Abstract
Background
The longitudinal association between cerebral amyloid-beta (Aβ) and change in gait, and whether this association is mediated by cortical thickness, has yet to be determined.
Methods
We included 439 clinically normal (CN) participants, aged 50–69 years and enrolled in the Mayo Clinic Study of Aging with cerebral Aβ, cortical thickness, and gait measurements. Cerebral Aβ deposition was assessed by Pittsburgh Compound B (PiB)-PET in multiple regions of interest (ROIs) (ie, frontal, orbitofrontal, parietal, temporal, anterior cingulate, posterior cingulate/precuneus, and motor). Cortical thickness was assessed on 3T MRI in corresponding ROIs. Gait parameters (gait speed, cadence, stride length, double support time, and covariance of stance time) were measured with GAITRite. Multivariate-adjusted two level structural equation models were used to examine the longitudinal association between PiB-PET, cortical thickness, and change in gait over a median 15.6 months.
Results
Higher PiB-PET in all ROIs was associated with decreasing cadence and increasing double support time, and in the temporal ROI was associated with declining gait speed. In sex-stratified analyses, higher PiB-PET in all ROIs was associated with declining performance on all gait parameters among women. In contrast, among men, the only association was with higher orbitofrontal ROI PiB-PET and declining cadence. None of the associations were mediated by cortical thickness or attenuated after adjustment of baseline cognition.
Conclusion
Higher PiB-PET was associated with declining gait, particularly among women in this middle-aged CN cohort, independent of cortical thickness and baseline cognitive. Elevated brain Aβ may play a critical role in age-related mobility decline.
Keywords: Alzheimer’s pathology, Functionality, Structural equation modeling
Gait disturbances in elderly adults are associated with an increased risk of cognitive decline and dementia (1,2). While several studies have examined the association between gait and cerebral vascular pathology (3,4), few studies have examined the relationship between gait and other types of neuropathology (eg, amyloid-beta [Aβ], neurofibrillary tangles). One autopsy study reported that greater decline in gait speed prior to death was associated with more Alzheimer’s disease (AD)-pathology (neuritic plaques, diffuse plaques, and neurofibrillary tangles) (5). Similarly, three cross-sectional studies and one longitudinal study of older adults (mean ages of 75–86 years) reported that higher Aβ PET was associated with slower gait speed (6–8). This association remained after adjustment for hippocampal volume and white matter hyperintensities (WMHs) (7).
Because gait abnormalities can be influenced by other brain pathologies, which increase with age, it is difficult to isolate the effects of brain Aβ deposition on gait. We therefore cross-sectionally examined the association between Aβ PET and gait among clinically normal (CN) 50–69-year olds. Among this middle-aged cohort, we also found that brain Aβ deposition across ROIs was associated with poorer performance on multiple gait parameters and was independent of AD-associated neurodegeneration (9). Further, in sex-stratified analyses, this association was only significant among women.
It is not yet understood whether brain Aβ is associated with change in gait. The aggregation of Aβ as a precursor to tau phosphorylation and eventual neurodegeneration has been posited as the process which leads to AD-associated cognitive and functional impairment (10). Therefore, the aim of the present analyses was to investigate the longitudinal association between Aβ PET and multiple gait parameters among CN participants aged 50–69, and whether this association was mediated by cortical thickness, a measure of neurodegeneration. Lastly, we again examined whether the association between amyloid PET and gait differed by sex.
Methods
Study Population
The MCSA is a prospective population-based cohort study. It began in 2004, and was designed to determine the incidence and prevalence of mild cognitive impairment in Olmsted County, MN. Using the medical records linkage system of the Rochester Epidemiology Project (REP), Olmsted County residents between the ages of 70 and 89 were identified in an age- and sex-stratified random sampling design so that men and women were equally represented in each 10-year age strata (11,12). Recruitment of participants aged 50 years and older began in 2012. Participants complete in-clinic visits approximately 15 months apart, which include a physician examination, an interview by a study coordinator, and neuropsychological testing. The present study included 439 CN participants, aged 50 to 69 years, with concurrent neuroimaging and gait measures, and follow-up gait measures. We excluded participants from the analyses with a history of stroke, alcoholism, Parkinson’s disease, subdural hematoma, traumatic brain injury, and/or normal pressure hydrocephalus. Study protocols were approved by the Mayo Clinic and Olmsted Medical Center Institutional Review Boards. All participants provided written informed consent.
Gait Assessment
Gait parameters were assessed with GAITRite instrumentation (CIR systems Inc., Havertown, PA), an electronic walkway 5.6 m in length and 0.9 m wide. Participants were instructed to walk at their normal pace without gait aids, initiating and terminating their walk 1 m before and after the walkway (13). We analyzed spatio-temporal, spatial, and temporal gait parameters, including participant gait speed (m/s); cadence (steps per minute); stride length, defined as the distance (cm) between successive heel contact points on the same foot; double support time, defined as the amount of time (seconds) that both feet are on the walkway; and intra-individual variation in stance time (coefficient of variation). Stride length and double support time were measured on each side (ie, left and right) for each step. We created a single average value across both sides and all steps for each of these parameters separately.
