Peripheral Lymphocyte-Derived and Lipoprotein-Derived Inflammatory Ratios as Blood-Based Biomarkers in Schizophrenia Spectrum Psychosis: Predictive Value and Association with Gender, Clinical, Metabolic, Substance Use and Psychopathological, Factors

1. Introduction

Schizophrenia spectrum and other psychotic disorders are defined by the presence of one or more of the following: delusions, hallucinations, disorganized thinking and/or behavior, and/or negative symptoms [1]. This condition is frequently associated with significant distress and impairment in personal, family, social, educational or occupational areas of life [2].

The etiology of schizophrenia has been attributed to a combination of genetic, neurochemical, and environmental factors. Some theories include generalized brain multi-dysconnectivity, abnormal brain development, deregulated neuronal migration, impaired spatial neural arrangement, and absence of gliosis [3]. Traditional models have emphasized dopaminergic dysfunction as well as alterations in NMDA signaling [4].

A neuroinflammatory hypothesis of schizophrenia has gained empirical support from various findings. Studies have demonstrated decreased levels of anti-inflammatory markers, such as interleukins IL-10 and IL-4, while pro-inflammatory factors, such as IL-6, are increased in patients with a first psychotic episode of schizophrenia [3]. Activated microglia, along with reactive astrocytes and chemokines, creates a neurotoxic environment that can disrupt synaptic plasticity, neurogenesis, and white matter integrity [4]. This disruption in the pro-/anti-inflammatory axis may have clinically relevant consequences, including cognitive impairment, violent behavior, greater symptom severity, and treatment resistance [3,5,6,7].

The search for peripheral biomarkers that enable objective evaluation in schizophrenia should become a priority. Biomarkers, defined as biological characteristics that are objectively measured and evaluated as indicators of physiological processes, pathological conditions, or pharmacological responses, can be classified as diagnostic, prognostic, and theragnostic. In schizophrenia, their potential utility lies in improving diagnostic accuracy, predicting treatment response, and monitoring clinical progression [8].

Systemic inflammatory indexes derived from complete blood counts have attracted attention due to their accessibility, low cost, and reproducibility. Markers such as the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), monocyte-to-lymphocyte ratio (MLR), systemic immune-inflammation index (SII), systemic inflammatory response index (SIRI), as well as HDL-related ratios like MHR, NHR, LHR, and PHR, have been evaluated in various psychiatric and somatic conditions with promising results [9,10,11].

Emerging indexes, such as the Pan-Immune-Inflammation Value (PIV) and the Neutrophil–Platelet–to–Lymphocyte–Hemoglobin Ratio (NPLHbR), have shown significant associations in studies involving other psychiatric disorders, including bipolar disorder [12,13]. However, research on PIV levels in schizophrenia has yielded inconsistent results, with some studies reporting contradictory findings [14].

Studies in schizophrenia have consistently reported elevated levels of NLR, PLR, and MLR compared to healthy controls, even in early stages of the disorder or in first psychotic episodes [6,10]. These indexes have shown correlations with classical inflammatory markers such as CRP, as well as with clinical measures of severity, functioning, and violent behavior [7]. Furthermore, their longitudinal stability and independence from sex and age position them as robust candidates for clinical use [10].

Moreover, logistic regression models demonstrated that combinations of these indexes can predict the presence of schizophrenia [11]. These markers are accessible, cost-effective, and clinically valuable, therefore, in the future they might be useful also to anticipate relapses or clinically relevant events. The utility of these markers has also been explored in therapeutic contexts, showing that long-acting injectable antipsychotic treatments are associated with lower NLR levels, which may reflect better adherence or indirect modulation of inflammation [15]. Systemic inflammatory indexes represent a potentially valuable tool for characterizing this spectrum, but they should be properly contextualized within the patient's clinical profile [15].

Nevertheless, significant challenges remain. Despite growing empirical support, the specificity of these biomarkers is still questioned, as they can be influenced by numerous non-psychiatric factors such as current infections, obesity, smoking, or autoimmune diseases. Additionally, the directionality of the relationship between inflammation and psychosis is not yet fully understood [10].

Considering this background, the objectives of this research was to examine differences in peripheral blood inflammatory biomarkers between individuals with psychosis and healthy controls, and to explore their associations with clinical, psychopathological, gender and other factors. We also evaluated the discriminatory ability of selected biomarkers for psychosis diagnosis. Finally, we explored the interrelationships between biomarkers and continuous variables such as age, BMI, functionality and substance use via correlation analyses.

