Psychomotor retardation and agitation (i.e. psychomotor disturbances) are core features of depression (one of nine primary phenomenology in Diagnostic and Statistical Manual 5) (APA, 2013) and reflect deficits in fine motor function as well as gross motor behavior (Sobin and Sackeim, Reference Sobin and Sackeim1997). Clinicians’ ratings of retardation or agitation are as frequent as 60–70% in acute depressive episodes, as indicated by the STAR*D trial (Novick et al., Reference Novick, Stewart, Wisniewski, Cook, Manev, Nierenberg, Rosenbaum, Shores-Wilson, Balasubramani, Biggs, Zisook and Rush2005). These disturbances are particularly frequent in episodes of severe depression (Parker, Reference Parker2000) and have been linked to poor outcome in treatment trials of depression (Ulbricht et al., Reference Ulbricht, Dumenci, Rothschild and Lapane2016, Reference Ulbricht, Dumenci, Rothschild and Lapane2018; Sakurai et al., Reference Sakurai, Suzuki, Yoshimura, Mimura and Uchida2017). Furthermore, clinician-rated psychomotor disturbances are suggested to distinguish the melancholic subtype of unipolar depression (Sobin and Sackeim, Reference Sobin and Sackeim1997; Parker, Reference Parker2000) and also distinguish major depressive disorder from bipolar disorder, with stronger effects for retardation than agitation (Leonpacher et al., Reference Leonpacher, Liebers, Pirooznia, Jancic, MacKinnon, Mondimore, Schweizer, Potash, Zandi and Goes2015). A longitudinal investigation of subjects with bipolar disorder reported that clinician-rated agitation was the best discriminator between mood episodes, performing better than retardation, and most importantly, much better than mood symptoms of common clinical rating scales for depression and mania (Cheniaux et al., Reference Cheniaux, Silva, Santana and Filgueiras2018). Other motor abnormalities such as neurological soft signs are associated with psychosis in multiple disorders, including depression (Owoeye et al., Reference Owoeye, Kingston, Scully, Baldwin, Browne, Kinsella, Russell, O'Callaghan and Waddington2013; Sagheer et al., Reference Sagheer, Assaad, Haddad, Hachem, Haddad and Hallit2018). In sum, psychomotor disturbances are of clinical value in determining poor outcome or treatment response.

Despite this importance, research on motor issues in depression is quite scarce. Some have argued that shifting the focus of psychopathology from (observable) signs to (self-reported) symptoms has led to an under-appreciation of overt behavior, a trend particularly fateful for motor disturbances (Kendler, Reference Kendler2016). However, in the last decade some groups have investigated the role of motor abnormalities in psychiatric disorders using new instrumental measures of motor behavior, allowing for objective and dimensional assessment (van Harten et al., Reference van Harten, Walther, Kent, Sponheim and Mittal2017). Today, with mobile technology being widely available, it is possible to collect large scale health instrumental assessments of motor behaviors. This approach to motor phenomena conceptualizes motor disturbance dimensionally, enables the integration with other types of motor assessment (e.g. lab-based tests, experiments with brain stimulation, ecological momentary assessment, and smartphone applications) and consequently, holds significant relevance for the NIMH Research Domain Criteria (RDoC).

These novel measures allow further insight into the neural mechanisms guiding multiple motor disturbances (Walther, Reference Walther2015; Walther and Mittal, Reference Walther and Mittal2017). Classical studies on motor disturbances in psychiatric disorders have usually applied observer-based rating scales covering one particular sign or a group of symptoms (e.g. tardive dyskinesia) in the context of one disorder, therefore, these measures fall short in describing the complete set of motor alterations possible. As a result, motor domains of unknown pathobiology such as agitation and retardation have been commonly lumped together, even though they may be mediated by different mechanisms, as they tend to do in neurological disorders (DeLong and Wichmann, Reference DeLong and Wichmann2009). Further, as self-reports of psychomotor disturbance are subject to patient insight and bias, and observer-based rating scales are reliant on the most readily observable or impactful disturbances, it is also unclear if these behaviors change across the course of depression. This lack of rigorous measures for motor disturbances in depression is particularly surprising for a core feature of the disease, as motor markers have been highly predictive and intimately tied to disease and course in other disorders such a psychosis, autism, and dementia (Scarmeas et al., Reference Scarmeas, Hadjigeorgiou, Papadimitriou, Dubois, Sarazin, Brandt, Albert, Marder, Bell, Honig, Wegesin and Stern2004; Jansiewicz et al., Reference Jansiewicz, Goldberg, Newschaffer, Denckla, Landa and Mostofsky2006; Mittal and Walker, Reference Mittal and Walker2007; Asgari and Shafran, Reference Asgari and Shafran2010; Peralta and Cuesta, Reference Peralta and Cuesta2017).

