Abstract
We mapped alterations of the functional structure of the cerebral cortex using a novel imaging approach in a sample of 160 obsessive–compulsive disorder (OCD) patients. Whole-brain functional connectivity maps were generated using multidistance measures of intracortical neural activity coupling defined within isodistant local areas. OCD patients demonstrated neural activity desynchronization within the orbitofrontal cortex and in primary somatosensory, auditory, visual, gustatory, and olfactory areas. Symptom severity was significantly associated with the degree of functional structure alteration in OCD-relevant brain regions. By means of a novel imaging perspective, we once again identified brain alterations in the orbitofrontal cortex, involving areas purportedly implicated in the pathophysiology of OCD. However, our results also indicated that weaker intracortical activity coupling is also present in each primary sensory area. On the basis of previous neurophysiological studies, such cortical activity desynchronization may best be interpreted as reflecting deficient inhibitory neuron activity and altered sensory filtering.
Introduction
Patients with obsessive compulsive disorder (OCD) experience distressing intrusive thoughts, images or urges combined with uncontrollable and repetitive actions. OCD symptoms may therefore express deficiencies in the inhibition of neural activity. Based on their prominent role in the control of behavior, frontostriatal brain circuits have been centrally implicated in the pathophysiology of OCD and, indeed, much research has demonstrated alterations at this level (Whiteside et al. 2004; Yücel et al. 2007; Menzies et al. 2008; Harrison et al. 2009; Fineberg et al. 2010; Radua et al. 2010; de Wit et al. 2014; Wood and Ahmari 2015; Frydman et al. 2016; van den Heuvel et al. 2016; Boedhoe et al. 2017). However, it has also become clear that brain alterations in OCD are not limited to the frontostriatal circuits, but extend to various other brain structures (Pujol et al. 2004; Menzies et al. 2008; Milad and Rauch 2012; Piras et al. 2015; Wood and Ahmari 2015; Frydman et al. 2016; Boedhoe et al. 2017, 2018). Different research methods have been used to characterize such alterations in OCD patients, each providing distinct perspectives regarding anatomical, metabolic, and functional correlates. In the present study, we examined the brain of OCD patients with a novel imaging approach designed to map alterations of the functional structure of the cerebral cortex at rest.
Essentially, we expanded well-established MRI measures of local functional connectivity in the cerebral cortex by considering the rich spatial information of cortical connections. For instance, existing measures consider the functional connectivity of a region with its local neighborhood according to a predefined radius in millimeters (local connectivity degree (Sepulcre et al. 2010)), a fixed number of adjacent voxels (regional homogeneity (Zang et al. 2004)) or the number of voxels above a defined connectivity threshold (local functional connectivity density (Tomasi and Volkow 2010)). Instead, our method targets local functional connectivity at varying distances by computing Iso-Distant Average Correlation (IDAC) measures. IDAC measures therefore represent the average functional MRI temporal correlation of a given brain unit, or voxel, with other units situated at increasingly separated isodistant intervals. In contrast to the other approaches, IDAC thus captures the graded change in local functional connectivity, which can be sampled at different isodistant intervals to jointly provide a more comprehensive depiction of the spatial structure of cortical connections. We have shown that resulting multidistance brain maps discriminate well between major classical anatomofunctional cortical areas (Macià et al. 2018).
On the basis of previous studies indicating multilevel functional connectivity alterations in OCD (e.g., global graph theory metrics (Göttlich et al. 2014; Reess et al. 2016; Moreira et al. 2017), network integration (Harrison et al. 2009; Hou et al. 2013; Jung et al. 2017), regional homogeneity (Yang et al. 2015; Chen et al. 2016; Niu et al. 2017), voxel connectivity degree (Beucke et al. 2013; Anticevic et al. 2014; Hou et al. 2014), and regional amplitude of low-frequency fluctuations (Giménez et al. 2017; Qiu et al. 2017; Zhao et al. 2017)), we predicted that the functional structure of the cerebral cortex would be anomalous in OCD, reflected in significant differences in IDAC measures. The whole cerebral cortex was mapped using this new imaging approach in a large sample of OCD patients and a comparative control subject group.