2. Materials and Methods

2.1. Study Design

This is a cross-sectional case-control study comparing individuals diagnosed with schizophrenia spectrum disorders (SSD, n = 354) and healthy controls (n = 76). Participants were recruited from the Parc Taulí University Hospital in Sabadell (Catalonia, Spain) between 2009 and 2026, within the framework of a broader ongoing research project on metacognitive, immunological, metabolic, therapeutic, and clinical parameters in psychotic disorders [16]. Healthy control participants were selected from hospital staff or casual visitors to non-psychiatric departments, ensuring the absence of current or previous personal psychiatric history or family antecedents of psychosis or bipolar disorders. All participants provided written informed consent, and the study protocol was approved by the Ethics and Research Committee of Parc Taulí Hospital, following national and international ethical standards.

2.2. Participant Selection

Inclusion criteria for the SSD group were: (a) age over 18 years, (b) diagnosis of schizophrenia or other psychotic disorders according to DSM-IV-TR1 or DSM-5 criteria [17], confirmed by trained psychiatrist using a structured clinical interview [18].

Healthy controls were included if they met the following criteria: (a) age over 18 years, (b) no current or past psychiatric diagnosis, (c) no current or past use of psychiatric medication, (d) no first- or second-degree relatives with schizophrenia spectrum disorders or bipolar disorder.

2.3. Psychopathological Assessment

Trained psychiatrists conducted structured interviews using the SCID-I for DSM-4-TR diagnoses18 or DSM-5 criteria [17]. Symptom severity was assessed using the Spanish version of the Positive and Negative Syndrome Scale (PANSS) [19,20]. Relevant sociodemographic and clinical information was recorded, including age at onset, duration of illness, number of psychotic previous episodes, previous hospitalizations, or current pharmacological treatments.

Healthy controls were evaluated with the Spanish version of the following tests: the 28-items General Health Questionnaire [21], Beck Depression Scale [22], the Spanish version of the Spielberg's Anxiety-State and Anxiety-Trait Scale [23], and the Mood Disorder Questionnaire [24]. Sociodemographic characteristics, anthropometric (including BMI) and use/abuse of alcohol and tobacco were identified in both groups. Global functioning was evaluated using the Global Assessment of Functioning (GAF) scale [1,25] in both groups.

2.4. Inflammatory Ratios and Laboratory Measures

Fasting blood samples were collected from all participants between 8:00 AM and 10:00 AM. Analyses were performed at the hospital’s clinical laboratory. Complete blood counts (CBCs), lipid profiles, thyroid hormones, and glucose levels were obtained.

The following inflammatory indexes were computed:

Lymphocyte-derived inflammatory ratios:

• Neutrophil-to-lymphocyte ratio (NLR): neutrophil count / lymphocyte count
• Monocyte-to-lymphocyte ratio (MLR): monocyte count / lymphocyte count
• Platelet-to-lymphocyte ratio (PLR): platelet count / lymphocyte count
• Systemic immune-inflammation index (SII), sometimes identified also as Neutrophil-to-lymphocyte and platelet ratio (NLPR): (platelets × neutrophils) / lymphocytes
• Systemic inflammatory response index (SIRI): (neutrophils × monocytes) / lymphocytes
• Pan-immune-inflammation value (PIV), sometimes identified also as Aggregate index of systemic inflammation (AISI): (neutrophils × monocytes x platelets) / lymphocytes
• Neutrophil–platelet–to–lymphocyte–hemoglobin ratio (NPLHbR): (Absolute number of neutrophils × Absolute number of platelets) / (Absolute number of lymphocytes × Hb level; g/dL)

Lipoprotein inflammatory ratios:

• Monocyte-to-HDL ratio (MHR): monocyte count / HDL value
• Neutrophil-to-HDL ratio (NHR): neutrophil count / HDL value
• Lymphocyte-to-HDL ratio (LHR): lymphocyte count / HDL value
• Platelet-to-HDL ratio (PHR): platelet count / HDL value

CBCs and the Friedewald formula were obtained with a Sysmex XE-2100 and lipid profiles were obtained using a Hitachi 917 (Roche Diagnostics). Indices were grouped into lymphocyte-derived (NLR, MLR, PLR, SII, SIRI) and lipoprotein-derived (MHR, NHR, LHR, PHR) inflammatory ratios.

2.5. Statistical Analyses

Descriptive analyses were conducted using Social Sciences version 29.0 (SPSS, Chicago, IL) and JASP version 0.19.3. Normality was tested with Shapiro–Wilk, Kolmogorov-Smirnov and visual inspection (histograms and normality plots).