Another limitation of prior work on motor disturbance in depression is the lack of a framework describing motor disturbances in depression, let alone a neurobiological model explaining these behaviors. The RDoC initiative proposes a framework consisting of five domains of constructs thought to be associated with psychopathology – negative valence, positive valence, cognitive systems, social processes, and arousal/regulatory systems – that can each be studied using multiple units of analysis (e.g. genes, molecules, cells, circuits, behavior, physiology, and self-report). For some of the constructs there is evidence of transdiagnostic principles guiding neural activation (McTeague et al., Reference McTeague, Huemer, Carreon, Jiang, Eickhoff and Etkin2017) or physiological reactions (Lang et al., Reference Lang, McTeague and Bradley2016) during tasks. However, an important domain that has been missing from the RDoC project is the motor domain, although NIMH is in the final stages of considering adding motor constructs to the RDoC framework (Bernard and Mittal, Reference Bernard and Mittal2015; Garvey and Cuthbert, Reference Garvey and Cuthbert2017; Mittal et al., Reference Mittal, Bernard and Northoff2017; Mittal and Wakschlag, Reference Mittal and Wakschlag2017). Studying the role of motor functioning (along with the other RDoC domains) has tremendous promise for yielding a more holistic understanding of depression. First, within the framework, distinct motor disturbances can be assessed dimensionally and inform on prevalence and course of motor disturbances in depression. Second, mapping these motor disturbances on brain circuits will prepare the field to develop specific treatment, e.g. using brain stimulation techniques. Finally, given the heterogeneity of depression (and the broader internalizing disorders in general), examining the specific role of motor symptoms can parse this heterogeneity, potentially identify more homogeneous subtypes, and yield novel and focal treatment targets that can be assessed with reliable and easily administered observational methods.

However, there are also significant challenges the motor symptoms domain will need to overcome. While other RDoC domains such as cognition had a tradition of testing multiple constructs with a battery of tests in the same individuals, the motor field has previously relied on investigations of single signs (e.g. tardive dyskinesia). Therefore, an RDoC motor domain will have the particular challenge of combining several motor subprocesses and integrating various very specific motor signs with unique definitions and referencing frameworks. On the other hand, instrumentation for the valid dimensional assessment of motor behavior is becoming increasingly available (van Harten et al., Reference van Harten, Walther, Kent, Sponheim and Mittal2017). An excellent example is the assessment of extrapyramidal signs in untreated psychosis, for which instrumental measures such as force variability or velocity scaling have much higher sensitivity and specificity than clinical rating scales (Cortese et al., Reference Cortese, Caligiuri, Malla, Manchanda, Takhar and Haricharan2005; van Harten et al., Reference van Harten, Walther, Kent, Sponheim and Mittal2017). The same improved sensitivity was reported in comparisons of instrumental measures v. rating scales in depression; 40% of depressed patients had motor disturbances according to clinicians’ ratings, but 60% according to instrumentation (Caligiuri and Ellwanger, Reference Caligiuri and Ellwanger2000). Likewise, computerized fine motor tasks reported motor dysfunction in both medicated and unmedicated depressed subjects, as well as changes in fine motor slowing after antidepressant pharmacotherapy (Sabbe et al., Reference Sabbe, Hulstijn, Van Hoof and Zitman1996a; Reference Sabbe, van Hoof, Hulstijn and Zitman1996b; Pier et al., Reference Pier, Hulstijn and Sabbe2004; Schrijvers et al., Reference Schrijvers, Maas, Pier, Madani, Hulstijn and Sabbe2009).

