Surface electromyography as a novel screening tool for early cognitive impairment: Exploring motor-cognitive interactions

Abstract

Background

Motor changes precede the emergence of cognitive impairment (CI), and Alzheimer's disease is the most common type of CI.

Objective

The study aimed to investigate the characteristics of surface electromyography (sEMG) in patients with CI and explore the ability of sEMG for CI detection.

Methods

639 participants were enrolled, including 284 patients with CI, and 355 controls. A series of motor and cognitive assessment were conducted. The sEMG examination under single and dual-task patterns were performed.

Results

The CI group exhibited significantly decreased motor function, including muscle bulk, swallowing, balance, and gait. Significant alterations of 20 sEMG features were detected in CI group compared to controls (p < 0.05), among them, the integrated electromyography feature of right gastrocnemius lateralis in dual-task was positively associated with plasma p-tau181 (r = 0.21, p = 0.017). Besides, the distinctive sEMG features were significantly correlated with cognitive scores (p < 0.05). Furthermore, the combination of single and dual-task sEMG model can effectively distinguish CI from controls with an area under the curve (AUC) of 0.849, and the AUC reached 0.948 after combining with p-tau181.

Conclusions

Motor dysfunction is common in patients with CI, and sEMG can serve as an effective tool for CI screening.

Keywords

Alzheimer's disease biomarkers cognitive impairment diagnostic screening dual-task assessment early detection motor disorders motor-cognitive interaction surface electromyography

Introduction

Dementia poses a serious social burden globally, and the number of individuals with dementia is expected to rise to 153 million worldwide by 2050 as the population ages.¹ Alzheimer's disease (AD) and vascular dementia are the most common causes of cognitive impairment (CI) in China.² AD is the most common type of dementia in the elderly population, characterized by extracellular amyloid-β (Aβ) deposition and intracellular tau protein hyperphosphorylation (p-tau). In fact, the pathologic alterations in AD occur 15–20 years before clinical symptoms arise.^3,4 Therefore, early diagnosis is particularly critical for achieving early intervention and treatment.

Multiple studies have demonstrated that motor disorders preceded the emergence of cognitive impairment by more than 10 years.^5–7 Besides, motor disorder could predict the onset of CI.⁸ In this study, motor disorders refer to the movement disorders that accompany cognitive impairment diseases, rather than those caused by other diseases that explicitly lead to movement disorders. Multiple studies have shown that movement disorders are common in patients with cognitive impairment, manifested in mild stages such as decreased grip strength⁹ and slowed pace.¹⁰ In the moderate to severe stage, it manifests as dysphagia, balance disorders, etc.¹¹ Assessment of motor function seems to be a viable method for early screening of CI. There has been some research on the motor assessment in dementia patients, including gait and grip strength assessment.^9,12 A longitudinal study found that slow gait speed could predict 10-year cognitive decline in Digit Symbol Substitution Test, while low grip strength could predict 10-year cognitive decline in Mini-Mental State Examination (MMSE) and Digit Symbol Substitution Test.¹³ Gait variability correlated with cognitive function and changed with cognitive decline.¹⁴ Furthermore, Aβ deposition in the brain has been found to be associated with slower gait speed in AD patients.¹⁰ Our previous findings from a large community cohort also showed that combining dual-task gait and dual-task eye-tracking analysis was feasible for detecting CI.¹⁵ However, gait examination has several limitations. To begin, it may be unable to detect tiny changes in muscle activity and functional status. Moreover, the massive size of the detection equipment is not suitable for large-scale community screening.

Surface electromyography (sEMG) is a technique for recording and quantitatively analyzing electrical activity signals of muscle motor units using surface electrodes.¹⁶ This technique has multiple advantages such as safety, non-invasiveness, easy operation, painlessness, reliability, objective quantification, and real-time dynamic multi-target evaluation and testing.¹⁷ sEMG technique, as a significant research tool for evaluating muscle status and motor function, has been widely used in clinical and fundamental scientific research in recent years, including the quantitative assessment of motor symptoms in Parkinson's disease.¹⁸ Moreover, electrophysiological examinations have played an important role in the detection of CI, including electroencephalography (EEG) and transcranial magnetic stimulation (TMS).^19–21 The most widely utilized approaches for assessing sEMG signals are time-domain and frequency-domain analysis. Time domain indicators describe the level of activation and recruitment of local muscular movement units, primarily used to assess muscle strength and coordination between muscles. Frequency domain indicators represent the fatigue level of local muscles and can be used to evaluate muscular fatigue.¹⁷ sEMG signals can accurately represent changes in muscle activity levels and functional status under well controlled conditions. However, there is currently no research into the use of sEMG technique to assess motor function in CI patients.

Although movement disorders precede and may predict cognitive disorders and sEMG is a non-invasive and convenient tool, there is currently no research employing sEMG technique to detect motor function in patients with CI. The aim of this study is to investigate the ability of sEMG technique under single and dual-task patterns for CI screening from the perspective of motor cognition.

