Behavioral inhibition/activation system in obsessive-compulsive disorder: brain morphometric correlates and treatment outcomes

Abstract

The behavioral inhibition/activation system (BIS/BAS) has been implicated in a range of psychiatric disorders, yet its role in obsessive-compulsive disorder (OCD) remains unclear. Integrating self-report measures, neuroimaging, and clinical interventions, we explored how BIS/BAS contribute to OCD psychopathology. Patients with OCD showed elevated BIS and reduced BAS drive sensitivity, with BIS tracking symptom severity. BIS was indirectly associated with obsession and compulsion symptoms via right supramarginal gyrus sulcal depth and right-lateralized thalamic volume, respectively, with the BIS-thalamus association emerging as a key brain morphometric association pattern distinguishing patients with OCD from healthy controls. In addition, BIS decreased following selective serotonin reuptake inhibitor treatment, with reductions correlating with OCD symptom improvement, and BAS increased following cognitive behavioral therapy. In summary, our findings suggest distinct contributions of BIS and BAS to OCD psychopathology and identify them as clinically relevant dimensions for future investigation.

Introduction

Obsessive-compulsive disorder (OCD) is a severe and disabling mental disorder characterized by intrusive thoughts and repetitive behaviors [1, 2], affecting approximately 4.1% of the global population [3], and imposing a substantial economic burden on society [4]. The behavioral inhibition system (BIS) and behavioral activation system (BAS) are key systems involved in the regulation of human emotion and behavior, conceptualized as personality dimensions reflecting individual differences in sensitivity to reward and punishment [5–7]. Heightened sensitivity of these systems is considered a risk factor for psychiatric disorders and has been closely associated with a range of affective symptoms [8]. Investigating BIS/BAS abnormalities in OCD therefore provides important insights into its underlying psychopathology.

Previous studies on BIS/BAS in OCD have primarily focused on BIS, which has been found to be elevated compared with healthy controls [9–12] and associated with OCD symptom severity [11–15], whereas findings on BAS remain inconsistent [11, 12, 14]. BIS abnormalities in OCD may reflect enhanced motivation for loss avoidance and threat avoidance [11]. Moreover, BIS sensitivity may play a potential bridging role in translating OCD-related risk factors, such as childhood trauma [15], and maladaptive subjective experiences, such as disgust [13], into OCD symptoms. These findings suggest a close association between BIS and vulnerability to OCD, highlighting the need for further investigation into how BIS contributes to OCD psychopathology.

Brain correlates associated with BIS/BAS abnormalities in OCD remain unclear, and to our knowledge, this question has not been directly investigated. Substantial prior work has characterized brain regions associated with BIS/BAS [16]. BIS has been associated with regions involved in threat detection and emotion processing, including the hippocampus, amygdala, thalamus, and prefrontal regions [16, 17]. By contrast, BAS has been more consistently associated with reward-related regions, particularly nucleus accumbens, prefrontal and ventral striatal areas within the mesocorticolimbic reward circuit [16, 18, 19]. Notably, these BIS/BAS-related regions show substantial overlap with the classical cortico-striato-thalamo-cortical (CSTC) circuit [1, 20] implicated in OCD. Studies in other psychiatric disorders have identified potential disorder-specific BIS/BAS-associated brain correlates, though simultaneously, shared brain correlates are also present. For example, in affective disorders such as depression and anxiety, BIS/BAS has been linked to fronto-limbic circuits [21, 22], whereas in impulsivity-related conditions, such as substance dependence, findings have often implicated the executive control network [23–25]. Taken together, investigating BIS/BAS-associated brain correlates in OCD may provide further insight into the specific contribution of BIS/BAS to OCD psychopathology.

