Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jul 1.
Published in final edited form as: Int J Geriatr Psychiatry. 2020 Oct 24;36(3):433–442. doi: 10.1002/gps.5441

Challenging the relationship of grip strength with cognitive status in older adults

Andrew Hooyman 1, Michael Malek-Ahmadi 2, Elizabeth B Fauth 3, Sydney Y Schaefer 4
PMCID: PMC13317022  NIHMSID: NIHMS2092621  PMID: 33027842

Abstract

Objective:

Grip strength is a widely-used motor assessment in aging research, and has repeatedly been shown to be associated with cognition. It has been proposed that grip strength could enhance cognitive screening in experimental or clinical research, but this paper uses multiple data-driven approaches to caution against this interpretation. Further, we introduce an alternative motor assessment, comparable to grip dynamometry, but has a more robust relationship with cognition among older adults.

Design:

Associations between grip strength and cognition (measured with the Montreal Cognitive Assessment) were analyzed cross-sectionally using multivariate regression in two datasets: 1) The Irish LongituDinal Study on Aging (TILDA; N=5,980, community-dwelling adults ages 49–80) and 2) an experimental dataset (N=250, community-dwelling adults aged 39–98). Additional statistical simulations on TILDA tested how ceiling effects or skewness in these variables influenced these associations for quality control.

Results:

Grip strength was significantly but weakly associated with cognition, consistent with previous studies. Simulations revealed this was not due to skewness/ceiling effects. Conversely, a new alternative motor assessment (functional reaching) had a stronger, more robust, and more sensitive relationship with cognition compared to grip strength.

Conclusions:

Grip strength should be cautiously interpreted as being associated with cognition. However, functional reaching may have a stronger and clinically useful relationship with cognition.

Keywords: Grip Strength, Physical Function, Cognition, Data Simulation

INTRODUCTION

Grip strength has been used widely in clinical research and epidemiological studies of aging to estimate frailty or muscle mass in older adults1. In fact, over 210 large-scale longitudinal studies archived in the Inter-university Consortium for Political and Social Research (ICPSR) report at least one wave of grip strength data in their samples. Although it is generally accepted as a measure of physical or motor function, there is a growing number of studies that have demonstrated its statistically significant relationship to cognition, either cross-sectionally/at baseline2–9 or longitudinally2,3,7,10–14 in both non-demented and demented samples. This statistical relationship has coined the phrase, “People who grip better, think better”8, and suggests that motor or physical function (i.e., neural drive to skeletal muscle) and certain cognitive functions share overlapping neurological processes. As a result, some studies have advocated for using motor measures to better identify dementia or Alzheimer’s disease risk2,10,15. For example, grip strength has been shown to predict decline in several cognitive abilities in non-demented older adults4,16. As a result, it has been interpreted that “maximum grip strength testing provides a discriminating measure of neurological function,” making it a potentially valuable motor measure that may track with2,7, or even precede2,3,12,16,17, declines in cognition, perhaps as early as middle-age11.

While the correlation between cognitive status and grip strength may be attractive from a clinical or theoretical perspective18,19 this statistically significant relationship remains weak in even large-scale studies and is confounded by ceiling effects in commonly-used cognitive measures (e.g., MMSE, MoCA) and skewed distributions of grip strength. Using publicly-available cross-sectional population data, new experimental data, and data simulations in community-dwelling adults, we quantify these limitations and propose an alternative upper extremity motor assessment, which is comparable to grip strength in terms of time and cost yet has a stronger and more sensitive relationship to cognitive status that make it a more viable measure for studying interactions between cognition and movement, as well as measuring physical function.

MATERIALS AND METHODS

Participants

Two cross-sectional datasets were used to illustrate relationships between cognitive status and motor function in community-dwelling samples: the publicly-available Irish Longitudinal Study of Ageing (TILDA)20, and a new experimental dataset. Additionally, we performed simulations modeling a ‘ground-truth’ relationship between grip strength and cognitive status with and without ceiling effects and skewness.

The Irish Longitudinal Study of Ageing (TILDA) dataset

The Irish Longitudinal Study of Ageing (TILDA)21–23 was first used in this report to examine the relationship with grip strength and cognition. To date, no studies have used TILDA for this purpose. We selected the first wave of TILDA for this reason, and because it is the largest publicly-available dataset in the ICPSR that has both grip strength and cognitive status measures. TILDA data are freely accessible through the ICPSR (https://www.icpsr.umich.edu/icpsrweb/ICPSR/). Cognition was reported in TILDA using the Montreal Cognitive Assessment (MoCA)24. The first wave of data contained 8,504 adults, and from this, 2,522 individuals were excluded due to a lack of MoCA or grip strength data. The resulting dataset was comprised of 5,980 adults (3,290, or 55%, female), with a mean±SD age of 62.29±9.06 years old (range: 49–80), total MoCA score of 24.78±3.72 (range: 2–30), and grip strength of 25.93±9.89 kilograms. Grip strength was measured via dynamometry within TILDA20 as the average of two attempts of maximal isometric dominant hand grip force.

