Hand motor function of patients with early- and late-onset Alzheimer's disease and its relation to activities of daily living

Abstract

Background

Alzheimer's disease (AD) is the most common cause of dementia, characterized by a progressive deterioration of cognitive capacity and the ability to carry out activities of daily living (ADL). Hand motor function may also be impaired, with slower and less automated movements during fine motor tasks. The severity of impairments may depend on task characteristics and the age of onset of AD.

Objective

This study investigated hand motor function of persons with dementia due to AD during handwriting and sequential hand movements, focusing on the impact of task complexity and age at onset.

Methods

Kinematic analysis of handwriting and sequential hand movements was carried out in 24 AD patients (early-onset AD (EOAD): n = 13; late-onset AD (LOAD): n = 11) and 23 controls (≤ 65 years: n = 12; > 65 years: n = 11). To estimate the impact on patients’ ADL performance, the Jebsen-Taylor Hand Function Test (JTHFT) was administered.

Results

AD patients exhibited a significantly lower handwriting and sequencing performance compared to healthy controls. Complexity effects were detectable for handwriting and sequencing; for z-standardized sequencing frequency, they were more prominent in AD patients than controls. Age at onset had no effect on handwriting or sequential hand movements; however, handwriting and sequencing frequency were predictive of JTHFT performance in EOAD, but not in LOAD.

Conclusions

Kinematic measures of both handwriting and sequential hand movements were able to capture impairments in AD patients. Deteriorations of frequencies seem to translate into deficits in simulated ADL performance only in EOAD.

Keywords

activities of daily living (ADL)Alzheimer's disease early-onset Alzheimer's disease fine motor control handwriting kinematic analysis sequential hand movements

Introduction

According to current estimates,¹ the number of adults with some type of dementia will increase to 152.8 million globally by the year 2050. Dementia is characterized by a progressive deterioration of cognitive capacity in multiple domains that exceeds the normal aging process and impedes the ability to independently carry out activities of daily living (ADL).^2,3 Approximately 60–80% of cases of dementia are attributable to Alzheimer's disease (AD),⁴ which is generally associated with impairments in memory, language, orientation, behavior, mood, and functional abilities.^4,5

According to age of onset dementia due to AD is subdivided into an early-onset form (EOAD) and a late-onset form (LOAD).³ EOAD is characterized by symptom onset before the age of 65³ and accounts for approximately 5–10% of cases.^6,7 Both EOAD and LOAD show the same neuropathological hallmarks,^6–8 and for the most part, no distinction is made between the two forms in the selection of diagnostic procedures or (pharmaco) therapies.⁸ However, in terms of clinical manifestation, course of disease, and genetic factors, research indicates clear differences between EOAD and LOAD. Memory symptoms are not as prevailing in EOAD as in LOAD, while impairments in language, attention, motor function, executive and visuospatial abilities occur more frequently, according to previous studies.^6,8

According to earlier diagnostic criteria, motor deficits were expected in the late stages of AD.^9,10 This perspective has changed, as impairments in grip strength,^11,12 and gait¹³ are now suggested as factors that may potentially be associated with an increased risk of dementia and even incident mild cognitive impairment (MCI).

As for handwriting, a skill requiring complex interactions of cognitive, kinaesthetic, and perceptual motor skills,^14,15 research indicates deficits in AD patients under various writing conditions. For example, Delazer and colleagues¹⁶ reported a lower frequency, automaticity, and velocity in AD patients compared to healthy controls in non-alphabetic tasks, which included upstroke and downstroke movements from the finger and wrist joints as well as their combination into circle movements (see also¹⁷). Deficits were further observable in AD patients during the writing of letter combinations,^16,18 single words,¹⁹ and sentences,^16,19,20 as well as during functional tasks, such as copying checks and entire paragraphs²⁰ or writing extended free texts.²¹

Further, there is evidence of significant differences between AD patients and healthy controls concerning sequential hand movements. For finger tapping tasks, previous findings have been mixed, with some (e.g.,^22–25) reporting differences between patients and controls but others not.^26,27 With regard to rapid pro-and supination movements, available research is more consistent, with reports of impaired performance,^22,23 lower frequency, and lower movement regularity.²⁷ In case of elaborate movement sequences such as the Luria test, which consists of the rapid execution of three different hand movements (Fist, Edge, Palm; see²⁸), apparent deficits in task performance were found in AD patients, making more errors than healthy older adults²⁸ and being significantly slower.²⁹

Role of task complexity

Studies that tested finger and wrist movements with different combinations¹⁷ and handwriting tasks with varying material^16,20 reported a decrement in fine motor control the more elaborate the task in both AD patients and healthy controls. While not statistically examined in the respective studies, the decrease in performance, i.e., an increase in movement time for handwriting-like movements as compared to simpler movements¹⁷ and a decrease in handwriting automaticity for alphabetic tasks as compared to non-alphabetic tasks¹⁶ was particularly evident in AD patients. This observation provides indices that the decrement in motor ability for more elaborate tasks may be more pronounced in AD patients than in controls. This is also underscored by results of classification analyses. There, AD patients and healthy controls could be better distinguished with respect to handwriting kinematics in a sentence task than in a simpler circle³⁰ or word¹⁹ task and with respect to sequential hand movements in complex motor tasks including diadochokinetic hand movements than in gross and single joint motor tasks.²³

Role of age at onset

As mentioned above, deficits in motor function may be more common in EOAD. Overall, supporting evidence is limited. In direct comparisons of higher aspects of motor control, EOAD patients exhibited significantly greater impairment in praxis than LOAD patients.^31–33 Apraxia even appeared as the most common non-memory deficit in EOAD patients.³³ For sequential hand movement tasks,³¹ as well as motor signs (especially the extrapyramidal signs rigidity, tremor, and bradykinesia^34,35) and axial features (including aspects such as gait and posture³⁴), however, previous research yielded no significant differences when directly comparing EOAD and LOAD. Yet, given the considerable age differences between EOAD and LOAD patients in these studies and the well-established decline in motor capacity with increasing age,^36,37 direct comparisons may be biased by age-effects.

Relationship with ADL performance

Per definition, dementia is linked to a deterioration in the ability to carry out ADL independently. Functional ability, or lack thereof, strongly impacts care needs and quality of life^38,39 and indicates the success of therapeutic interventions and the course and severity of AD.³⁸ First functional impairments are frequently observed as difficulties in performing complex ADL such as cooking in the MCI stage of AD and intensify as the disease progresses, with some patients eventually requiring assistance in accomplishing basic ADL.^38,40 Based on the close link between ADL with cognitive functions on the one hand and motor functions on the other hand, a careful examination of the relationship between manual fine motor skills and ADL performance seems warranted in patients with cognitive deficits.

In this regard, fine motor dexterity, as measured via performance in the Nine-Hole-Peg-Test⁴¹ or the Purdue Pegboard Test⁴² has been demonstrated to be predictive of self-care,⁴¹ domestic and complex ADL⁴² in MCI patients^41,42 and in patients with dementia due to AD.⁴¹ In addition, the presence of motor signs, particularly bradykinesia and rigidity, has also been found to be associated with a higher risk of functional decline in AD patients in a longitudinal investigation by Scarmeas and colleagues.⁴³ However, these studies^41–43 captured ADL performance using informant-based scales. Given that informant reports of functional abilities may be prone to bias,^39,40 the use of performance-based measures of ADL could provide more reliable information about the relationship between functional abilities and hand motor function in AD.

Aims and hypotheses

The overarching objective of this study was to investigate hand motor function in persons with dementia due to AD, focusing specifically on how it is affected by age at onset of AD and task complexity. We chose handwriting and sequential hand movements, since some data on AD patients’ performance exist, complexity can be scaled in both task types, and analysis can be based on precise performance measures and kinematics. In addition, we were interested in the relationship between fine motor deficits and ADL performance in AD patients.