Another important unresolved issue is whether BIS/BAS function as trait markers predisposing to OCD or as state-dependent indicators of symptom severity. Exploring how treatments influence BIS/BAS in OCD, and the coupling between changes in OCD symptoms and BIS/BAS, may help address this question. Current first-line treatments for OCD include pharmacological interventions, such as selective serotonin reuptake inhibitors (SSRIs), and psychotherapy, notably cognitive behavioral therapy (CBT) [26]. To our knowledge, no study has examined whether BIS/BAS would change after these two first-line interventions in patients with OCD. Evidence from other psychiatric disorders has documented potential associations between treatments and BIS/BAS. More specifically, BIS appears responsive to both SSRIs and CBT, whereas BAS shows a stronger association with CBT [27–29]. Given disorder-specific psychopathology and the heterogeneity of treatments across diagnoses, whether—and to what extent—SSRIs and CBT influence BIS/BAS in OCD remains uncertain.

In this study, we aimed to integrate self-report measures, neuroimaging, and clinical interventions to further investigate BIS/BAS abnormalities in OCD in a large sample of patients, while controlling for comorbidities and medication. Given the limited prior evidence, this study was conducted as an exploratory investigation. Our specific objectives are to: (1) Identify brain morphometric correlates of BIS/BAS abnormalities in OCD; (2) Investigate whether BIS/BAS change following first-line treatments in OCD.

Methods

Ethical considerations

The study obtained ethical approval from the Ethics Committee of the Shanghai Mental Health Center (2014-11), and all participants provided written informed consent before enrollment. All methods were performed in accordance with the relevant guidelines and regulations.

Study design

This study comprised two components: a cross-sectional case-control investigation and a longitudinal clinical intervention study, with data collected between 2014 and 2018. For the case-control investigation, both patients with OCD and HC completed assessments of psychopathological measures and BIS/BAS system, with a subset of participants additionally undergoing MRI scanning. The clinical intervention study is a secondary analysis of a prospective study in which patients with OCD received either SSRIs or CBT for 12 weeks. The original study was designed to conduct a randomized controlled trial to evaluate the comparative effectiveness of SSRIs and CBT. However, during implementation, many patients declined randomization. For ethical considerations, these individuals received their preferred treatment. This study included both participants who were randomized [30] and those allocated by preference, and therefore should not be considered a randomized controlled study. It should also be noted that the original study was not preregistered, as preregistration had not yet become common practice in 2014. This study should not be construed as a comparative test of SSRI versus CBT effectiveness; symptom scores are reported only to describe pre-post changes in psychopathological characteristics within each treatment group.

Participants

A total of 247 unmedicated patients with OCD and 113 HC were included in this study. The sample size for this secondary analysis was determined by the availability of eligible participants, rather than by a power calculation conducted before data collection. Among them, structural MRI was acquired from 154 patients with OCD and 93 HC, and a subset of 74 patients with OCD underwent 12-week first-line treatments—42 with SSRIs and 32 with CBT. Patients with OCD were recruited at the Shanghai Mental Health Center (Shanghai, China). All patients with OCD met DSM-IV criteria for OCD, as assessed with the Mini International Neuropsychiatric Interview (MINI) [31], had no MINI-defined psychiatric comorbidity, and had been unmedicated for at least 8 weeks before enrollment. HC were recruited from communities in Shanghai with no MINI-defined psychiatric disorders. Detailed inclusion and exclusion criteria are presented in Supplementary S1.

Measures

Well-trained psychiatrists and psychologists assessed symptoms of OCD, depression and anxiety in patients with OCD using the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS) [32], the 24-item Hamilton Depression Rating Scale (HAMD) [33], and the 14-item Hamilton Anxiety Rating Scale (HAMA) [34]. For HC, only depression and anxiety were assessed using the HAMD and HAMA. BIS/BAS were assessed with the BIS/BAS Scales [5, 35] (Supplementary S2). All raters received standardized training from a licensed senior psychiatrist and observed a sufficient number of assessments before conducting independent evaluations, until satisfactory inter-rater reliability had been achieved. In addition, all raters received regular supervision to ensure continued inter-rater reliability. Assessments in the intervention component were assessor-blinded.

Brain imaging data acquisition and preprocessing

Structural T1-weighted images were acquired on a 3T MRI scanner (Verio; Siemens) at the Shanghai Mental Health Center (Supplementary S3). Image preprocessing and morphometric analyses were conducted in MATLAB R2018b using Statistical Parametric Mapping (SPM12) and the Computational Anatomy Toolbox (CAT12), according to standard voxel-based morphometry (VBM) and surface-based morphometry (SBM) procedures (Supplementary S4).

