Skip to main content
NCBI home page
Brain Connectivity logoLink to Brain Connectivity
. 2016 Dec 1;6(10):795–803. doi: 10.1089/brain.2015.0401

Structural Brain Connectivity and the Sit-to-Stand-to-Sit Performance in Individuals with Nonspecific Low Back Pain: A Diffusion Magnetic Resonance Imaging-Based Network Analysis

Madelon Pijnenburg 1,2, SM Hadi Hosseini 3,*, Simon Brumagne 1, Lotte Janssens 1,4, Nina Goossens 1, Karen Caeyenberghs 5,✉
PMCID: PMC8024240  PMID: 27421840

Abstract

Individuals with nonspecific low back pain (NSLBP) show an impaired sensorimotor control. They need significantly more time to perform five consecutive sit-to-stand-to-sit (STSTS) movements compared with healthy controls. Optimal sensorimotor control depends on the coactivation of many brain regions, which have to operate as a coordinated network to achieve correct motor output. Therefore, the examination of brain connectivity from a network perspective is crucial for understanding the factors that drive sensorimotor control. In the current study, potential alterations in structural brain networks of individuals with NSLBP and the correlation with the performance of the STSTS task were investigated. Seventeen individuals with NSLBP and 17 healthy controls were instructed to perform five consecutive STSTS movements as fast as possible. In addition, data of diffusion magnetic resonance imaging were acquired and analyzed using a graph theoretical approach. Results showed that individuals with NSLBP needed significantly more time to perform the STSTS task compared with healthy controls (p < 0.05). Both groups exhibited small-world properties in their structural networks. However, local efficiency was significantly decreased in the patients with NSLBP compared with controls (p < 0.05, false discovery rate [FDR] corrected). Moreover, global efficiency was significantly correlated with the sensorimotor task performance within the NSLBP group (r = −0.73, p = 0.002). Our data show disrupted network organization of white matter networks in patients with NSLBP, which may contribute to their persistent pain and sensorimotor disabilities.

Keywords: : back pain, brain connectivity, graph theory, sensorimotor system, white matter

Introduction

Nonspecific low back pain (NSLBP) is an important socioeconomic health problem (Hoy et al., 2014). The lifetime prevalence of NSLBP is up to 84% (Airaksinen et al., 2006), and despite many efforts in the development of treatment strategies, up to 70% of individuals with NSLBP experience at least one recurrence (Pengel et al., 2003). Previous studies have shown that individuals with NSLBP exhibit an impaired sensorimotor control (Brumagne et al., 2008; Laird et al., 2014; Ruhe et al., 2011; Wand et al., 2011). For example, they need significantly more time to perform five consecutive sit-to-stand-to-sit (STSTS) movements compared with healthy controls (Claeys et al., 2012). The STSTS task necessitates optimal postural, pelvic, and lumbosacral coordination, which in turn requires efficient sensorimotor control (Hall et al., 2010).

Optimal sensorimotor control depends on the collaboration and coactivation of many brain regions, which need to operate as a coordinated network to achieve optimal motor output (Goodworth and Peterka, 2009). Therefore, the examination of brain connectivity from a network perspective is crucial in understanding the factors that drive sensorimotor control. Recent evidence suggests that NSLBP is associated with disrupted activity in the default mode network (Baliki et al., 2008; Tagliazucchi et al., 2010), with reduced gray matter density in sensorimotor-related areas (Apkarian et al., 2004; Ivo et al., 2013) and with decreased integrity of the left superior cerebellar peduncle (Pijnenburg et al., 2014). However, there is only limited knowledge on the topological organization of structural brain networks in individuals with NSLBP and the association with sensorimotor control.

Graph theoretical analysis is a multivariate mathematical framework for characterizing topological properties of brain networks (Bassett and Bullmore, 2006; Bullmore and Bassett, 2011; Bullmore and Sporns, 2009; Rubinov and Sporns, 2010; Sporns, 2011). This type of analysis moves away from the traditional neuroimaging approach of examining individual components of the brain, such as regions of interest, and instead concentrates on characterizing the regional and global structure of brain networks. In this way, graph theoretical analysis provides an interesting opportunity to demonstrate the efficiency of information transfer between brain regions and subtle neurobiological changes (McIntosh, 2000; Petrella, 2011; Sporns et al., 2004).

Changes in measures of local specialization and global integration have been associated with a wide range of neurological disorders (for an excellent review, see Griffa et al., 2013) such as schizophrenia (van den Heuvel et al., 2010), Alzheimer's disease (He et al., 2008), and multiple sclerosis (He et al., 2009). Furthermore, there is rapidly growing literature relating individual differences in cognitive performance, such as intelligence and memory, with graph theoretical network measures (Li et al., 2009; van den Heuvel et al., 2009). However, there is little evidence regarding brain network organization in individuals with disorders without typical neurological origin, such as NSLBP.

Only, a few studies investigated the association between motor performance and structural network measures (Caeyenberghs et al., 2012; Wang et al., 2013). Caeyenberghs et al. (2012) showed that poorer balance performance in children with traumatic brain injury is associated with decreased connectivity degree (i.e., number of connections) in the structural network. Wang et al. (2013) reported that global and local efficiency of the structural brain networks is significantly higher in world-class gymnasts compared with controls. However, to the authors' best knowledge, no research has been performed on the association between sensorimotor control and graph theoretical network measures in individuals with NSLBP.

In the present study, we aimed to investigate the organization of structural brain networks in individuals with NSLBP. Based on previous studies showing localized alterations in the NSLBP brain (Apkarian et al., 2004; Schmidt-Wilcke et al., 2006), including our previous diffusion magnetic resonance imaging (MRI) study (Pijnenburg et al., 2014), we expected subtle reductions in the graph theoretical network measures, that is, degree and efficiency, at regional level in the NSLBP group compared with healthy controls.

