Participation in Cognitive Activities is Associated with Foot Reaction Time and Gait Speed in Older Adults

Abstract

Background

Given the evidence of the links between cognition and mobility, participation in cognitive activities may benefit neuromotor performance and mobility in older adults.

Aims

To examine the association between participation in cognitive activities and foot reaction time (RT) and gait speed in community-dwelling older adults.

Methods

The MOBILIZE Boston Study II (MBSII) re-enrolled 354 community-dwelling older adults aged ≥ 70 years from the original MBS cohort. Of these, 310 completed the performance testing and we excluded 3 participants who had Parkinson’s disease. Cognitive Activities Scale (CAS) assessed participation in 17 cognitive activities. Simple and Choice foot RT (SRT, CRT, msec) and gait speed (m/s) were measured using a sensored GAITRite® gait mat.

Results

The average age of the 307 participants was 84 years; 79% were white and 65% were women. The average CAS score was 25.5±11.7 indicating participation in approximately 26 activities per week on average. The average foot SRT was 245±57msec and average CRT was 323±85msec. Usual-paced gait speed was 0.9±0.3 m/s on average. More frequent participation in cognitive activities was associated with shorter SRT (β = −0.759, p=0.015) and CRT (β= −1.125, p=0.013), and faster gait speed (β=0.003, p=0.026), after adjusting for potential confounders.

Discussion

Participation in cognitively stimulating activities may be beneficial for neuromotor performance and mobility in older adults.

Conclusions

Prospective and intervention studies are needed to determine whether participation in cognitive activities may prevent mobility decline over time, and thus reduce fall risk.

INTRODUCTION

Falls are a threat to the health of older adults, interfering with functional independence in everyday life. It is estimated that 35 to 45 percent of older adults living in the community have at least one fall each year [1]. Given the financial burden and increased mortality risk as a result of fall-related injuries, it becomes crucial to identify older adults at greatest risk for falls and implement fall prevention strategies [2, 3]. Falls are multifactorial, involving sensorimotor function and multisensory integration [4, 5]. As the quick response to unexpected hazards by gripping or stepping is critical to avoiding falls, foot reaction time (RT) testing has been used to assess fall risk and discriminate fallers and non-fallers [6]. Slower gait speed, which occurs with aging, also is identified as a strong indicator of increased fall risk [7].

Recent evidence demonstrates close links between cognition and mobility that become more evident with aging [8]. Older adults face challenges to mobility that are in part related to cognition, for example, slower walking while talking or while engaged in a cognitive task [9]. This age-associated change is associated with fall risk [9]. Motor control and gait performance involve intact neural processing from multiple brain regions [10–12]. Previous studies support that older adults with better cognitive performance commonly manifest better mobility and lower risk for falls [13, 14]. Cognitive impairment significantly impacts multisensory integration processes and motor function, which then in turn impacts gait performance and increases fall risk [15–18].

A growing body of evidence have showed that participation in cognitive activities is associated with better cognitive performance and reduced risk of dementia [19–21]. Given the link between cognition and mobility, participation in cognitive activities might benefit neuromotor performance and mobility in older adults [13, 14]. Previous studies reported that engagement in leisure activities was associated with better performance in mobility and physical function [22, 23]. This potential beneficial effects of leisure activities on mobility, however, could be due to either the physical activity component or the cognitive aspects of activities, or both. In order to understand whether primarily cognitively-based activities may be related to neuromotor performance and mobility, this study aims to examine the association between participation in cognitive activities and foot RT and gait speed in community-dwelling older adults. This exploratory study may provide insights in support of interventions involving cognitive activities to maintain or improve mobility and decrease fall risk among older adults.

METHODS

Study Design

The MOBILIZE Boston Study II (MBS II) is the third assessment wave of this population-based cohort. Approximately 6.5 years after the original MBS began with a cohort of 765 adults aged 70 and older living in the Boston area, the MBS II re-enrolled 354 participants who continued to live in the community. Details of the MBS study methods were published previously [24].