Clinical intervention

Both SSRI and CBT treatments lasted 12 weeks. For the SSRI treatment, patients received SSRIs approved by the China State Food and Drug Administration for OCD treatment (sertraline or fluvoxamine). For the CBT treatment, patients received psychotherapy, primarily emphasizing exposure and response prevention (ERP). Detailed intervention procedures are presented in Supplementary S5.

Statistics

All computations were carried out in SPSS (version 24), MATLAB (version R2018b), SPM12, CAT12. Statistical significance was set at p < 0.05 for two-tailed non-imaging tests. For neuroimaging analyses, significance was determined using a threshold of p < 0.001 at the voxel or vertex level, followed by cluster-level at p < 0.05 with familywise error (FWE) correction.

Descriptive statistics of demographic, psychopathological and BIS/BAS system characteristics were compared between two groups using independent-samples t-tests and chi-square tests. Considering not all variables met normality and homogeneity-of-variance assumptions, Mann-Whitney U tests were additionally conducted as non-parametric sensitivity analyses for independent-samples t-tests (Supplementary S6). Associations between different metrics were quantified using Pearson correlations within the OCD group.

For T1-weighted brain images, we first identified BIS/BAS-associated brain metrics using analyses of covariance (ANCOVAs) in SPM12, with sex, age, and education as covariates. For gray matter volume, total intracranial volume (TIV) was also included as an additional covariate, and a probability threshold of >0.2 was set as the definition of gray matter. We focused on the interaction between the group and the BIS/BAS scores, indicating whether the brain metric was significantly differentially associated with the BIS/BAS scores or not between the two groups, as well as the main effect of the BIS/BAS scores when the interaction effect was not significant, indicating whether the brain metric was significantly associated or not with the BIS/BAS scores, regardless of the group. For brain metrics showing significant interaction effects, Pearson correlations with BIS/BAS scores were examined separately within each group. For brain metrics showing significant main effects of BIS/BAS scores, correlations were assessed across the whole sample as well as within each group. To further clarify the associations between BIS/BAS system, brain and psychopathological metrics in patients with OCD, mediation models were conducted through the PROCESS toolbox [36] in SPSS. All mediation analyses were conducted using bias-corrected bootstrap procedures with 50,000 resamples, and all variables were standardized prior to analyses.

The intervention analyses were not designed to support a direct comparison of treatment-related changes between SSRIs and CBT. Changes between pre- and post-treatment in psychopathological metrics and BIS/BAS system were examined separately within the SSRIs and CBT groups using paired-samples t-tests. Considering not all variables met normality and homogeneity-of-variance assumptions, Wilcoxon signed-rank tests were additionally conducted as non-parametric sensitivity analyses for all paired-samples t-tests (Supplementary S6). For BIS/BAS dimensions showing significant pre-post changes, we examined within-group associations between changes in BIS/BAS scores and changes in Y-BOCS total scores using Pearson correlations. For exploratory purposes, we also computed correlations between the pre-treatment BIS/BAS scores and the Y-BOCS total scores change rate to examine whether BIS/BAS could predict treatment response. Y-BOCS total scores change rate was defined as (Y-BOCS_{pre-treatment} – Y-BOCS_{post-treatment})/Y-BOCS_{pre-treatment} × 100%.

Results

Descriptive statistics of demographic and psychopathological characteristics

A total of 247 unmedicated patients with OCD and 113 HC were included in this study. OCD and HC showed no significant differences in sex, age and education (p ≥ 0.161). Compared to HC, patients with OCD showed significantly higher HAMD scores (t₃₄₉ = 11.680, p < 0.001, Cohen’s d = 1.354) and HAMA scores (t₃₅₁ = 11.131, p < 0.001, Cohen’s d = 1.286) (Table 1).

Table 1.

Demographic Information and Psychopathological Characteristics (mean ± SD).

