Clinical dissociation and electromyographic convergence of paratonia in cognitive impairment including Alzheimer's disease: Evidence for a shared central inhibitory mechanism

Patients and healthy subjects

Patients with Alzheimer's disease (AD), mild cognitive impairment (MCI), and vascular dementia (VaD) were recruited from the outpatient clinic for cognitive disorders at the Geriatric Memory Clinic, Geriatric Clinic, University Hospital “IRCCS Ospedale Policlinico San Martino”, Genova, Italy.

The diagnosis of AD was made according to the National Institute of Neurological and Communicative Disorders and Stroke (NINCDS) Alzheimer Disease and Related Disorders Association (ADRDA) criteria. ¹⁷ Clinical diagnosis of MCI was made according to the revised Petersen criteria. ¹⁸ The diagnosis of VaD was made according to the International Society for Vascular Behavioral and Cognitive Disorders criteria. ¹⁹

A normal neurological examination was required for all recruited subjects, who should not present with coarse motor or sensory deficits, involuntary movements, ataxia, spasticity or signs suggesting parkinsonism.

Healthy subjects were recruited from caregivers and unit staff. All participants or their legal guardians gave informed consent for all study procedures, which were conducted in accordance with the Declaration of Helsinki. The study was approved by the “Comitato Etico Territoriale – Liguria” (15/2025 - id 14261).

Clinical evaluation of paratonia

In all participants (both cognitive impaired patients and healthy subjects), paratonia was assessed in the flexor/extensor muscles of the elbow on the non-dominant side using the Paratonia Scale and the modified Kral Procedure (m-Kral-P). ¹² During the assessment, instructions were given for the subjects to relax completely.

The Paratonia Scale was used to rate both facilitatory Paratonia (Facilitatory Paratonia Scale, FacPS) and oppositional paratonia (Oppositional Paratonia Scale, OppPS). Scoring is as follows: 0 = no paratonia; 1 = trace paratonia (minimal assistance or resistance offered to passive movement); 2 = moderate paratonia (moderate assistance or resistance to passive movement; 3 = severe paratonia (marked assistance or resistance to passive movement); 4 = extreme paratonia (full assistance or resistance to passive movement).

In the modified Kral Procedure (m-Kral-P), with the subject seated, the examiner performed three cycles of passive flexion-extension of the subject's forearm. At the end of the third cycle, the subject's limb was released at the level of the subject's thigh and the examiner's hand was withdrawn. Scoring is as follows: 0 = no movement; 1 = the forearm flexes, but not enough to lift the hand from the leg; 2 = the forearm flexes, lifting the forearm away from the leg but less than halfway to full flexion; 3 = the forearm flexes at least halfway, but not fully; 4 = the forearm flexes completely or repeated cycles of flexion and extension occur. ¹²

Mental status assessment

In all participants (both cognitive impaired patients and healthy subjects), mental status was assessed using the Mini-Mental State Examination (MMSE) with the score adjusted for age and educational level. ²⁰ Scores range from 0 to 30, with higher scores indicating less cognitive impairment.

EMG assessment of paratonia

Surface preamplified electrodes with fixed inter-electrode spacing (TSD150B, Biopac Systems Inc, USA) were placed over the muscle belly of biceps and triceps brachii of the left side according to SENIAM guidelines. ²¹ The distance between the electrodes placed on biceps and triceps muscles was maximized to reduce cross talk. ¹⁵ The elbow joint angle was monitored with an electronic goniometer (TSD130B, Biopac Systems Inc, USA) placed over the joint. All signals were recorded by a Biopac MP100 unit (Biopac Systems Inc, USA) for offline analysis.

