The Occipital-Fronto-Limbic Circuit of Body Dysmorphic Disorder: A Systematic Review and Meta-Analysis of Neuroimaging Studies

Abstract

Background

Body dysmorphic disorder (BDD) is a severe psychiatric disorder characterized by excessive preoccupation with perceived flaws in appearance. In this review, we aim to synthesize the latest findings on the neurobiology of BDD and propose an updated BDD neurocircuitry model.

Methods

A systematic search identified 38 peer-reviewed original articles, and the protocol was registered (CRD42024553665). Qualitative thematic analysis was conducted, and activation likelihood estimation (ALE) meta-analysis was performed to quantify brain activation patterns.

Results

Brain regions with significant morphometric divergence in patients with BDD were predominantly located in primary and secondary visual processing areas and temporal-limbic and frontal-striatal networks despite overall heterogeneous findings. The electroencephalography studies suggested early visual processing and attentional abnormalities. The ALE analysis revealed a general hyperactivation over the frontotemporal region and hypoactivation over parieto-occipital regions. Although BDD shared similar connectivity patterns in frontostriatal and arbitration networks with obsessive-compulsive disorder, it was further characterized by bottom-up and top-down interaction between the ventral visual stream and temporal-limbic network compared with anorexia nervosa. Neurotransmitters such as serotonin, dopamine, and oxytocin play a key role in the pathophysiology of BDD, suggesting a complex interplay of neural circuits and neurotransmitters underlying the disorder.

Conclusions

This comprehensive review of up-to-date neurobiological studies of individuals with BDD reveals differences in brain structure and functionality compared with control participants. The proposed neurocircuitry model expands on the previous understanding of BDD neurobiology and elucidates the interconnection between the visual processing, temporal-limbic, and frontostriatal networks and their clinical implications. This review provides theoretical support for future neuromodulation target identification.

Plain Language Summary

This systematic review and meta-analysis brings together the latest neurobiological research on body dysmorphic disorder (BDD), summarizing findings in brain morphometry, activation, connectivity, white matter networks, and functional neurochemistry. Building on these findings, the authors propose an updated neurobiological model—the occipital-fronto-limbic circuit—which illustrates how differences in brain activity and connectivity within visual, emotional, and executive regions may shape how individuals with BDD perceive their appearance.

Plain Language Summary

This systematic review and meta-analysis brings together the latest neurobiological research on body dysmorphic disorder (BDD), summarizing findings in brain morphometry, activation, connectivity, white matter networks, and functional neurochemistry. Building on these findings, the authors propose an updated neurobiological model—the occipital-fronto-limbic circuit—which illustrates how differences in brain activity and connectivity within visual, emotional, and executive regions may shape how individuals with BDD perceive their appearance.

Body dysmorphic disorder (BDD) is a severe neuropsychiatric disorder characterized by obsessive concerns about perceived appearance flaws and associated compulsive behaviors such as rechecking and camouflaging (1). More than 20% of patients experience delusional visual distortion of their appearance, often with poor insight (2,3). Although the lifetime prevalence is 1.7% to 2.9%, BDD is often underdiagnosed (4, 5, 6, 7). The lifetime suicide attempt rate can reach 25% among those diagnosed (8,9). Classified under the obsessive-compulsive and related disorder (OCRD) spectrum, BDD shares clinical and neuroimaging features with obsessive-compulsive disorder (OCD) (1). Although emerging OCD treatments, such as deep brain stimulation (DBS), are ongoing (10, 11, 12, 13), BDD treatment remains limited to conventional psychotherapy and medications due to limited understanding of its neurobiology (14,15).

A recent case report showing significant BDD symptom relief using ventral capsule/ventral striatum DBS suggested its potential for treatment-resistant patients with BDD and the importance of understanding BDD neurocircuitry to select effective modulation/surgical targets (16). Grace et al. (17) outlined BDD’s neurobiology, and recent neuroscience studies using advanced techniques have added further insights. We conducted a systematic review of neuroimaging and functional neurochemistry studies and performed a meta-analysis to examine BDD brain activation patterns, proposing an updated BDD neurocircuitry model.

