Abstract
Background
Slow gait speed has been consistently associated with an increased risk of dementia. This study examined whether measures of balance and lower limb strength are similarly related to the risk of developing dementia.
Methods
Participants from the Health and Retirement Study (HRS, N = 5 658, mean age = 73.23, standard deviation [SD] = 6.22) and the English Longitudinal Study of Ageing (ELSA, N = 3667, mean age = 69.90, SD = 7.02) completed measures of gait speed, semi-tandem balance, chair stand (ELSA only), and cognitive status at baseline. Cognitive status was assessed over up to 15 years.
Results
Baseline slower gait speed (hazard ratio [HR]HRS = 1.52, 95% confidence interval [CI] = 1.32–1.75, p < .001; HRELSA = 1.73, 95% CI = 1.37–2.18, p < .001); and balance impairment (HRHRS = 1.58, 95% CI = 1.26–1.96, p < .001; HRELSA = 1.97, 95% CI = 1.24–3.14, p < .01) were related to a higher risk of incident dementia, adjusting for demographic factors. The combination of slower gait and impaired balance was associated with a two-to-three times higher risk of dementia in HRS and ELSA. Worse performance on the chair stand at baseline was associated with a higher risk of dementia in ELSA (HR = 1.56, 95% CI = 1.23–1.99, p < .001). All performance measures remained significant when entered simultaneously and accounted for obesity, diabetes, blood pressure, physical activity, smoking, and depressive symptoms. There was little evidence that age, sex, or APOE ε4 moderated the association.
Conclusions
Similar to gait speed, measures of balance and strength are associated with a higher risk of incident dementia. The findings have implications for clinical practice, given that these routinely used geriatric assessment tools are similarly related to dementia risk.
Keywords: Balance, Dementia, Gait speed, Strength
Performance-based measures of physical function, particularly lower extremity function, are useful screening tools for detecting individuals at risk of dementia (1). Indeed, systematic reviews and meta-analyses provide consistent evidence that slower gait speed is associated with an increased risk of incident dementia (1–4). Gait speed is relatively easy to measure and predicts various health-related outcomes, including mortality (5). Gait speed figures prominently in more complex tasks, such as dual-task gait speed, which is also related to the risk of dementia (6). Furthermore, a growing literature reports that the risk of dementia is heightened when slow gait co-occurs with poor subjective cognition (7). However, comparatively fewer studies have tested other measures of lower-extremity physical function, such as standing balance and chair stand.
Despite some evidence for the association between better standing balance and lower risk of dementia (8–12), existing recommendations indicate that the evidence is not strong enough to recommend balance as a marker of future incident dementia, calling for further research on this association (3). Furthermore, most research on standing balance and incident dementia is limited by relatively small samples and short longitudinal follow-ups. In addition, the link between performance on chair stand and incident dementia has received little attention in existing literature. The few existing studies report mixed associations between performance on the five-times sit-to-stand test (FTSST) and incident dementia (9,12): 1 study found that better performance on this task was associated with a lower risk of incident dementia (12), whereas it was not significant in another study (9). More research is needed on the association between chair stand and incident dementia.
The present study examines the association between measures of lower extremity function and incident dementia over 15 years in 2 large longitudinal samples of older adults from the United States and England. The primary aim was to examine whether the typical association between gait speed and dementia risk could be observed with balance and lower limb strength measures. The study further assessed whether associations were independent or overlapping across these performance measures, and whether their combination was related to dementia risk. It was hypothesized that similar to slower gait speed, balance impairment, and slower chair stand performance would be related to an increased risk of incident dementia. This study further tested whether clinical (obesity, diabetes, blood pressure), behavioral (physical activity, smoking), and psychological (depressive symptoms) factors accounted for the association between lower extremity function and incident dementia. Exploratory analyses tested whether age, sex, and APOE ε4 moderated the link between gait speed, balance, chair stand, and incident dementia.
Method
Participants
The Health and Retirement Study (HRS) and the English Longitudinal Study of Ageing (ELSA) were used to test the association between lower extremity function and incident dementia. Balance and gait speed were available in both samples; chair stands were available only in ELSA. Informed consent was obtained from all participants in the 2 samples. The HRS was approved by the University of Michigan Institutional Review Board (IRB) and the National Research Ethics Service approved ELSA. The analyses in this article did not require a local IRB because HRS and ELSA are publicly available deidentified data sets.
