Is psychosis a multisystem disorder? A meta-review of central nervous system, immune, cardiometabolic, and endocrine alterations in first-episode psychosis and perspective on potential models

Abstract

People with psychotic disorders show abnormalities in several organ systems in addition to the central nervous system (CNS); and this contributes to excess mortality. However, it is unclear how strong the evidence is for alterations in non-CNS systems at the onset of psychosis, how the alterations in non-CNS systems compare to those in the CNS, or how they relate to symptoms. Here, we consider these questions, and suggest potential models to account for findings. We conducted a systematic meta-review to summarize effect sizes for both CNS (focusing on brain structural, neurophysiological, and neurochemical parameters) and non-CNS dysfunction (focusing on immune, cardiometabolic, and hypothalamic-pituitary-adrenal (HPA) systems) in first-episode psychosis (FEP). Relevant meta-analyses were identified in a systematic search of Pubmed and the methodological quality of these was assessed using the AMSTAR checklist (A Measurement Tool to Assess Systematic Reviews). Case-control data were extracted from studies included in these meta-analyses. Random effects meta-analyses were re-run and effect size magnitudes for individual parameters were calculated, as were summary effect sizes for each CNS and non-CNS system. We also grouped studies to obtain overall effect sizes for non-CNS and CNS alterations. Robustness of data for non-CNS and CNS parameters was assessed using Rosenthal's fail-safe N. We next statistically compared summary effect size for overall CNSand overall non-CNS alterations, as well as for each organ system separately. We also examined how non-CNS alterations correlate CNS alterations, and with psychopathological symptoms. Case-control data were extracted for 165 studies comprising a total sample size of 13,440. For people with first episode psychosis compared with healthy controls, we observed alterations in immune parameters (summary effect size: g = 1.19), cardiometabolic parameters (g = 0.23); HPA parameters (g = 0.68); brain structure (g = 0.40); neurophysiology (g = 0.80); and neurochemistry (g = 0.43). Grouping non-CNS organ systems together provided an effect size for overall non-CNS alterations in patients compared with controls (g = 0.58), which was not significantly different from the overall CNS alterations effect size (g = 0.50). However, the summary effect size for immune alterations was significantly greater than that for brain structural (P < 0.001) and neurochemical alterations (P < 0.001), while the summary effect size for cardiometabolic alterations was significantly lower than neurochemical (P = 0.04), neurophysiological (P < 0.001), and brain structural alterations (P = 0.001). The summary effect size for HPA alterations was not significantly different from brain structural (P = 0.14), neurophysiological (P = 0.54), or neurochemical alterations (P = 0.22). These outcomes remained similar in antipsychotic naive sensitivity analyses. We found some, but limited and inconsistent, evidence that non-CNS alterations were associated with CNS changes and symptoms in first episode psychosis. Our findings indicate that there are robust alterations in non-CNS systems in psychosis, and that these are broadly similar in magnitude to a range of CNS alterations. We consider models that could account for these findings and discuss implications for future research and treatment.

Fig. 1

An overview and comparison of CNS and non-CNS alterations in first-episode psychosis. Figure 1a: Forest plot for magnitude of immune, cardiometabolic, HPA, brain structural, neurophysiological, and neurochemical alterations in first-episode psychosis compared with healthy controls. Each line represents a summary effect size for a meta-analysis in one parameter: squares represent the summary effect size for that parameter, with the horizontal line running through each square illustrating the width of the overall 95% CI. Blue diamonds represent summary effect sizes for immune, cardiometabolic, HPA, structural, neurophysiological, and neurochemical systems: the middle of each diamond represents the summary effect size, and the width of the diamond depicts the width of the overall 95% CI. Red diamonds represent summary effect sizes and accompanying 95% CI for non-CNS and CNS effect sizes. ES effect size, CNS central nervous system, FEP first-episode psychosis, HPA hypothalamic pituitary adrenal axis, IL1β interleukin-1β, sIL2-R soluble interleukin-2 receptor, IL6 interleukin-6, TGFβ transforming growth factor-β, CRP C-reactive protein, NAA N-acetylaspartic acid, N number. Figure 1b: Heat map comparing relative magnitude of effect sizes (ES) for immune, hypothalamic pituitary adrenal (HPA) axis, cardiometabolic, brain structural, neurophysiological, and neurochemical alterations in first-episode psychosis (FEP). The map is read from left to right, comparing parameters on the y axis with parameters on the x axis. A negative Wald score (blue squares) demonstrates that the parameter ES on the y axis is numerically lower compared with the intersecting parameter ES on the x axis. A positive Wald score (red squares) demonstrates that the parameter ES on the y axis is numerically higher than the intersecting parameter ES on the x axis. Numbers within the squares are the P values that accompany the Wald score, e.g., structural abnormalities show significantly smaller patient-control differences compared to immune abnormalities, and significantly greater differences compared to cardiometabolic abnormalities

Fig. 2

A summary of non-CNS alterations in first-episode psychosis, and a consideration of potential pathoetiology. Figure 2a: First-episode psychosis shows alterations in multiple systems in addition to the central nervous system. OGTT oral glucose tolerance test, HDL high-density lipoprotein, LDL low-density lipoprotein. Figure 2b–d: Models of the relationship between psychosis and non-CNS dysfunction. Figure 2b: Model 1: A risk factor induces non-CNS dysfunction, which may consequently impact CNS function to increase the risk of psychosis. Figure 2c: Model 2: A risk factor induces CNS dysfunction and thence psychotic symptoms, which may consequently trigger non-CNS dysfunction. Figure 2d: Model 3: A shared risk factor may result in the development of psychosis and non-CNS dysfunction through independent mechanisms. CNS central nervous system