Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jun 15.
Published in final edited form as: Exp Gerontol. 2022 Mar 24;163:111773. doi: 10.1016/j.exger.2022.111773

Gait Subgroups Among Older Adults with Chronic Pain Differ in Cerebellum and Basal Ganglia Gray Matter Volumes

Ania L Lipat 1, David J Clark 2,3, Chris J Hass 1, Yenisel Cruz-Almeida 4,5
PMCID: PMC9948689  NIHMSID: NIHMS1870206  PMID: 35341939

Abstract

Background:

Current literature regarding morphological gray matter atrophy in chronic pain is mixed, inhibiting our ability to understand neurological mechanisms of chronic pain. The inconsistent findings may be due to the presence of subgroups within the older adult chronic pain population that differ in gait performance, as gait and gray matter have been previously associated. These gait subgroups, however, have been inadequately characterized in prior work and have not been compared across gray matter measures. Therefore, the purpose of this study was to identify and characterize gait subgroups within the older adult chronic pain population, and to evaluate differences in gray matter measures between subgroups.

Methods:

The present study was a secondary analysis of the Neuromodulatory Examination of Pain and Mobility Across the Lifespan (NEPAL) study. A subset of older participants (n = 40) completed assessments to evaluate psychological status, cognitive abilities, pain characteristics, and spatiotemporal gait performance using an instrumented gait mat. Gray matter measures were obtained from a T1-weighted anatomical scan using Freesurfer’s recon-all function.

Results:

After data reduction, a hierarchical cluster analysis identified three gait clusters: A Normal Gait cluster (n = 12), a Shuffle Gait cluster (n = 15), and an Unsteady Gait cluster (n = 13). Clusters differed in gait velocity, stride length, step width, double support percentages, and stride length variability. The Shuffle Gait cluster exhibited reduced gray matter volumes in the cerebellum, caudate, putamen, and pallidum, as well as a worse pain severity when compared to the Normal Gait cluster (p<0.05). The Shuffle Gait cluster also had less gray matter in the cerebellum and caudate when compared to the Unsteady Gait cluster (p<0.05).

Conclusions:

Our results confirm the existence of gait subgroups among the older adult chronic pain population and gray matter differences observed between groups support the need for the consideration of subgroups within this population for future pain, mobility, and aging studies.

Keywords: Chronic pain, gray matter, spatiotemporal gait, older adults

1. Introduction

Chronic pain, which is traditionally defined as “prolonged and persistent pain of at least three months in duration” (Gatchel et al., 2007), affects approximately one-fifth of adults in the United States, with prevalence rates highest among older adults (Dahlhamer et al., 2018). In addition to being a widespread condition, chronic pain is one of the leading causes of disability in the world (Blyth & Noguchi, 2017), and is associated with declines in mobility, cognition, and quality of life (Breivik et al., 2006; Eggermont et al., 2009; Moriarty et al., 2011). Given the extent of disability associated with chronic pain, research across multiple disciplines have sought to determine underlying mechanisms driving this condition. Brain imaging studies, in particular, have been useful in identifying neural structures involved in pain processing. According to these studies, individuals with chronic pain exhibit less gray matter in areas such as the anterior cingulate cortex, orbitofrontal cortex, insula, and basal ganglia compared to healthy controls (Baliki et al., 2011, Cauda et al., 2014, Yuan et al., 2017).

While gray matter atrophy is a consistent finding among these studies, there are many inconsistencies regarding which areas of the brain undergo atrophy in individuals with chronic pain. For instance, both Fritz et al. (2016) and Apkarian et al. (2004) found individuals with chronic low back pain had less gray matter in the prefrontal cortex when compared to controls, however, Fritz et al. found an additional group difference in the insula, whereas Apkarian et al., found an additional group difference in the thalamus. Differences such as these within the literature make it difficult to conclusively determine the role of gray matter and the impacted structures in chronic pain. As such, there exists a need to determine the cause of the variable results in the literature.

An understudied factor that may be driving the variable results in the literature could be the presence of distinct subgroups within the chronic pain population. Previous studies have identified chronic pain subgroups that exhibit unique pain sensitivities (Cardoso et al., 2016; Frey-Law et al., 2016), psychological and cognitive profiles (Cruz-Almeida et al., 2013, 2020), and physical performance abilities (Cruz-Almeida et al., 2017). Differences in gray matter measures between these subgroups remain unknown, yet differences are likely to exist given the role of the brain in regulating these processes. These differences could offer a potential explanation for the inconsistencies in gray matter atrophy patterns in chronic pain within the literature. Moreover, studying these differences could clarify the role of structures involved in the development of chronic pain and its associated conditions. For instance, gray matter comparisons between the previously identified physical performance subgroups could identify which structures solely contribute to chronic pain, solely contribute impaired mobility among individuals with chronic pain, or contribute to both conditions.

Evaluating gray matter differences between the physical performance subgroups, however, would be challenging as the assessment used to identify the subgroups, the Short Physical Performance Battery (SPPB), assessed multiple domains of lower extremity function (i.e., balance, gait, and chair stands). Identifying chronic pain subgroups using a single domain from the SPPB, however, would allow for less ambiguous interpretation of gray matter differences between groups. Of the three domains assessed in the SPPB, gait would be the most promising domain to further investigate as gait scores differed between groups in the original study (Cruz-Almeida et al., 2017a), and gait impairments have been associated with both chronic pain and reductions in gray matter (Cruz-Almeida et al., 2017b). Indeed, individuals with chronic pain have been shown to walk with slower gait speeds, longer stride lengths, wider step widths, longer double support times, and less variability compared to controls (Al-Zahrani & Bakheit, 2002; de Kruijf et al., 2015; Hicks et al., 2017), and these impairments have been associated with reduced gray matter in areas such the cerebellum, basal ganglia, and precentral gyrus in healthy older adults (Beauchet et al., 2014; Callisaya et al., 2014; Nadkarni et al., 2014; Rosano et al., 2007). To our knowledge, however, no study has investigated the relationship between gait performance and gray matter measures among older adults with chronic pain. Therefore, the purpose of this study was to identify and characterize gait subgroups within the older adult chronic pain population, and to determine how they differ in cortical and subcortical gray matter measures. We hypothesize that we will identify three gait subgroups that differ in gait impairment severity, and that groups with more severe gait impairment will also demonstrate more extensive gray matter atrophy in motor control regions of the brain.