Univariate Analyses: Mann–Whitney U tests were used for group comparisons of continuous variables. Chi-square and Fisher tests were used for categorical variables. Significance was set at p < 0.05.

Biomarker–Covariate Associations: To assess the influence of clinical and demographic variables on peripheral inflammation indices, Pearson’s correlation coefficients were calculated between continuous variables (age, BMI, alcohol and tobacco consumption) and each biomarker, separately for the psychosis and healthy control groups. Additionally, independent samples t-tests (or Mann–Whitney U tests) were performed within each group (psychosis and control) to compare biomarker levels across binary covariates (e.g., sex, tobacco use, alcohol use). Significance was set at p < 0.05.

Diagnostic Accuracy of Inflammatory Ratios: Receiver Operating Characteristic (ROC) curves were generated to evaluate the discriminatory power of inflammatory ratios between patients and controls. The area under the curve (AUC) was interpreted as follows: 0.5 = no discrimination, 0.7–0.8 = acceptable, 0.8–0.9 = excellent, > 0.9 = outstanding. Then, logistic regressions analyses were made using biomarker quartiles to predict psychosis diagnosis

Statistical consultation and data analysis support were provided by an AI-based research assistant (ChatGPT, OpenAI), which contributed to the organization of variables, selection of appropriate statistical tests, interpretation of results, and drafting of some preliminary sections of the manuscript. The assistant was used under the supervision of the researcher, the director, and following academic and ethical guidelines.

3. Results

3.1. Sample Characteristics

The sociodemographic characteristics of the sample are presented in Table 1. The mean age ± standard deviation (SD) of the mean age of participants with psychosis was 39.34 ± 16.8 years while the control group was 41.2 ± 13.1 years.

Regarding anthropometric measures, patients with psychosis showed higher body weight (75.7 ± 16.25 kg) and BMI (27.08 ± 5.85) compared to controls (68.48 ± 13.5 kg and 24.57 ± 4.5, respectively), these differences reached statistical significance.

Educational attainment was significantly lower among the psychosis group (p < 0.001), with fewer individuals attaining university-level education. Employment status also differed markedly between groups (p < 0.001), with higher unemployment and disability rates in the psychosis cohort.

Substance use patterns varied notably. The psychosis group reported significantly higher tobacco consumption (26.45 ± 19.7 vs. 14.5 ± 10.23 cigarettes/day; p = 0.009), with a higher proportion of smokers (p < 0.001). Although alcohol use was more prevalent in the control group (64.5% vs. 61.4%; p <0,001), alcohol consumption levels (in grams/week) tended to be higher in patients, though this did not reach significance (p = 0.053). No significant differences were observed in xanthine use (p = 0.18) or daily xanthine intake (p = 0.68).

3.2. Comparison of Blood Count Values, Lipoprotein Levels, and Inflammatory Ratios Between Psychotic Patients and Healthy Controls

Group comparisons revealed several significant differences in metabolic and hematological markers between SSD patients and healthy controls (see Table 2).

Among lipid parameters, individuals with psychosis exhibited significantly lower HDL cholesterol levels (46.27 ± 13.09 vs. 58.26 ± 18.62 mg/dL, p < 0.001), while total cholesterol was marginally lower in the SSD group but didn’t reach significance level (p = 0.053). Fasting glucose, LDL levels or triglycerides levels did not show any difference.

Hematological differences included elevated hemoglobin concentrations in the SSD group (140.83 ± 16.22 vs. 137.8 ± 13.6 g/L, p = 0.03), and a higher mean corpuscular hemoglobin concentration (MCHC; p = 0.01). SSD patients also showed significantly lower mean platelet volume (MPV; p = 0.01) and platelet distribution width (PDW; p = 0.01).

Monocytic parameters were elevated in the psychosis group, including both monocyte percentage (7.56 ± 2.12% vs. 6.76 ± 1.82%, p = 0.003) and absolute monocyte count (0.56 ± 0.21 vs. 0.48 ± 0.17 ×10⁹/L, p = 0.002), suggesting a possible low-grade pro-inflammatory state.

Several derived inflammatory indices also differed between groups. The monocyte-to-lymphocyte ratio (MLR) was significantly higher in patients (0.26 ± 0.13 vs. 0.23 ± 0.10, p = 0.01), as were the systemic inflammatory response index (SIRI; p = 0.04), lymphocyte-to-HDL ratio (LHR; p = 0.02), platelet-to-HDL ratio (PHR; p = 0.02), and especially the monocyte-to-HDL ratio (MHR; p < 0.001).