Furthermore, we expected that decreased connectivity degree/efficiency would be associated with poorer sensorimotor performance. We focused on the STSTS task because (1) it has good test–retest reliability (Simmonds et al., 1998; Smeets et al., 2006), (2) it can be easily used in clinical practice to evaluate functionality of individuals with NSLBP over time (Andersson et al., 2010; Smeets et al., 2006), and (3) it is a prototype of a complex sensorimotor task, requiring integration of sensory and motor information in the brain (Hall et al., 2010). We hypothesized that the network measures would be mainly associated with the STSTS performance on unstable support surface because this support surface enforces a reweighting of sensory signals due to the decreased reliability of sensory signals of the ankle region (Kiers et al., 2011).

Materials and Methods

Participants

A total of 34 individuals, aged between 20 and 50 years, were included in this study (Table 1). The participants were categorized into two groups; 17 individuals with NSLBP (11 women and 6 men) and 17 age-matched (±2 years) healthy individuals (12 women and 5 men). The individuals in the NSLBP group experienced at least 6 months of disabling NSLBP [Oswestry Disability Index, version 2 (adapted Dutch version) (ODI-2) >12% (Fairbank and Pynsent, 2000)]. The healthy individuals had no history of NSLBP and reported a score of 0% on the ODI-2. The exclusion criteria for both groups included a history of specific vestibular problems, cardiovascular and/or neurological disorders, neck problems [Neck Disability Index >6% (Vernon and Mior, 1991)], and previous major trauma and/or surgery of the spine or lower limbs. In addition, each individual had to meet the criteria for MRI-related research. Data from the same cohort regarding local analysis of MRI and proprioceptive reweighting capacity were previously reported (Pijnenburg et al., 2014).

Table 1.

Characteristics of the Participants

Characteristic NSLBP group (n = 17) Healthy group (n = 17) p
Age (years) 33.3 ± 7.9 31.8 ± 8.2 NS
Height (cm) 173.0 ± 6.4 169.1 ± 6.4 NS
Weight (kg) 72.7 ± 10.6 64.9 ± 10.2 0.036
BMI (kg/m2) 24.2 ± 2.8 22.6 ± 2.7 NS
ODI-2 20.6 ± 7.6 0 N/A
NSLBP duration (years) 9.8 ± 8.2 0 N/A
NRSback usual 4.5 ± 2.0 0 N/A
NRSback current 2.0 ± 2.0 0 N/A
Open in a new tab

Data are presented as mean ± standard deviation.

BMI, body–mass index; NRSback usual, back pain score on the numerical rating scale (0–10) during the last month; NRSback current, back pain score on the numerical rating scale (0–10) at the moment of testing; significance level (p < 0.05); NSLBP, nonspecific low back pain; ODI-2, Oswestry Disability Index, version 2 (adapted Dutch version); N/A, not applicable.

All procedures followed were in accordance with the ethical standards of the local Ethics Committee of Biomedical Sciences, KU Leuven, Belgium (s53802), and with the principles of the Declaration of Helsinki (1964) and its later amendments. In addition, the study was registered at www.clinicaltrails.gov with identification number NCT01540617. Informed consent was obtained from all patients for being included in the study.

MRI acquisition

MRI scanning was performed on a Philips 3 Tesla Achieva scanner (Philips) equipped with a 32-channel standard head coil. High-resolution, whole-brain, T1-weighted anatomical scans were obtained with a 3D-TFE sequence (voxel size of 0.98 × 0.98 × 1.2 mm3, repetition time of 9.59 msec, echo time of 4.6 msec, a flip angle of 8°, 182 coronal slices, field of view of 250 × 250 × 218 mm3, and a matrix of 256 × 256 mm2). All T1-weighted anatomical scans were checked by a radiologist to ensure that no brain lesions were present in the total study group. Diffusion-weighted single-shot spin-echo echo planar imaging (DTI SE-EPI) of the entire brain and brainstem was acquired using the following parameters: data acquisition matrix of 96 × 96 mm2, a field of view of 200 × 240 mm2, a repetition time of 7600 msec, and an echo time of 65 msec. Diffusion gradients were applied along 60 noncollinear directions at a b-value of 1300 s/mm2. In total, 58 contiguous sagittal slices (slice thickness = 2.5 mm, voxel size = 2.5 × 2.5 × 2.5 mm3) were acquired. In addition, an average of five volumes without diffusion weighting (b = 0 s/mm2) was obtained.

Diffusion MRI preprocessing

ExploreDTI 4.8.3 was used to process the diffusion MRI data (Jones and Leemans, 2011; Leemans et al., 2009; Pijnenburg et al., 2014). The following multistep procedure was used: (1) the quality of the diffusion MRI data was examined (e.g., looping through the diffusion MRI images at different views (the coronal, sagittal, and axial image plane), (2) corrections for subject motion and eddy current-induced geometrical distortions were performed, and (3) diffusion MRI data were coregistered to the Montreal Neurological Institute (MNI) space.

White matter tractography and network construction

Constrained spherical deconvolution (CSD) can offer a reliable reconstruction of fiber orientation within regions of complex fiber architecture (Jeurissen et al., 2011; Tournier et al., 2007, 2008). This method has been proposed as an approach aiming at solving the well-known critical issue of crossing fibers in studies using the tractography method based on diffusion tensor imaging (DTI). CSD-based tractography (Jeurissen et al., 2011) was performed with a uniform seed point resolution of 2 mm3 and an angle threshold of 30°. Data from the CSD tractography were parcellated using the Automated Anatomical Labeling atlas (AAL) (Tzourio-Mazoyer et al., 2002), resulting in 116 brain regions. Each brain region of the AAL template represented a node of the network. Inter-regional connectivity was then examined by determining the number of streamlines between any two nodes. This value became the edge weight in the 116 by 116 connectivity matrix. Furthermore, each individual association matrix was transformed into an unweighted binary adjacency matrix by changing the entry into one when the number of tracts was greater than or equal to three (these entries were called edges) and, respectively, into zero if the number of tracts was less than three fibers to avoid false-positive connections. Thresholding the connectivity matrices at an absolute threshold results in networks with different number of nodes and degrees (connections) that influence network measures and reduce interpretation of between-group results (van Wijk et al., 2010). Therefore, the constructed adjacency matrices were thresholded at a range of network densities for which the networks were not fragmented (each node had at least one connection with another node in the graph) and displayed small-world properties (nonrandom graphs). The network densities ranging from 0.01 to 0.09 fulfilled these criteria. We compared the networks in this density range in steps of 0.01. Comparing network measures at different network densities results in multiple comparisons. Therefore, we quantified a summary metric using area under the curve (AUC) analysis for each network measure (Bernhardt et al., 2011; Hosseini et al., 2012b). By performing AUC analysis, the comparison between network measures is less sensitive to the thresholding process.

