Lower resting-state functional connectivity between frontoparietal and sensory networks is associated with recent pain intensity in a community sample of youth Authors: Carmen I Bango, Scott A Jones, Angelica M Morales, Sara Shao, Dani Y Del Rubin, Arturo Lopez Flores, Amy L Holley, Bonnie J Nagel, Anna C Wilson Published: 2026 ABSTRACT AbstractSelf-reported pain symptoms are highly prevalent during adolescence and are associated with significant individual and societal burden; yet their neurobiological correlates remain insufficiently characterized, particularly in large, community-based cohorts. Identifying patterns in resting-state functional connectivity associated with adolescent pain characteristics may elucidate neural factors relevant to future pain and related health risks. The present cross-sectional study utilized linear mixed-effects models to examine associations between resting-state functional connectivity and past-month pain characteristics (average and worst pain intensity, pain-related limitations, daily pain duration, and number of pain body regions) in youth from Year 2 of the Adolescent Brain Cognitive Development Study (ABCD). No significant differences in connectivity were observed between youth with (n = 2,364, 47.7% female, average age = 12.04) or without (n = 4,447, 46.4% female, average age = 11.99) past-month pain. However, among youth reporting pain, higher average pain intensity was associated with lower connectivity between the frontoparietal network and sensorimotor hand, sensorimotor mouth, and auditory networks, as well as greater connectivity between auditory and visual networks, and between frontoparietal and cinguloparietal networks (FDR-corrected p < 0.01). Higher worst pain intensity showed a similar pattern. No significant associations were observed for pain-related limitations, daily pain duration, or number of pain body regions. Together, these findings highlight altered connectivity between frontoparietal and distributed sensory networks as a potential neural marker of higher pain intensity.Keywords: ABCD study, Resting-state functional connectivity, Adolescent pain, Pain intensity, Functional magnetic resonance imaging, Brain networks Introduction Pain symptoms are commonly reported by adolescents and linked to an increased risk of adverse health outcomes, including chronic pain, mental health conditions, long-term disability, and suicidality (Chambers et al., 2024; Walker et al., 2010; Hassett et al., 2013; Dudeney et al., 2024; Wildeboer et al., 2023; Murray et al., 2020; Mangerud et al., 2013; Noel et al., 2016). Pediatric pain also carries substantial societal and economic burden, estimated at $11.8-19.5 billion annually in the U.S. (Groenewald et al., 2014, 2015) Despite the prevalence and impact of pediatric pain, the neurobiological correlates of pain symptoms during adolescence, particularly in large, community cohorts, remain poorly understood. Adolescence is a developmental period characterized by ongoing maturation of brain systems involved in sensory processing, emotion regulation, and cognitive control (Menon, 2013; Jones et al., 2025a; Gogtay et al., 2004). Neuroimaging studies suggest that these systems are also implicated in pain processing and perception (Apkarian et al., 2005; Cauda et al., 2012; Bushnell et al., 2013; Fiuza-Fernandes et al., 2025; Hashmi et al., 2013; Lumley et al., 2011; Torta et al., 2017). Identifying neural correlates of self-reported pain characteristics in well-powered, community youth samples may provide insight into early neurobiological risk factors for persistent pain and related health risks. Resting-state functional magnetic resonance imaging (rsfMRI) measures brain activity while a person is not engaged in a specified task, allowing the investigation of intrinsic functional connectivity between different brain regions (De Luca et al., 2006). As resting-state connectivity is a clinically modifiable construct(Li et al., 2018; Yoshino et al., 2018; Prehn et al., 2019; Smith et al., 2021; Kong et al., 2019; Masoumbeigi et al., 2022), identifying functional correlates of pain characteristics, such as intensity and interference, may help inform future work on neural targets for pain-related interventions. Prior studies in youth have identified differences in resting-state functional connectivity (rsFC) associated with pain symptoms, particularly within and between the sensorimotor, default-mode, and frontoparietal networks. Alterations in the sensorimotor network, which supports sensory and motor processing and includes the primary motor (M1) and primary somatosensory cortices (S1), have been associated with multisite pain, sensory pain characteristics, and greater pain widespreadedness (Hidalgo-Lopez et al., 2025; Crawford et al., 2023; Hubbard et al., 2014; Kutch et al., 2017). The default-mode network, involved in self-referential thought and mind-wandering, and including key nodes of the ventromedial prefrontal cortex (vmPFC) and the posterior cingulate cortex (PCC), has also been associated with pain features such as frequency, intensity, and current pain presence (Jones et al., 2020; Alshelh et al., 2018; Ceko et al., 2020). The frontoparietal network, which underlies attention, working memory, and goal-directed behavior, and includes the dorsolateral prefrontal cortex (dlPFC), posterior parietal cortex (PPC), and lateral temporal cortex, has been similarly linked