Research

JAMA Psychiatry | Original Investigation

Differences in Blood Leukocyte Subpopulations in Schizophrenia

A Systematic Review and Meta-Analysis
Leon Dudeck, MS; Madeleine Nussbaumer; Thomas Nickl-Jockschat, MD; Paul C. Guest, PhD;
Henrik Dobrowolny, MScEng; Gabriela Meyer-Lotz; Zhongming Zhao, PhD, MS; Roland Jacobs, PhD;
Kolja Schiltz, MD; Brisa S. Fernandes, PhD, MD, MS; Johann Steiner, MD

Supplemental content
IMPORTANCE This study aims to provide robust evidence to support or challenge the immune
hypothesis of schizophrenia.

OBJECTIVE To conduct a meta-analysis of reports on blood leukocyte subpopulations in

schizophrenia vs healthy controls, examining disease- and treatment-related differences as
well as potential confounders.

DATA SOURCES Systematic database search for English and non-English peer-reviewed
articles in PubMed, Web of Science, Scopus, and Cochrane Library databases, with the last
search in January 2024.

STUDY SELECTION Cross-sectional, case-control, and longitudinal studies comparing leukocyte

numbers in patients with schizophrenia and healthy controls. After duplicates were removed,
3691 studies were identified for screening.

DATA EXTRACTION AND SYNTHESIS Data extraction and quality assessment were conducted
following PRISMA and MOOSE guidelines. Data were independently extracted by 2 authors
and pooled using random-effects models.

MAIN OUTCOMES AND MEASURES The planned primary outcomes were differences in
leukocyte subpopulation counts between individuals with schizophrenia and healthy controls
to increase our understanding of the immune system dysfunction in schizophrenia.

RESULTS Sixty-four relevant articles were identified (60 cross-sectional/case-control studies

and 4 longitudinal studies) with data on leukocyte numbers from 26 349 individuals with
schizophrenia and 16 379 healthy controls. Neutrophils (g = 0.69; 95% CI, 0.49 to 0.89;
Bonferroni-adjusted P < .001; n = 40 951 [47 between-group comparisons]) and monocytes
(g = 0.49; 95% CI, 0.24 to 0.75; Bonferroni-adjusted P < .001; n = 40 513 [44 between-group
comparisons]) were higher in schizophrenia compared with control participants. Differences
were greater in first-episode vs chronic schizophrenia and in patients who were not treated
vs treated with antipsychotic medication. There were no significant differences in eosinophils
(g = 0.02; 95% CI, −0.16 to 0.20; Bonferroni-adjusted P > .99; n = 3277 [18 between-group
comparisons]), basophils (g = 0.14; 95% CI, −0.06 to 0.34; Bonferroni-adjusted P = .85;
n = 2614 [13 between-group comparisons]), or lymphocytes (g = −0.08; 95% CI, −0.21 to
0.06; Bonferroni-adjusted P > .99; n = 41 693 [59 between-group comparisons]). Neutrophils
decreased longitudinally (g = −0.30; 95% CI, −0.45 to −0.15; Bonferroni-adjusted P < .001;
n = 896 [4 within-group comparisons]) and eosinophils increased longitudinally (g = 0.61;
95% CI, 0.52 to 0.71; Bonferroni-adjusted P < .001; n = 876 [3 within-group comparisons])
after successful treatment of acute psychosis.

CONCLUSIONS AND RELEVANCE Our findings of increased blood neutrophils and monocytes
support the immune hypothesis of schizophrenia, particularly highlighting the role of innate
immune activation. As these effects were more pronounced in early disease stages and also
reflected clinical improvement, they may pave the way for innovative treatment strategies
Author Affiliations: Author
based on immunological and inflammatory pathways and help revolutionize the treatment affiliations are listed at the end of this
landscape for schizophrenia. article.
Corresponding Author: Johann
Steiner, MD, Department of
Psychiatry and Psychotherapy,
University of Magdeburg,
Leipziger Str. 44, D-39120
JAMA Psychiatry. doi:10.1001/jamapsychiatry.2024.4941 Magdeburg, Germany
Published online March 5, 2025. ([email protected]).

