Predictive value of C-reactive protein and hematologic ratios for metabolic syndrome in schizophrenia

Abstract

Objectives

This study aimed to examine the association between the inflammatory biomarkers—C-reactive protein (CRP), neutrophil–lymphocyte ratio (NLR), platelet–lymphocyte-ratio (PLR), and monocyte–lymphocyte ratio (MLR)—and metabolic status in a sample of patients with schizophrenia. We tested the hypothesis that individuals with schizophrenia exhibit a high prevalence of metabolic alterations, and that these inflammatory biomarkers are associated with a greater number of metabolic risk factors and the presence of metabolic syndrome (MetS), after adjusting for potential confounders.

Methods

A total of 541 participants with a DSM-5 diagnosis of a schizophrenia spectrum disorder at any illness phase were included. We analyzed inflammatory markers and receiver operating characteristic (ROC) curves were used to determine their predictive utility for MetS.

Results

Patients with higher NLR and CRP levels and lower PLR values exhibited a more severe metabolic profile. ROC analysis identified CRP and PLR as the most effective predictors of MetS, with optimal cut-off points of 2.87 mg/L for CRP (sensitivity, 70%; specificity, 54%) and 80.7 for PLR (sensitivity, 71%; specificity, 60%).

Conclusions

Screening CRP levels and inflammatory ratios is a cost-effective and easily implementable practice that may facilitate the early identification of patients with schizophrenia spectrum disorders at risk of developing metabolic complications.

Introduction

Substantial evidence links low-grade chronic inflammation and oxidative stress to the pathophysiological mechanisms of schizophrenia,^1–3 metabolic syndrome (MetS), and cardiovascular (CV) disease.^4,5 It has been hypothesized that inflammatory dysregulation is a common underlying factor in all 3 conditions.⁶

Several peripheral blood inflammatory biomarkers have been investigated in individuals with schizophrenia. C-reactive protein (CRP), an acute-phase protein produced by hepatocytes in response to inflammation, is one of the most studied. Numerous meta-analyses and systematic reviews have consistently reported elevated CRP levels in patients with schizophrenia,^7,8 suggesting that anti-inflammatory strategies may help mitigate the disabling course of the illness.^1,2 High CRP levels have also been associated with an increased risk of MetS and other CV risk factors such as elevated body mass index (BMI), greater waist circumference, and obesity.^6,9,10 There is growing interest in the neutrophil–lymphocyte ratio (NLR), platelet–lymphocyte ratio (PLR), and monocyte–lymphocyte ratio (MLR) as cost-effective inflammatory biomarkers.¹¹ These indices can be easily calculated from standard white blood cell counts and are obtainable under routine laboratory conditions. Studies have demonstrated significant correlations between these ratios, CRP levels, and pro-inflammatory cytokines, supporting their potential utility in detecting systemic inflammation.¹² These ratios have been consistently associated with a higher risk and severity of MetS and CV risk factors in the general population, and with a higher cardiovascular mortality rate.^13–16 Several studies and meta-analyses have confirmed elevated levels of NLR, MLR, and PLR in patients with schizophrenia and first-episode psychosis,^17–22 some findings indicate that inflammatory biomarkers such as CRP or hematological ratios may be particularly relevant during relapse periods.^23–25 Therefore, inflammatory biomarkers may be influenced by two distinct factors: relapse or symptom severity, and the presence of MetS.

As far as we know, the relationship between inflammatory ratios and MetS in schizophrenia has not yet been investigated. Therefore, the aim of this study was to explore the association between the inflammatory biomarkers—CRP, NLR, PLR, and MLR—and metabolic status in a cohort of patients with schizophrenia, whether hospitalized in acute psychiatric units or treated in outpatient settings. This is the first study to employ receiver operating characteristic (ROC) curve analysis to determine optimal cut-off values for these biomarkers in detecting MetS.

Material and methods

Study design and patient sample

The sample for this study was drawn from an observational, cross-sectional, and multicenter study conducted at four centers in Spain (Project No. PI17/00246) from January 2019 to November 2022. The study objectives and protocol have been previously described.²⁶

We considered for inclusion a total of 637 individuals aged between 18 and 65 years, diagnosed with a schizophrenia spectrum disorder (including schizoaffective disorder and First-Episode Psychosis) according to DSM-5 criteria at any phase of illness, and who gave their prior written informed consent. For outpatients, inclusion in the long-term prescription program and regular attendance at clinical appointments (generating corresponding medical records) was also required.

