The Neural Underpinnings of Obsessive-Compulsive Disorder: From Adolescence to Adulthood

Abstract

This thesis investigates the neurocognitive mechanisms underpinning compulsive behaviour in obsessive-compulsive disorder (OCD), focusing on how these mechanisms evolve across development. OCD is a debilitating neuropsychiatric disorder typically emerging during adolescence, characterised by intrusive thoughts and repetitive behaviours that are difficult to control. Despite its early onset and chronic nature, the underlying neural and computational processes that drive compulsivity, particularly during the critical neurodevelopmental period, remain incompletely understood. To address this gap, I apply a multimodal approach combining ultra-high field magnetic resonance spectroscopy (7T MRS), resting-state functional MRI (rs-fMRI), and behavioural paradigms to identify neurochemical markers, network-level disruptions, and behavioural signatures of compulsivity in individuals aged 16–21. A particular focus is placed on the anterior cingulate cortex (ACC), supplementary motor area (SMA), and putamen—regions critically involved in behavioural regulation and decision-making. The central hypothesis is that compulsive behaviour in OCD arises from a failure in the arbitration between goal-directed and habitual systems, driven by disrupted excitation-inhibition (E/I) balance within corticostriatal circuits. This thesis builds directly on previous work in adult OCD, particularly the findings of Biria et al. (2023), which demonstrated altered glutamate and GABA concentrations in frontostriatal regions. A key innovation of the current research is its extension of this work to adolescents with OCD, enabling a developmental perspective on the emergence and consolidation of compulsive phenotypes. The inclusion of adolescents allows for a mechanistic investigation of how neural dysfunctions evolve over time and how early vulnerability may be identified. Chapter 2: General methods This chapter presents the multimodal methodological framework employed across all empirical studies. The research recruited 100 participants (50 OCD, 50 healthy volunteers) aged 16-64 years, stratified into adolescent (16-21 years) and adult (>21 years) cohorts to examine developmental trajectories. The protocol integrated ultra-high field 7T magnetic resonance spectroscopy to quantify GABA and glutamate in the anterior cingulate cortex, supplementary motor area, and left putamen—the latter involving a novel optimised protocol for subcortical spectroscopy. Resting-state fMRI assessed intrinsic and extrinsic functional connectivity within corticostriatal networks, while a contingency degradation task with metacognitive ratings evaluated the balance between goal-directed and habitual control. Comprehensive clinical phenotyping included the Y-BOCS, OCI, and dimensional measures of perfectionism, uncertainty intolerance, and habitual tendencies. Advanced analytical approaches encompassed tissue-corrected metabolite quantification, standardised preprocessing pipelines (fMRIPrep, XCP-D), and generalized additive modelling for nonlinear developmental trajectories, establishing a rigorous foundation for investigating how disrupted excitation-inhibition balance manifests across neural and behavioural levels in OCD. Chapter 3: Age-related neurochemical and connectivity alterations in OCD: A 7T-MRS & rs-fMRI investigation This chapter examines neurochemical alterations in OCD throughout development using 7T MRS and resting-state fMRI in a cross-sectional design. The adolescent cohort showed no significant differences in GABA, glutamate, or glutamate/GABA ratios between OCD participants and healthy volunteers across the ACC, SMA, and putamen. However, despite preserved mean neurochemical levels, adolescents with OCD showed markedly disrupted relationships between neurotransmitter concentrations and functional connectivity. Healthy adolescents demonstrated robust negative correlations between neurochemistry and connectivity, while adolescents with OCD showed significantly weakened coupling. Across the full age range, developmental modelling using generalized additive models revealed that neurochemical alterations in OCD emerge specifically during early adulthood. The ACC showed elevated excitation/inhibition balance in OCD driven primarily by glutamatergic rather than GABAergic alterations. Group × age interactions demonstrated that glutamate differences emerge in early adulthood. Notably, these neurochemical changes were regionally specific to the ACC, with no alterations observed in the SMA, suggesting targeted rather than global dysfunction. These findings support a developmental model where OCD begins with