Items	OCD	HC	χ2/t	p
Sex (female/male)	118/129	52/61	0.096	0.757
Age (years)	28.95 ± 6.79	27.86 ± 6.91	1.405	0.161
Education (years)	15.00 ± 2.25	15.36 ± 2.55	–1.347	0.179
Y-BOCS total	24.58 ± 6.03	/	/	/
Obsession	13.45 ± 3.33	/	/	/
Compulsion	11.13 ± 4.35	/	/	/
HAMD-24	14.01 ± 8.03a	3.36 ± 7.46b	11.680	<0.001
HAMA-14	9.07 ± 5.49c	2.15 ± 5.12d	11.131	<0.001

OCD obsessive-compulsive disorder, HC healthy controls, Y-BOCS yale-brown obsessive-compulsive scale, HAMD 24-item hamilton depression rating scale, HAMA 14-item hamilton anxiety rating scale.

^aData from 3 participants were unavailable.

^bData from 6 participants were unavailable.

^cData from 2 participants were unavailable.

^dData from 5 participants were unavailable.

BIS/BAS in OCD

To examine group differences in BIS/BAS system, independent-samples t-tests were performed for BIS scores, BAS total scores, and BAS subscale scores (reward responsiveness, drive, and fun seeking) (Table S1). Patients with OCD exhibited significantly higher BIS scores than healthy controls (t₃₅₈ = 8.235, p < 0.001, Cohen’s d = 0.935) (Fig. 1a), whereas BAS drive scores were significantly lower in the OCD group (t₃₅₈ = –2.000, p = 0.046, Cohen’s d = –0.227) (Fig. 1e). No significant group differences were found for BAS total scores, BAS reward responsiveness scores and BAS fun seeking scores (p ≥ 0.262) (Fig. 1c, d and f).

Fig. 1. BIS/BAS Characteristics and Symptom Associations in OCD.

a Differences in BIS Scores between OCD and HC. b The association between BIS and Y-BOCS total scores in the OCD group. Differences between OCD and HC in c BAS total scores, d BAS reward responsiveness scores, e BAS drive scores, f BAS fun seeking scores. ^***p < 0.001; ^**p < 0.01; ^*p < 0.05; ^n.sp ≥ 0.262^. Violin plots display distributions (box = median and IQR, white dot = mean). Lines in Fig. 1b are least-squares fits with 95% CI. Y-BOCS, Yale-Brown Obsessive-Compulsive Scale; BIS, behavioral inhibition system; BAS, behavioral activation system; BASR, BAS reward responsiveness; BASD, BAS drive; BASF, BAS fun seeking.

To further explore associations between BIS/BAS system and psychopathological symptoms in OCD group, correlation analyses were conducted between BIS/BAS scores and psychopathological symptoms. BIS scores were positively correlated with Y-BOCS total scores (r = 0.183, p = 0.004) (Fig. 1b), Y-BOCS obsession scores (r = 0.137, p = 0.031), Y-BOCS compulsion scores (r = 0.149, p = 0.020), HAMD scores (r = 0.165, p = 0.010), and HAMA scores (r = 0.153, p = 0.016). In contrast, no significant correlations were observed for BAS total scores or any of its subscale scores with these psychopathological symptoms (p ≥ 0.281).

Brain metrics associated with BIS/BAS system

Among the participants, 154 patients with OCD and 93 HC volunteered for MRI scanning. The demographic, psychopathological, and BIS/BAS metrics of the MRI subsample were generally representative of the overall sample (see Supplementary S7 and Table S2).

For BIS scores, the whole-brain ANCOVA revealed that the correlation between gray matter volume in the bilateral thalamus proper with right lateralization and BIS scores was significantly lower in the OCD group than in the HC group (peak t = 4.04, p = 0.022, familywise error corrected at the cluster level) (Fig. 2a and Table S3). No significant correlation between gray matter volume in the right-lateralized thalamus proper and BIS scores was observed when the two groups were combined. Further analyses showed that gray matter volume in the right-lateralized thalamus proper was negatively correlated with BIS scores in the OCD group (r = –0.222, p = 0.006), but positively in the HC group (r = 0.259, p = 0.012) (Fig. 2b). In addition, we found sulcal depth in the right supramarginal gyrus (SMG) was significantly positively correlated with BIS scores across two groups (peak t = 3.76, p = 0.043, familywise error corrected at the cluster level) (Fig. 2c and Table S3). Further analysis showed a significant correlation between BIS scores and sulcal depth in the right SMG for the whole sample (r = 0.239, p < 0.001) (Fig. 2d), the OCD group (r = 0.177, p = 0.028) and the HC group (r = 0.281, p = 0.006). No significant clusters were identified for BAS total scores.