Firstly, we aimed to examine the influence of task complexity on hand motor function of AD patients compared to healthy controls. Previous research on handwriting^16,17,19,30 and sequential hand movements²³ suggest a disproportionately large decline in performance in AD patients from simple to more elaborate tasks. Accordingly, we hypothesized a more pronounced increase in impairments in handwriting and sequential hand movements the more elaborate the task in AD patients. We based the classification of task complexity on the concept of Wood,⁴⁴ considering the number of unique subactions and coordinative requirements as critical elements.

Secondly, we aimed to explore whether the deficits in motor performance in AD patients are related to age at onset. While previously EOAD and LOAD patients’ motor performance was compared directly,^31–35 the present research seeks to compare the hand motor function of EOAD and LOAD patient groups with respective age-matched healthy control groups, to avoid distortions due to the age-related decline affecting both handwriting. (e.g.,^45,46) and sequential hand movements (e.g.,^47–49). Considering reports of an overall more aggressive course of disease in EOAD^6,7 and findings of more^31–33 or equally pronounced motor deficits^31,34,35 in EOAD when directly contrasted to LOAD, we anticipated larger discrepancies between EOAD patients and their age-matched controls than between LOAD patients and their controls.

Lastly, we aimed to investigate the relationship between hand motor function and ADL performance in AD patients. Given that previous research has identified an association between fine motor dexterity and informant-based ADL scales in AD patients,^41,42 we hypothesized hand motor function in handwriting tasks and sequential hand movements to be able to predict functional abilities in AD patients, as captured by a performance-based ADL measure.

Methods

Study design and participants

This study followed a cross-sectional design. Recruitment of patients was carried out through the Centre for Cognitive Disorders, Technical University of Munich (TUM) University Hospital. The control group involved family members (e.g., spouse) or caregivers who accompanied patients to appointments at the clinic. The study addressed adults aged 45 to 90 years. Patients had to be diagnosed with dementia due to AD based on the criteria by the National Institute on Aging – Alzheimer's Association workgroup.⁵⁰ Controls had to be without any neurological disease. Exclusion criteria for both groups were impairments in hearing or vision interfering with performance, neurological diseases (other than AD in patients) such as a stroke, traumatic brain injuries, or brain tumors, medical history or presence of psychiatric diseases and a score of 15 or less in the Mini-Mental Status Examination test (MMSE,⁵¹). Moreover, disorders known to affect fine motor control or sensitivity of the hand (rheumatologic, neurological, orthopedic, or traumatic conditions, complex regional pain syndrome, polyneuropathy) precluded from participating.

All participants provided written informed consent. The study was approved by the Ethics Commission of TUM (Reference No. 479/19 S-SR) and performed in accordance with the ethical standards specified in the Declaration of Helsinki.

Interventions and procedures

Participants’ handedness was assessed by the 10-item version of Oldfield's⁵² Edinburgh Handedness Inventory. Cognitive performance was assessed with the MMSE.⁵¹ Grip strength was evaluated using a digital Jamar hand dynamometer, measuring maximum isometric hand grip strength in kilograms. Participants were instructed to sit upright in a chair with shoulders adducted, elbows in a 90° flexed, and the wrist in a neutral position and squeeze the fixed lever of the dynamometer as tightly as possible for three seconds (s). Three trials were performed per hand, starting with the dominant hand. Breaks of approximately 10 s were given between trials. Considering the sensitivity of grip strength to age and gender,⁵³ the higher of the two average values (dominant versus nondominant hand) was z-standardized for each participant using Maitland's⁵⁴ z-score calculator.

Assessment of ADL performance

To examine fine and gross hand motor function during simulated ADL in an objective, standardized, and relatively short manner (15–45 min), the Jebsen-Taylor Hand Function Test (JTHFT) was administered.⁵⁵ Given that handwriting was examined separately in the current study, only six of the seven JTHFT tasks were to be performed, i.e., (1) turning cards, (2) picking up small objects, (3) stacking checkers, (4) simulated eating, moving (5) empty and (6) filled cans. Each task was executed with the non-dominant hand first, followed by the dominant hand. The total time needed to perform all JTHFT tasks with the dominant hand served as the outcome parameter for statistical analyses.

Kinematic analysis of handwriting

Handwriting traces were recorded using a graphics tablet (Wacom Intuos Pro M, Wacom Co., Ltd, Kazo, Japan) and a wireless, pressure-sensitive pen (Wacom Pro Pen 2, Wacom Co., Ltd, Kazo, Japan). The position of the pen tip was registered with a spatial resolution of 0.25 millimeters and a temporal resolution of up to 200 Hertz (Hz). Analysis of handwriting kinematics was performed with PC-based program CSWin (Version 2011, MedCom, Munich, Germany), a software enabling the automatic segmentation of the written trace into up- and downstrokes on the y-axis and the pen tip velocity to be computed utilizing kernel estimates.⁵⁶ Apart from frequency as the number of up-and downstrokes per s measured in Hz, handwriting characteristics extracted from CSWin included writing duration, pressure and numbers of inversion in velocity. We have chosen frequency as the parameter to be reported within the scope of this publication, given that a) this parameter was also available for sequential hand movements (described below) and b) frequency is a parameter commonly reported in similar research, thereby facilitating contextualization of our results.

Handwriting was assessed during three tasks performed with the dominant hand using the inking pen and writing on a blank sheet of paper (DIN A5), which was positioned on the graphics tablet. The test started with producing a (1) Circle, proceeding with (2) ll's, and concluding with a (3) Sentence task (see Table 1). During the Circle task combining the two basic handwriting elements, i.e., finger and wrist movements, participants were asked to write superimposed circles fluidly and swiftly within 3 s without instruction on movement direction (counterclockwise versus clockwise). The ll's task consisted of repeatedly writing connected loops (l's) in pairs of two. As many pairs as possible were to be written fluently and swiftly within 10 s. In the Sentence task, “Die Wellen schlagen hoch” (German for: the waves rise high), a sentence commonly used in kinematic analyses⁵⁷ had to be written down in an individual handwriting style at regular speed without concentrating on neatness or legibility.

Table 1.

Handwriting tasks.

Task	Duration	Repetitions
Circle	3 s	3
ll's	10 s	3
Sentence	unrestricted	3

Each task was executed three times, with the average value of the three trials being included in statistical analyses. All participants were given the opportunity to familiarize themselves with writing on the tablet; each task was explained, demonstrated, and could be tried out before data collection. Movement execution was monitored, with repetition of individual trials allowed in the instance of errors or technical difficulties.

Kinematic analysis of sequential hand movements

Sequential hand movements were investigated using Polhemus motion capture systems (Polhemus Inc., Colchester, USA), mainly the Patriot^TM (sampling frequency: 60 Hz) and for the last six participants the Viper^TM (sampling frequency: 120 Hz). The source produced a magnetic field in which position and orientation of the hand in space could be tracked via a standard sensor, which was attached to the dorsum of the participants’ dominant hand between the first and second metacarpal bone. Movements were displayed and recorded using the Polhemus PiMgr graphical user interface and subsequently saved. The files were then imported into 3DAWin (version 2.3., MedCom, Munich, Germany) to analyze movements along the vertical Z-coordinate (up and down movement of the hand). Of interest was the parameter frequency as the number of movement cycles per s. All trials were also captured on video.

Sequential hand movements were collected during three tasks, i.e., (1) Tapping, (2) Diadochokinesia, and (3) Luria's. (see Figure 1). In the Tapping task, the palm of the hand was to be repeatedly elevated and then tapped onto the tabletop. Movements were performed from the elbow with a fixed wrist. No instruction was given on the expected amplitude; the investigator demonstrated the task with an amplitude of around 8 centimeters. The Diadochokinesia task consisted of rapid alternating forearm pronation and supination movements from the wrist and forearm with the elbow fixed in an approximately 90° flexed position. The Luria's task was composed of three individual movements, namely Fist, Edge, and Palm (for execution see Figure 1(c)), which had to be repeated fluidly and precisely.

Figure 1.

Snapshots from the tested sequential hand movements for the Tapping (a), Diadochokinesia (b), and Luria's (c) task.

