Detecting Motor Slowing in Clinical High-Risk for Psychosis in a Computerized Finger Tapping Model

Abstract

Finger tapping is sensitive to motor slowing and emerging symptoms in individuals at Clinical High-Risk for psychosis (CHR). A sensitive, computerized finger tapping task would be beneficial in early psychosis screening batteries. The study included 41 CHR and 32 healthy volunteers, who completed a computerized finger tapping task and clinical interviews. This computerized finger tapping task was sensitive to slowing in the CHR group compared to healthy volunteers, and as expected negative but not positive symptoms related to motor slowing. Computerized finger tapping tasks may be an easily dispersible tool for early symptom detection battery relevant to emerging negative symptoms.

Introduction

Schizophrenia (SZ) is a major cause of disability worldwide, and in spite of existing treatment, patients continue to demonstrate significant disability [1, 2], which may be related to the duration of untreated psychosis [3, 4]. As a result, early detection and intervention in individuals at clinical high-risk (CHR) is critical for early intervention in order to improve the trajectory of psychosis progression. However, extant CHR screenings require clinical training and are impractical for widespread dissemination of CHR detection methods. Advancement and validation of online neuropsychological batteries, that engage mechanisms underlying emerging symptoms, may be a beneficial tool in CHR screening in clinical practice.

The finger tapping task is a well-established neuropsychological measure of timing and motor deficits that has been validated for online use [5, 6]. Critically, patients with psychosis show motor slowing [7, 8] that appears early in the course of psychosis development [9–17]. This motor slowing in psychosis may be related to coordination among basal ganglia, cerebellar, and cortico-motor circuits [18–20], which is disrupted across disease stage [10, 21–24]. The extant literature demonstrates that motor abnormalities are predictive of future conversion to psychotic disorders [25] and future symptom progression [26–28]. Relatedly, there is emerging evidence that speaks to an important role for slowing in CHR individuals [11] and strong links with negative symptoms in schizophrenia [29–31]. These clinical features make motor abnormalities a great candidate for early risk detection. Despite this potential for clinical utility, typical motor assessment approaches require trained administrators or expensive equipment that is not easily distributable, which limits the utility of motor slowing metrics in clinical practice. The present study examines the possibility that online finger tapping tasks may be sensitive to risk for psychosis, which would provide a time and cost-efficient measure that is easily distributed for clinical use.

Methods

Participants

A total of 73 participants (59% female; CHR=41, healthy volunteers (HV)=32) age 15–24 (M=20.66, SD=2.14) completed a computerized neuropsychiatry test battery and clinical interviews. All participants provided informed consent and data was collected in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. The majority of the sample was (90.4%) right-handed. All subjects were administered the Structured Clinical Interview for DSM-IV Axis I disorders (SCID) to rule out a psychosis diagnosis and assess history of other disorders [32]. The Structured Interview Structured Interview for Psychosis-Risk Syndromes (SIPRS) was administered to diagnose CHR subjects and rule out symptoms in HV [33]. Medication with antipsychotics was an exclusionary criterion as antipsychotics have been related to motor slowing [34], and so none of the current sample of CHR subjects were medicated with antipsychotics.

Finger Tapping Task

Participants completed the Computerized Finger-tapping Test (CTAP) [5, 35] from the Penn Computerized Neurocognitive Battery (CNB). They were instructed to press the spacebar as quickly as possible moving only the index finger for 10 seconds after their first button press. During the task, trial blocks alternate between the dominant and non-dominant hand over a total of 10 trials (5 trials per hand). Mean finger tapping behavior will be compared across groups; if there are group differences in finger tapping, then potential relationship to symptoms will be explored in a single model. These analyses were completed using SPSS 25.0.

Results

Participants

There were not significant differences in sex (CHR: 22F/19M; HV: 21F/11M), χ²=1.06, p=.30, or handedness χ²=0.56, p=.46, by group. There was not a significant difference in age between groups, t(71)=0.22, p=.82, and age did not significantly relate to finger tapping performance, F(1,71)=1.69, p=.20. As expected, the CHR group showed significantly higher total positive symptoms, t(71)=14.18, p<.001, and total negative symptoms, t(71)=5.92, p<.001, on the SIPRS compared to the HV group, Figure 1B.

Figure 1.