Methods

Participants

A total of 639 participants were enrolled, including 284 CI and 355 controls, who were from Jili County, Liuyang, China.²² The inclusion criteria for this study are: (1) completion of a thorough cognitive assessment; (2) informed consent signed by the participant or their guardian. The exclusion criteria include: (1) a history of neurological disorders that could impair the motor function, such as stroke, Parkinson's disease, corticobasal syndrome, Huntington disease, hepatolenticular degeneration, multiple system atrophy, traumatic brain injury, and brain surgery; (2) history of other systemic diseases that affect motor function, such as osteoarthritis, limb or spinal trauma; (3) subjects who were unable to walk independently for more than 50 meters. Following the Declaration of Helsinki, written informed consent was obtained from all participants or their guardians. This study was approved by the Ethics Committee of Xiangya Hospital of Central South University. The flow chart of the study was shown in Figure 1.

Figure 1.

Flow chart of the study.

Neuropsychological assessment

All participants underwent cognitive assessment, including the MMSE and the Clinical Dementia Rating (CDR). The MMSE was used to assess the general cognitive function of the subjects. We set the cut-off scores for CI of different educational backgrounds according to the criteria reported in the previous China Population Research Report (illiterate participants, ≤17 points; participants with primary education, ≤20 points; and participants with junior secondary school or higher education, ≤24 points).²³ Participants were classified as cognitive normal (CN) control group and cognitive impairment (CI) group based on the MMSE scores. The CI group was further classified as mild cognitive impairment (MCI) and dementia. The diagnostic criteria for MCI are as follows: (1) expression of cognition related concerns by the subject, informant, or physician (CDR ≥ 0.5); (2) evidence of objective cognitive impairment in one or more cognitive domains; (3) maintain independence in terms of functional capabilities; and (4) there is no evidence of any significant impairment in social or occupational functioning.²⁴ The diagnostic criteria for dementia are based on the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV).²⁵ It is worth noting that the final diagnosis of the subjects was determined by a consensus of three neurologists specialized in cognitive disorders based on medical history, education level, and neuropsychological scores.

Movement function assessment

The SARC-F combined with calf circumference (SARC-CalF) was utilized to assess the muscle bulk for screening sarcopenia.^26,27 The SARC-F questionnaire assesses five aspects, including strength (S), assistance walking (A), risk from a chair (R), climbing stairs (C), and falls (F).²⁸ The movement speed was evaluated using the 6-meter pace test. We marked the starting point, 3 meters, 9 meters, and endpoint on a 12-meter straight line. The subjects started walking from the starting point and started timing when they reached the 3-meter line at normal speed, and ended timing when they reached the 9-meter line. Each participant was tested three times, with the fastest walking speed taken into account. In addition, fine motor function and coordination was assessed by the spiral drawing test,^29,30 while swallowing function was evaluated by Water Swallowing Test.³¹ The patient was instructed to sit upright and drink 30 milliliters of warm water, and the time required and choking were observed. Swallowing function was divided into five levels. The Tinetti Performance Oriented Mobility Assessment balance and gait subscales were performed for comprehensive evaluation of balance function and gait.³²

sEMG test

The sEMG examination was performed utilizing a wireless surface electromyography acquisition system (Beijing Changfeng Technology Co., Ltd), with a filtering bandwidth of 10–480 Hz and a sampling frequency of 1000 Hz. The selected muscles for testing were right vastus medialis (RVM), left vastus medialis (LVM), right gastrocnemius lateralis (RGL), and left gastrocnemius lateralis (LGL). The evaluation metrics include time domain metrics, frequency domain metrics, time-frequency domain metrics and nonlinear parameters. The time domain indicators included root mean square (RMS), average electromyography (AEMG), integrated electromyography (IEMG), mean absolute value (MAV), slope sign change (SSC), zero crossing rate (ZCR), etc. Frequency domain indicators include mean power frequency (MPF) and median frequency (MF). The time-frequency domain metrics include power spectral density energy (PSDE). Nonlinear parameters include sample entropy (SE), approximate entropy (AE), and fuzzy entropy (FE). In the single-task sEMG test, participants were requested to walk two round trips at normal speed on a 5-meter sidewalk without using any walking aids, with the beginning and end of the test route clearly marked by a cordon. In the dual-task sEMG test, participants were required to complete the following cognitive tasks while walking another two round trips at normal speed: subtract 7 continuously from 100.³³

Plasma biomarker assay

Venous blood was collected in EDTA tubes and centrifuged at 3000 rpm for 15 min at 4°C within 2 h of collection. Obtained plasma samples were stored at −80°C and all samples were not subjected to a freeze-thaw cycle before testing. Plasma samples were quantified using the ultra-sensitive Simoa technology (Quanterix, MA, US) on a Simoa HD-X automated platform (GBIO, Hangzhou, China). Plasma phosphorylated tau 181 (p-tau181) concentrations were measured using the P-tau181 test kit (#104111, Simoa^®pau181 V2.1 kit) provided by Quanterix. The testing physician was unaware of the participant's disease status.

Statistical analysis

We performed statistical analyses using R4.4.1. Descriptive statistics for demographic characteristics were reported as mean and standard deviation (SD) for continuous variables or frequency (percentage) for categorical variables. Continuous variables that conformed to normal distribution were analyzed for differences using t-tests or analysis of covariance (ANCOVA), otherwise Mann-Whitney U-tests or Kruskal-Wallis tests were used for analysis. Differences in categorical variables were analyzed using the chi-square test.