Methods and Materials
Participants
Overall, 160 adult OCD outpatients were recruited from the Obsessive–Compulsive Disorders Unit of the University Hospital of Bellvitge (Barcelona). Patients were selected after having satisfied DSM-IV diagnostic criteria for OCD (for at least 1 year before the study), in the absence of relevant medical, neurologic, and other major psychiatric illness, as well as imaging data quality control checks (see below). Comorbidity with depression or anxiety was not considered an exclusion criterion provided that OCD was the primary diagnosis and the reason for seeking medical assistance. Diagnosis was confirmed by 2 senior psychiatrists through separate interviews 1 month apart, using the Structured Clinical Interview for DSM-IV Axis I Disorders (First et al. 1998).
The Yale–Brown Obsessive–Compulsive Scale (Y-BOCS) (Goodman et al. 1989) total global severity score was used to measure overall illness severity. The validated Spanish version of the Dimensional Y-BOCS (Rosario-Campos et al. 2006; Pertusa et al. 2011) was used to rate the severity of major obsessive–compulsive symptom dimensions: Contamination and Cleaning, Aggressive and Checking, Symmetry and Ordering, Sexual and Religious, and Hoarding dimensions (Dimensional Y-BOCS data were available for 110 patients). Comorbid depression and anxiety symptoms were measured using Hamilton Depression Rating Scale (HDRS) and Hamilton Anxiety Rating Scale. See Table 1 for further sample descriptions.
Sample characteristics
| OCD patients (n = 160) | Control subjects (n = 121) | |
|---|---|---|
| Age, years, mean (SD), range | 35.41 (9.7), 18–58 | 34.6 (10.2), 19–61 |
| Sex, M:F, No. | 86/74 | 66/55 |
| Handedness, right:left, No. | 146:14 | 111:10 |
| Education, years, mean (SD), range | 12.1 (3.1), 4–19 | 13.2 (3.6), 4–22 |
| Age at onset of OCD, years, mean (SD), range | 21.2 (8.3), 5–41 | – |
| Duration of illness, years, mean (SD), range | 11.7 (9.43), 1–45 | – |
| Y-BOCS total, mean (SD), range | 22.4 (6.3), 10–36 | – |
| HDRS, mean (SD), rangea | 10.1 (5.4), 0–28 | – |
| HARS, mean (SD), rangea | 12.7 (6.7), 0–35 | – |
| Medication at study time, No. (%) | ||
| Medication free (>4 wk) | 2 (1.25) | |
| SSRIs | 110 (68.75) | – |
| Clomipramine | 40 (25) | – |
| SSRI or Clomipramine + Antipsychotic augment. | 8 (5) | – |
| DY-BOCS severity, mean (SD), range (N = 110) | ||
| Contamination and cleaning | 4.1 (4.6), 0–14 | |
| Aggressive and checking | 5.7 (4.6), 0–15 | |
| Symmetry and ordering | 3.9 (4.9), 0–15 | |
| Sexual and religious | 2.0 (3.8), 0–14 | |
| Hoarding | 1.8 (3.2), 0–13 |
| OCD patients (n = 160) | Control subjects (n = 121) | |
|---|---|---|
| Age, years, mean (SD), range | 35.41 (9.7), 18–58 | 34.6 (10.2), 19–61 |
| Sex, M:F, No. | 86/74 | 66/55 |
| Handedness, right:left, No. | 146:14 | 111:10 |
| Education, years, mean (SD), range | 12.1 (3.1), 4–19 | 13.2 (3.6), 4–22 |
| Age at onset of OCD, years, mean (SD), range | 21.2 (8.3), 5–41 | – |
| Duration of illness, years, mean (SD), range | 11.7 (9.43), 1–45 | – |
| Y-BOCS total, mean (SD), range | 22.4 (6.3), 10–36 | – |
| HDRS, mean (SD), rangea | 10.1 (5.4), 0–28 | – |
| HARS, mean (SD), rangea | 12.7 (6.7), 0–35 | – |
| Medication at study time, No. (%) | ||
| Medication free (>4 wk) | 2 (1.25) | |
| SSRIs | 110 (68.75) | – |
| Clomipramine | 40 (25) | – |
| SSRI or Clomipramine + Antipsychotic augment. | 8 (5) | – |
| DY-BOCS severity, mean (SD), range (N = 110) | ||
| Contamination and cleaning | 4.1 (4.6), 0–14 | |
| Aggressive and checking | 5.7 (4.6), 0–15 | |
| Symmetry and ordering | 3.9 (4.9), 0–15 | |
| Sexual and religious | 2.0 (3.8), 0–14 | |
| Hoarding | 1.8 (3.2), 0–13 |
Abbreviations: HDRS, Hamilton Rating Scale for Anxiety; HARS, Hamilton Rating Scale for Depression; OCD, obsessive–compulsive disorder; SD, standard deviation; SSRI, selective serotonin reuptake inhibitors; Y-BOCS, Yale–Brown Obsessive–Compulsive Scale; DY-BOCS, Dimensional Y-BOCS. aMissing data for 2 OCD patients.
