The Relationship between Neurobiological Function and Inflammation in Depressed Children and Adolescents: A Scoping Review

Depression has increased dramatically in the last decade, currently affecting about 32% of children and adolescents globally [1, 2]. However, depression in children and adolescents remains underdiagnosed due to the heterogeneity of the illness and the developmental differences in presentation in childhood compared with that of adulthood [3]. Multiple biomarkers have been associated with depression in children and adolescents, of which neurobiological dysfunction may be a promising candidate as several patterns of structural and functional connectivity within the brain are associated with the heterogeneous presentation of child and adolescent depression [4‒6]. Research has shown that children and adolescents with depression display changes in neuronal structures and activity [7‒9]. For example, Miller and colleagues [9] conducted a meta-analysis of 14 studies to examine the activity of various brain areas in a total of 246 currently depressed adolescents and 274 healthy controls during multiple cognitive tasks. Consistent evidence for hyperactivity of the subgenual anterior cingulate cortex (sgACC) and ventrolateral prefrontal cortex (vlPFC) and hypoactivity of the caudate nucleus was found in depressed adolescents compared to control participants across all tasks. In a subsequent study of 36 depressed adolescent outpatients and 37 healthy controls, fMRI and diffusion tensor imaging revealed that depressed adolescents possessed decreased functional and structural connectivity, respectively, between the left amygdala and the left ventral prefrontal cortex, suggesting that this neuronal circuit may be involved in early-onset depression [7]. Recently, Auerbach and colleagues [8] reported reduced volume and reward-related activation of the nucleus accumbens in 129 depressed-anxious adolescents compared with 64 healthy controls.

While there is growing evidence that neurobiological markers are associated with child and adolescent depression, the underlying mechanism of these differences is unclear. Studies investigating the pathoetiology of brain dysfunction in depressed adults have suggested a potential role for inflammatory pathways [10, 11]. However, this proposed link has received limited attention in children and adolescents with depression. Nonetheless, inflammation appears to be an important factor in child and adolescent depression [5, 12]. Studies to date have reported alterations in concentrations of several inflammatory markers, including interleukin 6 (IL-6), C-reactive protein (CRP), tumor necrosis factor (TNF) alpha, and interferon (IFN) gamma in depressed children and adolescents compared with healthy controls [13‒15]. Moreover, Toenders and colleagues [16] recently proposed that depression in youth may be characterized by multiple different immunophenotypes that correlate with specific depression symptom profiles, further strengthening the idea that inflammatory markers may be instrumental in the heterogeneous profile of youth depression. Thus, dysregulation of inflammatory cascades should be considered a potential candidate for provoking neurobiological dysfunction in child and adolescent depression.

Taken together, depression in children and adolescents has been characterized by altered neuronal structure, neuronal activity, and markers of inflammation. Despite these findings, the relationship between these factors in children and adolescents with depression has yet to be examined [16‒18]. Thus, the aim of this scoping review was to provide an overview of current research examining the potential interaction between neurobiological function and inflammation in depressed children and adolescents.

This scoping review was conducted using the Joanna Briggs Institute methodology for scoping reviews and reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist (online suppl. Table 1; for all online suppl. material, see https://doi.org/10.1159/000538060) [19, 20]. A protocol for this scoping review was created and registered on the Open Science Framework [21].

A comprehensive literature search was conducted in February 2023 in the following databases: Ovid MEDLINE (1946-present, including Epub ahead of print, in-process, and other non-indexed citations), Ovid Embase (1947-present), Ovid APA PsycINFO (1806-present), EBSCO CINAHL Plus with Full Text (1981-present), Child Development and Adolescent Studies (EBSCOhost), and Scopus (Elsevier). The database search strategies were developed by an academic health sciences librarian (GBR) and peer-reviewed according to the Peer Review of Electronic Search Strategies (PRESS) for systematic review guidelines [22]. The full search strategies can be found in online supplementary Table 2. Review of reference lists of the included studies as well as search for gray literature via ProQuest were conducted to identify records from other resources.