Imaging
Aβ PET images were formed using PiB (14), and were obtained 40–60 minutes after injection. We analyzed SUVR from the MCSA AD-associated ROIs: prefrontal, orbitofrontal, parietal, temporal, anterior cingulate, and posterior cingulate/precuneus, and a motor-specific ROI which consisted of the precentral gyrus, postcentral gyrus, Rolandic operculum, and supplementary motor area (15). All ROIs were normalized to cerebellar grey matter uptake (16,17). PiB-PET images were partial volume corrected (18). PET atlas regions were localized using a rigid alignment to the corresponding MRI, which was segmented and parcellated using SPM5 (19) and an in-house template/atlas.
Participants completed MRI and PET scans at a single visit; CT was obtained for PET attenuation correction. All MRI scans were completed on one of three General Electric 3T scanners using a sagittal 3D magnetization prepared rapid acquisition gradient recalled echo (MP-RAGE) sequence. Repetition time (TR) was ≈2,300 ms, echo time (TE) ≈3 ms, and inversion time (TI) = 900 ms. Voxel dimensions were ≈1.2 × 1.015 × 1.015 mm. Gradient distortion in the sagittal plane was performed on-scanner, and through-plane correction was performed as part of image processing (20). Intensity inhomogeneity was corrected using first the N3 algorithm (21), followed by the SPM5-based bias correction (19). From these preprocessed images, the cortical surface was segmented and cortical thickness values were estimated using FreeSurfer version 5.3.0 (https://surfer.nmr.mgh.harvard.edu/) (22). Cortical thickness estimations were then resampled from FreeSurfer outputs to the input images’ native space using FreeSurfer tools (23). This allowed us to parcellate FreeSurfer thickness values into the same atlas regions used for our PET analyses, in order to perform regional correlations.
Covariates
Baseline participant characteristics including age, sex, education, and body mass index (BMI) were ascertained in-clinic. Participants completed the Beck Depression Inventory (BDI). A score ≥13 was considered evidence of depression (24). A blood draw was used to obtain APOE genotype. Medical conditions and the Charlson comorbidity index (25) were determined by medical record abstraction using the REP medical records-linkage system (11,26).
Cognition Diagnosis
Determination of cognitive status was based on consensus agreement between the study coordinator, examining physician, and neuropsychologist who evaluated the participant. The diagnosis considered education, prior occupation, visual or hearing deficits, informant interview, and all other participant clinical information (12). Cognitive test performance in four domains (memory, executive function, language, and visual-spatial) and an average of these four (global) was compared with the age-adjusted scores of CN individuals previously obtained using Mayo’s Older American Normative Studies (27). Participants with scores approximately −1.0 SD below the age-specific mean in the general population were considered for a clinical diagnosis of possible mild cognitive impairment. Individuals who performed in the normal range and did not meet criteria for mild cognitive impairment or dementia, which was diagnosed using Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition criteria (28), were deemed CN.
Statistical Analyses
Participant baseline characteristics were compared by sex using two-tailed Wilcoxon rank sum tests for continuous variables and Fisher exact test for dichotomous variables. We created z-scores for each of the gait parameters to make them comparable.
To determine the best approach for fitting structural equation models (SEMs), we first used mixed effects models to determine which covariates to include in the SEMs (mixed effects models are not presented). The mixed effects models examined the association between the baseline PiB-PET SUVR and change in each z-scored gait parameter treating subject-specific intercepts, but not time, as random effects. We added an interaction term between PiB-PET SUVR in each ROI and sex to determine whether sex modified the association between PiB-PET SUVR and gait; these interaction terms were significant at p less than .10 so were considered in the SEMs.
We next fit two-level SEMs to determine the associations between baseline PiB-PET SUVR, baseline cortical thickness, and change in gait. These two-level models each incorporated one “within subjects” level (with random effects) and one “between subjects” level, with the primary assessment of direct and indirect effects between subjects. Direct effects represented the direct association between PiB-PET SUVR and gait change adjusted for baseline age, sex, BMI, and APOE. Indirect effects represented the association between PiB-PET SUVR and gait mediated by cortical thickness in the corresponding ROI adjusted for age, sex, BMI, and APOE. The total effect was the summation of the direct and indirect effects. Because the interaction terms between sex and PiB-PET SUVR ROIs were significant in our mixed effects models, we additionally conducted sex-stratified analyses. These analyses were completed with MPLUS v. 7.3 (Muthén & Muthén, Los Angeles, CA). p values for the effects were calculated using the multivariate delta method (29). We set α to 0.05 to identify statistical significance. Comparisons of participants’ baseline characteristics and mixed effects modeling were completed using Stata Version 13.0 (Stata Corp, College Station, TX).
Results
Participant characteristics are presented in Table 1. Overall, participants were later middle aged (median = 62.6 years), had a median of 15.0 years of education, and a median BMI of 28.6 kg/m2. The median follow-up was 15.6 months (interquartile range [IQR] = 13.2, 28.8). Men had lower cadence, longer stride length, longer double support time, reduced cortical thickness, and poorer performance in all cognitive domains, except visuospatial, compared to women. There were no other significant differences between men and women at baseline.
Table 1.