to pain persistence, variability in pain intensity, and the cognitive modulation of pain (Crawford et al., 2023; Gim et al., 2024; Becerra et al., 2014; Erpelding et al., 2016; Zeng et al., 2025). These networks are also implicated across structural and task-based studies; specifically, lower cortical thickness and gray matter volume in the dlPFC (Erpelding et al., 2016; Bhatt et al., 2019; Hubbard et al., 2016), PCC (Erpelding et al., 2016; Hubbard et al., 2016), lateral temporal cortex (Erpelding et al., 2016), M1 (Erpelding et al., 2016; Bango et al., 2025), and S141 have been observed across pediatric pain settings. Task-based studies similarly highlight greater activation of S1 (Tong et al., 2023) and PCC (Jones et al., 2019) during evoked pain states for adolescents with persistent pain. Collectively, these studies highlight the sensorimotor, default-mode, and frontoparietal networks as potential key systems in adolescent pain, yet a gap still exists in understanding how these networks relate to dimensional characteristics of pain in broader community populations. The Adolescent Brain Cognitive Development (ABCD) Study is an ongoing longitudinal study examining brain development and health trajectories of nearly 12,000 adolescents across the United States (Casey et al., 2018). Within the ABCD study, pain is assessed using youth self-report measures that capture multiple dimensions of past-month pain. In this cohort, we previously found that greater pain characteristics (e.g., intensity, pain-related limitations) represented by a latent pain factor were associated with lower gray matter volume in predominantly sensory and parietal regions (Bango et al., 2025) and with altered white matter microstructure in tracts subserving the prefrontal cortex (Jones et al., 2025b). Previous ABCD studies using rsfMRI have found that greater functional connectivity between sensorimotor, default-mode and salience networks was predictive of parent-reported multisite pain a year later(Kaplan et al., 2022), and that presence of self-reported multisite pain was linked to lower within-sensorimotor network activity (Hidalgo-Lopez et al., 2025). However, due to differences in pain measures used, study timelines, and methodological variation, existing studies examining neurobiological correlates of pain characteristics in ABCD are difficult to directly compare. For example, while our structural findings were related to youth self-reported pain, other studies in ABCD utilized parent-reported assessments of youth pain or focused on the specific category of multisite pain (Hidalgo-Lopez et al., 2025; Kaplan et al., 2022). Additionally, in terms of resting-state data, analytic flexibility in processing pipelines and regions-of-interest (ROI) further complicates replicability (Botvinik-Nezer et al., 2020). Prior studies have either relied on standardized post-processed data in ABCD, or performed in-house preprocessing on minimally processed data, while also varying in atlas selection and use of a priori ROIs versus whole-brain approaches, hindering direct cross-study comparisons (Hidalgo-Lopez et al., 2025; Kaplan et al., 2022; Quidé et al., 2026). Using ABCD data as publicly released may better allow for repeatable and extendable analyses. Furthermore, the extent to which these findings generalize across pain types (not just multisite) or are associated with relevant dimensional characteristics of pain, such as pain intensity or interference, is still unclear. Assessing specific pain characteristics further enhances the precision of the analysis and may be particularly important in community cohorts like ABCD, where a binary variable of past-month pain is unlikely to capture the heterogeneity of pain symptoms in the sample (Bango et al., 2025). In order to better characterize the neural correlates of recent pain characteristics in the ABCD study, the current study uses postprocessed, publicly available ABCD neuroimaging data to investigate associations between cortical resting-state functional connectivity and several features of pain symptomology, including: the presence of any pain in the past month, average pain intensity, worst pain intensity, pain-related limitations, daily pain duration, and number of pain body regions in youth from the ABCD Study. Based on previous reports implicating the roles of sensorimotor, default-mode, and frontoparietal networks in adolescent pain states (Hidalgo-Lopez et al., 2025; Jones et al., 2020; Becerra et al., 2014; Kaplan et al., 2022; Sunol et al., 2024), we hypothesize that alterations to resting-state functional connectivity involving the sensorimotor, default-mode, and frontoparietal networks will be associated with higher average and worst pain intensity, higher pain-related limitations, greater daily pain duration, and greater number of pain body regions. Methods Participants This study utilized Year 2 data from the Adolescent Brain Cognitive Development (ABCD) Study. ABCD recruited 11,877 youth (ages 9-10 years old) at 21 research sites across the United States and is prospectively carrying out annual follow-up visits to assess changes in numerous domains, including physical and mental health (Barch et al., 2018), neurocognition (Luciana et al., 2018), substance use (Lisdahl et al., 2018), cultural and environmental factors (Zucker et al., 2018), and brain structure and function (Casey et al., 2018). Voluntary informed consent and