Each task was to be executed as rapidly as possible and was initiated on “Go”, with Tapping and Diadochokinesia carried out for 6 s each and Luria's for 20 s before being terminated on “Stop”. All sequencing tasks were performed with the dominant hand, in the same order, beginning with the Tapping task and concluding with the Luria's task. Three trials were performed each before moving on to the next task. An average value was calculated from the three trials and included in statistical analyses. Tasks were explained and demonstrated, and each participant was allowed a trial run. During the Luria's task, participants were additionally provided with a hard copy of the three hand movements. Investigators did not provide any further assistance during the execution of the task. Trials were repeated only in case of technical errors or gross comprehension difficulties.

As the movement recordings revealed technical errors and inconsistencies in some trials of the Diadochokinesia and the Luria's task, video recordings were consulted to determine frequency in these two tasks (for Tapping, both methods revealed virtually identical results with a mean frequency deviation of 0.09 Hz). For Luria's, one movement cycle was defined as a complete and correctly executed series consisting of Fist, Edge, and Palm.

Data and statistical analysis

Patients and controls were compared regarding age, age group, sex, z-standardized grip strength, JTHFT duration, MMSE score, as well as handwriting and sequencing kinematics. For the categorical variables sex and age group, between-group differences were identified using Fisher's exact tests. Differences between patients and controls regarding age were investigated using Two-Sample t-tests; Wilcoxon-Mann-Whitney tests were used for comparisons of grip strength, JTHFT time, and MMSE score, given their lack of normal distribution (Shapiro-Wilk test p < 0.05).

Patient and control groups were additionally split into two age groups. For patients, the allocation was based on age at onset into an EOAD (n = 13) and a LOAD (n = 11) group. To generate similarly aged and sized groups, controls aged 65 or younger served as the young controls (n = 12) and controls over 65 as the older controls (n = 11). Differences between EOAD and LOAD patients were assessed with a Two Sample t-test for the variables age, grip strength and MMSE score, and with Wilcoxon-Mann-Whitney tests for the non-normally distributed (Shapiro-Wilk test p < 0.05) variables JTHFT time and disease duration (time since diagnosis according to medical records in months). Potential age differences between each patient group and the corresponding control group were investigated using Two-Sample t-tests. In case of heterogeneity of variance (Levene's test p < 0.05), Welch's t-tests were applied.

Data for grip strength was missing in one EOAD patient for the non-dominant hand. For handwriting, data was missing for one trial of the Circle task for one LOAD patient; for sequencing, data was missing for one trial of the Tapping task and two trials of the Diadochokinesia task for one young control; for these participants, the averages for statistical analyses were calculated from the remaining trial(s) per task. Sequencing data were fully missing for one LOAD patient who was subsequently dropped from the analysis of sequential hand movements. For both handwriting and sequencing frequency, each combination of subject group, age group, and task was tested for normal distribution (Shapiro-Wilk test p > 0.05) and homogeneity of variance (Levene's test p > 0.05).

To investigate effects of age and task on handwriting and sequential hand movement frequency in AD patients and controls, three-way mixed analyses of variance (ANOVAs) were run with both subject group (1. AD patients; 2. controls), and age group (1. Young: patients: EOAD; controls: ≤ 65 years; 2. Old: patients: LOAD; controls: > 65 years) as between-subject factors and task (1. Circle/Tapping; 2. ll's/Diadochokinesia; 3. Sentence/Luria's) as the within-subject factor. Where the assumption of sphericity was violated (Mauchly-test p < 0.05), ANOVA results were reported with a Greenhouse-Geisser correction. Significant effects were followed up using Holm-adjusted t-tests, with paired t-tests for within-subject factors and Two Sample t-tests for between-subject factors.

Subsequently, the EOAD and LOAD patients’ performance on each task was standardized on the mean (M) and standard deviation (SD) of the younger and older control groups’ performance for handwriting and sequencing frequency, respectively. This was done to examine whether EOAD and LOAD groups significantly differed from their healthy counterparts across tasks regardless of measure, given results of the three-way ANOVA suggesting that particularly for sequencing, effects of large magnitude differences of the outcome parameter across tasks may have negatively influenced the meaningfulness of this main analysis. Z-standardization was followed by a two-way mixed ANOVA with the between-subject factor age group (two levels: EOAD, LOAD) and the within-subject factor task (three levels: Circle/Tapping, ll's/Diadochokinesia, Sentence/Luria's). Significant effects were explored with Holm-adjusted t-tests. Due to the chosen standardization procedure, subject group could not be included as a third factor in the ANOVAs. Therefore, the AD patients’ z-standardized handwriting and sequencing frequencies were tested against zero (with zero corresponding to the controls’ mean z-score) via One Sample t-tests to investigate whether the patients’ performance differed from the control group.

To assess whether motor performance in handwriting and sequential hand movements can predict ADL performance in AD patients, measures of handwriting and sequential hand movements were correlated with the JTHFT score for the dominant hand using Spearman's correlations. For both handwriting and sequencing frequency, the time needed to perform the JTHFT was correlated with the average of all three tasks in EOAD and LOAD patients separately. MMSE scores were also correlated with the JTHFT to determine the extent to which global cognitive impairment can predict ADL performance in EOAD and LOAD.

Effect sizes were reported for significant results; for t-tests and Wilcoxon-Mann-Whitney tests, Cohen's d was utilized, for significant ANOVA results, generalized eta squared ( $η_{G}^{2}$ ) was given. Data analysis and visualization were performed with R (version 4.2.2, R Core Team, 2022) at an α-level of 0.05.

Results

Participants

The study sample comprised 24 patients (8 females, 16 males) and 23 controls (16 females, 7 males), with a mean age of 68.4 (SD 9.1) and 65.0 (SD 9.2) years, respectively. An overview of participant characteristics is provided in Table 2. The AD patient and the control group had a similar age (t(45) = −1.28, p = 0.207) and a comparable grip strength (z = 0.61, p = 0.551). With more males than females in the patient group and vice versa in the control group, groups differed significantly for sex (p = 0.020, OR = 0.23, 95% CI 0.05–0.87). Disease duration averaged slightly over two years with a high SD (Table 2). Patients took significantly longer than controls to complete the JTHFT with the dominant hand, which was the right hand for all controls and 23 patients (z = −3.19, p = 0.001, d = 0.75). As expected, the MMSE score was significantly lower in patients than in controls (z = 4.71, p < 0.001, d = 1.67).

Table 2.

Characteristics of the overall study sample.

Characteristic	Patients, N = 24	Controls, N = 23	p
Age [years]	68.4 (9.1), 54–84	65.0 (9.2), 53–84	0.207^c
Young/EOAD	13	12	1.000^d
Old/LOAD	11	11
Sex			0 . 020 ^d
Female	8 (33.3%)	16 (69.6%)
Male	16 (66.7%)	7 (30.4%)
Grip Strength [z]^a	0.0 (0.8), −1.1–1.8	0.2 (0.7), −0.7–2.7	0.551^e
JTHFT Time [s]	42.6 (18.4), 27.5–118.7	32.4 (4.5), 26.8–43.9	0 . 001 ^e
Disease Duration [months]^b	24.4 (25.5), 0.8–102.9
MMSE Score	23.5 (3.9), 14–30	28.6 (1.7), 24–30	<0 . 001 ^e

Mean (SD), Range; n (%); Young/EOAD: controls ≤ 65 years of age/patients with early-onset Alzheimer's disease; Old/LOAD: controls > 65 years of age/patients with late-onset Alzheimer's disease; JTHFT: Jebsen-Taylor Hand Function Test; MMSE: Mini-Mental State Examination.

Z-Scores based on norm data by Dodds et al.⁵³

Time since diagnosis according to medical records.

Two-Sample t.

Fisher‘s exact.

Wilcoxon-Mann-Whitney.

bold indicates a significant p-value.