In contrast, no significant differences were observed for the neutrophil-to-lymphocyte ratio (NLR; p = 0.65), platelet-to-lymphocyte ratio (PLR; p = 0.059), systemic immune-inflammation index (SII - NLRP; p = 0.50), Pan-immune-inflammation value (PIV; p = 0.06), or Neutrophil–platelet–to–lymphocyte–hemoglobin ratio (NPLHbR; p = 0.32)

3.3. Differences in Biomarker Distributions by Sex/Gender

In the SSD group, significant sex-related differences were observed for the platelet-to-lymphocyte ratio (PLR; p <0.001), the Neutrophil–platelet–to–lymphocyte–hemoglobin ratio (NPLHbR; p <0.001) the monocyte-to-HDL ratio (MHR; p = 0.002) and the lymphocyte-to-HDL ratio (LHR; p = 0.002). Women showed higher PLR (127.89 ± 64.86) and NPLHbR (45.22 ± 33.94) values than men (PLR: 106.77 ± 44.69; NPLHbR: 34.54 ± 23.08); and lower MHR (0.01 ± 0.00) and LHR (0.04 ± 0.02) values than men (MHR: 0.01 ± 0.00; LHR: 0.05 ± 0.02). Among healthy controls, significant sex effects were observed for LHR (p = 0.005) and PHR (p = 0.01), with higher values in men (LHR: 0.05 ± 0.02; PHR: 5.73 ± 2.33). Results are shown in Table 3.

3.4. Differences in Biomarker Distributions by Substance Use

Tobacco Use/Abuse: In the psychosis group, patients who reported tobacco use had significantly lower PLR (p < 0,001), but higher MHR (p < 0.001), NHR (p = 0.006), LHR (p = 0,002) and PHR (p = 0.03) compared to non-users. This didn’t happen in the healthy controls (Table 3).

Cigarettes/day: When we analyze the relation between the amount of tobacco per day, in SSD group, and inflammatory biomarkers we see a statistically positive correlation with NHR (r = 0.29, p = 0.008), LHR (r = 0.25, p = 0.02), and PHR (r = 0.24, p = 0.03) (Table 4). No significant results were seen in the control group (Table 5).

Alcohol Use/Abuse: In the psychosis group, patients who reported alcohol use had significantly lower PLR (p = 0,007), SII-NLPR (p = 0.04) and NPLHbR (p = 0.01) compared to non-users. This, also, didn’t happen in the healthy controls (Table 3).

Alcohol/week: In the SSD group, MLR (r = 0.29, p = 0.03) and SIRI (r = 0.31, p = 0.02) were positively correlated with the amount of alcohol that patients consumed per week (Table 4). This positive correlation was even higher when considering LHR (r = 0.67, p = 0.03), in the control group (Table 5).

3.5. Differences in Biomarker Distributions by BMI and Age

In the SSD group, NLR, NPLHbR and NHR showed statistically significant but weak positive correlations with BMI (NLR: r = 0.15, p = 0.01; NPLHbR: r = 0.15, p = 0.01; NHR: r = 0.24, p = 0.03), suggesting only a minimal association between body mass index and systemic inflammatory activity. PLR and NPLHbR were also weakly correlated with age (PLR: r = 0.13, p = 0.01: NPLHbR: r = 0.13, p = 0.01). No other biomarkers were meaningfully associated with demographic variables. Even if statistically significant, these correlations were weak and probably not relevant from a clinical perspective (Table 4).

Among healthy controls, MHR, NHR, LHR, and PHR showed moderate and statistically significant positive correlations with BMI (MHR: r = 0.43, p = 0.02; NHR: r = 0.39, p = 0.03; LHR: r = 0.48, p = 0.008; PHR: r = 0.46, p = 0.01), suggesting that body mass index is moderately associated with these inflammation-based indices. Additionally, MHR and PHR correlated moderately with age (MHR: r = 0.45, p = 0.01; PHR: r = 0.44, p = 0.01). These results indicate that metabolic factors may exert a stronger influence on inflammatory ratios in healthy individuals than in patients with psychosis (Table 5).