assent were obtained from parents or caregivers and youth, respectively, and all study procedures were approved by the Institutional Review Board at each research site. All recruitment details and study procedures have been described previously (Garavan et al., 2018). Year 2 follow-up data were used for the current study, as it was the earliest visit (ages 11-13) in which self-reported pain characteristics were collected at the same time as neuroimaging sessions for the full sample. All data relevant for these analyses can be accessed via ABCD Data Release 6.0 (https://doi.org/10.82525/jy7n-g441) stored on the National Institute of Health Brain Development Cohorts (NBDC) Data Sharing Platform. Self-reported pain characteristics As part of a physical health questionnaire, all participants self-reported whether they had any pain in the last month (yes/no). If participants answered yes, they were subsequently asked to report for the past month the average intensity of that pain, the worst intensity of that pain, and the amount that pain limited their daily activities, all on a 0-10 numerical rating scale. They were also asked to report how long that pain lasted on a typical day via a categorical response option (“less than 1 h”, “a few hours”, “half of the day”, or “all day”). Lastly, participants indicated their pain location(s) using the Collaborative Health Outcomes Information Registry (CHOIR) electronic body map (Scherrer et al., 2021). Based on self-reported sex, participants were presented with either a male or female body outline and instructed to: “touch each body region where you have felt pain in the last month.” Participants were able to select up to 74 distinct pain location sites. To reduce dimensionality and enhance interpretability, and consistent with previous work from our lab (Bango et al., 2025), the sites were subsequently grouped into eight major regions: head/neck, abdomen/pelvis, chest, back, right upper extremity, left upper extremity, right lower extremity, and left lower extremity. A count variable (range: 1-8), was derived to represent the number of pain locations endorsed. Consistency of pain reporting was evaluated prior to analyses, and participants were excluded if they met any of the following criteria: 1) answered “yes” to having pain in the past month but reported an average or worst pain intensity of zero; 2) reported an average pain intensity score greater than their worst pain intensity score; or 3) answered “yes” to having pain in the past month but endorsed zero pain sites (Bango et al., 2025; Jones et al., 2025b). Associations among average pain intensity, worst pain intensity, pain-related limitations, daily pain duration, and number of pain body regions were assessed using Spearman's rank correlation coefficients, with pairwise complete observations used to account for missing data. Resting-state fMRI Publicly available postprocessed neuroimaging data from the ABCD study were utilized in the current analyses. Functional MRI scans were collected by the ABCD study using a 3T scanner platform (Siemens Prisma and Prisma Fit, General Electric 750 or Philips Achieva dStream and Ingenia), and multi-channel head coil. Full scan parameter details for all sites have been published previously (Casey et al., 2018). Data processing, detailed previously(Hagler et al., 2019), was carried out by the ABCD consortium before data release. Briefly, initial preprocessing included correction for motion, spatial distortion, and gradient nonlinearities, rigid registration to native structural anatomical space, voxel timeseries normalization, band-pass filtering, and linear regression to remove quadratic trends, motion time courses, and average timeseries from white matter, ventricle, and whole brain masks. Further, time points with framewise displacement (FD) values greater than 0.2 mm or periods with fewer than 5 contiguous sub-threshold frames were also excluded, and average FD was used as a covariate in all analytic models (see below). Both automated and manual quality control were carried out to ensure protocol compliance and check for gross artifacts in the data, as described previously (Hagler et al., 2019). Only scans that were marked by the consortium as passing quality control procedures were included in the current analyses. Connectivity values, as provided by the ABCD consortium, were generated using preprocessed timeseries sampled onto the cortical surface for each subject, and regions-of-interest (ROIs) were defined using the Gordon parcellation scheme (Gordon et al., 2016), with ROIs grouped into the following 12 networks: auditory, cingulo-opercular, cinguloparietal, default mode, dorsal attention, frontoparietal, retrosplenial temporal, salience, sensorimotor (hand), sensorimotor (mouth), ventral attention, and visual. Average between- and within-network connectivity values were calculated by averaging the Fisher-transformed correlation for each unique, pairwise combination of ROIs from each network, resulting in 12 within-network connectivity values and 66 between-network connectivity values. In total, 78 cortical connectivity values were included in the ABCD Data Release and utilized for the current analyses. All neuroimaging processing was conducted by the ABCD consortium, and connectivity values from the Gordon parcellation ROIs were used directly from the ABCD Data Release 6.0. No additional processing of the fMRI