For the original cohort, participants were included if aged 70 years or older, and able to communicate in English and walk 20 feet without personal assistance. Eligible spouses aged 65 and older were also enrolled. Older adults were excluded if they had terminal disease, severe vision or hearing deficits, or moderate to severe cognitive impairment measured by the Mini-Mental State Examination (MMSE) score < 18. For the current study, 310 participants completed the MBS II clinic assessment that included reaction time and gait testing. Three participants with Parkinson’s disease were excluded because of potential interference with neuromotor testing.

The MBS II assessment consisted of a telephone health interview followed by a 3-hour clinic-based assessment at the Institute for Aging Research at Hebrew SeniorLife in Boston, MA. An update on medical conditions and health behaviors was collected during the telephone interview, followed by the clinic assessment comprising health, function and performance measures including foot RT and gait speed. The study was approved by the Institutional Review Boards at Hebrew SeniorLife and the University of Massachusetts Boston.

Cognitive Activities Assessment

The modified version of the Cognitive Activities Scale (CAS) developed by Verghese and colleagues [19], assessed the frequency of participation in 17 cognitively stimulating activities: reading, playing games, using computers, and social and other activities in the previous month. Responses are coded to create a CAS score such that a participant receives 1 point for each activity that they participate in one day per week, 4 points for participating 2 to 6 days per week, and 7 points for daily participation. Total CAS score is the sum of points for all 17 activities, with a maximum score of 119. The scale has good validity and reliability in measuring cognitive activities in older adults [19].

Outcome Measures

Reaction Time

Simple Reaction Time (SRT) and Choice Reaction Time (CRT), were measured on a 16-ft sensored gait mat with an active area of 192 inches in length and 24 inches in width (GAITRite®, CIR Systems Inc., Franklin, NJ) with the PKMAS software (Protokinetics, Havertown, PA). The mat, connected to a Windows computer, measures the geometry and arrangement of the footfall. Participants were instructed to sit in a straight-backed chair with their feet placed flat on the mat behind a white line. In the SRT test, participants were instructed to pick up a self-selected foot and quickly tap a blue dot on the mat in response to a randomly intermittent light on the right side of the mat as quickly as possible. In the CRT test, participants responded to the random light on either side of the mat by quickly picking up the foot on the same side as the light and tapping the mat. We measured RT from the time the light turned on to the movement initiation of the correct foot. Participants performed 10 trials in both SRT and CRT tests, and the average of each series was used in the data analysis. To measure reaction time, the intermittent light is wired directly to the sensored mat instrumentation so the timing of the foot response to the light is embedded in the GAITRite equipment, synced directly with the mat sensor software. The assessors were trained to follow a scripted protocol, involving initial practice sessions with the participant to ensure they understood the test procedure and were able to perform the test correctly. The GAITRite® mat is widely used and its instrumentation has good validity and reliability in measuring temporal-spatial parameters in older adults [25, 26].

Gait Speed

In the clinical assessment, participants were instructed to walk at a usual pace on the gait mat for two trials of the 4-meter walk. Gait speed (meters/second) was determined using 4 meters divided by time to complete the fastest trial in seconds. Gait speed from the 4-meter walk is strongly predictive of disability in older adults [27].

Covariates

Sociodemographic characteristics (e.g., age, gender, race, education) were recorded in the MBS baseline home interview. In MBS II, chronic conditions such as diabetes, osteoarthritis, peripheral neuropathy, and peripheral artery disease were determined using disease algorithms [24]. Other physician-diagnosed chronic conditions such as heart disease were self-reported. Height and weight were measured using a stadiometer and balance scale during the clinic assessment. We calculated the body mass index (BMI) based on the NHLBI cut points. Mini-Mental State Exam (MMSE) assessed global cognitive function during the clinic assessment [28]. Mobility difficulty was determined by self-reported difficulty walking ¼ mile or walking upstairs to the second floor without personal assistance. The 12-item joint pain questionnaire assessed chronic musculoskeletal pain lasting for 3 or more months in the past year and present in the past month in hands/wrists, shoulders, back, hip, knee, and feet [29]. The 10-foot distant vision test using the Good-Lite Chart™ was used to determine vision impairment [30].

