The effect of excess weight on circulating inflammatory cytokines in drug-naïve first-episode psychosis individuals

Inflammation is usually a physiological response of the organism to harmful stimuli, whether physical, chemical, or biological. The inflammatory state that accompanies excess weight shows a peculiar presentation, as it is not accompanied by infection or sign of autoimmunity and no massive tissue injury seems to occur. In fact, it is often called “low-grade” chronic inflammation [17]. Cytokines are hormone-like proteins that enable immune cells to communicate and play an integral role in the initiation, perpetuation, and subsequent down-regulation of the immune response [28]. Several studies have found that overweight individuals (body mass index (BMI) ≥ 25) have higher plasma levels of some pro-inflammatory cytokines such as IL-6, TNF-α, and IL-1β than normal-weight individuals [16, 22].

Obesity is reported in approximately 50% of patients with schizophrenia [2, 3]. In addition, a pro-inflammatory state has been repeatedly associated with first-episode psychosis (FEP) [5, 11, 12, 15], especially in women, overweight patients, and tobacco users [9, 10]. However, no previous study has addressed the direct effect of BMI on basal serum cytokines in FEP individuals.

The aim of this study is to analyze the effect of BMI on basal serum cytokine levels in FEP patients and control subjects, separating the total sample into two groups: normal-weight and overweight individuals. Our hypothesis is that higher levels of inflammatory cytokines will be observed in patients than in controls in the overweight group with respect to this same comparison in the normal-weight group.

These findings support our main hypothesis that excess weight can alter the homeostasis of the immune system and therefore may have an additive pro-inflammatory effect on the one produced by psychosis in the central nervous system (CNS). As far as we know, this is the first research that studies the direct effect of excess weight on basal serum cytokine levels in individuals with FEP.

In the present study, we separated the subjects into two different groups according their BMI (Table 1). In the normal-weight group, only the IL-8 (chemokine) was higher in patients than in healthy volunteers, whereas in the overweight group, we found more differences, as expected: patients had higher levels than the control subjects of two pro-inflammatory cytokines (IL-6 and IL-1β), two chemokines (IL-8, MIP-1β), four Th-1 and Th-2 cytokines (IL-13, IL-2, IL-7, IL-12p70), as well as one Type-3 cytokine (IL-23) (Figs. 1 and 2). IL-8 seems to be the chemokine more related to psychosis and might be considered as a probably state/trait marker of the disease. This is supported by the fact that its levels are not influenced by weight (no differences were found between overweight patients and normal-weight patients, as shown in Fig. 4), as well as its levels are higher in patients than in healthy volunteers in both overweight comparison and in normal-weight comparison (see Figs. 1 and 2). In addition, one article by Hayes et al. (2014) found higher IL-8 levels in the cerebrospinal fluid (CSF) of schizophrenia patients than in healthy volunteers; they concluded that the differences observed were consistent, suggesting these may represent a trait marker associated with psychosis. Regarding the rest of altered cytokines between patients and controls in the overweight group, the cytokines which are more related to overweight are IL-6, TNF-α, and IL-1β. The fact that these cytokines are not higher in the overweight control group than in the normal-weight control group suggest that excess weight in isolation, without the synergistic pro-inflammatory effect of psychosis, might not be enough to alter these cytokines in some cases.

Previous studies carried out in FEP either have excluded the overweight individuals [7, 20, 23] from the analysis or have not considered the BMI in the analysis [6, 18] because the result was similar between patients and controls. Regarding the two articles that did not consider the effect of the BMI, Di Nicola et al. [6] and Noto et al. [18] observed that IL-6, TNF-α, and IL-8 were higher in patients than in healthy volunteers. On the other hand, regarding the three studies which excluded overweight individuals from the analysis, Ding et al. [7] found that IL-6 and IL-17 were increased in patients when compared to controls, Song et al. [23] found that IL-6, TNF-α, and IL-1β were elevated in patients compared to healthy volunteers, and [20] observed that IL-6 and IL-2 were higher in patients than in controls.

We suggest three possible reasons that may explain the discrepancies between our findings in the normal-weight group, in which only IL-8 was increased in patients as opposed to the control subjects, and previous studies that excluded the overweight individuals [7, 20, 23], where several pro-inflammatory cytokines as well as Th-1 and Th-2 cytokines are altered. First, whereas our study population is of Caucasian origin, the studies by Ding et al. [7] and Song et al. [23] were conducted in Asian populations, so some of the differences could be explained by genetic background. Second, cannabis, which is known to play a role in the homeostasis of the immune system [24], has not been considered as a possible confounding variable in the match of any of the three studies. Third, it is important to remark that individuals with identical BMI may have different body compositions regarding fat and muscle, so we cannot exclude the contribution of adipose tissue or of other lymphoid or non-lymphoid cells to the elevated levels of pro-inflammatory cytokines [6, 26].

