Sensorimotor Impairment and Incident Dementia in the US Medicare Beneficiaries

ABSTRACT

Background

Sensorimotor impairments, including sensory (vision and hearing) and motor (standing balance, gait, chair stands, grip strength) deficits, are prevalent in older adults and are associated with an increased risk of dementia. Understanding the individual and combined effects of these impairments on dementia risk may enhance early detection and preventive strategies. We examined the association between baseline sensory difficulties and motor impairments, individually and collectively, with dementia over up to 11 years.

Methods

We used data from the National Health and Aging Trends Study (NHATS), a nationally representative study of US community‐dwelling Medicare beneficiaries aged ≥ 65 years. Participants cognitively unimpaired at baseline in 2011 were followed annually through 2022. Baseline exposures included self‐reported sensory difficulties (vision, hearing) and objectively measured motor impairments (standing balance, gait speed, chair stand, grip strength). The primary outcome was incident dementia, assessed annually based on NHATS consensus procedures. Survey‐weighted Cox proportional hazards models assessed the association between each impairment and number of impairments and incident dementia, adjusting for demographics and clinical conditions.

Results

The analytic sample included 3847 participants (weighted average age 73.9 (95% CI, 73.7–74.1) years, 56% women). Over an average follow‐up of 5.2 years, vision difficulty (HR = 1.34; 95% CI, 1.13–1.60), standing balance impairment (HR = 1.23; 95% CI, 1.10–1.39), slow gait speed (HR = 1.49; 95% CI, 1.34–1.67), impaired chair stand performance (HR = 1.31; 95% CI, 1.18–1.47), and weak grip strength (HR = 1.31; 95% CI, 1.13–1.51) were each independently associated with higher dementia risk. Compared to no impairments, dementia risk increased significantly with three or more sensorimotor impairments (HR = 1.44 to 1.96).

Conclusions

In this nationally representative cohort, vision difficulty and motor impairments were independently associated with increased dementia risk over up to 11 years. The presence of multiple sensory difficulties and motor impairments substantially increased the risk of dementia, emphasizing the importance of their early detection and management to reduce dementia risk.

Summary

• Key points

  • ○
    Vision difficulty and individual motor impairments—including standing balance, gait speed, chair stands, and grip strength—were each independently associated with higher dementia risk over up to 11 years.
  • ○
    Dementia risk increased significantly with three or more sensorimotor impairments, with slow gait speed suggesting the highest risk.
• Why does this paper matter?

  • ○
    In this nationally representative cohort of older US adults, vision and motor impairments were linked to higher dementia risk, with greater risk for a higher number of impairments. Findings support incorporating simple vision and motor assessments into routine care to identify high‐risk individuals and inform prevention strategies.

1. Introduction

Sensory and motor (“sensorimotor”) impairments are increasingly prevalent at older ages [1, 2, 3]. A growing body of research suggests these impairments, including poor vision, hearing, walking, strength, and balance, are associated with low and declining cognitive performance, making them potential markers of adverse changes in cognitive health [4, 5, 6, 7, 8, 9]. Most current evidence has concentrated on the association of either sensory or motor performance with cognitive decline and dementia, and the collective presence of sensorimotor impairments has generally not been considered. Importantly, sensory difficulties and motor impairments are often modifiable through preventative mechanisms (e.g., physical therapy) or accommodations (e.g., hearing aids, cataract surgery), and may offer targets for interventions aimed at reducing dementia risk [10, 11, 12, 13]. With the aging of the population, a better understanding of the magnitude of the association of sensorimotor impairments with dementia risk is needed to identify and refine potential preventative strategies.

The connection between declining or impaired sensory and/or motor function and dementia is often multifactorial and not well understood. Sensory loss increases cognitive burden for understanding speed and processing auditory and visual information [14, 15]. In addition, both hearing and vision difficulty can limit engagement in social activities, leading to social isolation and depression, potentially contributing to brain atrophy and cognitive decline [16, 17, 18]. Impaired motor function may reduce physical activity levels, resulting in a more sedentary lifestyle with limited social participation, all of which increase dementia risk [19]. Further, both sensory difficulty and motor function impairments may be indicators of structural brain changes or accumulation of dementia pathology in brain regions responsible for sensory function and/or motor coordination [19, 20, 21, 22]. To this end, sensorimotor deficits may exacerbate risk factors for poor cognitive health (e.g., social isolation, low physical activity) and/or serve as markers of underlying neuropathology, contributing to cognitive decline and dementia [23]. Although there are many studies linking sensory function with dementia and motor function with dementia, to date, there are no studies examining the contribution of sensorimotor impairments to incident dementia in a large representative population of older adults.