We kept exclusion criteria minimal to ensure a heterogeneous and representative sample. Exclusions included (a) individuals on anti-inflammatory treatment or diagnosed with an acute or chronic inflammatory disease (including fever > 38 ̊C, infection within the past 2 weeks prior to the interview, or vaccination within the past 4 weeks; n = 27); (b) individuals older than 70 years (n = 18); and (c) those residing in long-term care facilities (n = 51). The final sample included a total of 541 patients.

All data were collected in full compliance with applicable legal requirements for confidentiality. Patient identifiers were removed to ensure complete dissociation between identifying information and clinical data, in accordance with Spanish legislation on health data protection. The study protocol was approved by the Clinical Research Ethics Committee of each participant center.

Data extraction

We obtained clinical data—including anthropometric measurements, age, gender, and smoking status—from either medical records or patient reports. Pharmacological data, including the use of antipsychotics, antidepressants, mood stabilizers, statins, glucose-lowering medications, and/or antihypertensive treatments, were retrieved from the regional prescription database (which includes all prescriptions for each subject) or medical records.

Anthropometric variables and biochemical analysis

We assessed participants in fasting conditions between 8:00 and 9:00 a.m. Anthropometric measurements included body weight, body mass index (BMI = weight in kilograms divided by the square of height in meters [kg/m²]), and waist circumference (measured 1 cm above the navel, in cm). Smoking status was recorded as yes/no.

We collected peripheral blood samples after a confirmed 8-h fast. Plasma glucose levels, total cholesterol, high-density lipoprotein cholesterol (HDL), triglycerides, insulin, platelet count, white blood cell counts, and CRP were analyzed at each center using standard laboratory procedures. NLR, MLR, and PLR were calculated from absolute neutrophil, lymphocyte, monocyte, and platelet counts (NLR, neutrophils/lymphocytes; MLR, monocytes/lymphocytes; PLR, platelets/lymphocytes). An inflammatory state was defined by a C-reactive protein (CRP) level > 3 mg/L, in accordance with thresholds previously reported in the literature.^27,28

HOMA, QUICKI, and MetS assessment

The Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) was calculated as: glucose (mmol/L) × insulin (mIU/mL)/22.5.²⁹ Higher values indicate greater insulin resistance. Although a universal HOMA-IR cut-off is not established, values > 2 generally suggest clinically significant insulin resistance.³⁰

The Quantitative Insulin Sensitivity Check Index (QUICKI) was calculated as: 1/[log(insulin [mIU/mL]) + log(glucose [mg/dL])], with higher values reflecting greater insulin sensitivity.²⁹ Suggested cut-offs are 0.382 for non-obese individuals, 0.331 for obese individuals, and 0.304 for diabetic patients.

MetS was diagnosed based on the modified criteria of the National Cholesterol Education Program (NCEP) Adult Treatment Panel (ATP) III. We defined MetS as the presence of at least 3 of the following 5 criteria: (a) serum triglycerides > 1.69 mmol/L or current treatment with fibrates/nicotinic acid; (b) HDL cholesterol < 1.0 mmol/L in men or <1.3 mmol/L in women; (c) systolic/diastolic blood pressure ≥ 130/85 mmHg or current antihypertensive drugs; (d) fasting glucose ≥ 5.55 mmol/L or current glucose-lowering treatment or diagnosis of diabetes mellitus; (e) abdominal obesity (waist circumference > 102 cm in men, >88 cm in women) or BMI > 30 kg/m². We regarded each of these criteria as a risk factor for MetS in the study. Although the diagnosis requires the presence of only 3 out of the 5 criteria, we assessed the total number of criteria present in each patient. Thus, we defined the number of MetS risk factors as the sum of the positive criteria for each patient, ranging from 0 to 5.

Healthy lifestyle habits

We assessed physical activity using the International Physical Activity Questionnaire—Short Form (IPAQ-SF), a self-administered tool for monitoring activity levels.³¹ It evaluates walking, moderate-intensity, and vigorous-intensity activities in terms of frequency (days/week) and duration (minutes/day).