2. Methods

2.1. Participants

This study was a secondary analysis of participants enrolled in the Neuromodulatory Examination of Pain and Mobility Across the Lifespan (NEPAL) study at the University of Florida. Participants included in the present study were all over the age of 60 (i.e., older adults) with self-reported chronic pain. Exclusion criteria included: 1) Neurological disease (e.g. Alzheimer’s disease, Parkinson’s disease); 2) serious psychiatric conditions (e.g., schizophrenia, major depression, bipolar disorder), 3) uncontrolled hypertension (blood pressure >150/95 mm Hg), heart failure, or history of acute myocardial infarction; 4) systemic rheumatic disorders (e.g., rheumatoid arthritis, systemic lupus erythematosus, fibromyalgia); 5) chronic opioid use; 6) magnetic resonance imaging (MRI) contraindications; 7) excessive anxiety regarding protocol procedures; 8) hospitalization within the preceding year for psychiatric illness; 9) HIV or AIDS; and 10) cognitive impairment (Modified Mini-Mental State Examination (3MS) score < 77 (Teng et al. 1987). Written consent was obtained for all participants. All protocols were reviewed and approved by the University of Florida’s Institutional Review Board.

2.2. Data Acquisition

2.2.1. Demographic, Psychological, Cognitive, and Pain Assessments

To identify potential confounding variables, several demographic, psychological, and cognitive data were collected. Demographic data (i.e., age, gender, race, education level) were collected via questionnaires. The Positive and Negative Affect Scale (PANAS) was used to assess positive and negative affect, with higher scores representing higher levels of each affect (Watson et al., 1988). The State-Trait Anxiety Inventory (STAI) assessed anxiety level through a 20-item questionnaire, with higher values corresponding with higher anxiety (Julian, 2011). Scores for the PANAS and STAI assessments were obtained for both state (i.e., due to a specific situation and trait (i.e., due to an individual’s personality) affect and anxiety. Participants additionally completed the Center for Epidemiologic Studies Depression Scale (CES-D) to determine frequency of depressive symptoms during the past week (Radloff, 1977), and the Montreal Cognitive Assessment (MoCA) to assess global cognitive function. Higher scores on the MoCA corresponded with better cognitive performance (Nasreddine et al., 2005).

Clinical pain severity was assessed using the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and the revised Short-Form McGill Pain Questionnaire (SF-MPQ-2) for each participant. WOMAC pain scores range from 0-20, with a higher score indicating a greater pain interference with daily activities, including walking, climbing stairs, and standing (Bellamy et al., 1988). The SF-MPQ-2 assess continuous, intermittent, neuropathic, and affective pain severity in the past week, and provides a total pain score that ranges from 0-10, with higher scores corresponding with worse pain severity (Lovejoy et al., 2012). Number of painful sites was also determined by having participants indicate anatomical locations where they felt pain (e.g., head, neck, shoulders, chest, stomach, upper and lower back, arms, hands, legs and feet) on a body manikin. The total number of sites indicated by the participants was used to represent this measure.

2.2.2. Spatiotemporal Gait Analysis

Spatiotemporal gait parameters were measured using the GAITRite Mat (Cir Systems Inc., Franklin, NJ, USA). Participants completed four walking trials in which they were instructed to walk at their normal, comfortable speed across the length of the walkway (7 meters). To ensure only steady state walking data were collected, participants began each walking trial 1.5 meters before the instrumented gait mat and were instructed to stop walking 1.5 meters after the gait mat ended. Data from the left and right legs were averaged across the three trials to calculate average gait velocity (m/s), stride length (m), step width (cm), double support % (% cycle), and stride length variability (standard deviation).

2.2.3. Neuroimaging

A T1-weighted anatomical scan was collected using a Phillips (Best, the Netherlands) 3-Tesla scanner using a 32-channel-radio-frequency coil at the University of Florida’s McKnight Brain Institute on the Advanced Magnetic Resonance Imaging and Spectroscopy facility. The following parameters were used for the scan: TR = 7.0 ms, TE = 3.2 ms, 170 slices acquired in a sagittal orientation, flip angle = 8 degrees, 1 resolution = 1 mm3. Cortical thickness (mm) measures for the left and right rostral middle frontal cortex, lateral orbitofrontal cortex, medial orbitofrontal cortex, rostral anterior cingulate cortex, caudal anterior cingulate, precentral gyrus, postcentral gyrus, and insula, as well as gray matter volumes (cm3) for the cerebellum, thalamus, caudate, putamen, pallidum, and accumbens area, and total estimated intracranial volume were obtained using Freesurfer’s recon all-function (Version 6.0.0, Boston, MA, USA). Cortical thickness measures were then averaged across hemispheres and subcortical volumes were summed across hemisphere to obtain a single value for each measure.

2.3. Data Reduction

Individuals with gait impairment were identified using statistical methods (i.e., principal component and cluster analyses) previously employed by studies that have identified homogeneous subgroups among individuals with chronic pain (Cardoso et al., 2016; Cruz-Almeida, Cardoso, et al., 2017b; Cruz-Almeida et al., 2013; Frey-Law et al., 2016). Principal component analysis (PCA) was used to reduce the number of variables that accounted for the maximum amount of variance, thus reducing the number of comparisons. Spatiotemporal gait parameters were reduced to principal components and both oblique and orthogonal rotations were examined and compared to ensure agreement of primary loadings for individual items. Once the factors were determined, gait scores were created from the individual measures by computing Z-scores for each gait parameter and averaging the Z-scores from measures that comprised the factors. All Z-scores were determined so that positive scores corresponded to larger gait values and negative scores corresponded to smaller gait values. Principal components or factors with eigenvalues greater than 1 were retained for interpretation.