Another strength lending to an RDoC motor domain rests in the well-defined brain circuitry underlying normative motor behavior. This can serve as an anchoring point for the circuit-centric nature of the RDoC framework. Furthermore, as the field starts to understand motor-brain mechanisms, this may help to shed light on psychiatric illnesses where these motor behaviors occur. Notably, certain instrumental assessments map on to well-defined motor circuitry, which can be related to behavior or psychopathology (Mittal et al., Reference Mittal, Bernard and Northoff2017): (1) the cortico-basal ganglia circuit from primary motor cortex (M1) via caudate, putamen, pallidum, and subthalamic nucleus to the thalamic motor nuclei and back to M1, which controls excitation and inhibition of movements (Aron and Poldrack, Reference Aron and Poldrack2006; Obeso et al., Reference Obeso, Rodriguez-Oroz, Stamelou, Bhatia and Burn2014). The function of this circuit can be tested using actigraphy or force variability, which map on components of this circuit (Bracht et al., Reference Bracht, Schnell, Federspiel, Razavi, Horn, Strik, Wiest, Dierks, Muller and Walther2013; Mittal et al., Reference Mittal, Orr, Turner, Pelletier, Dean, Lunsford-Avery and Gupta2013). (2) The cerebello-thalamo-motor circuit, which combines M1, thalamus, and the cerebellar cortex. This loop is suggested to coordinate sensorimotor dynamics (Bostan et al., Reference Bostan, Dum and Strick2010; Bernard et al., Reference Bernard, Orr and Mittal2016). Therefore, postural sway is an applicable instrumental measure capturing the function of this circuit (Bernard et al., Reference Bernard, Dean, Kent, Orr, Pelletier-Baldelli, Lunsford-Avery, Gupta and Mittal2014). And finally, (3) the cortico-cortical motor circuit includes premotor and motor cortex, medial prefrontal cortex as well as posterior parietal cortex. This loop is thought to regulate movement organization and speed (Picard and Strick, Reference Picard and Strick1996; Chouinard and Paus, Reference Chouinard and Paus2006). Quantitative video analysis of complex behaviors such as hand gesture may monitor the function of this circuit (Dutschke et al., Reference Dutschke, Stegmayer, Ramseyer, Bohlhalter, Vanbellingen, Strik and Walther2018). Notably, aberrant functional connectivity in exactly these abovementioned motor circuits was demonstrated in psychosis patients with evident motor abnormalities and subjects at risk for psychosis with distinct motor behavior profiles (Walther et al., Reference Walther, Stegmayer, Federspiel, Bohlhalter, Wiest and Viher2017; Dean et al., Reference Dean, Walther, Bernard and Mittal2018). Given the well-defined functional neuroanatomy of the motor system, it should be possible to identify the neural mechanisms of altered motor behavior in depression.

How could motor disturbances in depression be conceptualized within the proposed RDoC motor domain? Motor disturbances could be tested as a dimension of depression; for example, psychomotor retardation could be quantified by the total amount of movement, allowing for changes within the course of depressive episodes. Indeed, using actigraphy, studies have linked individual differences in total movement with current mood states in depression (Razavi et al., Reference Razavi, Horn, Koschorke, Hugli, Hofle, Muller, Strik and Walther2011; Bewernick et al., Reference Bewernick, Urbach, Broder, Kayser and Schlaepfer2017; van Diermen et al., Reference van Diermen, Walther, Cools, Fransen, Birkenhager, Sabbe and Schrijvers2018). Motor retardation in terms of fine motor slowing can be assessed by analysis of key strike delay on smart phones, which have been shown to longitudinally predict mood ratings (Stange et al., Reference Stange, Zulueta, Langenecker, Ryan, Piscitello, Duffecy, McInnis, Nelson, Ajilore and Leow2018). Likewise, fine motor slowing was evidenced in drawing and writing tasks may separate subgroups of depression (Sabbe et al., Reference Sabbe, Hulstijn, van Hoof, Tuynman-Qua and Zitman1999; Caligiuri and Ellwanger, Reference Caligiuri and Ellwanger2000; Pier et al., Reference Pier, Hulstijn and Sabbe2004). Furthermore, continuous recordings of gross motor activity and smart phone typing behavior can easily be assessed outside of the laboratory, adding ecological validity to their measurement.

Reliable assessment of agitation seems to be much more difficult, but could be accomplished by focusing on continuous actigraphy data or frequency of typing errors on mobile devices (Walther et al., Reference Walther, Ramseyer, Horn, Strik and Tschacher2014). In the laboratory, measures such as force variability could be useful to assess agitation (Mittal et al., Reference Mittal, Orr, Turner, Pelletier, Dean, Lunsford-Avery and Gupta2013; Willems et al., Reference Willems, Sommer, Tenback, Koning and van Harten2016).