Network measures

The Brain Connectivity Toolbox (Rubinov and Sporns, 2010; https://sites.google.com/site/bctnet/) and the Graph Analysis Toolbox (Hosseini et al., 2012a) were used for the graph theoretical analysis. Three small-world metrics (i.e., normalized path length, normalized clustering, and small-worldness), two commonly used measures of local segregation (i.e., local efficiency and clustering coefficient), one measure of global integration (i.e., global efficiency), and two regional measures for each of the selected anatomical AAL labels (i.e., nodal degree and local efficiency) were computed. In addition, network centrality was calculated to define network hubs within the brain of individuals with NSLBP and healthy controls separately. Comprehensive descriptions and equations for these measures can be found in Rubinov and Sporns (2010). In this study, we only provide brief definitions for each of the network properties used in this study.

Global network measures

Small-worldness is a property that facilitates an energy-efficient balance between network segregation and network integration. The normalized clustering coefficient and the characteristic path length are the two key measures of small-worldness (Bassett and Bullmore, 2006; Bullmore and Bassett, 2011; Bullmore and Sporns, 2009; Rubinov and Sporns, 2010; Sporns, 2011). The clustering coefficient of a node is a measure of the number of edges that exist between that node and its nearest neighbor nodes. The normalized clustering coefficient of the global network is equal to the average of clustering coefficients across nodes and is a measure of network segregation, which reflects the ability of a network to process information locally. The characteristic path length is the most commonly used measure of network integration and is defined by the average shortest path length between all pairs of nodes in the network. It characterizes the ability of a network in processing information on the global level. In a small-world network, the normalized clustering coefficient is significantly higher than that of random networks (C/Crandom ratio greater than one), while the characteristic path length is comparable with random networks (L/Lrandom ratio close to one) (Maslov and Sneppen, 2002; Milo et al., 2002), resulting in a small-world index greater than one. In summary, a small-world network reflects architecture with optimal balance between local and global information processing.

Regional network measures

Nodal degree is the degree calculated for each of the selected anatomical AAL labels and is considered as a measure of how interactive the node is within the network. Nodes with a high degree are interacting with many other nodes in the network (Hosseini et al., 2012b, 2013).

Local efficiency of a node characterizes the ability of that node for specialized processing of information at the local level. It measures how well the neighbors of a node communicate if that node is removed (Latora and Marchiori, 2001).

Network hubs

Betweenness centrality of a node is defined as the fraction of all shortest paths in the network that pass through a given node (Rubinov and Sporns, 2010). This measures the importance of nodes to overall network integrity. The nodes with the largest betweenness centrality are considered to be the most important regions in the brain network and are called hubs. Hubs are essential for coordinating brain functions through their connectivity with various brain regions (Cole et al., 2010) and facilitate efficient communication across the network. Specifically, a node is considered to be a hub if its nodal betweenness centrality is two standard deviations higher than the mean network betweenness centrality (Bassett et al., 2008; Hosseini et al., 2012b).

STSTS task

The equipment, paradigm parameters, and dependent variables of the STSTS task were identical to a previous study (Claeys et al., 2012). The participants were instructed to sit barefoot on a stool that was placed on a six-channel force plate (Bertec Corporation) with their arms relaxed along their body and their vision occluded by means of nontransparent goggles. The stool height was adjusted for each participant to assure an angle of 90° in both hips and knees. After 15 sec of usual sitting (no instructions on posture were given), participants were asked to perform five consecutive STSTS movements, with a full range of motion and as fast as possible. An investigator stood near the participant to prevent actual falls. The force plate registered anterior–posterior center of pressure displacements. The center of pressure displacements were sampled at 500 Hz using a Micro1401 data acquisition system and Spike2 software (Cambridge Electronic Design). This protocol was performed both with the feet placed on a stable (force plate itself) and on an unstable support surface (50 cm length × 41 cm width × 6 cm thickness, Airex balance pad elite) placed on the force plate. The total duration of the five consecutive STSTS movements was calculated based on the anterior–posterior center of pressure displacement. The starting and end-points of the task were defined by the mean value of the center of pressure during usual sitting before and after the task.

Statistical analyses

Unpaired t-tests were used to calculate group differences in characteristics of the participants.

Significance of between-group differences in small-world, global, and regional network measures was tested employing the AUC statistics within the Graph Analysis Toolbox (Hosseini et al., 2012a, 2012b). AUC accounts for multiple comparisons because the networks are not compared at different densities (Zhang et al., 2011). The curves extracted from thresholding across a range of densities were used. Each of these curves depicted the changes in a specific network measure (for each group) as a function of network density. To test the significance of the between-group differences in AUC of each network measure, the actual between-group difference in AUC for each network measure was placed in the corresponding permutation distribution and the p-value was calculated based on its percentile position. By performing AUC analysis, the comparison between network measures is less sensitive to the thresholding process. The p-values reported for regional differences between groups are corrected for multiple comparisons using the false discovery rate, with a statistical threshold of p < 0.05.

