Acute onset movement disorders in diabetes mellitus: A clinical series of 59 patients

Abstract Background and purpose No previous study has assessed the frequency and clinical–radiological characteristics of patients with diabetes mellitus (DM) and acute onset nonchoreic and nonballistic movements. We conducted a prospective study to investigate the spectrum of acute onset movement disorders in DM. Methods We recruited all the patients with acute onset movement disorders and hyperglycemia who attended the wards of three hospitals in West Bengal, India from August 2014 to July 2021. Results Among the 59 patients (mean age = 55.4 ± 14.3 years, 52.5% men) who were included, 41 (69.5%) had choreic or ballistic movements, and 18 (30.5%) had nonchoreic and nonballistic movements. Ballism was the most common movement disorder (n = 18, 30.5%), followed by pure chorea (n = 15, 25.4%), choreoathetosis (n = 8, 13.6%), tremor (n = 5, 8.5%), hemifacial spasm (n = 3, 5.1%), parkinsonism (n = 3, 5.1%), myoclonus (n = 3, 5.1%), dystonia (n = 2, 3.4%), and restless leg syndrome (n = 2, 3.4%). The mean duration of DM was 9.8 ± 11.4 years (89.8% of the patients had type 2 DM). Nonketotic hyperglycemia was frequently (76.3%) detected. The majority (55.9%) had no magnetic resonance imaging (MRI) changes; the remaining showed striatal hyperintensity. Eight patients with MRI changes exhibited discordance with sidedness of movements. Most of the patients (76.3%) recovered completely. Conclusions This is the largest clinical series depicting the clinical–radiological spectrum of acute onset movement disorders in DM. Of note was that almost one third of patients had nonchoreic and nonballistic movements. Our findings highlight the importance of a capillary blood glucose measurement in patients with acute or subacute onset movement disorders, irrespective of their past glycemic status.

There is a knowledge gap, as no previous study has assessed the frequency and clinical-radiological characteristics of patients with DM and acute onset movement disorders, including nonchoreic and nonballistic movements. A study on this topic would advance our knowledge of these conditions.
We conducted a prospective study to investigate the spectrum of acute onset movement disorders associated with DM in patients attending the wards of three Indian hospitals. Data on demographics, numerous comorbidities, laboratory values, neuroimaging, outcomes, and treatment were analyzed. This study, comprising 59 patients, is the largest observational study depicting the clinicalradiological spectrum of acute onset movement disorders associated with DM.

ME THODS
All patients with acute onset movement disorders who attended the wards of the Burdwan Medical College and Hospital (Neurology Superspecialty and Internal Medicine Wings), Bangur Institute of Neurosciences, and Berhampore Mental Hospital, in West Bengal, India, between 1 August 2014 and 1 July 2021, were screened with random capillary blood glucose, followed by confirmation with fasting blood glucose, postprandial blood glucose, and glycated hemoglobin according to the guidelines of the American Diabetes Association [17].
History of head injury, visual disturbances, altered sensorium, drowsiness, headache, convulsions, fever, addiction, DM and complicating comorbidities, similar illnesses in family members, atherosclerotic cardiovascular disease, seizures, and drug therapy (current prescription medications) was obtained.
A thorough neurological examination (higher cognitive functions, cranial nerves, fundoscopy, motor, sensory, and autonomic systems, reflexes, cerebellum, meningeal signs, and gait) and a review of other systems were performed. The semiology of the involuntary movements (evaluated in resting state and postural and action-induced whenever applicable) was also noted and classified accordingly as choreic or ballistic (pure chorea, choreoathetosis, and ballism), and nonchoreic and nonballistic movements (myoclonus, dystonia, hemifacial spasm, restless leg syndrome, tremor, and parkinsonism) TA B L E 1 Previous selected clinical series and systematic reviews of acute onset movement disorders in diabetes mellitus Patients with seizures or those with stroke, demyelination, structural intracranial lesion (other than striatopathy), pregnancy, major metabolic perturbation not related to hyperglycemia, history of preexisting movement disorders, and any other pre-existing neurological disorders were excluded ( Figure 1).
Patients with movement disorders were documented and followed up for at least 3 weeks. Those with complete resolution of movement disorders were assigned to good recovery, whereas those with persisting movements were considered incomplete recovery.
Fully recovered patients who were free from all movement disorders within 7 days were regarded as early responders, whereas those who recovered after 7 days were considered late responders. Because of the large number of statistical tests performed, we used the Benjamini-Hochberg procedure with a defined false discovery rate of 5% [20].

