Speech as an objective measure of psychomotor dysfunction in major depressive disorder: validation from non-speech motor measures.

Speech as an objective measure of psychomotor dysfunction in Major Depressive Disorder: Validation from non-speech motor measures - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Psychiatry Res . Author manuscript; available in PMC: 2026 Jun 2. Published in final edited form as: Psychiatry Res. 2026 Apr 22;362:117191. doi: 10.1016/j.psychres.2026.117191 Speech as an objective measure of psychomotor dysfunction in Major Depressive Disorder: Validation from non-speech motor measures Erika L Exton 1 , David Klemballa 2 , Sebastian Walther 3 , Allison M Letkiewicz 2 , Joseph Keshet 4 , Vijay A Mittal 5 , Stewart A Shankman 2 , Matthew Goldrick 1 Author information Article notes Copyright and License information 1 Department of Linguistics, Northwestern University 2 Department of Psychiatry and Behavioral Sciences, Northwestern University 3 Department of Psychiatry, Psychosomatics and Psychotherapy, Center of Mental Health, University Hospital of Würzburg, Würzburg, Germany; Translational Research Center, University Hospital of Psychiatry and Psychotherapy, University of Bern, Bern, Switzerland. 4 Andrew and Erna Viterbi Faculty of Electrical and Computer Engineering, Technion - Israel Institute of Technology 5 Department of Psychology, Northwestern University ✉ Corresponding author: Erika L. Exton, Ph.D., erika.exton@northwestern.edu , erika.exton1@gmail.com , Postdoctoral Scholar, Department of Linguistics, Northwestern University, 2016 Sheridan Rd., Evanston, IL, USA, 60208 Issue date 2026 Aug. PMC Copyright notice PMCID: PMC13225114  NIHMSID: NIHMS2172888  PMID: 42086001 The publisher's version of this article is available at Psychiatry Res Abstract Psychomotor dysfunction, which can manifest as slowing (psychomotor retardation; PmR) and jerkiness or restlessness (psychomotor agitation; PmA) often occurs in individuals with major depressive disorder (MDD). As psychomotor dysfunction predicts a worse treatment response, accurately measuring and tracking PmR and PmA may improve outcomes. Traditionally, psychomotor dysfunction has been assessed using self-report or observer-based methods, which are often insensitive to subtle but potentially relevant movement abnormalities. Instrumental probes often require specialized equipment. Speech presents a promising option as it requires a significant motor component and is easy to collect. This study serves as a preliminary test of the association between speech indicators of PmR and PmA and manual motor tasks (handwriting velocity and force variability). Participants with current MDD (n=36) and remitted MDD (n=78) completed a diadochokinetic speech task, quickly repeating syllables in a sequence (“pataka” and “katapa”), as well as a test of handwriting velocity while drawing loops (PmR) and a test of variability in force applied to a transducer (PmA). Velocity of speech production and speech rate variability were automatically measured. For current MDD individuals, speech velocity was positively associated with the manual PmR measure (t(29) = 2.59, p < 0.05), and speech rate variability was positively associated with the manual PmA measure (t(33) = 2.952, p < 0.01). Effects were not significant for remitted MDD individuals, suggesting that this method may detect PmR/PmA only in those with acute depressive symptoms. These results suggest that diadochokinetic speech is a promising, objective measure of psychomotor dysfunction. Keywords: MDD, depression, psychomotor, motor, speech motor, speech, diadochokinetic 1. Introduction Major Depressive Disorder (MDD), along with other psychiatric conditions such as bipolar disorder and schizophrenia, has long been associated with psychomotor symptoms in both gross and fine motor behaviors ( Day, 1999 ; Sobin & Sackeim, 1997 ). In MDD, psychomotor symptoms include slowed and decreased motor movements and thoughts, referred to as psychomotor retardation (PmR), and movements that occur more frequently, intensely, or erratically, referred to as psychomotor agitation (PmA) ( Walther, et al., 2019 ). Up to 70% of individuals with MDD may experience one or both of these symptoms ( Avery & Silverman, 1984 ; Kim et al., 2015 ). The presence of psychomotor symptoms is associated with treatment response and greater depression severity (i.e., Leventhal et al., 2008 ; Iwanami et al., 2015 ; van Diermen et al., 2019 ; Gómez-Cumplido et al., 2024 ). Given the clinical importance and high prevalence of psychomotor dysfunction in MDD, assessing psychomotor dysfunction is of high clinical and research utility. Currently, PmR and PmA are typically rated using behavioral rating scales ( Sobin et al., 1998 ; Dantchev & Widlöcher, 1998 ; Parker & McCraw, 2017 ). These scales rely on the judgment of highly-trained clinicians who observe an individual’s movements, behaviors, and responses during a clinical interview. While these scales exhibit strong inter-rater reliability ( Hickie et al., 1996 ; Sobin et al., 1998 ), they may be biased by the use of ordinal/categorical scales, a reliance on rater training, and a focus on overt psychomotor disturbance rather than subtle movement abnormalities. Current objective laboratory (instrumental) assessments are able to assess subtle motor disturbances, but require the use of specialized equipment to precisely measure motor movements, limiting their usefulness outside the lab or clinic. For example, PmR has been measured via speed of writing on a tablet (handwriting velocity); and PmA has been measured via