Studies were included in this review if they met the following criteria: (1) examined a pediatric population (mean age ≤18 years); (2) assessed participants with a current diagnosis of clinical depression according to standardized validated diagnostic criteria (i.e., International Classification of Diseases, Diagnostic and Statistical Manual of Mental Disorders) either through diagnostic interview or expert clinical assessment, including clinically diagnosed unipolar depression, major depressive disorders, and dysthymia [23, 24]. Studies that included participants with comorbidity were retained in the scoping review due to the frequent co-occurrence of depression with comorbidity in youth [25]; (3) measured neurobiological functioning via standardized methods (e.g., MRI) and measured the most commonly implicated inflammatory markers in child and adolescent depression (i.e., CRP, IL, TNF, IFN) [13, 26]; (4) were peer-reviewed cross-sectional, case-control, case series, randomized control trials, or cohort (longitudinal) studies with or without healthy controls. Studies were excluded if they (1) included participants diagnosed with a comorbid neurological condition, intellectual disability, or inflammatory disorder; (2) were case reports, narrative and systematic reviews, conference proceedings, commentaries, or letters to the editor. No limits to the publication date of records, publication language, geographical location, socio-demographic factors, or setting were imposed.

Search results generated by the systematic searches were stored in Covidence online systematic review software for management and duplication removal as well as the screening of remaining articles [27]. Screening was conducted independently by two reviewers (A.S., J.M., or S.C.C.) and performed in a two-step process: (1) citations were first screened by title and abstract; (2) articles meeting inclusion criteria were then reviewed in full text. In the event of discrepancies, consensus was obtained. The following data were extracted in duplicate (A.S., J.M.): first author, publication year, country, age, sample size, sex, setting, design, ethnicity, diagnostic tools for assessment of depression, depression outcomes (means, SD), diagnostic tools for assessment of neurobiological functioning, neurobiological outcomes (descriptive results), diagnostic tools for assessment of inflammatory markers, inflammation outcomes (means, SD), neuro-immune outcomes (descriptive results).

The literature search yielded 6,911 studies; following de-duplication and abstract screening, 44 remained. After full-text review, 40 studies were excluded for the following reasons: no combined measurement of neurobiological function and inflammation (n = 4); no clinical diagnosis of depression (n = 6); conference proceedings (n = 11); and wrong population (adult participants) (n = 19). Subsequently, four studies met the criteria and were included in this review [28‒31]. Details of the stages of screening and reasons for exclusion are shown in a PRISMA flowchart in Figure 1.

Studies were published between 2019 and 2022 and had sample sizes ranging from 46 to 75 participants. All studies were conducted in the USA. Participants’ age ranged from 12 to 20 years, with a mean age of 16.0 ± 0.6 years (62% female). Children and adolescents were recruited from community settings (n = 2) or combined clinical and community settings (n = 2). All included studies were cross-sectional in design. Of note, two studies [30, 31] were performed by the same study team in 2021 and 2022, and two studies [28, 29] were conducted within the same study setting; however, participants were not included in more than one study. Table 1 shows the characteristics of the four included studies. Due to the limited number of included studies, meta-analysis of the results could not be conducted.

Diagnosis of a depressive disorder in all studies was confirmed using the Kiddie Schedule for Affective Disorders and Schizophrenia, Present and Lifetime (KSADS-PL), a semi-structured diagnostic interview based on DSM-IV criteria [32]. Similarly, all studies utilized the Children’s Depression Rating Scale – Revised (CDRS-R), an interviewer-administered measure, to assess depressive symptoms [33]. Three studies also included self-reports of depression symptoms to assess depression severity, with two studies using the Beck Depression Inventory – Second Edition (BDI-2) [28, 29] and one study [30] using the self-reported Reynolds Adolescent Depression Scale – Second Edition (RADS-2) [34, 35]. See online supplementary Figure 1a for a graphical overview.