Experimental Dataset

To compare the relationship between cognitive status and an alternative upper extremity motor assessment, we used a new experimental dataset compiled in our laboratory that included MoCA and dominant hand grip strength data, as well as data from another motor assessment (functional reaching; see below for more details) from 250 adults (121, or 45%, female), with a mean±SD age of 73.12±8.21 years old (range: 39–98), total MoCA score of 25.24±2.7 (range: 17–30), and grip strength of 25.65±9.31 kilograms. In this dataset, grip strength was measured via dynamometry, and reported as the average of three attempts of maximal isometric grip force based on recommended protocol25. This study was approved by two university Institutional Review Boards, and all participants in this dataset provided informed consent. Both TILDA and the experimental datasets recruited community-dwelling older adults. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for reporting observational studies (see Supplementary Material).

Functional reaching: An alternative motor assessment

Despite the potential methodological concerns in relating grip strength to cognition, it is noted that other motor tasks may be of value, particularly since coordination of the upper extremity may be sensitive to cognitive aging26,27. Upper extremity tasks that involve functional movement likely rely on more cognitive processes than maximal force production, which recruits primarily sensorimotor cortical regions. Movements such as reaching, grasping, and object manipulation are fundamental to activities of daily living thereby suggesting that a task involving such motor components could better predict cognition than grip strength.

Our laboratory has developed a quick, easy-to-administer motor task that objectively evaluates upper extremity dexterity and coordination. This task has been validated against other established upper extremity tasks28 and is feasible for a wide range of ages and health conditions29–32. To summarize, participants are required to use a standard plastic spoon with their nondominant hand to acquire two raw kidney beans at a time from a central cup to one of three distal cups arranged at a radius of 16 cm at −40°, 0°, and 40° relative to the central cup, starting with the cup ipsilateral to their nondominant hand, then to the middle cup, then the contralateral cup. All cups were 9.5cm in diameter and 5.8cm deep. This sequence was repeated four more times for a total of 15 out-and-back movements. Performance was measured as trial time (in seconds; lower = better). A full visual description of this task can be viewed on Open Science Framework (https://osf.io/u7ynx/) and is described at length in28. Unlike grip strength, this task does not exhibit any sex differences (p=.67), highlighting its advantage as a motor measure. As noted above, functional reaching data were collected in our experimental dataset (n=250), along with grip strength and MoCA. Similar regression analyses were used to assess the relationship between functional reaching and MoCA.

Statistical analyses

Calculation of Predicted Range of MoCA scores

We first calculated the range of predicted MoCA scores based on actual grip strength (and functional reaching) values using both datasets. This modeled range illustrates the possible MoCA scores that grip strength (and functional reaching) could predict in each dataset, allowing us to determine the overall “coverage” of the dataset for each measure, i.e., the percentage of participants who would fall within the bounds of the fitted model based on their grip strength (or functional reaching) scores. To do so, linear regression between total MoCA score and grip strength (or functional reaching) were performed. The predicted fitted values for each regression were then calculated. To statistically compare grip strength and functional reaching models from our experimental dataset, we compared calculated Akaike information criteria (AIC) values, which estimate the in-sample prediction error. The magnitude of AIC difference between these models allows us to determine if model accuracy was significantly different between models. Next, the predictive range of MoCA for each independent variable was derived from the maximum and minimum fitted value from each resultant regression model. For example, the maximum and minimum predicted values from the regression model for MoCA were 23 and 27, respectively, which then allowed us to determine the actual percentage of participants that could be predicted by grip strength (or functional reaching). To ensure accuracy of the regression coefficient for grip strength, multivariate regression was also performed to control for age, sex, and education to ensure consistent relationships throughout. Additionally, published cut-off scores for the MoCA of 26 and 21 were used to illustrate the extent to which these variables (grip strength or functional reaching) could be used theoretically to identify mild impairment or suspected dementia, respectively24,33. Lastly, a binomial generalized linear regression model was performed to analyze and compare the classification accuracy of grip strength and reaching performance on MoCA score cut-offs of 26 and 21. Receiver operator characteristic (ROC) curves were used to determine the sensitivity/specificity of both grip strength and functional reaching as classifiers based on these cut-off MoCA scores. We acknowledge that a number of other co-morbid factors likely mediate the relationship between physical and cognitive measures, but we have restricted our analysis to key variables for purposes of model simplicity and clinical availability. All analyses were conducted in R, version 3.5.5.