The factor scores were then entered into a hierarchical cluster analysis using Ward’s clustering method with squared Euclidean distances. The optimal number of clusters was determined by examining the agglomeration coefficients for the point at which the percent change was largest between clusters and through analysis of the dendogram. A multivariate analysis of variance (MANOVA) was used to assess the internal validity of the final cluster solution on the spatiotemporal variables, and follow-up pairwise comparisons with Bonferroni confidence interval adjustments were used to determine how clusters differed.

2.4. Statistical Analysis

χ2 analyses were used to determine if identified clusters significantly differed in gender, race, and education (p < .05). A one-way analysis of variance (ANOVA) was used to compare age and WOMAC pain scores between clusters. Two multivariate analyses of variance (MANOVA) were used to compared psychological and cognitive assessment scores and pain measures between clusters. Two multivariate analyses of covariance (MANCOVA) were used to identify differences in regional cortical thicknesses and subcortical brain structure volumes. Estimated total intracranial volume was entered as a covariate for both MANCOVAs. For analyses with significant (p < .05) omnibus effects (Wilk’s lambda), follow-up univariate tests and pairwise comparisons with a Bonferroni confidence interval adjustment were used to identify significant differences between clusters (p < .05). For the cortical and subcortical comparisons, the Benjamini-Hochberg procedure was additionally performed at a false discovery rate less than 0.05. Lastly, partial correlations, controlling for estimated total intracranial volume, were run to assess associations between pain measures and gray matter measures. The Benjamini-Hochberg procedure was again performed at a false discovery rate less than 0.05 to correct for multiple comparisons in the partial correlations analysis.

3. Results

3.1. Cluster Analysis of Gait Scores

Examination of the agglomeration coefficients and the dendogram supported a three-cluster solution. Internal validation of the final three cluster solution detected a significant multivariate omnibus effect (F (5,33) = 8.950, p < 0.001), in addition to significant follow-up univariate tests (p < 0.05) for all gait variables. Pairwise comparisons indicated that Cluster 2 exhibited a slower velocity (p < 0.001), shorter stride length (p < 0.001), and larger double support percentage (p < 0.001) than Cluster 1. Cluster 2 additionally demonstrated a slower velocity (p = 0.002) and shorter stride length (p < 0.001) in comparison to Cluster 3. Pairwise comparisons further revealed that the Cluster 3 had a slower velocity (p = 0.011), wider step width (p < 0.001), and greater stride length variability (p < 0.0028) than Cluster 1, and a wider step width than Cluster 2 (p = 0.001). Results are summarized in Table 1. Based off these gait differences, the clusters were henceforth referred to as the Normal Gait Cluster (Cluster 1), the Shuffle Gait Cluster (Cluster 2), and the Unsteady Gait Cluster (Cluster 3).

Table 1.

Spatiotemporal gait measures across clusters (mean ± standard error)

Normal Gait Shuffle Gait Unsteady Gait p
n 13 15 12
Velocity (m/s) 1.30 ± 0.03 1.02 ± 0.03* 1.17 ± 0.03* < 0.001
Stride Length (m) 1.35 ± 0.03 1.11 ± 0.02** 1.29 ± 0.03 < 0.001
Step Width (cm) 8.11 ± 0.64 9.30 ± 0.59 12.76 ± 0.66** < 0.001
Double Support (% Gait Cycle) 25.88 ± 1.00 32.50 ± 0.93* 28.74 ± 1.04 < 0.001
Stride Length Variability (m) 0.024 ± 0.003 0.030 ± 0.003 0.035 ± 0.003* 0.029
Open in a new tab

Notes:

* =

significantly differs (p < .05) from Normal Gait cluster

** =

significantly differs (p < .05) from other two clusters

3.2. Differences in Demographics, Psychology, Cognition, and Pain Across Clusters

The χ2 analyses comparing sex, race, education level between groups found no differences for any variable (sex: (χ2 (2df, n = 40) = 4.697, p = 0.095), race: (χ 2(4df, n = 40) = 3.110, p = 0.540), education: (χ 2(8df, n = 40) = 9.514, p = 0.301)). The one-way ANOVA comparing age between groups did not reach statistical significance (F (2,37) = 1.454, p = 0.247), nor did the MANOVA comparing psychological and cognitive scores (F (16,60) = 1.010, p = 0.459), indicating clusters did not significantly differ across these variables. The MANOVA comparing pain measures between clusters did reach statistical significance (F (6,70) = 2.481, p = 0.031). A significant univariate effect was detected for WOMAC-Pain scores (F (2,40) = 4.323, p = 0.021), with post hoc analyses indicating that the Shuffle Gait Cluster had significantly higher WOMAC-Pain scores than the Normal Gait Cluster (p = 0.005). Results are summarized in Table 2.

Table 2.

Demographic, psychological, cognitive, and pain questionnaire measures across clusters

Normal Gait Shuffle Gait Unsteady Gait p
Age 70 ± 2 71 ± 2 75 ± 2 0.247
Gender (# females) 9 14 7 0.095
Race 0.427
 # Caucasian 11 14 12
 # Black 1 1 0
 # Other 1 0 0
Highest Education Level 0.310
 # High School 3 6 5
 # Associates degree 3 4 0
 # Bachelor’s Degree 3 3 1
 # Master’s Degree 3 2 3
 # Doctoral Degree 1 0 3
PANAS Positive Affect (State) 31 ± 6 38 ± 6 24 ± 7 0.299
PANAS Negative Affect (State) 11 ± 1 11 ± 1 10 ± 1 0.312
PANAS Positive Affect (Trait) 28 ± 4 32 ± 3 38 ± 4 0.173
PANAS Negative Affect (Trait) 9 ± 1 11 ± 1 11 ± 1 0.581
STAI (State) 29 ± 2 27 ± 2 20 ± 2 0.012
STAI (Trait) 22 ± 3 26 ± 3 27 ± 3 0.513
CES-D 8 ± 1 9 ± 1 6 ± 1 0.427
MoCA 27 ± 1 26 ± 1 26 ± 1 0.455
WOMAC-Pain 4 ± 1 7 ± 1* 4 ± 1 0.021
SF-MPQ-2 1 ± 0 1 ± 0 1 ± 0 0.566
# of Anatomical Pain Sites 5 ± 1 4 ± 1 3 ± 1 0.414
Open in a new tab

Notes: Age and assessment scores are reported as mean ± standard error.