data was carried out by the authors. Other analytic covariates Age and sex-assigned-at-birth were self-reported by the participant. Total household income was reported by participants’ parent/caregiver at the Year 2 visit and categorized into four categories: “less than $50,000”, “$50,000-$100,000”, “greater than $100,000”, and “Do Not Know/Refuse to Answer”. Final sample selection Of the 11,877 youth recruited by ABCD, 10,909 completed the Y2 pain questionnaire; however, 154 participants were excluded for discrepant pain reporting (described above in Self-reported pain characteristics), and 14 were excluded for missing, incomplete, or invalid body map data (Bango et al., 2025; Jones et al., 2025b).Additionally, 2998 were excluded for missing or incomplete resting state scan data, 923 were excluded based on ABCD's neuroimaging quality control procedures, and 9 were excluded for incomplete covariate information. This resulted in a final sample of 6811 total participants, with 2364 reporting past-month pain and 4447 reporting no past-month pain. In the sample excluded from analyses (largely due to missing or poor-quality imaging data), the proportion of participants reporting pain in the past month (32%), was not higher than the proportion of those reporting pain in the sample included in final analyses (35%), suggesting that the presence of pain itself was not a driver of neuroimaging data missingness. Statistical analysis Statistical analyses were conducted by the authors using the publicly available ABCD data described above. Linear mixed effects models, via the lme4 package (Bates et al., 2015) in R (v4.4.2), were used to examine the association between pain symptoms and within- and between-network connectivity values. All models included age, assigned sex-at-birth, household income, and motion during the scan as covariates, and random intercepts per family and data collection site, to account for relatedness due to the enrollment of siblings in the study and possible variability in connectivity values by site, respectively. Group differences in connectivity between those reporting past month pain and those without were examined in the full sample (n = 6811), and continuous (or categorical) associations between connectivity and average pain intensity, worst pain intensity, pain-related limitations, pain durations, and number of body map locations endorsed were examined only in those reporting pain in the past month (n = 2364). For all pain variables, a separate linear mixed effects model was fit with each of the 78 connectivity values as the outcome, and all p-values resulting from these models were FDR-corrected to account for 78 multiple comparisons. FDR-corrected p-values <0.05 were used to threshold significant group differences (between those reporting past month pain and those without) in the full sample, and FDR-corrected p-values <0.01 (to correct across the 5 predictors tested) were used to test for significant associations between pain variables and connectivity in those reporting past month pain. To examine potential sex-specific effects, additional linear mixed effects models were fit for each connection with an interaction between assigned-sex-at-birth and the respective pain variables, using the same covariates and random effects structure described above. Interaction term p-values were similarly adjusted for multiple comparisons (FDR-corrected p-values <0.05 for group differences; FDR-corrected p-values <0.01 for pain variables). Across all analyses, average pain intensity, worst pain intensity, pain-related limitations, and number of body map locations were treated as continuous variables. Daily pain duration was modeled as a categorical variable using dummy coding with four levels, as defined above in Self-reported pain characteristics. All continuous variables were mean-centered and scaled by their standard deviations before modeling so that the resulting coefficients represent standardized regression estimates (βs). All analytic scripts utilized in this study are available here: https://osf.io/w482t/overview?view_only=0c361d0ecd174976948f51c710469ab9. Results Sample characteristics There were 6811 participants included in the analysis looking at the effects of past-month pain on connectivity, and 2364 participants reporting past-month pain included in the analysis looking at the effects of various pain symptoms on connectivity. Sample demographics relevant for the current analyses are provided in Table 1. Extended sample demographics for the full ABCD sample have been published previously (Barch et al., 2021; Karcher and Barch, 2021). For the current study, in those reporting past-month pain, the mean average pain intensity was 4.3 (SD = 1.9), the mean worst pain intensity was 6.2 (SD = 2.2), and the mean pain-related limitations was 2.0 (SD = 2.4). For average daily pain duration, 43% of participants reported pain for “less than an hour”, 23% reported pain for “a few hours”, 16% reported pain for “half the day”, and 18% reported having pain “all day”. The median number of pain locations endorsed was 2 out of the 8 possible categories (head/neck, abdomen/pelvis, chest, back, right upper extremity, left upper extremity, right lower extremity and left lower extremity). In terms of body site location, 557 (23.6%) participants endorsed having head/neck pain, 570 (24.1%) participants endorsed having abdomen/pelvis pain, 142 (6.0%) participants