Anticipatory and Reactive Grip Force Control in Patients with Alzheimer’s Disease: A Pilot Study

Abstract

Background:

Alzheimer’s disease (AD) affects several cognitive functions and causes altered motor function. Fine motor deficits during object manipulation are evident in other neurological conditions, but have not been assessed in dementia patients yet.

Objective:

Investigate reactive and anticipatory grip force control in response to unexpected and expected load force perturbation in AD.

Methods:

Reactive and anticipatory grip force was investigated using a grip-device with force sensors. In this pilot study, fifteen AD patients and fourteen healthy controls performed a catching task. They held the device with one hand while a sandbag was dropped into an attached receptacle either by the experimenter or by the participant.

Results:

In contrast to studies of other neurological conditions, the majority of AD patients exerted lower static grip force levels than controls. Interestingly, patients who were slow in the Luria’s three-step test produced normal grip forces. The timing and magnitude of reactive grip force control were largely preserved in patients. In contrast, timing and extent of anticipatory grip forces were impaired in patients, although anticipatory control was generally preserved. These deficits were correlated with decreasing Mini-Mental State Examination scores. Apraxia scores, assessed by pantomime of tool-use, did not correlate with performance in the catching task.

Conclusion:

We interpreted the decreased grip force in AD in the context of loss of strength and lethargy, typical for patients with AD. The lower static grip force during object manipulation may emerge as a potential biomarker for early stages of AD, but more studies with larger sample sizes are necessary.

Keywords

Alzheimer’s disease dementia early diagnosis fine motor control hand strength

INTRODUCTION

Alzheimer’s disease (AD) is the most common dementia syndrome. It is a slowly progressive disorder in which pathophysiological changes precede clinical symptoms by many years [1]. There is growing evidence that increases in amyloid-β may be one of the triggers for AD setting off a chain of events that includes the accumulation of toxic forms of tau that eventually cause downstream neurodegeneration and dementia [2, 3]. Motor impairments in dementia are often overlooked because cognitive decline is more obvious, especially memory impairments. However, epidemiological studies of various aspects of motor control in cognitively healthy elderly cohorts have shown that subtle motor impairments are risk factors for imminent progression to AD and might serve as an early predictor [4 –6].

Apraxia is a common higher-order motor dysfunction characterized by the inability to correctly perform skilled limb movements, which is not due to sensorimotor or general cognitive deficits [7, 8]. Apraxic deficits can already be detected in patients with mild cognitive impairment (MCI) and become more frequent and severe in patients with AD [9]. AD patients fail typical apraxia tests, such as imitation of gestures or pantomime of tool use [9 –12] and make errors when manipulating familiar or novel tools [10 , 14]. AD patients may also present motor deficits in more elementary upper extremity motor tasks, such as finger tapping [15, 16], rapid pro- and supination movements [17], alternating bimanual finger movements [18], or aiming movements [19]. In addition, sequential hand movements, as performed in the Luria’s three-step test (fist-edge-palm) [20], or alternating thumb-to-fingers movements [17] may be particularly vulnerable in patients with dementia, compared to healthy control subjects. Furthermore, fine motor tasks such as repetitive grasping and placing of pegs on a pegboard [21] and handwriting analyses [22, 23] revealed deficits in both patients with MCI and AD.

The above-mentioned deficits in arm and hand function may contribute to impairments in basic and instrumental activities of daily living (ADL), which are present in AD patients [24 –26]. Apraxia exacerbates deficits in ADL [27]. In particular, patients suffering from both dementia and apraxia exhibit problems with correct tool use [28]. We have previously shown that performance deficits of AD patients in typical ADL can be partially predicted by kinematic analysis of goal-directed arm movements during a trail making task [29].

An elegant method to simultaneously assess elementary motor functions, dexterity, and anticipatory motor control is to analyze the coordination of fingertip-positions, forces, and torques during object manipulation [30]. Adequate control of finger movements and finger forces is essential for skilled tool use and fluent ADL performance [31]. Dexterous manipulation is characterized by grip forces (GF) that are accurately adjusted to the weight, fragility, and surface condition of the object to prevent slippage of the object, as well as damage to the grasped object and fatigue [32, 33]. In a previous series of studies we have shown that the analysis of finger forces in a catching task is a sensitive method to reveal deficits in reactive and anticipatory control of GF. In this task, a loading perturbation of a handheld object is induced by either the participants or the experimenter dropping a sandbag into a receptacle attached to a handheld experimental object. Patients with neurological disorders of the central nervous system (CNS), such as multiple sclerosis, cerebellar degeneration, and basal ganglia dysfunction, and patients with peripheral sensory neuropathy, produced higher static GF than healthy controls during the task [34 –36]. In contrast, a series of studies observed lower GF safety margins in patients with diabetes mellitus with and without polyneuropathy than in healthy controls [37 –39]. From these observations, we concluded [40] that increases in GF may be a general response to disturbance of the sensory and motor system of the CNS. Impaired force control due to incoordination, dystonia, or spasticity may lead to increases in GF to reduce the risk of object loss. In contrast, peripheral sensory neuropathies may initially lead to a reduced GF [37 –39] response and only in later stages to an exaggerated GF when holding an object [41, 42]. In addition to exaggerated GF, patients with cerebellar degeneration also failed to anticipate self-induced load perturbation by increasing GF in an anticipatory manner [35]. In contrast, no significant differences were found in stroke survivors compared to controls [31].

In the population of older adults, studies have shown a relationship between cognitive and physical functions, especially for physical functions of the lower limb [43]. Studies investigating the relationship between grip strength and cognitive decline yielded contradictory results. While some studies proposed that weaker grip strength was associated with cognitive decline and onset of dementia [44 –46], other studies found no such association [47, 48]. A recent meta-analysis confirmed that poorer grip strength was associated with a higher risk of both, cognitive decline and onset of dementia [49]. However, maximum whole-hand grip strength and the fine motor control of fingertip forces during object manipulation are fundamentally different aspects of motor control, precluding the generalization of findings related to grip strength. Whether fine motor control of digit forces in object manipulation is affected by cognitive decline has not been investigated, but may be important for subject’s ability to remain independent in daily activities. Anticipatory fine motor fingertip force control in dementia patients has been investigated in a pilot study. The authors showed that both MCI (n = 8) and AD (n = 8) patients were less successful than controls in learning arbitrary weight cues to scale fingertip forces [50]. However, control of fingertip forces during dynamic changes in load and object properties has not been investigated in AD patients yet and it is unknown whether and how these aspects of fine motor function are affected.