The HRS is a longitudinal panel study of a nationally representative sample of Americans aged 50 and older and their spouses/partners. Data on gait speed and standing balance were assessed as part of an enhanced face-to-face interview among half the sample in 2006 and for the other half in 2008. Gait speed was available for individuals aged ≥65 years at baseline. Across the combined 2006 and 2008 baseline waves, 7 883 participants had complete data on gait speed, balance, demographic factors, and cognitive status. Data on cognitive status were collected every 2 years up to the 2020 wave. Excluding participants with cognitive impairment and dementia at baseline (N = 1 858), cognitive status over the follow-up was available for 5 665 participants. With outliers on gait speed excluded (N = 7), the final analyzed sample had 5 658 participants. HRS data are publicly available at https://hrs.isr.umich.edu/data-products.
ELSA is a longitudinal study of a nationally representative sample of people aged 50 years and older living in England. Data on gait speed were collected among individuals aged 60 years and older. Baseline data on gait speed, balance, chair stand, demographic factors, and cognitive status were obtained from 4 104 participants during Wave 2 (2004/2005). Cognitive status was assessed every 2 years, up to Wave 9 (2018/2019). Excluding participants with dementia at baseline (N = 3), cognitive status over the follow-up was available for 3 691 participants. Excluding outliers on gait speed (N = 24), the final sample analyzed had 3 667 participants. ELSA data are publicly available at https://www.ukdataservice.ac.uk/.
Measures
Gait speed
Participants were asked to walk twice at their normal pace on a distance of 2.5-m course in the HRS, and a 2.4-m course in ELSA. The best of the 2 trials was taken. Gait speed was calculated as the distance (in meters) divided by the time taken (in seconds). Participants with performances of 3 standard deviations (SD) above the mean were excluded (NHRS = 7; NELSA = 24). A speed of <0.80 m/s was defined as slow gait speed (13) and coded as 1 (vs 0 for not slow gait speed). Participants who were unable to complete the task were coded as 1.
Standing balance
In the HRS, 3 separate stances assessed standing balance. The semi-tandem balance task was conducted first by asking respondents to stand with the side of the heel of 1 foot touching the toe of the other foot. Participants were asked to do so for about 10 seconds. If they were unable to complete this task for 10 seconds, they were asked to perform the side-by-side tandem task with both feet together side-by-side for 10 seconds. If participants were able to perform the semi-tandem task for 10 seconds, then they were asked to complete the full-tandem balance task, with the heel of 1 foot in front of and touching the toes of the other foot. Participants aged 65 years and older were asked to perform a 30-second full tandem task and those younger than 65 years old were asked to perform a 60-second task. In ELSA, participants were also asked to complete 3 stances, starting with the side-by-side task. If they were able to complete this stand for 10 seconds, they were asked to perform a 10-second semi-tandem stand. If participants were able to complete the semi-tandem task for 10 seconds, they were asked to perform a 30-second full tandem stand for participants younger than 70 years or a 10-second full tandem stand for those aged 70 years and older. Balance impairment was defined as the inability to complete the semi-tandem stance (coded as 1 vs 0 for semi-tandem completed) (14).
Chair stand
Chair stand was assessed using the FTSST and only in ELSA. Participants were asked to stand up as fast as possible from a chair to a full standing position 5 times without using their arms. The time taken to complete the 5 repetitions was recorded in seconds. Faster time indicates better performance. A time >15 seconds is indicative of worse lower limb strength and probable sarcopenia (15) and was coded as 1; ≤15 seconds was coded as 0. Participants who completed less than the requested rises, who tried but were unable to complete the task, or who could not hold the position unassisted were also coded as 1.
Dementia
The modified Telephone Interview for Cognitive Status (TICSm) (16) was used to assess cognitive status in the HRS. This measure includes immediate and delayed recall of 10 words (ranging from 0 to 20 points), serial 7 subtraction (0–5 points), and backward counting (0–2 points). The sum of 3 tasks was computed to obtain a 27-point TICSm score, with dementia defined by scores ≤6 (16). Self-reports of physician diagnosis of Alzheimer’s disease or dementia and scores on the 16-item version of the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE (17)) were used to classify dementia in ELSA (18,19). In the IQCODE, informants rated participants’ current functioning compared with 10 years ago on a scale from 1 (much improved) to 5 (much worse). The mean of the 16 items was taken, with higher means indicative of a steeper informant-rated decline in functioning. Based on existing research (18,19), the mean threshold higher than 3.38 indicated dementia.