Among the patient group, 13 patients were classified as EOAD and 11 as LOAD (see Table 3). At the time of testing, EOAD patients were approximately 15 years younger than LOAD patients, t(22) = −7.69, p < 0.001, d = 3.15. EOAD and LOAD groups did not significantly differ regarding sex, p = 0.679, OR = 1.63, 95% CI 0.22–14.28, grip strength, t(22) = −1.37, p = 0.184, and time to perform the JTHFT with the dominant hand (z = −0.61, p = 0.569). The average disease duration and MMSE score were slightly higher in LOAD than in EOAD patients, but differences were not significant (disease duration: z = −0.55, p = 0.609; MMSE: t(22) = −0.57, p = 0.574). EOAD and LOAD patients did not differ with respect to age from the matched subgroups of the control subjects (EOAD versus young (≤65) controls with mean age 57.8 years, SD 4.05, range 53–65 years, t(23) = −1.95, p = 0.06; LOAD patients versus older (> 65) controls with mean age 72.82 years, SD 6.16, range 67–84 years, t(20) = −1.67, p = 0.110).

Table 3.

Patient characteristics EOAD versus LOAD.

Characteristic	EOAD, N = 13	LOAD, N = 11	p
Age [years]	61.4 (5.2), 54–69	76.6 (4.4), 70–84	<0 . 001 ^c
Sex			0.679^d
Female	5 (38.5%)	3 (27.3%)
Male	8 (61.5%)	8 (72.7%)
Grip Strength [z]^a	−0.1 (0.8), −1.1–1.8	0.3 (0.7), −0.7–1.4	0.184^c
JTHFT Time [s]	43.5 (23.5), 27.5–118.7	41.5 (10.6), 30.0–61.4	0.569^e
Disease Duration [months]^b	22.0 (22.2), 0.8–81.3	27.1 (9.8), 1.7–102.9	0.609^e
MMSE Score	23.1 (4.6), 14–30	24.0 (3.0), 20–29	0.574^c

Mean (SD), Range; n (%); EOAD: early-onset Alzheimer's disease; LOAD: late-onset Alzheimer's disease; JTHFT: Jebsen-Taylor Hand Function Test; MMSE: Mini-Mental State Examination.

Z-Scores based on norm data by Dodds et al.⁵³

Time since diagnosis according to medical records.

Two-Sample t.

Fisher's exact.

Wilcoxon-Mann-Whitney.

bold indicates a significant p-value.

Influence of age and task on AD patients versus controls

The handwriting frequency of AD patients and controls for both age groups is shown in Figure 2 for each task. The three-way mixed ANOVA to investigate the effect of subject group, age group, and task on handwriting frequency (Table 4) revealed a significant main effect of subject group, F(1.00, 43.00) = 10.92, p = 0.002, $η_{G}^{2}$ = 0.137, with a significantly higher frequency in controls than in patients.

Figure 2.

Handwriting frequency for the Circle (a), ll's (b), and Sentence (c) task in patients with Alzheimer's disease (AD; blue) and healthy controls (green), grouped by age, with young corresponding to early-onset AD in patients and age ≤ 65 in controls and old corresponding to late-onset AD in patients and age > 65 years in controls. Boxplots were supplemented with individual data points (presented as dots).

Table 4.

Results of the three-way mixed ANOVA for handwriting frequency.

Effect	F	df	p	$η_{G}^{2}$
Subject Group	10.92	1.00, 43.00	0 . 002	0.137
Age Group	2.51	1.00, 43.00	0.121	0.035
Task	34.17	1.54, 66.07	<0 . 001	0.230
Subject Group × Age Group	0.49	1.00, 43.00	0.489	0.007
Age Group × Task	1.13	1.54, 66.07	0.318	0.010
Subject Group × Task	0.41	1.54, 66.07	0.613	0.004
Subject Group × Age Group × Task	1.96	1.54, 66.07	0.158	0.017

df: degrees of freedom; $η_{G}^{2}$ : generalized eta squared. Significant p-values are indicated in bold.

While there was no significant main effect of age group for handwriting frequency, there was a significant main effect of task F(1.54, 66.07) = 34.17, p < 0.001, $η_{G}^{2}$ = 0.230. Thereby, handwriting frequency was significantly higher in the Sentence task than in both the Circle (p < 0.001) and ll's (p < 0.001) tasks, as well as higher in the Circle than the ll's task (p < 0.001). With a nonsignificant two-way subject group × task interaction, however, the effect of task did not differ between patients and controls for handwriting frequency. The analysis revealed no significant main effect or interaction of the factor age group, indicating that both AD groups performed similarly in the tasks (see Table 4).

Subsequent z-transformations were carried out to examine whether EOAD and LOAD groups differed from their healthy counterparts across tasks regardless of the original unit of measure. AD patients’ handwriting frequency z-scores for the three handwriting tasks are illustrated in Figure 3, with the EOAD and LOAD patients’ performance on each task being standardized on the younger and older control groups’ performance, respectively. As indicated graphically, for two of the three tasks, patient groups performed approximately 1 SD or more than 1 SD worse than the corresponding control groups. Only for the Circle task, the difference was smaller. Despite these descriptive task differences, the two-way mixed ANOVA on z-standardized handwriting frequency showed neither significant main effects of age group, F(1.00, 22.00) = 0.00, p = 0.095, or task, F(2.00, 44.00) = 2.96, p = 0.062, nor an interaction between age group and task, F(2.00, 44.00) = 1.74, p = 0.118. However, overall subject group differences between AD patients and controls were apparent, with highly significant results in the one-sample t-test against zero, t(71) = −5.69, p < 0.001, d = −0.67.

Figure 3.

Z-scores for handwriting frequency in EOAD (early-onset Alzheimer's disease) and LOAD (late-onset Alzheimer's disease) patients for the three handwriting tasks. Error bars represent the standard error (SE ± 1).

Figure 4 provides an overview of the sequencing frequency of AD patients and controls grouped by age for the Tapping, Diadochokinesia, and Luria's task. The three-way mixed ANOVA to investigate the effect of subject group, age group, and task on sequencing frequency (Table 5) showed a trend for slower frequencies in the patient group, which, however, did not reach statistical significance (p = 0.088). While there was no significant main effect of age group, F(1.00, 42.00) = 0.06, p = 0.806, there was a significant main effect of task, F(1.07, 45.10) = 437.60, p < 0.001, $η_{G}^{2}$ = 0.844, with a significantly higher frequency for the Tapping than for both the Diadochokinesia (p < 0.001) and Luria's (p < 0.001) tasks and a significantly higher frequency for the Diadochokinesia task compared to the Luria's (p < 0.001). No interaction was observed between subject group and task. Further, the ANOVA yielded neither a significant interaction between subject group and age group, nor a significant three-way interaction between subject group, age group, and task (see Table 5).

Figure 4.

Sequencing frequency for the Tapping (a), Diadochokinesia (b), and the Luria's (c) task in patients with Alzheimer's disease (AD; blue) and healthy controls (green), grouped by age, with young corresponding to early-onset AD in patients and age ≤ 65 in controls and old corresponding to late-onset AD in patients and age > 65 years in controls. Boxplots were supplemented with individual data points (presented as dots).

Table 5.

Results of the three-way ANOVA for sequencing frequency.

Effect	F	df	p	$η_{G}^{2}$
Subject Group	3.06	1.00, 42.00	0.088	0.001
Age Group	0.06	1.00, 42.00	0.806	0.037
Task	437.60	1.07, 45.10	<0 . 001	0.844
Subject Group × Age Group	0.01	1.00, 42.00	0.814	0.000
Age Group × Task	0.28	1.07, 45.10	0.614	0.003
Subject Group × Task	0.01	1.07, 45.10	0.953	0.000
Subject Group × Age Group × Task	0.57	1.07, 45.10	0.464	0.007

df: degrees of freedom; $η_{G}^{2}$ : generalized eta squared. Significant p-values are indicated in bold.

With an F-value of over 400 for the main effect of task, concerns over the substantial differences in frequency between tasks negatively impacting the meaningfulness of our main analysis were particularly strong for sequencing frequency. Therefore, z-standardization was carried out. As with handwriting, the two sequencing tasks deemed more complex according to the criteria of Wood⁴⁴ resulted in performance decrements of around 1 or 1.5 SD. The difference between patients and controls was smaller for Tapping (see Figure 5).