Methods and Materials

Search Strategy and Selection Criteria

All studies regarding neuroimaging and functional neurochemistry research in BDD were considered. Studies were included if they met the following eligibility criteria: full-text availability in English, publication in a peer-reviewed journal, BDD diagnosed per DSM criteria (1), and demographically matched healthy control (HC) participants. Exclusion criteria included systematic reviews, meta-analyses, case reports, book chapters, conference abstracts, studies on subclinical BDD, or studies involving BDD participants without subsetting a specific BDD cohort.

The literature was searched using the PubMed/MEDLINE, PsycInfo, and Embase databases. Reference lists of relevant articles were also examined. The final search was conducted on May 15, 2025, with no date restrictions. Search terms were adapted from Grace et al. (17) with modifications: ((body dysmorphic disorder [MeSH Terms]) OR “body dysmorphic disorder” OR “body dysmorphia” OR “BDD” OR “body dysmorphic∗”) AND (neuro∗ OR brain OR neurobiology OR neuroimaging OR EEG OR electroencephalography OR MEG OR magnetoencephalography OR SPECT OR “single-photon emission computed tomography” OR PET OR “positron emission tomography” OR MRI OR fMRI OR “magnetic resonance imaging” OR DTI OR “diffusion tensor imaging” OR functional OR structural OR connectome OR network). Two authors independently screened titles and abstracts, and full texts of selected studies were reviewed for eligibility. The online tool Covidance (available at https://www.covidence.org/) was used for literature screening. A data extraction sheet was developed and refined to simplify the organization of information. All data extracted were confirmed twice to avoid errors. Studies from the same research group were carefully analyzed to identify repetitive recruitment of participants with BDD.

Risk of bias was assessed using the Newcastle-Ottawa Scale or the Cochrane Collaboration (18). The review followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines and was registered on PROSPERO (CRD420024553665) in June 2024 (19).

Synthesis of Results

A qualitative and quantitative approach was used to synthesize results for the review.

The activation likelihood estimation (ALE) meta-analysis was conducted to assess brain activation (20, 21, 22). All functional magnetic resonance imaging (fMRI) activity studies that used whole-brain activity analysis for image-matching tasks were included. The ALE algorithm, implemented in GingerALE 3.0.2 (available at: https://brainmap.org/tools.html), was used to perform meta-analysis based on extracted activation coordinates, standardized to Montreal Neurological Institute space.

The ALE analysis used a random-effects method that considered the uncertainty of the spatial distribution of the loci of activation. The uncertainty was quantitatively assessed based on the sample sizes of individual studies. This generated a modeled activation map for the activation coordinates, which was visualized using Mango 4.1 (available at: https://mangoviewer.com/). Voxelwise scores reflected the probability of activation across studies (20,23). The analysis was adjudicated at a cluster-level familywise error–corrected threshold of p < .05, using 1000 threshold permutations and the cluster significance set at a corrected p value of .05.

Results

A comprehensive search resulted in 3652 articles after deduplication. Of these, 38 met the predefined eligibility criteria and were included for data extraction, as detailed in a selection flowchart (Figure 1). Two articles from 2024 were excluded because they either lacked HC participants or the HC participants were not demographically matched (24,25). One article from 2025 was excluded due to not having a dedicated BDD subcohort in the analysis (26). Selected articles published from 2003 to 2023 primarily used MRI to study BDD neurocircuitry, with 27 using various MRI techniques. Electroencephalography (EEG) was also used to track temporal brain activity changes, enhancing the analysis of brain function dynamics in BDD. Some studies included patients with OCD or anorexia nervosa (AN) as the disease control group to explore BDD-specific neurocircuitry.

Figure 1.

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram of the article screening and selection process. HC, healthy control participants.