Differences in STSTS performance were analyzed with a 2 × 2 repeated measures analysis of variance (ANOVA) with group as between-subject factor (NSLBP group and healthy control group) and surface as within-subject factor (stable support surface and unstable support surface). The significance level for group differences in STSTS performance was set at p < 0.05.

Furthermore, the behavioral parameters, that is, the duration of the STSTS task on the stable and unstable support surfaces, were used to assess relationships between sensorimotor control and global network measures (including normalized path length, normalized clustering coefficient, small-worldness, global efficiency, clustering, and local efficiency) using partial correlations (while correcting for age) within each of the groups. The outliers labeling rule (with a g-factor of 2.2) (Hoaglin and Inglewicz, 1987) was applied to detect outliers from both the STSTS performance and the network measures. These values were excluded pairwise from the correlation analysis.

Finally, we conducted region-specific partial correlation analyses (with the effect of age removed) between the regional network metrics (i.e., local efficiency and degree) and STSTS performance within each group. Bonferroni corrections for multiple comparisons were made, hence p < 0.008 was considered significant following correction for the correlation analyses regarding the six global network metrics and p < 0.0004 for the node-specific correlation analyses regarding the 116 ROIs. The statistical analysis was performed with SPSS 19.

Results

Characteristics of the participants

No significant differences in demographic characteristics were found between groups (p > 0.05, as shown in Table 1), except for weight (p = 0.036). The weight of the participants was not significantly correlated with the total time needed to perform the STSTS task on stable and unstable support surfaces and with the global network measures (p > 0.05). Therefore, weight was not included in the subsequent analysis.

Small-world network measures

There were no significant group differences in the key measures of small-worldness between individuals with NSLBP and healthy controls over the range of densities (normalized clustering coefficient: pAUC = 0.26, normalized path length: pAUC = 0.90, small-world index: pAUC = 0.23). In both groups, normalized clustering coefficients were greater than one (random clustering) and normalized characteristic path lengths were close to one (near random path length). In other words, patients with NSLBP displayed small-world values close to the values of the brain networks of healthy controls, suggesting an intact overall organization of the structural brain network in patients with NSLBP (Fig. 1).

FIG. 1.

FIG. 1.

Open in a new tab

Group differences in graph metrics. Patients with NSLBP are indicated in blue, healthy controls are represented in red. NSLBP, nonspecific low back pain. Color images available online at www.liebertpub.com/brain

Global network measures

Compared with healthy controls, the structural network of individuals with NSLBP showed significantly smaller values of local efficiency in the NSLBP network (pAUC = 0.04). A marginal significant effect of group was also seen for the clustering coefficient (pAUC = 0.057). These results indicate that local segregation is affected in NSLBP. We did not find any significant group difference for the measure of global integration (global efficiency, pAUC = 0.71). The absence of this group effect suggests that global connectivity is relatively intact in individuals with NSLBP (Fig. 1).

Regional network measures

Nodal degree and local efficiency of the (sub)cortical regions in the white matter networks were examined between the two groups. We found a significant group difference in degree in the left precentral gyrus after correction for multiple comparisons (false discovery rate, p < 0.05). Compared with controls, patients with NSLBP showed significantly higher connectivity degree in the left precentral gyrus (pcorr = 0.023) (Fig. 1).

Network hubs

To identify the hub regions, we calculated the betweenness centrality for each node of each group. This analysis revealed that both groups exhibited hubs in the left medial part of the superior frontal gyrus, left and right lingual gyrus, the left and right precuneus, and the right putamen. Of note, the right superior frontal gyrus and the left and right Crus I of the cerebellum were only identified as hub nodes in the individuals with NSLBP, and the left and right thalamus were only identified as a hub in the healthy control group. This result is indicative of a difference in hub distribution between both groups (Fig. 2).

FIG. 2.

FIG. 2.

Open in a new tab

Group differences in connectivity degree (left panel). Healthy controls and (right panel) patients with NSLBP. Size of the regions of interest (spheres) represents connectivity degree. Blue sphere indicating higher connectivity degree in left precentral gyrus (FDR corrected) in patients with NSLBP. Moreover, the hubs regions are shown. For the NSLBP group, the identified hubs included right superior frontal gyrus and Cerebellum Crus I regions (indicated in magenta). The unique hub for the control group included the thalamus (indicated in cyan). The common hubs for both groups are indicated in yellow. FDR, false discovery rate. Color images available online at www.liebertpub.com/brain

STSTS task

Repeated measures ANOVA on the duration of the STSTS task revealed a significant main effect of the factors group (F = 11.348, p = 0.002) and surface (F = 6.29, p = 0.017). No interaction effect was present (p > 0.05). The main effect of group showed that individuals with NSLBP needed significantly more time to perform five consecutive STSTS movements on stable and unstable support surfaces (stable: 18.1 ±6.9 sec, unstable: 16.8 ± 5.9 sec) compared with healthy controls (stable: 12.9 ± 2.5 sec, unstable: 11.5 ± 2.0 sec) (p = 0.002). Unexpectedly, the main effect of surface showed a significant decrease in duration of the STSTS task on unstable support surface (NSLBP: 16.8 ± 5.9 sec, healthy controls: 11.5 ± 2.0 sec) compared with the stable condition (NSLBP: 18.1 ± 6.9 sec, healthy: 12.9 ± 2.5 sec), irrespective of group (p = 0.017).

Association between network measures and the STSTS task in individuals with NSLBP and healthy controls

To determine the relationships between sensorimotor control and graph metrics, partial correlation analysis with age as confounding covariate was separately performed for the NSLBP and control groups. In the control group, none of the STSTS task measures showed significant correlations with the network metrics. The following descriptions focus on the results in the NSLBP group. Global efficiency was significantly associated with the duration of five consecutive STSTS movements on unstable support surface (r = −0.73, p = 0.002, Fig. 3) in individuals with NSLBP, indicating that decreased efficiency was associated with a worse performance of the STSTS task on unstable support surface. No significant results were obtained with the other global network metrics. The relationships between regional network metrics and STSTS performance did not survive correction for multiple comparisons.