Figure 5.

Z-scores for sequencing frequency in early-onset Alzheimer's disease (EOAD) and late-onset Alzheimer's disease (LOAD) patients for the three sequential hand movement tasks. Error bars represent the standard error (SE ± 1). **p < 0.01. ***p < 0.001.

The two-way repeated measures ANOVA yielded a significant main effect of task, with F(2.00, 44.00) = 12.94, p < 0.001, $η_{G}^{2}$ = 0.098. Post-hoc tests found a significantly lower z-score for Diadochokinesia than for Tapping (p = 0.001) and a significantly lower z-score for Tapping than for Luria's (p < 0.001), but no significant difference in z-scores between Diadochokinesia and Luria's (p = 0.125). There was no significant effect of age group, F(1.00, 22.00) = 0.181, p = 0.674, and no interaction between task and age group, F(2.00, 44.00) = 0.790, p = 0.460. In contrast to the results of the three-way ANOVA, which did not yield a significant subject group effect, the patients showed abnormal performance when their z-standardized sequencing frequency was compared with zero (controls’ means) with a one-sample t-test, t(71) = −5.58, p < 0.001, d = −0.66.

Correlation between motor performance and ADL

As for the correlation between motor performance and ADL, the time needed to complete the JTHFT was significantly and strongly correlated with both the average handwriting frequency, R = −0.94, p < 0.001 (Figure 6(a)), and the average sequencing frequency, R = −0.78, p = 0.005 (Figure 6(b)) in EOAD patients. In LOAD patients, however, the JTHFT showed no correlation with average handwriting frequency, R = −0.40, p = 0.225 (Figure 6(c)), or with average sequencing frequency, R = −0.12, p = 0.759 (Figure 6(d)). The JTHFT performance did not correlate with the MMSE score, neither in EOAD (R = −0.26, p = 0.406), nor in LOAD (R = −0.09, p = 0.788) patients.

Figure 6.

Correlation between the time needed to complete the Jebsen-Taylor Hand Function Test (JTHFT) with the dominant hand and frequency averaged over all handwriting tasks and sequencing tasks in patients with early-onset (EOAD; (a) handwriting, (b) sequencing) and late-onset (LOAD; (c) handwriting, (d) sequencing) Alzheimer's disease. One outlier (JTHFT: 118.7 s) is not displayed.

Discussion

Fine motor control deficits in AD patients

We tested two fine motor activities in a group of 24 patients with mainly mild AD and 23 healthy controls. Analysis of kinematics revealed deteriorations of performance in the patient group. This was particularly clear for handwriting, with a significantly lower handwriting frequency in AD patients than controls, as demonstrated by both the main effect of subject group in the three-way ANOVA and the z-standardized frequency data. This is consistent with existing research demonstrating a significantly lower performance in AD patients than in healthy controls for various handwriting tasks such as writing single strokes,¹⁶ letters,^16,18 circles,^16,17 and words,¹⁹ as well as sentences^16,19 and longer texts.^20,21

For sequential hand movements, the standard ANOVA revealed only a trend, while the alternative statistical approach confirmed deteriorations of performance in the patients. This finding is in line with a substantial body of literature indicating significant differences between AD patients and healthy controls for various finger tapping tasks (e.g.,^22–25), pro-and supination tasks,^22,23,27 and the Luria test.^28,29

Role of task complexity

We hypothesized that the influence of task complexity on motor performance differs between AD patients and healthy controls with a more pronounced increase in impairments in handwriting and sequential hand movements the more elaborate the task in AD patients than in controls.

Task complexity was defined according to Wood⁴⁴ based on aspects such as the number of unique subactions and coordinative requirements, especially the interplay of finger, wrist, and forearm movements as well as task duration. Applying this definition to the handwriting tasks, Circle is the least complex task, Sentence is the most complex task, and the ll's task is intermediate. Examining the main effect of task for handwriting frequency in the three-way ANOVA, a significant effect was observed, with the lowest frequency in the ll's and the highest frequency in the Sentence task. Therefore, the magnitude of the parameter frequency does not necessarily reflect complexity. Yet, our results prove frequency as a highly sensitive parameter to identify performance abnormalities in AD patients.

For handwriting, with no subject group × task interaction in the three-way ANOVA for frequency and no main effect of task in the two-way ANOVA for z-transformed frequency, we found no task-dependency of deficits in the patients. This was unexpected, given the findings of an exploratory investigation of handwriting-like movements by Yan and colleagues,¹⁷ who documented a disproportionately large increase in movement time in AD patients compared to controls when performing more complex tasks requiring coordination of finger and wrist. Likewise, results of discrimination analyses^19,30 have pointed towards a more extensive increase in deficits from easier to more elaborate tasks in AD patients compared to healthy controls, with a higher number of correct classifications in more complex tasks, indicating larger between-group differences.

Our results suggest that additional factors other than complexity may influence the performance of AD patients. Explanations for the task effects seen for handwriting frequency could be shifts in attentional control or familiarity effects. As such, writing a sentence represents a heavily overlearned and automatized activity.⁵⁷ Given that for the ll's particularly, and to some extent also for the Circle task, attention was directed to isolated components such as single letters, this specific focus may have represented a contextual change with a higher need for attentional control. This shift could also be linked with task familiarity, as healthy adults have been found to need significantly less time to perform familiar tasks, i.e., writing words and a sentence, than to perform unfamiliar tasks, i.e., writing single letters.⁴⁵ Therefore, specific demands on attention related to task familiarity in the seemingly less complex tasks of Circle and ll's may have had similar effects than higher complexity in the Sentence task.

While for sequential hand movements, the three-way ANOVA did not find a significant interaction between task and subject group, subsequent analyses of z-standardized patient data demonstrated significant task effects in line with our hypothesis, i.e., a growing divergence in the performance of AD patients from that of the control group, from Tapping to Diadochokinesia to the Luria's task. Post-hoc analyses of the task effect for z-standardized frequency identified the highest scores for Tapping, with significantly lower scores for Diadochokinesia and the lowest scores for Luria's in AD patients. This was in line with previous discriminant function analyses²³ correctly distinguishing AD patients and healthy controls for basic motor function (e.g., finger tapping) in 76.6% of cases and for complex motor function (e.g., bimanual diadochokinesia) in 92.4% of cases and thereby suggesting greater deficits with advancing task difficulty in patients compared to controls. Nevertheless, the difference between Diadochokinesia and Luria's was not significant, which was surprising given that based on factors such as the number and sequence of movements and task duration, we would have deemed the Luria's task to be significantly more demanding than the Diadochokinesia task.

Role of age at onset

As for the role of age at onset, we hypothesized that the deficits of EOAD patients compared to age-matched young controls would be greater than those of LOAD patients compared to age-matched older controls. With no interactions between age group and subject group in the three-way ANOVAs for frequency and no significant main effect of age group in the two-way repeated measures ANOVAs on AD patients’ frequency z-scores, this could not be confirmed, neither for handwriting nor for sequential hand movements.

Research on motor impairments in EOAD versus LOAD is limited and primarily compares EOAD and LOAD patients without considering age differences, resulting in inconclusive findings. For finger tapping and bimanual hand pro-and supination, Sá and colleagues³¹ identified no differences between EOAD and LOAD patients. Wallin & Blennow³⁵ found bipyramidal signs, such as elevated tendon reflexes and a positive Babinski reflex, more frequently in EOAD than LOAD patients; for extrapyramidal signs, such as rigidity, tremor, and hypokinesia, however, the authors³⁵ reported no significant between-group differences. A more recent study³⁴ focused on tremor, rigidity, bradykinesia, and axial features as motor symptoms among EOAD and LOAD patients with a rather extensive assessment of bradykinesia, i.e., finger and foot tapping, hand pro- and supination, and axial features, i.e., posture and gait. Although there was a significant age difference between EOAD and LOAD patients with an average age of 56 years and 72.4 years, respectively, no between-group differences in the frequency of motor symptoms for axial features, tremor, rigidity, or bradykinesia could be detected.³⁴ Along with the absence of effects in our study, it seems likely that the lack of differences between EOAD and LOAD patients in these previous studies^31,34,35 cannot be attributed to the large age differences between patient groups. We therefore conclude that age at onset, i.e., EOAD versus LOAD may not critically influence fine motor deficits in AD patients.