Bradley et al. [29] recruited psychiatric adolescents with a diagnosis of a depressive disorder, bipolar disorder, anxiety disorder, attention-deficit-hyperactivity disorder, behavior disorder, or obsessive-compulsive disorder, as well as healthy non-psychiatric adolescents. Liu et al. [28] focused on adolescents with diverse psychiatric profiles and did not include healthy control participants in their study. Ho et al. [30, 31] (2021) examined depressed participants only, while the follow-up study in 2022 included a healthy control group that exhibited significantly fewer depressive symptoms compared with the depressed group.

Brain imaging in all included studies was conducted using a 3T MRI scanner. Whole-brain imaging was conducted in three studies to identify the brain areas that showed alterations in depressed children and adolescents [28‒30]. Region of interest (ROI) analyses were used to investigate changes in the ACC in three studies [28, 29, 31], to examine the basal ganglia, including the nucleus accumbens, caudate, and putamen, in two studies [28, 29], and to examine white matter tracts, including the genus of corpus callosum, corpus callosum splenium, and bilateral uncinate fasciculus, in one study [30]. See online supplementary Figure 1b for a graphical overview. Studies focused on reporting associations between neurobiological outcomes and inflammatory markers (see below Associations between Inflammation and Neurobiology), preventing specific examination of the association between brain alterations and depression.

Inflammatory markers IL-6 and TNF-α were measured in three studies [29‒31], while CRP and IL-1β were examined in two investigations each [28, 29, 31]. Inflammatory marker levels were assessed using immunoassays, and concentrations were reported in median fluorescence intensity or absolute concentration (in mg/L). See online supplementary Figure 1c for a graphical overview.

Among the included studies, the association between the inflammatory markers IL-6, TNF-α, CRP, and IL-1β and depression diagnosis or depressive symptoms in children and adolescents was examined.

The association between the inflammatory marker IL-6 and depression diagnosis or depressive symptoms, respectively, was investigated in two studies. In detail, Ho and colleagues [31] in 2021 measured IL-6 levels in depressed adolescents and found no correlation with depressive symptoms. However, in the follow-up study in 2022, Ho et al. [30] reported significantly higher IL-6 concentrations in depressed adolescents compared with healthy controls.

The association between TNF-α levels and depression diagnosis or depressive symptoms, respectively, was investigated in two studies. In detail, Ho et al. [31] (2021) found no correlation between TNF-α levels in depressed adolescents and depressive symptoms. In contrast, Ho et al. [30] reported significantly higher TNF-α levels in depressed adolescents compared with healthy controls.

Liu et al. [28] assessed the CRP concentration in the study sample as a whole and reported the inclusion of one outlier with an abnormally high CRP level. After the exclusion of the outlier, CRP levels were in the normal range and did not correlate with depressive symptoms.

Ho et al. [31] measured IL-1β levels in depressed adolescents and found no correlation with depressive symptoms.

A detailed overview of associations between inflammatory and neurobiological outcomes in depressed children and adolescents can be found in Table 2.

The association between IL-6 and neurobiological function was investigated in three studies. In detail, Ho and colleagues [31] in 2021 reported a positive association between IL-6 and dorsal ACC glutamate levels but noted that this correlation was no longer significant after removing an outlier with an abnormally high dorsal ACC glutamate level from the analysis. In their subsequent 2022 study, the team found negative correlations between IL-6 levels and connectivity within white matter tracts that were not moderated by depression diagnosis [30]. In contrast, Bradley et al. [29] did not find any association between IL-6 and neurobiological function.

Potential associations between TNF-α and brain function were examined in three studies, of which two found no significant correlations [29, 31]. However, similar to their findings with respect to IL-6 levels, Ho et al. [30] found a negative correlation between TNF-α levels and white matter connectivity that was not moderated by depression diagnosis.

Correlations between CRP levels and neurobiological function were examined in two studies, and findings were mixed. In detail, Liu et al. [28] found a negative correlation between CRP and bilateral dorsal ACC activity and a positive association between CRP and bilateral nucleus accumbens activity while performing a reward task. However, neither association was moderated by depressive symptoms. The team further saw positive correlations between CRP and parts of the visual and attentional brain network (i.e., left lateral occipital, posterior temporal, and superior parietal cortices) as well as the left cerebellum; however, these associations disappeared following the removal of an outlier (see above Associations between Inflammation and Depression) from the analysis. In contrast, Bradley and colleagues [29] did not observe any associations between CRP and neurobiological function.