* =

significantly differs (p < .05) from Normal Gait cluster.

3.3. Differences in Gray Matter Measures Across Clusters

The MANCOVA comparing regional cortical thicknesses between clusters failed to detect a significant multivariate omnibus effect (F (16,58) = 0.667, p = 0.814) (Table 3). The multivariate omnibus effect for the MANCOVA comparing subcortical brain structure volumes, however, reached significance (F (12,62) = 2.301, p = 0.017). A significant univariate effect was detected for cerebellum volume (F (2,40) = 6.227, p = 0.005), caudate volume (F (2,40) = 9.445, p < 0.001), putamen volume (F (2,40) = 5.065, p = 0.012), and accumbens area volume (F (2,40) = 3.357, p = 0.046). Follow-up pairwise comparisons indicated that the Shuffle Gait Cluster had a significantly smaller cerebellum volume (p = 0.007), caudate volume (p < 0.001), putamen volume (p = 0.010), and accumbens area volume (p = 0.042) when compared to the Normal Gait Cluster, and a significantly smaller cerebellum volume (p = 0.038) and caudate volume (p = 0.012) when compared to the Unsteady Gait Cluster (Table 4). After performing the Benjamini-Hochberg correction, cerebellum (p = 0.033) and caudate (p = 0.007) volumes remained significantly different between clusters.

Table 3.

Cortical thickness (mm) across clusters (mean ± standard error)

Normal Gait Shuffle Gait Unsteady Gait p-values Benjamini-Hochberg
p-values
Rostral Middle Frontal Cortex 2.28 ± 0.03 2.31 ± 0.03 2.31 ± 0.03 0.688 0.985
Lateral Orbitofrontal Cortex 2.59 ± 0.04 2.62 ± 0.03 2.57 ± 0.04 0.671 0.985
Medial Orbitofrontal Cortex 2.39 ± 0.04 2.40 ± 0.03 2.38 ± 0.04 0.919 0.990
Rostral Anterior Cingulate Cortex 2.76 ± 0.05 2.80 ± 0.05 2.78 ± 0.05 0.844 0.985
Caudal Anterior Cingulate Cortex 2.47 ± 0.06 2.65 ± 0.06 2.64 ± 0.06 0.055 0.129
Precentral Gyrus 2.52 ± 0.04 2.50 ± 0.03 2.52 ± 0.04 0.837 0.985
Postcentral Gyrus 2.01 ± 0.03 2.01 ± 0.03 2.01 ± 0.03 0.995 0.995
Insula 2.89 ± 0.04 2.91 ± 0.04 2.87 ± 0.04 0.727 0.985
Open in a new tab

Table 4.

Subcortical volumes (cm3) across clusters (mean ± standard error)

Normal Gait Shuffle Gait Unsteady Gait p-values Benjamini-Hochberg
p-values
Estimated Total Intracranial Volume 1401.40 ± 64.96 1444.70 ± 60.48 1409.00 ± 67.61 - -
Cerebellum 101.96 ± 2.38 91.21 ± 2.22** 99.94 ± 2.48 0.005 0.033
Thalamus 12.88 ± 0.34 12.01 ± 0.32 11.98 ± 0.35 0.117 0.234
Caudate 6.83 ± 0.19 5.78 ± 0.17** 6.58 ± 0.19 < 0.001 0.007
Putamen 8.76 ± 0.19 7.92 ± 0.18* 8.42 ± 0.20 0.012 0.054
Pallidum 3.75 ± 0.11 3.39 ± 0.10 3.45 ± 0.11 0.050 0.130
Accumbens Area 0.92 ± 0.03 0.81 ± 0.03* 0.86 ± 0.03 0.046 0.130
Open in a new tab

Notes:

* =

significantly differs (p < .05) from Normal Gait cluster

** =

significantly differs (p < .05) from Normal Gait cluster and Unsteady Gait cluster

3.4. Associations between Pain Measures and Gray Matter Measures

SF-MPQ-2 total pain scores were significantly associated with accumbens area volume (r = 0.317, p = 0.049). WOMAC-Pain scores and number of anatomical painful sites were not significantly associated with any gray matter measure (p > 0.05). After correcting for multiple comparisons, no pain measures were significantly associated with any gray matter measure (p > 0.004).

4. Discussion

This study sought to identify gait subgroups within the older adult chronic pain population and to evaluate cortical and subcortical gray matter differences between subgroups. A cluster analysis based on spatiotemporal gait parameters (i.e., velocity, stride length, step width, double support percentage, and stride length variability) identified three distinct gait subgroups with unique gait profiles. The clusters were then renamed to reflect their respective gait patterns. For instance, the Normal Gait Cluster was renamed so as it exhibited similar gait performance across the included measures when compared to literature on normative spatiotemporal gait parameters in older adults (Hollman et al., 2011). Moreover, the Shuffle Gait Cluster and the Unsteady Gait Cluster exhibited worse gait performance when compared to the Normal Gait Cluster, which suggests that individuals comprising the Normal Gait Cluster demonstrated unimpaired gait. The Shuffle Gait Cluster exhibited characteristics associated with a shuffle gait pattern (Snijders et al., 2007), such as reduced walking speed, shorter stride length, and a larger double support percentages when compared to the Normal Gait Cluster. Finally, the Unsteady Gait Cluster was renamed in this manner as it exhibited a wider step width and greater stride length variability compared to the Normal Gait Cluster, and these gait patterns have been shown to be a compensatory mechanism for balance deficits (Nonnekes et al., 2018) and a predictor of fall risk in older adults (Hausdorff et al., 2001; Verghese et al., 2009), respectively. The gait characteristics exhibited by Shuffle Gait and Unsteady Gait Clusters have been previously observed in individuals knee osteoarthritis and chronic low back pain (Al-Zahrani & Bakheit, 2002; Barzilay et al., 2016; Hamacher et al., 2014; Hicks et al., 2017), however, this study is the first to find that these gait characteristics are specific to distinct subgroups within the older chronic pain population.