These articles involved 1716 instances of participation, 816 from patients with BDD mainly concerned with facial flaws, with some repeated recruitments noted across studies (see Table S1) (27,28). Participants were predominantly female, ages 17 to 65 years, mostly medication-free, and not receiving cognitive-based therapy at the time of the study (Table S2). Many had comorbid mental disorders, such as major depressive disorder and generalized anxiety disorder, with BDD as the primary diagnosis. Overall, the quality of the selected studies was good (Table S3).

Table S4 details the methodology and summarizes findings from 7 MRI studies, with sample sizes between 8 and 65, totaling 370 participants—186 patients with BDD and 184 control participants. Five studies examined brain volume, one examined cortex thickness, and one examined both. Findings on brain volume, cortical thickness, and clinical correlations in BDD varied across studies. Table 1 lists major findings of these studies.

Table 1.

Brain Regions With Significant Morphometric Abnormalities or Correlated With Clinical Symptom Severity in Participants With BDD Compared With Healthy Control Participants

Brain Regions	Measurements	Changes
Brain Regions Affected in BDD
Total cerebrum, total WMa, total GMb, bilateral early extrastriatal cortex, right precuneusb, right OFCb, left ACC, left V4v, and left thalamus (trend)	Volume	Increased
Total GMb, left OFC, left ACC, right ACC, right thalamus, right OFCb, right precuneusb, right SPL, and bilateral fusiform gyri (trend)		Reduced
Left occipital, parietal, and frontal regions (trend)	Cortical thickness	Increased
Left inferior parietal and middle temporal region		Reduced
Brain Regions Correlated With Clinical Features (Changes Relate to Volume Unless Otherwise Specified)
Bilateral amygdalab and left IFG	BDD-YBOCS	Positive
Bilateral amygdalab and left early extrastriate GM		Negative
Left amygdala	HAM-D	Positive
Left superior temporal cortex GM thickness	HAM-A	Negative
Left OFC and right OFCa	Length of illness	Negative

The early extrastriatal GM included V2, V3d, V3v/VP, V3a, and V4v.

ACC, anterior cingulate cortex; BDD, body dysmorphic disorder; BDD-YBOCS, Yale-Brown Obsessive Compulsive Scale modified for body dysmorphic disorder; GM, gray matter; HAM-A, Hamilton Anxiety Rating Scale; HAM-D, Hamilton Depression Rating Scale; IFG, inferior frontal gyrus; OFC, orbitofrontal cortex; SPL, superior parietal lobule; WM, white matter.

^aFindings that were reported in at least 2 studies using different participant pools.

^bFindings that were contradictory between at least 2 studies using different participant pools.

An increase in global brain volume, including both white and gray matter (GM), has been reported in patients with BDD, although a reduction in total GM has also been noted (28, 29, 30). The right orbitofrontal cortex (OFC), precuneus, and left anterior cingulate cortex (ACC) are major contributors to this GM reduction (31). Atmaca et al. (29) reported decreased bilateral OFC and ACC volumes, similar to findings in OCD, where reduced OFC volume is common (32,33). Increased cortical thickness has been observed in the left occipital, parietal, and frontal regions, whereas reductions were seen in the left inferior parietal and middle temporal regions (27,32). Scores on the Yale-Brown Obsessive Compulsive Scale modified for body dysmorphic disorder (BDD-YBOCS), used to assess clinical BDD severity, were positively correlated with volumes in the left inferior frontal gyrus (IFG) and bilateral amygdala and negatively with cortical thickness in early extrastriate regions (28,32,34). Depression severity (measured by the Hamilton Depression Rating Scale) was positively correlated with left amygdala volume, whereas anxiety severity (measured by the Hamilton Anxiety Rating Scale) was negatively correlated with GM cortical thickness (34,35). Notably, bilateral OFC volume showed a negative correlation with illness duration, independent of age, across 2 BDD cohorts (28,29).