Covariates
Demographic factors were included as covariates in all analyses: age (in years), sex (1 for female, 0 for male), education, and race. Education was reported in years in the HRS and using a scale ranging from 1 (no qualification) to 7 (NVQ4/NVQ5/Degree or equivalent) in ELSA. NVQ4 and NVQ5 are equivalent to higher education certificate and diploma, respectively. Race was coded as 1 (African American/Black) and 0 (not African American/Black) in the HRS, and as 1 (non-White) and 0 (White) in ELSA. Ethnicity was also included in the HRS (1 = Hispanic or Latinx and 0 = non-Hispanic/Latinx). Assessment wave was controlled for in the HRS (1 for 2006, 0 for 2008). Additional analyses adjusted for body mass index (BMI), physical activity, smoking, depressive symptoms, diabetes, and high blood pressure. In both HRS and ELSA, staff-assessed weight and height were used to compute BMI in kg/m2. BMI ≥30 was indicative of obesity (1 for obesity, 0 for no obesity). Participants reported whether they ever had a medical diagnosis of diabetes and high blood pressure (1 for yes, 0 for no). The mean of 2 items on the frequency of vigorous and moderate physical activity was taken as a measure of physical activity. Smoking status was coded as 1 (for current/former smoker) and 0 (for never smoker). The 8-item version of the Center for Epidemiological Studies Depression Scale assessed depressive symptoms (20). Eight symptoms experienced (yes/no) for much of the past week were summed, and a score of 3 or higher was used to classify depression (1 for presence of depression, 0 for no depression). In the HRS, APOE status was coded as 1 = APOE ε4 carrier and 0 = noncarrier.
Data Analysis
The association between lower extremity function and incident dementia was tested using Cox proportional hazard models. Participants with dementia at baseline were excluded from these analyses. Time-to-incidence was coded in years from the baseline assessment to the first instance of dementia. Participants without incident dementia were censored at the last available assessment. In the primary analyses, gait speed, balance impairment, and chair stand were entered separately, controlling for age, sex, education, and race. The analyses of the HRS further included ethnicity and wave of assessment. A fully adjusted model included body mass index (BMI), diabetes, high blood pressure, physical activity, smoking, and depressive symptoms as additional covariates. Results for gait speed and balance impairment for the 2 samples were combined in a random-effect meta-analysis conducted with the comprehensive meta-analysis software.
All performance measures were then entered simultaneously in a single model to test whether the association of any marker was independent of the other. We also tested the association between a combined impairment index and incident dementia, in which participants with both slower gait speed and balance impairment were contrasted to those without any impairment. In ELSA, we further contrasted participants with slower gait speed, impaired balance, and poor chair stand performance to participants without impairment on any of the 3 measures. Sensitivity analyses excluded participants who developed dementia within the first 5 years of baseline. A second sensitivity analysis examined the associations using continuous measures of gait speed and balance. Moderation by age and sex was examined by testing the interaction between each performance measure and age or sex on dementia risk. The interaction between APOE status and both gait speed and balance impairment was also tested in HRS.
Results
The descriptive statistics for the 2 samples are in Table 1. The proportion of participants who developed dementia was 16% (N = 917) in HRS and 9% (N = 326) in ELSA over the 15 years follow-up (HRS: median follow-up = 10.17 years, 53 017 person-years; ELSA: median follow-up = 11.75 years, 36 174 person-years). In line with the predictions, slower gait speed at baseline was associated with an increased risk of incident dementia in both samples (Model 1, Table 2) and the meta-analysis (Model 1, Table 4). Participants with slower gait speed had 52% (HRS) and 73% (ELSA) higher risk of incident dementia. Also, consistent with expectations, baseline balance impairment was related to 58% (HRS) and 97% (ELSA) higher risk of incident dementia (Model 1, Table 3) and the meta-analysis (Model 1, Table 4). Worse chair stand performance at baseline was associated with a 56% increased risk of incident dementia in ELSA (Model 1, Table 5). Compared with participants without impairment in gait and balance, those with dual impairment had more than 2 times higher dementia risk in HRS (hazards ratio [HR] = 2.11, 95% confidence interval [CI] = 1.58–2.81, p < .001) and almost 3½ times higher dementia risk in ELSA (HR = 3.45, 95% CI = 2.01–5.94, p < .001; see Supplementary Table 1). In ELSA, participants with impairment in all 3 tasks had a 3 times higher risk of dementia than those without any impairment (HR = 3.04, 95% CI = 1.51–6.13, p = .002; Supplementary Table 1).
Table 1.