RESULTS
Of the 552 patients with acute onset movement disorders who attended the wards during the recruitment, 96 patients were diagnosed as diabetic ( Figure 1). Of those 96 diabetic patients, 29 were excluded, as they had coexisting other pathologies (20 with seizure disorders, four with stroke, three with a previous history of movement disorders, one with associated demyelination, and one with a space-occupying lesion). In addition, eight patients with epilepsia partialis continua were also excluded from the present study, as it is not categorized as a movement disorder conventionally; rather, it is categorized as a type of simple focal motor status epilepticus in which frequent repetitive muscle jerks, usually arrhythmic, continue over prolonged periods [21]. Moreover, these patients with epilepsia partialis continua had electroencephalographic changes. That is why they were excluded from the study.
Therefore, 59 diabetic patients with acute movement disorders were finally included in the study (   Table 3).  Table 4).

DISCUSS ION
Unlike previous clinical series and systematic reviews ( Table 1) [1-10], the current study also included patients with other movement disorders. Interestingly, almost one third of movement disorders were nonchoreic and nonballistic, which is not the common perception.
This underpins the importance of measuring capillary blood glucose immediately with any acute onset movement disorders, including those not classically described in diabetic striatopathy [13][14][15][16].
In a smaller case series from South India comprising 11 cases of diabetic chorea-ballism, the mean age was 58.7 years [7], a value much closer to the current study. Compared to most previous studies [1,3,6,8,10], our sample was younger. Whether ethnicity plays any role in this needs further exploration.
A clear predominance of women among patients with hyperglycemic choreic and ballistic movements has been observed across different studies (see Table 1  be indicative of gender bias at tertiary health care facilities and prevailing sex inequality in access to optimum health care [22]. Similar to previous studies [1], the vast majority (n = 53) of the patients in this study were diagnosed with type 2 DM. We observed high capillary blood glucose (mean 419.6 mg/dl) at presentation, suggestive of extremely poor recent control of glycemic status, a finding similar to other studies [1,3]. In our series, the mean duration of DM was 9.8 years, with three cases (5.1%) in whom movement disorders heralded previously undiagnosed DM. In contrast, Chua et al. [1] found that 17% of all reported diabetic striatopathy cases had previously undiagnosed DM. This discrepancy could be due to publication bias, as cases with a rare presentation of a common disease tend to be readily published.
Importantly, in our study, 48 (83.1%) patients had associated stigmata of diabetic microangiopathy (retinopathy, neuropathy, and nephropathy), suggesting long-standing poorly controlled DM [14,23,24]. However, acute disturbance of glycemic status may play a more important and consistent role than glycemic durability in movement disorder generation [25].
The majority of patients (76.3%) were negative for ketone bodies.
In the setting of nonketotic hyperglycemia, brain metabolism shifts to the alternative anaerobic pathway in the Krebs cycle, resulting in depletion of γ-aminobutyric acid (GABA), an inhibitory neurotransmitter [1,9,26]. This leads to attenuated inhibition of the thalamus by the medial globus pallidus, resulting in choreic movements in diabetic patients [1,9,26]. On the other hand, GABA can readily be resynthesized from acetoacetate in the ketotic milieu [1,12,14]. Hence, in the ketotic state, GABA level does not usually decrease too low to cause hyperkinetic movements unless other mechanisms, such as cerebrovascular insufficiency or ultrastructural changes in basal ganglia, are present [1,12,14].
In this study, most patients (55.9%) were devoid of any neuroradiological lesion. These data support that the presence of acute onset movement disorders with concurrent hyperglycemia warrants urgent management without waiting for neuroimaging [8]. The lagging of neuroradiological changes behind the clinical manifestations could be due to hyperglycemia-induced ultrastructural changes that occur much earlier than frank structural changes appreciable on MRI [1].
As in previous studies [1][2][3][4][5]7,8,9,10,23,26], our results confirm that, if neuroimaging is abnormal, putamen with or without caudate involvement was the most commonly affected area. Isolated caudate nucleus, globus pallidus, and rarely, subthalamic nucleus involvement have also been previously reported [1][2][3][4][5]7,8,9,10,23,26]; in our series, The underlying cause of clinical-radiological discordance, which defies the classical concept of neurological localization, needs further study and may be better studied with newer neuroimaging modalities, such as functional MRI [30]. This discordance may be a consequence of widespread involvement of brain circuitries and affliction at ultrastructural levels of neural substrates apart from what is observed in conventional neuroimaging [30]. Treatment with insulin alone helps alleviate abnormal movements, although the role of adding on other medications cannot be ignored.

CO N FLI C T O F I NTE R E S T
None of the authors has any conflict of interest to disclose.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are included in the article.