3.6. Differences in Biomarker Distributions by Other Clinical Variables

Spearman correlation analysis was conducted to examine the relationship between inflammatory biomarkers and clinical variables. Total cholesterol showed a positive and significant correlation with years of evolution of the disease (r = 0.31, p = < 0.001). HDL cholesterol showed a small negative but still significant correlation with the number of psychiatric hospitalizations (r = -0.24, p = 0.002). LDL cholesterol was positively and significantly correlated with years of the evolution of the disease (r = 0.27, p = < 0.001) (Table 6). MHR and LHR showed a positive and significant correlation with the number of lifetime psychiatric hospitalizations (MHR: r = 0.21, p = 0.008; LHR: r = 0.16, p = 0.04), suggesting that this lipoprotein-based inflammatory ratio may be associated with previous number of acute psychotic episodes or clinical decompensations (Table 6).

3.7. Differences in Biomarker Distributions and Psychopathological Variables

Spearman correlation analyses were also conducted to examine the associations between peripheral inflammatory biomarkers and PANSS scores in SSD individuals (Table 7). HDL cholesterol showed a weak but still significant negative correlation with Total PANSS, Total PANSS negative and General (r = -0.23, p = 0.001; r = -0.21, p = 0.003; r = -0.21, p = 0.003).

When analysing Lymphocyte-derived inflammatory ratios, NLR, MLR, SII, SIRI, PIV, NPLHbR, all showed a significant negative correlation with PANSS positive symptoms (NLR: r = -0.21, p = <0.001; MLR: r = -0.15, p = 0.004; SII: r = -0.18, p = <0.001; SIRI: r = -0.19, p = <0,001; PIV: r = -0.17, p = 0.001; NPLHbR:r = -0.17, p = 0.001). Among the Lipoprotein inflammatory ratios, LHR showed a significant positive correlation (r = 0.15, p = 0.02) with PANSS positive symptoms.

If we consider PANSS negative symptoms, they correlate negatively with HDL cholesterol (r = -0.21, p = 0.003) and positively with MHR (r = 0.14, p = 0.04); LHR (r = 0.16, p = 0.02) and PHR (r = 0.16, p = 0.02).

Total PANSS and Total PANSS general symptoms showed a negative correlation with HDL (r = -0.23, p = 0.001; r = -0.21, p = 0.003) and positive correlation with LHR (r = 0.2, p = 0.005; r = 0.19, p = 0.006) and PHR (r = 0.17, p = 0.01; r = 0.16, p = 0.02).

Functionality, measured by the GAF (Global Assessment of Functionality), was not related to any index in our sample (Table 7).

3.8. Predictive Values of Inflammatory Ratios for Patients with SSD

To evaluate the ability of peripheral inflammatory biomarkers to distinguish between patients with psychosis and healthy controls, logistic regression with receiver operating characteristic (ROC) curve analyses were performed.

Among the tested markers, the monocyte-to-HDL ratio (MHR) showed the best discriminative performance, with an AUC of 0.686 and a highly significant p-value (p < 0.001), indicating acceptable diagnostic accuracy. Additional ratios with statistically significant but modest discriminative ability included the lymphocyte-to-HDL ratio (LHR) (AUC = 0.629, p = 0.03) and the platelet-to-HDL ratio (PHR) (AUC = 0.632, p = 0.03).

The neutrophil-to-HDL ratio (NHR) (AUC = 0.572, p = 0.002), reached statistical significance but fell below the threshold for acceptable discrimination.

These findings support the potential of MHR, and to a lesser extent LHR and PHR, as accessible biomarkers to aid in the identification of patients with SSD (Table 8).

3.9. Predictive Values of Inflammatory Ratios for Patients with SSD Based on Biomarker Quartiles

Logistic regression models were conducted to assess whether inflammatory biomarkers, categorized into quartiles, were associated with the likelihood of an SSD diagnosis. Healthy controls and patients with psychosis were compared across quartiles for each biomarker (Table 9).

Among all models tested, Monocyte-based and lipoprotein-based ratios showed the strongest associations with psychosis diagnosis. Specifically, individuals in the highest quartile of Monocyte-to-Lymphocyte Ratio (MLR) had significantly increased odds of psychosis (OR = 2.16, p = 0.035), and Monocyte-to-HDL Ratio (MHR) demonstrated a robust association, with quartile 3 (OR = 5.89, p = 0.008) and quartile 4 (OR = 4.33, p = 0.015) showing significantly elevated odds compared to the reference group.