Finger Tapping Task: A. Finger Tapping (CTAP) Trial Structure B. Group differences in Mean Total Finger Presses (From Left to Right: Mean Number Presses by Group Separated by Hand Dominance, Model corrected Number of Button Presses by Group Across Handedness Conditions) C. Finger Tapping by SIPRS Total Negative (top panel) and Positive Symptoms Score (bottom panel)

Computerized Finger Tapping (CTAP) by Group

Finger tapping behavior was analyzed in a repeated-measure general linear model, which compared motor speed a cross hands (dominant, non-dominant) as the within-subject factor with group (CHR, HC) and sex (female, male) as between-subject factors. Sex was included as a factor because of known sex differences in handedness [5, 36]. There was a significant main effect of group, F(1,69)=4.80, p=.03, ɳ²_p=.07 with the CHR group (M=52.58, SEM=.91) showing more motor slowing compared to the HV group (M=55.68, SEM=1.08), Figure 1B. The group difference was present across hands and reflected a general motor slowing as there was no difference in handedness condition by group, p=.70. As expected, there was a significant hand by gender interaction, F(1,69)=4.33, p=.04, ɳ²_p=.06, driven by a greater motor speed difference across hands in women compared to males, but no significant main effect of gender was found, p=.08. There were no other significant main or interactive effects, p’s>.57. Follow-up analyses, found no main or interactive impact of block related to group, and that age had no impact on the significance or direction of the findings when added as a covariate.

Computerized Finger Tapping (CTAP) by Symptoms

Within the CHR group, a repeatedmeasure general linear model, which compared motor speed across hands (dominant, non-dominant) as the within-subject factor and sex (female, male), total positive symptoms, and total negative symptoms as between-subject factors. Motor speed was not related to SIPRS total positive symptom score, F(1,40)=3.19, p=.08, r= −.29, ɳ²_p=.08; which was consistent across hands with no interaction by hand, p=.92. Motor speed was related to SIPRS total negative symptom score, F(1,40)=5.75, p=.03, r=.37, ɳ²_p=.13; which was consistent across hands with no interaction by hand, p=.93. Again, there was a significant hand by gender interaction, F(1,40)=4.10, p=.05, ɳ²_p=.10. This interaction was driven by female participants showing a greater difference in tapping across hands compared to males, but no significant main effect of gender was found, p=.98. There were no other significant main or interactive effects, p’s>.92.

Discussion

CHR individuals showed significant motor slowing, i.e., a reduced number of button presses, compared to healthy volunteers regardless whether the hand was dominant or non-dominant, which is consistent previous psychosis literature [9, 13–16]. Additionally, female participants showed greater difference between dominant and non-dominant hands compared to males [5, 36], which did not vary by group. Within the CHR group, negative but not positive symptoms related to motor slowing. This finding is in line with other work from our lab [10, 11, 24], and others [15, 16], suggesting that motor slowing appears to be tied unique to negative symptoms. This is also consistent with a cognitive dysmetria framework [37, 38].

In summary, finger tapping was sensitive to the presence of attenuated symptoms in the CHR group, and reflected the magnitude of negative symptoms after accounting for positive symptoms. Although, motor slowing related to total negative symptom score, but it is possible that this relationship may reflect deficits in overlapping basal ganglia, cerebellar, and cortico-motor circuits [24, 39, 40]. Critically, these circuits underlie both motor and cognitive processes; disruptions in these circuits may result in motor and cognitive deficits that comprise negative symptoms [10, 24, 39, 40]. Future studies should examine the extent to which computerized finger tapping may be reflect underlying basal ganglia, cerebellar, and cortico-motor circuit dysconnectivity [10, 27].

Notably, this task is brief, and can be feasibly administered on any home computer keyboard (all that is required is an Internet connection and a keyboard). Therefore computerized, online finger tapping tasks may be a critical and easily dispersible tool for clinical practice that is also relevant to current emerging negative symptoms. As previously noted, motor symptoms are predictive of psychosis onset and symptom progression [25, 26, 28]. Accordingly, finger tapping may be also be a predictive, clinical tool that marks conversion. The utility of finger tapping as a predictive tool should be explored in future longitudinal analyses of conversion and symptom progression.

The current study highlights several important topics, but there are several limitations to consider. The current sample does not contain conversion or follow-up information, which could assess the sensitivity of motor slowing in predicting future symptoms. The sample size of the current study is consistent with much of the extant literature (Becker et al., 2010; Gschwandtner et al., 2006; Keefe et al., 2006; Lencz et al., 2006; Niendam et al., 2007, 2006; Woodberry et al., 2013), future studies would benefit from larger sample sizes. The computerized finger tapping task has been validated in other studies [6, 35], the current study did not validate the motor slowing with alternate finger tapping tasks within the current sample. The current study also focuses on a single, behavioral approach to assessing motor slowing; future studies may benefit from a multimodal approach to examine biological mechanisms. Finally, future research should examine if this finger tapping task can effectively distinguish individuals at clinical high-risk in large, population screenings.