Sample characteristics
| OCD patients (n = 160) | Control subjects (n = 121) | |
|---|---|---|
| Age, years, mean (SD), range | 35.41 (9.7), 18–58 | 34.6 (10.2), 19–61 |
| Sex, M:F, No. | 86/74 | 66/55 |
| Handedness, right:left, No. | 146:14 | 111:10 |
| Education, years, mean (SD), range | 12.1 (3.1), 4–19 | 13.2 (3.6), 4–22 |
| Age at onset of OCD, years, mean (SD), range | 21.2 (8.3), 5–41 | – |
| Duration of illness, years, mean (SD), range | 11.7 (9.43), 1–45 | – |
| Y-BOCS total, mean (SD), range | 22.4 (6.3), 10–36 | – |
| HDRS, mean (SD), rangea | 10.1 (5.4), 0–28 | – |
| HARS, mean (SD), rangea | 12.7 (6.7), 0–35 | – |
| Medication at study time, No. (%) | ||
| Medication free (>4 wk) | 2 (1.25) | |
| SSRIs | 110 (68.75) | – |
| Clomipramine | 40 (25) | – |
| SSRI or Clomipramine + Antipsychotic augment. | 8 (5) | – |
| DY-BOCS severity, mean (SD), range (N = 110) | ||
| Contamination and cleaning | 4.1 (4.6), 0–14 | |
| Aggressive and checking | 5.7 (4.6), 0–15 | |
| Symmetry and ordering | 3.9 (4.9), 0–15 | |
| Sexual and religious | 2.0 (3.8), 0–14 | |
| Hoarding | 1.8 (3.2), 0–13 |
| OCD patients (n = 160) | Control subjects (n = 121) | |
|---|---|---|
| Age, years, mean (SD), range | 35.41 (9.7), 18–58 | 34.6 (10.2), 19–61 |
| Sex, M:F, No. | 86/74 | 66/55 |
| Handedness, right:left, No. | 146:14 | 111:10 |
| Education, years, mean (SD), range | 12.1 (3.1), 4–19 | 13.2 (3.6), 4–22 |
| Age at onset of OCD, years, mean (SD), range | 21.2 (8.3), 5–41 | – |
| Duration of illness, years, mean (SD), range | 11.7 (9.43), 1–45 | – |
| Y-BOCS total, mean (SD), range | 22.4 (6.3), 10–36 | – |
| HDRS, mean (SD), rangea | 10.1 (5.4), 0–28 | – |
| HARS, mean (SD), rangea | 12.7 (6.7), 0–35 | – |
| Medication at study time, No. (%) | ||
| Medication free (>4 wk) | 2 (1.25) | |
| SSRIs | 110 (68.75) | – |
| Clomipramine | 40 (25) | – |
| SSRI or Clomipramine + Antipsychotic augment. | 8 (5) | – |
| DY-BOCS severity, mean (SD), range (N = 110) | ||
| Contamination and cleaning | 4.1 (4.6), 0–14 | |
| Aggressive and checking | 5.7 (4.6), 0–15 | |
| Symmetry and ordering | 3.9 (4.9), 0–15 | |
| Sexual and religious | 2.0 (3.8), 0–14 | |
| Hoarding | 1.8 (3.2), 0–13 |
Abbreviations: HDRS, Hamilton Rating Scale for Anxiety; HARS, Hamilton Rating Scale for Depression; OCD, obsessive–compulsive disorder; SD, standard deviation; SSRI, selective serotonin reuptake inhibitors; Y-BOCS, Yale–Brown Obsessive–Compulsive Scale; DY-BOCS, Dimensional Y-BOCS. aMissing data for 2 OCD patients.