Furthermore, we assessed Nutritional habits using the Mediterranean Diet Adherence Screener (MEDAS), a validated 14-item questionnaire.^32,33 It evaluates adherence to the Mediterranean diet based on the consumption and frequency of specific food items (e.g., olive oil, fruits, vegetables, fish, legumes, nuts, meat, and sweets). Each item was scored 1 (adherence) or 0 (non-adherence), yielding a total MEDAS score.

Statistical analysis

Differences between hospitalized patients and outpatients were determined using chi-square tests for categorical variables and the Student t test for continuous variables. Partial correlations were conducted to examine the relationships between inflammatory biomarkers and MetS risk factors, adjusting for age, sex, Positive and Negative Syndrome Scale (PANSS) total score, clinical setting, years of illness, MEDAS score, IPAQ score, and smoking status. We used the PANSS score as an indicator of symptom severity, as inflammatory biomarkers could be influenced by two separate factors: relapse or symptom severity, and the presence of MetS.

We used logistic regression to evaluate the predictive value of independent variables (NLR, PLR, MLR, CRP, sex, age, smoking status, PANSS total score, MEDAS score, IPAQ score) for the presence of MetS. Multiple linear regression with the same predictors was applied to determine which factors significantly contributed to the number of MetS risk factors. Dummy variables were created for categorical predictors (e.g., sex: male, 0; female, 1). Standardized beta coefficients were used to assess the strength of associations. ROC curve analysis was conducted to identify optimal cut-off points for inflammatory biomarkers in predicting MetS.

Statistical significance was set at a 2-tailed α = 0.05. All analyses and sample size estimations were conducted using SPSS for Windows, version 23.0.

Results

The final sample included a total of 541 patients (323 men, 218 women), 75.6% of whom were outpatients. Most were treated with one antipsychotic (62.3%), and 36.8% received 2 or more. Long-acting injectables (LAIs) were used by 29.6%. Clozapine use was significantly higher among inpatients (33.3%) than outpatients (15.6%) (p < 0.001). No significant differences in age, sex, smoking, illness duration, diagnosis, MEDAS, or IPAQ scores were found between settings (Table 1).

Table 1.

Demographic and clinical variables.

	Outpatients	Inpatients	Total sample	Test
	N = 409	N = 132	N = 541
Age (years)	41.8 ± 12.8	39.6 ± 11.4	41.3 ± 12.5	t = 1.754 NS
Sex (% male)	58.2%	64.4%	59.7%	X^2 = 1.349 NS
Schizophrenia	69.7%	65.2%	68.6%	t = 0.752 NS
Schizoaffective disorder	15.4%	22.0%	17.0%	t = 2.601 NS
First episode psychosis	15.9%	12.9%	15.2%	t = 0.490 NS
Smoking (% yes)	55.1%	52.3%	54.4%	X^2 = 0.212 NS
Years of illness	14.5 ± 11.4	14.9 ± 10.6	14.6 ± 11.1	t = −0.288 NS
PANSS total score	59.6 ± 17.1	68.2 ± 22.9	62.6 ± 19.8	t = −3.745***
PANSS-positive	11.9 ± 5.1	13.8 ± 6.2	12.5 ± 5.5	t = −3.167**
PANSS-negative	18.6 ± 7.3	21.6 ± 9.3	19.6 ± 8.1	t = −3.169**
PANSS-general	28.4 ± 8.5	32.8 ± 11.1	29.9 ± 9.6	t = −3.882***
CGI score	4.1 ± 2.8	4.2 ± 1.0	4.1 ± 2.4	t = −0.589 NS
GAF score	59.0 ± 13.8	56.8 ± 16.9	58.3 ± 14.9	t = 1.235 NS
Mediterranean diet (MEDAS) score	7.7 ± 2.3	7.6 ± 2.1	7.7 ± 2.2	t = 0.461 NS
IPAQ (exercise)	1879.6 ± 2121.2	1935.2 ± 2692.1	1895.3 ± 2291.9	t = −0.195 NS

PANSS, Positive and Negative Syndrome Scale; CGI, Clinical Global Impression; GAF, Global Assessment of Functioning; MEDAS, Mediterranean Diet Adherence Screener; IPAQ, International Physical Activity Questionnaire.

Quantitative variables values are expressed as mean ± SD.