The two studies that looked at associations between IL-1β levels and brain function found no correlations between these factors [29, 31].

In addition to the four most prominent inflammatory markers, Bradley and team examined a multitude of other cytokines that showed correlations with neurobiological function while performing a reward task [29]. In detail, they found negative correlations of the cytokines IL-2, IL-13, and IL-15 with the right basal ganglia that were moderated by psychiatric symptoms (see above Depression Outcomes). The group further reported negative correlations between IFN-α2, IFN-γ, IL-3, IL-4, and IL-7 with the bilateral precuneus/posterior cingulate cortex, as well as negative correlations between IL-1α, IL-2, IL-5, IL-9, IL-13, IL-15, and TNF-β with right angular gyrus activity. However, none of the cytokine correlations with components of the brain’s default mode network were moderated by psychiatric symptoms.

This scoping review examined findings from four cross-sectional studies investigating the association between inflammatory markers and neurobiological function in depressed children and adolescents. All studies were conducted in the USA within the last 4 years. Studies focused primarily on four inflammatory markers, IL-6, TNF-α, CRP, and IL-1β, and conducted exploratory whole-brain as well as ROI analyses of the ACC, basal ganglia, and white matter tracts. While most studies reported negative correlations between neurobiological function and inflammatory markers, moderating effects of depression diagnosis or depressive symptoms were not observed.

Previous work has shown that depressed children and adolescents exhibit higher levels of IL-6 and TNF-α, indicating a feasible role for these inflammatory markers in child and adolescent depression [13, 26]. In the current review, one study confirmed an increase in IL-6 and TNF-α levels in depressed adolescents, while another study reported no correlation between the inflammatory markers and depression diagnosis [30, 31]. When examining the association between IL-6, TNF-α, and neurobiological function, correlations between those factors were seen; however, the direction of the correlations varied. In detail, one study reported a negative correlation of IL-6 and TNF-α levels with white matter connectivity; another study found a positive correlation between IL-6 levels and dorsal ACC glutamate concentrations [30, 31]. Previous studies have shown that adolescents diagnosed with depression exhibit lower connectivity in different white matter structures compared to healthy controls and that decreased connectivity in tracts linking the corpus callosum to the ACC predicts higher risk for depression [36‒38]. However, while white matter connectivity is altered in child and adolescent depression, correlations between this neurobiological function and levels of IL-6 as well as TNF-α do not seem to be moderated by depression diagnosis. In detail, depression diagnosis did not directly moderate the relationship between increased levels of IL-6 and TNF-α and decreased white matter connectivity, leaving room for speculation if white matter connectivity represents a stronger biomarker for child and adolescent depression than inflammation or if replication of the study with increased sample size is needed to detect underlying depression-mediated effects [30]. Furthermore, while previous research has shown that increased bilateral ACC activity is part of adolescent depression, the positive correlation between peripheral IL-6 levels and dorsal ACC glutamate levels in depressed participants in Ho et al.’s [31] study (2021) was driven by an outlier with abnormally high glutamate levels in the dorsal ACC [31, 39]. As such, findings should be interpreted with caution, and more research is needed to determine if exceptionally high levels of the pro-inflammatory cytokine IL-6 are positively correlated with altered activity in the dorsal ACC in depressed adolescents.