functional disruptions in adolescence (altered neurochemistry-connectivity coupling) that precede the emergence of structural neurochemical alterations in early adulthood. The ACC-specific, glutamate-driven E/I imbalance that develops during the transition to adulthood may reflect compensatory mechanisms or consequences of chronic aberrant circuit activity. Chapter 4: Neural correlates of OCD symptom dimensions While Chapter 3 focuses on diagnostic group differences, Chapter 4 shifts the emphasis to dimensional symptomatology, asking whether individual variation in compulsive traits maps onto specific neural substrates. Using the same 7T MRS data, the chapter relates glutamate/GABA concentrations to scores on the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS) and Obsessive-Compulsive Inventory (OCI), as well as specific subdimensions such as ordering, and checking. Results show that associations between neurochemistry and symptoms differ across age groups. In adolescents, higher ACC glutamate/GABA ratios are strongly associated with greater OCD symptom severity. In contrast, adults exhibit a shifted pattern, with lower E/I ratios associated with higher symptom scores. This suggests a developmental reconfiguration in the relationship between brain chemistry and symptom expression, with glutamate and GABA markers playing different roles at different stages of the disorder. Moreover, the analysis of symptom subtypes reveals distinct neural associations with ACC neurochemistry. This chapter demonstrates that individual differences in symptom expression are tightly linked to region-specific neurochemical alterations, with patterns that shift as individuals move from adolescence into adulthood. Chapter 5: Contingency degradation and metacognitive awareness in OCD across development The final empirical chapter examines behavioural differences, using a contingency degradation task to examine goal-directed versus habitual control in OCD. Participants performed a task in which the relationship between their actions and outcomes was systematically degraded, allowing quantification of both behavioural inhibition and explicit awareness of action-outcome contingencies. Results reveal that adolescents with OCD exhibit impaired goal-directed behaviour: they persist in responding even when their actions no longer influence outcomes. However, they retain accurate metacognitive insight such that they can report that their actions are no longer effective, even though their behaviour does not reflect this knowledge. This dissociation between knowledge and action supports the hypothesis that OCD involves a disruption in arbitration, not learning per se. Notably, in adolescents, these behavioural patterns correlate with ACC glutamate levels, but only in relation to explicit knowledge, not overt behaviour. This suggests that neurochemical alterations in the ACC may impair the implementation of flexible control, even when the cognitive representation of contingencies remains intact. The findings also highlight early emergence of behavioural rigidity, consistent with prior findings that compulsivity may be rooted in developmental failure to engage goal-directed control systems. This chapter provides behavioural validation of the mechanistic model proposed in the earlier neuroimaging chapters, directly linking ACC dysfunction to arbitration deficits observable in real-time decision-making. General discussion Taken together, the chapters of this thesis present a convergent account of OCD as a disorder of developmentally mediated control system imbalance, marked by early connectivity dysfunction, later neurochemical disruption, and persistent behavioural inflexibility despite intact awareness. The findings support a neurodevelopmental model in which disruptions to E/I balance and network coordination in corticostriatal circuits begin in adolescence and evolve with chronicity. The inclusion of adolescents and young adults provides an opportunity to map the unfolding of OCD mechanisms across developmental stages, addressing a critical gap in the literature. Moreover, by integrating multiple levels of analysis—neurochemistry, functional connectivity, and behaviour—this work moves beyond categorical diagnosis and toward mechanistic, dimensional characterisations of compulsivity. The results have potential implications for early identification and intervention, especially in targeting preclinical connectivity alterations before more entrenched symptoms and neurochemical changes emerge. Finally, by building on and extending findings from adult-focused studies (notably Biria et al., 2023), this thesis lays the groundwork for a developmentally sensitive, circuit-based understanding of OCD.