The aim of the present study was to investigate reactive and anticipatory modes of GF control in response to unexpected and expected load force perturbation in AD patients. AD patients and age-matched healthy controls had to hold an object stationary while a sandbag was dropped into a receptacle attached to the object either unexpectedly by the experimenter or expectedly by the subject. We hypothesized to find exaggerated GF and deficits of anticipatory control in AD patients similar to the considerations and findings in patients with multiple sclerosis, cerebellar degeneration, and basal ganglia dysfunction mentioned above [34 –36]. Furthermore, we aimed to investigate the association of performance measures in the catching task with overall cognitive impairment, performance on the Luria’s three-step test, and symptoms of apraxia in AD patients.

METHODS

Participants

Due to a lack of comparable studies with dementia patients, the a-priori sample size calculation using G^*Power (G^*Power version 3.1., Heinrich-Heine-University Düsseldorf, Germany) [51] was based on the standardized effect size of 0.93 (Cohen’s d) for maximal GF, which was reported in a previous study with patients with multiple sclerosis [36]. With an addition of two participants per group to account for unforeseen events and with α= 0.05 and β= 0.8, the targeted sample size was 36 participants, with 18 participants in each group.

We could only recruit 29 participants, including 15 people with a diagnosis of dementia and 14 healthy controls. Therefore, this study is at risk of being underpowered and should be considered as an exploratory pilot study with the need to conduct further studies with larger sample sizes.

Patients were recruited from the memory clinic at the tertiary care hospital Klinikum Rechts der Isar (MRI) of the Technical University of Munich (TUM). The control group consisted of fourteen community dwelling seniors without memory complaints who scored in the normal range on the Mini-Mental State Examination (MMSE) (≥24).

All participants but one patient were right-handed according to the Edinburgh Handedness Inventory [52] (see Table 1) and had no concomitant other neurological diseases or musculoskeletal disorders of the involved upper limb, as assessed by a review of medical records and a structured questioning of the participants. They were not informed in advance of the purpose of the study. The study was conducted in accordance with the Declaration of Helsinki and approved by the local ethics board at the School of Medicine of TUM. Written informed consent was obtained from all participants. The study information was carefully reviewed with the AD patients, and all participants but one were able to provide informed consent themselves. In one case, informed consent was provided by a proxy.

Table 1

Demographics and results of the patients with Alzheimer's disease and the control group

Code	Gender	Age	Dementia	Other disease	Handed-ness	Medication	MMSE
D02	male	80	mild AD	–	right	Lithium	26
D03	female	74	mild AD &vascular pathology	–	right	–	24
D04	female	85	mild AD &vascular pathology	mild depr., migraine	right	–	25
D05	male	81	mild AD	mild depr., chr. headache	right	–	25
D06	female	69	mild AD	–	right	–	24
D07	male	62	mild AD	–	left	Quetiapin 25 mg	26
D08	male	62	mild AD	–	right	–	18
D09	female	80	mild AD	mild to mod. depr.	right	Quetiapin 12.5 mg	25
D10	female	53	mild AD	–	right	–	17
D11	female	79	mild AD	mild depr., vertigo	right	–	26
D12	female	72	moderate AD	–	right	–	19
D14	male	80	moderate AD	–	right	–	22
AD patients: mean ± SD
N = 12	7 female 5 male	73.1±9.8					23.1±3.3
N = 14	7 female 7 male	70.3±6.9					28.6±1.0

Other disease, other psychiatric or neurologic diseases (depr.: depression, chr.: chronic); AD, Alzheimer’s disease; MMSE, Mini-Mental State Examination; SD, standard deviation.

Experimental procedure

All tests were conducted in a single measurement session that lasted approximately 60 min. Participants performed all tests with their dominant hand. To avoid symptoms of fatigue, we allowed participants to take breaks on request. Assessments were performed in order of appearance.

MMSE

At the beginning of the experimental session, we administered the MMSE, a commonly used screening test for dementia, to assess general cognitive performance, including orientation, registration, attention and calculation, recall and language. A total score of 30 can be reached. The first part of the test requires verbal response only (maximum score of 21) and the second part tests the capability to name and follow verbal as well as written commands [53].

Pantomime of tool use

To detect symptoms of apraxia, participants were asked to pantomime the use of ten tools presented as photographic images [54]. Videos of the performance were recorded and scored to determine whether characteristic features of the pantomimes were present, as described by Goldenberg et al. [54]. A maximum of 24 points could be achieved and the cut-off value for apraxia was less than 22 points.

Luria’s three-step test

The Luria-three step test sequence is recommended to be included in dementia assessments to distinguish between normal elderly and patients with early stages of AD [20]. It examines the integrity of the action production system as well as motor coordination functions. Patients were requested to perform the hand movement sequence ‘fist-edge-palm’ as quickly as possible following a demonstration by one examiner. The time in seconds from the start to the end of the sequence was measured by a second examiner with a stopwatch and the execution was recorded by video. The errors were counted independently by both examiners via analyzing the video and the results were compared. If the results differed, the sequence was analyzed again to control for potential errors in the analysis. If differences persisted, the mean value of both results was taken. As recommended in prior studies, three trials were performed and the time to complete the sequence was averaged over trials [20 , 55].

Catching task

We tested reactive and predictive control of GF during object manipulation in a catching task (see Fig. 1). Participants held a wireless instrumented device (GF-Box, Messtechnik Andreas Häußler, GE, dimension 71×57×22 mm, weight 220 g) with the thumb (D1) and three fingers (D2-D4) in opposition. The grip surfaces (71×52 mm) were covered with sandpaper (grit 320). The GF-Box incorporated two force sensors (MAK 177, Rieger Sensortechnik, GE, range 0 –100 N, accuracy±0.1 N) measuring the GF, i.e., the orthogonal force exerted against the respective grip surface, and the load force, which is measured along the axis of the suspension (range±50 N, accuracy±0.1 N). In addition, the device was equipped with a 3D-acceleration sensor (AIS326DQ, ST Microelectronics GE, range±3 m/s², accuracy±0.02 m/s²). Signals were recorded offline with a frequency of 128 Hz and transmitted via Bluetooth to a computer for further processing with a custom-made software (GFWin, MedCom, GE). A receptacle was mounted 35 cm below the GF-Box and a plastic disk 15 cm above the receptacle defined the upper position of the sandbag before release (see Fig. 1).

Fig. 1

Apparatus used for the catching task. The instrumented handle measured grip- and load forces. A sandbag was dropped into a receptacle either by the blindfolded participant with the free hand (self-released condition) or unexpectedly by the experimenter (experimenter-released condition).

The catching task consisted of a predictive and reactive task condition. In both conditions, the participants sat at the side of a table and placed the elbow of the dominant hand in a comfortable position close to the edge of a table while holding the grip device. In a self-released condition (predictive task), the participant held a sandbag (150 g) with the free, non-dominant hand just below the upper plastic disc, and dropped the sandbag into the receptacle with closed eyes, after an acoustic signal was given. In the experimenter-released condition (reactive task) the examiner held the sandbag and dropped it into the receptacle at an unpredictable time point (0 to 2 s) after participants closed their eyes following an acoustic signal. After two practice trials for each condition, the procedure started with two self-released trials followed by four experimenter-released trials and ended with two self-released trials (A-B-B-A order) in order to control for sequence effects between tasks within participants. Data recording started immediately after the device was lifted and ended after 8 s. Three seconds after recording started, the tone signaled one of the two different events according to the condition. There was a pause of about 20 s between trials and the change between conditions lasted less than 1 min.

