Ongoing tireless research effort has been dedicated to elucidating the neurobiologic underpinnings of the etiology of schizophrenia for more than a half century. There has been tremendous renewed interest in the inflammation hypothesis of schizophrenia over the last two decades. The complex interactions between altered immune signaling and the brain were postulated in the development of schizophrenia [1,2]. This allegation came from the striking findings of elaborative studies that showed increased pro-inflammatory cytokine and decreased anti-inflammatory cytokine plasma levels [3], up-regulated inflammatory gene expression [4], evidence of inflammation in certain cortical areas in post-mortem brains [5], a higher prevalence of autoimmune disorder comorbidity [6], and symptomatologic improvement with anti-inflammatory drugs [7], in schizophrenia. Inflammation is a complex mechanism based on multiple biological pathways and closely linked to oxidative stress. Inflammatory processes both directly and indirectly interact with the central nervous system and result in neurodegeneration, microglial activation, dysfunction of hypothalamic-pituitary-adrenal (HPA) axis, disrupted glutamate metabolism and altered neurotransmission, all of which have been postulated to be involved in schizophrenia etiopathogenesis [8].

C-reactive protein (CRP) and albumin are acute phase reactants that respond to the systemic inflammation [9]. CRP is a well-documented circulating marker for systemic inflammation [10], and albumin is a negative acute-phase protein, serum levels of which are down-regulated in response to an inflammatory state [11]. Unlike proinflammatory cytokines, both CRP and albumin have longer half-lives and are easier to measure in a blood screening. Increased plasma CRP [12], and decreased serum albumin levels [13,14] have been shown in patients with schizophrenia in comparison with healthy controls. However, the CRP/albumin ratio (CAR) has been suggested to be a better indicator of an inflammatory response than CRP or albumin alone [15]. CAR has been proposed to reflect enhanced inflammatory tonus and increased levels of this biomarker are suggestive for poor prognosis and worse treatment outcomes in patients with systemic illnesses [15,16]. Previously, CAR has been studied in several neurological illnesses including Parkinson’s disease, migraine, and restless leg syndrome, and these studies have suggested that CAR might reflect increased inflammatory status in such neurological conditions [17-19]. In a recent study, the association between depression severity and CAR has been examined in patients diagnosed lung cancer [20]. However, to date, CAR has not been studied in schizophrenia or other psychotic disorders.

Peripheral biomarkers of inflammation derived from complete blood counts (CBC) have been investigated in numerous studies in recent years and suggested to reflect the pathogenesis of schizophrenia. The neutrophil/lymphocyte ratio (NLR), platelet/lymphocyte ratio (PLR), monocyte/lymphocyte ratio (MLR), red-cell distribution width (RDW), and mean platelet volume (MPV) are inexpensive and reproducible biomarkers of inflammation and easily measured in a CBC. Several lines of evidence showed elevated levels of these markers in both mood disorders [21,22] and schizophrenia [23]. However, studies are fraught with conflicting results, which are unable to unravel the complexity of the immunologic architecture of the clinical stages of schizophrenia, and raised questions as to whether acute relapse is associated with higher inflammatory tone and if such biomarkers are trait- or state-related [24].

In this study, alongside CAR and CBC-derived biomarkers of inflammation, we also focused on the neutrophil/albumin ratio (NAR), which has recently only been studied in patients with cancer, because higher NAR indicates an enhanced inflammatory status that worsens cancer prognosis and treatment response [25]. Likewise CAR, NAR has not been examined in psychiatric disorders to date. In this study, our aims were as follows: 1) to examine CAR and NAR as novel peripheral inflammatory markers in patients with both acutely exacerbated and remitted schizophrenia, and to compare them between patient and control groups; 2) to reveal whether peripheral inflammatory markers discriminated patients and controls and whether these markers varied between patients with acute exacerbation and remission; and 3) to determine optimum cutoff levels of CAR and NAR for the diagnosis of schizophrenia. We hypothesized that 1) regardless of being in relapse or remission, patients with schizophrenia have higher levels of CAR and NAR compared with healthy subjects, 2) CAR, NAR, and other inflammatory markers do not differ between relapsed and remitted schizophrenia groups, suggesting that inflammation is rather a trait marker for schizophrenia; and 3) there are positive correlations between CAR, NAR, and CBC-derived inflammatory markers. Thus, the results of this study could provide us to measure the inflammatory status between exacerbated and remitted patients in schizophrenia and specify a cut-off levels of CAR and NAR for the diagnosis of schizophrenia, which may be beneficial in clinical practice.