To further understand psychomotor disturbances in depression from an RDoC perspective, these instrumental dimensional measures could be measured longitudinally to determine whether these motor behaviors are stable or variable during the course of illness, i.e. tapping into state or trait characteristics of the disorder. Finally, to complement the lab-based and ecologically valid measures, new self-report instruments are needed assessing the subjective experience of agitation or retardation (beyond single items from scales such as the Hamilton Depression Rating Scale). Novel questionnaires assessing the subjective experience and everyday consequences of psychomotor retardation and agitation would be useful.

How would retardation or agitation in depression map on the cerebral motor circuits? Following the proposed function of the motor circuits, psychomotor retardation should be associated with structural and functional alterations in both the cortico-basal ganglia circuit and the cortico-cortical motor circuit. Preliminary evidence seems to support this view, demonstrating increased resting state perfusion in the motor cortex of patients with psychomotor retardation (Yin et al., Reference Yin, Wang, Wang, Xie, Zhang, Zhang, Zhang and Yuan2018). In addition, studies using actigraphy to assess gross motor activity reported an association of these dimensional measures with resting state perfusion and white matter properties in premotor cortices, external globus pallidus, and white matter motor tracts, all components of the cortico-basal-ganglia or cortico-cortical circuits (Bracht et al., Reference Bracht, Federspiel, Schnell, Horn, Hofle, Wiest, Dierks, Strik, Muller and Walther2012; Walther et al., Reference Walther, Hofle, Federspiel, Horn, Hugli, Wiest, Strik and Muller2012a, Reference Walther, Hugli, Hofle, Federspiel, Horn, Bracht, Wiest, Strik and Muller2012b; Cantisani et al., Reference Cantisani, Stegmayer, Bracht, Federspiel, Wiest, Horn, Muller, Schneider, Hofle, Strik and Walther2016). While these findings generally support the impact of the cerebral motor system in depression, the exact mechanism underlying psychomotor retardation still needs to be identified. In contrast, the neural correlates of agitation in depression have not been reported before. Lack of inhibitory control or functional imbalance in basal ganglia loops shifted toward motor excitation could be potential mechanisms, but that remains to be tested.

How could RDoC research on motor functioning in depression relate to treatment and research on other disorders? The proposed motor domain would allow for in depth dimensional assessment of motor behavior in depression. As these measures have been used with other disorders, these data can be combined with data on motor disturbance in other neuropsychiatric disorders such as schizophrenia, bipolar disorder, autism, or dementia in order to study the transdiagnostic nature of motor behavior (Bernard and Mittal, Reference Bernard and Mittal2015; Mittal et al., Reference Mittal, Bernard and Northoff2017). This would also help integrate motor behavior into the depressive syndrome cluster. The proposed RDoC motor domain has the potential to inform tests clarifying whether motor behavior has unique predictive value in the course and treatment response of depression. Current post-hoc analyses suggest that motor disturbances would indicate poor treatment response (Ulbricht et al., Reference Ulbricht, Dumenci, Rothschild and Lapane2016, Reference Ulbricht, Dumenci, Rothschild and Lapane2018; Sakurai et al., Reference Sakurai, Suzuki, Yoshimura, Mimura and Uchida2017). Again, instrumental measures are very likely to outperform self-rated motor disturbances and predictions could become even more precise (Caligiuri and Ellwanger, Reference Caligiuri and Ellwanger2000). Relatedly, in motor phenomena such as catatonia or Parkinsonism, researchers are increasingly focusing on the syndrome in various conditions, e.g. delirium, schizophrenia, depression, and bipolar disorder (Stuivenga and Morrens, Reference Stuivenga and Morrens2014; Wilson et al., Reference Wilson, Niu, Nicolson, Levine and Heckers2015). Along this line, the field could test whether psychomotor retardation and agitation in depression were a minor form of catatonia, a sign related to Parkinsonism, or phenomena distinct from catatonia and Parkinsonism. Furthermore, also brain imaging studies exploring motor disturbances in depression within this framework would be readily comparable with similar studies in other disorders. Given that similar mechanisms were involved across a spectrum of psychiatric disorders, specific treatments for motor inhibition or motor excitation could be tested and offered independent of the underlying psychiatric disorder. For example, studies could use noninvasive brain stimulation to target aberrant neural activity within cortical motor areas (Walther et al., Reference Walther, Stegmayer, Federspiel, Bohlhalter, Wiest and Viher2017) in order to ameliorate motor disturbance in depression. In sum, studying motor functioning in depression with an RDoC lens holds a great deal of promise for enhancing understanding (and ultimately treatment) for this important feature of depression.