Fig. 2. Brain Morphometric Correlates of BIS in OCD and HC.

a Interaction effect between BIS scores and group in the gray matter volume in the right-lateralized thalamus proper. b Associations between BIS scores and gray matter volume in the right-lateralized thalamus proper in both OCD and HC groups (line = linear fit; shading = 95% CI). c Main effect of BIS scores in the sulcal depth in the right SMG. d Associations between BIS and sulcal depth in the right SMG in the whole sample (line = linear fit; shading = 95% CI). ^***p < 0.001; ^**p < 0.01; ^*p < 0.05. SMG, supramarginal gyrus; BIS, behavioral inhibition system; OCD, obsessive-compulsive disorder; HC, healthy controls.

Finally, we conducted independent-samples t-tests to examine whether these two brain metrics (gray matter volume in the right-lateralized thalamus proper and sulcal depth in the right SMG) were different between the two groups. No significant effects were found (p ≥ 0.200).

Associations between BIS/BAS system, brain and OCD symptoms

To further investigate whether the identified BIS-associated brain regions mediated the relationship between BIS scores and OCD symptom severity in patients with OCD, we performed mediation models using BIS scores as independent variables, gray matter volume in the right-lateralized thalamus proper or sulcal depth in the right SMG as mediators, and Y-BOCS total scores, Y-BOCS obsession scores, or Y-BOCS compulsion scores as dependent variables, separately. Significant indirect effects of BIS on Y-BOCS total scores were observed through gray matter volume in the right-lateralized thalamus proper and through sulcal depth in the right SMG, respectively (indirect effect_thalamus = 0.0372, SE = 0.0248, 95% CI = [0.0024, 0.1042]; indirect effect_SMG = 0.0332, SE = 0.0235, 95% CI = [0.0013, 0.0970]) (Fig. 3a, b). When examining sub-dimensions of OCD symptoms, the indirect effect through gray matter volume in the right-lateralized thalamus proper was significant only for Y-BOCS compulsion scores (indirect effect = 0.0380, SE = 0.0266, 95% CI = [0.0017, 0.1109]), whereas the indirect effect through sulcal depth in the right SMG was significant only for Y-BOCS obsession scores (indirect effect = 0.0329, SE = 0.0204, 95% CI = [0.0038, 0.0875]) (Supplementary S8 and Fig. S1).

Fig. 3. Mediation Models for BIS, Brain Morphometric Metrics and OCD Symptoms.

Mediation models testing indirect effect between BIS scores and Y-BOCS total scores through a gray matter volume in the right-lateralized thalamus proper and b sulcal depth in the right SMG. Parallel mediation models for c Y-BOCS total scores, d Y-BOCS obsession scores, e Y-BOCS compulsion scores. ^**p < 0.01; ^*p < 0.05. SMG, supramarginal gyrus; BIS, behavioral inhibition system; Y-BOCS, Yale-Brown Obsessive-Compulsive Scale; obsession refers to the Y-BOCS obsession subscale scores; compulsion refers to the Y-BOCS compulsion subscale scores.

Building on these findings, we further constructed parallel mediation models including both brain mediators simultaneously. In the model for Y-BOCS total scores, the total indirect effect of BIS was significant, with significant specific indirect effects through gray matter volume in the right-lateralized thalamus proper and sulcal depth in the right SMG (indirect effect_total = 0.0778, SE = 0.0344, 95% CI = [0.0232, 0.1601]; indirect effect_thalamus = 0.0415, SE = 0.0259, 95% CI = [0.0053, 0.1116]; indirect effect_SMG = 0.0363, SE = 0.0236, 95% CI = [0.0031, 0.1005]) (Fig. 3c). In contrast, for Y-BOCS obsession scores, only sulcal depth in the right SMG exhibited a significant indirect effect (indirect effect = 0.0345, SE = 0.0207, 95% CI = [0.0044, 0.0892]), while the gray matter volume in the right-lateralized thalamus proper was not significant (indirect effect = 0.0210, SE = 0.0193, 95% CI = [–0.0074, 0.0719]) (Fig. 3d). Conversely, for Y-BOCS compulsion scores, only gray matter volume in the right-lateralized thalamus proper showed a significant indirect effect (indirect effect = 0.0407, SE = 0.0273, 95% CI = [0.0032, 0.1147]), whereas sulcal depth in the right SMG pathway was not significant (indirect effect = 0.0233, SE = 0.0200, 95% CI = [–0.0023, 0.0796]) (Fig. 3e).