*p < 0.05.

**

p < 0.01.

***

p < 0.001.

NS, non-significant.

Outpatients had a higher prevalence of MetS (31.8% vs 17.4%; p = 0.001) and more MetS risk factors (1.41 vs 1.21; p = 0.003). Statin use was also higher among outpatients (12.2% vs 6.1%; p = 0.031). No significant differences were found in insulin sensitivity/resistance, CRP levels, or NLR/MLR between groups. However, PLR was significantly higher in inpatients (p = 0.007) (Table 2).

Table 2.

Metabolic variables and inflammatory biomarkers.

	Outpatients	Inpatients	Total sample	Test
	N = 409	N = 132	N = 541
Metabolic syndrome variables
Metabolic syndrome (% yes)	31.8%	17.4%	28.3%	X^2 = 9.451**
No. of risk factors for MetS	1.8 ± 1.4	1.4 ± 1.2	1.6 ± 1.3	t = 2.965**
BMI (kg/m^2)	29.2 ± 5.7	26.4 ± 5.3	28.5 ± 5.7	t = 4.634***
Waist circumference (cm)	100.3 ± 16.3	96.5 ± 14.3	99.4 ± 15.9	t = 2.140*
Glucose (mmol/L)	5.2 ± 1.5	4.9 ± 1.0	5.2 ± 1.4	t = 2.450*
Hyperglycemia (% yes)	26.4%	17.4%	24.2%	X^2 = 3.911*
HDL (mg/dL)	1.2 ± 0.4	1.2 ± 0.4	1.2 ± 0.4	t = 0.507 NS
Low HDL (% yes)	42.4%	37.9%	41.3%	X^2 = 0.661
Triglycerides (mmol/L)	1.7 ± 1.0	1.3 ± 0.8	1.6 ± 1.0	t = 4.058***
Hypertriglyceridemia (% yes)	34.5%	21.2%	31.2%	X^2 = 7.566**
Systolic blood pressure (mmHg)	118.4 ± 17.2	114.1 ± 12.2	117.0 ± 15.9	t = 2.799**
Diastolic blood pressure (mmHg)	77.6 ± 9.7	75.7 ± 9.5	77.0 ± 9.6	t = 1.851 NS
Hypertension (% yes)	17.4%	18.9%	17.7%	X^2 = 0.375 NS
Insulin resistance/sensitivity variables
Insulin (mIU/L)	15.4 ± 16.9	14.1 ± 11.7	15.0 ± 15.7	t = 0.800 NS
HOMA-IR	3.83 ± 5.19	3.34 ± 3.33	3.7 ± 4.8	t = 0.993 NS
QUICKI	0.342 ± 0.056	0.341 ± 0.042	0.342 ± 0.053	t = 0.295 NS
Inflammatory biomarkers
CRP (mg/L)	5.0 ± 7.7	5.7 ± 6.6	5.2 ± 7.4	t = -0.900 NS
NLR	2.2 ± 1.1	2.5 ± 1.7	2.3 ± 1.3	t = -1.775 NS
PLR	116.4 ± 51.5	147.2 ± 125.0	124.2 ± 77.9	t = -2.756**
MLR	0.24 ± 0.12	0.26 ± 0.17	0.25 ± 0.14	t = -1.410 NS

MetS, metabolic syndrome; BMI, body mass index; HDL, high-density lipoprotein; HOMA-IR, Homeostatic Model Assessment for Insulin Resistance; QUICKI, Quantitative Insulin Sensitivity Check Index; CRP, C-reactive protein; NLR, neutrophil–lymphocyte ratio; PLR, platelet–lymphocyte ratio; MLR, monocyte–lymphocyte ratio.

Quantitative variables values are expressed as mean ± SD.

*

p < 0.05.

**

p < 0.01.

***

p < 0.001.

NS, non-significant.