Participant Characteristics at Baseline by Sex, Median (interquartile range), or n (%)
| All (n = 439) | Men (n = 233) | Women (n = 206) | p Value | |
|---|---|---|---|---|
| Age | 62.6 (57.2, 66.3) | 63.0 (58.2, 66.6) | 62.2 (56.4, 65.6) | .16 |
| Years of education | 15 (13,17) | 16 (13,17) | 14.5 (13,16) | .30 |
| APOE ε4 allele | 121 (28) | 64 (27) | 57 (28) | >.99 |
| Body mass index (kg/m2) | 28.6 (25.9, 32.4) | 29.1 (26.7, 32.0) | 28.0 (24.6, 33.1) | .07 |
| Gait speed (m/s) | 1.2 (1.1, 1.3) | 1.2 (1.1, 1.3) | 1.2 (1.1, 1.3) | .24 |
| Cadence (steps/min) | 108.5 (101.9, 114.1) | 105.7 (99.9, 110.4) | 112.2 (105.9, 118.7) | <.001 |
| Stride length (cm) | 135.1 (124.1, 145.3) | 141.2 (131.1, 149.7) | 129.8 (119.3, 137.8) | <.001 |
| Double support time (seconds) | 0.31 (0.28, 0.35) | 0.32 (0.30, 0.36) | 0.30 (0.27, 0.35) | <.001 |
| Log coefficient variation stance time | −3.7 (−4.0, −3.7) | −3.7 (−4.0, −3.4) | −3.4 (−4.1, −3.4) | .99 |
| Mean PiB-PET SUVR | 3.2 (3.1, 3.3) | 3.2 (3.1, 3.3) | 3.2 (3.1, 3.6) | .59 |
| Mean cortical thickness (mm) | 5.9 (5.7, 6.1) | 5.9 (5.7, 6.1) | 6.0 (5.8, 6.1) | .002 |
| Diabetes | 53 (12) | 33 (14) | 20 (10) | .19 |
| Hypertension | 194 (44) | 113 (49) | 81 (40) | .05 |
| Depression | 32 (7) | 19 (8) | 13 (6) | .58 |
| Charlson comorbidity index | 3 (2,5) | 3 (2,5) | 3 (2,4) | .43 |
| Memory (z-score) | 1.24 (0.60, 1.83) | 1.06 (0.30, 1.58) | 1.50 (0.86, 2.04) | <.001 |
| Attention (z-score) | 1.29 (0.74, 1.78) | 1.09 (0.63, 1.60) | 1.47 (0.97, 1.96) | <.001 |
| Language (z-score) | 0.99 (0.47, 1.48) | 0.83 (0.22, 1.32) | 1.11 (0.72, 1.62) | <.001 |
| Visuospatial (z-score) | 1.10 (0.44, 1.60) | 1.19 (0.54, 1.76) | 0.94 (0.37, 1.47) | 0.023 |
| Global (z-score) | 1.43 (0.91, 1.91) | 1.25 (0.68, 1.73) | 1.60 (1.05, 2.04) | <.001 |
| Follow-up time (months) | 15.6 (13.2, 28.8) | 16.8 (14.4, 28.8) | 15.6 (13.2, 28.8) | .13 |
Note: The values that are significant at a p-value <.05 are bolded. PiB = C11 Pittsburgh Compound B; ROI = Region of interest; SUVR = Standardized uptake volume ratio.
Mean PiB-PET SUVR and cortical thickness was calculated as the average of PiB-PET SUVR or cortical thickness across all ROIs.
Depression was determined by a score ≥13 on the Beck Depression Inventory.
In SEMs, we examined both the direct and indirect effects of PiB-PET SUVR on gait in relation to regional cortical thickness (Figure 1). Greater baseline PiB-PET SUVR in all ROIs was associated with a decrease in cadence and an increase in double support time (Table 2). Moreover, greater PiB-PET SUVR in the temporal ROI was associated with declining gait speed. The estimates in these models were significant for direct effects but not indirect effects. Thus, greater PiB-PET SUVR was directly associated with poorer gait parameters and this association was not mediated by cortical thickness in any region.
Figure 1.
Structural equation models (SEMs) path analyses. These pathways were tested in the SEMs. The pathway from amyloid to gait represents a direct effect, and the lines from amyloid to thickness to gait represent an indirect effect (mediation by thickness). Their sum represents the total effect of amyloid on gait.
Table 2.