The software detected characteristic events for each trial, which could be adjusted manually during visual inspection of the time series. The moment of sandbag contact on the receptacle was detected as the moment when the rate of the load force increase exceeded a value of 20 N/s. This threshold guaranteed the clear and precise detection of the impact since fluctuations of the load force rate resulting from unsteady holding the device always had smaller amplitudes. The following outcome parameters were calculated to represent the participant’s performance (see Fig. 2):

Fig. 2

Time courses of the grip force (GF) and the load force (LF) during two trials for the experimenter-released and the self-released condition of the catching task. Exemplarily, the performance of one control participant (G05) and two patients with Alzheimer’s disease (D08, D03) is shown. The vertical line indicates the time point of impact of the sandbag with the receptacle.

GFstat_pre: Static GF prior to impact –average GF in a time window 700 to 1000 ms prior to sandbag contact.

GFstat_post: Static GF after impact –average GF in a time window 1700 to 2000 ms after contact.

GFstat_mean: Mean of GFstat_pre and GFstat_post.

GFonset: The moment when GF exceeds GFstat_pre by 0.1 N. To distinguish short touches from GF increases an algorithm first detected a clear GF increase above 1 N, then a backward search localized the preceding minimum (which was around zero) and another forward-search the next GF above 0.1 N.

A different adapted procedure was used to detect the onset of the GF as compared to the impact.

GFmax: maximum GF after the sandbag contact.

ΔGF: Perturbation induced GF increase –difference between GFmax and GFonset.

Tpeak: Duration of GF increase –time interval between GFonset and GFmax.

Tanticip: Duration of anticipatory GF increase –time interval between GFonset and the moment of impact in the self-released condition.

Treact: GF reaction time –time interval between the impact and GFonset in the experimenter-released condition.

Statistical analysis

Statistical analyses were performed with SPSS 21 (IBM, NY). In tasks with repeated trials the individual average of the parameters was calculated across trials. We tested for normal distribution with the Shapiro-Wilk test, which showed that none of the parameters, except age, were normally distributed in the respective groups. Therefore, we used non-parametric Mann-Whitney-U-Test to analyze differences between the groups and Wilcoxon tests for dependent samples. Effect size r was evaluated for the non-parametric tests according to Fritz, Morris, and Richler (2012) [56] with the following guidelines: a large effect is 0.5, a medium effect is 0.3, and a small effect is 0.1. We tested correlations between the parameters of main interest of the catching task (GFstat_mean, ΔGF, Tanticip, Tpeak) and the MMSE, as well as the time in the Luria’s three-step test with Spearman rank correlation tests. To protect against Type 1 error, Bonferroni Correction was applied to the comparisons of MMSE with the measures from the catching task, and likewise for the comparisons of the Luria’s sequence results with the GF measures. Correlations were also performed within the parameters of the catching task to test their interrelationship. A p-value of 0.05 was considered statistically significant.

Data management

In seven individual trials of the catching task (three in controls, four in AD, 3.4% of total trials), GF did not increase after the sandbag contact due to technical failure. We excluded these trials from further analysis as Treact or Tanticip could not be calculated.

RESULTS

Participants

Of the fifteen patients with potential AD, only those twelve patients with biomarker evidence according to the diagnostic criteria of NIA-AA [57] and those who achieved a MMSE score > 15 were included in the final analysis. MMSE differed with a high statistical difference between groups (Mann-Whitney-U-Test: Z = –4.353, p < 0.001). Age did not differ between groups (U-Test: Z = –0.928, p = 0.374). Table 1 summarizes the demographics and test results in the MMSE of the AD patients and the control group.

Luria’s three-step test

Healthy controls required an average of 4.7 s (±1.1 s, range: 3.3–7.6 s) to complete the hand movement sequence of the Luria’s three-step test, whereas dementia patients required significantly longer (7.4 s±2.5 s, range: 3.8–12.0 s, U-Test: Z = –3.035, p = 0.002).

Apraxia

Applying the established cut-off score for identifying apraxia in the production scale of the pantomime of tool use test of < 22 points [58], five out of the twelve AD patients showed signs of apraxia, with three patient scoring below 20 (mean 21.4±2.5 s, range: 17–24). Four healthy controls scored borderline or just below the threshold (22.9±1.4, range: 20 –24). However, the Mann-Whitney-U-Test revealed no significant difference between groups (Z = –1.619, p = 0.118).

Catching task

Anticipatory and reactive GF scaling

Figure 2 shows the GF and load force signals for two exemplary trials in both experimental conditions in one control subject (G05) and in two patients. Patient D03 was mildly affected by dementia according to the relatively high MMSE score (see Table 1), while patient D08 was moderately affected. The vertical line and the rapid increase of load force in Fig. 2 indicate the moment of impact of the sandbag. As expected, in the experimenter-released condition, participants started to increase their GF with a short time lag after the impact of the sandbag (Treact > 0 s, see Fig. 2 and Fig. 3C). In contrast, participants in the self-released condition began to increase GF before the impact (Tanticip < 0 s, see Fig. 2 and Fig. 3C). In both conditions, GF was subsequently increased rapidly to a clear peak before GF decreased, indicating grip relaxation, until a new static GF level was achieved. As the GF increase started earlier in the self-released condition, the overall duration of the force increase (Tpeak) was longer than in the experimenter-released condition (see in particular G05 and D03 in Fig. 2 and Fig. 3C). GFstat_post was higher than GFstat_pre in control subjects in the self-released condition (Fig. 3A, Wilcoxon test, Z = –3.233, p = 0.001) and in the AD group in the experimenter-released condition (Wilcoxon test, Z = –1.961, p = 0.050), as should be expected since the sandbag’s weight added to the weight of the apparatus after contact. However, this GF scaling to the increased weight was not observed in the control subjects in the experimenter-released condition nor in the AD group in the self-released condition (Wilcoxon test, p > 0.2). GFstat_post and GFstat_pre were strongly correlated in both groups and both conditions (Spearman R≥0.811, p < 0.001). Therefore, we calculated the average GFstat_mean from GFstat_pre and GFstat_post and used the combined measure for statistical analyses.

Fig. 3

Performance of patients with Alzheimer’s disease (AD: blue bars and symbols) and control participants (NC: green bars and symbols) in the self-released (upper row) and the experimenter-released (lower row) condition of the catching task. A) Static grip forces before (GFstat_pre) and after the perturbation (GFstat_post); B) magnitude of grip force increase during the perturbation (ΔGF); C) duration of grip force increase during the perturbation (Tpeak, both conditions), grip force reaction time (Treact, only experimenter-released) and duration of anticipation (Tanticip, only self-released). Box plots as well as individual data points are shown.

