**4. Conclusions**

The current structured review provides an overview of the recent evidence presented on the influence of oxidative stress and inflammation in PTSD. The studies reviewed here demonstrate the alterations of specific peripheral inflammatory markers that may potentially be implemented as correlates of PTSD, including the elevated levels of serum proinflammatory cytokines IL-1β, IL-6, and TNF-<sup>α</sup>. Among these, IL-6 has been shown to be elevated or reduced in the serum of individuals exposed to trauma according to the source of oxidative stress, such as whether the trauma is physical in nature, includes the presence of TBI or loss of consciousness, or entails psychosocial trauma. This may in turn sugges<sup>t</sup> the potential of serum IL-6 level as a peripheral marker for PTSD based on trauma type. This review also discussed the three neuroimaging-based studies on the inflammatory response associated with PTSD within the brain, from which can be concluded that particular brain regions may be identifiable as neural correlates of PTSD, including the prefrontal cortex, hippocampus, and amygdala of the brain. Specifically, the evidence from the studies reviewed revealed reduction in cortical thickness among regions of the prefrontal cortex and hippocampal volume, as well as a potential link between amygdala activity and systemic inflammation in PTSD.

Given the current findings summarized, a potential pathway underlying inflammation in PTSD may be suggested. A major collective finding from this review was elevated serum IL-6 levels in all but one study [48] among the studies reviewed for alteration in proinflammatory cytokines in association with PTSD across various trauma types. Considering that IL-6 has been known to cross the blood-brain barrier (BBB) as an immune mediator [105], it may then be that psychosocial stress as a result of trauma induces elevated peripheral IL-6 concentrations, which then influences the inflammatory cytokines within the brain via crossing of the BBB. This potential pathway may also be in alignment with previous animal model study findings which reported increased proinflammatory cytokines within specific regions of the brain associated with PTSD including the hippocampus, amygdala, and prefrontal cortex [106], all of which are in alignment with the three neuroimaging-based studies reviewed in the current article also (Table 4). Moreover, the same study also reported reduced anti-inflammatory cytokines within the three brain regions [106], suggesting a similar pathway in the case of anti-inflammatory cytokines with respect to psychosocial stress in PTSD as well. The over-expression of IL-6 and other proinflammatory cytokines within the brain as a result of peripheral cytokines crossing the BBB in response to trauma and psychosocial stress may then lead to neurodegeneration as previously described [107], and thus leading to neural tissue loss followed by dysfunction in the respective brain regions. Dysfunction in the regions of the prefrontal cortex, amygdala, and hippocampus are then responsible for functions of executive control, emotional lability, fear reactivity, and retrieval of traumatic memories as demonstrated in PTSD. The dysfunction of the three brain regions according to inflammation may further be explored through detailed clinical assessments according to function, such as neurocognitive tests and PTSD symptom severity as measurable by CAPS-5. Here, the alteration in glucocorticoid receptor regulation in the three brain regions may also influence the neurochemical profile of the respective brain regions. For instance, alteration in proinflammatory cytokines including IL-6 as a result of psychosocial stress may disrupt the relationship between proinflammatory cytokines and glucocorticoid receptors, whose function is to inhibit and regulate proinflammatory cytokines [108]. Considering that disruption in glucocorticoid receptor signaling affects immune function as well as hypothalamus-pituitary axis including levels of cortisol [109], these alterations may further reinforce the related clinical symptoms of PTSD such as sustained fear and anxiety.

It is also important to note that, among trauma-exposed individuals, many individuals were found resilient and did not develop PTSD. The distinct pathways between trauma-exposed individuals who are diagnosed with PTSD versus trauma-exposed individuals who are resilient to PTSD may be broken down to two possibilities, one of which is having a preexisting proinflammatory state that distinguishes between individuals who are at risk or resilient to PTSD as described in previous studies [110,111]. Here, the premise is that individuals who are at risk for PTSD have a distinct proinflammatory state prior to any exposure to trauma as compared with resilient individuals. Another perspective is the distinct responsiveness towards trauma that may vary according to genetic factors [21], such as the *ALOX12* locus as described in Table 4 [74]. Since the activation of many genes including *ALOX12*, *ALOX15*, and *RORA* have shown to alter according to oxidative stress in a PTSD model [108], a targeted approach that observes for the relationship between inflammation and trauma exposure according to genetic variants may identify the risk or resilience factors of PTSD.

A few limitations should be noted. First of all, the current review provided an overview of some of the major peripheral inflammatory markers that are targeted in the field of PTSD research. As such, a large portion of inflammatory cytokines including chemokines have been excluded in this review in order to take a targeted, narrow approach to identifying the potential biomarkers of PTSD. In addition, considering previous literature which emphasized the significance of particular genes in relation to PTSD and inflammation beyond the currently reviewed genes, future reviews that include a wider scope of methodological approaches in this topic are warranted for a more supportive mechanistic

interpretation of the findings. Furthermore, despite previous studies which suggested that immune mediators such as the proinflammatory cytokines discussed in this review are able to directly influence the CNS including the hippocampus of the brain through the crossing of the BBB [105], the current structured review included a small number of neuroimaging-based studies as supportive evidence. Future directions in the study of inflammatory responses within the brain in PTSD as well as other non-communicable diseases may benefit in further exploring the detailed alteration of the brain in response to stress through the review of various neuroimaging techniques, such as magnetic resonance spectroscopy (MRS), di ffusion tensor imaging (DTI), functional MRI (fMRI), and perfusion MRI. It is also noteworthy that current literature in this topic of interest largely consist of male participants, due to the nature of conditions that are most prevalent in the exposure of traumatic events. In order to better distinguish the sex di fferences that had been discussed in some of the studies reviewed, future research that consider a wider distribution of target populations according to trauma type is warranted. Lastly, the literature selection criteria of the current review excluded a number of studies that provided significant evidence with regards to the pathophysiology of PTSD due to its comorbid sample population. As PTSD has a prevalence of comorbidity with a vast number of other chronic diseases [7–9,31,112,113], the segregation between PTSD with other medical conditions may limit our understanding of its pathophysiology if observed independently. Further research that investigate the pathogenesis of PTSD through a multi-level approach of inflammatory responses, clinical symptoms, and brain structural and functional changes, may help determine the detailed underlying pathway of PTSD, and open up novel strategies of assessment and intervention for the disorder.

**Author Contributions:** Conceptualization, S.Y.; writing—original draft preparation, S.Y., T.D.K., and S.L.; writing—review and editing, S.Y.; supervision, S.Y.; funding acquisition, S.Y. All authors proofread the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the Brain Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (2015M3C7A1028373) and the ICT R&D program of Institute for Information & Communications Technology Promotion (B0132-15-1001).

**Acknowledgments:** The authors thank Shinhye Kim for her technical support and contribution towards the preparation of the tables and figures.

**Conflicts of Interest:** The authors declare no conflict of interest.