Lymphocyte-to-HDL Ratio (LHR) was also significantly associated with psychosis, with quartile 3 (OR = 3.15, p = 0.04) predicting increased diagnostic likelihood. Similarly, Platelet-to-HDL Ratio (PHR) quartile 3 showed a significant association (OR = 4.32, p = 0.03). Notably, Platelet-to-Lymphocyte Ratio (PLR) showed a significant inverse association in quartile 3 (OR = 0.35, B = -1.030, SE = 0.390, Wald = 6.975, p = 0.008), suggesting a potentially protective role.

By contrast, Neutrophil-to-Lymphocyte Ratio (NLR), Systemic Immune-Inflammation Index (SII), Systemic Inflammatory Response Index (SIRI), and Neutrophil-to-HDL Ratio (NHR) did not significantly predict psychosis across quartiles. Models for these indices failed to reach statistical significance, and none of their individual quartiles showed reliable predictive value.

Additional indices including Pan-Inflammatory Value (PIV) - Aggregate index of systemic inflammation (AISI), or Neutrophil-Platelet-to-Hemoglobin Ratio (NPLHbR) similarly failed to show any significant associations. All models for these ratios were statistically non-significant, with low explained variance and no individual quartile yielding a meaningful odds ratio. 3.2. Tables.

4. Discussion

This study aimed to examine the differences in peripheral inflammatory biomarkers between individuals diagnosed with SSD and healthy controls, and to explore how these biomarkers relate to gender, age, substance use/abuse and clinical factors. Several relevant findings emerged.

SSD patients showed lower levels of HDL cholesterol, mean platelet volume and platelet distribution width, and higher levels of inflammatory biomarkers such as MLR, SIRI, MHR, LHR and PHR indexes in comparison with healthy controls. Suggesting increased peripheral inflammation levels in the psychotic patients.

Monocytes were significantly higher in young people at ultra-high risk for psychosis in comparison with healthy controls in recent studies [26], where white cells were significantly associated with remission status in people with higher risk of psychosis. This demonstrates that white blood cell proportions might be an interesting parameter to consider when evaluating ultra-risk people.

Monocytes were significantly higher also in psychotic patients compared to healthy controls. This finding aligns with a recent meta-analysis by Dudeck et al. [27], which reported elevated monocyte and neutrophil levels in individuals with schizophrenia relative to controls, interpreted as evidence of innate immune system activation. These results have a direct impact in MLR, MHR, SIRI indexes which were all significantly higher in psychotic patients compared to HC; consistent with previous studies where these indexes were evaluated as predictive factors for schizophrenia and bipolar disorder [11].

Among the lymphocyte-derived inflammatory ratios, the neutrophil-to-lymphocyte ratio (NLR) in our sample did not differ significantly between patients with psychosis and healthy controls. These findings stand in contrast to a recent study [28], which reported elevated NLR values in individuals with schizophrenia compared to controls. This inconsistency underscores the variability in inflammatory profiles across psychosis samples, and the importance of considering confounding factors, such as nationality/race, environmental/socioeconomic factors or medication they receive for mental health issues and other clinical treatments and pathologies. For instance, Sandberg et al. [29] found some interesting trends in this regard in their scoping review. They reported studies were first-episode psychosis and early-onset schizophrenia individuals showed higher neutrophil counts and NLR; attributing this to the disease being untreated. They also showed mixed results in studies where the effect of antipsychotics was considered.

Considering lipoprotein-based inflammatory ratios, Monocyte-to-HDL ratio (MHR), Lymphocyte-to-HDL ratio (LHR), and Platelet-to-HDL ratio (PHR) were elevated in psychotic patients. This finding is consistent with previous studies reporting lower HDL cholesterol levels in individuals with psychosis compared to healthy controls [30]. Moreover, increased levels of these biomarkers have been associated with aggressive behavior in psychotic populations [31].

Sex and substance use emerged as relevant factors influencing inflammatory biomarker levels, particularly among psychotic patients. In this group, men displayed significantly higher values of MHR and LHR compared to women, but lower values in PLR and NPLHbR. These sex-related differences were also evident in the healthy control group, though for LHR and PHR only, again with higher values in men. Another previous study also showed similar results analyzing NLR, MLR, PLR, and SII-NLPR and sex differences [32]. In fact, gender-differences in SSD is a highly relevant issue [33]. This suggests that sex-based physiological differences may have influence in inflammatory markers and should be considered in future related analyses.

Some biomarkers in the psychosis group, such as NLR, NPLHbR, and NHR, showed statistically significant correlations with BMI and in the case of PLR and NPLHbR the correlation was seen with age. These associations were weak and lacked clinical relevance. In contrast, among healthy controls, MHR, NHR, LHR, and PHR demonstrated moderate and significant correlations with BMI. Altogether, we see that inflammatory markers can vary for many reasons. Yet, this could also indicate that metabolic factors may exert a stronger and more consistent influence on inflammatory indices in healthy individuals than in patients with psychosis, where other biological or pharmacological variables may affect the metabolic outcomes [34].