In summary, these findings suggested distinct brain morphometric correlates of the indirect effects linking BIS to OCD symptoms. Only the right SMG showed a significant specific indirect effect for obsession scores and only the right-lateralized thalamus proper showed a significant specific indirect effect for compulsion scores.

Clinical interventions for BIS/BAS system

A total of 74 patients with OCD underwent 12-week first-line treatments, including 42 with SSRIs and 32 with CBT. Descriptive statistics of demographic, psychopathological and BIS/BAS characteristics for these patients are shown in Table S4, with no significant differences between the treated subgroup and the remaining OCD cohort (Supplementary S9, Table S5). All psychopathological characteristics were significantly decreased following treatments (p < 0.001; Supplementary S9, Table S6). Confirmatory analyses within the treated subgroup showed patterns broadly consistent with the main behavioral findings (Supplementary S10, Table S7). It should be noted that we only included participants who completed BIS/BAS scores pre- and post-treatment. This study should not be construed as a comparative test of SSRIs and CBT effectiveness and all symptom scores should only be regarded as the description of psychopathological characteristics.

For BIS, scores decreased significantly from pre- to post-treatment in the SSRI group (t₄₁ = –2.544, p = 0.015, Cohen’s d = –0.373), but did not change following CBT (t₃₁ = 1.109, p = 0.276, Cohen’s d = 0.174; Fig. 4a and Table S8). Furthermore, changes in BIS scores were significantly correlated with changes in Y-BOCS total scores following SSRI treatment (r = 0.338, p = 0.029; Fig. 4b).

Fig. 4. Changes in BIS/BAS following SSRI and CBT Treatments.

a Pre-post changes in BIS scores in the SSRI and CBT groups. b Association between pre-post changes in BIS scores and pre-post changes in Y-BOCS total scores in the SSRI group. Pre-post changes in c BAS, d BASR, e BASD, and f BASF scores in the SSRI and CBT groups (^**p < 0.01; ^*p < 0.05; ^n.sp ≥ 0.185). BIS, behavioral inhibition system; BAS, behavioral activation system; BASR, BAS reward responsiveness; BASD, BAS drive; BASF, BAS fun seeking; CBT, cognitive behavioral therapy; SSRI, selective serotonin reuptake inhibitor; Y-BOCS, Yale-Brown Obsessive-Compulsive Scale.

For BAS, BAS total scores increased significantly following CBT (t₃₁ = 3.133, p = 0.004, Cohen’s d = 0.410), but showed no significant change following SSRI treatment (t₄₁ = –0.394, p = 0.695, Cohen’s d = –0.052; Fig. 4c and Table S8). However, no significant correlation was found between changes in BAS total scores and Y-BOCS total scores pre- and post-CBT treatment (r = 0.012, p = 0.946). For BAS subscales, no significant pre- to post-treatment change was observed in BAS reward responsiveness in either the SSRI group (t₄₁ = –1.349, p = 0.185, Cohen’s d = –0.231) or the CBT group (t₃₁ = 1.256, p = 0.218, Cohen’s d = 0.221; Fig. 4d and Table S8). In contrast, both BAS drive and BAS fun seeking increased significantly following CBT (drive: t₃₁ = 2.273, p = 0.030, Cohen’s d = 0.349; fun seeking: t₃₁ = 2.830, p = 0.008, Cohen’s d = 0.447), whereas no significant change was observed following SSRI treatment (drive: t₄₁ = 0.788, p = 0.435, Cohen’s d = 0.103; fun seeking: t₄₁ = –0.431, p = 0.669, Cohen’s d = –0.053; Fig. 4e, f and Table S8). Changes in neither BAS drive nor BAS fun seeking were significantly correlated with changes in Y-BOCS total scores within the CBT group (both p ≥ 0.314).

