Excessive Checking in Obsessive-Compulsive Disorder: Neurochemical Correlates Revealed by 7T Magnetic Resonance Spectroscopy

Abstract

Background: Compulsive checking, a common symptom of obsessive-compulsive disorder (OCD), has been difficult to capture experimentally. Therefore, determination of its neural basis remains challenging despite some evidence suggesting that it is linked to dysfunction of cingulostriatal systems. This study introduces a novel experimental paradigm to measure excessive checking and its neurochemical correlates.

Methods: Thirty-one patients with OCD and 29 healthy volunteers performed a decision-making task requiring them to decide whether 2 perceptually similar visual representations were the same or different under a high-uncertainty condition without feedback. Both groups underwent 7T magnetic resonance spectroscopy scans on the same day. Correlations between out-of-scanner experimental measures of checking and the glutamate/GABA (gamma-aminobutyric acid) ratio in the anterior cingulate cortex, supplementary motor area, and occipital cortex were assessed. Their relationship with subjective ratings of doubt, anxiety, and confidence was also investigated.

Results: Patients with OCD exhibited excessive and dysfunctional checking, which was significantly correlated with changes in the glutamate/GABA ratio within the anterior cingulate cortex. No behavioral/neurochemical relationships were evident for either the supplementary motor area or occipital cortex. The excessive checking observed in patients was negatively correlated with their confidence levels and positively related to doubt, anxiety, and compulsivity traits.

Conclusions: We conclude that experimental measures of excessive and dysfunctional checking in OCD, which have been linked to increased doubt, anxiety, and lack of confidence, are related to an imbalance between excitatory and inhibitory neural activity within the anterior cingulate cortex. This study adds to our understanding of the role of this region in OCD by providing a laboratory model of the possible development of compulsive checking.

Keywords: 7T-MRS; Anterior cingulate cortex; Checking; GABA; Glutamate; OCD.

Figure 1

Image verification task. (A) An example of a trial, with 2 apples differing in angle, displayed for 1 second, with a white interstimulus interval presented for 800 ms in between. Participants could decide immediately whether the 2 images were the same or different by pressing the words on the screen or they could press the red << sign to go back and check the images again. Once a decision had been made, participants were asked to rate how confident they were about their answers using a 4-point scale as depicted in the third panel. (B) Three examples of black-and-white stimuli used in this task. From left to right: 2 bears (the right bear is bigger), ducks (the right duck is more crooked), and umbrellas (the 2 umbrellas are exactly the same).

Figure 2

Voxel positions and landmarks. The location of the 3 voxels is shown in a sagittal space while indicating the landmarks that were used to increase their consistent placement and to avoid an overlap between the anterior cingulate cortex and supplementary motor area (SMA) voxels. First, a horizontal line was drawn between the anterior and posterior commissure with 2 vertical lines going through each of them perpendicularly. These lines are depicted in red for the anterior cingulate cortex and SMA voxels. This figure is reproduced from Figure S1 in Biria et al. (44). (A) The anterior cingulate cortex box (in yellow) was placed in front of the line going through the anterior commissure, with the outer left corner of the box being in front of the genu of the corpus callosum. (B) The SMA box (in purple) was placed above the pons and between the 2 red lines. The upper side of the box was placed parallel with the skull above it. (C) Finally, the occipital cortex box (in orange) was placed in the outermost corner of the occipital lobes while avoiding the skull and sinuses, with the lower side of the box being parallel with the red line above the cerebellum. All 3 voxels were placed bilaterally, and the SMA box included regions from the SMA and the pre-SMA.

Figure 3

Image verification task results. Boxplots for healthy volunteers (HVs) (N = 29) in green and patients with obsessive-compulsive disorder (OCD) (N = 31) in blue on measures of (A) checking rate, (B) accuracy of choices (%), and (C) mean confidence ratings (ranging from 1 to 4). For visualization purposes, 2 patients who checked more than 2 SDs from the mean are excluded from these figures; however, they are still included in the analysis. The black circles show the individual data points; the boxes start from the first to the third quartile, with a horizontal line and a notch through the median. The whiskers go from each quartile to the minimum and maximum. The notch approximates a 95% CI for the median. If the notches of 2 boxes do not overlap, this suggests that the medians are significantly different. The points outside whiskers represent the outliers. ∗p < .05.

Figure 4

Checking rate relationships with accuracy and confidence. This figure depicts the checking rate relationship with both accuracy (left) and confidence ratings (right). Panels (A) and (B) show these relationships for healthy volunteers (HVs) (N = 29) in green, whereas panels (C) and (D) show the relationships for patients with obsessive-compulsive disorder (OCD) (N = 31) in blue. The fitted lines are drawn only for significant relationships. For visualization purposes, 2 patients who checked more than 2 SDs from the mean are excluded from these figures but not from the analysis. The line of best fit is shown with the 95% CIs for the regression estimate in translucent bands around the regression lines. r indicates the Pearson correlation coefficient; r_s indicates to the Spearman rank correlation coefficient, ∗p < .05, ∗∗p < .005. FDR, false discovery rate.

Figure 5

Relationship between image verification task checking rates and brain neurochemical measurements. (A) Examples of the LCModel analysis of in vivo ¹H magnetic resonance spectra acquired from a healthy volunteer (HV) at 7T (semi-LASER, echo time/repetition time = 1.99/4300 ms, from a 20 × 20 × 20 mm³ voxel placed bilaterally at the anterior cingulate cortex [ACC]). The acquired spectrum is plotted in black, and the fit is presented in red for GABA (gamma-aminobutyric acid) (left) and glutamate (Glu) (right). (B) The correlations between the levels of Glu/GABA concentrations in the ACC and checking rates (left) and accuracy (right). The blue color represents patients with obsessive-compulsive disorder (OCD) (N = 30), whereas green depicts the data for HVs (N = 29). The fitted lines are drawn only for significant relationships in the OCD group, with the 95% CIs for the regression estimate in translucent bands around the regression lines. The neurometabolites were normalized using (Cr+PCr), corrected for gray and white matter and cerebrospinal fluid of each individual voxel, within participants. r indicates Pearson correlation coefficient, and r_S indicates Spearman rank correlation coefficient. ∗p < .05. FDR, false discovery rate; ppm, parts per million.