Tobacco use was also associated with increased inflammatory activity in the psychosis group, as users showed significantly elevated PLR, MHR, NHR, LHR, and PHR levels. Notably, these associations were not observed in the control group. These findings are seen in previous research suggesting that tobacco use has a significant impact on lymphocyte-derived inflammatory ratios, particularly NLR, as reported by Bioque et al. [35]. Interestingly, in our study, tobacco use was not associated with NLR but instead showed a positive association with lipoprotein-derived inflammatory ratios, suggesting potentially distinct inflammatory pathways linked to smoking behavior. When considering alcohol use, consumption was associated with PLR, SII-NLPR, and NPLHbR in the SSD group. Again, these did not happen in the healthy control group.

We have also analyzed the inflammatory indexes with the amount of tobacco and alcohol they consume. Results showed a positively significant correlation between alcohol per week and MLR and SIRI indexes in the SSD group. This correlation was stronger in the healthy control for LHR index. When considering cigarettes per day, NHR, LHR and PHR showed a positive, significant correlation in the SSD group, but none of the indexes showed this among healthy controls.

The associations between inflammatory biomarkers and key clinical variables within the psychosis group, including years of illness evolution, Global Assessment of Functioning (GAF), and the number of lifetime psychiatric hospitalizations suggest that some lipid-related biomarkers, particularly cholesterol subtypes MHR and LHR, may be meaningfully linked to disease progression and clinical severity. Total cholesterol and LDL cholesterol demonstrate statistically significant positive correlations with years of illness evolution and HDL cholesterol showed a negative but still significant correlation with the number of psychiatric hospitalizations. This may reflect cumulative metabolic alterations associated with chronicity and repeated decompensations in psychotic disorders, suggesting that reduced anti-inflammatory lipid components may be linked to greater clinical instability. In alliance with other studies showing a relation between metabolic syndrome with the cognitive impairment in Schizophrenia [36] and contributing also with functional decline through the course of illness. Nonetheless, current evidence remains largely correlational, and the directionality of these associations has not been definitively established [37].

Among the derived inflammatory ratios, MHR and LHR were the only markers significantly associated with clinical outcomes, showing a positive correlation with the number of psychiatric hospitalizations, reinforcing its potential relevance as biomarkers of clinical severity or recurrence risk.

When considering current psychopathology (PANSS) as a variable, we saw that both lipoprotein-derived inflammatory ratios (LHR and PHR) were correlated with the Total PANSS domain. On the contrary, HDL cholesterol showed a small, negative, but still significant correlation with this domain. The positive PANSS domain exhibited the strongest and most consistent correlations with inflammatory indices. NLR, MLR, SIRI, SII-NLPR, PIV-AISI, NPLHbR they all showed negative and significant correlations. LHR was also significantly correlated, but positively. Regarding the Negative and General PANSS domain, HDL showed a negative correlation in both cases, and then LHR and PHR were positively correlated with both. Overall, the consistent associations between lymphocyte-derived and lipoprotein-derived ratios and multiple PANSS domains, especially positive symptoms, indicate that metabolic-inflammatory pathways may play a meaningful role in the clinical expression of psychosis. However, opposite results were reported in a previous study, which found no significant associations between changes in peripheral inflammatory markers and PANSS symptom domains [38].

The diagnostic accuracy analysis using ROC curves indicated that individual biomarkers had limited predictive value for psychosis, as AUC values across all indices remained below clinically relevant thresholds. However, our recent study in patients with Bipolar Disorder Type I found that MHR was a useful predictor in that specific population [39]. This discrepancy suggests that these biomarkers may have greater predictive utility within certain subtypes of SSD. Stratifying by diagnostic subtype could help clarify whether these markers are more informative in conditions characterized by affective symptomatology. Notably, widely used indices like NLR, PLR, SII-NLPR, and SIRI did not demonstrate sufficient discriminatory power in our sample (AUCs near 0.5, p > 0.05).

Among the biomarkers evaluated, lipoprotein-based MLR, MHR, LHR, and PHR emerged as significant predictors when modeled in quartiles, as in our previous study [39]. Conversely, commonly previous lymphocyte-based studied markers such as NLR, SII-NLPR, and SIRI failed to demonstrate predictive value, reinforcing the need to reassess which inflammatory indices are most informative in this clinical context. While these results are promising, the modest variance explained and inconsistent significance across quartiles underscore the need for further validations.