Significant findings in BDD centered on the frontal (OFC, ACC, IFG), temporal-limbic (amygdala, IFG), and visual processing (V4v, superior parietal lobule [SPL]) networks. However, small sample sizes (N ≤ 20) in most studies and repetitive recruitment of participants in Feusner et al. raise concerns about false positives. Inconsistencies across studies, such as diverse patient selection criteria, contributed to heterogeneity (29,30,32,35). For example, some studies overrepresented male patients with BDD despite equal prevalence in both sexes. Increased white matter volume may indicate developmental deficiencies, either from increased myelin per fiber or from glial proportion (32). Overall, these findings suggest that BDD lacks consistent morphological abnormalities.

Fifteen fMRI studies, primarily conducted at the University of California, Los Angeles, investigated brain activation and connectivity in 368 patients with BDD and 374 HC participants (Tables S5 and S6). These studies explored brain activation in the dorsal visual stream (DVS) and ventral visual stream (VVS) by presenting images of varying spatial resolutions of strangers’ faces, participants’ faces versus familiar faces, strangers’ bodies, and objects (36, 37, 38, 39). Some focused on temporal-limbic circuits modulating the emotional impact on visual processing, using participants’ faces, familiar faces, and fearful or negative emotional faces (40, 41, 42). Four studies also used psychiatric cohorts (AN and OCD) as disease controls to investigate BDD-specific brain abnormalities (42, 43, 44). Low spatial frequency (LSF) images aided holistic processing, whereas high spatial frequency (HSF) images facilitated detail processing, with normal spatial frequency images as controls. Prior research showed that HSF and LSF images engage distinct neural pathways, with HSF directed to the VVS for detailed analysis and LSF directed to the DVS for rapid, coarse processing (45).

Four studies, covering 74 patients with BDD and 53 foci, were included in the ALE-based meta-analysis (Table S5) (36, 37, 38,43). The study interventions involved visual stimuli such as strangers’ faces, familiar faces, bodies, and houses across different spatial frequencies. The ALE meta-analysis (Figure 2, Table 2) identified 8 significant clusters in the left hemisphere, with patients with BDD showing hyperactivation in the frontotemporal region (including the left IFG, claustrum, caudate, and superior temporal gyrus) and hypoactivation in the parieto-occipital region (left lingual gyrus, precuneus, cuneus, and SPL). This pattern indicated hypoactivity in primary visual processing and DVS areas, whereas hyperactivity occurred in frontotemporal higher-order regions. Bohon et al. (40) also used the VVS (e.g., fusiform cortex) as seeds for the region-of-interest analysis, revealing the right amygdala’s mediation effect on anxiety and VVS activity, upon the finding of a positive linear relationship between VVS and amygdala activity in both participants with BDD and HC participants. This highlighted a link between emotional face processing and the amygdala in patients with BDD.

Figure 2.

The visualized activation likelihood estimation analysis results of activated brain regions in patients with body dysmorphic disorder compared with healthy control participants when performing image-matching tasks. Hypoactivated clusters (including the left lingual gyrus, precuneus, cuneus, and superior parietal lobule) were green, whereas hyperactivated clusters (including the left inferior frontal gyrus, claustrum, and caudate) were orange.

Table 2.

ALE Meta-Analysis Results of Functional Magnetic Resonance Imaging Studies of Brain Activation for Visual Stimuli

	Cluster No.	x	y	z	ALE	p Value	z Score
Hyperactivation	1	−26	28	−18	0.008040248	4.97 × 10−05	3.8922446
	1	−10	16	2	0.008036619	4.97 × 10−05	3.8922446
	1	−30	18	−6	0.007146488	9.92 × 10−05	3.7211502
	1	−44	22	−20	0.007144364	1.44 × 10−04	3.6258767
Hypoactivation	1	13	−68.5	41	0.002630502	.007399478	2.4372618
	1	−20	−43.5	1	0.001997047	.010921425	2.2930896
	1	−15	−80	11	2.43E-04	.035349425	1.8074069
	2	−27.6	−67.2	51.2	0.002303292	.009116248	2.3608623

ALE, activation likelihood estimation.