difficulty maintaining a consistent amount of force on a strain gauge (force variability) ( Caligiuri et al., 1997 ; Dean et al., 2013 ; Lohr et al., 2013 ; van Harten et al., 2017 ). There is therefore a need for tasks that can objectively probe subtle psychomotor dysfunction, but require less specialized equipment ( Shankman et al., 2020 ). Developing more objective measures may improve the ability to monitor and assess psychomotor retardation and agitation in individuals with depression without the presence of a trained clinician, which could ultimately lead to more efficient treatment guided by detailed patient information. The current mixed evidence of convergent validity between clinical ratings and objective measures of psychomotor disturbance ( Razavi et al., 2011 ; Thomas-Ollivier, et al., 2017 ; van Diermen, et al., 2018 ; Paquet, et al., 2022 ) also highlights the importance of identifying sets of objective measures that are consistent within a construct in order to validate their approaches. Speech, which requires precise coordination of many articulators, may also provide information about psychomotor dysfunction. For example, PmR may be reflected in more pauses and a general slowing of speech articulators, resulting in a slower speech rate that can be measure acoustically ( Cannizzaro et al., 2004 ; Cummins et al., 2023 ) and variability in speech rate may reflect PmA (as suggested by results from individuals at clinical high risk for psychosis; Hitczenko et al., 2023 ). Critically, in contrast to other objective measures, speech can be recorded using inexpensive microphones in a variety of settings and reliably analyzed automatically, an approach that has shown promise in identifying risk state for Parkinson’s disease ( Montaña et al., 2018 ). While prior studies have examined speech abnormalities in depression (see Briganti & Lechien, 2025 ; Cummins et al., 2015 ), there are significant problems with extant speech tasks used to probe deficits. Previous research has often utilized free speech tasks (e.g., a clinical interview where the content of what is spoken varies greatly) and evaluated speech-related features (i.e. duration measures, pitch measures, spectral measures) as biomarkers for depression. Because speech acoustics are modulated by many factors, including talkers’ emotional ( Larrouy-Maestri et al., 2024 ) and cognitive ( Goldrick et al., 2019 ) states, it is difficult to attribute acoustic differences in free speech specifically to motor processes. To isolate these factors, several researchers have introduced more controlled speech tasks that are less likely to rely on cognitive planning, such as counting, reciting the alphabet, or repeating syllables ( Hoffman et al., 1985 ; Mundt et al., 2007 ; 2012 ). The present study therefore utilizes a task developed by speech-language pathologists to test individuals’ oromotor coordination, Diadochokinetic (DDK) speech. In DDK tasks, participants rapidly repeat a syllable or set of syllables (e.g. “…papapa…” or “…pataka…”). Because this a motorically but not cognitively demanding task, performance variation in DDK tasks is assumed to largely reflect motor processing and motor coordination ( Kent et al., 2022 ; Lancheros et al., 2022 ; see the supplemental materials for a description of the motoric demands and associated motor neurons of the DDK task). DDK task performance has been related to negative affective state in individuals with and without severe mental illness ( Granrud et al., 2025 ) and motor abnormalities in individuals at clinical high risk for psychosis ( Hitczenko et al., 2023 ). In the present work, we present preliminary findings on the relationship between the acoustic properties of speech in DDK tasks and instrumental non-speech motor measures of PmR (handwriting velocity) and PmA (force variability). Previous work has found that performance on these tasks relates to anhedonia ( Maher et al., 2025 ) and depressive state ( Pokorny et al., 2025 ) in similar populations, highlighting their appropriateness for validation in this sample. Additionally, given that speech measures may be more valid measures of psychomotor disturbance in individuals with more acute/active symptoms of depression, we examined whether associations between speech and non-speech motor measures varied for those with current vs. remitted MDD. 2. Methods Overview. Individuals with current and remitted Major Depressive Disorder (cMDD and rMDD, respectively) completed a series of speech and non-speech tasks designed to assess their motor functioning as part of a larger longitudinal study on psychomotor disturbances in depression ( anonymous for peer review , 2025; anonymous for peer review, 2025). This project was approved by the Institutional Review Board at at Northwestern University (project #STU00208972). 