The main limitation of this study is the cross-sectional design, which reveals associations but not causation. Inflammatory and chronic diseases were collected in the sample, but not further analysis were made, meaning that we can’t deny possible interactions between the biomarkers and preexisting conditions. Also, some patients may have presented with subclinical or undiagnosed acute inflammatory conditions at the time of sample collection. Additionally, healthy controls collection was voluntary, and the sample number was limited. Further studies should be made to replicate our findings.

The strengths of this study rely on the observational characteristics of the sample and the adequate number of cases we were able to include. Recruitment of patients had a naturalistic approach, with homogeneity of diagnosis. Another strength of this study is the inclusion of a broad panel of inflammatory biomarkers, allowing for a more comprehensive and detailed assessment of the participants' inflammatory profiles.

5. Conclusions

This study set out to examine the differences in peripheral blood inflammatory biomarkers between individuals with psychosis and healthy controls, and to explore how these biomarkers are influenced by demographic factors, substance use, and clinical variables. In addressing the first objective—comparing biomarker levels—our findings revealed that patients with psychosis showed significantly higher levels of monocytes and derived ratios such as MLR, MHR, LHR, and PHR. These results support the hypothesis that psychosis is associated with a heightened state of systemic inflammation, even after adjusting for potential confounding variables such as sex and substance use.

When assessing the impact of sociobiological factors, sex-based differences in inflammatory markers were evident, particularly within the SSD group, where SSD women showed elevated PLR and NPLHbR, while SSD men had higher MHR and LHR levels. These patterns were less pronounced in healthy controls, suggesting that sex may modulate immune-inflammatory responses differently in psychotic populations.

Regarding substance use/abuse and quantity, the results reveal that inflammatory responses to substance use may differ markedly between individuals with psychosis and healthy controls. In the SSD group, tobacco and alcohol use were associated with significant alterations in various inflammatory biomarkers—such as increases in MHR, NHR, LHR, and PHR for tobacco, and correlations between alcohol intake and markers like MLR and SIRI—while no such patterns were observed in healthy individuals. This divergence suggests that the biological impact of substances may be amplified or modulated by underlying psychiatric vulnerability. It reinforces the notion that psychosis is not only characterized by baseline immune dysregulation, but also by a distinct reactivity to external factors like substance use, which may contribute to clinical instability or progression.

The analysis revealed that lipid-related biomarkers may serve as indicators of illness progression in SSD. Total and LDL cholesterol were positively correlated with years of illness evolution, while HDL cholesterol was negatively associated with the number of lifetime psychiatric hospitalizations. Additionally, MHR and LHR showed positive correlations with hospitalization history, suggesting that these lipoprotein-based ratios could reflect cumulative clinical burden or recurrence risk in psychosis. Additionally, when examining psychopathological symptoms, lipoprotein-derived ratios—particularly LHR and PHR—were positively associated with current total PANSS scores and multiple symptom domains, including negative symptoms. In contrast, HDL cholesterol showed consistent negative correlations with total, general, and negative PANSS scores. Lymphocyte-derived and composite inflammatory markers such as NLR, SIRI, SII-NLPR, PIV-AISI, and NPLHbR were negatively correlated with positive PANSS symptoms. These patterns support the potential role of immune and metabolic dysregulation in shaping the clinical presentation of SSD, particularly in relation to symptom dimensions.

Regarding the discriminatory ability of selected biomarkers for psychosis through ROC curve analyses. While some indices showed statistically significant group differences, none reached clinically meaningful AUC thresholds. Although, some of them emerged as significant predictors when modeled in quartiles. These findings suggest that, when considered in isolation, peripheral inflammatory biomarkers have limited diagnostic utility for psychosis. However, evidence from previous studies in other psychiatric populations suggests that markers like MHR may have greater predictive power within specific subtypes.

In summary, the study confirms the presence of altered inflammatory profiles in individuals with psychosis, with MLR, SIRI, MHR, LHR, and PHR emerging as potentially relevant markers. However, their variability across demographic and behavioral subgroups and their limited predictive value highlights the complexity of using peripheral biomarkers in clinical settings. While our findings advance understanding of immune alterations in psychosis, they also underscore the need for more targeted, longitudinal, and diagnostic subgroup differentiation studies to clarify the role of systemic inflammation in psychiatric illness.