OCD patients were statistically matched for age, gender, and education level to 121 control subjects (Table 1). Each control subject underwent the Structured Clinical Interview for DSM-IV nonpatient version to exclude any Axis I or II psychiatric disorders. None had a personal history of neurologic or psychiatric illness.
Our study was conducted in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki). The study protocol was approved by the Institutional Review Board of the University Hospital of Bellvitge (Barcelona). Written informed consent was obtained from each participant.
MRI Acquisition
All participants of this study were uniformly assessed using a single MRI scanner. We used a 1.5-T Signa Excite system (General Electric, Milwaukee, WI) equipped with an 8-channel phased-array head coil and single-shot echo-planar imaging software. The functional sequence consisted of gradient recalled acquisition in the steady state (repetition time, 2000 ms; echo time, 50 ms; and pulse angle, 90°) in a 24-cm field of view, with a 64 × 64 pixel matrix and a slice thickness of 4 mm (interslice gap, 1 mm). Twenty-two interleaved sections, parallel to the anterior–posterior commissure line, were acquired to generate 120 whole-brain volumes (total duration of 4 min), excluding 4 initial additional dummy volumes. Participants were instructed to simply relax, stay awake, and to lie still without moving, while keeping their eyes closed throughout.
We also acquired a high resolution T1-weighted anatomical image for each subject using a 3D fast spoiled gradient inversion-recovery prepared sequence with 130 contiguous slices (repetition time, 11.8 ms; echo time, 4.2 ms; flip angle, 15°) in a 30-cm field of view, with a 256 × 256 pixel matrix and a slice thickness of 1.2 mm.
Iso-Distant Average Correlation Maps
Imaging data were processed using MATLAB version 2014b (The MathWorks Inc, Natick, Mass) and Statistical Parametric Mapping software (SPM8; The Wellcome Department of Imaging Neuroscience, London).
Whole-brain IDAC maps were generated by estimating the average temporal correlation of each voxel with all its neighboring voxels placed at increasingly separated isodistant intervals. Three IDAC maps were obtained at the distance intervals 5–10, 15–20, and 25–30 mm. Proof of concept and sensitivity analyses are extensively described in our early report (Macià et al. 2018). Precise image processing procedures adopted in the current study are described in full in the Supplementary Material. In short, important steps included comprehensive head motion correction, estimation of IDAC measures for each voxel in native space separately for each hemisphere, creation of group template from 3D anatomical individual acquisitions and normalization to the Montreal Neurological Institute (MNI) space using DARTEL (Ashburner 2007) to enable group inferences. Three participants were excluded from an initial sample of 163 OCD patients and 3 participants from a sample of 124 control subjects due to head motion during MRI acquisition (see Supplementary Material).
Multidistance IDAC color maps were generated from the overlay of the 3 IDAC maps obtained in 1-sample, 2-sample, and correlation analyses using an RGB color codification (Fig. 1). That is, RGB color channels permitted the display of 3 values at the same time. RED was chosen to display results from 5 to 10 mm IDAC map analyses, GREEN from 15 to 20 mm and BLUE from 25 to 30 mm. Overlapping these primary colors produces a full range of secondary colors, which are highly informative with regards to variations in the functional structure of the cerebral cortex, and of variations in both group differences and correlations (see Results).
Illustration of Iso-Distant Average Correlation (IDAC) brain mapping. The images represent the results obtained in 1-sample analyses from control subjects (n = 121) at 3 IDAC measures. The gray images (top) correspond to the 3 individual IDAC maps, which are scaled to their maximal t value using conventional, automated SPM tools. The color image (bottom) shows the result of superimposing the 3 IDAC maps using RGB (red, green, and blue) display and maintaining the original scaling. The final output is made up of primary RGB colors and their secondary combinations. Note that such a multidistance map is able to discriminate between various cortical areas.