Further evidence supporting this conclusion is provided by Licht and colleagues⁵⁸; while they demonstrated significant differences between EOAD and LOAD patients for the Luria test, with better performance in EOAD than LOAD patients, these differences disappeared as soon as age was included as a covariate. It should be noted, however, that they only included LOAD patients with an age at onset of at least 84 years.⁵⁸

Different from simple motor tasks and the Luria test, apraxia seems to be more prominent in EOAD patients when compared with LOAD patients with similar dementia severity. For instance, Suribhatla and colleagues³² observed a significantly worse praxis in EOAD than in LOAD patients, and Koedam and colleagues³³ even identified apraxia as the most frequent non-memory deficit, occurring significantly more frequently in the EOAD than in the LOAD group.

While our results of lacking performance differences between EOAD and LOAD support the conclusion that fine motor deficits are similar in both groups if age and general severity of dementia are controlled for, other explanations should nevertheless be considered. For instance, the absence of effects of age at onset might also be due to different symptom and disease durations in EOAD versus LOAD. With no documentation of the onset of first symptoms in the patients of the present research, symptom length could not be collected; instead, the more objective measure of disease duration as the time since formal diagnosis was utilized. Disease duration was slightly longer in the LOAD patients (M 27.1 months) than in the EOAD patients (M 22.0 months), which would fit this idea. Yet, the difference between the groups was not statistically significant, and importantly also the MMSE score did not differ between the AD subgroups.

The large heterogeneity of EOAD could be another reason for not detecting more severe motor deficits in young patients. For example, there is evidence^59,60 indicating that only certain forms of EOAD, such as autosomal dominant AD, accounting for around 10% of EOAD cases⁶ are associated with greater motor impairments.

Finally, the severity of motor symptoms in both groups may diverge in more advanced stages of cognitive impairment, which was relatively mild in both AD groups of the present study as reflected by the MMSE scores.

Relationship between motor function and ADL

Interestingly, motor performance in handwriting and sequential hand movements was strongly correlated with the time to carry out the JTHFT only in EOAD but not in LOAD patients. Thus, even though age at onset had no influence on handwriting and sequencing performance or on the time required to execute the JTHFT (see Table 3), the correlation analyses indicate differences between EOAD and LOAD, as only in EOAD both handwriting and sequencing frequency averaged over all tasks was predictive of ADL performance. One possible interpretation of this particular observation is that in EOAD, deficits in the coordination of sequential hand movements of various complexities are to a relevant degree responsible for deteriorations in manual activities of daily living, while in LOAD, such an association is absent or less direct.

We are unaware of other research examining the relationship between fine motor control and ADL performance in EOAD and LOAD specifically. However, a previous study including AD patients, but not providing details on age or age at onset found fine motor dexterity assessed via the Nine-Hole Peg Test to be a significant predictor of self-care ADL obtained using the General Activities of Daily Living Scale administered via caregiver interviews.⁴¹ While de Paula et al.⁴¹ did not find an association between fine motor control and domestic as well as complex ADL functions as measured with the General Activities of Daily Living Scale, a more recent study⁴² reported such a relationship in MCI patients aged 60–79 (M 68.38 SD 4.50) years with fine motor control measured via the Purdue Pegboard Test.

Notably, different from fine motor performance, the MMSE score was not able to predict the performance in the JTHFT, neither in EOAD, nor in LOAD patients. The relatively low variance of MMSE scores in our mainly mildly affected AD patients may contribute to this failure. Thus, our results also highlight limitations of the MMSE score.

It must be acknowledged that not capacity measures like speed of performance in the JTHFT, but rather measures like smoothness, agility, intensity, and volume of movements may be more suitable to detect relevant (in)dependencies in the performance of daily activities.⁶¹ Nevertheless, Gulde et al.⁶² showed in a group of AD patients that parameters derived from a timed motor task (variant of Trail-Making Test A) were indeed able to predict the performance in two ADL tasks (tea making and document filing) that were executed at self-selected speed.

Limitations

As for our study sample, the lack of an a priori sample size estimation must be acknowledged. With 13 EOAD patients and 12 young controls as well as 11 LOAD patients and 11 older controls, the sample may have been too small to detect significant effects. Some results of the present study, such as the lack of a main effect of group for the sequential movement and the lack of a task effect for the handwriting tasks may indeed be due to the sample size.

Additionally, we recognize that our alternative analyses of z-standardized frequency data were performed after the results of our main analysis, i.e., the three-way mixed ANOVAs, were known, thereby restricting generalizability of our findings.

The current research is further limited by the lack of data on participants’ educational background, occupation, medication use, and trial participation. It cannot be excluded that variations in educational background impacted task performance, however, previous research in AD found no evidence for a correlation between years of education and handwriting kinematics in AD patients.¹⁶ As for sequencing, Weiner and colleagues²⁸ yielded no educational differences between those who could perform the Luria test correctly and those who could not in a sample of healthy controls, AD, MCI, and frontotemporal dementia patients. Medication use and trial designation were not recorded in our study participants. As for medication, studies of handwriting in AD that included patients’ medications, particularly psychotropics, found no influence of these medications on handwriting performance.^18,63 In contrast, an association between AD medications and ADL performance has already been established; both memantine⁶⁴ and acetylcholinesterase inhibitors⁶⁵ can benefit ADL performance among AD patients. In addition to stable medications, novel treatments were provided within the context of the clinical trials from which our patients were recruited.

The significant sex differences between AD patients (8 females, 16 males) and controls (16 females, 7 males) could have potentially biased results. Gender effects were reported for some of the included tasks with studies demonstrating males to perform faster than females in finger tapping^48,49,66 and forearm diadochokinesia^47,48 as well as handwriting tasks like letter sequences⁴⁶ in adults without history of neurological disorders. However, this was not consistent, with a significantly higher performance in females than males reported for finger tapping⁶⁷ and other studies yielding no significant gender effects for finger tapping⁶⁸ and various handwriting tasks^69,70 in samples of healthy adults. Importantly, a male advantage, as reported in most studies that found gender effects, would have biased the results into the direction of better performance in the patient group, with relatively more male participants. Thus, an overestimation of group differences can be largely excluded. Rather, the lacking main effect of group in the sequencing tasks could be due to the particularity of our sample. Finally, the sex ratio was almost equal in the two subgroups of AD patients so that the corresponding results were not biased by sex effects.

Conclusion

This study aimed to elucidate the influence of task complexity and age at onset on hand motor function of patients with dementia due to AD and whether hand motor function can predict ADL performance in this patient group. Regarding complexity, a more substantial decrease in performance in AD patients as compared to controls from simpler to more complex tasks was seen only for z-standardized frequency data during sequential hand movements, but not for handwriting. Significant differences between AD patients and healthy controls for handwriting frequency as well as z-standardized handwriting and sequencing frequency were detected, with an overall significantly lower frequency in AD patients. Interestingly, the deficits in both activities were comparable, as can be gathered from the similar z-scores in the more complex task versions, even though handwriting is an overlearned motor skill, while sequencing consists of more basic (pro/supination) or newly learned tasks (Luria's). This finding points to a certain generalization of deficits within the fine motor domain. In general, kinematic analysis proved to be an appropriate tool for detecting hand motor impairment in AD. It could thereby represent a practical resource in the clinical assessment and diagnosis of AD. When deficits like these are observed, stimulation of fine motor skills may be beneficial. As such, evidence suggests that a structured training program including upper limb exercises over several weeks can enhance fine motor function in individuals with MCI.⁷¹

Age at onset of AD did not impact hand motor function in our sample, neither for handwriting nor for sequencing. We recommend a further investigation of potential age effects, ideally in a sample comprising more individuals and including a larger age range, and by considering the heterogeneity of EOAD and conducting adequate subgroup analyses. This may provide a better understanding of the role of hand motor impairments in EOAD and the extent to which motor testing may be beneficial in the diagnosis and disease monitoring of EOAD specifically. Nevertheless, differences between EOAD and LOAD were seen for the predictability of ADL performance as hand motor function during handwriting and sequential hand movements could predict ADL performance, quantified via time needed to carry out the JTHFT with the dominant hand only in EOAD and not in LOAD patients. Future research on this topic might also benefit from a more detailed investigation of the relationship between hand motor function and ADL performance utilizing performance-based measures requiring bimanual coordination, with a greater emphasis on parameters not solely reflecting speed but also smoothness of movement.