Association Between Baseline PiB-PET SUVR and Longitudinal Gait Parameters, Accounting for Mediation by Cortical Thickness (n = 440)
| Speed | Cadence | Stride Length | Double Support Time | Stance Time | ||
|---|---|---|---|---|---|---|
| Region | Aβ Effect on Gait | Estimate (SE) | Estimate (SE) | Estimate (SE) | Estimate (SE) | Estimate (SE) |
| Total | −0.19 (0.10) | −0.50 (0.11)*** | −0.13 (0.09) | 0.19 (0.06)** | 1.34 (0.98) | |
| Prefrontal | Direct | −0.16 (0.10) | −0.48 (0.11)*** | −0.14 (0.09) | 0.18 (0.06)** | 1.40 (0.98) |
| Indirect | −0.03 (0.02) | −0.02 (0.02) | 0.003 (0.02) | 0.007 (0.01) | −0.06 (0.06) | |
| Total | −0.36 (0.18)* | −0.93 (0.20)*** | −0.29 (0.16) | 0.37 (0.11)** | 2.59 (1.71) | |
| Orbitofrontal | Direct | −0.35 (0.18) | −0.92 (0.20)*** | −0.29 (0.16) | 0.37 (0.11)** | 2.58 (1.71) |
| Indirect | −0.006 (0.01) | −0.01 (0.01) | 0.001 (0.01) | 0.01 (0.01) | 0.01 (0.02) | |
| Total | −0.14 (0.10) | −0.46 (0.11) *** | −0.11 (0.08) | 0.17 (0.06)** | 1.33 (0.99) | |
| Parietal | Direct | −0.13 (0.10) | −0.45 (0.11) *** | −0.11 (0.08) | 0.16 (0.06)** | 1.35 (0.99) |
| Indirect | −0.01 (0.01) | −0.01 (0.01) | 0.001 (0.01) | 0.002 (0.01) | −0.01 (0.03) | |
| Total | −0.62 (0.26)* | −1.06 (0.29) *** | −0.41 (0.22) | 0.43 (0.15)** | 3.03 (2.02) | |
| Temporal | Direct | −0.62 (0.25)* | −1.06 (0.29) *** | −0.40 (0.22) | 0.43 (0.15)** | 3.06 (2.01) |
| Indirect | −0.01 (0.02) | −0.01 (0.02) | −0.003 (0.02) | 0.002 (0.01) | −0.03 (0.12) | |
| Anterior cingulate | Total | −0.29 (0.20) | −0.93 (0.23) *** | −0.24 (0.18) | 0.36 (0.13)** | 2.76 (1.97) |
| Direct | −0.27 (0.20) | −0.91 (0.23) *** | −0.25 (0.18) | 0.36 (0.13)** | 2.88 (1.97) | |
| Indirect | −0.02 (0.03) | −0.02 (0.03) | 0.01 (0.03) | 0.001 (0.01) | −0.12 (0.12) | |
| Posterior cingulate/precuneus | Total | −0.27 (0.18) | −0.83 (0.21) *** | −0.26 (0.16) | 0.32 (0.12)** | 2.42 (1.80) |
| Direct | −0.26 (0.18) | −0.82 (0.21) *** | −0.27 (0.16) | 0.32 (0.12)** | 2.41 (1.80) | |
| Indirect | −0.01 (0.02) | −0.01 (0.02) | 0.01 (0.02) | 0.003 (0.01) | 0.01 (0.05) | |
| Total | −0.04 (0.04) | −0.16 (0.04) *** | −0.03 (0.04) | 0.05 (0.02)* | 0.44 (0.31) | |
| Motor | Direct | −0.04 (0.04) | −0.15 (0.04) *** | −0.03 (0.04) | 0.05 (0.02)* | 0.44 (0.31) |
| Indirect | −0.01 (0.01) | −0.003 (0.01) | −0.003 (0.004) | 0.002 (0.002) | 0.001 (0.01) |
Note: The values that are significant at a p-value <.05 are bolded. PiB = C11 Pittsburgh Compound B; SE = Standard error; SUVR = Standardized uptake volume ratio.
Models adjusted for age, body mass index, and APOE.
*p ≤ .05, **p ≤ .01, ***p ≤ .001.
In sex-stratified analyses, among men, there was a total effect between the prefrontal ROI and cadence (estimate [SE], p value −0.34 [0.16], .029) and a direct association between greater PiB-PET SUVR in the orbitofrontal ROI and declining cadence (estimate [SE], p value −0.62 [0.28], .024; Table 3). Among women, greater PiB-PET SUVR in all ROIs was directly associated with declining performance on all gait parameters, with the exception that greater PiB-PET SUVR was not associated with stance time. There was no evidence that any of these associations were mediated by cortical thickness.
Table 3.
Association Between Baseline PiB-PET SUVR and Longitudinal Gait Parameters, Accounting for Mediation by Cortical Thickness by Sex
| Men (n = 234) | ||||||
|---|---|---|---|---|---|---|
| Speed | Cadence | Stride Length | Double Support Time | Stance Time | ||
| Region | Aβ Effect on Gait | Estimate (SE) | Estimate (SE) | Estimate (SE) | Estimate (SE) | Estimate (SE) |
| Prefrontal | Total | −0.05 (0.14) | −0.34 (0.16)* | 0.17 (0.11) | −0.01 (0.06) | −0.17 (0.23) |
| Direct | 0.04 (0.13) | −0.29 (0.15) | 0.17 (0.10) | −0.03 (0.06) | −0.08 (0.22) | |
| Indirect | −0.09 (0.04)* | −0.04 (0.03) | −0.001 (0.03) | 0.02 (0.01) | −0.09 (0.07) | |
| Orbitofrontal | Total | 0.22 (0.21) | −0.62 (0.28)* | 0.26 (0.18) | 0.10 (0.10) | −0.33 (0.43) |