To our knowledge, our study is also the first to investigate differences in cortical and subcortical gray matter between older adults with chronic pain that vary in gait performance. Through our gray matter comparisons, we found that the Shuffle Gait Cluster had reduced gray matter in the cerebellum and basal ganglia when compared to the Normal Gait and Unsteady Gait Clusters. The roles of the cerebellum and basal ganglia in motor control are well established (Takakusaki, 2017), therefore, morphological changes within each structure could contribute to the observed gait deficits in the Shuffle Gait Cluster. Moreover, studies in healthy older adults have previously demonstrated a relationship between gray matter declines in the cerebellum and basal ganglia and the gait deficits observed in the Shuffle Gait Cluster, with less gray matter in each structure corresponding to slower gait, shorter strides, and longer double support (Callisaya et al., 2014; Nadkarni et al., 2014; Rosano et al., 2007). Interestingly, both the cerebellum and basal ganglia also play a role in pain processing (Borsook et al., 2010; Coombes & Misra, 2016), and experience gray matter declines in chronic pain (As-Sanie et al., 2012; Moayedi et al., 2012; Obermann et al., 2013). As such, the Shuffle Gait Cluster may have experienced unique gray matter declines in these structures, leading to the development of these gait impairments, whereas the other chronic pain gait clusters did not, and consequently did not develop these gait impairments. The cause of these gray matter declines (e.g., age, chronic pain), however, remains unknown.

The Unsteady Gait Cluster, however, also exhibited gait impairment when compared to the Normal Gait cluster but did not differ across any of these gray matter measures. The lack of detected differences in gray matter measures between these clusters may be attributed to the limited cortical and subcortical structures included in our analysis, due to the preliminary nature of our investigation and smaller sample size. Comparing additional gray matter measures between the gait clusters may reveal gray matter differences between the Unsteady Gait and Normal Gait Clusters. For instance, gray matter declines in the parietal lobe, which was not included in the present analysis, has been associated with increased step width and larger stride time variability (Beauchet et al., 2014; Rosano et al., 2008), both of which were exhibited by the Unsteady Gait Cluster. Alternatively, the lack of gray matter differences between the Unsteady Gait and Normal Gait Clusters could also indicate that gray matter atrophy does not contribute to the gait impairments observed in the Unsteady Gait Cluster. Further investigation, however, is required to elucidate factors contributing to gait impairment in this cluster.

The differences in gray matter measures that were found between clusters offers a potential explanation for the inconsistencies in the literature regarding gray matter atrophy in chronic pain, particularly since gray matter atrophy in the cerebellum and basal ganglia are among these inconsistencies (Apkarian, 2004; As-Sanie et al., 2012; Fritz et al., 2016; Liao et al., 2018; Moayedi et al., 2012; Obermann et al., 2013). The discrepancies between study findings may be attributed to a disproportionate inclusion of chronic pain participants that belong to each of the gait subgroups. For instance, studies that have found chronic pain-related gray matter atrophy in the cerebellum and basal ganglia may have included many participants that could have been grouped into the Shuffle Gait Cluster. Conversely, studies that did not find chronic pain-related gray matter atrophy in these structures may have included participants that would have normal or unsteady gait patterns.

In addition to the gray matter differences, our study also found differences in WOMAC pain scores between clusters, with the Shuffle Gait Cluster reporting a higher WOMAC pain score than the Normal Gait Cluster. The higher pain scores may provide additional explanation for the gait impairments exhibited by the Shuffle Gait Cluster, particularly since the WOMAC quantifies pain intensity during movement (e.g., walking, stair climbing), and prior work has demonstrated a relationship between pain severity and gait (Barzilay et al., 2016; Elbaz et al., 2014). Findings from the present study, however, are the first to indicate that pain severity may not solely be responsible for gait impairments in this population, and that changes in brain morphology may also be a factor.

Despite identifying differences in both pain severity and gray matter measures between clusters, pain severity was not significantly associated with any gray matter measure. These findings were similarly observed by Fritz et al. (2016), and suggest that other chronic pain characteristics may contribute to gray matter atrophy, which, in turn, may contribute to gait impairments. Longer pain duration, for instance, has been associated with greater gray matter atrophy (Apkarian, 2004) in chronic pain, but was not assessed in the present study due to missing data. Future work should consider including additional pain measures, such as pain duration, to determine their roles in gray matter atrophy and gait in chronic pain.

Although this study provides promising initial evidence of brain morphological differences between gait subgroups in older persons with chronic pain, future work in larger samples should expand comparisons to additional brain regions. Additional comparisons will allow for a more complete understanding of the physiological mechanisms underpinning gait performance in these groups and may provide further explanation for the inconsistencies surrounding gray matter atrophy in chronic pain. Future work should also consider comparing gray matter measures between other previously identified subgroups within the chronic pain population and should investigate differences in other neurological factors affected by chronic pain (e.g., white matter, functional connectivity) between subgroups. Finally, the present cross-sectional study cannot determine any causal or temporal inferences, and whether pain leads to brain or gait changes or vice-versa. Future studies with larger sample sizes using longitudinal study designs are necessary to provide temporally-ordered evidence.