Descriptive Statistics for the 2 Samples
| HRS | ELSA | |||
|---|---|---|---|---|
| Variables | M/% (n) | SD | M/% (n) | SD |
| Age (years) | 73.23 | 6.22 | 69.90 | 7.02 |
| Sex (% women) | 58% (3 255) | — | 55% (2 006) | — |
| Race (% African American/Black) | 8% (449) | — | 2%* (56) | — |
| Ethnicity (% Hispanic) | 6% (318) | — | — | — |
| Education | 12.99 | 2.69 | 3.25 | 2.23 |
| Obesity (% yes)† | 36% (1 998) | — | 26% (955) | — |
| Diabetes (% yes)† | 20% (1 110) | — | 8% (278) | — |
| High blood pressure (% yes)† | 62% (3 526) | — | 46% (1 684) | — |
| Physical activity† | 2.45 | 1.06 | 2.63 | 0.89 |
| Depressive symptoms (% yes)† | 14% (805) | — | 18% (645) | — |
| Current smoking (% yes)b | 51% (2 870) | — | 62% (2 280) | — |
| Slow gait speed (% yes) | 42% (2 382) | — | 29% (1 066) | — |
| Balance impairment (% yes) | 6% (365) | — | 3% (101) | — |
| Poor chair stand (% yes) | — | — | 22% (790) | — |
Notes: HRS: N = 5 658; ELSA: N = 3 667. ELSA = English Longitudinal Study of Ageing; HRS = Health and Retirement Study; SD = standard deviation.
*Percent of non-White participants.
† N’s differ because of missing data on the variables.
Table 2.
Relationship Between Slow Gait Speed and Incident Dementia in HRS and ELSA
| HRS | ELSA | |||
|---|---|---|---|---|
| Predictor | Model 1 | Model 2 | Model 1 | Model 2 |
| Hazard Ratios (95% CI) | Hazard Ratios (95% CI) | Hazard Ratios (95% CI) | Hazard Ratios (95% CI) | |
| Age | 2.04 (1.91–2.18)*** | 2.07 (1.93–2.23)*** | 2.42 (2.15–2.72)*** | 2.35 (2.08–2.65)*** |
| Sex | 0.95 (0.83–1.09) | 0.98 (0.84–1.13) | 0.91 (0.72–1.14) | 0.88 (0.69–1.11) |
| Race | 1.85 (1.51–2.26)*** | 1.81 (1.46–2.23)*** | 2.00 (0.98–4.09) | 2.03 (0.99–4.16) |
| Ethnicity | 1.15 (0.88–1.50) | 1.11 (0.84–1.45) | — | — |
| Education | 0.77 (0.72–0.82)*** | 0.79 (0.74–0.85)*** | 0.90 (0.80–1.01) | 0.91 (0.81–1.03) |
| Gait speed | 1.52 (1.32–1.75)*** | 1.42 (1.23–1.64)*** | 1.73 (1.37–2.18)*** | 1.55 (1.21–1.97)*** |
| Diabetes | 1.37 (1.16–1.62)*** | 1.37 (0.93–2.03) | ||
| Blood pressure | 0.98 (0.85–1.12) | 0.91 (0.73–1.14) | ||
| Current smoking | 1.22 (1.05–1.41)** | 0.97 (0.77–1.23) | ||
| Obesity | 0.83 (0.72–0.96)* | 0.88 (0.68–1.14) | ||
| Physical activity | 0.90 (0.83–0.96)** | 0.82 (0.73–0.92)** | ||
| Depressive symptoms | 1.41 (1.19–1.67)*** | 1.18 (0.90–1.55) | ||
Notes: CI = confidence interval; ELSA = English Longitudinal Study of Ageing; HRS = Health and Retirement Study.
* p < .05. **p < .01. ***p < .001.
Table 4.
Meta-Analytic Summary of the Relationship Between Gait Speed and Balance Impairment and Incident Dementia
| Meta-analysis | ||||
|---|---|---|---|---|
| Predictor | Model 1† | Model 2‡ | ||
| Pooled Hazard Ratios (95% CI) | Heterogeneity I2 | Pooled Hazard Ratios (95% CI) | Heterogeneity I2 | |
| Gait speed | 1.57(1.40–1.78)*** | 0 | 1.45(1.28–1.64)*** | 0 |
| Balance impairment | 1.65 (1.35–2.01)*** | 0 | 1.48 (1.21–1.82)*** | 0 |
Notes: CI = confidence interval; ELSA = English Longitudinal Study of Ageing; HRS = Health and Retirement Study.
†Model 1 combined the hazard ratios from the Model 1 in HRS and ELSA.
‡Model 2 combined the hazard ratios from the Model 2 in HRS and ELSA.
*** p < .001.
Table 3.