| Direct | 0.21 (0.20) | −0.62 (0.28)* | 0.25 (0.18) | 0.10 (0.10) | −0.32 (0.43) | |
| Indirect | 0.01 (0.03) | 0.01 (0.03) | 0.004 (0.02) | −0.003 (0.01) | −0.003 (0.02) | |
| Parietal | Total | 0.16 (0.12) | −0.31 (0.15)* | 0.19 (0.11) | 0.01 (0.06) | −0.20 (0.27) |
| Direct | 0.21 (0.12) | −0.27 (0.15) | 0.21 (0.11) | −0.01 (0.06) | −0.16 (0.26) | |
| Indirect | −0.05 (0.03) | −0.04 (0.03) | −0.02 (0.02) | 0.02 (0.01) | −0.04 (0.05) | |
| Temporal | Total | 0.20 (0.33) | −0.59 (0.36) | 0.03 (0.32) | 0.13 (0.14) | −0.10 (0.69) |
| Direct | 0.20 (0.33) | −0.59 (0.36) | 0.03 (0.32) | 0.13 (0.14) | −0.11 (0.68) | |
| Indirect | 0.002 (0.02) | 0.003 (0.02) | 0.001 (0.004) | −0.001 (0.01) | −0.02 (0.11) | |
| Anterior cingulate | Total | 0.36 (0.21) | −0.43 (0.31) | 0.37 (0.22) | 0.02 (0.12) | −0.30 (0.50) |
| Direct | 0.39 (0.20) | −0.47 (0.30) | 0.39 (0.22) | 0.01 (0.12) | −0.21 (0.49) | |
| Indirect | −0.04 (0.04) | −0.05 (0.04) | −0.02 (0.03) | 0.01 (0.02) | −0.09 (0.08) | |
| Posterior cingulate/precuneus | Total | 0.34 (0.21) | −0.48 (0.28) | 0.36 (0.20) | −0.01 (0.11) | −0.54 (0.48) |
| Direct | 0.37 (0.21) | −0.46 (0.28) | 0.35 (0.20) | −0.01 (0.11) | −0.52 (0.47) | |
| Indirect | −0.03 (0.03) | −0.01 (0.03) | 0.01 (0.02) | 0.003 (0.01) | −0.02 (0.06) | |
| Motor | Total | 0.08 (0.04) | −0.08 (0.06) | 0.07 (0.04) | −0.003 (0.02) | −0.02 (0.11) |
| Direct | 0.09 (0.04)* | −0.08 (0.06) | 0.08 (0.04) | −0.01 (0.02) | −0.01 (0.11) | |
| Indirect | −0.01 (0.01) | −0.003 (0.01) | −0.004 (0.01) | 0.002 (0.002) | −0.01 (0.01) | |
| Women (n = 206) | ||||||
| Speed | Cadence | Stride length | Double support time | Stance time | ||
| Region | Aβ effect on gait | Estimate (SE) | Estimate (SE) | Estimate (SE) | Estimate (SE) | Estimate (SE) |
| Prefrontal | Total | −0.39 (0.12)*** | −0.60 (0.17)*** | −0.34 (0.11)** | 0.32 (0.11)** | 3.46 (2.07) |
| Direct | −0.43 (0.11)*** | −0.60 (0.17)*** | −0.35 (0.11)** | 0.32 (0.11)** | 3.51 (2.07) | |
| Indirect | 0.03 (0.03) | −0.001 (0.03) | 0.01 (0.03) | 0.001 (0.01) | −0.05 (0.14) | |
| Orbitofrontal | Total | −0.78 (0.20)*** | −1.18 (0.30)*** | −0.73 (0.22)*** | 0.63 (0.18)*** | 6.45 (3.01)* |
| Direct | −0.79 (0.19)*** | −1.16 (0.30)*** | −0.75 (0.22)*** | 0.62 (0.18)*** | 6.43 (3.01)* | |
| Indirect | 0.01 (0.02) | −0.02 (0.03) | 0.02 (0.02) | 0.01 (0.01) | 0.03 (0.05) | |
| Parietal | Total | −0.32 (0.12)*** | −0.53 (0.17)** | −0.28 (0.13)* | 0.29 (0.11)* | 3.10 (1.94) |
| Direct | −0.33 (0.12)*** | −0.53 (0.17)** | −0.29 (0.13)* | 0.29 (0.11)* | 3.12 (1.94) | |
| Indirect | 0.01 (0.01) | 0.004 (0.01) | 0.01 (0.02) | −0.002 (0.01) | −0.02 (0.05) | |
| Temporal | Total | −1.35 (0.31) *** | −1.48 (0.48)** | −0.73 (0.28)** | 0.75 (0.29)* | 7.39 (4.24) |
| Direct | −1.34 (0.31)*** | −1.46 (0.47)** | −0.72 (0.28)** | 0.74 (0.29)* | 7.46 (4.24) | |
| Indirect | −0.02 (0.05) | −0.01 (0.04) | −0.01 (0.04) | 0.004 (0.01) | −0.06 (0.20) | |
| Anterior cingulate | Total | −0.77 (0.25)** | −1.22 (0.33)*** | −0.73 (0.25)** | 0.65 (0.21)** | 7.17 (3.78) |
| Direct | −0.79 (0.24)** | −1.24 (0.33)*** | −0.78 (0.24)*** | 0.67 (0.21)** | 7.41 (3.77)* | |
| Indirect | 0.02 (0.04) | 0.02 (0.04) | 0.06 (0.04) | −0.02 (0.02) | −0.24 (0.28) | |
| Posterior cingulate/precuneus | Total | −0.64 (0.21)** | −1.00 (0.31)*** | −0.65 (0.23)** | 0.54 (0.19)** | 5.64 (3.36) |
| Direct | −0.66 (0.21)** | −1.00 (0.31)*** | −0.67 (0.22)** | 0.54 (0.19)** | 5.69 (3.36) | |
| Indirect | 0.02 (0.04) | 0.002 (0.04) | 0.02 (0.04) | 0.01 (0.02) | −0.05 (0.16) | |
| Motor | Total | −0.14 (0.05)** | −0.21 (0.07)** | −0.10 (0.05)* | 0.10 (0.04)* | 1.11 (0.72) |
| Direct | −0.15 (0.05)** | −0.21 (0.07)** | −0.10 (0.05)* | 0.10 (0.04)* | 1.10 (0.72) | |
| Indirect | 0.01 (0.01) | −0.001 (0.01) | 0.003 (0.01) | 0.001 (0.01) | 0.01 (0.03) | |
Note: The values that are significant at a p-value <.05 are bolded. PiB = C11 Pittsburgh Compound B; SE = Standard error; SUVR = Standardized uptake volume ratio.