We conducted partial correlations to initially assess the associations between inflammatory biomarkers and MetS risk factors, controlling for the following variables: age, sex, setting, duration of illness, total PANSS score, physical activity (IPAQ scale score), adherence to the Mediterranean diet (MEDAS score), and tobacco use. Regarding CRP levels, positive associations were found with BMI (r = 0.312; p < 0.001), abdominal circumference (r = 0.261; p < 0.001), number of MetS risk factors (r = 0.162; p = 0.008), and presence of MetS (r = 0.122; p = 0.045), as well as a negative association with the QUICKI index (r = −0.139; p = 0.022). We observed a positive association between NLR and diastolic blood pressure (r = 0.140; p = 0.022). PLR showed negative trends with the presence of MetS (r = −0.110; p = 0.071) and triglyceride levels (r = −0.110; p = 0.072), and a positive trend with HDL levels (r = 0.105; p = 0.084). Finally, we found a positive association between MLR and diastolic blood pressure (r = 0.123; p = 0.045).

To identify a model predicting the number of MetS risk factors, we conducted a multiple linear regression (backward method). The overall model was significant (F = 12.786; p < 0.001), explaining 22.1% of the variance. The final model included NLR (β = 0.148; p = 0.058), PLR (β = −0.166; p = 0.032), CRP (β = 0.151; p = 0.006), age (β = 0.328; p < 0.001), total PANSS score (β = 0.140; p = 0.013), and outpatient status (β = −0.114; p = 0.043). Therefore, higher NLR and CRP levels, lower PLR, older age, greater clinical severity (higher PANSS-total score), and outpatient status significantly predicted a higher number of MetS risk factors (Table 3).

Table 3.

Multivariate predictors of MetS risk factors and presence of MetS.

	Standardized coefficient β	p value
Multivariate predictors of MetS risk factors
Age	β = 0.328	p < 0.001*
PLR	β = −0.166	p = 0.032*
CRP	β = 0.151	p = 0.006*
NLR	β = 0.148	p = 0.058
PANSS total score	β = 0.140	p = 0.013*
Setting (outpatient)	β = −0.114	p = 0.043*
Overall model, F = 12.786, p < 0.001

Multivariate predictors of MetS
Age	OR, 1.046	p = 0.002*
PANSS total score	OR, 1.027	p = 0.007*
PLR	OR, 0.994	p = 0.101
Setting (outpatient)	OR, 0.317	p = 0.011*
Overall model, χ^2 (4, N = 223) = 31.979; p < 0.001

MetS, metabolic syndrome; PLR, platelet–lymphocyte ratio; CRP, C-reactive protein; NLR, neutrophil–lymphocyte ratio; PANSS, Positive and Negative Syndrome Scale.

The order of predictors reflects a descendent level of contribution to the overall model.

Furthermore, we performed a backward stepwise logistic regression to evaluate the impact of biomarkers on the likelihood of presenting MetS. The final model was statistically significant (Chi-square test [4; N = 223]) = 31 979; p < 0.001), explaining between 13.4% (Cox & Snell R²) and 21.1% (Nagelkerke R²) of the variance. The model correctly categorized 82.5% of cases. In the final step, PLR (p = 0.101), age (p = 0.002), total PANSS score (p = 0.007), and clinical setting (p = 0.011) made statistically significant contributions. Table 3 illustrates the odds ratios (ORs) for each variable. Accordingly, the probability of presenting MetS was greater in individuals with lower PLR, older age, higher clinical severity (higher PANSS-total score), and outpatient status. However, the effect sizes were small.

ROC curve analysis showed that C-reactive protein (CRP; AUC, p < .001) and the platelet-to-lymphocyte ratio (PLR; AUC, p = .013) were the strongest predictors. The optimal cutoff value for CRP was 2.87 mg/L (sensitivity, 70%; specificity, 54%), and for PLR was 80.7 (sensitivity, 71%; specificity, 60%) (Table 4 and Fig. 1).

Table 4.

ROC analyses of inflammatory biomarkers to identify metabolic syndrome.

	AUC	p value	95%CI
NLR	.507	.824	0.443–0.571
MLR	.466	.284	0.400–0.532
PLR	.422	.014	0.361–0.480
CRP	.638	.000	0.578–0.697

AUC, area under the curve; NLR, neutrophil–lymphocyte ratio; MLR, monocyte–lymphocyte ratio; PLR, platelet–lymphocyte ratio; CRP, C-reactive protein; CI, confidence interval.

Fig. 1.

ROC curve of NLR, CRP, PLR and MLR for MetS prediction. ROC, receiver operating characteristic; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; MLR, monocyte-to-lymphocyte ratio; CRP, C-reactive protein; MetS, metabolic syndrome.