Brain regions involved in various networks have been examined in connectivity studies (Table S6). Studies of higher-order visual processing, particularly in the temporal-limbic network, showed shared abnormalities between BDD and AN when exposed to strangers’ faces or fearful stimuli (39,42,43). Leow et al. (46) and Borgers et al. (41) explored the correlation between visual processing, the temporal-limbic network, and anxiety in BDD. Patients with BDD showed reduced information transfer between primary and secondary visual cortices and within the temporal lobe when exposed to LSF visual stimuli (46). Additionally, occipitotemporal face processing abnormalities were found to be BDD specific, independent of AN disease controls (39). Similar findings were reported by Moody et al. (43). BDD and AN shared connectivity patterns in the fusiform face area and higher-order regions, potentially explaining shared body image dissatisfaction traits (39). Borgers et al. (41) found increased amygdala and fusiform gyrus sensitivity to negative emotional stimuli in BDD, indicating enhanced negative emotional responses compared with HC participants. Rangaprakash et al. (42) revealed that patients with BDD, although retaining top-down connectivity, experience impaired bottom-up signaling, particularly between the medial prefrontal cortex (mPFC), rostral ACC (rACC), and amygdala, which may account for temporal-limbic hyporesponsiveness under anxiety (37). Fronto-striatal and arbitration-habitual systems were also examined in BDD, OCD, and HC cohorts. Participants with BDD and participants with OCD showed hyperactivity in the posterior-lateral OFC, with the hyperactivity in BDD leaning toward the posterior-central OFC. This OFC activity was correlated with symptom severity, suggesting shared frontostriatal network disruptions (47). Additionally, altered resting-state connectivity between the ventrolateral PFC and posterolateral putamen in both BDD and OCD correlated with obsessive-compulsive symptoms, pointing to a shared arbitration system abnormality (44).

Four studies investigated white matter networks, summarized in Table S7, involving 77 participants with BDD and 83 HC participants. All used the diffusion tensor model to analyze the diffusion-weighted images. Arienzo et al. (48) reported a higher mean clustering coefficient in BDD, suggesting abnormal network organization or information processing. This trend was weak in medication-naïve patients but significant in medicated ones, with a negative correlation between global efficiency and BDD-YBOCS scores. Buchanan et al. (31) reported lower fractional anisotropy (FA) in BDD, especially in major white matter tracts, indicating compromised white matter integrity. This was independent of BDD-YBOCS scores, but FA in the left superior longitudinal fasciculus was negatively correlated with a social interaction anxiety score. This reduction in FA was linked to possible white matter neurodevelopmental insufficiency and genetic predisposition. Feusner et al. (49) reported a similar FA reduction pattern. A connectome study also revealed abnormal modular involvement of the mPFC with striatal-cingulate coupling in both AN and BDD compared with HC participants (50).

Three studies have investigated BDD’s functional neurochemistry, analyzing neurotransmitter impacts on brain connectivity and symptom severity (Table S7). Fang et al. (51) observed that fasting serum oxytocin levels were elevated in patients with BDD more than in HC participants and correlated positively with BDD-YBOCS scores, linking this to stress response. Grace et al. (52) found that nasal oxytocin reduced the left amygdala’s connectivity to left frontotemporal regions in participants with BDD, in contrast to HC participants, suggesting oxytocin’s role in modulating key VVS regions and consistent with noted left hemisphere hyperactivation in BDD. However, the studies were limited by small samples and lacked behavioral correlations. Vulink et al. (53) reported reduced striatal dopamine D₂/D₃ receptor availability and binding potentials in BDD, similar to OCD (54, 55, 56), supporting the hypothesis of ventral striatal dysfunction and the effectiveness of selective serotonin reuptake inhibitors in OCRDs and Tourette syndrome (53,57,58). Zhu et al. (59) reported that psilocybin increased executive control network connectivity, providing symptom relief. These findings highlight neurotransmitter imbalances, including oxytocin, dopamine, and serotonin, in BDD’s pathology but underline the importance of more comprehensive studies to confirm clinical implications due to methodological and sample size limits.