2.1. Participants Thirty-six individuals with cMDD and 78 with rMDD participated in this study. All participants completed the Structured Clinical Interview for DSM-5 with a trained researcher who evaluated their eligibility for the study. Participants also completed the Inventory of Depression and Anxiety Symptoms – Extended (IDAS-II; Watson et al., 2012 ), from which the general depression subscore was calculated. Due to difficulty with recruitment, some participants taking medications known to affect dopaminergic functioning, such as bupropion, were included in the study. See the supplemental materials for analyses excluding these seven participants; the pattern of results does not change. Participants also withheld using certain other medications that temporarily impact dopamine or other motoric functions (e.g. antibiotics, antihistamines, antiemetics) for four weeks prior to completing the assessments. Finally, use of benzodiazepines or z-drugs for insomnia was limited to less than 3 times per week. Demographic and clinical characteristics of the sample are presented in Table 1 . There was no difference between the groups in gender (χ 2 (2,N=114) = 1.35, p = 0.51), age (t(112) = 0.02, p = 0.996), race (χ 2 (4,N=114) = 3.45, p = 0.485), ethnicity (χ 2 (1, N = 114) = 0.29, p = 0.589), or proportion taking medication, (z = 0.32, p = 0.746). Table 1. Demographic characteristics of participants. For continuous measures, range is given, followed by mean and SD in parentheses. As expected, the cMDD group had more severe depressive symptoms (t(112) = 11.53, p < 0.001), with a mean in the 89 th percentile on national norms ( Nelson et al., 2018 ). cMDD, N = 36 rMDD, N = 78 Age 19–36 ( M = 26.08, SD =5.49) 19–37 ( M = 26.06, SD =4.67) Gender 24 Female 10 Male 2 other or not reported 43 Female 29 Male 6 other or not reported Race 1 American Indian/Alaskan Native 4 Asian 5 Black or African American 21 White 5 More than one race/Unknown 11 Asian 7 Black or African American 52 White 8 More than one race/Unknown Ethnicity 7 Hispanic/Latino 29 Not Hispanic/Latino 12 Hispanic/Latino 66 Not Hispanic/Latino IDAS general depression 26–79 ( M = 60.6, SD =11.26) 20–59 ( M = 37.5, SD =9.28) Medication information: *percentages do not sum to total as some individuals were taking multiple medications Any medication for depression and/or anxiety: 27.78%* SSRI: 11.11% SNRI: 11.11% SARI: 2.78% NDRI: 5.56% Any medication for depression and/or anxiety: 30.77%* SSRI: 24.36% SNRI: 5.13% SARI: 5.13% NDRI: 6.41% Open in a new tab 2.2. Speech task: sequential motion rate diadochokinetic speech (DDK) Participants completed a sequential motion rate diadochokinetic speech task. Participants were instructed to first repeat the syllable sequence “pataka” 10 times as quickly and accurately as they could; this trial was repeated 10 times for a total of 100 repetitions of “pataka”. They then repeated the same procedure with the syllable sequence “katapa”. Given the absence of overt speech articulation problems in this population, we expect individual differences to be subtle, and therefore included a substantial number of repetitions to both tax the production system and increase statistical power. Speech samples were recorded in a quiet research lab room using a Zoom H2n Handy Recorder set to X/Y stereo, placed on a tabletop approximately 2 feet from the participant’s mouth. All recordings were manually screened for data collection errors (e.g. background noise, overlapping experimenter speech). Samples were analyzed using DDKtor ( Segal-Feldman et al., 2025 ), with a robust data filtering algorithm (described in the supplemental material ) to remove known errors from DDKtor output. We constructed composite measures of psychomotor retardation and agitation in speech because we have no strong theoretical basis to choose one individual feature of speech timing over any others. A composite measure of psychomotor retardation in speech was built from five features of speech duration: Voice onset time (VOT; the time from release of the consonant constriction to onset of the vowel), vowel duration , syllable duration (VOT + vowel), inter-syllable duration (vowel offset to onset of following VOT), and normalized syllable duration (seconds/syllable). Each feature was normalized by dividing by the average of the feature across all participants. The average of these five normalized feature values provided a composite measure of speech duration for each participant. Our measure of psychomotor retardation in speech, speech velocity , was defined as the inverse of this duration value; high-velocity individuals had smaller speech durations, and lower velocity individuals had longer speech durations. Psychomotor agitation was measured by the coefficient of variation (standard deviation/mean) within each participant for the five features above. Dividing by the mean allows us to isolate the variability in each of these features from their magnitudes. Individuals who produced the sequences quickly at a consistent rate therefore obtain low scores for speech PmA, while those who produced the sequences at a varying rate obtain high scores for speech PmA, regardless of their average speed. The average of the five coefficients of variation provided a composite measure of psychomotor agitation in speech. The speech indicators of retardation and agitation were not significantly correlated, r = 0.089, t(110) = 0.937, p = 0.351. See the supplemental materials for more details regarding how these features were measured and correlation matrices of the features that contributed to each composite measure. 