To address this research gap, we examined the association between baseline sensory difficulty and motor impairments and incident dementia over up to 11 years of follow‐up using data from the National Health and Aging Trends Study (NHATS), a nationally representative panel study of US Medicare beneficiaries aged 65 years and older. We hypothesized that older adults with sensory difficulty and/or motor function impairments are more likely to develop dementia and that a higher number of sensorimotor impairments is associated with a higher risk of dementia. A better understanding of the associations between sensorimotor impairments and incident dementia over a long follow‐up period may further refine dementia risk factors and assist with the development of targeted interventions.

2. Methods

2.1. Study Participants

Participants were recruited into NHATS in 2011 (Round 1, study initiation/baseline). The current study uses 11 years of longitudinal data collected through 2022 (Round 12). At baseline, 8245 participants were sampled, and annual interview data were collected by direct contact with either the participant or, for participants unable to answer for themselves, a proxy interview was conducted. We included community‐living participants and those residing in assisted living (non‐nursing home) facilities who had available data on dementia status, sensory difficulty, and motor function, and relevant covariates [24]. The study was approved by the Johns Hopkins Bloomberg School of Public Health Institutional Review Board, and all participants provided written informed consent.

2.2. Measures

2.2.1. Main Outcome

2.2.1.1. Dementia

Dementia outcome determination was performed using the NHATS dementia classification algorithm [25]. At each round, individuals were classified as having probable dementia if they met any of the following criteria: (1) a doctor had ever reportedly told the participant he/she had dementia or Alzheimer's disease; (2) a proxy rating of ≥ 2 on the eight‐item ad8 screener; or (3) had a score ≥ 1.5 SDs below the mean in at least two cognitive domains based on test items that evaluate memory, orientation, and executive function. Participants were classified as having possible dementia if they scored ≥ 1.5 standard deviations below the mean in one out of three cognitive domains in the neuropsychological tests. All other participants were classified as not having dementia. In the current analysis, dementia was defined as either possible or probable dementia. The dementia outcome in round 10 was not included in the current analysis due to restrictions on in‐person data collection from the COVID‐19 pandemic.

2.2.2. Sensory Difficulty

Sensory function was assessed using interview questions specific to participants' ability to see and hear. Participants were classified as having vision difficulty if they self‐reported blindness or an inability to see well enough to recognize someone across the street or to read a newspaper, even when wearing glasses [26]. Participants were classified as having hearing difficulty if they self‐reported deafness, hearing aid use, inability to hear well enough to use the telephone or to carry on a conversation in a room with the radio or television on [26]. Consistent with prior research in NHATS, hearing aid users were included as having hearing difficulty.

2.2.3. Motor Impairment

Motor function was assessed using objective measures of balance, gait speed, chair stands, and grip strength. A more detailed summary of data collection procedures is provided in the NHATS user guide and technical papers [27]. Participants who did not complete a functional test because they were not eligible or did not attempt for safety reasons, or attempted but could not complete the test, were assigned a NHATS SPPB score of 0 for that test and were classified as impaired. Participants who did not perform the test and were not assigned a score of 0 based on these criteria were classified as missing.

2.2.3.1. Standing Balance

Participants were asked to attempt 5 standing balance activities—side‐by‐side, semi‐tandem, full tandem, one leg eyes open, and one leg eyes closed. The standing balance components were coded according to the NHATS Short Physical Performance Battery (SPPB) [28, 29], ranging from 0 to 4, with higher scores indicating better standing balance. Standing balance impairment was considered a score of 0 to 2, based on the data distribution.

2.2.3.2. Gait Speed

Gait speed was assessed using two usual‐paced tests over a 3‐m walking course. We used the average speed for the two tests, with higher values indicating faster gait speed. Consistent with normative data, impairment was defined as gait speed < 0.8 m/s [30].

2.2.3.3. Chair Stands

We used the speed to complete the chair stand test for five repeated chair stands (chair stands/s), with higher values indicating a longer time to complete. Consistent with normative data, impairment was defined as chair stand pace < 0.42 stands/s [31].

2.2.3.4. Grip Strength

We used the average of the two grip strength tests, with higher values indicating better grip strength performance. Consistent with normative data, impairment was defined as < 32 kg for men and < 20 kg for women [31].

2.2.4. Sensorimotor Impairment Score

A cumulative score was created that summed the total number of sensory difficulties (range 0–2) with the total number of motor impairments (range 0–4) to create a sensorimotor impairment score (range 0–6).