Features of sEMG were standardized by log10 transformation, followed by ANCOVA for multiple comparisons with gender, age, and BMI as covariates. To develop the sEMG model, we included all statistically significant sEMG features (p < 0.05) along with risk factors such as gender, age, BMI, and education level. We then plotted the receiver operating characteristic (ROC) curve and calculated the area under the curve (AUC) to assess the ability of sEMG model for distinguishing CI from controls. In addition, the correlations between sEMG parameters with MMSE scores and p-tau181 were assessed using a general linear regression model, adjusting for gender, age, education level, and BMI. p < 0.05 was considered statistically different between the groups.

Results

Demographic characteristics

639 participants were enrolled, including 213 with dementia, 71 with MCI, and 355 CN controls. The MMSE scores of the dementia and MCI group were considerably lower than those of the CN group (p < 0.001). There was no significant difference in age among the three group (p > 0.05). Plasma p-tau181 levels were significantly elevated in dementia and MCI group compared to controls (p = 0.02). The demographic characteristics of the subjects are summarized in Table 1.

Table 1.

Characteristics of clinical features of enrolled participants.

	Dementia	MCI	CN	p
N	213	71	355
Gender (Female%)	64.79%	47.89%	58.59%	0.04
Age	69.97 ± 5.97	70.62 ± 6.70	69.75 ± 5.28	0.49
Education (illiteracy and primary (%))	61.3%	45.07%	60.56	0.02
BMI	24.54 ± 3.16	24.32 ± 2.95	24.63 ± 2.86	0.72
MMSE	18.51 ± 4.09	23.69 ± 2.28	26.72 ± 2.31	<0.001
Plasma p-tau181 (pg/mL)	18.30 ± 6.85	16.38 ± 3.80	15.16 ± 5.21	0.02

MCI: mild cognitive impairment; CN: cognitive normal; BMI: body mass index; MMSE: Mini-Mental State Examination; p-tau181: phosphorylated tau protein 181.

Motor disorders occurred in CI group

Muscle bulk features, such as SARC-F, calf circumference scores, and SARC-CaIF scores, were substantially increased in the CI group compared to the CN group (p < 0.05), implying that CI group has a higher risk of sarcopenia. The CI group scored considerably higher on the water swallowing test than control group, suggesting a significant deterioration in swallowing function of CI group (p = 0.01). Furthermore, the Tinetti Balance and Gait Rating Scale results revealed lower balance score and gait score in the CI group than control group (p = 0.02, p = 0.01). The movement speed, fine motor function and joint mobility of CI group were likewise reduced, although the differences were not statistically significant (p > 0.05, Table 2).

Table 2.

Differential analysis of motor assessment.

Category	Item	CI	CN	p
Muscle bulk	SARC_F	0.59 ± 1.39	0.30 ± 0.72	0.03
	Calf circumference score	1.32 ± 3.40	0.65 ± 2.46	0.03
	SARC_CaIF	1.91 ± 4.16	0.97 ± 2.94	0.03
Movement speed	Fastest walking speed(m/s)	1.47 ± 0.15	1.49 ± 0.10	0.43
Fine Motor	Spiral drawing	0.12 ± 0.58	0.04 ± 0.24	0.27
Joint activity	Joints with restricted mobility	1.00 ± 2.11	0.83 ± 1.55	0.43
Swallowing	Water Swallowing Test	0.21 ± 0.81	0.05 ± 0.38	0.01
Balance and gait	Balance subscale	15.61 ± 1.96	15.96 ± 0.34	0.02
	Gait subscale	11.69 ± 1.52	11.94 ± 0.15	0.01
	Balance and Gait subscales	27.30 ± 3.39	27.94 ± 0.47	8.64E-03

CI: cognitive impairment; CN: cognitive normal; SARC_CaIF: SARC-F combined with calf circumference.

sEMG key features were identified in CI group

A total of 96 sEMG parameters were extracted, with 12 sEMG features indicating statistical differences between the CI and CN groups based on Bonferroni corrected p value (p < 5 × 10⁻⁴, Supplemental Table 1). We listed the 12 distinctive indicators in Figure 2, including single-task sEMG features like RVM_SSC, RGL_SSC, LGL_SSC, LVM_SSC, as well as dual-task sEMG features like 7Back_RVM_SSC, 7Back_LGL_IEMG, 7Back_RGL_SSC, 7Back_LGL_SSC, 7Back_LVM_SSC, 7Back_RGL_IEMG, 7Back_LVM_IEMG, 7Back_RVM_IEMG. More importantly, the distinctive sEMG features of dual-task were significantly higher than those of single-task (p < 0.05, Supplemental Table 2), which indicating the potential effect of cognitive load on motor output.

Figure 2.

Top 12 distinctive features of sEMG between the dementia, MCI and CN control groups. sEMG: surface electromyography; MCI: mild cognitive impairment; CN: cognitive normal; RVM: right vastus medialis; LVM: left vastus medialis; RGL: right gastrocnemius lateralis; LGL: left gastrocnemius lateralis; 7Back: serial seven subtractions; IEMG: integrated electromyography; SSC: slope sign change. *p < 0.05; **p < 0.01; **p < 0.001; ****p < 0.0001.