2.3. Non-speech assessment of psychomotor retardation: handwriting velocity Psychomotor retardation in non-speech tasks was assessed using a handwriting task on a Fujitsu T901 tablet and analyzed using Neuroscript MovAlyzeR software ( Maher et al., 2025 ). Participants were instructed to draw circles and/or loops using their dominant hand (right in all participants) on a tablet that measures the location of the stylus over time. The absolute velocity of movement was calculated while participants completed three task conditions: 8 connected 4cm loops, as if writing cursive “l” repeatedly; 8 connected 1cm loops, as if writing cursive “e” repeatedly; 8 connected loops alternating height, as if writing cursive “lleellee”. Any leading or trailing strokes (e.g. a leading downward stroke at the beginning of a trial) were excluded from analysis. The average absolute velocity of movement from each stroke was averaged across all strokes in a trial to obtain the average velocity per condition, and the velocity across the three conditions was averaged to create a composite measure of psychomotor retardation in non-speech tasks. We use average velocity instead of velocity scaling ( Lohr et al., 2013 ; Maher et al., 2025 ) because it serves as a better parallel to the speech task. 2.4. Non-speech assessment of psychomotor agitation: force variability Psychomotor agitation in non-speech tasks was assessed using a force variability task ( Lohr et al., 2013 ; Pokorny et al., 2025 ). Participants were instructed to press on a force transducer with a sampling rate of 10 Hz for 45 seconds, using their index finger and a consistent amount of force. They completed this task with each hand and under three levels of force (high – 800 cN, medium – 600 cN, low – 400 cN). Participants saw a visual indication of the target degree of force and their output on a digital screen during each trial. The degree of force during the middle 80% of each trial was extracted, and the coefficient of variability of force during this range was calculated. The average coefficient of variability for high and medium force conditions was averaged to create a composite measure of non-speech psychomotor agitation for each participant (low force conditions were excluded due to low variability). The non-speech indicators of PmR and PmA were significantly negatively correlated, albeit with a small effect size, r = −0.203, t(104) = −2.116, p = 0.037. 2.5. Statistical approach Results from the PmR and PmA tasks were analyzed separately. To assess the relationship between speech and non-speech measures, we conducted multiple linear regression models with each non-speech measure (handwriting velocity and force variability) as the dependent variable and the speech measure (speech velocity or speech rate variability, respectively), effects-coded group (cMDD = 1, rMDD = −1) and their interaction (speech measure × group) as predictors. Speech velocity and speech rate variability were grand-mean centered for ease of interpretation, and handwriting velocity and force variability were log-transformed to improve normality of the residuals. 3. Results 3.1. Psychomotor retardation Handwriting velocity data was missing from 7 participants (4 cMDD, 3 rMDD), and one participant (cMDD) whose composite speech velocity score deviated from the mean by over 4 standard deviations was removed from the dataset, yielding n = 106. In the multiple linear regression model, there were no main effects of speech velocity, B = 0.255, SE = 0.144, t(102) = 1.777, p = 0.079, nor participant group on handwriting velocity, B = 0.011, SE = 0.026, t(102) = 0.436, p = 0.664. However, the interaction between speech velocity and participant group was significant, B = 0.316, SE = 0.144, t(102) = 2.197, p < 0.05, indicating a difference in the relationship between speech velocity and handwriting velocity between the two groups. The relationship between speech velocity and log-transformed handwriting velocity in each participant group is visualized in Figure 1 . Fig.1. Open in a new tab Composite measure of speech velocity in DDK task vs. composite measure of handwriting velocity (log transformed) for each participant, separated by group. Lines show linear regression fits (shaded region shows 95% confidence intervals). To follow up the group X speech velocity interaction, we conducted separate simple regression models for each participant group. In the cMDD group, speech velocity was a statistically significant predictor of handwriting velocity, B = 0.571, SE = 0.221, t(29) = 2.59, p < 0.05. In the rMDD group, speech velocity was not a significant predictor, B = −0.060, SE = 0.160, t = −0.378, p = 0.706. Thus, there was a positive relationship between speech velocity and handwriting velocity in the cMDD group but no significant relationship between these variables in the rMDD group. 3.2. Psychomotor agitation One participant (rMDD) whose speech rate variability score exceeded the mean by over 7.5 standard deviations and one participant (cMDD) whose force variability score exceeded the mean by over 3 SDs were removed from the dataset, yielding n = 112. In the multiple linear regression model, there was a main effect of speech rate variability, B = 2.947, SE = 0.889, t(108) = 3.315, p < 0.01, as well as a main effect of participant group on force variability, B = 0.148, SE = 0.065, t(108) = 2.273, p < 0.05. The interaction between participant group and speech rate variability shows a trend, B = 1.758, SE = 0.889, t(108) = 1.977, p = 0.051. The relationship between speech variability and log-transformed force variability in each participant group is shown in Figure 2 . Fig.2. Open in a new tab Composite measure of speech rate variability vs. composite measure of force velocity (log transformed) for each participant, separated by group. Lines show linear regression fits (shaded region shows 95% confidence intervals). To follow-up the group X speech rate variability interaction, we conducted separate simple regression models for each participant group. Speech rate variability was a statistically significant predictor of force variability in the cMDD group, B = 4.704, SE = 1.594, t(33) = 2.952, p < 0.01. However, in the rMDD group, the effect of speech rate variability was not statistically significant, B = 1.189, SE = 0.955, t(75) = 1.245, p = 0.217. 