Participants and the examiner repeated a previously described experimental paradigm.^6,22 Briefly, the examiner applied to the participant repeated consecutive passive elbow flexion-extension movements (sinusoidal continuous movements) paced by consecutive metronome tones, such that maximal elbow flexion and subsequent maximal elbow extension positions corresponded to two consecutive metronome beats. Changing the number of beats per minute (BPM) resulted in different passive movement velocities. To obtain non-sinusoidal, discontinuous movements (also called linear movements), the examiner waited a few metronome beats (randomly 1 to 4) at the maximal flexion or extension position before performing the upcoming movement at the same speed determined by BPM. The following 4 blocks of 15 consecutive flexion-extension movements were collected: 1) 15 sinusoidal flexion-extension movements at 60 BPM; 2) 15 sinusoidal flexion-extension movements at 100 BPM; 3) 15 linear flexion-extension movements at 60 BPM; 4) 15 linear flexion-extension movements at 100 BPM. Blocks were randomized for each participant who received a total of 60 flexion-extension movements.

EMG analysis: FacEMG score and OppEMG score

After visual inspection of raw data to eliminate artefacts, EMG data were filtered (band-pass 20–250 Hz) and rectified. For each of the 15 flexion-extension movements of a block, EMG activity of the flexion phase (from the point of maximal extension to that of maximal flexion) and EMG activity of the extension phase (from the point of maximal flexion to that of maximal extension) were measured in both muscles using the “mean EMG” function of the AcqKnowledge software (version 4.2, Biopac Systems Inc, USA).

During the flexion phase, the activity of the biceps was considered facilitatory (FacEMG biceps), while that of the triceps oppositional (OppEMG triceps). During the extension phase, the opposite was true: biceps activation was considered oppositional (OppEMG biceps) and triceps activation was considered facilitatory (FacEMG triceps). Considering all 4 blocks, 60 FacEMG biceps, 60 FacEMG triceps, 60 OppEMG biceps and 60 OppEMG triceps were measured in each subject. In each subject, the average of the 60 FacEMG biceps + the 60 FacEMG triceps was calculated to obtain an overall FacEMG score. Similarly, the mean of the 60 OppEMG biceps + the 60 OppEMG triceps was calculated to obtain an overall OppEMG score.

Statistics

Spearman correlation coefficients were calculated to assess the relationship between pairs of variables. The strength and direction of these correlations were interpreted according to the Rule of Thumb for Interpreting the Size of a Correlation Coefficient.^23,24 Specifically, coefficients between 0.00 and 0.30 (or −0.30 and 0.00) were considered negligible; between 0.30 and 0.50 (or −0.50 and −0.30) as low; between 0.50 and 0.70 (or −0.70 and −0.50) as moderate; between 0.70 and 0.90 (or −0.90 and −0.70) as high; and between 0.90 and 1.00 (or −1.00 and −0.90) as very high.

To evaluate the ability to discriminate cognitively impaired participants (AD, MCI, vascular) from healthy controls, receiver operating characteristic (ROC) curve analyses were conducted for FacEMG, OppEMG, FacPS, and OppPS. The area under the ROC curve (AUC) was calculated as a global measure of discriminative performance for each parameter. ²⁵ The discriminative performance of ROC analyses was interpreted according to commonly used criteria: an AUC of 0.50 indicating no discriminative ability, values between 0.60 and 0.70 indicating modest discrimination, values between 0.70 and 0.80 indicating good discrimination, and values above 0.80 indicating very good discrimination.^25–27 Differences between correlated AUCs were formally tested using DeLong's test.

All statistical analyses were performed using Stata (version 18; StataCorp, College Station, TX, USA) and R (version 4.2.1). A two-sided p-value < 0.05 was considered statistically significant.

Subjects

Eighty-six subjects (31 males, 55 females) were included in the study, with a median age of 79 years (interquartile range [IQR] 75–82.5). Forty-six participants (53.5%; 14 males, 32 females) had cognitive impairment (18 AD, 21 MCI, and 7 VaD), while the remaining 40 (46.5%; 17 males, 23 females) participants were cognitively healthy subjects.