Simulation of TILDA Dataset: Proof-of-concept

Since previous studies of other motor assessments have demonstrated that the presence of a ‘hard ceiling’ (upper bound)35,36 or skewness37 in a set of variables can distort the amount of explained variance and coefficients in linear regression, it is plausible that since the MoCA has a maximum score of 30, it underestimates a true relationship between grip strength and cognition whereby using a different cognitive test with theoretically minimal upper bound might show a stronger relationship with grip strength. It is also plausible that the skewed distribution of grip strength may also underestimate this association, whereby a stronger relationship with cognition would be observed if a more normally-distributed sample of grip strengths were evaluated. To test (and rule out) these two measurement concerns while considering multiple evaluations of the data for transparency34, we performed a simulation where we generated a linear relationship between a normally distributed independent variable (X) and a resulting dependent variable (Y), containing no ceiling effect, with a known slope of 100 and known intercept of 1,000. The simulated data set was composed of 5000 points, and the variance explained was similar to that of TILDA (R2 = 0.04). We refer to this base model, Y=100*X+1000, as the “ground truth” relationship (Fig. 1C). We then iteratively applied 1000 simulations with varying degrees of positive and negative skew to the independent variable, X: simulated grip strength, with the known dependent variable, Y: simulated MoCA, to test whether less skew and minimal ceiling effects improved the overall variance explained (R2) and the resultant regression coefficient of X. To further test the influence of ceiling effects, we reran the simulation, but with the addition of a ceiling effect on the dependent variable. A ceiling effect was imposed by calculating the upper bound of the third quantile, and then setting all values above that to the bound value itself. The regression coefficients and variance explained, R2 values, for each simulation were recorded and used for visualization purposes.

Figure 1.

Figure 1.

Open in a new tab

Effects of skewness and ceiling effect on resultant variance explained, regression coefficient, prediction accuracy and predictive range inclusion. A) Resulting regression coefficients and B) R2 values for linear models with varying degrees of positive skew in the independent variable (IV) and the presence of a ceiling effect in the dependent variable (DV) at the third quartile range in the independent variable. Points i, ii, iii and iv in these figures represent example cases of the True relationship (i), and a skewed relationship with and without ceiling effects (ii, iii and iv) like those seen in the TILDA dataset. C) Each example case has the histogram of independent and dependent variable plotted along its border in this panel. i) True relationship with both independent and dependent variables being normally distributed and no ceiling effect of the dependent variable. ii) Relationship and resulting linear model of the independent variable with positive skew and no ceiling effect of the dependent variable. iii) Relationship and resulting linear model of the independent variable with no skew and a ceiling effect on the dependent variable. iv) Relationship and resulting linear model of the independent variable with positive skew and a ceiling effect on the dependent variable.

RESULTS

Relationship between Grip Strength and MoCA

Consistent with previous findings19 regression analysis revealed a significant linear relationship between grip strength and MoCA scores in TILDA with (coefficient=.052; R2=.02, p<.001; 95%CI [.037, .061]) and without (coefficient=0.062; R2=0.027; p<.001; 95%CI [.052, .071]) controlling for age, sex, and education. A linear relationship was also observed in our experimental dataset when controlling for these factors (coefficient=0.03; R2=0.01; p<.05, 95%CI [5.1e-5, .071]). For reference, these coefficients are similar to those from other published studies3,10. Confidence intervals from each regression analysis demonstrate that the true relationship (slope) between grip strength and cognition likely exists within this narrow range (5.1e-5 to .071), which suggests that the inclusion of additional data would not substantially increase the regression coefficient. Based on these regression results, the range of MoCA scores predicted by actual grip strength values were calculated for both datasets, as indicated by shaded gray rectangles in Figure 2A and B. Notably, only about half of the MoCA scores (53%) across the two datasets fall into this predicted range that spans scores of 23 to 27 only. This is primarily due to the low grip strength coefficient for TILDA (0.062; 95%CI [.052, .071]), as well as for our own dataset (0.03; 95%CI [5.1e-5, .071]). Results indicated that even considering the maximum possible coefficient for grip strength of 0.071 would only predict a one-point change in MoCA score with every 14-kg change in grip strength. For reference, grip strength for a typical female age over 75 years of age is 42.6 kg (range: 25 kg to 65 kg)38. This also suggests that grip strength would be incapable of identifying a person with either mild impairment or suspected dementia based on optimal cut-off scores of 25 or 21 respectively33.

Figure 2.

Figure 2.

Open in a new tab

Linear fits of grip strength or functional reaching (FR) task against total MoCA score among community dwelling middle aged and older adults. Dashed line represents MoCA cutoff score for individuals with mild impairment (i.e., below 26). Long dashed line represents MoCA cutoff score for individuals with suspected dementia (i.e., below 21), according to33. Gray shaded box in each graph represent the predictive range of each model fit given the data in each graph. In all cases, values towards the left side of the x-axis indicate worse upper-extremity function. The x-axis in graph C is reversed for visualization purposes for comparison of model slope to other model slopes in graph A and B. Jitter has been applied to data points so distribution of data can be better visualized.