2.3. Covariates

Sociodemographic covariates included age, sex, race, and education level. Height and weight were self‐reported, and body mass index (BMI) was calculated as kilograms per meter squared. Clinical conditions included self‐reported history (yes/no) of heart attack, heart disease, lung disease, stroke, cancer, diabetes, depression, and hypertension. The sum of the number of clinical conditions was calculated to create a composite clinical conditions score (0, 1, 2, 3, ≥ 4). History of regular smoking was also included (yes/no).

2.4. Statistical Analysis

Participants with missing dementia diagnosis or a possible or probable dementia diagnosis at baseline were excluded from the analysis. To compare the participant characteristics between those who developed dementia during the follow‐up period and those who did not, the weighted distribution of covariates (age, sex, race, education, BMI, history of regular smoking, and number of clinical conditions) was calculated by final dementia status, and their statistical differences were examined using t‐tests (continuous variables) and chi‐squared tests (categorical variables).

Survey‐weighted cause‐specific Cox proportional hazards regression models were performed to evaluate the association of each difficulty/impairment (in separate models) and the number of impairments with incident dementia. We performed weighted estimations to account for NHATS' complex survey design using the 2011 survey weights. The proportional hazard assumption for each variable of interest was checked using log–log plots. Models were adjusted for participants' baseline characteristics, including age, sex, race, education, BMI, history of regular smoking, and number of clinical conditions. To address potential reverse causality, we conducted a sensitivity analysis excluding participants who were classified as having dementia at the second visit. This approach aimed to reduce the possibility that sensory difficulties or motor impairments were influenced by pre‐existing cognitive decline and the risk of ratings of sensory difficulties being influenced by poor cognitive status. We also repeated our main analysis by including participants with proxy respondents. In another sensitivity analysis, we used Fine‐Gray competing risk regression to estimate the subdistribution hazard of incident dementia, accounting for the competing risk of death [32]. Analyses were performed using R4.3.1. Statistical significance was set at p‐value < 0.05.

3. Results

Of the 8245 participants interviewed at baseline, 636 (7.7%) were missing a dementia diagnosis and 2034 (24.7%) were classified as having either possible or probable dementia and were therefore removed from the analytic sample. Furthermore, we removed 103 (1.2%) participants who used a proxy respondent to minimize potential misclassification of sensory difficulty. We also excluded 941 (11.4%) with incomplete dementia data, and 17 (0.2%), 592 (7.2%), and 75 (0.9%) with missing sensory difficulty, motor function, and covariate information, respectively (Figure S1, Table S1), leaving 3847 participants in the analytic sample.

Baseline characteristics of the analytic sample are presented in Table 1 as weighted estimates. More than half of the study population was female (56.0% [95% CI: 54.1–57.9]) with an average age of 73.9 [95% CI: 73.7–74.1] years, and the majority were non‐Hispanic White (85.4% [95% CI: 83.6–87.2]). Participants who developed dementia were on average older, less likely to be non‐Hispanic White, had fewer years of education, were less likely to have smoked, and had a higher number of comorbidities (p < 0.001 for all). Those who developed dementia had a higher prevalence of vision difficulty and hearing difficulty (p < 0.0001 for both). Those who developed dementia also had a higher prevalence of impairments in standing balance, gait speed, chair stand ability, grip strength, and a greater number of overall sensorimotor impairments (p < 0.0001 for all).

TABLE 1.

Weighted characteristics of participants by final dementia status, National Health and Aging Trends Study, Round 1, 2011.