Despite these common behaviors, differences were observed that turned out to be partially typical for the patient group (see below). Thus, the GF response of patient D08 to the self-triggered impact was rather weak and the GF started to increase quite late just before the impact. His response to the external-triggered perturbation was weaker than in the control subject in one trial, but quite strong in the other trial, indicating a high variability. The GF increases of patient D03 were weaker than those in the control subjects, but this was particularly notable because of the low level of static GF before and after the perturbation.

Group differences

Despite similar time courses of GF signals, we found distinct differences between the group of patients with AD and the control group.

Despite Fig. 3A shows lower median values of the static GF in AD patients compared to the control subjects in both conditions and before as well as after the perturbation, no statistically significant group differences of the averaged static GF were found for the experimenter-released condition (U-Test, GFstatic_mean: Z = –1.594, p = 0.118) and for the self-released condition (U-Test, GFstatic_mean: Z = –1.492, p = 0.145). Inspection of the data revealed that AD patients with relatively low MMSE scores and correspondingly more severe dementia tended to exert higher static GF than the patients with higher MMSE scores. We therefore repeated the analysis after excluding the AD patients with MMSE score lower than 20 (D08, D10, D12, see Table 1). For this analysis, both comparisons revealed significant differences between the remaining nine AD patients and the control group (GFstatic_mean U-Tests, experimenter-released, Z = –2.142, p = 0.033, effect size r = –0.447; self-released, Z = –2.583, p = 0.009, effect size r = –0.539) suggesting that low static GF were particularly common in patients with relatively mild dementia.

The level of GF increase in response to the perturbation was significantly lower in the dementia patients compared to the control subjects in the self-released condition (U-Test, ΔGF: Z = –2.006, p = 0.046, effect size r = –0.393, Fig. 3B). To assess whether the latter finding was related to lower static GF in the patients, we conducted exploratory simple regression analyses, which showed no statistically significant correlation (ΔGF versus GFstatic_pre/post: |R|≤0.322, p≥0.308). In the experimenter-released condition no significant group difference was found for the GF amplitude (U-Test, ΔGF: Z = –1.337, p = 0.193).

In the experimenter-released condition, groups did not differ significantly in the GF reaction time (U-Test, Treact: Z = –0.412, p = 0.705, see Fig. 3C). The time required to reach the GF peak in this condition was significantly longer in the AD group (U-Test, Tpeak: Z = –2.007, p = 0.046, effect size r = –0.394).

In the self-released condition, the onset of anticipatory GF increase is relevant. A smaller negative value of the anticipation duration in dementia patients indicated that they started to increase the GF later in preparation for the impact. However, this result was not statistically significant (U-Test, Tanticip: Z = –1.826, p = 0.067, see Fig. 3C). The total duration of the force increase was significantly shorter in dementia patients than in the healthy control group (U-Test, Tpeak: Z = –2.984, p = 0.002, effect size r = –0.585, Fig. 3C). Anticipation timing and the duration of GF increase were not strongly coupled in the patients as revealed by a non-significant correlation (Tanticip versus Tpeak: R = –0.448, p = 0.145). Similarly, the duration of GF increase did not correlate with the GF amplitude (Tpeak versus ΔGF: R = 0.119, p = 0.713).

Correlation analyses

We found only scarce correlations between the MMSE and performance parameters. In particular, the MMSE did not correlate with the pantomime score (R = –0.178, p = 0.580) and the Luria’s three-step test duration (R = –0.135, p = 0.676). Considering Bonferroni correction due to multiple testing between MMSE and the GF measures (p≤0.013), no comparison reached statistical significance. For exemplary purposes, Fig. 4 shows the MMSE correlation with the highest coefficient, which is the anticipation timing (Tanticip versus MMSE: R = –0.579, p = 0.048, see Fig. 4). From the figure it seems that in patients with lower MMSE the anticipatory GF response was delayed in the predictive condition. However, as indicted above, the correlation did not pass the adjusted threshold due to multiple testing. Luria’s three-step test performance of dementia patients correlated positively with the static GF in the catching task. Slower performance in the Luria’s three-step test was associated with higher static GF in the experimenter-released condition (Luria’s three step test duration versus GFstatic_mean R = 0.692, p = 0.013, see Fig. 4). For the self-released condition, the correlation did not reach statistical significance when the threshold for multiple testing was applied (GFstatic_mean: R = 0.601, p = 0.039). In contrast, no comparable correlation was detected in the group of healthy participants (p > 0.05, see Fig. 4). The Luria’s three step test performance did not significantly correlate with any other performance parameter (all p≥0.245). Similarly, the pantomime test for apraxia was not significantly correlated with other performance parameters.

Fig. 4

Correlation analyses in patients with Alzheimer's disease between the (A) Mini-Mental State Examination (MMSE) and the duration of anticipatory grip force increase (T_anticip), (B) between the time to complete the Luria’s three-step test (Luria 3-Step Movement) and the mean static grip force (GFstat_mean) in the experimenter-released condition. Filled blue symbols represent data of AD participants, open green symbols the data of healthy control participants. The linear regression lines, the coefficient of correlation as well as the levels of significance are indicated for the control group, no significant correlation was found for the AD group. Three AD patients with the lowest MMSE and who were excluded for the additional group comparison of the mean static grip force (GFstat_mean) are indicated (D08, D10, D12, see text).

DISCUSSION

In this study, we compared anticipatory and reactive control of GF between AD patients and healthy controls when an object was held stationary and load perturbation was applied either unexpectedly by another person or predictably by the subject. Many AD patients exerted lower static GF than healthy controls when holding the device before and after perturbation. While this was not significant when the whole patient group was considered, it turned into a statistically significant finding with a large (self-released condition) and a median effect size (experimenter-released condition) when three patients with low MMSE scores were excluded from the analysis. Therefore, GF decrease was obvious in patients with mild dementia. Interestingly, the static GF pre/post perturbations were positively correlated with time required for patients to execute the Luria’s three-step test. While reactive force increases in the experimenter-released condition were similar to control participants, patients’ GF response during the self-induced load perturbation was prolonged. Association with the MMSE scores were weak to absent.