Statistical Analysis

Descriptive analyses determined the means and standard deviations or percentages for all variables. Reaction time variables were winsorized, with outliers recoded at the ninety-ninth percentile of SRT and CRT. Chi-square tests and t-tests assessed group differences of sociodemographic characteristics, medical conditions, and other covariates according to tertiles of CAS score. Generalized Linear Models (GLM) tested the trend of age-adjusted least squares means of RT and gait speed according to tertiles of CAS score. Multivariable linear regression models were performed to evaluate associations between CAS score and outcomes of RT and gait speed, adjusted for potential confounders including sociodemographic characteristics, obesity, diabetes, peripheral arterial disease, peripheral neuropathy, mobility difficulty, pain distribution, and vision impairment. In a final step, we examined the potentially moderating effects of cognitive function, testing the interaction between CAS and MMSE scores (dichotomized at the median, MMSE=26) in relation to RT and gait speed outcomes in the fully adjusted models. The significance level was alpha=0.05. Data analyses were performed using SAS software 9.4 (SAS Institute Inc., Cary, NC).

RESULTS

The final sample consisted of 307 older adults with an average age of 84.0 years (SD=4.4), ranging from 71 to 101 years old. Most participants (79%) were white and 65% were women. The average CAS score was 25.5 (SD=11.7) indicating that participants engaged in approximately 26 activities per week on average. A number of sociodemographic and health characteristics were associated with CAS score (Table 1).

Table 1.

Sociodemographic Characteristics, Chronic Conditions, and Pain Characteristics according to the Cognitive Activity Scale (CAS)^a, 307 Community-living Older Adults, MOBILIZE Boston II, 2011-2015.

	n	Low CAS (n = 109) %	Med. CAS (n = 94) %	High CAS (n =104) %	Trendb (p-value)
Age (years)
<80	38	6(5.5)	11(11.7)	21(20.2)
80-84	150	52(47.7)	42(44.7)	56(53.9)
85-89	78	32(29.4)	28(29.8)	18(17.3)
≥90	41	19(17.4)	13(13.8)	9(8.7)	<0.001
Gender
Male	109	37(33.9)	39(41.5)	33(31.7)
Female	198	72(66.1)	55(58.5)	71(68.3)	0.327
Race
White	244	75(68.8)	73(77.7)	96(92.3)
African American	47	32(29.4)	12(12.8)	3(2.9)
Other	16	2(1.8)	9(9.6)	5(4.8)	<0.001
Education
< College graduate	72	48(44.4)	10(10.64)	14(13.5)
College graduate	234	60(55.6)	84(89.4)	90(86.5)	<0.001
Body mass index
< 25	116	44(41.1)	28(30.4)	44(43.1)
25 – 29	122	44(41.1)	38(41.3)	40(39.2)
≥ 30	63	19(17.8)	26(28.3)	18(17.7)	0.863
Mobility difficulty	144	64(58.7)	44(46.8)	36(34.6)	<0.001
MMSE <24	44	32(31.7)	12(13.0)	0	<0.001
Heart disease	125	47(44.8)	38(42.2)	40(39.6)	0.455
Diabetes	35	16(14.7)	11(11.7)	8(7.7)	0.110
Peripheral neuropathy	58	22(21.4)	21(23.6)	15(14.9)	0.247
Peripheral artery disease	35	18(16.5)	12(12.8)	5(4.8)	0.008
Vision impairment	55	26(24.5)	20(21.5)	9(8.7)	0.003
Pain sites
None	100	30(27.5)	28(29.8)	42(40.4)
Single-site	73	21(19.3)	28(29.8)	24(23.1)
Multisite	95	39(35.8)	24(25.5)	32(30.8)
Widespread	39	19(17.4)	14(14.9)	6(5.8)	0.004

Note.