Discussion

In this study, we explored the relationship between 4 inflammatory markers—CRP, NLR, PLR, and MLR—and MetS in a large and clinically representative sample of patients diagnosed with schizophrenia spectrum disorders. This is the first study to perform a ROC curve analysis to identify optimal cut-off values of the inflammatory biomarkers to detect MetS. Our findings confirm and extend previous evidence that a chronic low-grade inflammatory state is present in schizophrenia and is closely linked to a more adverse metabolic profile.

Consistent with previous reports,^7,9,34 we observed that CRP levels were significantly associated with BMI, abdominal circumference, insulin resistance (as shown by lower QUICKI scores), the number of MetS risk factors, and the presence of MetS. These associations reinforce the role of CRP as not only a marker of inflammation but also a potential predictor of cardiovascular risk in schizophrenia—a population already at markedly increased risk for premature mortality due to CV disease.^35,36

Moreover, the NLR was associated with diastolic blood pressure, a finding consistent with previous literature linking NLR to disease severity, metabolic burden, and behavioral manifestations such as aggression.^18,37 Although NLR did not reach statistical significance as a predictor of metabolic syndrome in the logistic regression analysis, it remained in the final model predicting the number of metabolic risk factors, suggesting potential utility as a composite marker of systemic stress and inflammation.

Of note, the PLR showed an inverse relationship with MetS and related metabolic components (triglycerides, HDL), indicating that lower PLR values may be indicative of a pro-metabolic syndrome profile. This finding contrasts with some studies suggesting elevated PLR in MetS,^38,39 yet aligns with the hypothesis of a dynamic staging role proposed by Mertoglu and Gunay,⁴⁰ in which PLR may behave differently in early vs late metabolic deterioration.

In addition, MLR was also positively associated with diastolic blood pressure. While fewer studies have explored MLR in psychiatric populations, our findings support former reports noting increased MLR in schizophrenia^19,21 and suggest it may be another accessible biomarker of vascular and metabolic stress.

A novel contribution of our study is the use of ROC curve analysis to define threshold values for these biomarkers. We identified CRP and PLR as the best predictors of MetS. A C-reactive protein cutoff of 2.87 mg/L and a platelet-to-lymphocyte ratio cutoff of 80.7 were associated with sensitivities of 70% and 71%, respectively. Although these sensitivity values indicate a reasonable ability to identify true-positive cases, the relatively modest specificities (54% and 60%) suggest that these biomarkers may be more suitable for screening rather than diagnostic purposes, particularly given their simplicity, cost-effectiveness, and wide availability in routine clinical practice.

From a clinical perspective, these results have important implications. Early identification of metabolic risk in schizophrenia is critical, particularly given the known delays in diagnosis and treatment of physical comorbidities in psychiatric populations. The inclusion of inflammatory markers in routine blood work may serve as a practical and scalable approach to stratify patients based on metabolic vulnerability.

Notably, our results highlight a surprising trend: outpatients had a significantly worse metabolic profile vs hospitalized individuals. A possible explanation for this finding is that the MEDAS and IPAQ scales, which evaluate the patients’ habitual behaviors, were completed based on their usual routines at home and may therefore fail to capture the behavioral changes that occur during hospitalizations. The MEDAS and IPAQ scores did not differ significantly across groups, which supports this hypothesis; consequently, these instruments may not adequately capture short-term lifestyle changes occurring during inpatient stays. Although hospitalization is often associated with weight gain due to pharmacologic treatment and reduced physical activity, our hypothesis suggests that short-term changes in routine, diet, and daily structure may confer a protective metabolic effect. Hospitalized patients may benefit from more balanced nutrition, regular schedules, and closer clinical supervision, which could indirectly mitigate metabolic risk. This hypothesis is supported by previous findings.^41,42 This hypothesis emerged after data analysis and could not be confirmed within the limitations of this study; however, it would be valuable to incorporate it in future research. This underscores a broader point: lifestyle interventions, even if short-term and low intensity, can significantly impact metabolic health, especially in vulnerable psychiatric populations. Our findings support previous meta-analyses in the general population showing the effectiveness of lifestyle changes in reducing MetS.^43,44 In schizophrenia, structured interventions promoting physical activity and Mediterranean-style dietary habits should be prioritized in outpatient care.