Static grip force

Unexpectedly, the median exerted static GF was lower in AD patients than in controls. This is a novel finding for patients with a CNS disease, as previous studies of GF uniformly reported excessive GF in patients with wide array of CNS diseases, including stroke, Parkinson’s disease, Huntington’s disease, cerebellar degeneration, or infantile cerebral palsy [34 , 59–61]. This GF increase in patients is often considered a strategy to reach stability and prevent the dropping of an object despite sensorimotor deficits, or as an indicator of incorrect gating, i.e., filtering of sensory information [34 , 60]. Importantly, an increase of GF is also associated with aging [62 –65]. It is, however, an open question whether GF increases with age are solely due to decreasing friction at the fingertip-surface interface due to age-related changes in skin properties. Interestingly and similar to our findings, patients with chronic mild peripheral sensory impairment due to diabetes also exerted lower GF than healthy controls when holding an object stationary. In contrast, healthy adults increase their GF in response to a reduced skin perception following digital anesthesia [66] or when wearing gloves [67]. Alternatively, deteriorations in central and peripheral processing of sensory information as well as adaptation of a more conservative, i.e., safety-concerned, grip strategy might also contribute to the observed higher force safety-margins in elderly [68, 69]. Comparing our present findings in the elderly control participants with previous studies in younger cohorts using similar methods, we found that our present control group produced substantially higher GF than previous, younger cohorts (e.g., control subjects aging 45.5 years produced average static GF between 6.0 N and 8.2 N with the same apparatus in Allgöwer, et al. [36]).

We will discuss altered strength and changes in control strategy as possible explanations for the observed decrease in static GF in the group of dementia patients. Loss of motor function parallels cognitive decline already during prodromal AD [5]. Most importantly, low grip strength was repeatedly shown to be a strong risk factor of incident AD in large scale epidemiological studies [70 –72]. However, one must keep in mind that the GF levels exerted by healthy controls are substantially lower than the maximum pinch force levels found in participants who developed AD (44±23 lbs equaling 196±102 N) [73]. Moreover, GF levels in fine-motor tasks do not necessarily reflect maximum grip strengths, as it has been shown in studies examining the paretic hand of stroke survivors [60 , 75]. Unfortunately, it is a limitation of the current study that we did not assess maximum GF. Moreover, as we did not measure the friction at the digit-object interface or the slip-force, i.e., the GF at which object slip occurs, we cannot infer the employed safety margins.

In addition to reduced strength, decreased GF may be related to apathy in AD. Apathy, a reduction in goal-directed behavior [76], affects 49% of people with AD [77] and is caused by dysfunctions or lesions in brain regions responsible for goal-directed behavior and motivation, namely the basal ganglia, the frontal lobe, and the fronto-basal ganglia circuits [76, 78]. Apathy may therefore be a consequence of impairments in cognitive functions necessary for an action plan [79]. This remains just a speculation, since no data about apathy were collected. A recent study found apathy to be associated with reduced precision of prior beliefs about action outcomes in healthy adults [80]. Predicting the threat of object slippage due to the impact of the falling sandbag is critical for eliciting a GF response to safeguard a stable grip. Similarly, we just hypothesized that lower GF levels might be associated with a dysfunction of the salience network, which is crucial for complex executive functions, such as stimuli perception and recruitment of relevant functional networks [81 –83]. Dysfunctions of this large-scale brain network composed mainly of the anterior insula and the dorsal anterior cingulate cortex, are common in psychiatric disorders, frontotemporal dementia, and AD [84]. It is the central brain system for complex executive functions, such as stimuli perception and recruitment of relevant functional networks that are critical for the task [81–83 , 85]. Yuen et al. [86] reported that the salience network plays an important role in apathy and depression in elderly and emphasized its function for motivational behavior. Accordingly, healthy elderly may be driven by higher intrinsic motivation to perform the task without object slippage compared to patients with dementia [77]. Therefore, the elderly controls produced higher GF levels. However, it must be considered that patients with more severe dementia did not show this effect. Thus, the effect of apathy on lower GF in more severe dementia may be suspended and replaced by a progressive increase in GF levels to protect against sensorimotor deficits (see Introduction). However, despite lower GF in dementia patients, we never observed object slippage in the current study.

Regardless of the underlying mechanisms, the low static GF observed particularly in patients with mild dementia is an important finding of the present study. It can be speculated that this extends to other object manipulation tasks, but this remains to be shown. An interesting finding in handwriting indicates that this may indeed be the case. Werner et al. [23] showed that patients with MCI and mild AD exerted less pressure with the pen on the writing surface when writing letters and text than a healthy control group.

Reactive and anticipatory modes of GF increase

In the experimenter-released condition, AD patients required only 68 ms to respond with an increase in GF to the sudden and unpredictable increase in load. This duration was similar to that of the control subjects and is within the range of fast automatic response reported in comparable studies [87]. Apart from a prolonged time to reach peak GF with a medium effect size, no other parameter for the experimenter-released condition differed significantly between patients and control participants or showed a significant correlation with the MMSE score. Therefore, we found no clear evidence of impaired reactive GF control in AD patients.

In the self-released condition, a shortened force increase duration was observed with a large effect size. In addition, force amplitude was decreased with a medium effect size. Descriptive analysis (Fig. 5A) suggested that a delay in GF increase was correlated with lower MMSE scores, which would indicate a gradual impairment in anticipatory control with AD disease progression. However, this observation was not statistically supported. Importantly, though, the GF increase in all AD patients began before the impact of the sandbag, suggesting that the fundamental mechanism of anticipatory GF control is still intact.

Luria’s three-step test performance

Surprisingly, AD patients who performed worse on the Luria’s three-step test had relatively high levels of static GF, as indicated by significant correlations between both measures (see Fig. 5). The Luria’s three-step test can be considered a measure of higher aspects of motor control and coordination [20 , 55]. As a consequence, Luria’s three-step test might have identified patients with poor motor control and coordination, leading to a relative increase of GF levels as a compensatory mechanism typically seen in patients with neurological diseases such as stroke—namely an excess of GF beyond normal. As a consequence, one could speculate that while a decrease in GF levels is the main characteristic of AD patients, a proportion of AD patients might have high GF levels along progressive motor deficits. The relationship with the MMSE score seems to be ambiguous, because on the one hand, removing three patients with low MMSE resulted in clear group differences for static GF, and on the other hand, MMSE for the whole group did not correlate with either static GF or performance on the Luria’s three-step test.

Apraxia

The found prevalence of apraxia is 41.7% (5 out of 12) assessed by a tool-use-pantomime test in our sample of patients with mild AD (median MMSE 23.1, IQR 4). This number is somewhat higher than the rate of 32.2% (n = 69, MMSE = 17.16±8.45) reported by Ozkan et al. [9] and the 35% rate (n = 158, MMSE: 20±5) reported by Smits et al. [88], but clearly lower than the prevalence of 69% of signs of limb apraxia in amnestic AD patients (n = 29) reported by Ahmed et al. [11]. However, at the group level, pantomime scores were not significantly different from controls, which might be due to insufficient statistical power given our small sample size. Moreover, no significant correlations were observed with the parameters of the catching task. The latter finding is consistent with the absence of significant correlations of pantomime scores with anticipatory GF scaling when lifting everyday objects in left-hemispheric stroke patients found in a previous study [89]. In contrast, however, hand imitation scores, which were not assessed in the current study, were significantly correlated with predictive GF control. Therefore, our findings support the notion of limited predictive value of pantomime, as abstract knowledge of object use may be impaired before actual, i.e., physical, tool use functions deteriorate [89, 90].