Variable	All (N = 3847)	Dementia (N = 1412)	No dementia (N = 2435)	p
Age (years), mean (95% CI)	73.9 (73.7, 74.1)	76.6 (76.2, 77.0)	72.7 (72.4, 72.9)	< 0.0001
Sex, % (95% CI)				0.19
Male	44.0 (42.1, 46.0)	42.3 (39.7, 44.9)	44.8 (42.2, 47.4)
Female	56.0 (54.1, 57.9)	57.7 (55.1, 60.3)	55.2 (52.6, 57.8)
Race, % (95% CI)				< 0.001
Non‐Hispanic white	85.4 (83.6, 87.2)	81.8 (79.0, 84.5)	87.1 (85.2, 88.9)
Non‐Hispanic black	6.8 (5.9, 7.8)	8.7 (7.3, 10.0)	6.0 (5.1, 6.9)
Hispanic	5.3 (4.1, 6.5)	7.4 (5.4, 9.3)	4.4 (3.1, 5.6)
Other	2.5 (1.8, 3.2)	2.2 (1.1, 3.3)	2.6 (1.8, 3.4)
Education, % (95% CI)				< 0.0001
Up to high school	41.9 (39.4, 44.3)	51.0 (46.9, 55.1)	37.7 (35.2, 40.1)
Some college	28.2 (26.8, 29.7)	26.3 (23.1, 29.4)	29.1 (27.2, 31.1)
Bachelor's and above	29.9 (27.3, 32.5)	22.7 (19.1, 26.3)	33.2 (30.2, 36.2)
BMI a , % (95% CI)				0.15
Healthy weight	32.0 (30.0, 34.0)	33.7 (31.2, 36.2)	31.2 (28.9, 33.5)
Overweight	39.3 (37.5, 41.1)	37.5 (35.2, 39.8)	40.1 (37.9, 42.4)
Obese	28.7 (27.1, 30.4)	28.8 (26.1, 31.5)	28.7 (26.8, 30.6)
Smoking history b				< 0.01
No (never)	47.2 (45.2, 49.2)	50.3 (47.6, 53.1)	45.8 (43.5, 48.1)
Yes (ever)	52.8 (50.8, 54.8)	49.7 (46.9, 52.4)	54.2 (51.9, 56.5)
Number of comorbidities, % (95% CI)				< 0.001
0	16.5 (15.0, 18.0)	12.9 (10.7, 15.1)	18.2 (16.3, 20.0)
1	28.3 (26.6, 30.0)	27.6 (24.6, 30.5)	28.6 (26.4, 30.8)
2	26.6 (25.1, 28.2)	26.0 (23.5, 28.4)	26.9 (24.9, 28.9)
3	16.6 (15.1, 18.1)	19.3 (16.8, 21.8)	15.3 (13.7, 17.0)
≥ 4	12.0 (10.9, 13.2)	14.2 (12.3, 16.2)	11.0 (9.6, 12.4)
Standing balance score, mean (95% CI)	2.7 (2.7, 2.7)	2.4 (2.3, 2.5)	2.8 (2.8, 2.9)	< 0.0001
Gait speed (m/s), mean (95% CI)	0.8 (0.8, 0.8)	0.8 (0.7, 0.8)	0.9 (0.8, 0.9)	< 0.0001
Chair stand speed (stands/s), mean (95% CI)	0.5 (0.5, 0.5)	0.4 (0.4, 0.5)	0.5 (0.5, 0.5)	0.12
Grip Strength (kg), mean (95% CI)	27.4 (26.9, 27.8)	25.3 (24.6, 3)	28.3 (27.7, 28.9)	< 0.0001
Balance impairment category c , % (95% CI)				< 0.0001
Not impaired	58.7 (57.0, 60.5)	48.4 (45.6, 51.2)	63.5 (61.4, 65.6)
Impaired	41.3 (39.6, 43.0)	51.6 (48.8, 54.5)	36.5 (34.4, 38.6)
Gait speed impairment category d , % (95% CI)				< 0.0001
Not impaired	54.9 (52.8, 57.0)	42.1 (39.5, 44.8)	60.8 (58.2, 63.3)
Impaired	45.1 (43.0, 47.2)	57.9 (55.3, 60.5)	39.2 (36.7, 41.8)
Chair stand impairment category e , % (95% CI)				< 0.0001
Not impaired	50.9 (48.7, 53.1)	40.6 (37.6, 43.5)	55.7 (53.2, 58.2)
Impaired	49.1 (46.9, 51.3)	59.4 (56.5, 62.4)	44.4 (41.9, 46.9)
Grip Strength impairment category f , % (95% CI)				< 0.0001
Not impaired	54.2 (52.3, 56.0)	42.7 (39.4, 46.0)	59.4 (57.3, 61.6)
Impaired	45.9 (44.0, 47.7)	57.3 (54.0, 60.6)	40.6 (38.5, 42.7)
Vision difficulty binary g , % (95% CI)				< 0.0001
Not having difficulty	94.8 (94.0, 95.7)	92.3 (90.8, 93.8)	96.0 (95.2, 96.7)
Having difficulty	5.2 (4.3, 6.1)	7.7 (6.2, 9.2)	4.1 (3.3, 4.8)
Hearing difficulty binary h , % (95% CI)				< 0.0001
Not having difficulty	80.8 (79.3, 82.3)	75.8 (73.9, 77.7)	83.1 (81.1, 85.1)
Having difficulty	19.2 (17.8, 20.7)	24.2 (22.3, 26.1)	16.9 (14.9, 18.9)
Number of sensorimotor impairments				< 0.0001
0	18.0 (16.4, 19.6)	11.3 (8.8, 13.7)	21.0 (19.2, 22.9)
1	23.1 (21.4, 24.8)	16.3 (14.3, 18.3)	26.2 (24.2, 28.2)
2	21.6 (20.3, 23.0)	20.0 (17.3, 22.7)	22.4 (20.7, 24.0)
3	16.8 (15.7, 18.0)	20.0 (17.8, 22.2)	15.3 (14.1, 16.6)
4	14.4 (13.3, 15.4)	22.0 (20.1, 23.9)	10.9 (9.8, 12.0)
5+	6.1 (5.3, 7.0)	10.5 (8.6, 12.4)	4.2 (3.4, 5.0)