A recent meta-analysis reported increased CRP levels in depressed children and adolescents compared to controls, suggesting an important role of this inflammatory marker in early-onset depression [13]. The current review could not confirm this observation, with one study reporting no correlation between CRP levels and depressive symptoms [28]. However, when examining the associations between CRP levels and neurobiological function, the same study reported neuro-immune correlations that varied in their directions. In detail, the study found a positive association between CRP levels and bilateral nucleus accumbens activity during positive prediction error (reward was better than predicted) and a negative correlation between CRP levels and bilateral dorsal ACC activity during reward anticipation [28]. Previous research has shown that both the ACC and nucleus accumbens are involved in depression, with depressed adolescents displaying increased bilateral ACC activity as well as reduced bilateral nucleus accumbens volume and activity compared with healthy controls [8, 39]. However, while Liu et al. [28] reported increased CRP levels in psychiatric adolescents, they found reduced dorsal ACC and increased nucleus accumbens activity, contradicting previous neurobiological findings. Furthermore, Liu and colleagues [28] did not see mediating effects of depressive symptoms on the observed neuro-immune correlations, indicating that the various psychiatric symptoms reported in the study, rather than depressive symptoms alone, may have impacted the brain alterations. Additional studies focusing on depressive symptoms are needed to further investigate the potential role of CRP on reward-related alterations in the ACC and nucleus accumbens in depressed children and adolescents.

The role of IL-1β in child and adolescent depression is less clear, with few studies investigating the role of this inflammatory marker and reporting inconclusive results [26]. In this review, studies reported no association between IL-1β levels and depression diagnosis and no correlation between IL-1β levels and brain function [29, 31]. It is apparent that more studies are needed to further examine the role of IL-1β in child and adolescent depression. Interestingly, Bradley et al. [29] reported associations between other inflammatory markers and various brain regions in psychiatric as well as healthy adolescents. In detail, the research team found negative correlations between levels of multiple inflammatory markers (i.e., IFN-α2, IFN-γ, IL-1α, IL-2, IL-3, IL-4, IL-5, IL-7, IL-9, IL-13, IL-15, TNF-β) and activation of the posterior default mode network during reward anticipation and attainment, which were not mediated by psychiatric symptoms. They further showed negative correlations between levels of IL-2, IL-13, and IL-15 with right basal ganglia activation during reward attainment in the psychiatric group only, indicating a mediating effect of psychiatric symptoms on this association. The outcomes are in line with previous studies, showing that adolescents with mental health disorders display (1) decreased striatal brain activity, (2) increased default mode network connectivity, and (3) reward-related increased cytokine levels [40‒43]. However, Bradley and colleagues compared neurobiological function during reward anticipation and attainment against an implicit (inactive) baseline, making it difficult to interpret what exactly their neurobiological findings reflect. Thus, further research using an active baseline as a comparator when measuring brain activity during reward processing is needed to investigate the association between inflammation and reward-related neurobiological changes in psychiatric adolescents.

This scoping review was, to the best of our knowledge, the first to examine the evidence regarding the potential interaction between inflammation and neurobiological function in children and adolescents with depression. This review reported various correlations between inflammatory markers and neurobiological function in children and adolescents; however, current depression diagnosis or depressive symptoms did not moderate the associations. The results contradict findings of studies examining adult depression that suggest that altered inflammation may affect neurobiological function and connectivity in depression via dysregulated neurotransmitter concentrations. In detail, a systematic review in 2016 indicated that modified levels of inflammatory markers correlate with dopaminergic and glutamatergic dysregulation in subcortical and cortical regions in depressed adults, resulting in rest- as well as reward-related neurobiological dysfunction [10]. In a follow-up review, Han and Ham [44] proposed that peripheral inflammation causes functional and structural abnormalities in the brains of depressed adults through a specific sequential biological mechanism. In short, chronic stress leads to dysregulation of the hypothalamic-pituitary-adrenal axis that, in turn, causes an increased release of peripheral inflammatory markers and enhanced blood-brain barrier permeability. Subsequently, microglial cell activation provokes a neurotoxic effect by upregulating kynurenine-pathway metabolites, causing glutamate-mediated excitotoxicity in the neurons. The neurotoxic effect directly affects neural circuits involved in depressive-like behavior, inducing typical characteristics of depression. The missing moderating effect of depression on the neuro-immune correlations in children and adolescents may be due to the early stage of the mental health disorder. In detail, depressive symptoms in early-onset depression may be too subtle to be correlated with biological mechanisms in depressed children and adolescents, indicating that neuro-immune alterations may be present prior to depression onset. This speculation would be in line with Han and Ham’s [44] hypothesis, stating that peripheral inflammation leads to functional and structural abnormalities in the brain, which in turn induce depressive symptoms. However, this is purely speculative and needs to be further examined in future research, specifically targeting neuro-immune associations linked to early-onset depression.