Limitations

Finally, a number of potential limitations need to be considered: We did not measure static coefficients of friction at the fingertip-object interface in this study. Therefore, we cannot exclude the possibility that differences in static GF levels might be related to differences between groups in skin properties that could affect friction.

The sample of included patients with AD was quite small considering the variability and complexity of results. As we did not reach the sample size, indicated by the G^*power estimation, the present study is underpowered and follow-up studies in a larger sample of AD patients are needed to better identify the disease characteristics that influence patients’ performance on the catching task and to understand the opposing trend of lowered static grip in a large proportion of the sample associated with an increase in static force with decreased performance on the Luria’s three-step test.

Contribution to the field statement

Activities of daily living and fine motor function are affected early in the course of AD. However, the mechanisms underlying impaired fine motor deficits are still poorly understood. This study was the first in the field to investigate reactive and anticipatory modes of GF production in a natural load perturbation task in AD patients. Our main findings were lower static GF levels in a majority of AD patients. We interpreted this finding in the context of a loss of strength, lethargy and salience network perturbation in AD. Interestingly, AD patients who were slow on the Luria’s three-step test deviated from this group finding and presented GF in the normal range. The timing and magnitude of reactive GF control did hardly deviate from controls. In contrast, timing and magnitude of anticipatory GF were impaired in AD patients despite a generally preserved anticipatory mode of control. Because low static GF was particularly prominent in patients with mild dementia, this GF measure during object manipulation may emerge as a potential biomarker for early stages of AD. Future research involving other fine-motor tasks in larger patient groups should further elaborate on this possibility for a rapid, non-invasive biomarker of early signs of AD.

Footnotes

ACKNOWLEDGMENTS

The authors like to thank Veit Kraft for the advisory function during measurement conduction. Our thanks also go to the patients and healthy controls for participating in our study. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Authors’ disclosures available online ().

References

Jack

Jr. , Knopman

, Jagust

, Shaw

, Aisen

, Weiner

, Petersen

, Trojanowski

(2010) Hypothetical model of dynamic biomarkers of the Alzheimer’s pathological cascade. Lancet Neurol 9, 119–128.

Pereira

, Janelidze

, Ossenkoppele

, Kvartsberg

, Brinkmalm

, Mattsson-Carlgren

, Stomrud

, Smith

, Zetterberg

, Blennow

, Hansson

(2021) Untangling the association of amyloid-β and tau with synaptic and axonal loss in Alzheimer’s disease. Brain 144, 310–324.

Edwards

(2019) A unifying hypothesis for Alzheimer’s disease: From plaques to neurodegeneration. Trends Neurosci 42, 310–322.

Montero-Odasso

, Speechley

, Muir-Hunter

, Sarquis-Adamson

, Sposato

, Hachinski

, Borrie

, Wells

, Black

, Sejdić

(2018) Motor and cognitive trajectories before dementia: Results from gait and brain study. J Am Geriatr Soc 66, 1676–1683.

Buchman

, Bennett

(2011) Loss of motor function in preclinical Alzheimer’s disease. Expert Rev Neurother 11, 665–676.

Schaefer

, Hooyman

, Duff

(2020) Using a timed motor task to predict one-year functional decline in amnestic mild cognitive impairment. J Alzheimers Dis 77, 53–58.

Goldenberg

(2014) Apraxia - The cognitive side of motor control. Cortex 57, 270–274.

Vanbellingen

, Kersten

, Van de Winckel

, Bellion

, Baronti

, Muri

, Bohlhalter

(2011) A new bedside test of gestures in stroke: The apraxia screen of TULIA (AST). J Neurol Neurosurg Psychiatry 82, 389–392.

Ozkan

, Adapinar

, Elmaci

, Arslantas

(2013) Apraxia for differentiating Alzheimer’s disease from subcortical vascular dementia and mild cognitive impairment. Neuropsychiatr Dis Treat 9, 947–951.

10.

Lesourd

, Le Gall

, Baumard

, Croisile

, Jarry

, Osiurak

(2013) Apraxia and Alzheimer’s disease: Review and perspectives. Neuropsychol Rev 23, 234–256.

11.

Ahmed

, Baker

, Thompson

, Husain

, Butler

(2016) Utility of testing for apraxia and associated features in dementia. J Neurol Neurosurg Psychiatry 87, 1158–1162.

12.

Johnen

, Frommeyer

, Modes

, Wiendl

, Duning

, Lohmann

(2016) Dementia Apraxia Test (DATE): A brief tool to differentiate behavioral variant frontotemporal dementia from Alzheimer’s dementia based on apraxia profiles. J Alzheimers Dis 49, 593–605.

13.

Baumard

, Lesourd

, Remigereau

, Jarry

, Etcharry-Bouyx

, Chauvire

, Osiurak

, Le Gall

(2018) Tool use in neurodegenerative diseases: Planning or technical reasoning? J Neuropsychol 12, 409–426.

14.

Lesourd

, Baumard

, Jarry

, Etcharry-Bouyx

, Belliard

, Moreaud

, Croisile

, Chauvire

, Granjon

, Le Gall

, Osiurak

(2016) Mechanical problem-solving strategies in Alzheimer’s disease and semantic dementia. Neuropsychology 30, 612–623.

15.

Rabinowitz

, Lavner

(2014) Association between finger tapping, attention, memory, and cognitive diagnosis in elderly patients. Percept Mot Skills 119, 259–278.

16.

Roalf

, Rupert

, Mechanic-Hamilton

, Brennan

, Duda

, Weintraub

, Trojanowski

, Wolk

, Moberg

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

17.

Franssen

, Souren

, Torossian

, Reisberg

(1999) Equilibrium and limb coordination in mild cognitive impairment and mild Alzheimer’s disease. J Am Geriatr Soc 47, 463–469.

18.

Suzumura

, Osawa

, Nagahama

, Kondo

, Sano

, Kandori

(2016) Assessment of finger motor skills in individuals with mild cognitive impairment and patients with Alzheimer’s disease: Relationship between finger-to-thumb tapping and cognitive function. Japanese J Compr Rehabil Sci 7, 19–28.

19.

Lin

, Li

, Wei

(2019) Sensorimotor control of reach-to pinch kinematics in individuals with Alzheimer’s disease. In 2019 12th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI), pp. 1-4.

20.

Weiner

, Hynan

, Rossetti

, Falkowski

(2011) Luria’s three-step test: What is it and what does it tell us? Int Psychogeriatr 23, 1602–1606.

21.

de Paula

, Albuquerque

, Lage

, Bicalho

, Romano-Silva

, Malloy-Diniz

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

22.