Simulated Results of Skewness and Ceiling Effects on Regression Outcomes

Because grip strength values in TILDA were moderately positively skewed (~0.5), we ran our simulation on data transformations of skewness ranging from −0.35 to 1.8. As shown in the Figure 1A and B, black points are used for reference to the ground-truth simulation and a simulated equivalent grip strength skewness found in TILDA with and without dependent variable ceiling effects. We chose to model a linear relationship, as most studies relating grip strength and cognitive status do so2,3,12. Simulation results revealed two key findings regarding the “ground-truth” relationship between simulated grip strength and simulated MoCA as skewness in the grip strength variable increases: 1) the regression slope approaches zero, and 2) the amount of variance explained is reduced. While these findings do suggest that the degree of skewness does in fact impact a variable’s ability to predict another (see gray line in Fig. 1A and B), our simulation indicates that the coefficient and variance explained remain low even when skew is equivalent to that of grip strength in TILDA (point iv and v in Fig. 1A and B). This indicates that transforming grip strength data to be normally distributed (or sampling from a more normally distributed cohort) would likely not increase its prediction of cognitive status. In other words, the weak association of grip strength with cognition is not attributed to the fact that grip strength is often skewed to the right in community-dwelling samples.

The simulation also investigated how ceiling effects in the dependent variable (i.e., MoCA) affect the regression coefficient and variance explained. It appeared that a ceiling effect reduced the overall regression coefficient by approximately 10%, regardless of how skewed the independent variable was. This suggests that if a cognitive assessment with no hard ceiling were used, then the grip strength regression coefficient would increase only by .006. In other words, the weak association between grip strength and MoCA is not attributable simply to the fact that the MoCA has a ceiling effect. Comparison of each case with and without TILDA-equivalent skewness (0.5) and a ceiling effect can be visualized in the Figure 1C i, ii, iii, iv respectively.

Relationship between Functional Reaching Task and MoCA

Regression analysis of our experimental dataset indicated that the functional reaching task was significantly related to MoCA score (p<.001, R2=0.08), functional reaching coefficient=−0.084, 95%CI [−0.12, −0.05]). Even when controlling for factors of age, sex, education, and grip strength, functional reaching is still significantly associated with total MoCA score (p<.001, coefficient=−.06, 95%CI [−.1, −.03]). Comparing AIC values between the multivariate regression models for grip strength (AIC=1177) and functional reaching (AIC=1166) indicates that functional reaching explained a larger portion of the variance compared to grip strength (since lower AIC values are better). Based on39, a difference of ≥10 in AIC values between models indicates that the grip strength model is less accurate than that of the functional reaching model. Overall, functional reaching explained a larger portion of the variance compared to grip strength in both datasets. Moreover, the predicted range of MoCA scores calculated from actual functional reaching scores (see Fig. 2C) was wider (scores from 21 to 30) and included a much larger percentage of the participants compared to that from actual grip strength scores (95% vs. 31%). Furthermore, the regression coefficient of −.084 indicated that an 11-second change on the functional reaching task would yield a one-point change in MoCA score between individuals. In the experimental dataset, mean functional reaching performance was 51.89 seconds with a range of 36.97 to 98.62 seconds. Lastly, binary MoCA classification based on cut-off scores of 26 or 21 using either grip strength or functional reaching performance as the dependent variable further demonstrated the limited capability of classifying cognitive status using grip strength (Table 1). Instead, functional reaching was superior to grip strength in both TILDA and experimental datasets for classifying both MoCA cut-off scores (Fig. 3). Thus, compared to grip strength, the functional reaching task is more sensitive to differences in cognition. It is important to note that in the experimental dataset, grip strength, and functional reaching were not correlated (r=.018; p=.77), suggesting that functional reaching is related to cognition through a pathway that is independent of grip strength. This also ensures that these two motor assessments are not redundant or collinear with one another.

Table 1: Results of MoCA classification by Dataset and Motor Task.

Comparison of classification results (statistical significance, odds ratio and area under the curve, AUC) for grip strength and functional reaching for MoCA score cutoffs of 26 and 21.

Dataset Measure Generalized Linear Model MoCA26 p-value Generalized Linear Model MoCA21 p-value Odds Ratio MoCA26 [95% CI] Odds Ratio MoCA21 [95% CI] AUC MoCA26 AUC MoCA21
TILDA Grip strength <.001 <.001 1.02 [1.01 – 1.03] 1.05 [1.04 – 1.06] .57 .63
Experimental Grip strength >.05 >.05 1.03 [.99 – 1.05] 1.04 [.98 – 1.21] .57 .59
Experimental Functional reaching <.001 <.01 1.05 [1.02 – 1.08] 1.06 [1.02 – 1.11] .64 .71
Open in a new tab

Figure 3.

Figure 3.

Open in a new tab

Receiver operating characteristics for classification of MoCA cut-off scores of (A) 26 and (B) 21 using grip strength from the TILDA dataset (light gray) and the experimental dataset (black), as well as functional reaching performance from the experimental dataset (dark gray). Solid lines represent the Receiver operating characteristics curves for each measure. Dashed line represents a reference line for an area under the curve (AUC) of .5, which would be classification accuracy equivalent to chance.