(Reprinted) E1
© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Research Original Investigation Differences in Blood Leukocyte Subpopulations in Schizophrenia

L
arge epidemiological studies have shown that severe
infections are significant risk factors for schizophrenia Key Points
and affective disorders,1,2 with bacterial infections pos-
Question Are there differences in blood leukocyte
ing higher risk than viral and other types.3 Meta-analyses subpopulations in schizophrenia?
consistently reported elevated proinflammatory cytokines in
Findings In this meta-analysis of data from 26 349 participants
blood and cerebrospinal fluid of patients with schizophrenia,4,5
with schizophrenia and 16 379 healthy controls, neutrophils and
suggesting infections could trigger or exacerbate mental ill-
monocytes were elevated in schizophrenia.
ness through cytokine action in genetically predisposed indi-
viduals or those with brain developmental risk.6 Meaning The findings of increased blood neutrophils and
monocytes support the immune hypothesis of schizophrenia
Recently, we observed elevated neutrophil counts in
with a focus on innate immune system activation.
acutely ill unmedicated individuals with schizophrenia and
major depression compared with controls.7,8 Additionally,
increased monocyte counts were found in schizophrenia. Building on emerging evidence, we hypothesized that
A decline in neutrophils during psychopharmacotherapy specific WBC populations:
correlated with improvement in Positive and Negative • Are increased in schizophrenia compared with healthy
Syndrome Scale (PANSS) and Hamilton Depression 21 controls.8,11-14
scores. Elevated neutrophil counts in acute schizophrenia • Show greater inc reases in first-episode vs chronic
and major depression imply that bacteria may act as disease schizophrenia.15
triggers, as neutrophils are key responders to bacterial • Show greater increases in untreated vs treated patients.8,16,17
infections.9 • Are linked to disease severity.8,18,19
Meta-analyses of differential white blood cell (WBC) counts • Normalize after treatment with antipsychotic medication.8,16-19
in psychosis are limited. We found only 4 systematic reviews
that were published before 2024 and included 8 to 24
studies.10-13 Two of these assessed neutrophil-lymphocyte or
monocyte-lymphocyte ratios,11,13 but reports on relative mea-
Methods
sures may mask absolute cell count changes. Another of these This meta-analysis compared absolute blood leukocyte counts
studies focused on monocyte counts only.12 A more recent between individuals with schizophrenia and healthy con-
meta-analysis analyzed studies on absolute routine blood cell trols, adhering to Preferred Reporting Items for Systematic
counts with a maximum of 33 analyzed studies per cell type and, Reviews and Meta-Analyses (PRISMA) and Meta-Analysis of
regarding lymphocyte counts, mainly focused on studies using Observational Studies in Epidemiology (MOOSE) reporting
immunophenotyping.14 Reports indicate that during initial dis- guidelines (eAppendix in Supplement 1) and the Cochrane
ease onset (first episode), immune activation may play a more Handbook. Figure 1 outlines the study selection process,
prominent role, aligning with findings that inflammatory blood covering identification, screening, eligibility, and inclusion.
cytokine levels are even more increased during the first epi- Two authors (L.D. and M.N.) independently assessed study
sode of psychosis.15 Antipsychotic drugs appear to modulate eligibility and performed data extraction in duplicate, with
immune cell counts and reduce inflammation,8,16,17 poten- discrepancies resolved by a third author (J.S.).
tially reflected by specific immune cell elevations in patients
who are antipsychotic naive or antipsychotic free. Elevated Search Strategy
immune activity, indicated by higher leukocyte counts, may We systematically searched PubMed, Web of Science, Scopus,
also be linked to more severe psychotic symptoms, 8,18,19 and Cochrane library databases for peer-reviewed articles. The
supporting the hypothesis that immune cell counts correlate Boolean search terms were (schizophreni* OR psychosis) AND
with symptom severity and normalize after antipsychotic (leukocyt* OR granulocyt* OR neutrophil* OR eosinophil* OR
treatment. basophil* OR monocyt* OR lymphocyt*). No restrictions were
We conducted this meta-analysis to determine if our pre- placed on year or country of publication with the last search
viously reported findings8 align with other studies. We aimed performed in January 2024. Articles in languages other than
to surpass previous research by analyzing absolute WBC num- English or German were translated using the DeepL Transla-
bers, systematically contacting authors for missing raw data, tor. Also, we examined reference sections of retrieved publi-
and exploring disease- and treatment-related changes and the cations and other meta-analyses to identify additional ar-
influence of potential confounding factors like age, sex, body ticles. After eliminating reviews and meta-analyses, we
mass index (BMI), and smoking. We compared acutely ill pa- narrowed the results according to the following inclusion and
tient subgroups with and without antipsychotic treatment, exclusion criteria:
patients with first-episode vs chronic schizophrenia, and Studies were included if they (1) included individuals 18
healthy controls to assess if blood count changes occur only years or older with schizophrenia or schizoaffective disorder
in specific disease stages or independently of treatment. Ad- based on DSM-III through DSM-5, International Statistical Clas-
ditionally, we included longitudinal studies to observe changes sification of Diseases and Related Health Problems, Tenth Re-
in differential blood counts during acute psychosis treat- vision, or Experiential World Inventory or Research Diagnos-
ment. Our main goal was to provide robust evidence for or tic Criteria (late 1970s standardized criteria for mental disorders
challenge the immune hypothesis of schizophrenia. influencing DSM-III); (2) involved pairwise comparisons with