Additionally, there are nominally significant differences in four sEMG features between MCI and dementia group, such as 7Back_RVM_SSC (p = 0.036), 7Back_LVM_SSC (p = 0.008), 7Back_RGL_SSC (p = 0.006), 7Back_LGL_SSC (p = 0.030).

Correlation between distinctive sEMG parameters and MMSE scores

General linear regression models were employed to assess the correlation between distinctive sEMG indicators and MMSE scores in the CI group, adjusting for covariates such as gender, age, education level, and BMI. The results revealed that the distinctive features of sEMG were significantly correlated with MMSE scores, including 7Back_RVM_SSC (r = 0.522, p = 0.042), 7Back_LVM_SSC (r = 0.522, p = 0.025), 7Back_RGL_SSC (r = 0.521, p = 0.033), 7Back_LGL_SSC (r = 0.524, p = 0.013), 7Back_RVM_IEMG (r = 0.526, p = 0.008), RVM_SSC (r = 0.507, p = 0.045), LVM_SSC (r = 0.508, p = 0.035), RGL_SSC (r = 0.51, p = 0.018), LGL_SSC (r = 0.511, p = 0.014) (Figure 3). The findings indicated that sEMG features, particularly SSC values, were related to the severity of dementia.

Figure 3.

Linear correlations between sEMG features and MMSE scores. sEMG: surface electromyography; MMSE: Mini-Mental State Examination; RVM: right vastus medialis; LVM: left vastus medialis; RGL: right gastrocnemius lateralis; LGL: left gastrocnemius lateralis; 7Back: serial seven subtractions; IEMG: integrated electromyography; SSC: slope sign change.

Correlation between distinctive sEMG features and plasma p-tau181

This study analyzed the correlation between the distinctive sEMG indicators and plasma p-tau181 levels. We discovered a significant positive correlation between 7Back_RGL_IEMG and p-tau181 (r = 0.21, p = 0.017, Figure 4). Due to plasma p-tau181 being a diagnostic biomarker for CI,^34–36 this shows that sEMG technique played a role in identifying CI.

Figure 4.

Correlations between sEMG features and plasma p-tau181 levels. sEMG: surface electromyography; RVM: right vastus medialis; LVM: left vastus medialis; RGL: right gastrocnemius lateralis; LGL: left gastrocnemius lateralis; 7Back: serial seven subtractions; IEMG: integrated electromyography; SSC: slope sign change; p-tau181: phosphorylated tau protein 181. *p < 0.05.

Diagnostic efficiency of the sEMG model

Diagnostic accuracy of the sEMG model for distinguishing CI from CN

To evaluate the efficacy of sEMG model for discriminating CI from CN, we constructed the sEMG model by incorporating all statistically significant parameters (p < 0.05) of sEMG as well as risk factors as gender, age, BMI, and education level. We found that the dual-task sEMG model (AUC = 0.801) outperformed the single-task sEMG model (AUC = 0.770) in terms of CI detection. After integrating the single and dual-task sEMG models, the AUC value increased to 0.849 (Figure 5(a)).

Figure 5.

Diagnostic accuracy of the sEMG model. (a) Diagnostic accuracy of the sEMG model for distinguishing CI from CN; (b) Diagnostic accuracy of the sEMG model for distinguishing MCI from CN; (c) Diagnostic accuracy of the sEMG model for distinguishing MCI from dementia; (d) Diagnostic accuracy of the sEMG model and p-tau181 in distinguishing CI from CN. MCI: mild cognitive impairment; CN: cognitive normal; sEMG: surface electromyography; AUC: area under the curve; p-tau181: phosphorylated tau protein 181.

Diagnostic accuracy of the sEMG model for distinguishing MCI from CN and dementia

We further explored the ability of the sEMG model to distinguish MCI from CN and dementia. The dual-task sEMG model was better at distinguishing MCI from controls (AUC = 0.811) than the single-task sEMG model (AUC = 0.774). Surprisingly, the AUC value for discriminating MCI from controls using the combined single and dual-task sEMG model was as high as 0.875 (Figure 5(b)). Besides, the AUC value for distinguishing MCI from dementia using the joint single and dual-task sEMG models reached 0.864 (Figure 5(c)). These findings indicated that the combined sEMG models could effectively distinguish patients with MCI from those with CN or dementia.

Diagnostic accuracy of the sEMG model and p-tau181 in distinguishing CI from CN

Furthermore, we evaluated the capability of sEMG model with plasma markers to discriminate CI from CN. The findings revealed that the ability for CI detection using dual-task sEMG model and p-tau181 (AUC = 0.871) was superior to that using single_task sEMG model and p-tau181 (AUC = 0.841). The AUC value was as high as 0.948 after combining single and dual-task sEMG models with p-tau181 (Figure 5(d)), implying that combining plasma p-tau181 levels contributes to the efficacy of the sEMG model in detecting CI.