Table S8 summarizes the methodologies of 8 EEG studies, comprising 98 BDD and 94 HC participation instances. Table S1 details the repeated patient recruitments across studies. Four studies examined event-related potentials (ERPs), with one integrating ERPs and fMRI and another combining ERPs with global field power (GFP). The remaining 3 studies focused on GFP and behavioral changes. When participants performed matching tasks for images of human faces at varying spatial frequencies, lower N170 amplitudes were correlated with poor insight, suggesting that reduced face encoding was related to more perceptual distortions in patients with BDD (60). Additionally, combined fMRI and EEG data indicated DVS hypoactivity (including precuneus, SPL, and lateral occipital cortex) in BDD and AN, supporting deficits in holistic processing in primary visual areas starting from 100 ms and extending to dorsal areas by 170 ms (36,61). However, this was not seen with face inversion (62). Compared with patients with AN, patients with BDD displayed a greater imbalance in detail versus holistic processing without differences in detail processing (61).

In auditory studies, patients with BDD exhibited unique brain activation patterns, including higher theta-1 and lower beta-1 oscillations, suggesting divergent attentional mechanisms from HC participants (63,64). ERPs analysis revealed reduced N100 amplitudes following auditory startle stimuli (63). Giannopoulos et al. (65) explored illusory perception and brain activity in BDD, finding that these patients had lower confidence in incorrect decisions and reduced alpha-1 power in frontocentral and parietal regions during optical illusion tasks. The reduction in alpha-1 power, linked to difficulty inhibiting distractions, was negatively correlated with BDD-YBOCS scores (66,67). Additionally, reduced directed connectivity from the left to the right hemisphere in BDD suggested potential reductions in interhemispheric communication (68). However, no differences in susceptibility to optical illusions were detected. Deficiencies in holistic visual processing across various stimuli were found in patients with BDD, indicating a widespread abnormality in their visual processing network, although their optical illusion judgment remained unaffected. These patients also exhibited abnormalities in attentional orientation and allocation. A limitation of EEG studies is their limited spatial resolution. Li et al. (61), which combined EEG and fMRI, did not standardize intervals between these 2 modalities, introducing potential bias. Additionally, participant overlap in studies from the same institute may have overrepresented the results. Overall, patients with BDD showed visual processing and higher cognitive function aberrations.

Discussion

This systematic review and meta-analysis on brain structure and activity in BDD expands our understanding of its neurobiology with up-to-date findings in brain morphometry, activity, connectivity, and neurotransmitter activity. The broad range of imaging modalities revealed an underlying abnormality in primary and secondary visual processing as well as temporal-limbic and frontostriatal networks compared with HC and disease (notably OCD) control groups. The distinctive body image visual processing imbalance in addition to overlap in frontostriatal networks with OCD and sharing OCD traits raised questions about whether, under the broad spectrum of OCRD, BDD is a form of OCD with specific provocation from visual-sensory disturbances. The integration of qualitative and quantitative analysis allows us to interpret BDD functional and structural findings in the context of a wider scheme. However, due to the nature of the cross-sectional design in most of the studies, the cause-and-effect relationships between observed neuroimaging findings and BDD clinical symptoms need further investigation. Nonetheless, the abnormal networks involved in BDD pathophysiology have potential clinical implications.

Our findings expanded and detailed a current model of BDD comprising large neural networks involving visual, limbic, and frontostriatal system dysfunction (17,69). Furthermore, we described the relationship between the visual, limbic, and frontostriatal systems and the interplay between them, how this was correlated with the clinical symptoms of BDD, and potential therapeutic interventions (Figure 3A). BDD is a complex disorder involving cognitive and emotional dysfunction (Figure 3B). However, more subtle cognitive abnormalities, including attention orientation and allocation, also participate in the neurocircuitry of BDD (63,65,70, 71, 72).

Figure 3.