Footnotes

Acknowledgements

The authors would like to thank On Ki Chong, Korbinian Hannes, Han-Yu Shih, Laura Wild, Denise Roman, and Martina Roosz for their help in data collection. We would also like to extend our gratitude to all participants for taking part in the study.

ORCID iDs

Laura Stetter

Lucas Wolski

Ethical considerations

The study was approved by the Ethics Commission of TUM (Reference No. 479/19 S-SR) and performed in accordance with the ethical standards specified in the Declaration of Helsinki.

Consent to participate

All participants provided written informed consent.

Consent for publication

Not applicable

Author contribution(s)

Laura Stetter: Formal analysis; Investigation; Visualization; Writing – original draft.

Lucas Wolski: Formal analysis; Resources; Writing – original draft.

Timo Grimmer: Project administration; Resources; Supervision; Writing – review & editing.

Janine Diehl-Schmid: Conceptualization; Project administration; Resources; Writing – review & editing.

Joachim Hermsdörfer: Conceptualization; Data curation; Formal analysis; Funding acquisition; Methodology; Project administration; Resources; Software; Supervision; Writing – original draft.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was part of the Innovation Network eXprt of the Technical University of Munich, funded by the Federal Ministry of Education and Research and the Free State of Bavaria under the Excellence Strategy of the Federal Government and the States.

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: L.S. declares no conflicting interest. L.W. declares no conflicting interest. J.D-S declares no conflicting interest. J.H. declares no conflicting interest. T.G. declares no conflicting interest with regards to the submitted work. Outside the submitted work T.G. reported receiving consulting fees from Acumen, Advantage Ther, Alector, Anavex, Biogen, BMS; Cogthera, Eisai, Eli Lilly, Functional Neuromod., Grifols, Janssen, Neurimmune, Noselab, Novo Nordisk, Roche Diagnostics, and Roche Pharma; lecture fees from Cogthera, Eisai, Eli Lilly. FEO, Grifols, Pfizer, Roche Pharma, Schwabe, and Synlab; and has received grants to his institution from Biogen and Eisai.

Data availability statement

Due to confidentiality, the data generated and analyzed within this study are not publicly accessible but will be made available by the corresponding author upon reasonable request.

References

Nichols

Steinmetz

Vollset

, et al. Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the Global Burden of Disease Study 2019. Lancet Public Health 2022; 7: e105–e125.

Institute for Health Metrics and Evaluation. Alzheimer's disease and other dementias—Level 3 cause, https://www.healthdata.org/research-analysis/diseases-injuries-risks/factsheets/2021-alzheimers-disease-and-other-dementias (2024, accessed 28 January 2025).

World Health Organization . International classification of diseases for mortality and morbidity statistics. 11th Revision. Geneva: World Health Organization, 2019, https://icd.who.int/ .

Alzheimer’s Association . 2024 Alzheimer's disease facts and figures. Alzheimers Dement 2024; 20: 3708–3821.

Porsteinsson

Isaacson

Knox

, et al. Diagnosis of early Alzheimer’s disease: clinical practice in 2021. J Prev Alzheimers Dis 2021; 8: 371–386.

Sirkis

Bonham

Johnson

, et al. Dissecting the clinical heterogeneity of early-onset Alzheimer’s disease. Mol Psychiatry 2022; 27: 2674–2688.

Valdez-Gaxiola

Rosales-Leycegui

Gaxiola-Rubio

, et al.

Early- and late-onset Alzheimer’s disease: two sides of the same coin?

Diseases 2024; 12: 110.

Natale

Wohlrab

Förtsch

, et al. Alzheimer-Demenz mit präsenilem Beginn – Besonderheiten in Diagnostik, Therapie und Management. In: Förstl

(eds) Demenzen in Theorie und Praxis. Berlin, Heidelberg: Springer, 2011, pp.73–92.

McKhann

Drachman

Folstein

, et al. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA work group under the auspices of department of health and human services task force on Alzheimer's disease. Neurology 1984; 34: 939–944.

10.

Pettersson

Olsson

Wahlund

L-O

. Motor function in subjects with mild cognitive impairment and early Alzheimer’s disease. Dement Geriatr Cogn Disord 2005; 19: 299–304.

11.

Cui

Zhang

Liu

, et al. Grip strength and the risk of cognitive decline and dementia: a systematic review and meta-analysis of longitudinal cohort studies. Front Aging Neurosci 2021; 13: 625551.

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.

Wisniewski

Masurkar

. Gait dysfunction in Alzheimer disease. Handb Clin Neurol 2023; 196: 267–274.

14.

Impedovo

Pirlo

. Dynamic handwriting analysis for the assessment of neurodegenerative diseases: a pattern recognition perspective. IEEE Rev Biomed Eng 2019; 12: 209–220.

15.

van Galen

. Handwriting: issues for a psychomotor theory. Hum Mov Sci 1991; 10: 165–191.

16.

Delazer

Zamarian

Djamshidian

. Handwriting in Alzheimer’s disease. J Alzheimers Dis 2021; 82: 727–735.

17.

Yan

Rountree

Massman

, et al. Alzheimer’s disease and mild cognitive impairment deteriorate fine movement control. J Psychiatr Res 2008; 42: 1203–1212.

18.

Slavin

Phillips

Bradshaw

, et al. Consistency of handwriting movements in dementia of the Alzheimer's type: a comparison with Huntington's and Parkinson's diseases. J Int Neuropsychol Soc 1999; 5: 20–25.

19.

Zhang

Wei

, et al. A study of auxiliary screening for Alzheimer's disease based on handwriting characteristics. Front Aging Neurosci 2023; 15: 1117250.

20.

Werner

Rosenblum

Bar-On

, et al. Handwriting process variables discriminating mild Alzheimer's disease and mild cognitive impairment. J Gerontol B Psychol Sci Soc Sci 2006; 61: 228–236.

21.

Chernov

. Handwriting markers for the onset of Alzheimer’s disease. Curr Alzheimer Res 2024; 20: 791–801.

22.

Franssen

Somen

LEM

Torossian

, et al. Equilibrium and limb coordination in mild cognitive impairment and mild Alzheimer's disease. J Am Geriatr Soc 1999; 47: 463–469.

23.

Kluger

Gianutsos

Golomb

, et al. Patterns of motor impairment in normal aging, mild cognitive decline, and early Alzheimer's disease. J Gerontol B Psychol Sci Soc Sci 1997; 52B: 28–39.

24.

Koppelmans

Ruitenberg

MFL

Schaefer

, et al. Delayed and more variable unimanual and bimanual finger tapping in Alzheimer's disease: associations with biomarkers and applications for classification. J Alzheimers Dis 2023; 95: 1233–1252.

25.

Roalf

Rupert

Mechanic-Hamilton

, et al. Quantitative assessment of finger tapping characteristics in mild cognitive impairment, Alzheimer’s disease, and Parkinson’s disease. J Neurol 2018; 265: 1365–1375.

26.

Goldman

Baty

Buckles

, et al. Motor dysfunction in mildly demented AD individuals without extrapyramidal signs. Neurology 1999; 53: 956–962.

27.

Jeppesen Kragh

Bruun

Budtz-Jørgensen

, et al. Quantitative measurements of motor function in Alzheimer’s disease, frontotemporal dementia, and dementia with Lewy bodies: a proof-of-concept study. Dement Geriatr Cogn Disord 2018; 46: 168–179.

28.

Weiner

Hynan

Rossetti

, et al.