Relationship Between Balance Impairment and Incident Dementia in HRS and ELSA
| HRS | ELSA | |||
|---|---|---|---|---|
| Predictor | Model 1 | Model 2 | Model 1 | Model 2 |
| Hazard Ratios (95% CI) | Hazard Ratios (95% CI) | Hazard Ratios (95% CI) | Hazard Ratios (95% CI) | |
| Age | 2.07 (1.94–2.21)*** | 2.09 (1.94–2.25)*** | 2.52 (2.24–2.83)*** | 2.42 (2.15–2.73)*** |
| Sex | 0.98 (0.86–1.12) | 1.00 (0.86–1.15) | 0.93 (0.74–1.17) | 0.89 (0.70–1.13) |
| Race | 2.02 (1.65–2.46)*** | 1.93 (1.57–2.38)*** | 2.31 (1.14–4.69)* | 2.18 (1.07–4.46)* |
| Ethnicity | 1.23 (0.94–1.60) | 1.16 (0.89–1.53) | — | — |
| Education | 0.75 (0.71–0.80)*** | 0.78 (0.73–0.83)*** | 0.87 (0.77–0.98)* | 0.90 (0.80–1.02) |
| Balance impairment | 1.58 (1.26–1.96)*** | 1.44 (1.15–1.81)** | 1.97 (1.24–3.14)** | 1.70 (1.04–2.79)* |
| Diabetes | 1.36 (1.15–1.61)*** | 1.44 (0.98–2.12) | ||
| Blood pressure | 0.98 (0.85–1.13) | 0.91 (0.73–1.14) | ||
| Current smoking | 1.21 (1.04–1.40)* | 0.99 (0.78–1.25) | ||
| Obesity | 0.85 (0.74–0.99)* | 0.90 (0.69–1.16) | ||
| Physical activity | 0.89 (0.83–0.95)** | 0.79 (0.70–0.89)*** | ||
| Depressive symptoms | 1.44 (1.21–1.71)*** | 1.25 (0.96–1.63) | ||
Notes: CI = confidence interval; ELSA = English Longitudinal Study of Ageing; HRS = Health and Retirement Study.
* p < .05. **p < .01. ***p < .001.
Table 5.
Relationship Between Chair Stands and Incident Dementia in ELSA
| Predictor | ELSA | |
|---|---|---|
| Model 1 | Model 2 | |
| Hazard Ratios (95% CI) | Hazard Ratios (95% CI) | |
| Age | 2.47 (2.19–2.77)*** | 2.39 (2.12–2.69)*** |
| Sex | 0.91 (0.73–1.14) | 0.88 (0.69–1.12) |
| Race | 2.19 (1.08–4.46)* | 2.13 (1.04–4.35)* |
| Education | 0.87 (0.77–0.98)* | 0.90 (0.79–1.01) |
| Chair stand | 1.56 (1.23–1.99)*** | 1.41 (1.11–1.80)*** |
| Diabetes | 1.44 (0.98–2.12) | |
| Blood pressure | 0.92 (0.73–1.15) | |
| Current smoking | 0.99 (0.79–1.25) | |
| Obesity | 0.91 (0.70–1.17) | |
| Physical activity | 0.81 (0.72–0.92)** | |
| Depressive symptoms | 1.23 (0.94–1.61) | |
Notes: CI = confidence interval; ELSA = English Longitudinal Study of Ageing.
* p < .05. **p < .01. ***p < .001.
Additional analyses indicated that clinical, behavioral, and psychological factors partially accounted for the association between lower extremity function and incident dementia in both samples and the meta-analysis (Model 2, Tables 2–5). The association between gait speed and dementia persisted but was reduced by 19% ([(HR Model 1—HR Model 2)/(HR Model 1 − 1)] × 100) in HRS and 25% in ELSA (Model 2, Table 2). The association between balance impairment and incident dementia was attenuated by 24% in the HRS and 28% in ELSA but remained significant (Model 2, Table 3). These results were confirmed by the meta-analysis (Model 2, Table 4). In ELSA, the association between chair stand and incident dementia was reduced by 27% but persisted when clinical, behavioral, and psychological factors were included in the analysis (Model 2, Table 5).
In analyses with measures of lower extremity function examined simultaneously, gait speed (HR = 1.49, 95% CI = 1.30–1.71, p < .001) and balance impairment (HR = 1.48, 95% CI = 1.19–1.85, p = .001) remained significant predictors of incident dementia in the HRS (Supplementary Table 2). Similarly, in ELSA, gait speed (HR = 1.60, 95% CI = 1.26–2.03, p < .001), balance impairment (HR = 1.70, 95% CI = 1.07–2.72, p = .025), and poor chair stand performance (HR = 1.40, 95% CI = 1.09–1.78, p = .008) were all independent predictors of incident dementia (Supplementary Table 2).