Associations between paratonia measures and cognitive status

According to Spearman's rank correlation analysis, MMSE showed a low negative correlation with FacPS (ρ = −0.37, p < 0.001) and with m-Kral-P (ρ = −0.33, p = 0.002), whereas the association between MMSE and OppPS was negligible (ρ = −0.24, p = 0.031).

In contrast, MMSE showed a moderate negative correlation with both FacEMG (ρ = −0.51, p < 0.001) and OppEMG (ρ = −0.54, p < 0.001) (Figure 1).

OPEN IN VIEWER

Figure 1.

Correlation between paratonia measures and cognitive status. Scatter plots illustrating the association between Mini-Mental State Examination (MMSE) scores and clinical paratonia scales (FacPS, OppPS, and m-Kral-P) as well as EMG-derived paratonia measures (FacEMG and OppEMG). Individual data points are color-coded according to clinical group (Control, MCI, AD, and VD). The solid black line represents the overall linear regression trend across all participants, while the surrounding gray shaded area indicates the 95% confidence interval of the regression estimate. The bubble plots shows Spearman correlation coefficients (larger circles indicate stronger associations).

Receiver operating characteristic analysis

FacEMG, OppEMG, and FacPS showed good overall discriminative performance in distinguishing cognitively healthy subjects from patients with cognitive impairment. Specifically, the AUC was 0.78 for OppEMG, 0.76 for FacEMG, and 0.76 for FacPS. In contrast, the discriminative accuracy of OppPS and m-Kral-P was modest, with AUC values of 0.63 and 0.70, respectively (Figure 2).

OPEN IN VIEWER

Figure 2.

Discriminative performance of paratonia measures. Receiver operating characteristic (ROC) curves illustrating the ability of clinical and EMG-based paratonia measures to discriminate cognitively healthy controls from patients with cognitive impairment. OppEMG, FacEMG, and FacPS showed good overall discriminative performance, whereas OppPS and m-Kral-P exhibited lower accuracy.

Pairwise comparisons of ROC curves revealed that OppEMG showed higher discriminative performance than OppPS (ΔAUC = 0.15, z = 2.16, p = 0.031). No other pairwise comparisons reached statistical significance.

From clinical dissociation to EMG convergence

When paratonia was assessed using clinical scales, a clear distinction emerged between the facilitatory and oppositional components. Oppositional paratonia, as clinically assessed (OppPS), showed a weaker association with global cognitive impairment than measures targeting the facilitatory component (FacPS and m-Kral-P), in terms of both correlation with cognitive status (Figure 1) and discriminative ability between cognitively healthy subjects and patients with cognitive impairment (Figure 2). Overall, these findings suggest that facilitatory paratonia is more closely related to cognitive decline and provides better discrimination than the oppositional component.

This pattern is consistent with previous observations by Beversdorf and Heilman, who reported stronger associations between cognitive status and facilitatory paratonia than oppositional paratonia.

However, this dissociation is no longer evident when paratonia is assessed using EMG. When muscle activity is quantified directly, both facilitatory (FacEMG) and oppositional (OppEMG) paratonia show comparable associations with cognitive status (Figure 1) and similar discriminative performance (Figure 2). This suggests that the clinical distinction between facilitatory and oppositional paratonia does not reflect a true pathophysiological separation but rather arises from limitations inherent to the clinical evaluation of muscle tone.

Biomechanical contributions to clinical oppositional paratonia

The reduced cognitive specificity of clinically assessed oppositional paratonia (OppPS) can be explained by the contribution of non-neural mechanical factors to the resistance perceived during manual examination. In OppPS, the force encountered by the examiner reflects not only centrally driven involuntary muscle activation, but also the passive stiffness of the muscle–joint system, which largely operates independently of cognitive control mechanisms.