Models adjusted for age, body mass index, and APOE.
*p ≤ .05, **p ≤ .01, ***p ≤ .001.
In sensitivity analyses, we examined whether individual medical comorbidities (eg, diabetes, hypertension), cognition (global cognitive z-score), and the Charlson comorbidity index affected the associations between PiB-PET SUVR and gait, but found they did not.
Discussion
In this study of CN adults, aged 50–69 years, greater baseline PiB-PET SUVR was associated with declines on multiple gait parameters over a median follow-up of 15.6 months. In sex-stratified analyses, these associations were strongest among women. There was no evidence that the association between PiB-PET SUVR and gait was mediated by cortical thickness in corresponding ROIs. These results were not attenuated after adjusting for baseline cognition.
Previous cross-sectional studies reported that greater Aβ burden in the precuneus, posterior cingulate cortex, and whole brain was associated with lower gait speed among both CN and cognitively impaired elderly participants (6). These findings were independent of hippocampal volume and WMHs but were attenuated after adjusting for cognitive test performance. An initial longitudinal study among a smaller sample of 59 CN older adults (mean = 75 years, range = 56–89 years) reported that higher PiB-PET SUVR was associated with declining gait speed and mobility (ie, chair stands, standing balance, timed usual pace and narrow walk tests) (8). These findings remained significant after adjusting for medical comorbidities, cerebrovascular pathology, and cognitive test performance.
Our longitudinal findings in a younger and larger CN sample additionally demonstrate that higher baseline PiB-PET SUVR is associated with declines in multiple gait parameters. These results suggest that elevated brain Aβ across multiple ROIs directly contribute to declines in gait parameters and may play a critical role in age-related mobility decline. Further, among individuals with elevated brain Aβ, changes in gait may be an earlier manifestation of disease pathology than cognitive impairment. Indeed, gait decline has been shown to precede cognitive decline (2), but not vice versa (30). Alternatively, it may be that reserve in our younger, healthy, CN sample is masking any decline in, or attenuation by, cognitive test performance.
We hypothesized that the association between PiB-PET SUVR and gait would be mediated by cortical thickness in the same ROIs. However, our findings were independent of cortical thickness, including in the motor cortex. It is possible that subtle atrophy in this middle-aged population (median of 62.6 years) is not detectable with current imaging methods and that this subtle atrophy is influencing gait. Alternatively, Aβ-related neurotoxicity may contribute to gait declines through different pathways than those measured.
Like our cross-sectional findings, we found the association between PiB-PET SUVR and change in gait was stronger among women than men. Women show a steeper decline in gait speed with age (31), which may contribute to the observed sex-specific findings. Alternatively, the observed sex-specific associations could suggest that women may be more susceptible to the negative effects of AD pathology (32).
This study has multiple strengths, including its longitudinal design, thorough phenotyping of participants, and population-based sample. However, the findings of the study must also be viewed within the scope of its limitations. We were unable to adjust for other types of brain pathology (eg, phosphorylated tau, TAR DNA-binding protein 43 [TDP43]). Findings from the present study suggest that cortical thickness in corresponding ROIs does not mediate the association between greater Aβ and decline in multiple measures of gait. Moreover, evidence from other studies suggest that other measures of neuropathology or neurodegeneration, including WMHs, hippocampal volume, glucose metabolism, and cortical thickness (7,9,33), do not attenuate the association between greater Aβ and disrupted mobility. However, it may be that other types of pathology, such as tau and TDP43, which are present even in these younger age groups (34,35) lie on the pathway between Aβ aggregation and gait disruption. Additionally, data on functional measures (eg, muscle strength, history of recurrent falls) associated with declining gait are not available in the MCSA; therefore, we were unable to adjust for these potential confounders.
The present findings bolster the evidence from previous cross-sectional studies (6,7,9,33), and suggest that Aβ accumulation is associated with declines in multiple gait parameters and is independent of cortical thickness. These associations were significant even in this relatively young, middle-aged sample. There are multiple implications of these results. First, the study further suggests that elevated brain Aβ may play a critical role in age-related mobility decline. Second, these findings may have implications for screening and identifying patients at risk of substantial Aβ burden, which is becoming increasingly important as therapeutics aimed at Aβ are being developed. Studies with longer follow-up are needed to further investigate the association between brain Aβ and gait. Moreover, as tau imaging data become available and as tools to measure other types of neuropathology are developed, future research should investigate how these types of pathology may mediate the association between Aβ and mobility.
