of arms/hands fingers/wrists), neurological soft signs, postural sway, laterality through kinematic handwriting analysis, and gesture behaviors, to name a few, are paving the way for better understanding of the psychosis-risk construct. This body of work has offered insight into neurodevelopmental models of psychosis, providing clues to specific behaviors, neural regions, circuitry, and processes associated with CHR status (see also Pantelis, this issue).

Despite this work, the use of motor assessment in prediction of psychosis among those at CHR is limited. Mittal et al⁷ used discriminant function analysis based on motor and neuropsychological tests to predict transition to psychosis, with upper body movements showing the strongest relation. Effect sizes in this study were in the medium to large range. In a similar study of movement abnormalities and prediction in those at CHR, Callaway et al,⁸ asserted the value of the Abnormal Involuntary Movement Scale (AIMS) at baseline in predicting transition to psychosis at follow-up. Although confidence in findings from these studies is limited by small sample sizes, restricted range of motor abnormalities, and no assurances that variance accounted for is unique relative to other easily accessible predictors, this type of work provides a foundation for exploring baseline motor issues to predict future conversion among those at risk.

These studies offer promise, but it should be noted that not all projects in this field have yielded positive findings, and the possibility of a “file drawer” issue has to be entertained. Further, the limitations inherent in all psychosis-prediction work hold for motor abnormality prediction efforts as well. Factors such as small sample sizes, within (eg, APS vs BIPS; APS based on minimal threshold endorsement of a single scale vs APS based on ceiling endorsement across multiple scale) and between (North America vs Europe vs Australia) group heterogeneity, cultural differences, duration of follow-up, control groups (typically functioning controls vs help-seeking controls),¹ common comorbidity (eg, anxiety, depression, substance use), sampling variability (eg, specialty care seeking vs general help seeking vs general population; method of outreach), poor operationalization of “conversion,” and lack of focus on functional outcomes⁹ all threaten predictive efforts. Given the role of medication in movement, psychopharmacology needs to be considered in this work as well. Additionally, effect sizes either need to be large and/or interactions need to be identified such that we can maximize the precision of prediction. Until these factors can be overcome, clinically significant prediction will be limited. Public policies that affirm prevention efforts, recognize psychosis as an impairing yet treatable condition, reduce stigma, and encourage applied services research collaboration can all help set the stage for these efforts.

To date motor variables are relatively neglected in efforts to predict psychosis. Integrating new movement paradigm technologies (eg, accelerometers, force gauges, electromyography, motion energy analysis) alongside more classic approaches may enhance prediction. Motor pathology may also provide insights into pathophysiology and identify new therapeutic targets. The integration of this work in the context of existing large scale consortia has the capacity to add reliability through scale. Additionally, encouraging creative work outside of large consortiums can help foster innovation. All of these efforts, when combined with other biological, psychological, and social initiatives, have the potential to enhance the capacity for clinically relevant and scientifically sound prediction of psychosis and/or functional outcomes.