FIG. 3.

FIG. 3.

Open in a new tab

Plot indicating the relationship (while correcting for age) between global efficiency (AUC) and the duration of five consecutive STSTS movements on unstable support surface. AUC, area under the curve; STSTS, sit-to-stand-to-sit. Color images available online at www.liebertpub.com/brain

Discussion

To the authors' best knowledge, this is the first study investigating alterations in white matter networks of individuals with NSLBP using graph theoretical analysis and the correlation with the performance of a complex sensorimotor task. Individuals with NSLBP needed significantly more time to perform five consecutive STSTS movements on stable and unstable support surfaces compared with healthy controls. Both groups exhibited small-world properties in their structural networks. However, local efficiency was significantly decreased in the patients with NSLBP compared with controls. Moreover, global efficiency was significantly correlated with the sensorimotor task performance within the NSLBP group. Our data show disrupted network organization of white matter networks in patients with NSLBP, which may contribute to their persistent pain and sensorimotor disabilities.

Group differences in structural white matter networks

The structural brain networks of both individuals with NSLBP and healthy controls exhibited a higher normalized cluster coefficient, but a similar characteristic path length compared with random networks. This result is indicative of a comparable small-world architecture of structural brain networks in individuals with NSLBP and healthy individuals. These results suggest an effective information transfer within the brain, with a relatively low wiring cost in both groups (He et al., 2009). This is in accordance with previous studies showing a small-world architecture not only in the brain networks of healthy individuals (Achard et al., 2006; Hagmann et al., 2007; He et al., 2007; Salvador et al., 2005) but also in other chronic pain conditions, such as patients with migraine (Liu et al., 2012) and female patients with irritable bowel syndrome (Labus et al., 2014). Furthermore, the absence of group effects on global efficiency marks the fact that NSLBP does not alter/affect the global organization of the brain networks.

In contrast, statistically significant group differences were found for network indices of local efficiency of information transfer between regions. Structural networks of individuals with NSLBP exhibited decreases in local efficiency and marginal decreases in clustering coefficient, indicating that the effects on local efficiency (and clustering coefficient) are likely to be related to a reduced local connectivity rather than global organizational effects.

With regard to regional network properties, we found an increased connectivity degree in the left precentral gyrus in the NSLBP group compared with healthy controls. Although the primary motor cortex (M1) is not part of the pain matrix (Lee and Tracey, 2013), M1 is considered to play a key role in modulation of pain in different chronic pain conditions (for reviews, see Castillo et al., 2014; Parker et al., 2016). Changes in M1 have also been found in animal models of pain (Lucas et al., 2011). Thus, the present study complements clinical and experimental studies of chronic pain and emphasizes the maladaptive plasticity of the primary motor cortex. The observed increased connectivity degree of the primary motor cortex in individuals with NSLBP may indicate a possible compensatory mechanism for decreased local connectivity (i.e., decreased local efficiency).

Importance of specific brain regions for patients with NSLBP

A difference between individuals with NSLBP and healthy controls was observed in the distribution of hub regions with high betweenness centrality. Compared with healthy controls, the right superior frontal gyrus and bilateral cerebellum Crus I were found to be more critical for controlling information in the individuals with NSLBP. The superior frontal gyrus is attributed to a variety of functions in both cognitive and motor control (Chouinard and Paus, 2006; Li et al., 2013). Contemporary clinical studies suggest that the posterior cerebellum contributes to complex cognitive operations (e.g., Schmahmann et al., 2007). Conversely, in the healthy control group, an additional hub was identified in the left and right thalamus. The thalamus is generally considered as a major relay area in the brain for the transmission of a variety of signals such as sensory, motor, and cognitive signals (Herrero et al., 2002). This difference in hub distribution between the groups indicates a subtle reorganization of the most important regions in the structural network of individuals with NSLBP compared with healthy controls.

Group differences in sensorimotor control

In accordance with a previous study (Claeys et al., 2012), the current group of individuals with NSLBP needed significantly more time to perform five consecutive STSTS movements on both stable and unstable support surfaces compared with the healthy controls. A decreased ability to weight proprioceptive signals adaptively, as part of sensorimotor control, was observed to be responsible for this impaired STSTS performance (Brumagne et al., 2008; Claeys et al., 2012). This impaired proprioceptive weighting capacity is particularly prominent on the unstable support surface because this condition enforces an additional reweighting of proprioceptive signals due to the decreased reliability of proprioceptive signals from the ankle region (Kiers et al., 2011). Our hypothesis is supported by a recent study that demonstrated an association between the proprioceptive reweighting capacity and microstructural integrity of the superior cerebellar peduncle in individuals with NSLBP and healthy controls (Pijnenburg et al., 2014). However, this is the first study that investigated the association between the sensorimotor STSTS task and the topological properties of structural brain networks in individuals with NSLBP. The behavioral relevance of these topological properties is discussed in the next paragraph.

Behavioral relevance of network alterations in individuals with NSLBP

Previously unreported, we have now shown associations between sensorimotor deficits and structural brain organization in patients with NSLBP from a network perspective. To our knowledge, only two studies have examined the relationship between topological parameters and motor functioning (Caeyenberghs et al., 2012; Wang et al., 2013). In the present study, an increased duration of the sensorimotor task (worse performance) on the unstable support surface was associated with decreased global efficiency in individuals with NSLBP. This is consistent with the notion that successful STSTS performance requires integration of sensory and motor information between remote regions in the brain (Hall et al., 2010).