Funding
This study was supported by National Institutes of Health grants U01 AG006786, R01 AG011378, R01 AG041851, P50 AG044170, P50 AG016574, R01 NS097495, and R01 AG049704; the GHR Foundation and the Mayo Foundation for Medical Education and Research, and was made possible by the Rochester Epidemiology Project (R01 AG034676).
Conflict of Interest
A.M.V.W., T.G.L., C.G.S., R.S., and C.E.H. report no disclosures. R.O.R. receives funding from the National Institutes of Health. D.S.K. serves as Deputy Editor for Neurology; serves on a Data Safety Monitoring Board for Lundbeck Pharmaceuticals and for the DIAN study; is an investigator in clinical trials sponsored by TauRX Pharmaceuticals, Lilly Pharmaceuticals and the Alzheimer’s Disease Cooperative Study; and receives research support from the National Institutes of Health. P.V. receives funding from the National Institutes of Health. C.R.J. has provided consulting services for Eli Lilly. He receives research funding from the National Institutes of Health, and the Alexander Family Alzheimer’s Disease Research Professorship of the Mayo Clinic. R.C.P. is a consultant for Roche, Inc, Merck, Inc., Biogen, Inc. and Eli Lilly and Company Genentech, Inc; receives publishing royalties from Mild Cognitive Impairment (Oxford University Press, 2003), and receives research support from the National Institutes of Health. M.M.M. served as a consultant to Eli Lilly and Lysosomal Therapeutics, Inc. She receives research support from the National Institute on Aging, National Institutes of Health and unrestricted research grants from Biogen, Lundbeck, and Roche.
References
- 1. Abellan van Kan G, Rolland Y, Andrieu S et al. Gait speed at usual pace as a predictor of adverse outcomes in community-dwelling older people an International Academy on Nutrition and Aging (IANA) Task Force. J Nutr Health Aging. 2009;13:881–889. [DOI] [PubMed] [Google Scholar]
- 2. Savica R, Wennberg AM, Hagen C et al. Comparison of gait parameters for predicting cognitive decline: the Mayo Clinic Study of Aging. J Alzheimers Dis. 2017;55:559–567. doi:10.3233/JAD-160697 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Annweiler C, Montero-Odasso M. Vascular burden as a substrate for higher-level gait disorders in older adults. a review of brain mapping literature. Panminerva Med. 2012;54:189–204. [PubMed] [Google Scholar]
- 4. Zheng JJ, Delbaere K, Close JC, Sachdev PS, Lord SR. Impact of white matter lesions on physical functioning and fall risk in older people: a systematic review. Stroke. 2011;42:2086–2090. doi:10.1161/STROKEAHA.110.610360 [DOI] [PubMed] [Google Scholar]
- 5. Buchman AS, Yu L, Wilson RS, Schneider JA, Bennett DA. Association of brain pathology with the progression of frailty in older adults. Neurology. 2013;80:2055–2061. doi:10.1212/WNL.0b013e318294b462 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Del Campo N, Payoux P, Djilali A et al. ; MAPT/DSA Study Group Relationship of regional brain β-amyloid to gait speed. Neurology. 2016;86:36–43. doi:10.1212/WNL.0000000000002235 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Nadkarni NK, Perera S, Snitz BE et al. Association of brain amyloid-β with slow gait in elderly individuals without dementia: influence of cognition and apolipoprotein E ε4 genotype. JAMA Neurol. 2017;74:82–90. doi:10.1001/jamaneurol.2016.3474 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Tian Q, Resnick SM, Bilgel M, Wong DF, Ferrucci L, Studenski SA. beta-amyloid burden predicts lower extremity performance decline in cognitively unimpaired older adults. J Gerontol A Biol Sci Med Sci. 2017;72:716–723. doi:10.1093/gerona/glw183 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Wennberg AMV, Savica R, Hagen CE et al. Cerebral amyloid deposition is associated with gait parameters in the mayo clinic study of aging. J Am Geriatr Soc. 2017;65:792–799. doi:10.1111/jgs.14670 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Jack CR Jr, Knopman DS, Jagust WJ et al. Tracking pathophysiological processes in Alzheimer’s disease: an updated hypothetical model of dynamic biomarkers. Lancet Neurol. 2013;12:207–216. doi:10.1016/S1474-4422(12)70291-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. St Sauver JL, Grossardt BR, Yawn BP et al. Data resource profile: the Rochester Epidemiology Project (REP) medical records-linkage system. Int J Epidemiol. 2012;41:1614–1624. doi:10.1093/ije/dys195 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Roberts RO, Geda YE, Knopman DS et al. The mayo clinic study of aging: design and sampling, participation, baseline measures and sample characteristics. Neuroepidemiology. 2008;30:58–69. doi:10.1159/000115751 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Hollman JH, McDade EM, Petersen RC. Normative spatiotemporal gait parameters in older adults. Gait Posture. 2011;34:111–118. doi:10.1016/j.gaitpost.2011.03.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Klunk WE, Engler H, Nordberg A et al. Imaging brain amyloid in Alzheimer’s disease with Pittsburgh Compound-B. Ann Neurol. 2004;55:306–319. doi:10.1002/ana.20009 [DOI] [PubMed] [Google Scholar]