Discussion

This is the first study to utilize sEMG technique for CI screening from the motor cognition perspective. Our study discovered that the CI group exhibited significantly decreased motor function, including muscle bulk, swallowing function, balancing function, and gait. Substantial alterations in sEMG parameters were detected in CI patients compared to controls. The sEMG features were significantly correlated with cognitive scores, while the integrated electromyography feature of right gastrocnemius lateralis of dual-task sEMG were considerably positively related to plasma p-tau181 levels. Moreover, the combination of single and dual-task sEMG model could effectively distinguish CI and MCI patients from controls. The capacity of CI detection was greatly enhanced after combining with plasma p-tau181. Given that sEMG technique is non-invasive and easy to operate, this tool is expected to be applied for large-scale screening of early CI in the elderly population.

We found that CI patients exhibited decreased motor function, including muscle bulk, swallowing function, balancing function, and gait, indicating the presence of motor disorders in CI patients. Research have shown that patients with sarcopenia have a relatively high incidence of MCI, and sarcopenia may be a risk factor for MCI.³⁷ A meta-analysis revealed a strong link between sarcopenia and MCI, AD, and other types of dementia.³⁸ Dysphagia, gait and balance disorders are common in dementia patients, raising the risk of aspiration and falls.^11,39 These data suggest that cognitive impairment always coexists with motor disorders.

We explored the alterations of sEMG features in patients with CI for the first time. Our findings revealed that the SSC values of each channel were considerably higher in the CI group than the CN group. The SSC parameter refers to the number of times the sign of the rate of change of amplitude changes in the waveform of sEMG signals, and it is frequently used to indicate the state of impending fluctuating changes in the sEMG signal. Patients with CI had considerably higher SSC values, indicating a greater frequency of fluctuations in sEMG signals. Similarly, gait variability provides real-time information about subtle changes between strides. Multiple studies have shown that gait variability is a sensitive marker of neurological dysfunction, predicting the incidence of dementia^40,41 and falls.⁴² A recent study showed that high gait variability is a hallmark of cognitive cortical dysfunction and can assist diagnose AD.⁴³ Notably, a study found that gait variability in dementia patients was structurally and functionally associated with the primary motor cortex, prefrontal cortex, basal ganglia, and hippocampus.⁴⁴ The hippocampus and prefrontal cortex are the areas most affected by AD pathology, and their structural and functional similarities may help explain the association between gait variability and dementia. The SSC values, like gait variability, represent the number of fluctuations in EMG signals and the variability of muscle activity. Our research demonstrated that the SSC values of sEMG in CI patients were substantially higher, indicating that the frequency of muscle activity fluctuations in CI patients were significantly increased.

In our cohort, the IEMG value of sEMG in the CI group was considerably higher than that in the control group. The IEMG value refers to the total area under a certain EMG signal, reflecting the contraction characteristics of the corresponding muscle over a period of time. The magnitude of IEMG value depends on the change of EMG amplitude, which correlates with the function of neuromuscular fibers, and the higher the IEMG value, the stronger the strength of muscle contraction is indicated. Multiple investigations have shown that increased excitability of the motor cortex is a common feature of dementia.²¹ TMS can be used to assess cortical network function, and research have found that resting motor threshold and active motor threshold are reduced in AD patients, indicating an increase in motor cortical excitability in AD patients. This reflects the functional alterations in cortical neurotransmission leading to an imbalance between excitatory and inhibitory activity.⁴⁵ A study employing sEMG to detect “paratonia” in patients with dementia found that paratonia increased with normal aging and cognitive deterioration.⁴⁶ Paratonia is thought to be a frontal lobe inhibitory release signal produced by the cerebral cortex that inhibits individuals from relaxing their muscles in the presence of CI.^46,47 Therefore, we conjecture that the increased IEMG values in CI patients may stem from increased excitability of the motor cortex.

In the present study, we found that the SSC values of sEMG were significantly correlated with MMSE scores, implying that the SSC values of sEMG were associated with dementia severity. A previous study discovered that gait variability altered with continuous cognitive deterioration and proposed that gait variability could be utilized as a predictor to identify elderly individuals at potential risk for cognitive impairment.¹⁴ Pieruccini-Faria et al. also reported that increased gait variability was associated with worse cognitive performance.⁴³ Likewise, our study discovered that higher SSC values of sEMG were associated with lower cognitive assessment. Furthermore, several features of sEMG differed significantly between MCI and dementia group. These findings indicated that sEMG technique may be utilized to evaluate the progression of cognitive impairment diseases.

Furthermore, this study revealed a substantial correlation between the RGL_IEMG values of dual-task sEMG and plasma p-tau181. Previous brain autopsy investigation showed that physical weakness among the elderly was associated with AD pathology.⁴⁸ Nadkarni et al. demonstrated that the deposition of Aβ in the brain was associated with reduced gait speed in elderly adults.¹⁰ The correlation between RGL_IEMG values of sEMG and plasma p-tau181 discovered in this study also reflects the correlation between parameters of sEMG technique and pathology. Plasma p-tau181 is regarded a biomarker for CI.^34–36,49 Therefore, our results suggest that RGL-IEMG values of sEMG are associated with CI, and sEMG is expected to develop into a new diagnostic tool for CI.