The overall pattern of association was similar when cases occurring in the first 5 years were excluded (Supplementary Table 3). When entered as a continuous measure, faster gait speed was related to lower risk of incident dementia in both samples (HRHRS = 0.77, 95% CI = 0.71–0.83, p < .001; HRELSA = 0.70, 95% CI = 0.62–0.79, p < .001) (Supplementary Table 4). The continuous measure of chair stands was also related to higher dementia risk in ELSA (HRELSA = 1.12, 95% CI = 1.04–1.21, p = .003; Supplementary Table 4). Finally, there was little evidence of moderation by age, with 1 exception. The association between balance impairment and an increased risk of incident dementia was stronger among relatively younger participants in the HRS (HRinteraction = 0.72, 95% CI = 0.60–0.86, p < .001). There was no moderation by sex in either sample or APOE status in the HRS.
Discussion
The present study examined the association between measures of lower extremity function and incident dementia over 15 years in 2 longitudinal samples of older adults. As predicted, impaired balance and lower chair stand performance were associated with a higher risk of incident dementia, similar to slower gait speed. Furthermore, the combination of slower gait speed and balance impairment was associated with an up to 3½ higher risk of incident dementia. The associations were replicated across 2 samples from different cultures and using different assessment methods. The associations were robust because they persisted when demographic, clinical, behavioral, and psychological factors were controlled for and were generalizable across age, sex, and genetic groups.
The association between slow gait speed and higher risk of incident dementia is consistent with existing evidence (1–4). The effect sizes observed in this study are similar to estimates from meta-analyses (relative risk = 1.66, 4; HR = 1.53) (21). This study extends existing knowledge by identifying this association over a longer follow-up (15 years) than in most research (2). This study further revealed that beyond gait speed, other measures of lower extremity function, such as standing balance and chair stand, were associated with risk of incident dementia. Past research reported weak evidence for the predictive role of balance, and thus caution was recommended about using balance assessment as a marker of future dementia risk (3). The present study identifies a replicable and robust association between balance impairment and a higher risk of incident dementia. Of note, the effect size for the association between balance impairment and incident dementia was similar or even stronger than for gait speed. The significant association between poor chair stand performance, which reflects weaker lower limb strength and probable sarcopenia, and higher risk of incident dementia extends a limited body of research that has reported mixed findings (9,12). This finding is also consistent with evidence that grip strength is associated with dementia risk (22), which suggests that the association between muscle strength and incident dementia may generalize across measures. Taken as a whole, the present study contributes to existing knowledge on the motor markers of the risk of dementia (1,3), by providing robust evidence of the predictive, and independent, role of different measures of lower extremity function and incident dementia.
Several potential mechanisms have been proposed to explain the association between lower extremity function and incident dementia. One conceptual model suggests that gait speed, standing balance, and chair stand could be early signs of neurodegenerative processes that lead to dementia (2,7). For example, β-amyloid (Aβ) burden has been associated with impaired lower extremity functions among cognitively normal individuals (23,24). Besides Aβ, poor performance on these measures may also reflect general pathological processes, including cortical and peripheral physiological and structural changes. However, when cases within the first 5 years were excluded, the results were similar, and thus contrary to reverse causality. A second interpretation is that performance measures may be indirectly related to dementia risk because they share common clinical, behavioral, and psychological risk factors (2,7). Consistent with this hypothesis, the present study found that obesity, diabetes, blood pressure, physical activity, smoking, and depressive symptoms partially accounted for the associations. These pathways, however, did not account for all the associations. A third conceptual model suggests that lower extremity function may play a causal role in a chain of processes leading to dementia. For example, slower gait, impaired balance, and weaker leg strength may limit involvement in physical, cognitive, and social activities, ultimately resulting in a higher risk of dementia. These interpretations are not mutually exclusive, and each one may contribute to the observed associations.
The present study has several strengths, including the 2 large longitudinal samples of older adults, the use of 3 measures of lower extremity function, a meta-analysis, the 15-year longitudinal follow-up, and the inclusion of demographic, clinical, behavioral, and psychological factors. There are several limitations to consider when interpreting the results. First, dementia was not based on a clinical assessment, which may have increased noise in our findings. Second, data on chair stand were only available in ELSA, and thus the significant association still needs to be replicated in another sample. Third, the present study tested baseline assessments of lower extremity function; future research could test whether changes in gait speed, standing balance, and chair stand are related to risk of incident dementia. Finally, the samples were from high-income countries; future research should examine these associations in samples from countries with fewer resources to evaluate the generalizability of the observed associations.