3.3. Specificity of effects to psychomotor retardation vs. psychomotor agitation To test whether these correlations are specific to the construct in question (PmR or PmA), we tested the correlation between speech velocity and force variability, as well as the correlation between speech rate variability and handwriting velocity, in the cMDD group. There was no evidence of a relationship between speech PmR and force variability, r = −0.128, t(33) = −0.743, p = 0.463, nor between speech PmA and handwriting velocity, r = −0.027, t(30) = −0.148, p = 0.883. 4. Discussion Speech may offer an easily acquired objective measure of psychomotor symptoms in Major Depressive Disorder. To provide an initial assessment of the feasibility of this approach, we tested whether automatically analyzed measures of speech in a task targeting motor control (the diadochokinetic speech task) correlated with non-speech motor measures of psychomotor retardation and psychomotor agitation. We found that among individuals currently experiencing MDD, speech velocity was significantly positively correlated with handwriting velocity. Similarly, speech rate variability was positively correlated with the variability in force applied to a transducer by those same individuals. No significant correlations between the speech and non-speech motor tasks were found for individuals whose MDD was in remission. Further, the relationship between speech and non-speech measures was specific to the respective psychomotor construct, as speech rate variability in currently depressed individuals was related to force variability but not handwriting velocity, and speech velocity was related to handwriting velocity but not force variability. There are several possible reasons that speech was an indicator of manual psychomotor dysfunction in current, but not remitted, MDD. First, speech may only detect PmR/PmA in individuals with acute or severe symptoms of depression. Severe depression may relate to significant disruption to central processes tied to psychomotor retardation and agitation, such as the functioning of cerebellar-thalamic circuits ( Mittal et al., 2020 ), processes which have been shown to differ in current vs. remitted MDD ( Wüthrich et al., 2024 ). In the absence of such strong neural disruptions, variation in speech and motor tasks may primarily be driven by variability in the modality-specific neural mechanisms and muscle groups that lead to a given motor action ( Peter & Stoel-Gammon, 2008 ). As an example, during the DDK task, a participant may experience disruption to brainstem processes associated with the cranial nerves, a lack of airflow due to poor breath planning, or problems in coordination of the tongue and jaw. These difficulties would affect their speech velocity and variability but not their handwriting velocity or force variability. Second, it is possible that effects are present, but weaker in remitted MDD, necessitating a larger sample to detect valid psychomotor indicators in speech in that population. This work also highlights the difficulty in defining PmA in particular ( Day, 1999 ). The symptom of agitation as it is rated by clinical observers (driven by restlessness and intense or erratic movements) may be influenced by multiple individual motor behaviors. A single instrumental approach, such as force variability or speech rate variability as used here, likely cannot capture all the different manifestations of agitation; single measures may instead partially reflect a latent agitation construct leading to erratic behavior. Future studies that assess multiple aspects of agitation from speech, fine motor, and gross motor movements could utilize factor analysis to examine shared variance across different aspects of agitation. In order to replicate and extend these findings, future work should continue to investigate the relationship between measures of PmR and PmA across task demands and involved motor functions. DDK speech aims to stress the oral-motor system with minimal cognitive or emotional demands. However, as psychomotor dysfunction is driven by the interaction between mental activity (i.e. cognitive and emotional) and motor demands, it may be that the ability to detect PmR or PmA in a given task is impacted by demands on both components. To explore this issue, future studies could compare motorically challenging tasks like DDK to more cognitively challenging tasks that require linguistic planning (e.g., describing an image or telling a story). Future work should also consider the role of neural connectivity on performance on these speech and non-speech motor tasks in order to assess whether the functioning of specific pathways is related to the observed behaviors. For example, recent research examining resting state functional connectivity in MDD found differences in thalamo-cortical, pallido-cortical, and cortico-cortical connectivity, as well as motor network topology, among current and remitted depressed MDD with and without psychomotor disturbance ( Wüthrich et al., 2024 ; Liang et al., 2025 ). Relating such connectivity to specific behavioral measures would allow us to better assess potential neural mechanisms of psychomotor dysfunction ( Northoff et al., 2021 ). For example, resting-state connectivity in psychomotor slowing in psychosis has been linked to aberrant GABAergic neurophysiology in the primary motor cortex and to actigraphy measures of physical activity ( Lefebvre et al., 2024 ). As polypharmacy is frequent in this population, future work should also systematically test the possible impacts of medication class and examine these measures in medication naïve participants. Our sensitivity analysis