More importantly, our study showed that combining single and dual-task sEMG models could effectively distinguish CI and MCI patients from controls. The gold standard for AD diagnosis, CSF examination and PET scans, are either invasive or expensive. Besides, the validity of routine MMSE in differentiating MCI is limited, and educational bias lead to the underdiagnosis of cognitive assessments in the diagnosis of CI.⁵⁰ Thence, there is an urgent need for an objective, non-invasive, convenient and cost-effective CI screening technique. In recent years, the analysis of electrophysiological signals such as EEG and TMS have played an essential role in the early identification and diagnosis of dementia. Our team previously revealed that EEG could be utilized for the diagnosis and disease progression evaluation of MCI and AD.⁵¹ TMS has been employed in the differential diagnosis of AD patients, and the combination of TMS indicators can effectively distinguish AD patients from patients with frontotemporal lobe dementia and controls.⁵² Besides, TMS study revealed an increased excitability of the motor cortex in dementia patients.⁴⁵The pathological proteins of AD have been widely reported to be found in many cortical regions, including primary motor cortex and auxiliary motor areas.⁵³ Previous research has demonstrated that motor deficits precede cognitive impairment by more than ten years,⁷ and motor function assessment techniques hold promise as an early screening tool for cognitive deficits. We explored the efficacy of sEMG model in screening CI patients for the first time. The results showed that the combination of single and dual-task sEMG model could effectively discriminate CI and MCI patients from CN, which provides new evidence for the role of electrophysiological signal analysis in early identification and diagnosis of dementia. Overall, these findings suggest that the combination of single and dual-task sEMG examination can be utilized for large-scale screening of CI patients in the community.

This is the first research to investigate sEMG characteristics in individuals with CI, and we explored the prospect of using sEMG to screen CI patients in the community. sEMG is low-cost, relatively easy to obtain, non-invasive, portable, and easy to implement. However, sEMG can only evaluate superficial muscles and cannot accurately distinguish target muscles from adjacent muscles due to its use of surface electrodes.⁵⁴ This study had some limitations. First, as our study was conducted in a community queue, blood tests and cognitive assessments were primarily used to discriminate patients with cognitive impairment from controls, and it is difficult to classify the patients into the dementia subgroup. Second, our queue employed the MMSE and CDR as measurement indicators for cognitive assessment, and more specific cognitive measurement methods would be more meaningful. Third, sEMG can serve as an effective screening tool for CI requires validation in diverse populations and real-world clinical settings.

Conclusions

In summary, this is the first study to apply sEMG technique in CI screening. We find that motor dysfunction is common in patients with CI, and sEMG technique holds promise for the large-scale screening of early CI. There is a need for longitudinal studies in multicenter hospital settings to confirm these findings. Future research should focus on developing standardized, cost-effective protocols to make sEMG a viable and accessible screening tool for cognitive impairment across diverse populations and in different clinical settings.

Supplemental Material

sj-docx-1-alz-10.1177_13872877251361853 - Supplemental material for Surface electromyography as a novel screening tool for early cognitive impairment: Exploring motor-cognitive interactions

Supplemental material, sj-docx-1-alz-10.1177_13872877251361853 for Surface electromyography as a novel screening tool for early cognitive impairment: Exploring motor-cognitive interactions by Xiaoli Hao, Xuan Yang, Fei Wang, Xuewen Xiao, Shuliang Chen, Ziyu Ouyang, Yingzi Liu, Junyin Lu, Yiliang Liu, Tianyan Xu, Li Yuan, Yuzhang Bei, Hasiyeti Yibulaiyin, Shilin Luo, Beisha Tang, Lu Shen and Bin Jiao in Journal of Alzheimer's Disease

Footnotes

Acknowledgements

We sincerely thank all the subjects for their participation.

ORCID iD

Lu Shen

Ethical considerations

The study was approved by the Ethics Committee of Xiangya Hospital of the Central South University in China (equivalent to an Institutional Review Board) with the ethics number 2022020483.

Consent to participate

Written informed consent was obtained from each participant or their legal representatives.

Author contribution(s)

Xiaoli Hao: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing.

Xuan Yang: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing.

Fei Wang: Data curation, Investigation.

Xuewen Xiao: Data curation, Investigation.

Shuliang Chen: Data curation, Investigation.

Ziyu Ouyang: Data curation, Investigation.

Yingzi Liu: Data curation, Investigation.

Junyin Lu: Data curation, Investigation.

Yiliang Liu: Data curation, Investigation.

Tianyan Xu: Data curation, Investigation.

Li Yuan: Data curation, Investigation.

Yuzhang Bei: Data curation, Investigation.

Hasiyeti Yibulaiyin: Data curation, Investigation.

Shilin Luo: Resources, Writing – review & editing.

Beisha Tang: Resources, Writing – review & editing.

Lu Shen: Conceptualization, Funding acquisition, Resources, Writing – review & editing.

Bin Jiao: Conceptualization, Funding acquisition, Resources, Writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Key R&D Program of China (2023YFC3603700), the National Natural Science Foundation of China (No. U22A20300, 82371434), the Science and Technology Major Project of Hunan Province (2021SK1020), Outstanding Youth Fund of Hunan Provincial Natural Science Foundation (2024JJ2097), Hunan Health Commission (20232460), Hunan Provincial Natural Science Foundation (2025JJ60696).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data are available on reasonable request to the corresponding author.

Supplemental material

Supplemental material for this article is available online.