In sum, the present study found that measures of lower extremity function are related to incident dementia. Slower gait speed, balance impairment, and worse lower limb strength were associated with an increased risk of incident dementia over 15 years. Measures of lower extremity function are useful markers for risk of developing dementia and may help identify individuals with greater needs for preventative interventions.
Supplementary Material
Contributor Information
Yannick Stephan, Euromov, University of Montpellier, Montpellier, France.
Angelina R Sutin, Department of Behavioral Sciences and Social Medicine, College of Medicine, Florida State University, Tallahassee, Florida, USA.
Martina Luchetti, Department of Behavioral Sciences and Social Medicine, College of Medicine, Florida State University, Tallahassee, Florida, USA.
Damaris Aschwanden, Center for the Interdisciplinary Study of Gerontology and Vulnerability, University of Geneva, Geneva, Switzerland; Department of Geriatrics, College of Medicine, Florida State University, Tallahassee, Florida, USA.
Selin Karakose, Department of Geriatrics, College of Medicine, Florida State University, Tallahassee, Florida, USA.
Antonio Terracciano, Department of Geriatrics, College of Medicine, Florida State University, Tallahassee, Florida, USA.
Lewis A Lipsitz, (Medical Sciences Section).
Funding
The research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health (grant numbers R01AG068093, R01AG053297). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The Health and Retirement Study (HRS) is sponsored by the National Institute on Aging (NIA-U01AG009740) and conducted by the University of Michigan. Funding for the English Longitudinal Study of Ageing is provided by the National Institute of Aging (grants 2RO1AG7644-01A1 and 2RO1AG017644) and a consortium of UK government departments coordinated by the Office for National Statistics.
Conflict of Interest
None.
Data Availability
HRS data is publicly available at https://hrs.isr.umich.edu/data-products. ELSA data are available from the UK Data Service (UKDS, https://www.ukdataservice.ac.uk/).
Author Contributions
Y.S.: Conceptualization, methodology, formal analysis, writing—original draft, writing—review & editing, visualization, supervision, project administration. A.R.S.: Conceptualization, methodology, writing—review & editing, visualization. M.L.: Conceptualization, methodology, writing—review & editing. D.A.: Methodology, writing—review & editing, visualization. S.K.: Writing—review & editing, visualization. A.T.: Conceptualization, methodology, formal analysis, writing—original draft, writing—review & editing, visualization, supervision
References
- 1. Kueper JK, Speechley M, Lingum NR, Montero-Odasso M.. Motor function and incident dementia: a systematic review and meta-analysis. Age Ageing. 2017;46(5):729–738. 10.1093/ageing/afx084 [DOI] [PubMed] [Google Scholar]
- 2. Grande G, Triolo F, Nuara A, Welmer AK, Fratiglioni L, Vetrano DL.. Measuring gait speed to better identify prodromal dementia. Exp Gerontol. 2019;124:110625. 10.1016/j.exger.2019.05.014 [DOI] [PubMed] [Google Scholar]
- 3. Montero-Odasso M, Pieruccini-Faria F, Ismail Z, et al. CCCDTD5 recommendations on early non cognitive markers of dementia: a Canadian consensus. Alzheimers Dement (N Y). 2020;6(1):e12068. 10.1002/trc2.12068 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Quan M, Xun P, Chen C, et al. Walking pace and the risk of cognitive decline and dementia in elderly populations: a meta-analysis of prospective cohort studies. J Gerontol A Biol Sci Med Sci. 2017;72(2):266–270. 10.1093/gerona/glw121 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Dommershuijsen LJ, Isik BM, Darweesh SKL, van der Geest JN, Ikram MK, Ikram MA.. Unraveling the association between gait and mortality-one step at a time. J Gerontol A Biol Sci Med Sci. 2020;75(6):1184–1190. 10.1093/gerona/glz282 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Montero-Odasso MM, Sarquis-Adamson Y, Speechley M, et al. Association of dual-task gait with incident dementia in mild cognitive impairment: results from the Gait and Brain Study. JAMA Neurol. 2017;74(7):857–865. 10.1001/jamaneurol.2017.0643 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Semba RD, Tian Q, Carlson MC, Xue QL, Ferrucci L.. Motoric cognitive risk syndrome: integration of two early harbingers of dementia in older adults. Ageing Res Rev. 2020;58:101022. 