indicated similar results when excluding individuals on bupropion, suggesting that use of a medication which directly impacts dopamine did not have a meaningful effect on the results. In summary, the present study found preliminary evidence that DDK speech tasks can elicit specific measures that align with PmR and PmA in current MDD. These pilot results are important given the clinical relevance of psychomotor disturbance ( Schrijvers et al., 2008 ). Future work with a larger sample that includes healthy controls, stronger reliability testing within the PmR and PmA constructs, and replication across samples will be important to establish robust validation of these speech-based measures, as well as to examine whether the speech task performance tracks with treatment-induced changes in depressive severity over time and/or serves as an easily assessable baseline risk factor for worse course of illness. Supplementary Material 1 NIHMS2172888-supplement-1.docx (30.1KB, docx) Highlights. Diadochokinetic speech may be objective measure of psychomotor dysfunction in MDD Speech velocity correlates with handwriting velocity in current MDD Speech rate variability correlates with force velocity in current MDD Acknowledgment This work was funded by NIH grants R01MH118741 and R01MH134369. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Thanks to Maya Kraidy, Hallee Simmons, and Victor Pokorny for assistance in data analysis and to Lauren Grzelak, Riley Maher, and Megan McFarland with assistance in data collection. Footnotes Declaration of interests The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Stewart Shankman, Vijay Mittal, Matthew Goldrick, Sebastian Walther reports financial support was provided by National Institute of Mental Health. Sebastian Walther reports a relationship with Neurolite that includes: speaking and lecture fees. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. CrediT author statement Erika L. Exton: conceptualization, formal analysis, writing - original draft, writing – review & editing, visualization David Klemballa: investigation, project administration, writing – review & editing Sebastian Walther: conceptualization, methodology, funding acquisition, writing – review & editing Allison M. Letkiewicz: conceptualization, methodology, writing – review & editing Joseph Keshet: software, writing – review & editing Vijay A. Mittal: conceptualization, methodology, funding acquisition, writing – review & editing Stewart A. Shankman: conceptualization, methodology, funding acquisition, writing – review & editing Matthew Goldrick: conceptualization, methodology, funding acquisition, supervision, writing – review & editing Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. References Avery D, Silverman J Psychomotor retardation and agitation in depression, (1984). Relationship to age, sex, and response to treatment. J Affect Disord. 7(1), 67–76. doi: 10.1016/0165-0327(84)90066-1. [ DOI ] [ PubMed ] [ Google Scholar ] Briganti G, & Lechien JR (2025). Speech and Voice Quality as Digital Biomarkers in Depression: A Systematic Review. J Voice, S0892–1997(25)00187–0. 10.1016/j.jvoice.2025.05.002 [ DOI ] [ PubMed ] [ Google Scholar ] Caligiuri MP, Lohr JB, Rotrosen J, Adler L, Lavori P, Edson R, Tracy K, & Veterans Affairs Cooperative Study #394 Study Group (1997). Reliability of an instrumental assessment of tardive dyskinesia: results from VA Cooperative Study #394, Psychopharmacology, 132, 61–66. [ DOI ] [ PubMed ] [ Google Scholar ] Cannizzaro M Harel B, Reilly N, Chappell P, Snyder PJ (2004). Voice acoustical measurement of the severity of major depression. Brain Cogn. 56(1), 30–35. doi: 10.1016/j.bandc.2004.05.003. [ DOI ] [ PubMed ] [ Google Scholar ] Cummins N, Scherer S, Krajewski J, Schnieder S, Epps J, Quatieri TF (2015). A review of depression and suicide risk assessment using speech analysis. Speech Communication, 71, 10–49. doi: 10.1016/j.specom.2015.03.004. [ DOI ] [ Google Scholar ] Cummins N, Dineley J, Conde P, Matcham F, Siddi S, Lamers F, Carr E, Lavelle G, Leightley D, White KM, Oetzmann C, Campbell EL, Simblett S, Bruce S, Haro JM, Penninx BWJH, Ranjan Y, Rashid Z, Stewart C, … The RADAR-CNS Consortium. (2023). Multilingual markers of depression in remotely collected speech samples: a preliminary analysis. J Affect Disord. 341, 128–136. doi: 10.1016/j.jad.2023.08.097. [ DOI ] [ PubMed ] [ Google Scholar ] Dantchev N & Widlöcher D (1998). The measurement of retardation in depression. J Clin Psychiatry. 59[suppl 14], 19–25. [ PubMed ] [ Google Scholar ] Day RK (1999). Psychomotor agitation: poorly defined and badly measured. J Affect Disord. 55(2–3), 89–98. Doi: 10.1016/S0165-0327(99)00010-5. [ DOI ] [ PubMed ] [ Google Scholar ] Dean DJ, Teulings H-L, Caligiuri M, & Mittal VA (2013). Handwriting analysis indicates spontaneous dyskinesias in neuroleptic naïve adolescents at high risk for psychosis. J Vis Exp. 81, e50852, doi: 10.3791/50832. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Goldrick M, McClain R, Cibelli E, Adi Y, Gustafson E, Moers C, Keshet J (2019). The influence of lexical selection disruptions on articulation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 45(6), 1107–1141. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Gómez-Cumplido J, Izquierdo A, Tobiasz-Adamczyk B, Koskinen S, Haro JM, Ayuso-Mateos JL, Cabello M, (2024). Do depressed people with subjective psychomotor retardation show a different symptomatic pattern? A network analysis approach using a cross-national sample. Eur J Psychiatry. 