(A) Schematic representation of the correlation and interplay between the visual, temporal-limbic, and frontal areas described in the studies. The key regions in the occipital-fronto-limbic network include the fusiform gyrus, amygdala, prefrontal cortex (PFC), and orbitofrontal cortex (OFC). Meanwhile, regional activity modulation interventions were described above, including nasal oxytocin, deep brain stimulation (DBS), and noninvasive transcranial magnetic stimulation. (B) Schematic representation of the connectivity and activation aberrance between body dysmorphic disorder (BDD) brain regions from the connectivity studies and the activation likelihood estimation analysis. A general hypoconnectivity and hypoactivation located around the primary and secondary visual processing areas. Complex connectivity changes were found around temporal-limbic and prefrontal areas. BDD-YBOCS, Yale-Brown Obsessive Compulsive Scale modified for body dysmorphic disorder; HC, healthy control; mPFC, medial prefrontal cortex; rACC, rostral anterior cingulate cortex; SPL, superior parietal lobule; vlPFC, ventral lateral prefrontal cortex; VC/VS, ventral capsule/ventral striatum.

The DVS and VVS involvement in visual processing in BDD have been well summarized by Grace et al. (17), illustrating the DVS’s involvement in whole picture processing, engaging regions including the superior occipital gyrus, SPL, and precuneus. The VVS focused on detail processing, involving regions including the inferior occipital gyrus, lingual gyrus, fusiform gyrus, and inferior temporal gyrus (73). The observed imbalance and hypoactivation within the VVS and DVS may underlie the preoccupation with visual details in patients with BDD. Importantly, these visual processing abnormalities extend beyond the facial or bodily stimuli to the object, manifesting shortly after exposure to visual stimuli. This suggests that such visual processing abnormalities constitute a primary dysfunction, likely functioning independently of the higher-order cognitive system. Reports of altered brain connectivity and activation patterns in response to object visual stimuli in BDD further supported this hypothesis (36,61). Thus, we hypothesize that facial/body visual processing and higher-order, top-down visual processing may be the triggers for BDD clinical presentation, including body image dysmorphia and obsessive-compulsive behaviors, against a background of imbalanced holistic and detail processing. This hypothesis was supported by a significant improvement in body image scores after theta burst stimulation at the lateral parietal cortical regions in patients with BDD. This indicated that DVS activity modulation was associated with improvement in body image dissatisfaction. Furthermore, pretherapy BDD-YBOCS scores were found to be significantly negatively correlated with DVS region activity, which implied that DVS engagement may be directly associated with BDD symptom severity (25). Besides brain stimulation, visual attention modulation, by encouraging more holistic viewing in patients with BDD, was also associated with insight improvement and had a carryover effect for later natural viewing (73). These preliminary modulations of the DVS activity suggested potential clinical significance. However, a possible causal relationship between DVS modulation and BDD symptom improvement requires further scrutiny.

The fusiform gyrus and amygdala are 2 key regions responsible for facial detail processing and negative emotional arousal. The occipital-fronto-limbic network was shown to be relevant to BDD-specific neurocircuitry, involving regions such as the fusiform face area, amygdala, mPFC, and OFC (Figure 3A, B). Bohon et al. (40) first noted that the VVS exhibited hyperconnectivity with the amygdala in BDD. The right amygdala not only mediated anxiety but also predicted VVS activity, suggesting possible emotional modulation of face processing in BDD. Comparative analysis revealed that similar to BDD, AN exhibited abnormal connectivity between higher-order regions and the right fusiform gyrus (39). However, BDD distinctively showed disruptions in the occipitotemporal face processing network, suggesting a closer interplay between the temporal-limbic system and occipital visual processing in BDD. The fusiform gyrus, particularly the right fusiform facial processing area, may serve as a crucial node for facial visual input to the temporal-limbic network, bridging visual processing and higher-order visual-emotion modulation. Furthermore, patients with BDD demonstrated dysfunctional bottom-up, feed-forward signaling in the frontolimbic system. The left amygdala exhibited hyperactivation and hyperconnectivity with the right frontal lobe and basal ganglia regions in response to negative emotional stimuli (42). Oxytocin, an important neuromodulator in the amygdala, could increase anxiety levels and social inhibition by enhancing amygdala activity (74). Grace et al. (52) reported that a single dose of nasal oxytocin could modulate resting-state amygdala-temporal lobe connectivity. However, no measurable therapeutic effects on symptoms were reported in this study, and others have raised concerns over possible negative effects of oxytocin in BDD (51). This raised questions about whether amygdala hypoconnectivity is a cause or a consequence of the detailed preoccupation and obsessive-compulsive symptoms in BDD. This gap suggests a direction for future research in exploring how oxytocin may modulate amygdala connectivity and activity under negative emotion arousal in patients with BDD.