Note: Bold values denote statistical significance at the p < 0.05 level.

^a BMI (body mass index) categories defined as underweight: BMI < 18.5, healthy weight: 18.5 ≤ BMI < 25, overweight: 25 ≤ BMI < 30, obese: BMI ≥ 30.

^b Smoking history is defined as a self‐reported history of smoking regularly.

^c Standing balance impairment defined as NHATS SPPB balance component < 3.

^d Gait speed impairment is defined as gait speed < 0.8 m/s, or NHATS SPPB walking = 0.

^e Chair speed impairment defined as chair stand speed < 0.42 chair stand/s, or NHATS SPPB chair stand = 0.

^f Grip strength impairment is defined as grip strength < 32 for men and grip strength < 20 for women, or NHATS SPPB grip = 0.

^g Vision difficulty is defined as self‐reported blindness, an inability to see well enough to recognize someone across the street, or an inability to see well enough to read a newspaper.

^h Hearing difficulty is defined as self‐reported deafness, hearing aid use, inability to hear well enough to use the telephone, or inability to hear well enough to carry on a conversation in a room with the radio or television on.

For the entire sample, the average length of follow‐up was 5.2 (±3.9) years; 4.0 (±3.1) years for those with dementia and 5.9 (±4.1) years for those without dementia. Figures 1, 2, 3 present the Kaplan–Meier survival curves showing the survival probability for incident dementia by each type of sensory difficulty (Figure 1), motor impairment (Figure 2), and number of sensory difficulties and motor impairments (Figure 3, Figures S2 and S3) at baseline. The probability of remaining free of dementia decreased over time across all impairments and was lower for those with a higher number of impairments. The median dementia‐free survival time was 5 years for visual difficulty, 8 years for hearing difficulty, and 7–8 years for motor impairments.

FIGURE 1.

Kaplan–Meier survival curves for incident dementia by baseline sensory difficulty: (A) Vision difficulty, and (B) Hearing difficulty. Each survey round is separated by 1 year.

FIGURE 2.

Kaplan–Meier survival curves for incident dementia by baseline motor function impairment: (A) Standing balance impairment, (B) Gait speed impairment, (C) Chair stand impairment, and (D) Grip strength impairment. Each survey round is separated by 1 year.

FIGURE 3.

Kaplan–Meier survival curves for incident dementia by baseline number of sensorimotor impairments. Each survey round is separated by 1 year.

Results from Cox proportional hazard models estimating the hazard of incident dementia for each baseline impairment are presented in Table 2, with progressive covariate adjustment. In unadjusted models, vision difficulty and hearing difficulty were each independently associated with an increased risk of dementia, compared to those without difficulty (HR [95% CI]: 1.91 [1.62–2.27] and 1.48 [1.27–1.74]), respectively. After adjusting for demographics, BMI, history of smoking, and number of clinical conditions, the association of vision difficulty with incident dementia persisted (HR: 1.34 [1.13–1.60]), but the association of hearing difficulty with incident dementia was attenuated (HR: 1.02 [0.89–1.16]). Additionally, having two sensory difficulties, compared to no sensory difficulty, was associated with a higher risk of incident dementia; however, this association did not remain significant after adjusting for demographics, BMI, smoking history, and number of clinical conditions.

TABLE 2.

Cox proportional hazards analysis modeling the association between sensory difficulties and motor impairments, and number of impairments, with incident dementia, National Health and Aging Trends Study (Rounds 1–12, 2011–2022). Each row represents a separate model.