This study has several limitations that should be acknowledged. First, its cross-sectional design precludes conclusions about causality between variables. Although we excluded participants with overt inflammatory conditions or on anti-inflammatory drugs, we did not control for other medical comorbidities that may influence inflammatory or metabolic parameters. A major limitation of our study is the incomplete information available regarding antipsychotic treatment. Although we collected data on the use of one or more antipsychotic drugs, LAIs, and the use of clozapine, the complexity of psychiatric pharmacotherapy was not fully documented or accounted for in our analyses—particularly the extent of polypharmacy, detailed data on drug dosages, specific drug combinations, and the duration of exposure to each drug or combination. While existing evidence suggests that antipsychotics would not significantly affect markers such as CRP and NLR,^45,46 it is known that antipsychotic drugs can influence the prevalence of MetS. The specific relevance of the impact of medication on inflammation and MetS prevalence and severity remains an important area for future research, and incorporating such data would provide valuable insights. Additionally, a more detailed assessment of dietary changes and physical activity across different care settings would improve understanding of potential strategies for enhancing metabolic health. Future research should also incorporate longitudinal designs and more comprehensive clinical, pharmacological, and lifestyle data to clarify the complex interactions between inflammation and metabolic status in individuals with psychotic disorders.

Nonetheless, the strengths of the study are substantial. We examined a large, naturalistic sample of schizophrenia patients with minimal exclusion criteria, enhancing generalizability. Moreover, we analyzed a comprehensive set of variables—including clinical, metabolic, and lifestyle data—while introducing ROC-based biomarker cut-offs as a practical tool for future clinical applications.

In conclusion, our findings supportsthe presence of inflammation and high rates of MetS in patients with schizophrenia, particularly in outpatient settings. Elevated CRP and NLR and reduced PLR are associated with more severe metabolic profiles, suggesting a potential link between systemic inflammation and metabolic alterations in this population. CRP and PLR may serve as valuable, cost-effective biomarkers for early identification of individuals at increased risk for MetS. Integrating the routine screening of these markers into psychiatric care could enable earlier interventions to reduce cardiometabolic burden and facilitate preventive strategies. Finally, our results underscore the relevance of lifestyle-based strategies—including structured diet and exercise—as complementary approaches to pharmacological treatment. Although further research should confirm the extent of their impact, our study suggests that such interventions may imply benefits even in the short term. Future research should aim to clarify our understanding of the complex interplay between physical and mental health in schizophrenia and psychotic disorders and allow us to provide more holistic models of care.

Authors’ contributions

BA and MSA designed the project. BA, GS, MB, and MPGP coordinated the project development. BA and MSA drafted the manuscript. GA, SA, MA, CH, GS, MA, MSA, LGB, MPGP, and BA participated in the recruitment. BA and MSA performed the statistical analyses. All authors reviewed and approved the final version of the manuscript.

Funding

This study was funded by Fondo de Investigaciones Sanitarias, grant No. PI17/00246.

Conflicts of interest

Dr. Anmella reports having received continuing medical education-related honoraria or consulting fees from Janssen-Cilag, Lundbeck, Lundbeck/Otsuka, and Angelini, with no financial or other relationships relevant to the subject of this article. Dr. Amoretti reports having served as a consultant for and/or having received honoraria or grants from Otsuka–Lundbeck, with no financial or other relationships relevant to the subject of this article. Dr. García-Portilla reports having served as a consultant for and/or having received honoraria or grants from Adamed, Angelini, Casen Recordati, Alianza Otsuka–Lundbeck, Janssen-Cilag, Lundbeck, Otsuka, and SAGE Therapeutics. Dr. Bernardo reports having served as a consultant for, received research grants or support from, and received honoraria from, as well as having participated on speakers’ or advisory boards for, ABBiotics, Adamed, Angelini, Casen Recordati, Janssen-Cilag, Lundbeck, Otsuka, Menarini, and Takeda; these activities were unrelated to the present work. Dr. Arranz reports having served as a consultant for and/or having received honoraria or grants from Adamed, Janssen-Cilag, Lundbeck, Otsuka, and Takeda, with no conflicts of interest related to this manuscript. All other authors report no financial or other relationships relevant to the subject of this article.

Acknowledgements

The authors thank the patients and nurses that participated in the study for their valuable contribution.