Resting-state Connectivity and Individual Differences in Affective Placebo Responsiveness

Overview

Expectation effects on affective states are central to mechanisms relevant for mood disorders and their treatment. This study will test whether inter-individual differences in resting-state functional connectivity (rsFC) within emotion regulation networks is associated with affective placebo responsiveness in healthy volunteers. Approximately 50 healthy participants will be pooled from two cross-over placebo studies in which participants receive intranasal saline either labeled as "oxytocin" in the placebo condition or as saline in the control condition. Primary indices of placebo responsiveness are placebo-control difference scores of mood state, emotional paradigm responses, as well as expectation and experience ratings. rsFC will be estimated using the CONN toolbox. Confirmatory analyses will use an a priori seed-based approach, with prefrontal seeds derived from previous task-based fMRI findings on affective placebo effects. These analyses will be complemented by secondary summaries of within- and between-network connectivity across the saliency network, the fronto-parietal control network, and the default mode network. We hypothesize that interindividual differences in target rsFC networks will be associated with affective placebo effects, indicating general, trait-like mechanisms of placebo responsiveness.

Background Understanding the neurobiological mechanisms underlying expectation effects in the affective domain can provide important insights into therapeutic principles relevant to mood disorders. Placebo and expectancy effects are substantial in mood-related interventions and can be elicited experimentally in healthy samples using standardized, instruction-based expectation inductions combined with sham treatment. In related experimental paradigms based on the same broader framework, experimentally induced positive expectations have been shown to enhance subjective mood and bias emotional information processing toward positivity even in the absence of an active pharmacological agent. Recent work further emphasizes marked interindividual variability in these effects, suggesting that affective placebo effects are stronger when more cognitive resources are available and are accompanied by lateral prefrontal engagement together with stronger coupling in prefrontal-limbic networks. This pattern motivates an individual-differences approach in which expectancy effects depend, at least in part, on the capacity to initiate and maintain instructed treatment beliefs and to translate them into downstream affective processing. Resting-state fMRI (rsfMRI) offers a well-established way to quantify intrinsic functional coupling among large-scale brain systems. Resting-state functional connectivity (rsFC) is commonly used to characterize individual-specific and partly trait-like components of functional network organization, while also encompassing state-dependent variance and measurement constraints. Such individual differences in rsFC have been linked to variability in executive functioning and broader phenotypic differences across healthy and clinical populations. In the context of placebo mechanisms, rsFC involving frontoparietal control circuitry has been associated with individual differences in cue- and expectancy-related modulation (e.g., cognitive modulation of pain). In clinical settings, baseline connectivity patterns can prospectively predict placebo response magnitude in chronic pain trials and placebo-related symptom change in depression, supporting the premise that intrinsic network organization may shape responsiveness to expectancy manipulations. Extending these baseline-predictor accounts, recent work in antidepressant placebo models highlights salience network-default mode network (SN-DMN) coupling as a candidate mechanism linking learned expectancies to both acute mood responses and longer-term belief-related symptom trajectories. In the present study, rsFC will be used to test whether interindividual differences in the intrinsic functional architecture of prefrontal-downstream affective circuitry are associated with affective placebo effects. In addition, canonical large-scale networks will be examined to characterize intrinsic functional architecture relevant to salience detection, cognitive control, and self-referential or internally oriented processing. All participants included in this study have completed one of the preceding fMRI placebo studies and agreed to attend a separate MRI session including structural and rsfMRI measurements. Objectives and hypotheses This rsfMRI study will investigate whether interindividual differences in intrinsic functional coupling are related to primary markers of affective placebo responses. The lateral PFC will serve as the seed region due to its proposed role in maintaining instructed beliefs and translating treatment expectations into limbic valuation networks. Canonical large-scale network summaries (SN, FPCN, DMN) will additionally be used as complementary, systems-level descriptors of the broader organizational context in which these targeted pathways are embedded. The following hypotheses are proposed: * Interindividual differences in rsFC within prefrontal-emotion-related circuitry are associated with placebo changes in mood. * Interindividual differences in rsFC within prefrontal-emotion-related circuitry are associated with placebo changes in task-derived emotional processing. * Interindividual differences in rsFC within prefrontal-emotion-related circuitry are associated with placebo changes in expectancy and experienced benefit. * Interindividual differences in within-network and between-network rsFC across canonical large-scale networks implicated in cognitive control, salience processing, and internally oriented or self-referential processing are associated with primary outcomes of placebo treatment. Methods Ethics information The study was approved by the Ethics Committee of the Hamburg Medical Association (PV 7141) and will be conducted in accordance with the Declaration of Helsinki, with written informed consent, financial compensation, and authorized deception as part of the expectancy manipulation. MRI acquisition parameters MRI data will be acquired on a Siemens Prisma\_fit 3T scanner using a 64-channel head/neck coil. Resting-state fMRI will be acquired with a multiband T2\*-weighted gradient-echo EPI sequence (TR = 1.33 s, TE = 30 ms, flip angle = 60 degrees, multiband factor = 2, 44 slices, matrix = 76 x 76, voxel size = 3 x 3 x 3 mm). Structural images will be acquired using a high-resolution T1-weighted 3D MPRAGE sequence (TR = 2.53 s, TE = 2.34 ms, TI = 1.10 s, flip angle = 7 degrees, voxel size = 1 mm isotropic, matrix = 256 x 256) for anatomical localization, coregistration, and normalization. During resting-state acquisition, participants will be instructed to keep their eyes open, fixate on a central cross, remain still, and not perform any explicit task. Sampling plan The pooled sample is expected to comprise approximately 50 eligible volunteers across the two matched cross-over datasets. An a priori G\*Power analysis showed that N = 44 provides approximately 80% power to detect medium effect sizes (r = .40) at alpha = .05. Analysis plan Resting-state fMRI data will be processed in CONN (RRID: SCR\_009550; v25.b, 22 Feb 2026) in combination with SPM12 (RRID: SCR\_007037; revision r7771). Structural images will be segmented and normalized using SPM's unified segmentation, and functional images will undergo standard preprocessing, including realignment, slice-timing correction, coregistration, normalization to MNI space, and 8 mm FWHM smoothing. Denoising will use CompCor with white matter and cerebrospinal fluid components, motion regressors and their derivatives, ART-based scrubbing regressors, and band-pass filtering (0.008-0.09 Hz). General linear models will be used to examine associations between placebo-related outcomes and interindividual differences in resting-state functional connectivity, behavioral indices, and standardized neural measures. Continuous variables will be mean-centered or z-standardized as appropriate, and standardized effect estimates with 95% confidence intervals will be reported. Resting-state functional connectivity will be quantified as Fisher r-to-z-transformed correlations between BOLD time series and examined using seed-to-voxel and ROI-to-ROI approaches based on prespecified seed and ROI definitions. Seed-to-voxel maps will be tested using cluster-based inference with a voxel-level threshold of p \< .001, uncorrected, and cluster-level FWE correction at p \< .05. ROI-to-ROI analyses will assess connectivity between prespecified prefrontal seeds and downstream mesolimbic regions of interest, with significance evaluated using small-volume-corrected FWE control at p \< .05. Complementary systems-level analyses will derive within-network and between-network connectivity summaries for the salience, frontoparietal control, and default mode networks from pairwise Fisher z-transformed edges. The study is part of the collaborative research center (CRC) SFB/TRR289 and is funded by the Deutsche Forschungsgemeinschaft (DFG, ID: 422744262).