Limitations and conclusions

Limitations of our study include a relatively small sample size. However, our results survived corrections for multiple comparisons and thus can be considered robust. In addition, the DWI acquisition was suboptimal for CSD-based tractography (Tournier et al., 2008) as the diffusion weighting was relatively low (1300 s/mm2). Nevertheless, fiber crossings, present in 60–90% of the white matter voxels (Jeurissen et al., 2013), can reliably be identified with CSD (Jeurissen et al., 2011; Tournier et al., 2007, 2008), and thus, using CSD is highly beneficial in comparison with DTI, where fiber crossings cannot be identified. Finally, the correlation between our observed network alterations based on diffusion MRI and morphological (based on gray matter changes) and functional network (such as resting state) alterations based on other MRI modalities remain to be further investigated.

Taking into account these limitations, the present study is the first report illustrating associations between network metrics and sensorimotor control in individuals with NSLBP. Although small-world properties were present for both NSLBP and control networks, local segregation was significantly altered in individuals with NSLBP. Specifically, individuals with NSLBP showed lower values of local efficiency. At the nodal level, we observed increased connectivity degree of the primary motor cortex in individuals with NSLBP, representing an adaptive mechanism for decreased local connectivity. Furthermore, decreased global efficiency was correlated with a longer duration to perform five consecutive STSTS movements in the NSLBP group. Thus, our results suggest an altered local connectivity in the large-scale brain systems in patients with NSLBP and provide structural insights into the chronic pain connectome.

Acknowledgments

This work was supported by the Agency for Innovation by Science and Technology–Flanders (IWT) (PhD fellowships M.P. and N.G., grant number 141037) (www.iwt.be) and The Research Foundation–Flanders (FWO) (postdoctoral fellowship L.J., grant number: 12M9815N). The funders had no role in study design; collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication. The authors are grateful to the study volunteers for their participation and to Alexander Leemans for his help with ExploreDTI.

Author Disclosure Statement

No competing financial interests exist.