5. Conclusion

Three distinct gait subgroups with varying gait performances (i.e., normal gait, shuffle gait, and unsteady gait) exist within the older adult chronic pain population. These subgroups also exhibit differences in subcortical gray matter volumes, indicating a potential neural mechanism may drive gait impairment in this population. The presence of these subgroups and the gray matter differences between them also provide a potential explanation for inconsistencies in the literature regarding gray matter atrophy in chronic pain. Future work should consider the presence of these subgroups in relation to brain structure and function to gain a more complete understanding of the complex interactions between the brain and gait patterns in persons with chronic pain.

Table 5.

Associations with pain measures and gray matter measures.

WOMAC-Pain # Anatomical Pain Sites SF-MPQ-2
r p-values Benjamini-Hochberg
p-values 		r p-values Benjamini-Hochberg
p-values 		r p-values Benjamini-Hochberg
p-values
Rostral Middle Frontal Cortex 0.059 0.721 0.889 −0.205 0.210 1.052 −0.086 0.601 0.980
Lateral Orbitofrontal Cortex −0.009 0.958 0.958 −0.256 0.116 1.052 −0.091 0.581 0.851
Medial Orbitofrontal Cortex −0.116 0.482 1.052 0.050 0.763 1.052 −0.113 0.495 0.889
Rostral Anterior Cingulate Cortex −0.054 0.743 0.889 0.097 0.556 0.851 −0.235 0.149 1.052
Caudal Anterior Cingulate Cortex 0.237 0.147 0.851 −0.033 0.840 1.052 0.122 0.459 0.905
Precentral Gyrus −0.266 0.102 0.851 −0.148 0.368 1.052 −0.130 0.429 1.052
Postcentral Gyrus −0.069 0.675 0.889 −0.195 0.234 0.969 −0.002 0.992 0.983
Insula 0.084 0.611 0.889 0.084 0.612 0.958 0.016 0.923 0.889
Cerebellum −0.070 0.672 0.889 −0.058 0.728 0.889 0.045 0.783 0.889
Thalamus 0.099 0.547 1.052 0.175 0.287 1.052 0.139 0.400 1.052
Caudate −0.228 0.162 0.851 −0.278 0.087 0.958 0.010 0.953 0.851
Putamen −0.035 0.830 0.905 0.102 0.535 0.851 0.299 0.064 1.052
Pallidum 0.153 0.354 1.052 0.067 0.684 0.889 0.067 0.687 0.889
Accumbens Area 0.072 0.664 0.889 0.149 0.365 0.851 0.317 0.049 1.052
Open in a new tab

Acknowledgments

The authors are very grateful to our older adult community volunteers for their participation in the NEPAL study. This work was supported by the NIH (NIA grants K01AG048259, R01AG059809, R01AG067757 to YCA), the University of Florida Claude D. Pepper Center (P30AG028740), and the University of Florida McKnight Brain Research Foundation and Center for Cognitive Aging and Memory.