The principal findings of the present study were that CAR, NAR, NLR, PLR, MLR, RDW, and MPV were peripheral inflammatory markers that were increased in a relatively large sample of patients with schizophrenia compared with healthy controls. Despite CAR values having been recently reported in patients with Parkinson disease [19], restless leg syndrome [18], and migraine [17], this is the first study to examine CAR and NAR in patients with with schizophrenia.

In terms of the other inflammatory markers we studied, our findings were consistent with previous research. Two recent meta-analyses demonstrated that CBC-derived biomarkers of inflammation including NLR, PLR, and MLR were higher in patients with schizophrenia than in healthy subjects [23,27]; however, they reported no significant alterations of these markers between patients with acute and chronic disease.

CRP and albumin screening are readily available, usually within minutes of acute admission, and are often performed automatically as part of an admission profile. Albumin has a myriad of substantial physiologic effects that are essential for normal health. It has been well documented that serum albumin levels are correlated strongly with dietary protein intake [28]; however, recently, researchers and physicians have appreciated the influence of inflammation on the serum albumin concentration. Given that albumin is a negative acute- phase protein, inflammation leads to a decrease in serum albumin levels because of several factors, including reduced synthesis, increased catabolism, and translocation of albumin to extravascular pools from vessels [29,30]. Proinflammatory cytokines including tumor necrosis factor (TNF)-alpha and interleukin-6 (IL-6) are thought to induce albumin catabolism [31], and TNF-alpha has been shown to inhibit albumin gene expression in a murine model [32]. The CAR is a novel biochemical marker of systemic inflammatory response and has been associated with poor survival [16,33], and treatment response [34] in various cancer types, inflammatory activity in autoimmune diseases [35,36], and sepsis [15]. CAR has been suggested to be a better inflammatory marker of prognosis than CRP or albumin alone [15]. In addition, CAR has a weak but significant correlation with NLR [37]; however, in our study, CAR was strongly correlated with other peripheral inflammatory markers including NLR.

We also found that none of the peripheral systemic inflammation markers, except NAR, differed between patients with acute exacerbation and remitted schizophrenia, which may support the notion that inflammation is rather a trait-related phenomenon in schizophrenia. Chronic inflammation usually includes a lower grade inflammatory response and it plays a role in the pathophysiology of many neuropsychiatric diseases [38]. Despite there being a long-standing debate as to the extent to which systemic inflammation resembles neuroinflammation, it is now recognized that the brain is not entirely isolated from the peripheral inflammatory response; cytokines and other circulating inflammatory mediators can reach and influence the brain through several complex mechanisms that interact with the blood-brain barrier [39]. The blood concentrations of some inflammatory molecules appear to vary with the clinical status of schizophrenia; IL-1-beta, IL-6, and transforming growth factor (TGF)-beta were found to be state-related, whereas IL-12, interferon (IFN)-gamma, and TNF-alpha were seen to be trait markers [40]. However, the term ‘state’ is not clear enough to define whether it refers to relapse or clinical severity. For instance, an increase in IL-6, a so-called state marker, is positively associated with both positive and negative symptoms as assessed using the PANSS; however, such a clinical condition is not necessarily required to meet a clinical relapse threshold [41]. It is suggested that greater systemic inflammation generates a tendency towards a full-blown psychosis among individuals at clinical high-risk for psychosis [42]. First-episode psychosis is closely associated with increased inflammatory markers, despite the fact that such an association has not been explicitly observed in patients with schizophrenia with a clinical relapse [43,44]. Disturbances in blood cell counts, increased CBC-derived inflammatory parameters and elevated levels of circulating inflammatory mediators including CRP, were reported in schizophrenia independent of clinical course [45,46]. On the other hand, several studies showed that more severe psychopathology and clinical symptoms in schizophrenia could be related with increased peripheral inflammatory markers [47-49]. In light of previous findings, one can postulate that it remains inconclusive as to whether an acute relapse is directly associated with increased inflammatory signaling. In fact, increased inflammatory tonus, whether state- or trait-related, has been suggested to be involved in the pathogenesis of schizophrenia through dopaminergic and glutamatergic transmission alterations [2]. Our findings are consistent with the insight that inflammation is not directly associated with clinical relapse; it rather reflects a pathophysiologic process that may be related to the occurrence of schizophrenia. In our opinion, this may be because of that as the duration of the disease gets longer, the inflammatory status tends to remain stable despite exacerbations. However, we should admit that, the difference in NAR values between the acute exacerbation and remission groups is worthy of further research because a rational explanation is needed.