Region-of-Interest Functional Connectivity Analysis
Conventional region-of-interest functional connectivity analyses were performed to explore the distant connectivity of areas showing local connectivity anomalies. The “seed” regions used to generate whole-brain separate connectivity maps were centered in the orbitofrontal, auditory, visual, and sensorimotor areas showing significant between-group differences in IDAC functional connectivity measures (see further). The procedures used for generating the seed maps have been reported (Harrison et al. 2009; Pujol et al. 2016) and are and described in Supplementary Methods.
Statistical Analysis
Participants’ IDAC connectivity maps were included in SPM group-wise random-effects analyzes adopting a 2 × 3 mixed ANOVA (ANCOVA) model (group [patient, control] by distance [5–10, 15–20, and 25–30 mm]). A motion summary measure (interframe motion (Pujol et al. 2014)) for each participant was included as a covariate. We tested for bidirectional group differences across the 3 distances (OCD > controls [SPM contrast 1 −1 1 −1 1 −1] and OCD < controls [SPM contrast −1 1 −1 1 −1 1]). Between-group differences in connectivity were then estimated for each of the 3 IDAC maps.
In addition, we estimated voxel-wise correlations between Y-BOCS scores and whole-brain functional connectivity for each of the 3 IDAC maps at distances 5–10, 15–20, and 25–30 mm in the OCD group. Similar voxel-wise correlation analyses were performed using Dimensional Y-BOCS scores for Contamination and Cleaning, Aggressive and Checking, Symmetry and Ordering, Sexual and Religious, and Hoarding dimensions.
As to the region-of-interest (seed) analysis, groups were compared in SPM using 2-sample t-tests (limited to voxels positively correlated with the seed region).
In all analyses, results were considered significant when clusters formed at a threshold of P < 0.005 survived whole-brain family-wise error (FWE) correction (P < 0.05), calculated using SPM.
Results
Cerebral Cortex Functional Connectivity Using IDAC Measures
Whole-brain maps were generated using distinct IDAC functional connectivity measures and combined to illustrate the local functional structure of the cerebral cortex. Figure 1 shows results from the control subject group illustrating the RGB composition display on the right hemisphere lateral view generated from IDAC maps at 3 distances. Consistent with results from our early study (Macià et al. 2018), the maps were able to discriminate between major anatomofunctional cortical areas. For example, in the visual association cortex, the dominant pattern involves high connectivity at all local distance ranges. The angular and supramarginal gyri of the inferior parietal lobule are both mostly connected at short and medium distances. Also, a large part of the prefrontal cortex is dominantly connected at short distance, whereas the insula and the central opercular region are mostly connected at long distance.
Group Differences in the Functional Structure of the Cerebral Cortex
We firstly tested for potential differences between OCD patients and control subjects in IDAC functional connectivity measures across the 3 distances (Effect of Group). Significant results were all obtained in the contrast OCD < controls. OCD patients showed weaker IDAC functional connectivity across the 3 distances in the bilateral sensorimotor cortex, bilateral visual cortex, right auditory cortex, left anterior insula, and bilateral orbitofrontal cortex (Fig. 2 and Suppl. Table S1). The identified anterior insula region overlaps with the primary gustatory cortex (Rolls 2016), and the posterior extent of the orbitofrontal cortex includes structures of the primary olfactory cortex (Gottfried 2010). The analysis, therefore, identified local functional connectivity anomalies implicating all sensory cortex modalities in addition to the orbitofrontal cortex. No group differences were observed in the contrast OCD > controls.
Group differences in Iso-Distant Average Correlation (IDAC) measures across distance maps. Differences between OCD patients and control subjects in functional connectivity were tested using an ANOVA that included the 3 IDAC maps (5–10, 15–20, and 25–30 mm) for both study groups. The brain views illustrate significant findings obtained in the contrast OCD < controls (negative group effect—weaker functional connectivity). Note that all sensory cortex modalities were implicated to some extent (somatosensory, visual, auditory, gustatory, and olfactory; see text) in addition to the orbitofrontal cortex.
