Evolving Concepts on Bradykinesia

Matteo Bologna; Giulia Paparella; Alfonso Fasano; Mark Hallett; Alfredo Berardelli


Brain. 2020;143(3):727-750. 

In This Article


The term bradykinesia has historically been used to indicate slowness in the execution of voluntary movement in patients affected by Parkinson's disease. Clinical and experimental studies, however, have suggested the use of the term bradykinesia to encompass other motor abnormalities, including low amplitude movement (hypokinesia), absence of movement (akinesia) and amplitude reduction during movement repetition (sequence effect) (Berardelli et al., 2013; Postuma et al., 2015; Schilder et al., 2017). Neurophysiological investigations indicated that slowness of movement and reduced movement amplitude rely on different mechanisms than those underlying the sequence effect (Espay et al., 2009, 2011; Wu et al., 2016; Bologna et al., 2018). The various neurophysiological abnormalities are not always present together in the same patient (Kang et al., 2011; Lee et al., 2014; Bologna et al., 2016a; Wu et al., 2016) and they not necessarily involve all the body segments. Bradykinesia features also vary according to the disease stage (less frequent sequence effect in advanced Parkinson's disease) (Bologna et al., 2016a) and the different types of parkinsonism. Some evidence suggests that the lack of sequence effect may help distinguishing PSP from Parkinson's disease and MSA (Ling et al., 2012; Djurić-Jovičić et al., 2016). Further compelling evidence for lack of unity is the differential effects that levodopa exerts on the various bradykinesia features. Levodopa improves slowed and reduced amplitude movements (Hallett et al., 1977; Berardelli et al., 1986; Castiello and Bennett, 1997; Alberts et al., 2000; Castiello et al., 2000; Espay et al., 2009, 2011) but not the sequence effect (Agostino et al., 2003; Kang et al., 2010, 2011; Bologna et al., 2016a, 2018; Wu et al., 2016). Levodopa may exert different modulation on basal ganglia circuits and on their interaction with cortical and cerebellar areas underlying the various motor abnormalities. DBS also improves bradykinesia, but its effect is more complex as it depends on exact electrode location, specific adopted frequency and volume of tissue activated (Fasano and Lozano, 2014; di Biase and Fasano, 2016; Strotzer et al., 2019).

In Parkinson's disease, earlier studies have demonstrated abnormalities in basal ganglia firing rate (Wichmann and DeLong, 1996), firing pattern (Ellens and Leventhal, 2013) and oscillatory activity, with an increase of beta band power in the GPi and STN nuclei (Brown et al., 2001; Little et al., 2012; Torrecillos et al., 2018), suggesting a key role of basal ganglia. Disrupted basal ganglia plasticity has been also pointed out as a possible mechanism of bradykinesia (Yttri and Dudman, 2018; Milosevic et al., 2019). More recent experimental studies now indicate that mechanisms other than basal ganglia may be involved in the pathophysiology of bradykinesia in Parkinson's disease. These include excitability and plasticity changes of M1 and other non-primary motor areas (Ridding et al., 1995; Berardelli et al., 1996b; Sabatini et al., 2000; Buhmann et al., 2004; Mir et al., 2005; Wu and Hallett, 2005; Lyoo et al., 2007; Suppa et al., 2010, 2011; Bologna et al., 2016d, 2018; Hu et al., 2017). Recent evidence has demonstrated a correlation between M1 excitability changes and motor slowness, and between M1 plasticity abnormalities and the sequence effect (Bologna et al., 2018). Whether M1 abnormalities are primary mechanisms or, alternatively, compensatory adaptations to the disease still need to be clarified (Ni et al., 2013; Blesa et al., 2017). Hyperactivity of cortical motor areas has been also described and interpreted as a possible compensatory mechanism for the defective basal ganglia function (Rascol et al., 1997; Sabatini et al., 2000; Thobois et al., 2000; Ceballos-Baumann 2003; Wu and Hallett, 2005), although it has been argued that hyperactivation could be related to other motor abnormalities, such as rigidity and levodopa-induced dyskinesias (Ridding et al., 1995; Kleine et al., 2001; Pierantozzi et al., 2001; Yu et al., 2007). Noteworthy, part of the beneficial effects of DBS resides within its decoupling property within these hyperactivated networks (de Hemptinne et al., 2015, Tinkhauser et al., 2017). The cerebellum may also be involved in the pathophysiology of bradykinesia. As the activation of the cerebello-thalamo-cortical circuit increases with Parkinson's disease progression severity (Wu et al., 2009, 2010; Sen et al., 2010), the cerebellum probably compensates for the defective activity in basal ganglia and M1 activation (Cerasa et al., 2006; Yu et al., 2007; Wu and Hallett, 2013; Fasano et al., 2017). An alternative possibility is that the increased cerebellar activity in Parkinson's disease reflects a primary pathophysiological change, i.e. failure to inhibit inappropriate basal ganglia outflow (Turner et al., 2003; Grafton et al., 2006). Changes at the level of sensory cortical areas are also involved in generating bradykinesia. The abnormal sensorimotor processing, as demonstrated by tactile sensory discrimination studies (Rocchi et al., 2013; Conte et al., 2018), suggested that the normal filtering of sensory information exerted by basal ganglia is lost. How this translates to bradykinesia is unknown, but it is likely that an abnormal sensory integration is also present at the level of sensory cortical area.