Independent variable	Model 1 c		Model 2 d		Model 3 e
	HR (95% CI)	p	HR (95% CI)	p	HR (95% CI)	p
Hearing: having difficulty a	1.48 [1.27–1.74]	< 0.0001	1.03 [0.90–1.17]	0.70	1.02 [0.89–1.16]	0.82
Vision: having difficulty a	1.91 [1.62–2.27]	< 0.0001	1.38 [1.15–1.64]	< 0.001	1.34 [1.13–1.60]	< 0.01
One sensory function having difficulties b	1.53 [1.3–1.79]	< 0.0001	1.12 [0.98–1.28]	0.10	1.1 [0.96–1.26]	0.18
Two sensory functions having difficulties b	2.50 [1.85–3.40]	< 0.0001	1.16 [0.78–1.71]	0.47	1.15 [0.80–1.66]	0.44
Balance impaired a	1.95 [1.73–2.20]	< 0.0001	1.27 [1.14–1.43]	< 0.0001	1.23 [1.10–1.39]	< 0.001
Gait speed impaired a	2.29 [2.04–2.56]	< 0.0001	1.54 [1.37–1.72]	< 0.0001	1.49 [1.34–1.67]	< 0.0001
Chair stand speed impaired a	1.82 [1.62–2.03]	< 0.0001	1.36 [1.22–1.52]	< 0.0001	1.31 [1.18–1.47]	< 0.0001
Grip strength impaired a	1.93 [1.68–2.21]	< 0.0001	1.35 [1.17–1.57]	< 0.001	1.31 [1.13–1.51]	< 0.001
One motor impairment b	1.17 [0.95–1.46]	0.14	1.01 [0.81–1.26]	0.93	0.99 [0.79–1.24]	0.92
Two motor impairments b	1.73 [1.37–2.20]	< 0.0001	1.26 [1.00–1.57]	0.04	1.21 [0.97–1.51]	0.09
Three motor impairments b	2.71 [2.23–3.29]	< 0.0001	1.69 [1.39–2.05]	< 0.0001	1.61 [1.33–1.95]	< 0.0001
Four motor impairments b	4.24 [3.43–5.24]	< 0.0001	2.11 [1.67–2.66]	< 0.0001	1.97 [1.57–2.48]	< 0.0001
Number of sensorimotor impairments b
One	1.17 [0.91–1.51]	0.22	1.00 [0.78–1.29]	0.98	0.98 [0.77–1.25]	0.87
Two	1.73 [1.32–2.27]	< 0.001	1.26 [0.98–1.62]	0.08	1.21 [0.94–1.57]	0.13
Three	2.46 [1.94–3.11]	< 0.0001	1.51 [1.20–1.92]	< 0.001	1.44 [1.14–1.82]	< 0.01
Four	3.95 [3.11–5.02]	< 0.0001	2.06 [1.59–2.65]	< 0.0001	1.92 [1.50–2.47]	< 0.0001
Five or more	5.07 [3.92–6.57]	< 0.0001	2.12 [1.63–2.75]	< 0.0001	1.96 [1.50–2.56]	< 0.0001

Note: Bold values denote statistical significance at the p < 0.05 level.

^a Reference is “not impaired.”

^b Reference is 0 impairments.

^c Unadjusted model.

^d Adjusted for age, sex, BMI (body mass index), levels of education, and race.

^e Adjusted for age, sex, BMI, levels of education, race, history of regular smoking, and number of clinical conditions.

Each baseline motor impairment—balance, gait speed, chair stand speed, and grip strength—was associated with incident dementia in unadjusted models, with each motor function examined in a separate model (Table 2). Each motor impairment association persisted after adjusting for covariates. Compared to those without standing balance impairment, the HR for incident dementia for those with balance impairment was 1.23 (95% CI: 1.10–1.39). Compared to those without gait speed impairment, the HR for incident dementia for those with gait speed impairment was 1.49 (95% CI: 1.34–1.67). Compared to those without chair stand impairment, the HR for incident dementia for those with chair stand impairment was 1.31 (95% CI: 1.18–1.47). Compared to those without grip strength impairment, the HR for incident dementia for those with grip strength impairment was 1.31 (95% CI: 1.13–1.51, p < 0.001). When considering number of motor impairments, the HR for incident dementia was greater for those with three (HR: 1.61 [95% CI: 1.33–1.95]) or four (HR: 1.97 [95% CI: 1.57–2.48]) motor impairments, compared to those with no motor impairments.

When number of sensorimotor impairments were examined collectively, the risk of incident dementia was higher after the number of impairments exceeded two, with a HR of 1.44 (95% CI 1.14–1.82, p = 0.003) for those with three impairments, 1.92 (95% CI 1.50–2.47, p < 0.0001) for those with four impairments, and 1.96 (95% CI 1.50–2.56, p < 0.0001) for those with five or more impairments (Table 2).