This scoping review has several strengths, including a rigorous search strategy that captured a large body of evidence without limiting the search by date, country, setting, or language. Importantly, only studies including depressed children and adolescents ≤18 years were selected to ensure the investigation of a pediatric population. However, there are also limitations to consider. The four studies included in the current review examined depressed participants diagnosed with one or more comorbid psychiatric diagnoses (e.g., anxiety disorder, attention-deficit-hyperactivity disorder), which could have impacted the reported associations between neurobiological function and inflammatory markers. However, due to the frequent co-occurrence of depression with other mental disorders in children and adolescents, the findings of this review are more generalizable and therefore represent depressed child and adolescent populations found in clinical and community settings [25]. Furthermore, the current review restricted inclusion to studies that looked at four inflammatory markers (i.e., CRP, IL, TNF, IFN). While there is growing evidence that these are the most frequently implicated markers of inflammation in child and adolescent depression, it is possible that other components of the immune system that were not examined in the current review play a role in early-onset depression. Additionally, the neuroimaging methods used in the four included studies differed substantially, using whole brain as well as ROI imaging to examine gray and white matter activity and volume. The lack of commonly agreed brain areas and inflammatory markers across different studies limited comparability and interpretation, calling for a more uniform approach to screen for neuro-immune associations among depressed children and adolescents. Future studies are encouraged to focus on the most prominent neurobiological components and inflammatory markers involved in child and adolescent depression to further examine associations between the brain and the peripheral immune system in child and adolescent depression.

Research to date examining the association between inflammation and neurobiological dysfunction in depressed children and adolescents is scarce, and study designs are heterogeneous. Future research would benefit from incorporating standardized neuroimaging protocols for scanning, data pre-processing, and analytical pathways to enhance the uniformity and compatibility of study outcomes. Where applicable, well-established behavioral tasks should also be used in conjunction with imaging to provide insight into interactions with cognition (e.g., motivational processing) and facilitate direct comparisons with findings from the adult literature. Future research would also benefit from focusing on key inflammatory markers and brain regions that have been independently shown to be altered in child and adolescent depression. For example, there is reliable evidence that IL-6, CRP, and TNF-α levels are modified in depression among children and adolescents compared with healthy controls, such that these inflammatory markers should be included in future research in this area [13, 26]. Similarly, research to date suggests that the ACC, ventral prefrontal cortex, striatum, and amygdala may play a more distinct role in depressed children and adolescents compared to other brain areas [9, 45]. By ensuring that most frequently implicated inflammatory markers and brain regions are included, future research will be more likely to be able to comment on the reliability and validity of study results with greater confidence.

In summary, this scoping review suggests that children and adolescents display associations between neurobiological dysfunction and inflammatory activity without the moderating effects of depression diagnosis or depressive symptoms. Current research is restricted to a small number of highly heterogeneous studies with various limitations, such as small sample sizes, different brain regions of interest, and the assessment of multiple inflammatory markers, leading to inconclusive findings of neuro-immune associations in depressed children and adolescents. Further research with larger sample sizes and consistent methodological approaches is needed to advance the knowledge of potential neuro-immune interactions early in the course of depression.

An ethics statement is not applicable because this study is based exclusively on published literature.

The authors have no conflicts of interest to declare.

This study was not supported by any sponsor or funder.

A.S., J.M., S.C.C., A.C.H.L., and D.J.K. conceptualized the study design. G.B.-R. developed the search strategy and conducted the search in all databases. A.S., J.M., and S.C.C. screened records. A.S. and J.M. extracted results from the included studies. A.S. wrote and revised all drafts of the manuscript. D.J.K. and A.C.H.L. provided senior supervision for all aspects of the study. All authors assisted in data interpretation and critically reviewed and approved the final manuscript.

Protocol registration: Open Science Framework [21].

All data generated or analyzed during this study are included in this article and its supplementary material files. Further inquiries can be directed to the corresponding author.