A shared neural pathway in frontostriatal and arbitration networks, involving the OFC and ventral lateral PFC, correlated with obsessive-compulsive symptoms in both BDD and OCD, distinct from other symptoms such as anxiety and depression (44,47). Structurally, OFC volume exhibited a negative correlation with the length of BDD illness (28,29). This overlap between OCD and BDD suggests that similar therapeutic strategies to those used for OCD might be effective in BDD. Indeed, Baldermann et al. (16) reported a reduction in BDD-YBOCS scores following DBS to the ventral capsule/ventral striatum in a refractory patient with BDD; BDD-YBOCS scores also rebounded after switching off the device. This report suggests that frontostriatal network modulation in patients with BDD may also have clinical significance. However, further research should investigate emotional and body image perception postoperation to examine the interconnectivity of the occipital-fronto-limbic network. A preliminary repetitive transcranial magnetic stimulation study on the dorsolateral PFC (dlPFC), which is the most investigated target for noninvasive neuromodulation in OCD, improved BDD severity and anxiety in some patients (75,76). This suggests that the dlPFC and the dorsal cognitive circuit may be involved in the neurobiology of BDD. Further studies should investigate executive function in BDD and its correlation with dorsal cognitive circuits’ (including the dlPFC, dorsomedial PFC, and pre–supplementary motor area) activity and connectivity, preferably in longitudinal studies to probe possible causal mechanisms.

Limitations

Apart from the shortcomings of the cross-sectional design of reviewed studies, another major limitation was the limited sample size, which could underpower detailed analysis. Furthermore, key criteria for patient selection, such as antipsychotic medication status and comorbid psychiatric disorders, were inconsistent across studies from different institutes, contributing to result heterogeneity. Specific to this review, the synthesis and analysis of results were predominantly qualitative. However, we conducted a quantitative meta-analysis on a subset of activity studies that were eligible for ALE analysis. Although the analysis is preliminary and exploratory, it is hoped that future analyses will be enhanced by incorporating additional brain activation coordinates from image-matching visual stimuli experiments.

Future Implications

In addition to implications for future network studies, expanding the scope of neurobiological research by including psychiatric disorders with obsessive-compulsive or body image dysmorphia traits as multiple disease control cohorts could potentially enhance our understanding of the distinctions and similarities in neurobiology among these disorders. From a neurochemistry perspective, additional research into central nervous system neurotransmitter levels, such as cerebrospinal fluid, further positron emission tomography ligand, and magnetic resonance spectroscopy studies to explore cortical glutamate and GABA (gamma-aminobutyric acid) levels, which are disturbed in OCD, is warranted (77). In general, power calculation and a more rigorous methodology should be used to investigate morphometric features in BDD. The precise relationship between the apparent visual processing impairments in BDD and the impairments in executive or arbitration control in frontostriatal networks also requires investigation. For example, the former may well precede the latter and provide an early biomarker for risk for BDD. More generally, OCD itself is associated with some sensory deficits in posterior cortical regions, which may also provide triggers for the development of the disorder.

Conclusions

This comprehensive review of the latest neurocircuitry studies revealed functional and structural brain differences in participants with BDD compared with healthy and disease control groups. The proposed neurocircuitry model expanded on the previous understanding of BDD neurobiology and elucidated the interconnection between the visual processing, temporal-limbic, and frontostriatal networks with their clinical implications. It further described the detailed connectivity and activation aberrancies in BDD neurocircuitry.