For exploratory purposes, we examined whether baseline BIS/BAS sensitivity was associated with Y-BOCS total score change rate, and no significant correlations were found (SSRIs: p ≥ 0.490; CBT: p ≥ 0.241).

Discussion

In this study, we identified brain morphometric correlates of BIS/BAS abnormalities in OCD and explored whether BIS/BAS changed following first-line treatments in OCD. Patients with OCD exhibited elevated BIS and reduced BAS drive sensitivity, with BIS correlating with symptom severity. At the neural level, gray matter volume in the right-lateralized thalamus proper was negatively associated with BIS in OCD but positively associated in HC, whereas sulcal depth in the right SMG showed a positive association with BIS across both groups. Parallel mediation analyses indicated distinct indirect association patterns linking BIS to OCD symptom dimensions, with thalamic gray matter volume linked to compulsive behaviors and sulcal depth in the right SMG linked to obsessive thoughts. At the level of longitudinal intervention, BIS decreased following SSRI treatment, with changes in BIS following SSRI treatment positively associated with improvements in OCD symptoms, whereas BAS increased following CBT. We believe this study provides preliminary insights into how BIS/BAS contribute to OCD psychopathology.

Consistent with prior studies [9, 11, 13, 15], our findings were more strongly related to BIS than to BAS. The SMG and thalamus, both components of the ventral attention network, are involved in the integration and processing of sensory information [37]. The different associations of the SMG and thalamus with obsession and compulsion may be attributable to their distinct roles within the ventral attention network. The right SMG processes and integrates sensory information from the external environment [38], which may be associated with intrusive thoughts elicited by external threat-related cues. By contrast, the thalamus processes interoceptive signals [39], which may be associated with compulsive behaviors driven by internally aversive or inappropriate feelings. Moreover, the BIS-thalamus association showed a significant difference between OCD patients and healthy controls, possibly reflecting an aberrant internal response to external threat cues in OCD.

Beyond the cross-sectional analyses of BIS/BAS in OCD, we further examined their associations with symptom improvement in a longitudinal intervention study. The close associations between BIS/BAS and first-line treatments for OCD further reflect their contribution to OCD psychopathology. BIS decreased following SSRI treatment, and its reduction was associated with OCD symptom improvement. SSRIs can reduce excessive attention to threat cues [40]. Although our study did not investigate responses in BIS-associated brain regions, including the SMG and thalamus, following SSRI treatment, prior research has indicated that SSRIs may alter the responses of these regions to threat cues [41]. In addition, BAS, especially the fun seeking dimension, increased following CBT. CBT for OCD facilitates safety learning and fear extinction via exposure and response prevention [42], potentially engaging reward-seeking processes. It should be emphasized that the potential differential relationships of BIS and BAS with SSRIs and CBT interventions should be interpreted cautiously. Because the study was not designed as a randomized controlled trial, these findings do not support causal inference regarding treatment-specific modulation of BIS/BAS.

The BIS-related behavioral and brain associations were statistically significant but modest in magnitude, with correlation coefficients ranging from 0.18 to 0.34. According to Cohen’s conventional benchmarks, the upper end of the present associations approached the medium range (0.30 as medium effects) [43]. Nevertheless, empirical benchmarks from psychological research indicate that correlations around 0.20 are common and meaningful [44], but should still be interpreted cautiously, as reflected in the focus of the present study on BIS as part of a broader brain-symptom association pattern. These results also suggest that the association between BIS and OCD may be shaped by a broader range of factors, which remains to be further explored.