endorsed having chest pain, 482 (20.4%) participants endorsed having back pain, 552 (23.4%) participants endorsed having right upper extremity pain, 528 (22.3%) participants endorsed having left upper extremity pain, 1108 (46.9%) participants endorsed having right lower extremity pain, and 1132 (47.9%) participants endorsed having left lower extremity pain. Table 1Sample characteristics.Table 1No PainPainTotal Participants44472364Age, years (SD)11.99 (0.65)12.04 (0.66)Female, N (%)2063 (46.4%)1128 (47.7%)Race/Ethnicity N (%)White2434 (54.73%)1394 (58.97%)Black623 (14.01%)249 (10.53%)Hispanic or Latino879 (19.77)447 (18.91)Asian79 (1.78)28 (1.18)Other432 (9.71)246 (10.41)Household income, N (%)Less than $50,000997 (22.42)475 (20.09)50,000 to 100,0001144 (25.73)632 (26.73)Greater than $100,0001961 (44.1)1101 (46.57)Don't know or decline to answer345 (7.76)156 (6.6)NOTE: Racial identity categories are self-reported, and several categories have been collapsed as follows: Asian includes participants who self-identify as Asian Indian, Chinese, Filipino, Japanese, Korean, Vietnamese, or other Asian; Other includes categories collapsed across Other, Multi-racial and Native Hawaiian and/or Pacific Islander (participants who self-identify as Native Hawaiian, Guamanian, Samoan, or other Pacific Islander). Categories do not sum to 100% of the sample, because participants identifying with multiple racial or ethnic identities are included in multiple groups. Sample characteristics. NOTE: Racial identity categories are self-reported, and several categories have been collapsed as follows: Asian includes participants who self-identify as Asian Indian, Chinese, Filipino, Japanese, Korean, Vietnamese, or other Asian; Other includes categories collapsed across Other, Multi-racial and Native Hawaiian and/or Pacific Islander (participants who self-identify as Native Hawaiian, Guamanian, Samoan, or other Pacific Islander). Categories do not sum to 100% of the sample, because participants identifying with multiple racial or ethnic identities are included in multiple groups. Correlations among the five pain measures ranged from r = 0.06 to r = 0.82, with all associations reaching statistically significance (p < 0.05). Average and worst pain intensity were strongly correlated (r = 0.82). Associations between average pain intensity and pain-related limitations (r = 0.43) and worst pain intensity and pain-related limitations (r = 0.46) were moderate. All other correlations were weak (r = 0.06 – 0.22). Results are presented in a correlation matrix (see Supplemental Fig. S13). Past-month pain There were no significant (FDR-corrected p < 0.05) group differences in rsFC between those with and without past-month pain across all 78 tested connections (Supplemental Table S1). Pain intensity Analyses in the sub-sample reporting past-month pain found that pain intensity (average pain intensity and/or worst pain intensity) was significantly associated (FDR-corrected p < 0.01) with rsFC differences in 5 of the 78 connections examined (Fig. 1a and b). Specifically, higher average pain intensity was associated with less connectivity between the frontoparietal network and sensorimotor (hand) (β = −0.082), sensorimotor (mouth) (β = −0.090), and auditory (β = −0.077) networks, as well as greater connectivity between the auditory and visual networks (β = 0.073), and greater connectivity between the cinguloparietal and frontoparietal networks (β = 0.074) (Fig. 2a–e). Similarly, greater worst pain intensity was also associated with less connectivity between the frontoparietal and auditory networks (β = −0.074) and greater connectivity between the auditory and visual networks (β = 0.082) (Fig. 2f and g). Model results for all connections in association with average and worst pain intensity are in Supplemental Tables S2 and S3, respectively. Associations between pain symptoms and network connectivity in individuals reporting past-month pain. Connectivity matrices depict the association (β – standardized regression coefficient) between network-to-network connectivity and average pain intensity (A) and worst pain intensity (B). Cells noted with an asterisk (∗) represent connections where there is a statistically significant association between connectivity and the pain variable (FDR-corrected p < 0.01). Significant Associations between pain symptoms and network connectivity in individuals reporting past-month pain. A-G. Scatterplots display the relationship between Average (A-E) and Worst (F-G) Pain Intensity and between network connectivity values (Fisher's z-transformed correlation coefficients) (All FDR-corrected p < 0.01). Each point represents an individual participant. The blue line indicates the fitted regression line. Panels illustrate connectivity between: (A, F) auditory and frontoparietal networks, (B, G) auditory and visual networks, (C) frontoparietal and cingulo-opercular networks, (D) sensorimotor hand and frontoparietal networks, and (E) sensorimotor mouth and frontoparietal networks. Pain-related limitations, daily pain duration and number of pain body regions Analyses in the sub-sample reporting past-month pain found no significant associations (FDR-corrected p < 0.01) between functional connectivity and pain-related limitations (Supplemental Table S4), daily pain duration (Supplemental Table S5) or number of body regions with pain endorsed (Supplemental Table S6) across all 78 tested