^aCognitive Activity Scale, grouped into tertiles.

^bMantel-Haenzel chi-square test for trend (1 d.f.), except for race (2 d.f.).

The average foot RT was 245 msec (SD=57) for SRT and 323 msec (SD=85) for CRT. Average gait speed was 0.9 m/s (SD=0.3). Simple and choice foot RT were strongly inversely associated with CAS scores and faster gait speed also was strongly associated with CAS (Fig. 1). These relationships persisted after adjusting for sociodemographic characteristics, vision impairment, obesity, chronic conditions, and mobility difficulty (SRT, p=0.015; CRT, p=0.013; and gait speed, p=0.026) (Table 2). The association between CAS and RT was strongest among participants who had lower MMSE scores (Table 3). In testing for interaction, MMSE score moderated the effect of CAS on SRT and CRT but not gait speed (test of interaction p=0.005, p=0.04, p=0.30, respectively).

Fig. 1.

Age-adjusted means for simple reaction time (SRT), choice reaction time (CRT) and gait speed, according to tertiles of cognitive activity scale (CAS) score. Least squares means derived from generalized linear models (GLM); p-value indicates test for trend (1 d.f.).

Table 2.

Relationship between Cognitive Activity Scale, Foot Reaction Time and Gait Speed, 307 Community-living Older Adults, MOBILIZE Boston II, 2011-2015.

Outcomes	Mean (SD)	Model 1a			Model 2b
		B	SE B	p	B	SE B	p
SRT (msec)	245 (57)	−0.782	0.302	0.010	−0.759	0.309	0.015
CRT (msec)	323 (85)	−1.251	0.442	0.005	−1.125	0.451	0.013
Gait Speed (m/s)	0.9 (0.3)	0.004	0.001	0.005	0.003	0.001	0.026

Note. SRT=Simple foot reaction time, CRT=Choice foot reaction time.

^aMultivariable linear regression, model 1 adjusted for age, gender, race, and education.

^bModel 2 additionally includes obesity, diabetes, peripheral arterial disease, peripheral neuropathy, mobility difficulty, pain distribution, and vision impairment.

Table 3.

Association between Cognitive Activities with Foot Reaction Time and Gait Speed according to MMSE score, 307 Community-living Older Adults, MOBILIZE Boston II, 2011-2015.

	Low MMSE score (≤26)a (N=155)			High MMSE score (>26)a (N=152)
Outcomes	B	SE B	p	B	SE B	p
SRT (msec)	−1.477	0.523	0.006	−0.086	0.388	0.825
CRT (msec)	−1.737	0.748	0.022	−1.023	0.571	0.076
Gait Speed (m/s)	0.004	0.002	0.046	0.002	0.002	0.474

Note. SRT=Simple foot reaction time, CRT=Choice foot reaction time.

^aMMSE dichotomized at the median score of the sample. Multivariable linear regression model adjusted for age, gender, race, education, obesity, diabetes, peripheral arterial disease, peripheral neuropathy, mobility difficulty, pain distribution, and vision impairment.

DISCUSSION

To our knowledge, this is the first study to examine the association between participation in cognitive activities and foot reaction time and gait speed in community-dwelling older adults. The study found that more frequent participation in cognitive activities is associated with shorter foot RT and faster gait speed. Specifically, participants with the greatest participation in cognitive activities had an approximately 0.12 m/s faster gait speed compared to those in the least cognitively active group. In prospective studies, a change in gait speed of 0.1m/s is considered to be a substantial meaningful change associated with functional decline and survival in older adults [31, 32]. Thus, the differences we observed in gait speed related to cognitive activity were not only statistically significant, but clinically meaningful. These results support the idea that participation in cognitively stimulating activities may be beneficial for neuromotor performance and mobility especially for those with poorer cognitive function, with implications for possibly reducing fall risk in older adults living in community.