38(3), 100253. [ Google Scholar ] Granrud OE, Corporandy A, Holmlund TB, Elvevåg B, Foltz PW, Cowan T, Masucci MD, & Cohen AS (2025). Dynamic or not, here I come: Ambulatory assessment of psychomotor ability for clinical predictions. Schizophrenia Research, 282, 49–55. doi: 10.1016/j.schres.2025.05.029. [ DOI ] [ PubMed ] [ Google Scholar ] Hickie I, Mason C, Parker G (1996). Comparative validity of two measures of psychomotor function in patients with severe depression. J Affect Disord. 37(2–3), 143–149. doi: 10.1016/0165-0327(95)00087-9. [ DOI ] [ PubMed ] [ Google Scholar ] Hitczenko K, Segal Y, Keshet J, Goldrick M, Mittal VA (2023). Speech characteristics yield important clues about motor function: speech variability in individuals at clinical high-risk for psychosis. Schizophr. 9, 60. doi: 10.1038/s41537-023-00382-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Hoffman GMA, Gonze JC, Mendlewicz J. Speech Pause Time as a Method for the Evaluation of Psychomotor Retardation in Depressive Illness. British Journal of Psychiatry. 1985;146(5):535–538. doi: 10.1192/bjp.146.5.535 [ DOI ] [ PubMed ] [ Google Scholar ] Iwanami T, Maeshima H, Baba H, Satomura E, Namekawa Y, Shimano T, Suzuki T, Arai H, (2015). Psychomotor agitation in major depressive disorder is a predictive factor of mood-switching. J Affect Disord. 170, 185–189. doi: 10.1016/j.jad.2014.09.001. [ DOI ] [ PubMed ] [ Google Scholar ] Kent RD, Kim Y, Chen L-M. (2022). Oral and laryngeal diadochokinesis across the life span: a scoping review of methods, reference data, and clinical applications. J Speech Lang Hear Res. 65(2), 574–623. doi: 10.1044/2021_JSLHR-21-00396. [ DOI ] [ PubMed ] [ Google Scholar ] Kim J-H, Cho MJ, Hong JP, Bae JN, Cho SJ, Hahm BJ, Lee DW, Park JI, Lee JY, Jeon HJ & Chang SM (2015). Gender differences in depressive symptom profile: results from nationwide general population surveys in Korea. J Korean Med Sci. 30(11), 1659–1666. Doi: 10.3346/jkms.2015.30.11.1659. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Larrouy-Maestri P, Poeppel D, Pell MD (2024). The sound of emotional prosody: nearly 3 decades of research and future directions. Perspectives on Psychological Science, 20(4), 623–638. doi: 10.1177/17456916231217722 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Lancheros M, Pernon M, Laganaro M (2022). Is there a continuum between speech and other oromotor tasks? Evidence from motor speech disorders. Aphasiology, 37(5), 715–734. doi: 10.1080/02687038.2022.2038367. [ DOI ] [ Google Scholar ] Lefebvre S, Gehrig G, Nadesalingam N, Nuoffer MG, Kyrou A, Wüthrich F, & Walther S (2024). The pathobiology of psychomotor slowing in psychosis: altered cortical excitability and connectivity. Brain, 147(4), 1423–1435. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Leventhal AM, Pettit JW and Lewinsohn PM (2008), Characterizing major depression phenotypes by presence and type of psychomotor disturbance in adolescents and young adults. Depress. Anxiety, 25, 575–592. 10.1002/da.20328 [ DOI ] [ PubMed ] [ Google Scholar ] Liang Q, Zhou Z, Li Y, Chen S, Lin S, Lin X, Zhang Y, Peng B, Xu Z, Hou G Qiu, Y. (2025). Motor circuits and beyond: Functional connectivity related to psychomotor syndromes in depression. Psychol Med, 55, e285. doi: 10.1017/S0033291725101852. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Lohr JB, May T, & Caligiuri MP (2013). Quantitative assessment of motor abnormalities in untreated patients with major depressive disorder. J Affect Disord. 146(1), 84–90. doi: 10.1016/j.jad.2012.08.043. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Maher RE, Klemballa DM, Walther S, Mittal VA, Shankman SA, & Letkiewicz AM (2025). Anhedonia relates to increased psychomotor retardation using an instrumental handwriting-based measure. J Psychopathol Behav Assess. 47, 74. Doi: 10.1007/s10862-025-10249-1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Mittal VA, Bernard J, Walther S (2020). Cerebellar-thalamic circuits play a critical role in psychomotor function. Mol Psychiatry, 26(8), 3666–3668. doi: 10.1038/s41380/020-00935-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Montaña D, Campos-Roca Y, & Pérez CJ (2018). A diadochokinesis-based expert system considering articulatory features of plosive consonants for early detection of Parkinson’s disease. Computer Methods and Programs in Biomedicine, 154, 89–97. doi: 10.1016/j.cmpb.2017.11.010. [ DOI ] [ PubMed ] [ Google Scholar ] Mundt JC, Snyder PJ, Cannizzaro MS, Chappie K, Geralts DS (2007). Voice acoustic measures of depression severity and treatment response collected via interactive voice response (IVR) technology. J Neurolinguistics. 20(1), 50–64. doi: 10.1016/j.neuroling.2006.04.001. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Mundt JC, Vogel AP, Feltner DE, Lenderking WR (2012). Vocal acoustic biomarkers of depression severity and treatment response. Biol Psychiatry, 72(7), 580–587. doi: 10.1016/j.biopsych.2012.03.015. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Nelson GH, O’Hara MW, Watson D (2018). National norms for the expanded version of the inventory of depression and anxiety symptoms (IDAS-II). Journal of Clinical Psychology, 74(6), 953–968. doi: 10.1002/jclp.22560. [ DOI ] [ PubMed ] [ Google Scholar ] Northoff G, Hirjak D, Wolf RC, Magioncalda P, Martino M (2021). All roads lead to the motor cortex: psychomotor mechanisms and their biochemical modulation in psychiatric disorders. Mol Psychiatry, 26, 92–102. doi: 10.1038/s41380-020-0814-5. [ DOI ] [ PubMed ] [ Google Scholar ] Paquet A, Lacroix A, Calvet B, & Girard M (2022). Psychomotor semiology in depression: a standardized clinical psychomotor approach. BMC Psychiatry, 22(1), 474. doi: 10.1186/s12888-022-04086-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Parker G & McCraw S (2017). The properties and utility of the CORE measure of melancholia. J Affect Disord. 