Luria's three-step test: what is it and what does it tell us?

Int Psychogeriatr 2011; 23: 1602–1606.

29.

Gabriel

Lehner

Höhler

, et al. Anticipatory and reactive grip force control in patients with Alzheimer’s disease: a pilot study. J Alzheimers Dis 2021; 82: 1651–1665.

30.

Garre-Olmo

Faúndez-Zanuy

López-de-Ipiña

, et al. Kinematic and pressure features of handwriting and drawing: preliminary results between patients with mild cognitive impairment, Alzheimer disease and healthy controls. Curr Alzheimer Res 2017; 14: 960–968.

31.

Sá

Pinto

Cunha

, et al. Differences between early and late-onset Alzheimer’s disease in neuropsychological tests. Front Neurol 2012; 3: 81.

32.

Suribhatla

Baillon

Dennis

, et al. Neuropsychological performance in early and late onset Alzheimer's disease: comparisons in a memory clinic population. Int J Geriatr Psychiatry 2004; 19: 1140–1147.

33.

Koedam

ELGE

Lauffer

van der Vlies

, et al. Early-versus late-onset Alzheimer's disease: more than age alone. J Alzheimers Dis 2010; 19: 1401–1408.

34.

Kim

Jang

Park

Y-H

, et al. Motor symptoms in early- versus late-onset Alzheimer’s disease. J Alzheimers Dis 2023; 91: 345–354.

35.

Wallin

Blennow

. Neurologic motor signs in early and late onset Alzheimer's disease. Dementia 2010; 3: 314–319.

36.

Scherder

Dekker

Eggermont

. Higher-level hand motor function in aging and (preclinical) dementia: its relationship with (instrumental) activities of daily life – a mini-review. Gerontology 2008; 54: 333–341.

37.

van der Willik

Licher

Vinke

, et al. Trajectories of cognitive and motor function between ages 45 and 90 years: a population-based study. J Gerontol A Biol Sci Med Sci 2021; 76: 297–306.

38.

Desai

Grossberg

Sheth

. Activities of daily living in patients with dementia. CNS Drugs 2004; 18: 853–875.

39.

Luttenberger

Reppermund

Schmiedeberg-Sohn

, et al. Validation of the Erlangen test of activities of daily living in persons with mild dementia or mild cognitive impairment (ETAM). BMC Geriatr 2016; 16: 111.

40.

Jekel

Damian

Wattmo

, et al. Mild cognitive impairment and deficits in instrumental activities of daily living: a systematic review. Alzheimers Res Ther 2015; 7: 17.

41.

de Paula

Albuquerque

Lage

, et al. Impairment of fine motor dexterity in mild cognitive impairment and Alzheimer’s disease dementia: association with activities of daily living. Braz J Psychiatry 2016; 38: 235–238.

42.

Rattanawan

. Correlations between hand dexterity and bimanual coordination on the activities of daily living in older adults with mild cognitive impairment. Dement Geriatr Cogn Dis Extra 2022; 12: 24–32.

43.

Scarmeas

Albert

Brandt

, et al. Motor signs predict poor outcomes in Alzheimer disease. Neurology 2005; 64: 1696–1703.

44.

Wood

. Task complexity: definition of the construct. Organ Behav Hum Decis Process 1986; 37: 60–82.

45.

Dixon

Kurzman

Friesen

. Handwriting performance in younger and older adults: age, familiarity, and practice effects. Psychol Aging 1993; 8: 360–370.

46.

Mergl

Tigges

Schröter

, et al. Digitized analysis of handwriting and drawing movements in healthy subjects: methods, results and perspectives. J Neurosci Methods 1999; 90: 157–169.

47.

Hermsdörfer

Marquardt

Wack

, et al. Comparative analysis of diadochokinetic movements. J Electromyogr Kinesiol 1999; 9: 283–295.

48.

Jiménez-Jiménez

Calleja

Alonso-Navarro

, et al. Influence of age and gender in motor performance in healthy subjects. J Neurol Sci 2011; 302: 72–80.

49.

Shimoyama

Ninchoji

Uemura

. The finger-tapping test: a quantitative analysis. Arch Neurol 1990; 47: 681–684.

50.

McKhann

Knopman

Chertkow

, et al. The diagnosis of dementia due to Alzheimer's disease: recommendations from the national institute on aging-Alzheimer's association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement 2011; 7: 263–269.

51.

Folstein

McHugh

. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189–198.

52.

Oldfield

. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 1971; 9: 97–113.

53.

Dodds

Syddall

Cooper

, et al. Grip strength across the life course: normative data from twelve British studies. PLoS One 2014; 9: e113637.

54.

Maitland

. A tool to calculate grip strength z score, https://github.com/StuMaitland/grip_strength.git (2018, accessed 28 January 2025).

55.

Jebsen

Taylor

Trieschmann

, et al. An objective and standardized test of hand function. Arch Phys Med Rehabil 1969; 50: 311–319.

56.

Marquardt

Mai

. A computational procedure for movement analysis in handwriting. J Neurosci Methods 1994; 52: 39–45.

57.

Marquardt

C. Schreibanalyse.

Nowak

In: DA

(ed.) Handfunktionsstörungen in der Neurologie: Klinik und Rehabilitation. Berlin, Heidelberg: Springer, 2011, pp.379–394.

58.

Licht

McMurtray

Saul

, et al. Cognitive differences between early- and late-onset Alzheimer's disease. Am J Alzheimers Dis Other Demen 2007; 22: 218–222.

59.

Llibre-Guerra

Iaccarino

Coble

, et al. Longitudinal clinical, cognitive and biomarker profiles in dominantly inherited versus sporadic early-onset Alzheimer’s disease. Brain Commun 2023; 5: fcad280.

60.

Vöglein

Paumier

Jucker

, et al. Clinical, pathophysiological and genetic features of motor symptoms in autosomal dominant Alzheimer’s disease. Brain 2019; 142: 1429–1440.

61.

Schmidle

Gulde

Herdegen

, et al. Kinematic analysis of activities of daily living performance in frail elderly. BMC Geriatr 2022; 22: 244.

62.

Gulde

Leippold

Kohl

, et al. Step by step: kinematics of the reciprocal trail making task predict slowness of activities of daily living performance in Alzheimer’s disease. Front Neurol 2018; 9: 140.

63.

Schröter

Mergl

Bürger

, et al. Kinematic analysis of handwriting movements in patients with Alzheimer’s disease, mild cognitive impairment, depression and healthy subjects. Dement Geriatr Cogn Disord 2003; 15: 132–142.

64.

McShane

Westby

Roberts

, et al. Memantine for dementia. Cochrane Database Syst Rev 2019; 3: CD003154.

65.

Zhang

Y-H

Zhang

, et al. Meta-analysis of randomized controlled trials on the efficacy and safety of donepezil, galantamine, rivastigmine, and memantine for the treatment of Alzheimer’s disease. Front Neurosci 2019; 13: 472.

66.

Ruff

Parker

. Gender- and age-specific changes in motor speed and eye-hand coordination in adults: normative values for the finger tapping and grooved pegboard tests. Percept Mot Skills 1993; 76: 1219–1230.

67.

Sebastjan

Skrzek

Ignasiak

, et al. The effects of age and sex on hand movement structure. Aging Clin Exp Res 2017; 29: 1221–1229.

68.

McElvogue

Steffes

Burke

, et al. 56 Predictors of finger tapping variability in older adults evaluated for a neurodegenerative memory disorder. J Int Neuropsychol Soc 2023; 29: 567–568.

69.

Burger

McCluskey

. Australian Norms for handwriting speed in healthy adults aged 60–99 years. Aust Occup Ther J 2011; 58: 355–363.

70.

Rosenblum

Werner

. Assessing the handwriting process in healthy elderly persons using a computerized system. Aging Clin Exp Res 2006; 18: 433–439.

71.

Jiang

Chen

Wang

, et al. An investigation of limbs exercise as a treatment in improving the psychomotor speed in older adults with mild cognitive impairment. Brain Sci 2019; 9: 277.