Yan

, Rountree

, Massman

, Doody

, Li

(2008) Alzheimer’s disease and mild cognitive impairment deteriorate fine movement control. J Psychiatr Res 42, 1203–1212.

23.

Werner

, Rosenblum

, Bar-On

, Heinik

, Korczyn

(2006) Handwriting process variables discriminating mild Alzheimer’s disease and mild cognitive impairment. J Gerontol B Psychol Sci Soc Sci 61, P228–P236.

24.

Marshall

, Amariglio

, Sperling

, Rentz

(2012) Activities of daily living: Where do they fit in the diagnosis of Alzheimer’s disease? Neurodegener Dis Manag 2, 483–491.

25.

Burns

, Iliffe

(2009) Dementia. BMJ 338, b75.

26.

Fymat

(2019) Dementia: A review. J Clin Psychiatry Neurosci 1, 27–34.

27.

Cassidy

(2016) The clinical assessment of apraxia. Pract Neurol 16, 317–322.

28.

Lesourd

, Baumard

, Jarry

, Etcharry-Bouyx

, Belliard

, Moreaud

, Croisile

, Chauvire

, Granjon

, Le Gall

, Osiurak

(2017) Rethinking the cognitive mechanisms underlying pantomime of tool use: Evidence from Alzheimer’s disease and semantic dementia. J Int Neuropsychol Soc 23, 128–138.

29.

Gulde

, Leippold

, Kohl

, Grimmer

, Diehl-Schmid

, Armstrong

, Hermsdörfer

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

30.

Schneider

, Hermsdörfer

(2016) Anticipation in object manipulation: Behavioral and neural correlates. In Advances in Experimental Medicine and Biology. Progress in Motor Control, Laczko J, Latash M, eds. Springer, pp. 173–194.

31.

Allgöwer

, Hermsdörfer

(2017) Fine motor skills predict performance in the Jebsen Taylor Hand Function Test after stroke. Clin Neurophysiol 128, 1858–1871.

32.

Johansson

, Westling

(1984) Roles of glabrous skin receptors and sensorimotor memory control of precision grip when lifting rougher or more slippery objects. Exp Brain Res 56, 550–564.

33.

Johansson

, Westling

(1988) Coordinated isometric muscle commands adequately and erroneously programmed for the weight during lifting task with precision grip. Exp Brain Res 71, 59–71.

34.

Nowak

, Hermsdörfer

(2006) Predictive and reactive control of grasping forces: On the role of the basal ganglia and sensory feedback. Exp Brain Res 173, 650–660.

35.

Nowak

, Hermsdörfer

, Rost

, Timmann

, Topka

(2004) Predictive and reactive finger force control during catching in cerebellar degeneration. Cerebellum 3, 227–235.

36.

Allgöwer

, Kern

, Hermsdörfer

(2017) Predictive and reactive grip force responses to rapid load increases in people with multiple sclerosis. Arch Phys Med Rehabil 98, 525–533.

37.

Lima

KCA

, Santos

GOC

, Donato

SSV

, Borges

, Hatanaka

, de Freitas

(2021) Grip and load force control and coordination in individuals with diabetes in different manipulation tasks. Hum Mov Sci 77, 102793.

38.

Lima

KCA

, Borges

LDS

, Hatanaka

, Rolim

, de Freitas

(2017) Grip force control and hand dexterity are impaired in individuals with diabetic peripheral neuropathy. Neurosci Lett 659, 54–59.

39.

de Freitas

, Lima

KCA

(2013) Grip force control during simple manipulation tasks in non-neuropathic diabetic individuals. Clin Neurophysiol 124, 1904–1910.

40.

Nowak

, Hermsdörfer

(2005) Grip force behavior during object manipulation in neurological disorders: Toward an objective evaluation of manual performance deficits. Mov Disord 20, 11–25.

41.

Zhang

, Johnston

, Ross

, Sanniec

, Gleason

, Dueck

, Santello

(2013) Effects of carpal tunnel syndrome on dexterous manipulation are grip type-dependent. PLoS One 8, e53751.

42.

Lowe

, Freivalds

(1999) Effect of carpal tunnel syndrome on grip force coordination on hand tools. Ergonomics 42, 550–564.

43.

Handing

, Leng

, Kritchevsky

, Craft

(2020) Association between physical performance and cognitive function in older adults across multiple studies: A pooled analysis study. Innov Aging 4, igaa050.

44.

Heward

, Stone

, Paddick

, Mkenda

, Gray

, Dotchin

, Kissima

, Collingwood

, Swai

, Walker

(2018) A longitudinal study of cognitive decline in rural Tanzania: Rates and potentially modifiable risk factors. Int Psychogeriatr 30, 1333–1343.

45.

Jeong

, Choi

, Kim

, Park

(2018) Association among handgrip strength, body mass index and decline in cognitive function among the elderly women. BMC Geriatr 18, 225.

46.

Jeong

, Kim

(2018) Prospective association of handgrip strength with risk of new-onset cognitive dysfunction in Korean adults: A 6-year national cohort study. Tohoku J Exp Med 244, 83–91.

47.

Sibbett

, Russ

, Allerhand

, Deary

, Starr

(2018) Physical fitness and dementia risk in the very old: A study of the Lothian Birth Cohort 1921. BMC Psychiatry 18, 285.

48.

Doi

, Tsutsumimoto

, Nakakubo

, Kim

, Kurita

, Hotta

, Shimada

(2019) Physical performance predictors for incident dementia among Japanese community-dwelling older adults. Phys Ther 99, 1132–1140.

49.

Cui

, Zhang

, Liu

, Gang

, Wang

(2021) Grip strength and the risk of cognitive decline and dementia: A systematic review and meta-analysis of longitudinal cohort studies. Front Aging Neurosci 13, 625551.

50.

Ameli

, Kemper

, Sarfeld

, Kessler

, Fink

, Nowak

(2011) Arbitrary visuo-motor mapping during object manipulation in mild cognitive impairment and Alzheimer’s disease: A pilot study. Clin Neurol Neurosurg 113, 453–458.

51.

Faul

, Erdfelder

, Lang

A-G

, Buchner

(2007) G^* Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods 39, 175–191.

52.

Oldfield

(1971) The assessment and analysis of handedness: The Edinburgh inventory. Neuropsychologia 9, 97–113.

53.

Folstein

, Folstein

, McHugh

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

54.

Goldenberg

, Hartmann

, Schlott

(2003) Defective pantomime of object use in left brain damage: Apraxia or asymbolia? Neuropsychologia 41, 1565–1573.

55.

Guedalia

, Finkelstein

, Drukker

, Frishberg

(2000) The use of Luria’s method for the neurobehavioral assessment of encephalopathy in an adolescent: Application in a rehabilitation setting. Arch Clin Neuropsychol 15, 177–184.