E2 JAMA Psychiatry Published online March 5, 2025 (Reprinted) jamapsychiatry.com

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Differences in Blood Leukocyte Subpopulations in Schizophrenia Original Investigation Research

Figure 1. Flow Diagram According to Preferred Reporting Items for Systematic Reviews and Meta-Analyses
(PRISMA) Criteria Showing the Process of Study Selection

5372 Records identified through database

searching
2420 PubMed
1916 Web of Science
877 Scopus
159 Cochrane Library

1681 Records excluded

1647 Duplicates
34 Trial registration data (ICTRP, CT.gov)

3691 Records screened

548 Records excluded for depicting reviews or meta-analyses

3143 Records with abstract, title, and

methods screened

3083 Records excluded

3042 Excluded after screening abstract, title, and methods for the
adjoining exclusion criteria
1234 Did not investigate SCZ
1109 Major WBC subpopulation count not determined
253 Case reports
143 Without original data (eg, meeting abstracts, editorials, or book sections)
76 Cross-sectional comparison without a control group (eg, comparison of
multiple psychiatric disorders)
73 Animal studies
47 Postmortem studies
44 Pediatric and adolescent psychiatry Sixty-four articles19,28-89 were
17 Disease controls enrolled instead of healthy controls included in our meta-analyses,
14 Major WBC subpopulation count not assessed for all enrolled groups comprising 60 cross-sectional/case-
10 WBC counts distorted by comorbidities or non-antipsychotic medication
control studies and 4 longitudinal
9 Focused on WBC count changes in treatment-resistant or clozapine-treated
SCZ over several months studies (eTables 2 and 3 in
7 Follow-up >12 wk and no focus on acuteness Supplement 1). Two longitudinal
4 Focused on changes in WBC count after drug switch or combination studies34,47 were excluded because
2 WBC counts assessed by immunophenotyping or flow cytometry of follow-up periods longer than
(CD-markers) or in vitro techniques 12 weeks and lack of focus on acute
41 Authors could not provide necessary missing data
treatment of schizophrenia (SCZ).
These studies were still included in
4 Studies identified by checking reference sections the between-group analyses because
of retrieved publications they included control groups.
CT.gov indicates ClinicalTrials.gov;
ICTRP, International Clinical Trials
64 Studies included in meta-analysis
Registry Platform; WBC, white
blood cell.

a healthy control group for between-group or pairwise com- infections, immunological/autoimmune diseases) or nonan-
parisons before and after antipsychotic treatment for longitu- tipsychotic (eg, immunosuppressant) therapies; (15) longitu-
dinal within-group analyses; and (3) assessed leukocyte sub- dinal studies using switching or combination therapies,
population counts in human blood samples in vivo. (16) longitudinal studies on clozapine adverse effects in treat-
Exclusion criteria were: (1) duplicate articles; (2) reviews/ ment-resistant schizophrenia (to avoid confounding by cloza-
meta-analyses; (3) animal studies; (4) postmortem studies; pine’s known adverse effect of neutropenia); and (17) longitu-
(5) studies on pediatric/adolescent psychiatric populations dinal studies with more than 12 weeks of follow-up not focused
(to avoid confounding specific to early-onset schizophrenia); on acute illness treatment (to avoid confounding by increased
(6) schizophrenia not investigated; (7) articles lacking origi- variability in adherence and clinical course) (Figure 1).
nal data; (8) case reports; (9) cross-sectional studies without
healthy controls (eg, comparisons between neuropsychiatric Data Extraction
disorders); (10) case-control studies with disease controls (ie, Extracted data included diagnostic status, possible confound-
participants with psychiatric or somatic illnesses); (11) stud- ing factors (sex, age, disease duration, age at onset, smoking),
ies not reporting cell counts; (12) studies assessing leukocyte leukocyte counts, medication status (medicated, antipsy-
counts using only in vitro methods; (13) studies not reporting chotic medication–naive or -free), and symptom/disease sever-
leukocyte subpopulation counts for all groups; (14) studies with ity scores (PANSS, Scale for the Assessment of Positive Symp-
leukocyte subpopulation counts affected by comorbidities (eg, toms [SAPS], Scale for the Assessment of Negative Symptoms