DISCUSSION

Based on publicly-available cross-sectional population data, new experimental data, and data simulations in community-dwelling adults, findings from this report demonstrate limitations in clinically interpreting the relationship between grip strength and cognition. These results point to the limited association between cognition and grip strength based small regression slopes regardless of grip strength distribution skewness (as supported by our simulations). For example, the entire range of grip strength values in multiple datasets of older adults spans only 3 to 4 points on the MoCA, which is particularly problematic since the MoCA’s reported standard deviation is ±340. This report also expands on previous work that has investigated grip strength and cognitive relationships in only a single sex, ethnic group, or age4,12 by utilizing two independent datasets that include diverse groups of older adults based on age, sex, and education. Although findings from this report focus on the relationship between grip strength and cognitive status cross-sectionally, it is noted that grip strength has also been used as a prognostic tool to predict future risk of developing dementia, AD, or general cognitive decline2,10,15. The concerns raised by this report regarding the use of grip strength to predict cognitive status or decline also generalize to prospective studies, particularly since the hazard ratios in these studies (with exception to Boyle et al., 2009) remained very close to 1. While grip strength may be associated with cognition (both at a single time point and longitudinally), this association is not strong enough to justify its use as a correlate of cognition or perhaps even a predictor of cognitive decline. Even though the voluntary maximal recruitment of hand muscles may engage some degree of cognitive processing19, it does not appear to be as sensitive to cognition as more complex motor behaviors (like functional reaching) are.

This study was not designed to undermine the experimental and clinical value of measuring muscle strength1,41–43, nor to advocate for using motor measure to replace cognitive screening. It is even noted that in the context of cognitive status and decline, muscle strength measures can provide insight when gathered from both peripheral and axial muscles10. However, Boyle and colleagues evaluated 11 muscle groups to calculate Z-scores and hazard ratios, making it much more challenging to translate this approach to a clinical setting with limited time and equipment. Additionally, their results did not indicate if it were possible to predict individuals with lower Z-score values, much like how we show here that grip strength cannot predict MoCA scores that are on the lower end.

We acknowledge that using grip strength to study the overlap between cognition and motor function is attractive since there is a need for motor assessments designed and validated for MCI44, particularly ones that are easy, fast, and inexpensive to administer. However, grip strength can vary widely and, as described throughout this study, has a very limited range in which it is related to cognition and is therefore not very sensitive. Other well-established motor assessments such as the Timed Up and Go (TUG) also have a very limited relationship with cognition in community-dwelling older adults45. Instead, our data support the use of an alternative motor task that assesses physical function in a way that is highly relevant to activities of daily living46 (more so than maximal grip force), and that draws more heavily on cognitive processes than grip strength or the TUG30,47,48 Recent studies are beginning to identify which specific aspects of cognition are probed by this task49, but future work is necessary to understand the underlying mechanisms that explain such associations.

We acknowledge the following limitations within our data simulation: 1) the estimation of unexplained variance within the relationship and 2) the placement of the ceiling on the dependent variable. In the case of variance explained within our true relationship, we chose to input noise that would yield a similar relationship specific to grip strength and MoCA in the TILDA dataset. Thus, our simulation is specific to the underlying relationship investigated and may not generalize to other relationships with stronger or weaker ground-truth relationships. In terms of a dependent variable ceiling effect, we assumed that the ceiling effect would occur somewhere in the upper quartile of our true relationship rather than the median or below the median. Since there is no known cognitive measure that is not limited to such an effect, we placed our ceiling based on visual inspection of our grip strength to MoCA relationship from the TILDA data. The placement of ceiling can only be confirmed if an alternative numerical cognitive assessment were compared to MoCA, which is measured on an ordinal scale.

Conclusion

It is clear from previous studies that physical function (i.e., neural drive to muscle) and cognitive function share common neurological processes, which points to the potential and value of studying associations between cognitive and motor function particularly in older adults where one may precede the other. Though grip strength is a widely-used assessment of physical/motor function, this report demonstrates the limitations of using it as a correlate of cognition based on multiple data and analytic approaches. This report also provides empirical support for an alternative measure of motor function that is comparable to grip strength in terms of cost and time but with fewer statistical limitations.

Supplementary Material

Supplementary Material

Key points:

  • Grip strength is significantly yet weakly related to cognitive status.

  • This relationship is not due simply to measurement features like ceiling effects.

  • A new motor assessment proposed here is more robust and sensitive to cognitive status than grip strength.

Acknowledgements:

We would like to thank the New Adventures in Learning program for their assistance and input. This work was supported by the National Institutes of Health [grant number K01AG047926 to SYS] and the Marriner S. Eccles Foundation. These sponsors did not have any role in the following: the study design; the collection, analysis, or interpretation of data; the writing of the paper; nor the decision to submit the paper for publication.

Footnotes

Conflict of Interest: The authors have no disclosures to report.

Contributor Information

Andrew Hooyman, Department of Biological and Health Systems Engineering, Arizona State University, Tempe, Arizona, USA.

Michael Malek-Ahmadi, Banner Alzheimer’s Institute, Phoenix, Arizona, USA..

Elizabeth B. Fauth, Department of Human Development and Family Studies, Utah State University, Logan, Utah, USA.