Consequently, OppPS is susceptible to contamination by non-neural components of resistance, resulting in reduced specificity with respect to cognitive decline. In contrast, OppEMG directly quantifies involuntary muscle activation during passive movement and is unaffected by passive mechanical resistance. This provides a more physiologically specific estimate of centrally mediated paratonia. This methodological difference offers a plausible explanation for the stronger association with cognitive status and the improved discriminative performance observed for OppEMG compared to OppPS.

Facilitatory paratonia as a low-threshold marker of cognitive impairment

The stronger association between facilitatory paratonia, as clinically assessed, and cognitive status observed in the present study probably reflects the fact that the assistive force produced involuntarily by the patient during passive mobilization is easier and more reproducible to assess than resistive force. Facilitatory paratonia involves the involuntary activation of muscles in the direction of passive movement (i.e., muscle shortening), which cannot be affected by passive mechanical resistance.

In contrast, the responses encountered during manual examination that are categorized as oppositional may be influenced by several non-neural factors, including physiological muscle stiffness, which varies between individuals and muscle groups, as well as intrinsic muscle alterations such as fibrosis and shortening. Consequently, clinicians are often required to distinguish between true paratonic resistance and mechanically driven opposition, particularly when only minimal resistance is present. This interpretative ambiguity is largely absent in the case of facilitatory paratonia, where even subtle involuntary motor activation in the direction of movement is more readily attributable to paratonia.

As a result, clinically assessed facilitatory paratonia may represent a more sensitive and earlier detectable behavioral marker of frontal inhibitory dysfunction in patients with cognitive impairment.

Limitations

A limitation of the present study is the heterogeneity of the clinical sample, which included patients with different etiologies of cognitive impairment (AD, MCI, and VaD). Although this reflects real-world clinical populations, disease-specific mechanisms may differentially influence paratonia patterns, and future studies in larger and more etiologically homogeneous cohorts are needed to confirm the generalizability of the present findings.

A potential source of apparent inconsistency between the present findings and our previous work warrants clarification. In our earlier study, in which we first characterized the electromyographic and clinical features of paratonia across cognitively healthy and impaired individuals, both facilitatory and oppositional components were found to be more prevalent and more severe in cognitively impaired patients compared to healthy subjects. ⁶ In that context, oppositional paratonia showed higher values particularly in patients with AD, although without a clear separation from the mild cognitive impairment group, while facilitatory paratonia was already prominent in earlier stages of cognitive decline (MCI). At first glance, these findings may appear partially discrepant with the present observation that clinically assessed oppositional paratonia shows a weaker association with global cognitive status compared to facilitatory paratonia. However, this apparent discrepancy can be explained by the use of different methods of analysis. The previous study was based on comparisons between diagnostic categories, capturing differences in average paratonia severity across groups, whereas the present study evaluates the relationship between paratonia measures and cognitive status at the individual level, using MMSE as a continuous variable. These approaches are not equivalent and may yield partially divergent results, as differences observed between groups do not necessarily translate into strong associations at the individual level. ²⁸ In this sense, the present findings should be interpreted as complementary rather than contradictory to our previous observations.

Conclusions

This study provides two main findings with distinct clinical and pathophysiological implications. From a clinical perspective, our results indicate that facilitatory paratonia assessed with standard bedside scales represents the most informative clinical marker of cognitive impairment, showing stronger associations with global cognitive status and better discriminative ability between cognitively healthy subjects and patients with cognitive decline. These findings support the use of facilitatory paratonia as a practical and sensitive behavioral indicator of frontal-related cognitive dysfunction in routine clinical settings.

From a pathophysiological perspective, however, the apparent clinical superiority of facilitatory paratonia should not be interpreted as evidence of a true biological dissociation between facilitatory and oppositional components. The convergence observed when paratonia is assessed using EMG-based measures indicates that both motor components reflect different behavioral expressions of a shared central inhibitory deficit. Together, these results highlight the critical role of measurement methodology in shaping the clinical interpretation of paratonia and support a unified neurophysiological framework underlying facilitatory and oppositional paratonia.