Excess weight may produce a pro-inflammatory effect when BMI is equal or over 25 [16, 22]. Thus, although there is a statistically significant difference between BMI of healthy volunteers and patients in the normal-weight group (see Table 1), as BMI is under 25 in both subgroups, the difference observed should not assume any effect on the cytokines studied.

In this study, the overweight controls used more cannabis than the overweight patients (p = 0.001), as shown in Table 1. Cannabis may have an anti-inflammatory effect [24] that could explain why more differences have been found in the overweight group with regard to the normal-weight group. Thus, we have repeated the previous analysis excluding cannabis-use individuals and found that the results remain similar. Because the cannabis variable is qualitative (consumption yes/no), we may speculate that our findings seem to support the notion that overweight controls might consume small amounts of cannabis and therefore it seems not to have any anti-inflammatory effect on the cytokines studied.

Regarding the likely effect of age in our study, patients in the overweight group are older than the controls (p = 0.017), as shown in Table 1. Álvarez-Rodríguez et al. [1] described a positive correlation between age and some circulating pro-inflammatory cytokines, such as TNF-α, IL-1β, IL-6, and IL-12. However, this correlation was observed for individuals with a mean age difference much greater than the one observed in our sample (25 years of age difference VS 5 years of age difference, respectively). In addition, as shown in Fig. 2, patients in the overweight group had higher levels of other cytokines besides the pro-inflammatory ones that have not been previously correlated with age, such as two chemokines (IL-8 and MIP-1β) and also several Th-1 and Th-2 cytokines (IL-13, IL-2, and IL-7).

Three original studies [8, 13, 21] and one meta-analysis [27] have found elevated mRNA expressions of several pro-inflammatory cytokines and Toll-like receptors in post-mortem brains of individuals with schizophrenia. Moreover, as it was mentioned above, Hayes et al. [14] found higher levels of IL-8 in the CSF of psychosis individuals compared to healthy volunteers. However, peripheral blood is a biological matrix not sensitive enough to detect immune disturbances in the CNS. Furthermore, it is possible that CNS immune changes may be only observed after a second hit. Thus, the inflammation induced by the overweight might be the spark that initiates the fire and the inflammation caused by psychosis would be the second hit [16, 22, 25]. The fact that the largest differences in the basal cytokine levels between patients and healthy volunteers were found in the overweight group might be explained by the additive pro-inflammatory effect of both excess weight and psychosis [4, 22]. This is supported by the analyses comparing the basal serum cytokine levels between normal-weight and overweight individuals within the groups of healthy volunteers and patients; Fig. 3 shows no differences in the cytokine levels between overweight and normal-weight healthy volunteers, whereas in Fig. 4, it is depicted that overweight patients have higher inflammatory cytokine and chemokine levels than the normal-weight patients. Keeping this in mind, we suggest that cytokines are probably not the best biomarkers to detect the inflammation that is taking place in the CNS because they are influenced by several confounding factors such as BMI, cannabis use, the taking of analgesics and anti-inflammatory drugs, stress, and age.

This study has several limitations that should be considered: (1) stress level, which is known to influence inflammatory status, was not measured; (2) we have not been able to gauge the possible influence of adipose tissue on the pro-inflammatory cytokines; (3) we were not able to consider the consumption of classical analgesic or anti-inflammatory drugs as possible confounding factors (however, it is important to remark that the subjects from this study are young, healthy individuals without chronic diseases or chronic treatments); and (4) analyses were performed on peripheral blood. We assume that these findings are an indirect reflection of the CNS events; therefore, we would expect to find a greater disturbance of these cytokines in the CNS.

We found that the most remarkable differences in the basal serum cytokine levels between patients and healthy volunteers were found in the overweight group. In addition, the overweight patients had higher levels of several inflammatory cytokines and chemokines than the normal-weight patients, whereas no differences were found between the normal-weight controls and the overweight controls. Considering these findings, we suggest that excess weight may help to detect immune alterations in peripheral blood that are taking place in the CNS by psychosis, due to a possible additive pro-inflammatory effect that alters the homeostasis of the immune system.