References

Livingston

Huntley

Liu

, et al. Dementia prevention, intervention, and care: 2024 report of the lancet standing commission. Lancet 2024; 404: 572–628.

Jia

Wang

Wei

, et al. The prevalence of dementia in urban and rural areas of China. Alzheimers Dement 2014; 10: 1–9.

Drew

. An age-old story of dementia. Nature 2018; 559: S2–s3.

Xiao

Liu

Yao

, et al. LMTK2 And CRB1 are two novel risk genes for Alzheimer's disease in Han Chinese. J Prev Alzheimers Dis 2025; 12: 100087.

Oveisgharan

Wang

Barnes

, et al. The time course of motor and cognitive decline in older adults and their associations with brain pathologies: a multicohort study. Lancet Healthy Longev 2024; 5: e336–e345.

Camicioli

Howieson

Oken

, et al. Motor slowing precedes cognitive impairment in the oldest old. Neurology 1998; 50: 1496–1498.

Buracchio

Dodge

Howieson

, et al. The trajectory of gait speed preceding mild cognitive impairment. Arch Neurol 2010; 67: 980–986.

Wilkins

Roe

Morris

, et al. Mild physical impairment predicts future diagnosis of dementia of the Alzheimer's type. J Am Geriatr Soc 2013; 61: 1055–1059.

Kuo

Zhang

Chen

, et al. Associations of grip strength, walking pace, and the risk of incident dementia: a prospective cohort study of 340212 participants. Alzheimers Dement 2023; 19: 1415–1427.

10.

Nadkarni

Perera

Snitz

, et al. Association of brain amyloid-β with slow gait in elderly individuals without dementia: influence of cognition and apolipoprotein E ε4 genotype. JAMA Neurol 2017; 74: 82–90.

11.

Makhnevich

Perrin

Talukder

, et al. Thick liquids and clinical outcomes in hospitalized patients with Alzheimer disease and related dementias and dysphagia. JAMA Intern Med 2024; 184: 778–785.

12.

Duchowny

Ackley

Brenowitz

, et al. Associations between handgrip strength and dementia risk, cognition, and neuroimaging outcomes in the UK biobank cohort study. JAMA Netw Open 2022; 5: e2218314.

13.

Chou

Nishita

Nakagawa

, et al. Role of gait speed and grip strength in predicting 10-year cognitive decline among community-dwelling older people. BMC Geriatr 2019; 19: 186.

14.

Kwak

Kostic

Kim

. Gait variability-based classification of the stages of the cognitive decline using partial least squares-discriminant analysis. Sci Prog 2023; 106: 368504231218604.

15.

Lin

Yang

, et al. A detection model of cognitive impairment via the integrated gait and eye movement analysis from a large Chinese community cohort. Alzheimers Dement 2024; 20: 1089–1101.

16.

Pullman

Goodin

Marquinez

, et al. Clinical utility of surface EMG: report of the therapeutics and technology assessment subcommittee of the American academy of neurology. Neurology 2000; 55: 171–177.

17.

Alcan

Zinnuroğlu

. Current developments in surface electromyography. Turk J Med Sci 2023; 53: 1019–1031.

18.

Sande de Souza

Dionísio

Almeida

. Multi-joint movements with reversal in Parkinson's disease: kinematics and electromyography. J Electromyogr Kinesiol 2011; 21: 376–383.

19.

Jiang

, et al. Differentiating between Alzheimer's disease and frontotemporal dementia based on the resting-state multilayer EEG network. IEEE Trans Neural Syst Rehabil Eng 2023; 31: 4521–4527.

20.

Tăuƫan

Casula

Pellicciari

, et al. TMS-EEG perturbation biomarkers for Alzheimer's disease patients classification. Sci Rep 2023; 13: 7667.

21.

Koch

Martorana

Caltagirone

. Transcranial magnetic stimulation: emerging biomarkers and novel therapeutics in Alzheimer's disease. Neurosci Lett 2020; 719: 134355.

22.

Yuan

, et al. The prevalence and risk factors study of cognitive impairment: analysis of the elderly population of han nationality in hunan province, China. CNS Neurosci Ther 2024; 30: e14478.

23.

Katzman

Zhang

Ouang Ya

, et al. A Chinese version of the Mini-mental state examination; impact of illiteracy in a Shanghai dementia survey. J Clin Epidemiol 1988; 41: 971–978.

24.

Langa

Levine

. The diagnosis and management of mild cognitive impairment: a clinical review. JAMA 2014; 312: 2551–2561.

25.

Jiao

Zhang

Bei

, et al. A detection model for cognitive dysfunction based on volatile organic compounds from a large Chinese community cohort. Alzheimers Dement 2023; 19: 4852–4862.

26.

Chen

Woo

Assantachai

, et al. Asian working group for sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J Am Med Dir Assoc 2020; 21: 300–307.e302.

27.

Yang

Cao

, et al. Reliability and validity of the Tinetti performance oriented mobility assessment in Chinese community-dwelling older adults. Geriatr Nurs 2023; 53: 85–89.

28.

Nishikawa

Asai

Fukunishi

, et al. Screening tools for sarcopenia. In Vivo 2021; 35: 3001–3009.

29.