jamapsychiatry.com (Reprinted) JAMA Psychiatry Published online March 5, 2025 E3

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Research Original Investigation Differences in Blood Leukocyte Subpopulations in Schizophrenia

[SANS], Brief Psychiatric Rating Scale [BPRS] and Clinical Global phocytes). Statistical significance was defined as Bonferroni-
Impression [CGI]). Baseline symptom severity scores were ana- adjusted P values less than .05.
lyzed for longitudinal studies. Subgroup analyses considered an- We assessed study heterogeneity using Cochrane Q, which
tipsychotic status (taking antipsychotic medication, antipsy- measures deviation of single effect sizes from overall effect size
chotic-free for at least 2 weeks, antipsychotic-naive at blood (P < .10).23 We also applied the I2 metric with 50% to 75% in-
sampling), disease state (first-episode vs chronic schizophre- dicating moderate and greater than 75% high heterogeneity.24
nia), study setting (clinical vs population based), and second- Random-effects models accounted for variability.25
ary clinical conditions. We considered studies clinical if they Meta-regression analyses used the unrestricted maxi-
recruited participants in controlled settings (eg, hospitals), and mum likelihood random-effects model25 to evaluate how age,
population based if they used broader recruitment (eg, cities, BMI, percentage of males, smoker percentage, follow-up length,
multiple hospitals) mostly via advertising. Clinical conditions disease severity (PANSS/SAPS/SANS/BPRS/CGI scores), age at
were defined as comorbidities, including chronic diseases (eg, onset, disease duration, sample size, and NOS scores influ-
diabetes, hypertension), acute conditions (eg, infections, inju- enced effect sizes. Studies were weighted by precision (95%
ries), mental disorders, or other medical states requiring inter- CI and sample size). These analyses were considered explor-
vention. When data were unavailable, we contacted authors. atory; no Bonferroni correction was applied.
For articles about the same cohort, we selected the one with the Median and quartiles were converted to mean and stan-
largest sample to maximize statistical power. dard deviation, assuming normal distribution. In studies with
missing PANSS scores but reporting BPRS values, we applied
Quality Assessment the linear regression approach by Leucht et al26 to estimate
We assessed study quality using the Newcastle-Ottawa Scale PANSS total scores. Similarly, SANS and SAPS values were con-
(NOS)20 with separate scales for case-control and cohort stud- verted to PANSS negative and PANSS positive scores using the
ies in between-group and within-group meta-analyses. The regression model by van Erp et al.27
NOS evaluates selection, comparability, exposure (for case- The meta-analysis comprised (1) overall analysis as a ba-
control studies), and outcome (for cohort studies). Items in- sis for subsequent stages, (2) sensitivity evaluation via leave-
dicating high quality were marked with stars. Each category 1-out analysis (a sensitivity method excluding 1 study at a time
has a maximum star count (selection = 4, comparability = 2, to assess result robustness and identify studies disproportion-
exposure = 3, and outcome = 3), yielding overall scores of 0 ately affecting effect size or heterogeneity), and (3) meta-
to 9 stars.21 For case-control studies, we awarded 1 star for regression to identify potential effect size moderators.
“nonresponse rate” and “representativeness of the cases” even
if not explicitly stated. Quality scores for included studies
ranged from 4 to 9 stars (eTable 1 in Supplement 1). An NOS
score of 4 or higher (at least moderate quality) was consid-
Results
ered sufficient for inclusion in this meta-analysis. Database Search Outcomes
We included 64 publications in the meta-analysis: 60 cross-