207, 128–135. doi: 10.1016/j.jad.2016.09.029. [ DOI ] [ PubMed ] [ Google Scholar ] Peter B & Stoel-Gammon C (2008). Central timing deficits in subtypes of primary speech disorders. Clinical Linguistics & Phonetics, 22(3), 171–198. Doi: 10.1080/02699200701799825 [ DOI ] [ PubMed ] [ Google Scholar ] Pokorny VJ, Anderson Z, Letkiewicz AM, Hirjak D, Walther S, Shankman S, Mittal V (2025). Visuomotor dysconnectivity as a candidate mechanism of psychomotor agitation in major depression. Psychological Medicine, 55, e363, 1–10. 10.1017/S0033291725102638 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Razavi N, Horn H, Koschorke P, Hügli S, Höfle O, Müller T, Strik W, & Walther S (2011). Measuring motor activity in major depression: the association between the Hamilton Depression Rating Scale and actigraphy. Psychiatry Res, 190(2–3), 212–216. doi: 10.1016/j.psychres.2011.05.028. [ DOI ] [ PubMed ] [ Google Scholar ] Schrijvers D, Hulstijn W, Sabbe BGC (2008). Psychomotor symptoms in depression: a diagnostic, pathophysiological and therapeutic tool. J Affect Disord, 109(1–2), 1–20. doi: 10.1016/j.jad.2007.10.019 [ DOI ] [ PubMed ] [ Google Scholar ] Segal-Feldman Y, Hitczenko K, Goldrick M, Buchwald A, Roberts A, Keshet J (2025). Enhancing analysis of diadochokinetic speech using deep neural networks. Computer Speech & Language. 90, 101715. doi: 10.1016/j.csl.2024.101715. [ DOI ] [ Google Scholar ] Shankman SA, Mittal VA, & Walther S (2020). An examination of psychomotor disturbance in current and remitted MDD: an RDoC study. J Psychiatr Brain Sci. 5, e200007. Doi: 10.20900/jpbs.20200007. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Shankman SA, Glazer JE, Rappaport BI, Li LY, Wüthrich F, Grzelak LN, Walther S, & Mittal VA (2025). Disentangling the effects of daily physical activity and natural white light exposure on affect. J Psychopathol Clin Sci. 134(5), 520–526. Doi: 10.1037/abn0000995. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Sobin C, Sackeim HA, (1997). Psychomotor symptoms of depression. Am J Psychiatry. 154(1), 4–17. doi: 10.1176/ajp.154.1.4. [ DOI ] [ PubMed ] [ Google Scholar ] Sobin C, Mayer L, Endicott J (1998). The motor agitation and retardation scale: a scale for the assessment of motor abnormalities in depressed patients. J Neuropsychiatry Clin Neurosci. 10(1), 85–92. doi: 10.1176/jnp.10.1.85. [ DOI ] [ PubMed ] [ Google Scholar ] Thomas-Ollivier V, Foyer E, Bulteau S, Pichot A, Valriviere P, Sauvaget A, Deschamps T (2017). Psychiatry Clin Neurosci. 71(9), 612–623. Doi: 10.1111/pcn.12529. [ DOI ] [ PubMed ] [ Google Scholar ] van Diermen L, Walther S, Cools O, Fransen E, Birkenhäger TK, Sabbe BCG, & Schrijvers D (2018). Observer-rated retardation but not agitation corresponds to objective motor measures in depression. Acta Neuropsychiatrica, 30(6), 359–364. doi: 10.1017/neu.2018.21. [ DOI ] [ PubMed ] [ Google Scholar ] van Diermen L, Vanmarcke S, Walther S, Moens H, Veltman E, Fransen E, Sabbe B, van der Mast R, Birkenhäger T, Schrijvers D, (2019). Can psychomotor disturbance predict ect outcome in depression? J Psychiatr Res. 117, 122–128. doi: 10.1016/j.jpsychires.2019.07.009. [ DOI ] [ PubMed ] [ Google Scholar ] van Harten PN, Walther S, Kent JS, Sponheim SR, Mittal VA (2017). The clinical and prognostic value of motor abnormalities in psychosis, and the importance of instrumental assessment. Neuroscience & Biobehavioral Reviews, 80, 476–487. doi: 10.1016/j.neubiorev.2017.06.007. [ DOI ] [ PubMed ] [ Google Scholar ] Walther S, Bernard JA, Mittal VA, & Shankman SA (2019). The utility of an RDoC motor domain to understand psychomotor symptoms in depression. Psychol Med, 42(2), 212–216. doi: 10.1017/S0033291718003033. [ DOI ] [ PubMed ] [ Google Scholar ] Watson D, O’Hara MW, Naragon-Gainey K, Koffel E, Chmielewski M, Kotov R, Stasik SM & Ruggero CJ (2012). Development and validation of new anxiety and bipolar symptom scales for an expanded version of the IDAS (the IDAS-II). Assessment, 19(4), 399–420. 10.1177/1073191112449857. [ DOI ] [ PubMed ] [ Google Scholar ] Wüthrich F, Lefebvre S, Mittal VA, Shankman SA, Alexander N, Brosch K, Flinkenflügel K, Goltermann J, Grotegerd D, Hahn T, Jamalabadi H, Jansen A, Leehr EJ, Meinert S, Nenadić I, Nitsch R, Stein F, Straube B, Teutenberg L, … Walther S (2024). The neural signature of psychomotor disturbance in depression. Mol Psychiatry, 29, 317–326. 10.1038/s41380-023-02327-1 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Ziegler W & Ackermann H (2014). ‘Neural bases of phonological and articulatory processing’ in Goldrick M, Ferreira VS, & Miozzo M (eds), The Oxford Handbook of Language Production, Oxford Library of Psychology. 10.1093/oxfordhb/9780199735471.013.016 [ DOI ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials 1 NIHMS2172888-supplement-1.docx (30.1KB, docx) ACTIONS View on publisher site PDF (565.1 KB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Auditory Masking Effects on Speech Fluency in Aphasia and Apraxia of Speech#969524
• Mismatch Responses to Speech Contrasts in Preschoolers with and without Developmental Language Disorder#277482
• Ukulele Playing to Improve Cognition in People with Multiple Sclerosis: a Feasibility Study#1124415
• Damped Physical Activity and Unstable Sleep: Fitbit-Derived Rest-Activity Phenotypes in Inter-episode Mood Disorders#682332
• Virtual Reality Intervention for Speech Anxiety#1008809