A map of between-group differences in local functional connectivity was then generated for each of the 3 IDAC maps corresponding to functional connectivity distances 5–10, 15–20, and 25–30 mm (Fig. 3). Overall, this set of analyses showed the sensory cortices and the orbitofrontal cortex to have weaker functional connectivity at each distance, but notable differences were identified across the maps. See Figure 4 and Suppl. Figure S1, Suppl. Table S2 for group-by-distance interactions. For instance, group differences in sensorimotor cortex and primary auditory cortex functional connectivity were more notable at short distances (5–10 and 15–20 mm); whereas in the visual cortex, differences were more evident at long distances (15–20 and 25–30 mm). As for the orbitofrontal cortex, the most extensive differences were observed at mid-distance (15–20 mm). Interestingly, there was a gradient effect in the orbitofrontal region, with a more prominent short-distance alteration in the anterior part versus more prominent long-distance alteration in the posterior part (Figs 3 and 4). An integrated RGB display of the 3 distance map analyses is provided in Figure 5 to summarize the functional structure alterations identified in OCD.
Group differences in IDAC measurements at 3 functional connectivity distances (contrast OCD < controls). OCD patients showed lower IDAC functional connectivity measures involving sensory cortices and the orbitofrontal cortex in the 3 analyses. However, the effect was not identical for each distance.
Group by distance interactions as to local functional connectivity IDAC measures. Top, short distance greater effect (OCD < Controls at short distances) > (OCD < controls at long distance). Bottom, long distance greater effect (OCD < controls at long distances) > (OCD < controls at short distance).
RGB (red, green, and blue) display summarizing alterations in the local architecture of functional connectivity in OCD. The color map corresponds to the superimposition of between-group differences in IDAC measurements at functional connectivity distances 5–10 mm (red), 15–20 mm (green), and 25–30 mm (blue). That is, RGB combination of the 3 sets of results shown in Figure 3. The final output is made up of primary RGB colors and their secondary combinations.
Correlations Between Symptom Severity and IDAC Measures
A correlation analysis between Y-BOCS scores and IDAC measures was generated for each of the 3 whole-brain IDAC maps at distances 5–10, 15–20, and 25–30 mm in the OCD group. All the findings were observed in the direction of symptom severity associated with weaker functional connectivity. Regions showing significant correlation included the auditory cortex and a region extending from the orbitofrontal cortex to the medial frontal cortex and the subgenual portion of the anterior cingulate cortex. Correlations were also significant in the temporal pole and angular gyrus (Fig. 6 and Suppl. Table S3). A distance effect was also observed in the correlation pattern. The inverse correlation between Y-BOCS scores and IDAC measures was significantly greater at short versus long distances (correlation at short distances > correlation at long distance) in the orbitofrontal/frontal medial cortex, the temporal pole, and the angular gyrus. A subthreshold correlation interaction was observed in the auditory cortex.
Correlations between symptom severity (Y-BOCS scores) and IDAC functional connectivity measures. Top; regions showing significant inverse correlations between severity and connectivity in different IDAC maps. Middle; all the results obtained in separate maps superimposed using RGB display. Bottom; short distance greater effect (inverse correlation at short distances > correlation at long distance). Subthreshold findings are displayed when implicated our regions of interest.
Region-of-Interest Functional Connectivity Analysis
Conventional region-of-interest functional connectivity analyses were performed to examine the relationships of identified local functional connectivity anomalies with other brain regions. The “seed” regions used to generate whole-brain connectivity maps were centered in the orbitofrontal, auditory, visual, and sensorimotor areas showing significant between-group differences in IDAC functional connectivity measures. In each map, OCD patients showed weaker functional connectivity between the region-of-interest (seed) and a region pertaining to the same cortical area modality (Suppl. Table S4), which is consistent with the primary study findings relating to weaker functional connectivity in local IDAC measures. In the sensorimotor cortex map, OCD patients showed in addition stronger functional connectivity than control subjects between the sensorimotor region and a large region implicating the basal ganglia and thalamus bilaterally (Suppl. Table S4 and Suppl. Fig. S2).