Rapid foot movement measured by foot reaction time is essential for avoiding hazards and regaining balance when confronted with an obstacle or in the process of a slip or trip. As one of the components of the Physiological Profile Assessment (PPA) to assess fall risk, slower foot RT has been identified as an important factor contributing to falls [6, 33, 34]. Reaction time is also a cognitive-based test, in which processing speed and attentional capabilities are needed to initiate movement. Others have shown that participation in cognitive activities is associated with better cognitive performance in both global cognitive tests and selected cognitive domains such as memory and processing speed [35, 36]. Also, cognitive activities have been found to have long-term beneficial effects on reduced risk of cognitive decline and dementia in older population over a 5-year follow-up [19, 37]. Thus, it is plausible that the beneficial effects of cognitive activities on cognitive function extend to faster reaction time.

Studies link cognitive function to performance-based physical function including gait speed and patient reported functional measures, and suggest a bidirectional relationship between cognition and mobility [8, 16]. In addition, cognitive training interventions demonstrate modest effects on simple and complex walking performance [38]. Participation in a range of cognitive activities may be important, and it is possible that the added social component of many cognitively challenging activities is an important motivator for older adults. Notably, these activities may benefit those most at risk for cognitive decline as evidenced by our findings showing those with lower MMSE scores demonstrated the strongest association between CAS and RT. Previous research also demonstrated incorporating cognitive training focusing on attention and executive function to physical exercise interventions significantly improved gait performance in older adults with cognitive impairment [39]. Our study may potentially support the beneficial effects of home-based interventions on cognitive leisure activities such as reading and playing games that are feasible for older adults living in the community. Nonetheless, future work could determine which cognitive activities may be most beneficial for improving RT and gait, and potentially reducing fall risk.

Several potential underlying mechanisms could support the pathway from cognitive activities to improved mobility-related performance. Researchers reported that engagement in cognitive activities is associated with greater gray matter volume [35]. The central nervous system (CNS) plays an important role in lower extremity motor control involving speed and initiation of movement; and abnormalities in the prefrontal cortex and basal ganglia are associated with gait speed [40, 41]. Evidence supports the association between brain changes such as white matter hyperintensities and physical function and fall risk in older adults [41, 42]. Further, gait speed and cognitive function may share the same brain substrates [43]. Although substantial research evidence is emerging on the cognition-mobility relationship, the underlying mechanisms between brain function and impaired mobility are not well understood [40].

There are several limitations to the study. First, the temporal relationship between cognitive activities and foot RT or gait speed cannot be inferred due to the cross-sectional design of this study. An alternative explanation could be that older adults with better mobility and physical performance are more likely to participate in cognitive activities. Second, because RT was measured in a seated position, the test may capture more of the cognitive aspects of foot RT, than if performed from a standing position which would more closely capture impacts on mobility. Notably, we found associations with gait speed were also strong and others have found that seated foot RT is associated with fall risk [33]. Importantly, using the seated approach allowed our very elderly participants to safely complete the testing.

In conclusion, our results suggest that participation in cognitive activities may contribute to better neuromotor performance and mobility, thus potentially decreasing fall risk in older adults. Further research is needed to determine the longitudinal relationship between participation in cognitive leisure activities and preservation of neuromotor function and mobility with aging. If confirmed, the evidence would support intervention studies of cognitive activities to improve mobility and decrease fall risk in vulnerable older adults especially for those with cognitive impairment. Future home-based and community-based interventions adding cognitive components need to be considered toward the goal of maintaining physical and cognitive functioning and quality of life with advancing age.

Availability of data and material:

Not applicable.