Sydney Y. Schaefer, Department of Biological and Health Systems Engineering, Arizona State University, Tempe, Arizona, USA

DATA AVAILABILITY STATEMENT

Some of the data that support the findings of this study are available in the Inter-university Consortium for Political and Social Research at doi: 10.3886/ICPSR34315.v2, reference number 34315. These data were derived from the following resources available in the public domain: https://www.icpsr.umich.edu/web/NACDA/studies/34315. The remaining data that support the findings of this study are available from the corresponding author upon reasonable request.

REFERENCES

  • 1.Giampaoli S Hand-grip strength predicts incident disability in non-disabled older men. Age Ageing. 1999;28(3):283–288. doi: 10.1093/ageing/28.3.283 [DOI] [PubMed] [Google Scholar]
  • 2.Buchman AS, Wilson RS, Boyle PA, Bienias JL, Bennett DA. Grip Strength and the Risk of Incident Alzheimer’s Disease. Neuroepidemiology. 2007;29(1–2):66–73. doi: 10.1159/000109498 [DOI] [PubMed] [Google Scholar]
  • 3.Sternäng O, Reynolds CA, Finkel D, Ernsth-Bravell M, Pedersen NL, Dahl Aslan AK. Grip Strength and Cognitive Abilities: Associations in Old Age. Journals Gerontol Ser B. 2016;71(5):841–848. doi: 10.1093/geronb/gbv017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Praetorius Björk M, Johansson B, Hassing LB. I forgot when I lost my grip-strong associations between cognition and grip strength in level of performance and change across time in relation to impending death. Neurobiol Aging. 2016;38:68–72. doi: 10.1016/j.neurobiolaging.2015.11.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Clouston SAP, Brewster P, Kuh D, et al. The Dynamic Relationship Between Physical Function and Cognition in Longitudinal Aging Cohorts. Epidemiol Rev. 2013;35(1):33–50. doi: 10.1093/epirev/mxs004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Shin H-Y, Kim S-W, Kim J-M, Shin I-S, Yoon J-S. Association of grip strength with dementia in a Korean older population. Int J Geriatr Psychiatry. 2012;27(5):500–505. doi: 10.1002/gps.2742 [DOI] [PubMed] [Google Scholar]
  • 7.Sprague BN, Phillips CB, Ross LA. Age-Varying Relationships Between Physical Function and Cognition in Older Adulthood. Journals Gerontol Ser B. 2019;74(5):772–784. doi: 10.1093/geronb/gbx126 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Deary IJ, Johnson W, Gow AJ, et al. Losing One’s Grip: A Bivariate Growth Curve Model of Grip Strength and Nonverbal Reasoning From Age 79 to 87 Years in the Lothian Birth Cohort 1921. Journals Gerontol Ser B Psychol Sci Soc Sci. 2011;66B(6):699–707. doi: 10.1093/geronb/gbr059 [DOI] [PubMed] [Google Scholar]
  • 9.Vancampfort D, Stubbs B, Firth J, Smith L, Swinnen N, Koyanagi A. Associations between handgrip strength and mild cognitive impairment in middle-aged and older adults in six low- and middle-income countries. Int J Geriatr Psychiatry. 2019;34(4):609–616. doi: 10.1002/gps.5061 [DOI] [PubMed] [Google Scholar]
  • 10.Boyle PA, Buchman AS, Wilson RS, Leurgans SE, Bennett DA. Association of Muscle Strength With the Risk of Alzheimer Disease and the Rate of Cognitive Decline in Community-Dwelling Older Persons. Arch Neurol. 2009;66(11):1339–1344. doi: 10.1001/archneurol.2009.240 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Liu Y, Cao X, Gu N, Yang B, Wang J, Li C. A Prospective Study on the Association Between Grip Strength and Cognitive Function Among Middle-Aged and Elderly Chinese Participants. Front Aging Neurosci. 2019;11. doi: 10.3389/fnagi.2019.00250 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Alfaro-Acha A, Snih SA, Raji MA, Kuo Y-F, Markides KS, Ottenbacher KJ. Handgrip Strength and Cognitive Decline in Older Mexican Americans. Journals Gerontol Ser A Biol Sci Med Sci. 2006;61(8):859–865. doi: 10.1093/gerona/61.8.859 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Kim KH, Park SK, Lee DR, Lee J. The Relationship between Handgrip Strength and Cognitive Function in Elderly Koreans over 8 Years: A Prospective Population-Based Study Using Korean Longitudinal Study of Ageing. Korean J Fam Med. 2019;40(1):9–15. doi: 10.4082/kjfm.17.0074 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Robitaille A, Piccinin AM, Hofer SM, Johansson B, Muniz Terrera G. An examination of the heterogeneity in the pattern and association between rates of change in grip strength and global cognition in late life. A multivariate growth mixture modelling approach. Age Ageing. 2018;47(5):692–697. doi: 10.1093/ageing/afy048 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Wang L, Larson EB, Bowen JD, van Belle G. Performance-Based Physical Function and Future Dementia in Older People. Arch Intern Med. 2006;166(10):1115. doi: 10.1001/archinte.166.10.1115 [DOI] [PubMed] [Google Scholar]