10.1016/j.arr.2020.101022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Ahn S, Chung JW, Crouter SE, Lee JA, Lee CE, Anderson JG.. Gait and/or balance disturbances associated with Alzheimer’s dementia among older adults with amnestic mild cognitive impairment: a longitudinal observational study. J Adv Nurs. 2023;79(12):4815–4827. 10.1111/jan.15768 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Bullain SS, Corrada MM, Shah BA, Mozaffar FH, Panzenboeck M, Kawas CH.. Poor physical performance and dementia in the oldest old: the 90+ study. JAMA Neurol. 2013;70(1):107–113. 10.1001/jamaneurol.2013.583 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Kim HJ, Jeong S, Oh YH, Suh MJ.. Association of balance impairment with risk of incident dementia among older adults. J Prev Alzheimers Dis. 2024;11(1):130–137. 10.14283/jpad.2023.79 [DOI] [PubMed] [Google Scholar]
- 11. Lee AT, Richards M, Chan WC, Chiu HF, Lee RS, Lam LC.. Poor balance as a noncognitive predictor of incident dementia. J Am Geriatr Soc. 2015;63(8):1701–1702. 10.1111/jgs.13587 [DOI] [PubMed] [Google Scholar]
- 12. 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):1115–1120. 10.1001/archinte.166.10.1115 [DOI] [PubMed] [Google Scholar]
- 13. Studenski S, Perera S, Patel K, et al. Gait speed and survival in older adults. J Am Med Assoc. 2011;305(1):50–58. 10.1001/jama.2010.1923 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Tu R, Wang S, He H, et al. Association between subjective cognitive complaints, balance impairment and disability among middle-aged and older adults: evidence from a population-based cohort study. Geriatr Gerontol Int. 2022;22(12):1025–1031. 10.1111/ggi.14501 [DOI] [PubMed] [Google Scholar]
- 15. Cruz-Jentoft AJ, Bahat G, Bauer J, et al. ; Writing Group for the European Working Group on Sarcopenia in Older People 2 (EWGSOP2), and the Extended Group for EWGSOP2. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing. 2019;48(1):16–31. 10.1093/ageing/afy169 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Crimmins EM, Kim JK, Langa KM, Weir DR.. Assessment of cognition using surveys and neuropsychological assessment: the Health and Retirement Study and the Aging, Demographics, and Memory Study. J Gerontol B Psychol Sci Soc Sci. 2011;66(Suppl 1):i162–i171. 10.1093/geronb/gbr048 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Jorm AF. A short form of the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE): development and cross-validation [published correction appears in Psychol Med 1995 Mar;25(2):437]. Psychol Med. 1994;24(1):145–153. 10.1017/s003329170002691x [DOI] [PubMed] [Google Scholar]
- 18. Almeida-Meza P, Steptoe A, Cadar D.. Markers of cognitive reserve and dementia incidence in the English Longitudinal Study of Ageing. Br J Psychiatry. 2021;218(5):243–251. 10.1192/bjp.2020.54 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Cadar D, Lassale C, Davies H, Llewellyn DJ, Batty GD, Steptoe A.. Individual and area-based socioeconomic factors associated with dementia incidence in England: evidence from a 12-year follow-up in the English Longitudinal Study of Ageing. JAMA Psychiatry. 2018;75(7):723–732. 10.1001/jamapsychiatry.2018.1012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Wallace R, Herzog AR, Ofstedal MB, Steffick D, Fonda S, Langa K.. Documentation of Affective Functioning Measures in the Health and Retirement Study. Survey Research Center, University of Michigan; 2000. [Google Scholar]
- 21. Beauchet O, Annweiler C, Callisaya ML, et al. Poor gait performance and prediction of dementia: results from a meta-analysis. J Am Med Dir Assoc. 2016;17(6):482–490. 10.1016/j.jamda.2015.12.092 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Cui M, Zhang S, Liu Y, Gang X, Wang G.. Grip strength and the risk of cognitive decline and dementia: a systematic review and meta-analysis of longitudinal cohort studies. Front Aging Neurosci. 2021;13:625551. 10.3389/fnagi.2021.625551 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Dougherty RJ, Ramachandran J, Liu F, et al. Association of walking energetics with amyloid beta status: findings from the Baltimore Longitudinal Study of Aging. Alzheimers Dement (Amst). 2021;13(1):e12228. 10.1002/dad2.12228 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Tian Q, Resnick SM, Bilgel M, Wong DF, Ferrucci L, Studenski SA.. β-Amyloid burden predicts lower extremity performance decline in cognitively unimpaired older adults. J Gerontol A Biol Sci Med Sci. 2017;72(5):716–723. 10.1093/gerona/glw183 [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
Data Availability Statement
HRS data is publicly available at https://hrs.isr.umich.edu/data-products. ELSA data are available from the UK Data Service (UKDS, https://www.ukdataservice.ac.uk/).