References

  1. Al-Zahrani KS, & Bakheit AMO (2002). A study of the gait characteristics of patients with chronic osteoarthritis of the knee. Disability and Rehabilitation, 24(5), 275–280. 10.1080/09638280110087098 [DOI] [PubMed] [Google Scholar]
  2. Apkarian AV (2004). Chronic Back Pain Is Associated with Decreased Prefrontal and Thalamic Gray Matter Density. Journal of Neuroscience, 24(46), 10410–10415. 10.1523/JNEUROSCI.2541-04.2004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. As-Sanie S, Harris RE, Napadow V, Kim J, Neshewat G, Kairys A, Williams D, Clauw DJ, & Schmidt-Wilcke T (2012). Changes in regional gray matter volume in women with chronic pelvic pain: A voxel-based morphometry study. Pain, 153(5), 1006–1014. 10.1016/j.pain.2012.01.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Baliki MN, Schnitzer TJ, Bauer WR, & Apkarian AV (2011). Brain Morphological Signatures for Chronic Pain. PLoS ONE, 6(10), e26010. 10.1371/journal.pone.0026010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Barzilay Y, Segal G, Lotan R, Regev G, Beer Y, Lonner BS, Mor A, & Elbaz A (2016). Patients with chronic non-specific low back pain who reported reduction in pain and improvement in function also demonstrated an improvement in gait pattern. European Spine Journal, 25(9), 2761–2766. 10.1007/s00586-015-4004-0 [DOI] [PubMed] [Google Scholar]
  6. Beauchet O, Annweiler C, Celle S, Bartha R, Barthélémy J-C, & Roche F (2014). Higher Gait Variability is Associated with Decreased Parietal Gray Matter Volume Among Healthy Older Adults. Brain Topography, 27(2), 293–295. 10.1007/s10548-013-0293-y [DOI] [PubMed] [Google Scholar]
  7. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, & Stitt LW (1988). Validation study of WOMAC: A health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. Journal of Rheumatology, 15(12), 1833–1840. [PubMed] [Google Scholar]
  8. Blyth FM, & Noguchi N (2017). Chronic musculoskeletal pain and its impact on older people. Best Practice & Research Clinical Rheumatology, 31(2), 160–168. 10.1016/j.berh.2017.10.004 [DOI] [PubMed] [Google Scholar]
  9. Borsook D, Upadhyay J, Chudler EH, & Becerra L (2010). A Key Role of the Basal Ganglia in Pain and Analgesia—Insights Gained through Human Functional Imaging. Molecular Pain, 6, 1744-8069-6-27. 10.1186/1744-8069-6-27 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Breivik H, Collett B, Ventafridda V, Cohen R, & Gallacher D (2006). Survey of chronic pain in Europe: Prevalence, impact on daily life, and treatment. European Journal of Pain, 10(4), 287–287. 10.1016/j.ejpain.2005.06.009 [DOI] [PubMed] [Google Scholar]
  11. Callisaya ML, Beare R, Phan TG, Chen J, & Srikanth VK (2014). Global and Regional Associations of Smaller Cerebral Gray and White Matter Volumes with Gait in Older People. PLoS ONE, 9(1), e84909. 10.1371/journal.pone.0084909 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Cardoso JS, Riley JL, Glover T, Sibille KT, Bartley EJ, Goodin BR, Bulls HW, Herbert M, Addison AS, Staud R, Redden DT, Bradley LA, Fillingim RB, & Cruz-Almeida Y (2016). Experimental pain phenotyping in community-dwelling individuals with knee osteoarthritis. Pain, 157(9), 2104–2114. 10.1097/j.pain.0000000000000625 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cauda F, Palermo S, Costa T, Torta R, Duca S, Vercelli U, Geminiani G, & Torta DME (2014). Gray matter alterations in chronic pain: A network-oriented meta-analytic approach. NeuroImage: Clinical, 4, 676–686. 10.1016/j.nicl.2014.04.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Coombes SA, & Misra G (2016). Pain and motor processing in the human cerebellum. Pain, 157(1), 117–127. 10.1097/j.pain.0000000000000337 [DOI] [PubMed] [Google Scholar]
  15. Cruz-Almeida Y, Cardoso J, Riley JL, Goodin B, King CD, Petrov M, Bartley EJ, Sibille KT, Glover TL, Herbert MS, Bulls HW, Addison A, Staud R, Redden D, Bradley LA, & Fillingim RB (2017a). Physical performance and movement-evoked pain profiles in community-dwelling individuals at risk for knee osteoarthritis. Experimental Gerontology, 98, 186–191. 10.1016/j.exger.2017.08.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Cruz-Almeida Y, Cardoso J, Riley JL, Goodin B, King CD, Petrov M, Bartley EJ, Sibille KT, Glover TL, Herbert MS, Bulls HW, Addison A, Staud R, Redden D, Bradley LA, & Fillingim RB (2017b). Physical performance and movement-evoked pain profiles in community-dwelling individuals at risk for knee osteoarthritis. Experimental Gerontology, 98, 186–191. 10.1016/j.exger.2017.08.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Cruz-Almeida Y, Crowley SJ, Tanner J, & Price CC (2020). Pain Severity and Interference in Different Parkinson’s Disease Cognitive Phenotypes. Journal of Pain Research, Volume 13, 3493–3497. 10.2147/JPR.S270669 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Cruz-Almeida Y, King CD, Goodin BR, Sibille KT, Glover TL, Riley JL, Sotolongo A, Herbert MS, Schmidt J, Fessler BJ, Redden DT, Staud R, Bradley LA, & Fillingim RB (2013). Psychological Profiles and Pain Characteristics of Older Adults With Knee Osteoarthritis: Identification of Psychological Profiles in Knee OA Patients. Arthritis Care & Research, 65(11), 1786–1794. 10.1002/acr.22070 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Cruz-Almeida Y, Rosso A, Marcum Z, Harris T, Newman AB, Nevitt M, Satterfield S, Yaffe K, Rosano C, & for the Health ABC Study. (2017). Associations of Musculoskeletal Pain With Mobility in Older Adults: Potential Cerebral Mechanisms. The Journals of Gerontology: Series A, 72(9), 1270–1276. 10.1093/gerona/glx084 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Dahlhamer J, Lucas J, Zelaya C, Nahin R, Mackey S, DeBar L, Kerns R, Von Korff M, Porter L, & Helmick C (2018). Prevalence of Chronic Pain and High-Impact Chronic Pain Among Adults—United States, 2016. MMWR. Morbidity and Mortality Weekly Report, 67(36), 1001–1006. 10.15585/mmwr.mm6736a2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. de Kruijf M, Verlinden VJA, Huygen FJPM, Hofman A, van der Geest JN, Uitterlinden AG, Bierma-Zeinstra SMA, Ikram MA, & van Meurs JBJ (2015). Chronic joint pain in the lower body is associated with gait differences independent from radiographic osteoarthritis. Gait & Posture, 42(3), 354–359. 10.1016/j.gaitpost.2015.06.193 [DOI] [PubMed] [Google Scholar]
  22. Eggermont LHP, Bean JF, Guralnik JM, & Leveille SG (2009). Comparing Pain Severity Versus Pain Location in the MOBILIZE Boston Study: Chronic Pain and Lower Extremity Function. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 64A(7), 763–770. 10.1093/gerona/glp016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Elbaz A, Mor A, Segal G, Debi R, Shazar N, & Herman A (2014). Novel classification of knee osteoarthritis severity based on spatiotemporal gait analysis. Osteoarthritis and Cartilage, 22(3), 457–463. 10.1016/j.joca.2013.12.015 [DOI] [PubMed] [Google Scholar]