connections for each pain variable. Exploratory pain-by-sex interaction models Exploratory analyses of pain-by-sex interactions revealed no statistically significant sex-specific effects across all 78 tested connections for each pain variable (FDR-corrected p < 0.05 for group differences, FDR-corrected p < 0.01 for the five pain variables). Full results are included in Supplemental Table S7–S12. Discussion The current study sought to investigate rsFC patterns associated with self-reported pain characteristics in a large, nonclinical, community sample of youth. Our findings partially support our original hypothesis: we observed associations between pain characteristics and rsFC involving sensorimotor and frontoparietal networks. However, only models including average and worst pain intensity resulted in significant findings, and the expected patterns involving the default-mode network were not observed. Additionally, unexpected and novel associations emerged involving the cinguloparietal network, auditory, and visual networks. Broadly, the current study's findings differ somewhat from prior ABCD work examining relationships between rsFC and pain characteristics (Hidalgo-Lopez et al., 2025; Kaplan et al., 2022). These discrepancies may be attributable to differences in pain measures, analytic decisions, and study design. Prior ABCD studies have focused on multisite pain, employed longitudinal designs in which participants were pain-free at the imaging visit, and/or relied on parent-reported measures of pain (Hidalgo-Lopez et al., 2025; Kaplan et al., 2022). No prior study, to our knowledge, has investigated associations between self-reported dimensional pain characteristics and rsFC among participants in the ABCD cohort. The current study also used postprocessed ABCD data, whereas previous work have typically extracted minimally processed data followed by study-specific processing pipelines (Hidalgo-Lopez et al., 2025; Kaplan et al., 2022). Differences in downstream processing decisions, including atlas selection, parcellation strategy, motion thresholds and other quality-control criteria may further account for variability in reported findings (Botvinik-Nezer et al., 2020). By leveraging publicly released, postprocessed ABCD data generated through standardized pipelines, the present study aims to facilitate reproducibility and methodological consistency relative to prior work. Nevertheless, residual methodological differences across existing studies, likely continue to contribute to inconsistencies in this literature and limit cross-study comparisons. Prominent in the current study's findings is the pattern of lower connectivity between frontoparietal and sensorimotor networks associated with higher average pain intensity. The frontoparietal network is thought to play a critical role in integrating contextual, expectancy-related, and attentional information with sensory information, such as stimulus intensity, to shape pain perception (Gim et al., 2024; Geuter et al., 2020; Kong et al., 2013). In healthy adult populations, receiving noxious/painful stimuli has been associated with increased connectivity between frontoparietal and somatosensory networks, which may represent an adaptive executive prioritization of sensory signaling to facilitate protective threat responses (Zheng et al., 2020; Li et al., 2022). In contrast, lower connectivity between frontoparietal and sensorimotor networks may reflect diminished executive regulation of sensory information during sustained pain states. This pattern is consistent with models of persistent pain development in which lower engagement of prefrontal modulatory regions leaves sensory networks more susceptible to nociceptive input or pain-induced neuroplasticity (Crawford et al., 2023; Nara et al., 2025; Ong et al., 2019; Xie et al., 2026). Critically, the dorsolateral prefrontal cortex (dlPFC), a central node of the frontoparietal network, is implicated in descending pain modulation(Atlas et al., 2010; Seminowicz and Moayedi, 2017), and is a common target in noninvasive brain stimulation techniques for pain. Similarly, the motor cortex (M1), a key region of the sensorimotor network, has also been targeted. Research demonstrates promising effects of dlPFC stimulation,(Song et al., 2025), (Sahu et al., 2019) (Che et al., 2021; Nardone et al., 2017; Vats et al., 2024) as well as M1 stimulation (Attal et al., 2021; Ngernyam et al., 2015; Altas et al., 2019; Valle et al., 2009; Fregni et al., 2006), for pain reduction across varied conditions. In healthy populations, stimulation of both dlPFC and M1 has been shown to modulate pain thresholds and perception (Boggio et al., 2008; Dehghani et al., 2025; Mylius et al., 2012). The current study's findings further highlight these pathways and regions as potential targets for pain modulation and treatment across settings. Higher past-month average and worst pain intensity were also associated with lower connectivity between frontoparietal and auditory networks, and greater connectivity between auditory and visual networks. This finding, though novel in adolescents, is in line with theories of heightened multi-sensory arousal or central sensitization processes contributing to pain perception (Andias and Silva, 2022; Harte et al., 2018). Multi-sensory processing difficulties, manifested as increased sensitivity to stimuli such as light, sound, touch, smell and taste, has been