References

  1. Achard S, Salvador R, Whitcher B, Suckling J, Bullmore E. 2006. A resilient, low-frequency, small-world human brain functional network with highly connected association cortical hubs. J Neurosci 26:63–72 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Airaksinen O, Brox JI, Cedraschi C, Hildebrandt J, Klaber-Moffett J, Kovacs F, et al. 2006. Chapter 4. European guidelines for the management of chronic nonspecific low back pain. Eur Spine J 15 Suppl 2:S192–S300 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Andersson EI, Lin CC, Smeets RJ. 2010. Performance tests in people with chronic low back pain: responsiveness and minimal clinically important change. Spine (Phila Pa 1976) 35:E1559–E1563 [DOI] [PubMed] [Google Scholar]
  4. Apkarian AV, Sosa Y, Sonty S, Levy RM, Harden RN, Parrish TB, Gitelman DR. 2004. Chronic back pain is associated with decreased prefrontal and thalamic gray matter density. J Neurosci 24:10410–10415 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Baliki MN, Geha PY, Apkarian AV, Chialvo DR. 2008. Beyond feeling: chronic pain hurts the brain, disrupting the default-mode network dynamics. J Neurosci 28:1398–1403 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bassett DS, Bullmore E. 2006. Small-world brain networks. Neuroscientist 12:512–523 [DOI] [PubMed] [Google Scholar]
  7. Bassett DS, Bullmore E, Verchinski BA, Mattay VS, Weinberger DR, Meyer-Lindenberg A. 2008. Hierarchical organization of human cortical networks in health and schizophrenia. J Neurosci 28:9239–9248 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bernhardt BC, Chen Z, He Y, Evans AC, Bernasconi N. 2011. Graph-theoretical analysis reveals disrupted small-world organization of cortical thickness correlation networks in temporal lobe epilepsy. Cereb Cortex 21:2147–2157 [DOI] [PubMed] [Google Scholar]
  9. Brumagne S, Janssens L, Knapen S, Claeys K, Suuden-Johanson E. 2008. Persons with recurrent low back pain exhibit a rigid postural control strategy. Eur Spine J 17:1177–1184 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bullmore E, Sporns O. 2009. Complex brain networks: graph theoretical analysis of structural and functional systems. Nat Rev Neurosci 10:186–198 [DOI] [PubMed] [Google Scholar]
  11. Bullmore ET, Bassett DS. 2011. Brain graphs: graphical models of the human brain connectome. Annu Rev Clin Psychol 7:113–140 [DOI] [PubMed] [Google Scholar]
  12. Caeyenberghs K, Leemans A, De Decker C, Heitger M, Drijkoningen D, Linden CV, et al. 2012. Brain connectivity and postural control in young traumatic brain injury patients: a diffusion MRI based network analysis. Neuroimage Clin 1:106–115 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Castillo Saavedra L, Mendonca M, Fregni F. 2014. Role of the primary motor cortex in the maintenance and treatment of pain in fibromyalgia. Med Hypotheses 83:332–336 [DOI] [PubMed] [Google Scholar]
  14. Chouinard PA, Paus T. 2006. The primary motor and premotor areas of the human cerebral cortex. Neuroscientist 12:143–152 [DOI] [PubMed] [Google Scholar]
  15. Claeys K, Dankaerts W, Janssens L, Brumagne S. 2012. Altered preparatory pelvic control during the sit-to-stance-to-sit movement in people with non-specific low back pain. J Electromyogr Kinesiol 22:821–828 [DOI] [PubMed] [Google Scholar]
  16. Cole MW, Pathak S, Schneider W. 2010. Identifying the brain's most globally connected regions. Neuroimage 49:3132–3148 [DOI] [PubMed] [Google Scholar]
  17. Fairbank JC, Pynsent PB. 2000. The Oswestry Disability Index. Spine (Phila Pa 1976) 25:2940–2952 [DOI] [PubMed] [Google Scholar]
  18. Goodworth AD, Peterka RJ. 2009. Contribution of sensorimotor integration to spinal stabilization in humans. J Neurophysiol 102:496–512 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Griffa A, Baumann PS, Thiran JP, Hagmann P. 2013. Structural connectomics in brain diseases. Neuroimage 80:515–526 [DOI] [PubMed] [Google Scholar]
  20. Hagmann P, Kurant M, Gigandet X, Thiran P, Wedeen VJ, Meuli R, Thiran JP. 2007. Mapping human whole-brain structural networks with diffusion MRI. PLoS One 2:e597. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hall LM, Brauer S, Horak F, Hodges PW. 2010. Adaptive changes in anticipatory postural adjustments with novel and familiar postural supports. J Neurophysiol 103:968–976 [DOI] [PubMed] [Google Scholar]
  22. He Y, Chen Z, Evans A. 2008. Structural insights into aberrant topological patterns of large-scale cortical networks in Alzheimer's disease. J Neurosci 28:4756–4766 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. He Y, Chen ZJ, Evans AC. 2007. Small-world anatomical networks in the human brain revealed by cortical thickness from MRI. Cereb Cortex 17:2407–2419 [DOI] [PubMed] [Google Scholar]
  24. He Y, Dagher A, Chen Z, Charil A, Zijdenbos A, Worsley K, Evans A. 2009. Impaired small-world efficiency in structural cortical networks in multiple sclerosis associated with white matter lesion load. Brain 132:3366–3379 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Herrero MT, Barcia C, Navarro JM. 2002. Functional anatomy of thalamus and basal ganglia. Childs Nerv Syst 18:386–404 [DOI] [PubMed] [Google Scholar]
  26. Hoaglin DC, Inglewicz B. 1987. Fine tuning some resistant rules for outlier labeling. J Am Stat Assoc 82:1147–1149 [Google Scholar]
  27. Hosseini SM, Black JM, Soriano T, Bugescu N, Martinez R, Raman MM, et al. 2013. Topological properties of large-scale structural brain networks in children with familial risk for reading difficulties. Neuroimage 71:260–274 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hosseini SM, Hoeft F, Kesler SR. 2012a. GAT: a graph-theoretical analysis toolbox for analyzing between-group differences in large-scale structural and functional brain networks. PLoS One 7:e40709. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Hosseini SM, Koovakkattu D, Kesler SR. 2012b. Altered small-world properties of gray matter networks in breast cancer. BMC Neurol 12:28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Hoy D, March L, Brooks P, Blyth F, Woolf A, Bain C, et al. 2014. The global burden of low back pain: estimates from the Global Burden of Disease 2010 study. Ann Rheum Dis 73:968–974 [DOI] [PubMed] [Google Scholar]
  31. Ivo R, Nicklas A, Dargel J, Sobottke R, Delank KS, Eysel P, Weber B. 2013. Brain structural and psychometric alterations in chronic low back pain. Eur Spine J 22:1958–1964 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Jeurissen B, Leemans A, Jones DK, Tournier JD, Sijbers J. 2011. Probabilistic fiber tracking using the residual bootstrap with constrained spherical deconvolution. Hum Brain Mapp 32:461–479 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Jeurissen B, Leemans A, Tournier JD, Jones DK, Sijbers J. 2013. Investigating the prevalence of complex fiber configurations in white matter tissue with diffusion magnetic resonance imaging. Hum Brain Mapp 34:2747–2766 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Jones DK, Leemans A. 2011. Diffusion tensor imaging. Methods Mol Biol 711:127–144 [DOI] [PubMed] [Google Scholar]
  35. Kiers H, Brumagne S, van Dieen J, van der Wees P, Vanhees L. 2011. Ankle proprioception is not targeted by exercises on an unstable surface. Eur J Appl Physiol 112:1577–1585 [DOI] [PubMed] [Google Scholar]