In sensitivity analyses excluding those who developed dementia in the year after baseline assessment (Round 12), results remained similar (Table S2). In sensitivity analyses using Fine‐Gray regression models to estimate the subdistribution hazard for incident dementia accounting for the competing risk of death, results remained similar as well (Table S3). When adding participants with proxy respondents, the results remained similar (Table S4).

4. Discussion

In this nationally representative sample of older adults aged 65 years and older, we found that having visual difficulty or motor impairments at baseline significantly increased the risk of incident dementia over up to 11 years of follow‐up. We also found that being impaired in three or more sensorimotor domains was associated with an increased risk of incident dementia over time. Collectively, these results suggest that the presence of sensory difficulty or motor impairment may act as a warning sign of deteriorating physical and cognitive health that warrants clinical intervention and follow‐up, especially for those with multiple sensorimotor impairments.

Our findings are consistent with previous work examining self‐reported vision difficulty and dementia incidence in older adults [26, 33]. A meta‐analysis of 14 studies in adults aged 50 years and older found a pooled relative risk of 1.47 (95% CI, 1.36–1.60) for incident dementia over follow‐up periods ranging from 3.7 to 14.5 years among those with vision difficulty (determined using self‐reported data and/or medical chart review) [34]. Our hazard ratio of 1.34 over 11 years reinforces these findings and highlights a median dementia‐free survival time of 5 years for those with uncorrected or under‐corrected visual difficulty.

Although hearing difficulty has emerged in recent years as an important risk factor for dementia, previous findings on the association of self‐reported hearing difficulty with incident dementia have been mixed. Specifically, our current findings are in agreement with several similar studies reporting non‐significant associations among self‐reported hearing and dementia risk. Research on self‐reported hearing difficulty and dementia incidence in older adults found no association in the Ginkgo Evaluation of Memory (GEM) study over an 8‐year follow‐up [35], the Health and Retirement Study (HRS) over an 18‐year follow‐up [36], and NHATS over a 7‐year follow‐up [26]. However, self‐reported hearing difficulty was associated with incident dementia over an 11‐year period in the English Longitudinal Study of Ageing (ELSA) [37]. The demographic and methodological differences between NHATS and the ELSA study make it difficult to compare findings, as the population was generally younger (50–69 years) with differing definitions of hearing difficulty and dementia. Although vision appeared to be a significant predictor, it should be noted that the association was no longer present when also considering hearing difficulty, which may in part be explained by the limited number of participants with two sensory difficulties.

Previous research also suggests motor impairments are associated with increased dementia risk. Balance impairment was associated with a higher risk of incident dementia over 10–15 years of follow‐up in studies from HRS (HR = 1.58 [95% CI, 1.26–1.96]) [38], ELSA (HR = 1.97 [95% CI, 1.24–3.14]) [38], and among 143,788 Korean older adults (HR = 1.83 [95% CI, 1.69–2.00]) [39]. Slow gait speed was consistently associated with a higher risk of dementia over 10–15 years of follow‐up in HRS (HR = 1.42 [95% CI, 1.23–1.64]) [38], ELSA (HR = 1.55 [95% CI, 1.21–1.97]) [38], and a meta‐analysis of 12 studies with follow‐up periods ranging from 3 to 9 years (pooled HR and RR = 1.53 [95% CI, 1.42–1.65]) [40]. Slower performance on chair stand tests was associated with a higher risk of dementia in ELSA participants over a 15‐year follow‐up period (HR = 1.56 [95% CI, 1.23–1.99]) [38] and being in the lowest sex‐specific quartile of performance was associated with a higher risk of incident dementia in a study of Japanese older adults (HR = 1.69 [95% CI, 1.10–2.59]) [41]. Finally, lower grip strength was associated with a higher risk of dementia in the UK Biobank over a mean follow‐up of 9.1 years (HR = 1.84 [95% CI, 1.66–2.05]) [42] and in a meta‐analysis of 10 studies with follow‐up periods ranging from 3.6 to 16 years (HR = 1.54 [95% CI, 1.32–1.79]) [5]. Although these results are consistent with our findings, the current study is the first to examine all motor function components in the same analytic sample. Furthermore, we observed evidence of a graded relationship when investigating the number of present motor impairments with future dementia risk. It should be noted that definitions of dementia and motor impairment(s) differ by study, making cross‐study comparisons challenging.