documented in adolescents with diverse pain conditions (Koechlin et al., 2023), and in adults, particularly in patients with fibromyalgia or multisite pain (Wang and Frey-Law, 2023; Inoue et al., 2025; Geisser et al., 2008; Staud et al., 2021). The current findings suggest that youth with higher pain intensity show greater coordinated activity across sensory networks, alongside less integration with higher-order regulatory systems. Future studies may examine whether greater functional connectivity between sensory networks is related to hyper-responsivity to different sensory stimuli in this population. Our findings also reveal greater rsFC between frontoparietal and cinguloparietal networks associated with higher average pain intensity in youth. This is, to our knowledge, a novel finding not previously published in either small clinical studies or large community cohorts. This may partly reflect the specificity of the cinguloparietal network to the Gordon Atlas, as other commonly used atlases may assign cinguloparietal regions to broader networks, such as the default-mode network, rather than treating them as a distinct network (Wainberg et al., 2022; Yeo et al., 2011). Nevertheless, as prior work has suggested that cinguloparietal-related activity is related to internal monitoring of bodily and emotional states (i.e., interoception) (Cauda et al., 2012), and encoding pain intensity levels (Favilla et al., 2014), its greater connectivity with the frontoparietal network in youth reporting higher average pain intensity may reflect heightened salience of sensory and emotion-related pain information. However, these interpretations should be made cautiously, as the present study is not able to address underlying mechanisms of pain processing. Further research is needed to better characterize the connectivity between frontoparietal and cinguloparietal networks in relation to pain intensity. Within a developmental context, research suggests that sensory-related networks reach full development first, while cognitive networks do not fully develop until closer to adulthood (Jones et al., 2025a). If situated within this research, the lower connectivity observed between the frontoparietal and sensory networks may suggest a developmental lag in the integration of regulatory systems with sensory processing that could contribute to an underlying vulnerability to higher pain intensity. Alternatively, these findings may reflect neuroplastic changes in response to persistent pain, resulting in diminished regulation of sensory systems due to processes such as central sensitization (McCarberg and Peppin, 2019). An additional possibility is that the observed lower rsFC between frontoparietal and sensorimotor networks associated with higher pain intensity reflects altered motor system engagement. Pain has been associated with increased movement inhibition or protective motor responses in both acute and chronic contexts (Bank et al., 2013; Kantak et al., 2022; Chowdhury et al., 2022). Lower connectivity between these networks may therefore relate to diminished integration between cognitive control and motor systems. Consistent with this interpretation, prior work suggests that stimulation of the cortical motor system can alleviate pain, potentially by restoring frontoparietal-sensorimotor connectivity (Giannoni-Luza et al., 2020; Betancur et al., 2025; Urien and Wang, 2019). Future analyses may consider inclusion of movement-related data to further explore this theory. However, these conclusions regarding the developmental context of our findings are speculative given the single time-point design of the current study. Further longitudinal research is needed to better understand these patterns over time. The present study contributes to the literature on the relationship between resting-state functional connectivity and pain characteristics in a nonclinical, community sample of youth. However, several limitations should be discussed as they relate to the interpretability of the findings. Like previous reports examining brain structure in this cohort (Bango et al., 2025; Jones et al., 2025b), we observed no significant group differences in resting-state connectivity between those with and without past-month pain. As noted in prior work(Bango et al., 2025; Jones et al., 2025b), the absence of significant group differences may reflect the limitations of using a nonspecific binary item – reporting any aches or pains in the past-month – which likely fails to capture meaningful variability in pain severity in this nonclinical sample. While this measure may serve as clinically relevant information in a patient population, it is less effective in distinguishing every day, acute pain, (e.g., a stubbed toe) from more persistent or impactful pain in a nonclinical context. Accordingly, this null finding should be interpreted with caution, and more granular measures, such as pain intensity or pain-related limitations, may better characterize neural patterns associated with pain characteristics in this group. Additionally, the null findings with the number of body regions may relate to the coarse measure of body sites included in this study, which operationalized the 74 possible body sites into 8 major independent regions, and may obscure potentially meaningful variation. While this study focuses on continuous measures of self-reported pain, future studies could consider creating pain-related