Overall, our findings highlight a close association between BIS and OCD across both cross-sectional and longitudinal dimensions, and provide a possible dual-pathway framework for understanding the BIS, ventral attention network, and OCD symptoms. Together, these findings offer new insights into how heightened sensitivity to threat may contribute to OCD psychopathology. Although BIS/BAS sensitivity may represent a shared vulnerability across psychiatric disorders, its neural correlates may vary in a disorder-specific manner. Unlike findings in other psychiatric disorders, which have primarily emphasized reward-related or executive-control networks [21–25], our study points to a potential role of the ventral attention network in BIS abnormalities in OCD. To the best of our knowledge, few studies have reported associations of the SMG and thalamus with BIS/BAS in psychiatric samples, thereby suggesting that the neural correlates of BIS abnormalities in OCD may differ from those observed in other disorders. Taken together with previous research, our findings are consistent with the view that BIS/BAS may constitute a vulnerability factor across multiple psychiatric disorders, including OCD, although the underlying neural substrates may vary in a disorder-specific manner. However, all of these findings should be regarded as preliminary and correlational, and future longitudinal and experimental studies are needed to clarify the temporal and causal significance of these associations.

Limitations

This study has several important limitations to consider. Firstly, although all patients were medication-free for at least 8 weeks prior to enrollment, detailed records of their previous treatments were unavailable, and self-reported medication histories could not be fully verified. The 8-week washout period was intended to minimize residual drug effects [45]; however, potential changes in BIS/BAS associated with prior medication may not have fully reverted to baseline. Secondly, the clinical intervention part should not be considered a randomized controlled study, and the original study was not preregistered, as preregistration had not yet become common practice in 2014. In addition, the treatment analyses were based on a smaller subgroup, which may have limited statistical power. Therefore, these findings should be considered preliminary. Thirdly, the neural conclusions of this study were derived from cross-sectional analyses, and longitudinal research is still needed to clarify the potential causal relationships between these neural metrics and BIS in OCD. Finally, we only included self-report scales to assess BIS/BAS. Although such measures are well-validated, further inclusion of behavioral or computational measures of reward and punishment sensitivity would have strengthened construct validity and translational implications.

Conclusion

In summary, our study integrates self-report measures, neuroimaging, and clinical interventions to explore how BIS/BAS contribute to OCD psychopathology. The results highlight a close association between BIS and OCD psychopathology across cross-sectional and longitudinal analyses and further reveal a possible dual-pathway model linking BIS, the ventral attention network, and OCD symptoms.

Author contributions

Chen Zhang: Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Visualization, Writing-original draft, Writing-review and editing. Binyu Xiang: Data curation, Formal analysis, Methodology, Visualization, Writing-original draft, Writing-review and editing. Zongfeng Zhang: Data curation, Funding acquisition, Investigation, Project administration. Yongjie Li: Formal analysis. Jiajia Zhou: Conceptualization, Visualization. Yongjun Chen: Investigation. Rui Gao: Investigation. Xuan Cao: Investigation. Xiaochen Zhang: Formal analysis, Writing-review and editing. Qing Fan: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing-original draft, Writing-review and editing.

Funding

Qing Fan was supported by the National Natural Science Foundation of China (82571766), Shanghai 2024 “Science and Technology Innovation Action Plan” Technical Standard Project (24DZ2201400), Project of Yuanshen Rehabilitation Institute, Shanghai Jiao Tong University School of Medicine (yskf1-23-1107-1), Collaborative Innovation Cluster Project of Shanghai Municipal Health Commission (2024CXJQ03), Shanghai Hospital Development Center Foundation—Shanghai Municipal Hospital Rehabilitation Medicine Specialty Alliance (SHDC22023304), Key Scientific and Technological Project of Chinese Association of Rehabilitation Medicine. Xiaochen Zhang was supported by the National Natural Science Foundation of China (82571756, 32100885), Natural Science Foundation of Shanghai (25ZR1402450). No funding source mentioned above has any form of involvement in study design, data collection, data analyses, data interpretation, report writing, or the decision to submit the article for publication.

Data availability

Because data of psychiatric patients are highly sensitive, data sharing requires individual evaluation by the Ethics Committee of the Shanghai Mental Health Center. Data can be accessed upon reasonable request to QF, all requests will be evaluated case by case.

Code availability

All data analyses used standard pipelines from SPM12 and CAT12, SPSS (version 24) and MATLAB (version R2018b) implementations of openly available algorithms.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41398-026-04207-y.