  • 16.Okely JA, Deary IJ. Associations Between Declining Physical and Cognitive Functions in the Lothian Birth Cohort 1936. Newman A, ed. Journals Gerontol Ser A. January 2020. doi: 10.1093/gerona/glaa023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Camargo EC, Weinstein G, Beiser AS, et al. Association of Physical Function with Clinical and Subclinical Brain Disease: The Framingham Offspring Study. J Alzheimer’s Dis. 2016;53(4):1597–1608. doi: 10.3233/jad-160229 [DOI] [PubMed] [Google Scholar]
  • 18.Carson RG. What accounts for the association between grip strength and mental functioning in aging people? Maturitas. 2020;0(0). doi: 10.1016/j.maturitas.2020.01.010 [DOI] [PubMed] [Google Scholar]
  • 19.Carson RG. Get a grip: individual variations in grip strength are a marker of brain health. Neurobiol Aging. 2018;71:189–222. doi: 10.1016/j.neurobiolaging.2018.07.023 [DOI] [PubMed] [Google Scholar]
  • 20.Kenny RA. The Irish Longitudinal Study on Ageing, 2009 – 2011. Inter-university Consort Polit Soc Res. 2018. [Google Scholar]
  • 21.Kearney PM, Cronin H, O’Regan C, et al. Cohort Profile: The Irish Longitudinal Study on Ageing. Int J Epidemiol. 2011;40(4):877–884. doi: 10.1093/ije/dyr116 [DOI] [PubMed] [Google Scholar]
  • 22.Whelan BJ, Savva GM. Design and Methodology of The Irish Longitudinal Study on Ageing. J Am Geriatr Soc. 2013;61:S265–S268. doi: 10.1111/jgs.12199 [DOI] [PubMed] [Google Scholar]
  • 23.Cronin H, O’Regan C, Finucane C, Kearney P, Kenny RA. Health and Aging: Development of The Irish Longitudinal Study on Ageing Health Assessment. J Am Geriatr Soc. 2013;61:S269–S278. doi: 10.1111/jgs.12197 [DOI] [PubMed] [Google Scholar]
  • 24.Nasreddine ZS, Phillips NA, Bedirian V, et al. The Montreal Cognitive Assessment, MoCA: A Brief Screening Tool For Mild Cognitive Impairment. J Am Geriatr Soc. 2005;53(4):695–699. doi: 10.1111/j.1532-5415.2005.53221.x [DOI] [PubMed] [Google Scholar]
  • 25.van der Ploeg RJ, Fidler V, Oosterhuis HJ. Hand-held myometry: reference values. J Neurol Neurosurg Psychiatry. 1991;54(3):244–247. doi: 10.1136/jnnp.54.3.244 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Rogojin A, Gorbet DJ, Hawkins KM, Sergio LE. Cognitive-Motor Integration Performance Is Affected by Sex, APOE Status, and Family History of Dementia. J Alzheimer’s Dis. 2019;71(2):685–701. doi: 10.3233/jad-190403 [DOI] [PubMed] [Google Scholar]
  • 27.Toosizadeh N, Ehsani H, Wendel C, Zamrini E, Connor KO, Mohler J. Screening older adults for amnestic mild cognitive impairment and early-stage Alzheimer’s disease using upper-extremity dual-tasking. Sci Rep. 2019;9(1):10911. doi: 10.1038/s41598-019-46925-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Schaefer SY, Hengge CR. Testing the concurrent validity of a naturalistic upper extremity reaching task. Exp Brain Res. 2016;234(1):229–240. doi: 10.1007/s00221-015-4454-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Schaefer SY, Dibble LE, Duff K. Efficacy and Feasibility of Functional Upper Extremity Task-Specific Training for Older Adults With and Without Cognitive Impairment. Neurorehabil Neural Repair. 2015;29(7):636–644. doi: 10.1177/1545968314558604 [DOI] [PubMed] [Google Scholar]
  • 30.Schaefer SY, Patterson CB, Lang CE. Transfer of training between distinct motor tasks after stroke: implications for task-specific approaches to upper-extremity neurorehabilitation. Neurorehabil Neural Repair. 2013;27(7):602–612. doi: 10.1177/1545968313481279 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Paul SS, Dibble LE, Olivier GN, Walter C, Duff K, Schaefer SY. Dopamine replacement improves motor learning of an upper extremity task in people with Parkinson disease. Behav Brain Res. 2020;377:112213. doi: 10.1016/j.bbr.2019.112213 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Schaefer SY, Duff K. Within-session and one-week practice effects on a motor task in amnestic mild cognitive impairment. J Clin Exp Neuropsychol. 2017;39(5):473–484. doi: 10.1080/13803395.2016.1236905 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Hoops S, Nazem S, Siderowf AD, et al. Validity of the MoCA and MMSE in the detection of MCI and dementia in Parkinson disease. Neurology. 2009;73(21):1738–1745. doi: 10.1212/wnl.0b013e3181c34b47 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Landis SC, Amara SG, Asadullah K, et al. A call for transparent reporting to optimize the predictive value of preclinical research. Nature. 2012;490(7419):187–191. doi: 10.1038/nature11556 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Austin PC, Brunner LJ. Type I Error Inflation in the Presence of a Ceiling Effect. Am Stat. 2003;57(2):97–104. doi: 10.1198/0003130031450 [DOI] [Google Scholar]