56.

Fritz

, Morris

, Richler

(2012) Effect size estimates: Current use, calculations, and interpretation. J Exp Psychol Gen 141, 2–18.

57.

McKhann

, Knopman

, Chertkow

, Hyman

, Jack

Jr. , Kawas

, Klunk

, Koroshetz

, Manly

, Mayeux

, Mohs

, Morris

, Rossor

, Scheltens

, Carrillo

, Thies

, Weintraub

, Phelps

(2011) 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 7, 263–269.

58.

Randerath

, Buchmann

, Liepert

, Büsching

(2017) Diagnostic Instrument for Limb Apraxia : Short Version (DILA-S), ät Konstanz & Lurija Institute.

59.

Eliasson

, Gordon

, Forssberg

(1991) Basic co-ordination of manipulative forces of children with cerebral palsy. Dev Med Child Neurol 33, 661–670.

60.

Hermsdörfer

, Hagl

, Nowak

, Marquardt

(2003) Grip force control during object manipulation in cerebral stroke. Clin Neurophysiol 114, 915–929.

61.

Quaney

, Perera

, Maletsky

, Luchies

, Nudo

(2005) Impaired grip force modulation in the ipsilesional hand after unilateral middle cerebral artery stroke. Neurorehabil Neural Repair 19, 338–349.

62.

Cole

, Rotella

, Harper

(1999) Mechanisms for age-related changes of fingertip forces during precision gripping and lifting in adults. J Neurosci 19, 3238–3247.

63.

Gilles

, Wing

(2003) Age-related changes in grip force and dynamics of hand movement. J Mot Behav 35, 79–85.

64.

Kinoshita

, Francis

(1996) A comparison of prehension force control in young and elderly individuals. Eur J Appl Physiol Occup Physiol 74, 450–460.

65.

Parikh

, Cole

(2012) Handling objects in old age: Forces and moments acting on the object. J Appl Physiol 112, 1095–1104.

66.

Nowak

, Hermsdörfer

, Glasauer

, Philipp

, Meyer

, Mai

(2001) The effects of digital anaesthesia on predictive grip force adjustments during vertical movements of a grasped object. Eur J Neurosci 14, 756–762.

67.

Kinoshita

(1999) Effect of gloves on prehensile forces during lifting and holding tasks. Ergonomics 42, 1372–1385.

68.

Cole

, Rotella

(2001) Old age affects fingertip forces when restraining an unpredictably loaded object. Exp Brain Res 136, 535–542.

69.

Diermayr

, McIsaac

, Gordon

(2011) Finger force coordination underlying object manipulation in the elderly - a mini-review. Gerontology 57, 217–227.

70.

Shin

H-Y

, Kim

S-W

, Kim

J-M

, Shin

I-S

, Yoon

J-S

(2012) Association of grip strength with dementia in a Korean older population. Int J Geriatr Psychiatry 27, 500–505.

71.

McGrath

, Robinson-Lane

, Cook

, Clark

, Herrmann

, O’Connor

, Hackney

(2019) Handgrip strength is associated with poorer cognitive functioning in aging Americans. J Alzheimers Dis 70, 1187–1196.

72.

Buchman

, Wilson

, Boyle

, Bienias

, Bennett

(2007) Grip strength and the risk of incident Alzheimer’s disease. Neuroepidemiology 29, 66–73.

73.

Boyle

, Buchman

, Wilson

, Leurgans

, Bennett

(2009) Association of muscle strength with the risk of Alzheimer disease and the rate of cognitive decline in community-dwelling older persons. Arch Neurol 66, 1339–1344.

74.

Nowak

, Hermsdörfer

, Topka

(2003) Deficits of predictive grip force control during object manipulation in acute stroke. J Neurol 250, 850–860.

75.

Blennerhassett

, Carey

, Matyas

(2006) Grip force regulation during pinch grip lifts under somatosensory guidance: Comparison between people with stroke and healthy controls. Arch Phys Med Rehabil 87, 418–429.

76.

Levy

, Czernecki

(2006) Apathy and the basal ganglia. J Neurol 253, vii54–vii61.

77.

Nobis

, Husain

(2018) Apathy in Alzheimer’s disease. Curr Opin Behav Sci 22, 7–13.

78.

Fazio

, Logroscino

, Taurisano

, Amico

, Quarto

, Antonucci

, Barulli

, Mancini

, Gelao

, Ferranti

(2016) Prefrontal activity and connectivity with the basal ganglia during performance of complex cognitive tasks is associated with apathy in healthy subjects. PLoS One 11, e0165301.

79.

Levy

, Dubois

(2006) Apathy and the functional anatomy of the prefrontal cortex–basal ganglia circuits. Cereb Cortex 16, 916–928.

80.

Hezemans

, Wolpe

, Rowe

(2020) Apathy is associated with reduced precision of prior beliefs about action outcomes. J Exp Psychol Gen 149, 1767–1777.

81.

Menon

, Uddin

(2010) Saliency, switching, attention and control: A network model of insula function. Brain Struct Funct 214, 655–667.

82.

Peters

, Dunlop

, Downar

(2016) Cortico-striatal-thalamic loop circuits of the salience network: A central pathway in psychiatric disease and treatment. Front Syst Neurosci 10, 104.

83.

Kotani

, Ohgami

, Yoshida

, Kiryu

, Inoue

(2017) Anticipation process of the human brain measured by stimulus-preceding negativity (SPN). J Phys Fit Sports Med 6, 7–14.

84.

Menon

(2011) Large-scale brain networks and psychopathology: A unifying triple network model. Trends Cogn Sci 15, 483–506.

85.

Menon

(2015) Salience network. In Brain Mapping, Toga AW, ed. Academic Press, Waltham, pp. 597–611.

86.

Yuen

, Gunning-Dixon

, Hoptman

, AbdelMalak

, McGovern

, Seirup

, Alexopoulos

(2014) The salience network in the apathy of late-life depression. Int J Geriatr Psychiatry 29, 1116–1124.

87.

Johansson

, Westling

(1988) Programmed and triggered actions to rapid load changes during precision grip. Exp Brain Res 71, 72–86.

88.

Smits

, Flapper

, Sistermans

, Pijnenburg

YAL

, Scheltens

, van der Flier

(2014) Apraxia in mild cognitive impairment and Alzheimer’s disease: Validity and reliability of the Van Heugten Test for apraxia. Dement Geriatr Cogn Disord 38, 55–64.

89.

Eidenmüller

, Randerath

, Goldenberg

, Li

, Hermsdörfer

(2014) The impact of unilateral brain damage on anticipatory grip force scaling when lifting everyday objects. Neuropsychologia 61, 222–234.

90.

Goldenberg

, Hentze

, Hermsdörfer

(2004) The effect of tactile feedback on pantomime of tool use in apraxia. Neurology 63, 1863–1867.