  36. Labus JS, Dinov ID, Jiang Z, Ashe-McNalley C, Zamanyan A, Shi Y, et al. 2014. Irritable bowel syndrome in female patients is associated with alterations in structural brain networks. Pain 155:137–149 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Laird RA, Gilbert J, Kent P, Keating JL. 2014. Comparing lumbo-pelvic kinematics in people with and without back pain: a systematic review and meta-analysis. BMC Musculoskelet Disord 15:229. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Latora V, Marchiori M. 2001. Efficient behavior of small-world networks. Phys Rev Lett 87:198701. [DOI] [PubMed] [Google Scholar]
  39. Lee MC, Tracey I. 2013. Imaging pain: a potent means for investigating pain mechanisms in patients. Br J Anaesth 111:64–72 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Leemans A, Jeurissen B, Sijbers J, Jones DK. 2009. ExploreDTI: a graphical toolbox for processing, analyzing, and visualizing diffusion MR data. In: 17th Annual Meeting of Intl Soc Mag Reson Med, Hawaii, USA, p. 3537 [Google Scholar]
  41. Li W, Qin W, Liu H, Fan L, Wang J, Jiang T, Yu C. 2013. Subregions of the human superior frontal gyrus and their connections. Neuroimage 78:46–58 [DOI] [PubMed] [Google Scholar]
  42. Li Y, Liu Y, Li J, Qin W, Li K, Yu C, Jiang T. 2009. Brain anatomical network and intelligence. PLoS Comput Biol 5:e1000395. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Liu J, Zhao L, Li G, Xiong S, Nan J, Li J, et al. 2012. Hierarchical alteration of brain structural and functional networks in female migraine sufferers. PLoS One 7:e51250. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Lucas JM, Ji Y, Masri R. 2011. Motor cortex stimulation reduces hyperalgesia in an animal model of central pain. Pain 152:1398–1407 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Maslov S, Sneppen K. 2002. Specificity and stability in topology of protein networks. Science 296:910–913 [DOI] [PubMed] [Google Scholar]
  46. McIntosh AR. 2000. Towards a network theory of cognition. Neural Netw 13:861–870 [DOI] [PubMed] [Google Scholar]
  47. Milo R, Shen-Orr S, Itzkovitz S, Kashtan N, Chklovskii D, Alon U. 2002. Network motifs: simple building blocks of complex networks. Science 298:824–827 [DOI] [PubMed] [Google Scholar]
  48. Parker RS, Lewis GN, Rice DA, McNair PJ. 2016. Is motor cortical excitability altered in people with chronic pain? A systematic review and meta-analysis. Brain Stimul 9:488–500 [DOI] [PubMed] [Google Scholar]
  49. Pengel LH, Herbert RD, Maher CG, Refshauge KM. 2003. Acute low back pain: systematic review of its prognosis. BMJ 327:323. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Petrella JR. 2011. Use of graph theory to evaluate brain networks: a clinical tool for a small world? Radiology 259:317–320 [DOI] [PubMed] [Google Scholar]
  51. Pijnenburg M, Caeyenberghs K, Janssens L, Goossens N, Swinnen SP, Sunaert S, Brumagne S. 2014. Microstructural integrity of the superior cerebellar peduncle is associated with an impaired proprioceptive weighting capacity in individuals with non-specific low back pain. PLoS One 9:e100666. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Rubinov M, Sporns O. 2010. Complex network measures of brain connectivity: uses and interpretations. Neuroimage 52:1059–1069 [DOI] [PubMed] [Google Scholar]
  53. Ruhe A, Fejer R, Walker B. 2011. Center of pressure excursion as a measure of balance performance in patients with non-specific low back pain compared to healthy controls: a systematic review of the literature. Eur Spine J 20:358–368 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Salvador R, Suckling J, Coleman MR, Pickard JD, Menon D, Bullmore E. 2005. Neurophysiological architecture of functional magnetic resonance images of human brain. Cereb Cortex 15:1332–1342 [DOI] [PubMed] [Google Scholar]
  55. Schmahmann JD, Weilburg JB, Sherman JC. 2007. The neuropsychiatry of the cerebellum—insights from the clinic. Cerebellum 6:254–267 [DOI] [PubMed] [Google Scholar]
  56. Schmidt-Wilcke T, Leinisch E, Ganssbauer S, Draganski B, Bogdahn U, Altmeppen J, May A. 2006. Affective components and intensity of pain correlate with structural differences in gray matter in chronic back pain patients. Pain 125:89–97 [DOI] [PubMed] [Google Scholar]
  57. Simmonds MJ, Olson SL, Jones S, Hussein T, Lee CE, Novy D, Radwan H. 1998. Psychometric characteristics and clinical usefulness of physical performance tests in patients with low back pain. Spine (Phila Pa 1976) 23:2412–2421 [DOI] [PubMed] [Google Scholar]
  58. Smeets RJ, Hijdra HJ, Kester AD, Hitters MW, Knottnerus JA. 2006. The usability of six physical performance tasks in a rehabilitation population with chronic low back pain. Clin Rehabil 20:989–997 [DOI] [PubMed] [Google Scholar]
  59. Sporns O. 2011. The human connectome: a complex network. Ann N Y Acad Sci 1224:109–125 [DOI] [PubMed] [Google Scholar]
  60. Sporns O, Chialvo DR, Kaiser M, Hilgetag CC. 2004. Organization, development and function of complex brain networks. Trends Cogn Sci 8:418–425 [DOI] [PubMed] [Google Scholar]
  61. Tagliazucchi E, Balenzuela P, Fraiman D, Chialvo DR. 2010. Brain resting state is disrupted in chronic back pain patients. Neurosci Lett 485:26–31 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Tzourio-Mazoyer N, Landeau B, Papathanassiou D, Crivello F, Etard O, Delcroix N, et al. 2002. Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. Neuroimage 15:273–289 [DOI] [PubMed] [Google Scholar]
  63. Tournier JD, Calamante F, Connelly A. 2007. Robust determination of the fibre orientation distribution in diffusion MRI: non-negativity constrained super-resolved spherical deconvolution. Neuroimage 35:1459–1472 [DOI] [PubMed] [Google Scholar]
  64. Tournier JD, Yeh CH, Calamante F, Cho KH, Connelly A, Lin CP. 2008. Resolving crossing fibres using constrained spherical deconvolution: validation using diffusion-weighted imaging phantom data. Neuroimage 42:617–625 [DOI] [PubMed] [Google Scholar]
  65. van den Heuvel MP, Mandl RC, Stam CJ, Kahn RS, Hulshoff Pol HE. 2010. Aberrant frontal and temporal complex network structure in schizophrenia: a graph theoretical analysis. J Neurosci 30:15915–15926 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. van den Heuvel MP, Stam CJ, Kahn RS, Hulshoff Pol HE. 2009. Efficiency of functional brain networks and intellectual performance. J Neurosci 29:7619–7624 [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. van Wijk BC, Stam CJ, Daffertshofer A. 2010. Comparing brain networks of different size and connectivity density using graph theory. PLoS One 5:e13701. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Vernon H, Mior S. 1991. The Neck Disability Index: a study of reliability and validity. J Manipulative Physiol Ther 14:409–415 [PubMed] [Google Scholar]
  69. Wand BM, Parkitny L, O'Connell NE, Luomajoki H, McAuley JH, Thacker M, Moseley GL. 2011. Cortical changes in chronic low back pain: current state of the art and implications for clinical practice. Man Ther 16:15–20 [DOI] [PubMed] [Google Scholar]
  70. Wang B, Fan Y, Lu M, Li S, Song Z, Peng X, et al. 2013. Brain anatomical networks in world class gymnasts: a DTI tractography study. Neuroimage 65:476–487 [DOI] [PubMed] [Google Scholar]
  71. Zhang J, Wang J, Wu Q, Kuang W, Huang X, He Y, Gong Q. 2011. Disrupted brain connectivity networks in drug-naive, first-episode major depressive disorder. Biol Psychiatry 70:334–342 [DOI] [PubMed] [Google Scholar]

Articles from Brain Connectivity are provided here courtesy of Mary Ann Liebert, Inc.

RESOURCES