  • 36.Hawe RL, Scott SH, Dukelow SP. Taking Proportional Out of Stroke Recovery. Stroke. 2019;50(1):204–211. doi: 10.1161/strokeaha.118.023006 [DOI] [PubMed] [Google Scholar]
  • 37.Ho AD, Yu CC. Descriptive Statistics for Modern Test Score Distributions. Educ Psychol Meas. 2015;75(3):365–388. doi: 10.1177/0013164414548576 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Mathiowetz V, Kashman N, Volland G, Weber K, Dowe M, Rogers S. Grip and pinch strength: normative data for adults. Arch Phys Med Rehabil. 1985;66(2):69–74. [PubMed] [Google Scholar]
  • 39.Burnham KP, Anderson DR. Multimodel Inference: Understanding AIC and BIC in Model Selection. Sociol Methods Res. 2004;33(2):261–304. doi: 10.1177/0049124104268644 [DOI] [Google Scholar]
  • 40.Malek-Ahmadi M, O’Connor K, Schofield S, Coon DW, Zamrini E. Trajectory and variability characterization of the Montreal cognitive assessment in older adults. Aging Clin Exp Res. 2018;30(8):993–998. doi: 10.1007/s40520-017-0865-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Roberts HC, Denison HJ, Martin HJ, et al. A review of the measurement of grip strength in clinical and epidemiological studies: towards a standardised approach. Age Ageing. 2011;40(4):423–429. doi: 10.1093/ageing/afr051 [DOI] [PubMed] [Google Scholar]
  • 42.Sallinen J, Stenholm S, Rantanen T, Heliövaara M, Sainio P, Koskinen S. Hand-Grip Strength Cut Points to Screen Older Persons at Risk for Mobility Limitation. J Am Geriatr Soc. 2010;58(9):1721–1726. doi: 10.1111/j.1532-5415.2010.03035.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Wickham C, Cooper C, Margetts BM, Barker DJP. Muscle Strength, Activity, Housing and the Risk of Falls in Elderly People. Age Ageing. 1989;18(1):47–51. doi: 10.1093/ageing/18.1.47 [DOI] [PubMed] [Google Scholar]
  • 44.Jekel K, Damian M, Wattmo C, et al. Mild cognitive impairment and deficits in instrumental activities of daily living: a systematic review. Alzheimers Res Ther. 2015;7(1):17. doi: 10.1186/s13195-015-0099-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Donoghue O, Feeney J, O’Leary N, Kenny RA. Baseline Mobility is Not Associated with Decline in Cognitive Function in Healthy Community-Dwelling Older Adults: Findings From The Irish Longitudinal Study on Ageing (TILDA). Am J Geriatr Psychiatry. 2018;26(4):438–448. doi: 10.1016/j.jagp.2017.11.006 [DOI] [PubMed] [Google Scholar]
  • 46.Katz S, Downs TD, Cash HR, Grotz RC. Progress in development of the index of ADL. Gerontologist. 1970;10(1):20–30. doi: 10.1093/geront/10.1_part_1.20 [DOI] [PubMed] [Google Scholar]
  • 47.Lingo VanGilder J, Walter CS, Hengge CR, Schaefer SY. Exploring the relationship between visuospatial function and age-related deficits in motor skill transfer. Aging Clin Exp Res. September 2019. doi: 10.1007/s40520-019-01345-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Fauth EB, Schwartz S, Tschanz JT, Østbye T, Corcoran C, Norton MC. Baseline disability in activities of daily living predicts dementia risk even after controlling for baseline global cognitive ability and depressive symptoms. Int J Geriatr Psychiatry. 2013;28(6):597–606. doi: 10.1002/gps.3865 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Kobayashi-Cuya KE, Sakurai R, Sakuma N, et al. Hand dexterity, not handgrip strength, is associated with executive function in Japanese community-dwelling older adults: a cross-sectional study. BMC Geriatr. 2018;18(1):192. doi: 10.1186/s12877-018-0880-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material
NIHMS2092621-supplement-Supplementary_Material.docx (521.4KB, docx)

Data Availability Statement

Some of the data that support the findings of this study are available in the Inter-university Consortium for Political and Social Research at doi: 10.3886/ICPSR34315.v2, reference number 34315. These data were derived from the following resources available in the public domain: https://www.icpsr.umich.edu/web/NACDA/studies/34315. The remaining data that support the findings of this study are available from the corresponding author upon reasonable request.

RESOURCES