We conducted a cross-sectional study of community-dwelling participants aged 60 years or older. Participants with a negative medical recommendation to exercise were excluded from the study. A total of 174 participants (age: 73.88, SD 6.05 years; weight: 70.93, SD 13.07 kg; height: 159.07, SD 9.47 cm) divided into men (n=66; age: 73.63, SD 5.96 years; weight: 71.47, SD 13.16 kg; height: 159.56, SD 9.52 cm) and women (n=108; age: 73.90, SD 6.06 years; weight: 70.85, SD 13.10 kg; height: 158.95, SD 9.44 cm), participated in this study.

Anthropometric data, including height assessed with a wall-mounted stadiometer (Seca 220) and weight assessed with a scale (Tanita, MC 780-P MA), were collected. Fat-free mass and fat mass were assessed using a bioimpedance analysis system, the BIA 101 Biva Pro (Akern). This provided detailed body composition measurements for the study. Participants were assessed barefoot and in minimal clothing. They were also instructed to refrain from consuming alcohol for 48 hours before the assessments, avoid intense physical activity for at least 12 hours before the assessments, and not wear any metallic objects during the evaluation.

The main variables obtained from each handwriting task were as follows:

To assess classification, the pentagon copying test from the Montreal Cognitive Assessment [36] was administered. In this task, participants were instructed to replicate a specific geometric figure, a pentagon. The test evaluates visuospatial abilities, executive functioning, and attention. Participants were provided with a digital sheet featuring a predrawn pentagon and instructed to replicate the figure by connecting its vertices using straight lines. The task aimed to ensure the accurate completion of the shape in accordance with the specified guidelines.

The test was scored based on the accuracy of the drawing and adherence to the instructions. Errors such as incorrect line placement, additional lines, or failure to follow instructions may indicate impairments in visuospatial processing and executive functioning, which are cognitive domains commonly affected in neurological conditions. Most experts assign a binary score of 0 or 1, with a correct response defined as producing 2 pentagons, each with 5 sides, and a correctly intersected region between them. If these criteria are not met, the score is 0. This scoring process is both straightforward and efficient.

In some contexts, a more granular scoring approach is used, particularly in computerized versions of the test. For example, Nagaratnam et al [19] proposed a scoring system ranging from 1 to 10 based on the quality of the drawing. In line with the original criteria, factors such as figure rotation and tremor were disregarded. In this study, a binary scoring system was adopted because the aim was to classify participants according to pentagon-copy performance rather than to perform a detailed visuoconstructive assessment. Importantly, pentagon-copy performance was used as a task-specific grouping variable and not as a diagnostic criterion or cognitive screening reference standard for cognitive impairment.

The same database used in this study was previously used to investigate cognitive impairment detection through a frontal camera during handwriting tasks, where we reported up to 83% accuracy based on head movement analysis, highlighting the potential of this approach as an automatic assessment method [37].

To evaluate muscular strength and power, a series of standardized tests targeting both upper- and lower-body performance were conducted. These tests included measures of isometric, dynamic, and explosive strength, as described below.

The handgrip strength test was used as a measure of muscular strength [38], particularly for assessing overall upper-body strength. Participants were instructed to grip a Jamar hand dynamometer (Patterson Medical) with maximal force for several seconds. The test was performed on both hands, and the highest value, measured in kilograms, was recorded as the result.

The 30-second arm curl test was used to evaluate upper-body strength and endurance [39]. Participants performed as many bicep curls as possible within 30 seconds using a 2.3 kg dumbbell for women and a 3.6 kg dumbbell for men, with the total number of repetitions recorded as the score.

The midthigh pull test was used to assess lower-body strength [40]. For this test, participants gripped a fixed bar connected to a strength sensor (Chronojump) positioned at midthigh height while maintaining a partial squat posture. They were instructed to pull upward with maximal effort for 3 to 5 seconds, with the peak force generated, measured in Newtons, recorded as a measure of lower-body strength.

The CMJ was used to dynamically assess lower-body explosive strength [41]. Participants started from an upright standing position, performed a rapid downward movement, and immediately jumped vertically as high as possible. Jump height (cm), flight time (ms), and power output (W/kg) were recorded using a force platform (MuscleLab). Each participant performed 3 trials with a 1-minute rest interval, and the highest value was used for analysis.

Cardiorespiratory fitness was evaluated using the 6MWT [42]. The test was conducted on a hardened surface along a 30-m straight-line circuit. Participants were instructed to walk at their fastest pace, without transitioning to running, to cover the greatest possible distance within 6 minutes. They wore comfortable clothing and footwear and were allowed to rest or stop if necessary. To objectively measure ventilatory parameters, participants were equipped with a portable gas analyzer (K5 COSMED). Peak oxygen (peak VO₂, mL/kg/min) uptake was recorded as an indicator of cardiorespiratory fitness, alongside peak carbon dioxide (peak VCO₂, mL/kg/min) production and pulmonary ventilation (VE; L/min) to assess ventilatory efficiency.

To evaluate participants’ functional mobility, balance, and strength, a series of standardized tests were used, including the TUG test, 400-m walking test, and SPPB. These assessments are designed to measure key aspects of physical performance, such as gait speed over short and medium distances, balance, and the ability to perform dynamic movements. The results from these tests help classify individuals based on their functional abilities and identify potential areas of decline in mobility and stability.

The TUG test involves measuring the time required for an individual to rise from a chair, walk 3 m as quickly as possible, turn around, and return to sit down. A completion time of less than 10 seconds is generally considered to be within normal limits. The onset of functional decline is indicated by times ranging from 10 to 20 seconds, although the individual may still be capable of performing basic transfers independently [43].

The 400-m walking test was incorporated within the 6MWT protocol, during which the total time taken by participants to complete a distance of 400 m was recorded. This measurement provides a practical indicator of functional walking capacity and endurance [44].

The SPPB comprises 3 components: a balance assessment, a 4-m gait speed test, and a sit-to-stand test involving 5 repetitions.

The balance assessment evaluates the participant’s ability to maintain specific standing positions for 10 seconds each: feet side by side, semitandem, and tandem. For the feet side by side and semitandem positions, participants received scores of 0 or 1 based on their ability to hold the position (0 points for less than 10 seconds and 1 point for 10 seconds or more). For the tandem position, scoring ranges from 0 to 2 points (0 points for less than 3 seconds, 1 point for 3 to less than 10 seconds, and 2 points for 10 seconds or more) [45]. The total balance assessment score ranged from 0 to 4 points.

Gait speed was measured over an 8-m course, which included 2 m for acceleration, 4 m for timing, and 2 m for deceleration, thereby minimizing early braking bias. Participants were instructed to walk at their usual pace, as if walking on a street. Timing started on the command “Go,” following a 3-second countdown, and concluded as the participant crossed the 4-m mark. Scoring was determined as follows: 0 points if the participant was unable to perform the test, 1 point for times exceeding 6.52 seconds, 2 points for times between 4.66 and 6.52 seconds, 3 points for times between 3.62 and 4.65 seconds, and 4 points for times below 3.62 seconds [45].

The sit-to-stand test involved a straight-backed chair stabilized against a wall. Participants were instructed to cross their arms over their chest and stand up and sit down 5 times as quickly as possible. Timing began from the initial seated position and ended when the participant reached the final standing position after the fifth repetition. The scoring for this component was based on the total time taken to complete the task: 0 points for more than 60 seconds or inability to perform the test, 1 point for between 16.7 and 59 seconds, 2 points for between 13.70 and 16.69 seconds, 3 points for between 11.20 and 13.69 seconds, and 4 points for less than 11.19 seconds [45].

A composite score ranging from 0 to 12 points was derived from all components, categorizing individuals into the following groups: severe limitations (0-3 points), moderate limitations (4-6 points), mild limitations (7-9 points), and minimal limitations (10-12 points) [46].

Static balance was assessed using a force platform (MuscleLab) [47]. Participants stood barefoot on both feet for up to 30 seconds across 3 trials, each separated by a 1-minute rest interval, without external support. The test was performed bilaterally to evaluate balance ability. Center of pressure parameters were recorded, including mean path velocity (mm/s), anteroposterior and mediolateral path velocities (mm/s), anteroposterior and mediolateral oscillation frequencies (Hz), and mean center of pressure displacement (mm).

Continuous variables were expressed as mean (SD) and categorical variables as frequencies (%). Initial between-group comparisons were performed using 2-tailed Student t tests or Mann–Whitney U test, as appropriate, and chi-square tests for categorical variables.

Adjusted analyses were conducted using linear regression models, including group as the main predictor and age and sex as covariates. Regression coefficients (β) with 95% CIs were reported.

Variables with nonnormal distributions were log transformed before analysis.

Given the number of comparisons, the false discovery rate (FDR) was controlled using the Benjamini-Hochberg procedure, and both raw and adjusted P values are presented. Statistical significance was set at P<.05.

All participants were informed about the aims and procedures of the study and provided written informed consent before participation. The study protocol was approved by the Ethics Committee of Universitat TecnoCampus Mataró-Maresme (CEI1/2022) and was conducted in accordance with the Declaration of Helsinki. To ensure privacy and confidentiality, all participant data were anonymized before analysis and stored securely in password-protected files accessible only to authorized members of the research team. No financial or other compensation was provided to participants for taking part in the study.

Group classification was based on the results of the pentagon copying test. All drawings were manually reviewed, and participants were classified as either having normal pentagon-copy performance or altered pentagon-copy performance, depending on whether the drawing met the criteria: the presence of 2 pentagons, each with 5 sides, and a correctly formed intersection region between them. Participants who fulfilled these criteria received 1 point; those who did not were given a score of 0. Figure 3 presents the results of this task, illustrating the classification of 4 participants into 2 distinct groups: normal pentagon-copy performance and altered pentagon-copy performance.

Of the total sample, 93 participants were classified as having normal pentagon-copy performance (mean age 69.6, SD 5.9 years) and 81 as having altered pentagon-copy performance (mean age 72.9, SD 5.6 years). Among them, 56.1% (37/66) men and 51.9% (56/108) women were classified as having normal pentagon-copy performance, whereas 43.9% (29/66) men and 48.1% (52/108) women were classified as having altered pentagon-copy performance. Differences in pentagon-copy performance across age groups are presented in Table 1.

After adjustment for age and sex and correction for multiple comparisons using the FDR, several handwriting variables remained significantly associated with group classification based on pentagon-copy performance. The number of statistically significant differences, as well as mean values, regression coefficients, confidence intervals, and P values, are presented in Tables 2-5.

The most consistent differences between groups were observed in temporal variables. In particular, time down was significantly higher in participants with altered pentagon-copy performance in tasks 1, 2, 8, and 9. In addition, time on air was significantly higher in this group in tasks 8 and 9.

Significant between-group differences were also found in pressure-related variables. Mean pressure was significantly lower in participants with altered pentagon-copy performance in tasks 1 and 8 after FDR correction.

For kinematic variables, significant differences after FDR adjustment were identified in selected tasks only. In task 6, participants with altered pentagon-copy performance showed significantly higher mean acceleration. In task 8, this group showed significantly higher maximum speed and maximum acceleration.

Overall, the greatest number of significant differences after FDR correction was observed in tasks 8 and 9. No other handwriting variables remained statistically significant after adjustment for multiple comparisons.

In relation to physical performance tests, the unadjusted analyses showed statistically significant differences between participants with normal and altered pentagon-copy performance in selected variables, including CMJ height and flight time, anteroposterior axis balance stability, and 400-m walking time. However, after adjustment for age and sex, these differences were no longer statistically significant. Likewise, after controlling for multiple comparisons using the FDR, none of the physical performance variables remained statistically significant. No statistically significant differences were observed between the groups in other strength tests, such as handgrip strength, the 30-second arm curl test, or chair stand tests, nor in 4-m gait speed, total distance covered in the 6MWT, the TUG test, or metabolic variables such as peak VO₂. The adjusted results for physical performance variables are presented in Table S1 in Multimedia Appendix 1. The number of participants differed across some tests because certain individuals were unable to complete them.

The objective of this study was to examine whether handwriting kinematic and physical performance variables differ according to pentagon-copy performance classification in community-dwelling adults aged 60 years or older. After adjustment for age, sex, and multiple comparisons, only selected handwriting-derived variables remained significantly associated with altered pentagon-copy performance, whereas no physical performance variables remained significant.

The findings of this study should be interpreted within the context of a community-based sample without confirmed clinical diagnoses of cognitive impairment, in which group classification was based on pentagon-copy performance as a marker of task performance potentially related to cognitive vulnerability [20] rather than on a formal neuropsychological or clinical diagnostic assessment. Although failure on the intersecting-pentagons task is generally considered suggestive of underlying neurocognitive dysfunction, particularly in visuospatial and constructional domains, the literature supports its use primarily as a screening marker rather than as an isolated diagnostic standard [19,20]. Accordingly, altered pentagon-copy performance in this study should be interpreted as a marker of probable cognitive vulnerability within a community-based sample, but not as sufficient evidence on its own to establish clinical cognitive impairment. This study was not designed as a cognitive screening validation study; rather, it examined associations between handwriting and physical performance variables and group classification based on pentagon-copy performance in a community-based sample. Within this group-classification framework, differences were observed between individuals with normal and altered pentagon-copy performance in selected kinematic parameters recorded during digital handwriting tasks, whereas no corresponding differences were found in physical performance measures after adjusted analyses. These results are consistent with the hypothesis that early cognitive-related alterations may be reflected in fine motor execution [48], although not in overall physical performance in the present sample, in contrast to previous literature [49].

In the cognitively demanding handwriting tasks, which require mental planning, sustained attention, and working memory, the most consistent differences were observed between the 2 groups. Participants with altered pentagon-copy performance exhibited a significantly prolonged contact time with the screen and a markedly lower mean writing pressure. This pattern may reflect, among other possible explanations, greater disruption in movement planning during the writing process, a finding consistent with previous studies that have documented task execution slowing in complex sequential activities, as well as reductions in applied writing pressure among individuals with cognitive impairment [30,50]. Although velocity values in these tasks were not consistently lower in the altered pentagon-copy performance group, a nonsignificant tendency toward lower mean velocity was observed in some recordings, which may be indicative of less stable writing performance.

In fine motor control tasks, the altered pentagon-copy performance group exhibited higher values in mean acceleration. Contrary to what might be interpreted as an indicator of superior performance, this increase in acceleration likely reflects, among other possible explanations, reduced motor self-regulation and a more impulsive and inefficient execution. Previous evidence suggests that in contexts of cognitive impairment, individuals tend to adopt more erratic motor patterns, characterized by lower temporal and spatial consistency in stroke execution [30]. Higher acceleration in tasks demanding graphical precision may reflect less regulated motor execution.

The mechanically oriented tasks, designed to assess motor execution without additional cognitive demands, also revealed noteworthy differences. In tasks 8 and 9, the altered pentagon-copy performance group demonstrated a significantly longer total execution time (8749 ms vs 4839.2 ms and 5611 ms vs 3005.5, respectively) and higher time on air (7146.5 ms vs 5456.3 ms and 4840.9 ms vs 3599.7 ms, respectively). In addition, the altered pentagon-copy performance group showed lower mean writing pressure (6579.1 au vs 7855.5 au), higher maximum writing speed (10252.7 mm/s vs 9060.5 mm/s), and higher maximum acceleration (10252.8 mm/s² vs 9059.4 mm/s²) in task 8. These results suggest reduced motor efficiency even in low-complexity tasks, consistent with previous evidence indicating that cognitive decline may compromise the basal function of motor networks involved in handwriting [27,51], even in the absence of overt clinical signs detectable through conventional neurological assessments.

The findings for physical performance should also be interpreted with caution. Although the unadjusted analyses initially suggested between-group differences in selected variables, including CMJ height and flight time, anteroposterior balance parameters, and 400-m walking time, these associations were no longer statistically significant after adjustment for age and sex. Furthermore, none of the physical performance variables remained significant after correction for multiple comparisons using the FDR, indicating that the initial differences were not robust after statistical adjustment. These findings suggest that, in this community-based sample, conventional physical performance measures may lack sensitivity to detect subtle functional differences associated with pentagon-copy performance classification. Previous studies have identified reduced explosive force production as a robust indicator of cognitively healthy aging [52], and associations have also been reported between MCI and other physical performance measures, including handgrip strength, TUG performance, gait speed, postural control, and cardiorespiratory fitness [53-59]. In this study, however, physical performance measures did not provide support as independent markers of group classification based on pentagon-copy performance. This discrepancy may reflect the relatively subtle nature of the classification approach used, the older age of participants with altered pentagon-copy performance, and other differences in study design, participant characteristics, cognitive reference standards, and severity of impairment across studies. It is also possible that such associations may become more evident in populations with more advanced cognitive impairment.

From a methodological perspective, this study provides a novel contribution by combining noninvasive digital tools for the precise quantification of handwriting kinematics with standardized assessments of physical performance in a community-based sample of older adults. This multimodal approach enabled the simultaneous evaluation of fine motor and functional measures in relation to group classification based on pentagon-copy performance. After adjustment for age, sex, and multiple comparisons, the most consistent associations were observed in selected handwriting-derived variables, particularly temporal measures, whereas physical performance outcomes did not remain statistically significant. In this context, digital handwriting analysis may represent a sensitive and scalable complementary approach for detecting subtle functional differences in community-based settings.

Several limitations should be acknowledged. First, the cross-sectional design does not allow conclusions regarding causality or predictive value. Second, group classification was based on pentagon-copy performance and does not constitute a clinical diagnosis of cognitive impairment. Third, participants with altered pentagon-copy performance were older on average, which may have contributed to the initial unadjusted differences despite statistical adjustment. Fourth, pentagon-copy performance may be influenced by factors other than cognitive status, such as visuoconstructional deficits, fine motor impairment, tremor, or visual difficulties. Finally, the study was conducted in a real-world community setting, where comprehensive neuropsychological evaluation was not feasible.

Future studies should examine these associations using more robust and clinically validated cognitive reference standards, such as the Montreal Cognitive Assessment total score, the Mini-Mental State Examination, or formal neuropsychological assessment. Longitudinal designs will also be needed to determine the predictive value of the handwriting variables identified here and to establish whether these measures may contribute to the development of digital screening tools or automated risk assessment models in primary care and community-based settings.

In this community-based sample of older adults, only a limited subset of digital handwriting variables remained significantly associated with group classification based on pentagon-copy performance after adjustment for age, sex, and multiple comparisons. The most consistent associations were observed in temporal handwriting measures, particularly time down and time on air, together with selected pressure and kinematic parameters, whereas none of the assessed physical performance variables remained statistically significant after adjusted analyses.

These findings suggest that digital handwriting analysis may be a more sensitive complementary approach than conventional physical performance tests for identifying subtle functional differences associated with altered pentagon-copy performance in older adults. However, because group classification was based on a single visuoconstructional task rather than on a formal clinical or neuropsychological diagnosis, the results should be interpreted with caution and should not be considered evidence of diagnostic screening performance for cognitive impairment.

Overall, this study supports the potential value of digital handwriting analysis as a noninvasive and scalable complementary approach for community-based assessment. Future research should use validated cognitive reference standards and longitudinal designs to determine the clinical relevance, predictive value, and potential application of these handwriting-derived measures in digital screening strategies for older adults.