Several pathways may explain potential mechanisms of the association between sensorimotor impairments with increased dementia incidence. First, difficulties with standing balance, walking speed, chair stands, and/or grip strength may reflect a decline in function in the peripheral and central nervous systems, which may share pathways with cognitive processes. For example, maintaining balance is a complex task that involves visual and motor brain regions, including those responsible for coordination and spatial awareness [43]. Some difficulties in motor function, such as balance and gait, may represent manifestations of underlying neurodegenerative disorders such as Parkinson's disease, which increase the risk of developing dementia [19, 44, 45, 46]. Second, difficulties in sensory and impairments in motor functioning can lead to reduced physical function and lower levels of physical activity and social engagement, which are important for maintaining cognitive health [47, 48]. Motor function impairments, such as difficulty rising from a chair or grasping objects, represent a decline in overall physical function that can impact an individual's ability to leave the home and perform daily activities independently. Finally, common underlying factors such as cardiovascular disease and inflammation can simultaneously impair motor function and increase the vulnerability of cerebrovascular damage (e.g., white matter hyperintensities), increasing the risk of developing dementia [49]. Although the current study highlights important connections between sensorimotor impairments and dementia incidence, more research is needed to discern specific biological and social/behavioral mechanisms linking physical and cognitive impairments in later life, which may identify opportunities for intervention and treatment.

To our knowledge, this is the first study to investigate the cumulative impact of sensory difficulties and motor impairments on dementia incidence in a nationally representative population of US older adults. In addition, we explored the cumulative impact of multiple impairments, providing evidence of a graded relationship between an increasing number of impairments and dementia. Our findings indicate that vision difficulties and gait speed characterize the highest risk of incident dementia, and that multiple impairments in sensorimotor function convey a higher burden of dementia risk. Routine assessment of sensory difficulty and motor function is lacking in clinical practice. Assessing motor function is inexpensive and requires minimal equipment and staff training. The current results suggest implementation of motor function assessment into routine clinical practice for adults older than 65 years may provide important insights into future cognitive health. Notably, the median dementia‐free survival time was 7–8 years for those with motor impairments, suggesting an important time window for initiating dementia screening and prevention efforts.

A few limitations should be noted. First, the NHATS definitions of possible and probable dementia are not clinically adjudicated; however, the algorithm has been validated against the widely used Health and Retirement Study Aging, Demographics, and Memory Study (HRS‐ADAMS) and has a sensitivity of 71.8% and specificity of 83.7% for possible and probable dementia, respectively [50]. Second, the analytic sample was limited to those who were eligible to participate in motor function testing, which excluded non‐ambulatory participants or those who declined the tests, who may be at higher risk of dementia. Third, we used measures of vision and hearing captured via self‐reported sensory difficulty in everyday life. While this is reflective of how participants perceive their sensory abilities, previous research suggests that older adults may underreport/overreport vision and hearing difficulty [51, 52], which may influence the sensory difficulties associations. Moreover, hearing aid use was included in the definition of hearing difficulty. It is important to note, however, that while amplification may improve audibility, it does not fully correct underlying auditory dysfunction or meaningfully alter dementia risk [12], and that excluding those who use hearing aids would exclude those with the highest degree of hearing loss. More research is needed to study the longitudinal effect of objectively measured vision and hearing impairment and the resulting sensorimotor impairment on dementia risk in older adults. Moreover, we did not account for genetic risk factors (such as APOE ε4 status), which may influence dementia risk and modify the observed longitudinal associations.

5. Conclusion

In this nationally representative sample of US Medicare beneficiaries with sensory difficulty and motor impairments, those with difficulty in vision and impaired motor function (standing balance, gait speed, chair stands, and grip strength) were at a higher risk of incident dementia over 11 years than those without sensory difficulty or motor impairments. These findings suggest that screening for and addressing sensory difficulty and motor function impairments in older adults may be crucial for dementia prevention, highlighting the need for early detection and intervention efforts for managing dementia.

Author Contributions

J.A.S. contributed to data collection. A.D. and J.A.S. developed the study design. A.D. performed data analysis and interpretation. A.D. in collaboration with J.A.S. wrote the first draft. All authors contributed to the revision of the manuscript. J.A.S. provided supervision, and J.A.S. and Y.A. provided funding. Everyone who contributed significantly to this article has been listed.

Disclosure

The analyses, conclusions, opinions, and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred.

Conflicts of Interest

A.A.W. and J.A.S. received an honorarium from The Villages Inc.

Supporting information

Data S1: jgs70250‐sup‐0001‐Supinfo.pdf.

Davoudi A., Dougherty R. J., Wanigatunga A. A., et al., “Sensorimotor Impairment and Incident Dementia in the US Medicare Beneficiaries,” Journal of the American Geriatrics Society 74, no. 2 (2026): 458–469, 10.1111/jgs.70250.

Data Availability Statement

NHATS data is publicly available at www.nhats.org.