All these data indicate that bradykinesia should be interpreted as the result of a network dysfunction, including basal ganglia, sensorimotor cortical areas, and the cerebellum, rather than a consequence of one single system default. The role of each structure in determining bradykinesia is still unclear. One hypothesis is that basal ganglia are responsible for aiding the joint processes of movement selection and inhibition, through the direct and indirect pathways. Recent evidence suggests that the basal ganglia-cortical loops in a parallel fashion contribute to different functions, including speed within the motor loop. The energy for these functions appears to come from dopamine. With dopamine loss, the system slows down, reaction times increase, movement speed and amplitude decrease. A signature of this loss is the increase of beta oscillation in the indirect pathway. The scaling problems observed in patients with bradykinesia may arise at least in part from a fault in the sensory loop, responsible of sensorimotor integration. This function involves cerebellar structures, which may play a role in movement feedback, particularly important for continued and repetitive movements. Other cerebellar components, involved in motor loops, try then to compensate. If they cannot fully compensate, then the sequence effect develops. Other compensatory mechanisms could finally involve primary and non-primary motor cortex functional changes. When interpreted in a network perspective, distinct bradykinesia features are likely to be mediated by a variable involvement of the various nodes in the network (Figure 1).

Figure 1.

The Network hypothesis for bradykinesia pathophysiology. Basal ganglia dysfunction is responsible for altered movement selection. The altered movement scaling at basal ganglia level may also arise from a fault in the sensory loop, responsible for sensorimotor integration. Other mechanisms involve sensorimotor areas. Cerebellar structures likely play a role in movement feedback, particularly important for continued and repetitive movements. The energy for all basal ganglia, sensorimotor areas, and the cerebellar functions appear to come from dopamine. With dopamine loss, the system slows down, reaction times increase, movement speed and amplitude decrease, and the sequence effect develops. Dotted lines indicate possible compensatory mechanisms.

Neurophysiological insight is relevant to improve the appropriateness of the terminological use of bradykinesia in Parkinson's disease and atypical parkinsonism. Features of bradykinesia are slowness of voluntary movements together with a decrement in amplitude or speed during repetitive or continued movements (sequence effect), small (i.e. under-scaled, low amplitude) movements (hypokinesia), and loss of spontaneous/automatic movements or absence of movement or difficulty initiating a movement such as in freezing, motor blocks, or hesitations (akinesia). In this perspective, we suggest that the terms bradykinesia (slowness of movement), hypokinesia (reduced amplitude of movement), akinesia (lack of movement), and sequence effect (progressive decrement of amplitude and velocity of movement) should be defined separately in the description of clinical phenotypes, saving the Greek origin of the words, rather than encompass them in one word. This is supported by the evidence that each single abnormality has a specific pathophysiological mechanism and can occur in isolation in one given patient.

Understanding the features of bradykinesia has very important implications when it comes to enrolment criteria for research studies and, more importantly, diagnostic criteria of parkinsonism, e.g. in the differential diagnosis with pyramidal slowness or functional hypokinesia (Thenganatt and Jankovic, 2016), Huntington disease (Berardelli et al., 1999), cerebellar disorders (Manto et al., 2012), and other neurological conditions in which a slowness of movement is present. Specific pathophysiological mechanisms likely explain the pathophysiology of bradykinesia which is present in the various pathological conditions. Increasing insight into bradykinesia pathophysiology in Parkinson's disease, atypical parkinsonisms and other movement disorders, will serve as a new starting point for clinical and experimental purposes.