subgroups (e.g., high impact pain, or by specific pain locations) as another method for understanding pain presentations in this sample. Critically, although self-reported pain measures are commonly used in both research and clinical practice, they are subject to recall bias, particularly when assessed over extended periods such as the past month. Future studies employing ecological momentary assessment (EMA) of pain would strengthen the existing rsFC pain literature in youth. Furthermore, due to the cross-sectional nature of this study, we are unable to determine the causal direction between resting-state connectivity differences and pain symptoms. Lastly, as we relied on secondary data analysis of the ABCD study, wherein pain was not the focus, we were limited in our assessment of adolescent pain symptoms. Among other metrics, we lacked information regarding the etiology of pain symptoms or relevant diagnoses, whether pain relates to a chronic or recurrent pain condition, or if any participants with past-month pain received medical care for their pain. Despite these limitations, this study provides valuable insights into the neurobiological markers of pain characteristics in youth and represents a potential starting point for future work conducting longitudinal analyses in this sample and beyond. Collectively, our results describe a pattern of higher pain intensity characterized by lower connectivity between frontoparietal and dispersed sensory networks, as well as greater connectivity between frontoparietal and cinguloparietal and between auditory and visual networks. Importantly, these findings were observed in a nonclinical, community sample, and may serve as early features or neural risk factors for clinical pain outcomes. Further longitudinal research is needed to explore the trajectories of these networks over time and determine if they are predictive of future pain reports, and if interventions targeting the connectivity between regulatory and sensory systems may help in the prevention and treatment of pain in youth. Ethics statement This project uses publicly available data from the Adolescent Brain Cognitive Development (ABCD) Study, for which all procedures were approved by the Institutional Review Boards of participating institutions. Parents and guardians of participants provided consent, while participants provided assent. Funding This work was funded by the Eunice Kennedy Shriver National Institute of Child Health & Human Development [R21 HD112210, PI: Wilson], and supported by philanthropic funding through the Center for Mental Health Innovation (CMHI) at Oregon Health & Science University (OHSU). This work was also supported by the resources provided by the OHSU Advanced Imaging Research Center and the Oregon opportunity partnership for advancing biomedical research, as well as by NIH S10OD021701, which funded the 3T Siemens Prisma MRI instrument, and NIH S10OD016356, which funded a 32-channel radio-frequency head coil, both housed in OHSU's Advanced Imaging Research Center. The research reported in this publication used computational infrastructure supported by the Office of Research Infrastructure Programs, Office of the Director, of the National Institutes of Health under Award Number S10OD034224. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. CRediT authorship contribution statement Carmen I. Bango: Conceptualization, Formal analysis, Methodology, Visualization, Writing – original draft, Writing – review & editing. Scott A. Jones: Conceptualization, Data curation, Formal analysis, Methodology, Visualization, Writing – review & editing. Angelica M. Morales: Conceptualization, Supervision, Writing – review & editing. Sara Shao: Data curation, Formal analysis, Methodology, Validation, Visualization, Writing – review & editing. Dani Y. Del Rubin: Data curation, Formal analysis, Methodology, Validation, Visualization, Writing – review & editing. Arturo Lopez Flores: Investigation. Amy L. Holley: Conceptualization, Supervision, Writing – review & editing. Bonnie J. Nagel: Conceptualization, Supervision, Writing – review & editing. Anna C. Wilson: Conceptualization, Funding acquisition, Supervision, Writing – review & editing. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.ynirp.2026.100366. Supplementary data The following are the Supplementary data to this article: Multimedia component 1 Multimedia component 2 Multimedia component 3 Multimedia component 4 Multimedia component 5 Multimedia component 6 Multimedia component 7 Multimedia component 8 Multimedia component 9 Multimedia component 10 Multimedia component 11 Multimedia component 12 Multimedia component 13 Data availability Data were obtained from the ABCD Study (https://abcdstudy.org/) held in the NBCD Data Hub. Project Analysis code: https://osf.io/w482t/overview?view_only=0c361d0ecd174976948f51c710469ab9 References Associated Data Supplementary Materials Multimedia component 1 Multimedia component 2 Multimedia component 3 Multimedia component 4 Multimedia component 5 Multimedia component 6 Multimedia component 7 Multimedia component 8 Multimedia component 9 Multimedia component 10 Multimedia component 11 Multimedia component 12 Multimedia component 13 Data Availability Statement Data were obtained from the ABCD Study (https://abcdstudy.org/) held in the NBCD Data Hub. Project Analysis code: https://osf.io/w482t/overview?view_only=0c361d0ecd174976948f51c710469ab9