The Influence of Major OCD Symptom Dimensions
The influence of major OCD symptom dimensions on the functional structure of the cerebral cortex was assessed by estimating the correlations between IDAC measures and symptom severity measured by the Dimensional Y-BOCS. No significant result was appreciated in this analysis. At lenient thresholds (e.g., P < 0.005 and cluster size > 0.8 mL, 30 voxels) results were also scarce (Suppl. Table S5).
Discussion
Our functional connectivity mapping was able to parcel the cerebral cortex into distinct regions replicating the results described in our early report (Macià et al. 2018). The degree of correspondence with classical anatomical–functional areas suggests that multidistance IDAC maps capture primary features of the functional structure of the cerebral cortex. OCD patients showed functional structure anomalies characterized by weaker multidistance functional connectivity within the orbitofrontal cortex and in all sensory cortex modalities. There was a notable distance effect, with a different degree of alteration in several areas in the short, middle, and long-distance maps. Symptom severity was associated with the degree of functional structure alteration in OCD-relevant brain regions.
Previous research has demonstrated orbitofrontal cortex alterations in anatomy, metabolism, and function in patients with OCD (Whiteside et al. 2004; Menzies et al. 2008; Fineberg et al. 2010; Pauls et al. 2014; Ahmari and Dougherty 2015; van den Heuvel et al. 2016; Jung et al. 2017; van der Straten et al. 2017). Our results now show that this part of the brain is also anomalous in terms of local functional structure with a pattern of weaker intracortical neural coupling. A reduction in local functional connectivity in a brain region characteristically showing high metabolism at rest in OCD (Whiteside et al. 2004; van der Straten et al. 2017) may be best interpreted as reflecting deficient cortical inhibition. Indeed, lower local functional connectivity may result from both a reduction in neural activity associated with a decrease in the number of synchronized neurons (Niessing et al. 2005) and cortical inhibitory neuron deficiency associated with the desynchronization (i.e., uncoupling) of principal neuron populations (Buzsáki and Watson 2012; Mathalon and Sohal 2015; Turkheimer et al. 2015). Our results are in line with the latter possibility. According to basic neurophysiological research, most network activity (and MRI signal) oscillations are based on inhibitory neurons that synchronize assemblies of excitatory/principal neurons (Buzsáki and Watson 2012; Mathalon and Sohal 2015).
The pattern of results shown in our study indicates that orbitofrontal cortex alterations are not homogeneous. Functional connectivity differences between OCD patients and control subjects were more evident at shorter distances in the anterior part of the orbitofrontal cortex and at longer distances in the posterior part. The posterior part includes structures collectively referred to as the primary olfactory cortex (piriform cortex, anterior olfactory nucleus, and olfactory tubercle) (Gottfried 2010). Interestingly, abnormally large gray matter volume has been reported in this region in OCD patients with anomalous odor identification (Segalàs et al. 2014), whereas the anterior orbitofrontal cortex is characterized by gray matter volume reduction (Pujol et al. 2004; Menzies et al. 2008; Hou et al. 2013; Frydman et al. 2016). Also, greater symptom severity in the present study was associated with altered functional connectivity in the lateral orbitofrontal cortex (and in the anterior cingulate cortex region). Therefore, anomalies in the orbitofrontal cortex seem to have subcomponents with potentially distinct pathological significance.
In addition to the orbitofrontal cortex, we identified broad alterations in sensory cortices. Figure 2 provides compelling evidence as to the involvement of primary somatosensory, auditory, visual, gustatory, and olfactory areas. As in the orbitofrontal cortex, findings were in the direction of reduced intracortical functional coupling, which may similarly express inhibitory neuron dysfunction. Reduced intracortical inhibition has indeed been directly demonstrated in OCD patients via transcranial magnetic stimulation (Richter et al. 2012; Radhu et al. 2013; Bunse et al. 2014; Russo et al. 2014). Significant impairment has also been demonstrated in neurophysiological studies examining sensory filtering. Specifically, OCD patients showed deficient prepulse inhibition of both acoustic startle responses (Ahmari et al. 2012; Kohl et al. 2013) and somatosensory evoked potentials (Rossi et al. 2005).