  24. Frey-Law LA, Bohr NL, Sluka KA, Herr K, Clark CR, Noiseux NO, Callaghan JJ, Zimmerman MB, & Rakel BA (2016). Pain sensitivity profiles in patients with advanced knee osteoarthritis. Pain, 157(9), 1988–1999. 10.1097/j.pain.0000000000000603 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Fritz H-C, McAuley JH, Wittfeld K, Hegenscheid K, Schmidt CO, Langner S, & Lotze M (2016). Chronic Back Pain Is Associated With Decreased Prefrontal and Anterior Insular Gray Matter: Results From a Population-Based Cohort Study. The Journal of Pain, 17(1), 111–118. 10.1016/j.jpain.2015.10.003 [DOI] [PubMed] [Google Scholar]
  26. Gatchel RJ, Peng YB, Peters ML, Fuchs PN, & Turk DC (2007). The biopsychosocial approach to chronic pain: Scientific advances and future directions. Psychological Bulletin, 133(4), 581–624. 10.1037/0033-2909.133.4.581 [DOI] [PubMed] [Google Scholar]
  27. Hamacher D, Hamacher D, & Schega L (2014). A cognitive dual task affects gait variability in patients suffering from chronic low back pain. Experimental Brain Research, 232(11), 3509–3513. 10.1007/s00221-014-4039-1 [DOI] [PubMed] [Google Scholar]
  28. Hausdorff JM, Rios DA, & Edelberg HK (2001). Gait variability and fall risk in community-living older adults: A 1-year prospective study. Archives of Physical Medicine and Rehabilitation, 82(8), 1050–1056. 10.1053/apmr.2001.24893 [DOI] [PubMed] [Google Scholar]
  29. Hicks GE, Sions JM, Coyle PC, & Pohlig RT (2017). Altered spatiotemporal characteristics of gait in older adults with chronic low back pain. Gait & Posture, 55, 172–176. 10.1016/j.gaitpost.2017.04.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Hollman JH, McDade EM, & Petersen RC (2011). Normative spatiotemporal gait parameters in older adults. Gait & Posture, 34(1), 111–118. 10.1016/j.gaitpost.2011.03.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Julian LJ (2011). Measures of anxiety: State-Trait Anxiety Inventory (STAI), Beck Anxiety Inventory (BAI), and Hospital Anxiety and Depression Scale-Anxiety (HADS-A). Arthritis Care & Research, 63(S11), S467–S472. 10.1002/acr.20561 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Liao X, Mao C, Wang Y, Zhang Q, Cao D, Seminowicz DA, Zhang M, & Yang X (2018). Brain gray matter alterations in Chinese patients with chronic knee osteoarthritis pain based on voxel-based morphometry: Medicine, 97(12), e0145. 10.1097/MD.0000000000010145 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Lovejoy TI, Turk DC, & Morasco BJ (2012). Evaluation of the Psychometric Properties of the Revised Short-Form McGill Pain Questionnaire. The Journal of Pain, 13(12), 1250–1257. 10.1016/j.jpain.2012.09.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Moayedi M, Weissman-Fogel I, Salomons TV, Crawley AP, Goldberg MB, Freeman BV, Tenenbaum HC, & Davis KD (2012). Abnormal gray matter aging in chronic pain patients. Brain Research, 1456, 82–93. 10.1016/j.brainres.2012.03.040 [DOI] [PubMed] [Google Scholar]
  35. Moriarty O, McGuire BE, & Finn DP (2011). The effect of pain on cognitive function: A review of clinical and preclinical research. Progress in Neurobiology, 93(3), 385–404. 10.1016/j.pneurobio.2011.01.002 [DOI] [PubMed] [Google Scholar]
  36. Nadkarni NK, Nunley KA, Aizenstein H, Harris TB, Yaffe K, Satterfield S, Newman AB, Rosano C, & for the Health ABC Study. (2014). Association Between Cerebellar Gray Matter Volumes, Gait Speed, and Information-Processing Ability in Older Adults Enrolled in the Health ABC Study. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 69(8), 996–1003. 10.1093/gerona/glt151 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Nasreddine ZS, Phillips NA, Bédirian V, Charbonneau S, Whitehead V, Collin I, Cummings JL, & Chertkow H (2005). The Montreal Cognitive Assessment, MoCA: A Brief Screening Tool For Mild Cognitive Impairment: MOCA: A BRIEF SCREENING TOOL FOR MCI. Journal of the American Geriatrics Society, 53(4), 695–699. 10.1111/j.1532-5415.2005.53221.x [DOI] [PubMed] [Google Scholar]
  38. Nonnekes J, Goselink RJM, Růžička E, Fasano A, Nutt JG, & Bloem BR (2018). Neurological disorders of gait, balance and posture: A sign-based approach. Nature Reviews Neurology, 14(3), 183–189. 10.1038/nrneurol.2017.178 [DOI] [PubMed] [Google Scholar]
  39. Obermann M, Rodriguez-Raecke R, Naegel S, Holle D, Mueller D, Yoon M-S, Theysohn N, Blex S, Diener H-C, & Katsarava Z (2013). Gray matter volume reduction reflects chronic pain in trigeminal neuralgia. NeuroImage, 74, 352–358. 10.1016/j.neuroimage.2013.02.029 [DOI] [PubMed] [Google Scholar]
  40. Radloff LS (1977). The CES-D Scale: A Self-Report Depression Scale for Research in the General Population. Applied Psychological Measurement, 1(3), 385–401. 10.1177/014662167700100306 [DOI] [Google Scholar]
  41. Rosano C, Aizenstein H, Brach J, Longenberger A, Studenski S, & Newman AB (2008). Special Article: Gait Measures Indicate Underlying Focal Gray Matter Atrophy in the Brain of Older Adults. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 63(12), 1380–1388. 10.1093/gerona/63.12.1380 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Rosano C, Aizenstein HJ, Studenski S, & Newman AB (2007). A Regions-of-Interest Volumetric Analysis of Mobility Limitations in Community-Dwelling Older Adults. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 62(9), 1048–1055. 10.1093/gerona/62.9.1048 [DOI] [PubMed] [Google Scholar]
  43. Snijders AH, van de Warrenburg BP, Giladi N, & Bloem BR (2007). Neurological gait disorders in elderly people: Clinical approach and classification. The Lancet Neurology, 6(1), 63–74. 10.1016/S1474-4422(06)70678-0 [DOI] [PubMed] [Google Scholar]
  44. Takakusaki K (2017). Functional Neuroanatomy for Posture and Gait Control. Journal of Movement Disorders, 10(1), 1–17. 10.14802/jmd.16062 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Verghese J, Holtzer R, Lipton RB, & Wang C (2009). Quantitative Gait Markers and Incident Fall Risk in Older Adults. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 64A(8), 896–901. 10.1093/gerona/glp033 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Watson D, Clark LA, & Tellegen A (1988. 1101). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063. 10.1037/0022-3514.54.6.1063 [DOI] [PubMed] [Google Scholar]
  47. Yuan C, Shi H, Pan P, Dai Z, Zhong J, Ma H, & Sheng L (2017). Gray Matter Abnormalities Associated With Chronic Back Pain: A Meta-Analysis of Voxel-based Morphometric Studies. The Clinical Journal of Pain, 33(11), 983–990. 10.1097/AJP.0000000000000489 [DOI] [PubMed] [Google Scholar]

RESOURCES