We found no association between clinical characteristics (illness duration, number of hospitalizations, and total duration of hospital stays) and inflammatory markers, except for a negative correlation between MPV and number of hospital stays and total duration of hospital stays. Miller and Goldsmith [42] stated that the chronicity of schizophrenia might be associated with increased inflammatory markers. However, they also reported that newly-diagnosed first-episode psychosis was also linked with heightened inflammatory markers. Anti-inflammatory agents are more efficacious in first-episode psychosis, but not in the latter episodes [50], which supports the idea that anti-inflammatory therapy would decelerate the clinical progression of the disorder at the early stage, but may not succeed in treating acute exacerbations of schizophrenia. There is also wide recognition that longer duration of disease is associated with metabolic disturbances, poor prognosis, and worse treatment response [51]; however, in our opinion, extended illness duration may be related with these features through different mechanisms other than increased inflammation in schizophrenia. Furthermore, we only found two of seven markers correlated with age, which has already been found positively correlated with inflammation in the general population [52]. Our results confirm prior findings that illness duration [47,53], and the number of hospitalizations [24,54] have not been clearly associated with peripheral inflammatory markers. As inpatient admission is mainly required for an acute exacerbation of patients in our hospital, the data of the number of hospital stays may reflect the number of acute relapses of the patients. These findings may indicate that increased inflammatory markers should not be solely attributed to age, illness duration or more relapses in schizophrenia.

We constructed ROC curves to determine optimum cutoff levels of CAR and NAR for the diagnosis schizophrenia. The cut-off value of CAR found in our study was 0.388 with a sensitivity of 81% and a specificity of 81%, and 0.885 as a cut-off value for NAR had lower sensitivity and specificity. To date, few studies have reported cut-off levels of peripheral inflammatory biomarkers such as NLR [55], IL-6 [56], and the high sensitivity-CRP/IL-10 ratio [57]; however; insufficient power of sensitivity and/or specificity for these markers were denoted. Despite there still existing copious room for research to appreciate its diagnostic use, our finding suggests that CAR, as an inflammatory biomarker, may be used to support the diagnosis of schizophrenia.

According to the logistic regression analyses, CAR, NAR, and MPV were significantly predictive of a diagnosis of schizophrenia which are novel findings. In the study of Wysokiński and Szczepocka [58], logistic regression analysis failed to demonstrate a significant predictive value of MPV for the diagnosis of schizophrenia, and the predictive values of CAR and NAR for schizophrenia have not yet been studied. Our findings suggest that a combination of inflammatory markers including CAR and NAR could be used to reflect the increased inflammatory status in schizophrenia.

We have to mention several limitations about the study. First, despite there being a line of evidence that demonstrates abnormal inflammatory parameters in drug-naive patients with first-episode psychosis [59], antipsychotics increase the risk of metabolic conditions, which are associated with inflammation, as such we cannot rule out antipsychotic medication’s influence on inflammatory markers. Second, age, a possible confounding factor, significantly differed between the groups which may be a result of the consecutive nature of the data acquisition in the study. It is has been suggested that advanced age is associated with chronic low grade inflammation and the number of people affected by age-related diseases including cardiovascular diseases, diabetes mellitus, cancer and dementia which are closely associated with inflammation, increases with age in both men and women [60]. With this regard, not matching group ages may not be considered a major limitation in terms of age-inflammation interplay because we have already excluded the patients with age-related diseases. Besides, regression analyses in recent studies revealed significant predictive values of inflammatory biomarkers for psychotic disorders, although the mean ages of the patient and control groups were significantly different [45]. However, we should disclose that mean age differences between our study groups might have an impact on serum levels of inflammatory markers. Third, despite we excluded obesity or being underweight and heavy smoking, BMI and smoking status of the patients might have affected our results. Finally, we did not examine relationship between psychopathological features and inflammatory markers.

In conclusion, this cross-sectional retrospective study showed that, except NAR, peripheral inflammatory markers did not significantly vary between patients with relapse and remission, suggesting that inflammation is rather a trait marker for schizophrenia. Additionally, we brought CAR and NAR as peripheral trait biomarkers, which may reflect enhanced inflammatory signalling in schizophrenia, to the relevant literature. On the other hand, both biomarkers should be further examined in larger patient groups, including their relationships with clinical symptoms and psychopathology. Yet, it is largely unknown as to whether peripheral inflammatory markers reflect or potentially have an impact on the prognosis and course of schizophrenia. It seems key to differentiate inflammation, either as a trait or state feature for schizophrenia, which is increasingly recognised as a pleiotropic disorder. Indeed, much more comprehensive and well-designed biomarker studies are warranted to yield an invaluable insight for characterizing inflammatory markers, whether they reflect the pathogenesis, or the clinical status of schizophrenia, or both, and for developing new potential diagnostic and advanced intervention strategies to reduce risk, treat, and improve quality of life in schizophrenia.