An important question to understand, therefore, is how dysfunction at primary sensory areas may contribute to OCD. Activity in sensory cortices, both from primary sensory inputs and from mental representations, are relevant information sources to the orbitofrontal cortex controlling motivated behavior (Menzies et al. 2008; Rolls and Grabenhorst 2008; Pauls et al. 2014; Wood and Ahmari 2015). Thus, at a simple level, deficient inhibition within sensory cortices may add noise to the system and potentiate the urge to respond to augmented sensations, for example, like the urge to scratch pruritic skin. More generally, deficient multisensory inhibition may potentiate repetitive behavior across the various disorders of the obsessive–compulsive spectrum.
Sensory cortex deficient inhibition may be particularly relevant in skin-picking and hair-pulling disorders (Fineberg et al. 2018). We had previously observed that the degree of functional connectivity between the somatosensory cortex and basal ganglia was selectively associated with the severity of self-picking behavior in Prader Willi syndrome (Pujol et al. 2016). Interestingly, we also observed “sensory phenomena” in typical OCD patients associated with gray matter volume increase precisely in the sensorimotor cortex (Subirà et al. 2015). Sensory phenomena preceding or accompanying compulsions are present in 65% of OCD patients (Ferrão et al. 2012) and relate, for instance, to uncomfortable tactile sensations and visually or aurally triggered “just-right” perceptions.
We also found a distance effect in primary sensory areas. Group differences in sensorimotor cortex and primary auditory cortex functional connectivity were more notable at short distance; whereas in the visual cortex, differences were more evident at long distance. The correlation with symptom severity was generally stronger at short distance. Such distance effects may be related to regional differences in the complex functional anatomy of cortical connections, which appears to be altered in primary sensory areas and orbitofrontal cortex in OCD. It may be pertinent now to explore whether IDAC measures are sensitive enough to capture subtle changes in the functional structure of the cerebral cortex in other psychiatric and neurological disorders. Of particular interest may be to explore conditions where neuronal inhibition is thought to be dysfunctional such as schizophrenia, autism spectrum disorders, and epilepsy (Turkheimer et al. 2015).
All in all, we need to be cautious in suggesting pathogenic implications on the grounds of our observations, as the link between basic neurophysiology and functional connectivity measures from imaging studies has only been partially established (Niessing et al. 2005; Mathalon and Sohal 2015). Also, a potential effect of patient medication on the functional structure of the cerebral cortex cannot be ruled out. Indeed, most patients were under a stable medication regime. However, it is important to indicate that our correlation analysis was able to identify a significant association between symptom severity and part of the alterations, which does not support the medication effect. More importantly, the full range of functional connectivity alterations (Sakai et al. 2011; Beucke et al. 2013; Hou et al. 2013; Göttlich et al. 2014; Shin et al. 2014; Jung et al. 2017; Moreira et al. 2017; Niu et al. 2017; Qiu et al. 2017; Zhao et al. 2017) and deficient cerebral cortex inhibition (Ahmari et al. 2012; Richter et al. 2012; Russo et al. 2014) have both been consistently reported in unmedicated OCD patients.
In conclusion, we have assessed the functional structure of the cerebral cortex using a novel imaging approach that provided a distinct perspective of brain dysfunction in OCD. The new measures identified anomalies once again in the orbitofrontal cortex, now in the form of weaker intracortical activity coupling, but also in each primary sensory cortex. Based on existing neurophysiological studies, such cortical activity desynchronization may be interpreted as indicating inhibitory neuron deficiency and deficient filtering of sensory stimuli. However, future studies are required to definitively establish the specific role of the identified anomalies in the pathophysiology of OCD and isolate the potential effects of medication.
Funding
Carlos III Health Institute (Grants PI13/01958 and PI16/00889).
Notes
We thank the Agency of University and Research Funding Management of the Catalonia Government for their participation in the context of Research Groups SGR2017-1198, SGR2017-0134, and SGR2017-1247. Dr Soriano-Mas thanks the support of the Miguel Servet contract from the Carlos III Health Institute (CPII16/00048). Conflicts of interest: Drs Jesus Pujol, Laura Blanco-Hinojo, Didac Macia, Pino Alonso, Ben J. Harrison, Gerard Martínez-Vilavella, Joan Deus, José M. Menchón, Narcís Cardoner, and Carles Soriano-Mas report no competing interests.