- 15. Jack CR Jr, Wiste HJ, Lesnick TG et al. Brain β-amyloid load approaches a plateau. Neurology. 2013;80:890–896. doi:10.1212/WNL.0b013e3182840bbe [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Jack CR Jr, Knopman DS, Weigand SD et al. An operational approach to national institute on aging-alzheimer’s association criteria for preclinical Alzheimer disease. Ann Neurol. 2012;71:765–775. doi:10.1002/ana.22628 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Senjem ML, Gunter JL, Shiung MM, Petersen RC, Jack CR Jr. Comparison of different methodological implementations of voxel-based morphometry in neurodegenerative disease. Neuroimage. 2005;26:600–608. doi:10.1016/j.neuroimage.2005.02.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Meltzer CC, Leal JP, Mayberg HS, Wagner HN Jr, Frost JJ. Correction of PET data for partial volume effects in human cerebral cortex by MR imaging. J Comput Assist Tomogr. 1990;14:561–570. [DOI] [PubMed] [Google Scholar]
- 19. Ashburner J, Friston KJ. Unified segmentation. Neuroimage. 2005;26:839–851. doi:10.1016/j.neuroimage.2005.02.018 [DOI] [PubMed] [Google Scholar]
- 20. Jovicich J, Czanner S, Greve D et al. Reliability in multi-site structural MRI studies: effects of gradient non-linearity correction on phantom and human data. Neuroimage. 2006;30:436–443. doi:10.1016/j.neuroimage.2005.09.046 [DOI] [PubMed] [Google Scholar]
- 21. Sled JG, Zijdenbos AP, Evans AC. A nonparametric method for automatic correction of intensity nonuniformity in MRI data. IEEE Trans Med Imaging. 1998;17:87–97. doi:10.1109/42.668698 [DOI] [PubMed] [Google Scholar]
- 22. Fischl B. FreeSurfer. Neuroimage. 2012;62:774–781. doi:10.1016/j.neuroimage.2012.01.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. FreeSurfer. How to Convert from Freesurfer Space Back to Native Anatomical Space. 2015. Retrieved from: http://surfer.nmr.mgh.harvard.edu/fswiki/FsAnat-to-NativeAnat. [Google Scholar]
- 24. Beck AT, Epstein N, Brown G, Steer RA. An inventory for measuring clinical anxiety: psychometric properties. J Consult Clin Psychol. 1988;56:893–897. [DOI] [PubMed] [Google Scholar]
- 25. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–383. [DOI] [PubMed] [Google Scholar]
- 26. Rocca WA, Yawn BP, St Sauver JL, Grossardt BR, Melton LJ 3rd. History of the rochester epidemiology project: half a century of medical records linkage in a US population. Mayo Clin Proc. 2012;87:1202–1213. doi:10.1016/j.mayocp.2012.08.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Ivnik RJ, Malec JF, Smith GE et al. Mayo’s older americans normative studies: WAIS-R norms for ages 56 to 97. Clinical Neuropsychologist. 1992;6:1–30. [Google Scholar]
- 28. American Psychiatric Association. Diagnostic and Statistical manual of mental disorders (DSM-IV). Washington, DC: American Psychiatric Association; 1994. [Google Scholar]
- 29. Bauer DJ, Preacher KJ, Gil KM. Conceptualizing and testing random indirect effects and moderated mediation in multilevel models: new procedures and recommendations. Psychol Methods. 2006;11:142–163. doi:10.1037/1082-989X.11.2.142 [DOI] [PubMed] [Google Scholar]
- 30. Mielke MM, Roberts RO, Savica R et al. Assessing the temporal relationship between cognition and gait: slow gait predicts cognitive decline in the Mayo Clinic Study of Aging. J Gerontol A Biol Sci Med Sci. 2013;68:929–937. doi:10.1093/gerona/gls256 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Frimenko R, Goodyear C, Bruening D. Interactions of sex and aging on spatiotemporal metrics in non-pathological gait: a descriptive meta-analysis. Physiotherapy. 2015;101:266–272. doi:10.1016/j.physio.2015.01.003 [DOI] [PubMed] [Google Scholar]
- 32. Barnes LL, Wilson RS, Bienias JL, Schneider JA, Evans DA, Bennett DA. Sex differences in the clinical manifestations of Alzheimer disease pathology. Arch Gen Psychiatry. 2005;62:685–691. doi:10.1001/archpsyc.62.6.685 [DOI] [PubMed] [Google Scholar]
- 33. Nadkarni NK, Lopez OL, Perera S et al. Cerebral amyloid deposition and dual-tasking in cognitively normal, mobility unimpaired older adults. J Gerontol A Biol Sci Med Sci. 2017;72:431–437. doi:10.1093/gerona/glw211 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Elobeid A, Libard S, Leino M, Popova SN, Alafuzoff I. Altered proteins in the aging brain. J Neuropathol Exp Neurol. 2016;75:316–325. doi:10.1093/jnen/nlw002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Jack CR Jr, Wiste HJ, Weigand SD et al. Age-specific and sex-specific prevalence of cerebral β-amyloidosis, tauopathy, and neurodegeneration in cognitively unimpaired individuals aged 50-95 years: a cross-sectional study. Lancet Neurol. 2017;16:435–444. doi:10.1016/S1474-4422(17)30077-7 [DOI] [PMC free article] [PubMed] [Google Scholar]