Longstaff

Heath

. Spiral drawing performance as an indicator of fine motor function in people with multiple sclerosis. Hum Mov Sci 2006; 25: 474–491.

30.

Memedi

Westin

Nyholm

. Spiral drawing during self-rated dyskinesia is more impaired than during self-rated Off. Parkinsonism Relat Disord 2013; 19: 553–556.

31.

Osawa

Maeshima

Tanahashi

. Water-swallowing test: screening for aspiration in stroke patients. Cerebrovasc Dis 2013; 35: 276–281.

32.

Tinetti

. Performance-oriented assessment of mobility problems in elderly patients. J Am Geriatr Soc 1986; 34: 119–126.

33.

Montero-Odasso

Muir

Speechley

. Dual-task complexity affects gait in people with mild cognitive impairment: the interplay between gait variability, dual tasking, and risk of falls. Arch Phys Med Rehabil 2012; 93: 293–299.

34.

Janelidze

Mattsson

Palmqvist

, et al. Plasma P-tau181 in Alzheimer's disease: relationship to other biomarkers, differential diagnosis, neuropathology and longitudinal progression to Alzheimer's dementia. Nat Med 2020; 26: 379–386.

35.

Weng

Zhang

, et al. Genetic spectrum features and diagnostic accuracy of four plasma biomarkers in 248 Chinese patients with frontotemporal dementia. Alzheimers Dement 2024; 20: 7281–7295.

36.

Gonzalez

Ashton

Gomes

, et al. Association of plasma p-tau181 and p-tau231 concentrations with cognitive decline in patients with probable dementia with Lewy bodies. JAMA Neurol 2022; 79: 32–37.

37.

Yang

Xiao

Leng

, et al. A systematic review and meta-analysis of the prevalence and correlation of mild cognitive impairment in sarcopenia. J Cachexia Sarcopenia Muscle 2023; 14: 45–56.

38.

Amini

Ibn Hach

Lapauw

, et al. Meta-analysis on the interrelationship between sarcopenia and mild cognitive impairment, Alzheimer's disease and other forms of dementia. J Cachexia Sarcopenia Muscle 2024; 15: 1240–1253.

39.

Taylor

JCT

. Dementia. Handb Clin Neurol 2018; 159: 303–321.

40.

Darweesh

SKL

Licher

Wolters

, et al. Quantitative gait, cognitive decline, and incident dementia: the Rotterdam study. Alzheimers Dement 2019; 15: 1264–1273.

41.

Ceïde

Ayers

Lipton

, et al. Walking while talking and risk of incident dementia. Am J Geriatr Psychiatry 2018; 26: 580–588.

42.

Hausdorff

Rios

Edelberg

. Gait variability and fall risk in community-living older adults: a 1-year prospective study. Arch Phys Med Rehabil 2001; 82: 1050–1056.

43.

Pieruccini-Faria

Black

Masellis

, et al. Gait variability across neurodegenerative and cognitive disorders: results from the Canadian consortium of neurodegeneration in aging (CCNA) and the gait and brain study. Alzheimers Dement 2021; 17: 1317–1328.

44.

Tian

Chastan

Bair

, et al. The brain map of gait variability in aging, cognitive impairment and dementia-A systematic review. Neurosci Biobehav Rev 2017; 74: 149–162.

45.

Oberman

Benussi

. Transcranial magnetic stimulation across the lifespan: impact of developmental and degenerative processes. Biol Psychiatry 2024; 95: 581–591.

46.

Marinelli

Trompetto

Puce

, et al. Electromyographic patterns of paratonia in normal subjects and in patients with mild cognitive impairment or Alzheimer's disease. J Alzheimers Dis 2022; 87: 1065–1077.

47.

Drenth

Zuidema

Bautmans

, et al. Paratonia in dementia: a systematic review. J Alzheimers Dis 2020; 78: 1615–1637.

48.

Buchman

Schneider

Leurgans

, et al. Physical frailty in older persons is associated with Alzheimer disease pathology. Neurology 2008; 71: 499–504.

49.

Brickman

Manly

Honig

, et al. Plasma p-tau181, p-tau217, and other blood-based Alzheimer's disease biomarkers in a multi-ethnic, community study. Alzheimers Dement 2021; 17: 1353–1364.

50.

Rosli

Tan

Gray

, et al. Cognitive assessment tools in Asia: a systematic review. Int Psychogeriatr 2016; 28: 189–210.

51.

Jiao

Zhou

, et al. Neural biomarker diagnosis and prediction to mild cognitive impairment and Alzheimer's disease using EEG technology. Alzheimers Res Ther 2023; 15: 32.

52.

Vucic

Stanley Chen

Kiernan

, et al. Clinical diagnostic utility of transcranial magnetic stimulation in neurological disorders. Updated report of an IFCN committee. Clin Neurophysiol 2023; 150: 131–175.

53.

Albers

Gilmore

Kaye

, et al. At the interface of sensory and motor dysfunctions and Alzheimer's disease. Alzheimers Dement 2015; 11: 70–98.

54.

Suchowersky

Reich

Perlmutter

, et al. Practice parameter: diagnosis and prognosis of new onset Parkinson disease (an evidence-based review): report of the quality standards subcommittee of the American academy of neurology. Neurology 2006; 66: 968–975.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB