Dose-Response Association between Hand Dexterity and Functional Disability: A Longitudinal Study from the Kasama Study
Article information
Abstract
Background
Poor hand dexterity may increase the risk of functional disability; however, few studies have examined the relationship between hand dexterity and incident functional disability. The aim of this study was to prospectively investigate the dose-response association of hand dexterity with incident functional disability in community-dwelling older adults.
Methods
This study included 1,069 older adults aged ≥65 years in Kasama City, Japan. Peg-moving and circle-drawing tasks were used to evaluate hand dexterity. Functional disability was identified using the Japanese Long-Term Care Insurance System database. Restricted cubic spline analysis was performed to investigate the dose-response association between hand dexterity and incident functional disability.
Results
During a mean follow-up of 8.5 years (maximum 14.0 years), 248 participants (23.2%) developed functional disability. The lowest performance group in each hand dexterity test had a significantly higher risk of functional disability than the highest performance group—peg-moving (hazard ratio [HR]=1.92, 95% confidence interval [CI] 1.29–2.87) and circle drawing (HR=1.66, 95% CI 1.15–2.41). Spline analysis confirmed curvilinear dose-response associations between hand dexterity and incident functional disability. Increased risk was observed when participants performed worse than the cut points (peg-moving, 37.9/38.0 seconds; circle drawing, 21/20 points), and no decreased risk was observed for those who performed better than these cut points.
Conclusion
Easily evaluated hand dexterity tests may be valuable for predicting functional disability in older adults. Curvilinear dose-response associations suggest that maintaining adequate hand dexterity could be a key strategy to support functional independence.
INTRODUCTION
An aging global population is a critical social and demographic issue worldwide. In 2021, 10% of the global population was aged ≥65 years,1) and this is projected to rise to 25% by 2050.2) With this growth, the number of individuals with functional disabilities certified for long-term care insurance (LTCI) is also increasing in Japan, imposing substantial burdens on social, economic, and healthcare systems.3) Therefore, preventing or reducing functional disability is essential.
Physical performance tests, including the 5-meter habitual walk (5-mHW),4) single-leg balance with eyes open (SLB),5) timed up and go (TUG),6) five-repetition sit-to-stand (5-STS),7) and grip strength,8) reflect the physical function required for activities of daily living (ADL) and are known predictors of functional disability in older adults.9) However, hand dexterity, which requires more refined movements, is also vital for ADL and instrumental ADL (IADL).10,11) For example, hand dexterity is essential for tasks such as eating, dressing, writing, housekeeping, laundry, and shopping. Declines in hand dexterity can impair ADL and IADL performance, thereby increasing the risk of functional disability.
Skilled hand movements involve brain–nervous system activity,12,13) and poor finger function has been linked to higher dementia risk,14) a key factor underlying long-term care (LTC) needs.15) Previous studies have reported associations between poor hand dexterity and increased risk of functional disability.16,17) For example, a Dutch study linked poor performance on block-moving tasks to reduced self-reported ADL levels,17) while a 5-year follow-up study of African Americans reported that higher hand dexterity is strongly associated with a lower risk of incident ADL disability.16) However, these studies had a limitation in that ADL limitations were defined using self-reported assessments.16,17) Additionally, whether these findings from Europe and America are applicable to Asian older adults is uncertain. The dose-response association between hand dexterity and functional disability also remains unclear. Determining whether hand dexterity is dose-responsively associated with functional disability is important to develop effective prevention strategies for functional disability.
Exploring simple assessments for hand dexterity that predict future functional disability is vital for public health. The peg-moving task is a well-established, effective method for assessing hand dexterity, offering a precise and standardized measure.18) Conversely, the circle-drawing task, used in some Japanese studies,19,20) is a simpler alternative, requiring only basic materials such as a pen, paper, and a stopwatch. If both tasks reliably predict future incidents of functional disability, researchers and public health practitioners will have options for assessing hand dexterity based on available tools. Therefore, we aimed to prospectively investigate the association between hand dexterity, assessed by peg-moving and circle-drawing tasks, and incident functional disability over a maximum of 14 years in community-dwelling older adults, applying dose-response analysis.
MATERIALS AND METHODS
Participants
Study data were sourced from the “Kasama study,” an open cohort study conducted annually since 2009 in Kasama City, Ibaraki Prefecture, Japan.21) The cohort included participants aged 65–85 years randomly drawn from the Basic Resident Registrar. Participant flow is shown in Supplementary Fig. S1. As the baseline survey, 7,500 individuals were randomly invited from 2009 to 2019, and 1,195 participated. To avoid overlapping data when individuals participated multiple times, the earliest valid data without missing values were prioritized. Exclusions included individuals with incomplete data (n=70), those untraceable in the LTCI database (n=28), and those previously enrolled as LTCI beneficiaries before the baseline survey (n=26). Additionally, two individuals were excluded because they moved to another city before the earliest case of functional disability, a requirement for conducting Cox proportional hazards analysis. Ultimately, 1,069 individuals were included.
All participants were fully informed about the study details and provided written informed consent. The study was approved by the University of Tsukuba Ethics Committee (Tai 30-5).
Hand Dexterity Test
Peg-moving and circle-drawing tasks were used to evaluate hand dexterity.
(1) Peg-moving task: Participants stood in front of a pegboard with 48 pegs arranged in a six-by-eight matrix (TKK1306; Takei Scientific Instruments Co. Ltd., Niigata, Japan). At the beginning of the task, all pegs were inserted into the holes on the distal board. Participants were instructed to simultaneously move two pegs, with one peg held in each hand, from the distal to the proximal board as quickly as possible. The task began from the far-right columns of the distal board and proceeded leftward in two-column increments (Supplementary Fig. S2). After practicing with 12 pegs, we measured the time (seconds) required to move 48 pegs. This method was used in a previous longitudinal study that reported a significant association with cognitive function.22)
(2) Circle-drawing task: Participants circled numbers 1–80 in sequential order as quickly as possible. The number of circles completed in 15 seconds was recorded as points (Supplementary Fig. S2). This test, developed as part of the Japanese cognitive assessment tool “Five Cog Test,”20) was significantly associated with cognitive function in a cross-sectional study.19)
Functional Disability
Functional disability was identified using the nationally uniform LTCI database provided by the Kasama City municipal database. LTCI, a mandatory social insurance, provides benefits for Japanese adults aged ≥65 years based on physical and mental disabilities.23) Details of LTCI certification have been described previously.3) LTCI certification includes support levels 1 and 2 and care levels 1 (least disabled) to 5 (most disabled), reflecting the increasing severity of conditions. The higher the level of care needed, the more severe the conditions. Functional disability was defined as the onset of LTC needs at care level 2 or above, corresponding to assistance in ADL and IADL.24) Participants were followed from the baseline survey until one of the following: LTC needs, death, relocation, or July 2023, whichever came first. Death or relocation information was provided by Kasama City’s municipal government via the Basic Resident Registrar.
Covariates
Covariates were selected based on prior studies.25,26) Demographic variables collected via self-reported questionnaires included age (continuous), sex (male or female), education level (compulsory, high school, college/junior college, or no answer), and clinical history of stroke, heart disease, hypertension, hyperlipidemia, and diabetes (yes or no). Body mass index (BMI, kg/m2) was categorized into four groups based on World Health Organization BMI definitions27): underweight (<18.5 kg/m2), normal weight (18.5–24.9 kg/m2), overweight (25.0–29.9 kg/m2), and obesity (≥30.0 kg/m2). Physical activity levels were assessed using the Physical Activity Scale for the Elderly (PASE).28,29)
Statistical Analyses
Participant characteristics were compared by tertile for each hand dexterity test using chi-squared tests for categorical variables and one-way analysis of variance for continuous variables. For the peg-moving task, tertiles 1, 2, and 3 were ≤34.49, 34.50–38.78, and ≥38.79 seconds, respectively. For the circle-drawing task, tertiles 1, 2, and 3 were ≥27, 21–26, and ≤20 points, respectively. Lower tertiles (tertile 1) represented better performance, whereas higher tertiles (tertile 3) indicated poorer performance.
Cox proportional hazards analysis with robust errors was used to investigate the association between hand dexterity and incident functional disability, with results presented as hazard ratios (HR) and 95% confidence intervals (CI). Model 1 adjusted for age, sex, and baseline years. Model 2 additionally adjusted for education level, BMI, stroke, heart disease, hypertension, hyperlipidemia, diabetes, and physical activity. Kaplan–Meier curves based on Model 2 compared performance levels of the hand dexterity test in relation to the incidence of functional disability.
To investigate dose-response associations between hand dexterity and incident functional disability, restricted cubic spline curves were conducted based on the Model 2. Three knots were placed at the 10th, 50th, and 90th percentiles.30) To deal with extremity values, the data within the worst 0.5 percentile (n=5) for each hand dexterity test were excluded from the spline analysis. Receiver operating characteristic (ROC) analysis was performed to set the reference values, which were used as the optimal cut points. The optimal cut point for balancing sensitivity and specificity was defined using the Youden index (sensitivity + specificity − 1), with maximum index values indicating optimal cut points.31) Participants who moved to other cities (n=10) or died (n=79) before event observation or end of follow-up were excluded from ROC analysis. Statistical analyses were conducted using Stata/SE 18 (Stata Corp., College Station, TX, USA). Statistical significance was set at p<0.05.
RESULTS
Tables 1 and 2 summarize the baseline characteristics of participants according to the tertiles of the peg-moving and circle-drawing tasks, respectively. Participants in higher hand dexterity tertiles tended to be younger, female, have fewer clinical histories, and be more physically active. The characteristics of all participants at baseline are summarized in Supplementary Table S1.
Participant characteristics by tertile in the peg-moving task
Participant characteristics by tertile in the circle-drawing task
Over a mean follow-up of 8.5 years (maximum 14.0 years), 248 participants (23.2%) developed functional disabilities. The total follow-up duration was 9,044 person-years, with an incidence rate of 27.4 per 1,000 person-years. Table 3 outlines the associations between hand dexterity and incident functional disability and Supplementary Tables S2 and S3 present the HR for the covariates used in each model. In Model 2, participants in tertile 3 had significantly higher risks of incident functional disability than those in tertile 1—peg-moving (HR=1.92, 95% CI 1.29–2.87) and circle drawing (HR=1.66, 95% CI 1.15–2.41). Tertile 2 was not significantly associated with incident functional disability—peg-moving (HR=1.13, 95% CI 0.74–1.71) and circle drawing (HR=1.06, 95% CI 0.71–1.58). Supplementary Fig. S3 presents Kaplan–Meier curves adjusted for Model 2, showing a higher cumulative incidence of functional disability in tertile 3 throughout the follow-up period. No violations of the proportional hazard assumption, such as crossing lines, were observed.
Hazard ratios for incident functional disability linked to hand dexterity
To establish reference values for the cubic spline analysis, ROC analysis was conducted (Table 4 and Supplementary Fig. S4). Significant areas under the curve (AUC) were observed for both tasks—peg-moving (0.74) and circle drawing (0.70). The optimal cut points were 37.9/38.0 seconds for the peg-moving task and 21/20 points for the circle-drawing task. In the adjusted Cox model, participants performing worse than each cut point had significantly higher risks of functional disability—peg-moving (HR=1.96, 95% CI 1.46–2.64) and circle drawing (HR=1.52, 95% CI 1.14–2.02)—than those performing better.
Cut points for peg-moving and circle-drawing tasks (n=980)
Fig. 1 illustrates cubic spline analysis results based on Model 2. A curvilinear dose-response association was observed, where higher hand dexterity (better performance) corresponded to a lower risk of functional disability. Risks increased when participants performed worse than the cut points, and no reductions in risk were observed for those performing better.
Cubic spline curves for the incident functional disability by peg-moving (A) and circle-drawing tasks (B). The cubic spline curves were performed based on Model 2 in Table 3. The solid line represents hazard ratios, dashed line represents 95% confidence intervals, and bars represent histograms. Results are trimmed at the worst 0.5 percentile (n=5) for each hand dexterity test. The optimal cut points for each test (peg-moving: 37.9/38.0 seconds; circle drawing: 21/20 points) were used as reference value.
DISCUSSION
Following a cohort of community-dwelling older Japanese adults for up to 14 years, we found that poor hand dexterity was significantly associated with an elevated risk of functional disability. Moreover, a curvilinear dose-response association was noted between hand dexterity, as assessed through peg-moving and circle-drawing tasks, and incident functional disability. Participants with values worse than the cut points demonstrated increased risks. These findings highlight the utility of simple hand dexterity tests in predicting functional disability and suggest that maintaining adequate dexterity is a potential strategy for supporting functional independence, alongside other rehabilitation approaches. To our knowledge, this study is the first to investigate a dose-response association between hand dexterity and incident functional disability.
Previous studies have reported similar findings to our study. For example, a cross-sectional study comparing older Japanese adults with and without LTCI benefits showed that dependent older adults have poorer hand dexterity than their independent counterparts.32) Even a comparison using Fried’s definition showed that frail older adults have poorer hand dexterity than non-frail older adults.32) Prospective cohort studies from the United States and the Netherlands similarly identified poor hand dexterity as a predictor of self-reported functional disability among older adults.16,17) Extending these findings, our study assessed two hand dexterity tests and analyzed their relationship with incident functional disability using a standardized municipal database in Japan. Poor performance on both tests was significantly associated with a higher risk of functional disability, supporting and advancing the existing evidence through a longitudinal study design.
The association between hand dexterity and functional disability likely stems from the essential role of hand function in performing ADLs and IADLs. Tasks such as using utensils or chopsticks and brushing teeth require fine hand function, and the loss of these abilities can necessitate external assistance.33) Therefore, the deterioration of hand dexterity can result in individuals requiring assistance with daily activities, such as eating or bathing independently. While impaired mobility is a primary contributor to functional disability,7) impaired ability with ADL/IADL, such as eating or oral care, also plays a significant role. As hand dexterity begins to decline around age 65,34) regular assessments of hand function and mobility are essential to prevent functional disability.
The hands serve as motor organs that execute commands from the brain while providing critical feedback for movement.13) Fine hand movements depend on intricate brain–nervous system integration, particularly the activation of the primary motor cortex. Hand control signals from this region are transmitted via spinal interneurons to the nerves and muscles responsible for hand function.35) Consequently, smooth brain–nervous system activity is vital for maintaining fine motor skills. Older adults with poor hand function often exhibit cognitive impairments, which can lead to dementia. The deterioration of hand function is associated with progression of cognitive decline22) and recognized as a risk factor for dementia, a key contributor to functional disability.14,15) Interestingly, recent studies have shown that repetitive hand movement training can enhance both hand dexterity and cognitive function in older adults.36,37)
Based on the association between hand dexterity and incident functional disability in this study, we investigated the dose-response relationship to gain more meaningful insights for LTC prevention strategies. We identified this association for the first time, noting that individuals performing worse than the cut points (peg-moving, 37.9/38.0 seconds; circle drawing, 21/20 points) had an increased risk of functional disability. However, values better than these cut points did not significantly reduce the risk of disability. Although improved hand dexterity may enhance the ability to perform ADL, it does not imply a continuous preventive effect on functional disability. This result indicates a minimal hand dexterity standard required for independent ADL. Previous studies corroborate that a certain level of hand dexterity is crucial for maintaining independence in daily life.11,38) To enhance the effectiveness of LTC prevention among older adults with hand dexterity better than the cut point, other physical functions, particularly lower limb function and gait ability, which have traditionally been associated with functional decline,4,7) may be considered. Future research should recognize hand function as an important factor in preventing the need for LTC and evaluate it comprehensively with these other physical measures. We suggest that maintaining hand dexterity better than the cut point in both tests can help support functional independence, and that LTC prevention interventions should prioritize older adults with worse than the cut point.
We utilized peg-moving and circle-drawing tasks to assess hand dexterity. Comparing the AUC of these tests for predicting functional disability, the peg-moving task showed slightly higher predictive power. This task involves complex, multi-joint movements, including grasping pegs with both hands from a distal board and simultaneously inserting them into small holes on a proximal board. This requires precise coordination of wrist, elbow, shoulder, and finger joint movements. In contrast, the circle-drawing task involves simpler movements, such as holding a pen and drawing circles around numbers, with minimal elbow and shoulder joint involvement. Considering that ADL and IADL require the coordination of multiple joints and fine hand function, the complexity of the peg-moving task likely enhances its predictive power for functional disability. Nonetheless, the simplicity of the circle-drawing task, requiring only minimal materials, such as a pen and paper, indicates that it may be a practical and cost-effective option for evaluating hand dexterity.
The strengths of this study include a high follow-up rate and the use of a nationally standardized LTCI certification database over 14 years in Japan, ensuring objective and accurate disability data. Additionally, we used cubic spline analysis to explore the dose-response association between hand dexterity and functional disability. However, this study also had several limitations. The low survey participation rate (15.9%) may introduce sampling bias, with participants likely being healthier older adults. The mean PASE score in this study (123.7 points) was slightly higher than that in a previous Japanese study (114.9 points),29) limiting generalization to less active or less healthy older adults. Moreover, as this study was conducted in a rural area, hand dexterity may be affected by occupational and lifestyle factors,39) and the results may not generalize to urban populations.
In conclusion, throughout the 14-year follow-up, both the peg-moving and circle-drawing tasks were significantly associated with a reduced risk of functional disability and demonstrated a curvilinear dose-response association among older Japanese adults. The spline analysis indicated increased risk when participants performed worse than the cut points (peg-moving, 37.9/38.0 seconds; circle drawing, 21/20 points), and no reduction in risk was observed for those who performed better than these points. These findings suggest that maintaining a sufficient level of hand dexterity can help support functional independence and that LTC prevention interventions should focus on older adults whose performance is worse than these cut points. Given the limited information on interventions for improving hand dexterity, further research is required to develop strategies for maintaining or enhancing this function.
Notes
We would like to thank the study participants and Kasama city officials for their cooperation. We thank Editage (www.editage.com) for the English language editing.
CONFLICT OF INTEREST
The researchers claim no conflicts of interest.
FUNDING
This study was supported by grants from Japan Science and Technology Agency (JPMJPF2017) and Japan Sport Association.
AUTHOR CONTRIBUTIONS
Conceptualization, NL, KTsunoda; Formal analysis, NL, KTsunoda; Data curation, NL, KTsunoda, JS, YA, KN, TT, KF, YF, KTeraoka, TO; Funding acquisition, TO; Project administration, TO; Supervision, KTsunoda, TO; Writing-original draft, NL; Writing-review & editing, KTsunoda, JS, YA, KN, TT, KF, YF, KTeraoka, TO.
DATA AVAILABILITY STATEMENT
The data are not publicly available because of privacy and ethical restrictions. Data supporting the findings of this study are available from the corresponding author upon request.
SUPPLEMENTARY MATERIALS
Supplementary materials can be found via https://doi.org/10.4235/agmr.25.0075.
Participant characteristics (n=1,069)
Hazard ratios for incident functional disability according to peg-moving task and covariates across models
Hazard ratios for incident functional disability according to circle-drawing task and covariates across models
Participant flow. a)To avoid overlapping data when individuals participated multiple times, the earliest valid data without missing values were prioritized. b)Participants were followed up using the city’s municipal database until July 2023.
Hand dexterity test. (A) Peg-moving task: Participants were instructed to move two pegs simultaneously from the distal to the proximal board using both hands as quickly as possible. The time (in seconds) required to move all 48 pegs was recorded. (a-1) Start position: 48 pegs are inserted into the distal board, and participants hold two pegs with both hands to begin the task. (a-2) Participants move two pegs simultaneously using both hands from the distal to the proximal board. (a-3) The task begins from the far-right columns of the distal board and proceeds leftward in two-column increments. (a-4) Task completion: all pegs have been moved to the proximal board. (B) Circle-drawing task: Participants were instructed to circle numbers 1 to 80 in sequential order as quickly as possible. The number of circles completed in 15 seconds was recorded as points.
Adjusted Kaplan–Meier curves for incidence of functional disability by peg-moving (A) and circle-drawing task (B) based on Model 2 of Table 3. Range of peg-moving task is ≤34.49 seconds for tertile 1, 34.50–38.78 seconds for tertile 2, and ≥38.79 seconds for tertile 3. Range of circle-drawing task is ≥27 points for tertile 1, 21–26 points for tertile 2, and ≤20 points for tertile 3. HR, hazard ratio; CI, confidence interval.
Receiver operating characteristic curves for predicting functional disability based on peg-moving task (A) and circle-drawing task (n=980). The area under the curves (AUC) were 0.74 for the peg-moving task and 0.70 for the circle-drawing task. The optimal cutoff points were 37.9/38.0 seconds for the peg-moving task and 21/20 points for the circle-drawing task, respectively.