Publication Bias sectional/case-control studies and 4 longitudinal studies. The
Publication bias originates from favored publication of posi- study selection process is shown in Figure 1.
tive results. This was assessed using funnel plots that plot
single-study effect size against a parameter representing study Study Characteristics
size, as well as Egger and Orwin fail-safe N tests. The latter es- The 64 studies provided 62 pairwise comparisons of schizo-
timates how many negative studies would be needed to make phrenia vs controls for between-group meta-analysis19,28-86 and
results nonsignificant (P > .05).22 2 longitudinal studies without controls87,88 (eTables 2 and 3 in
Supplement 1). These studies, published from 1972 to 2024,
Statistical Analysis included 26 349 individuals with schizophrenia and 16 379
Data analyses were performed using R (version 4.3.1; metafor healthy controls, with a mean age of 23 to 51 years. Forty-three
package version 4.6.0; R Foundation) and SPSS Statistics (ver- publications28-30, 32, 33, 35, 38, 40, 42-44, 47-49, 51, 52, 54-56, 58, 59, 61, 62,
65-68, 70-73, 75, 76, 78-81, 83-88
sion 28; IBM). We measured effect sizes using Hedges ad- focused on chronic schizophrenia,
justed g with 95% CIs to assess differences in leukocyte counts. 18 on first-episode schizophrenia,19, 34, 36, 37, 39, 41, 45, 46, 50, 53, 57,
60, 63, 64, 69, 74, 77, 82
Hedges g is preferred for low bias and adjustability for small and 3 publications8,31,89 on both groups.
8,19,34,45-47,49,53,55,56,64,66,88
samples, with effect sizes classed as small (0.2), moderate (0.5), Thirteen articles reported all major
or large (0.8).20 Hedges g was calculated if 3 or more studies absolute leukocyte subpopulation counts, and 39 matched
were available. For between-group analyses, positive effect patients and controls for age and sex.8,19,28-30,33,34,38,41-43,45-47,
49, 51-57, 59, 64, 66, 69, 71-76, 78, 80, 81, 83-86
sizes indicated higher leukocyte counts. For within-group lon- BMI data were available in
gitudinal comparisons, negative effect sizes indicated lower 23 articles,8, 19, 30, 31, 42, 45, 47, 49, 51, 53-57, 59, 60, 64, 73, 74, 76, 78, 82, 84
counts after medication. To control for familywise type I er- 12 included patients without antipsychotic treatment,28, 31, 33,
36, 37, 45, 50, 57, 62, 63, 74, 75, 77, 82, 86
ror in the primary outcome parameters, we applied Bonfer- and 42 exclusively examined
roni adjustments to the P values by multiplying raw P values patients taking medication.19,29,30,32,34,35,38-44,46-49,51,52,54-56,
59-61, 64-73, 76, 78-81, 83-85
by 5, corresponding to the 5 analyzed leukocyte subpopula- Three studies53,58,86 analyzed mixed
tions (neutrophils, eosinophils, basophils, monocytes, and lym- cohorts of antipsychotic-naive/-free and medicated patients.

E4 JAMA Psychiatry Published online March 5, 2025 (Reprinted) jamapsychiatry.com

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Differences in Blood Leukocyte Subpopulations in Schizophrenia Original Investigation Research

Four longitudinal studies examined patients who were antipsy- cyte count was related to BMI in patients who were antipsy-
chotic-naive or -free at baseline and after treatment. Only 20 chotic-naive or -free (Table 2).
studies reported smoking data8, 34, 41, 42, 45, 46, 51, 54, 55, 57, 60, 63, 64,
69, 72-74, 76, 78, 84
and 31 used automated cell count methods.8, 19, Longitudinal Changes
28, 29, 31, 33, 36-38, 41, 43-45, 48, 49, 51, 54-56, 58, 59, 64, 66, 68, 72, 78, 82, 84-86, 88
Four longitudinal studies met inclusion criteria. Two of
these8,89 analyzed first-episode and chronic schizophrenia
Schizophrenia vs Controls separately, leading to 6 pairwise comparisons of baseline vs
Random-effects meta-analysis found higher neutrophil and follow-up data. Because of the limited number of studies, no
monocyte counts in schizophrenia compared with controls separate moderator analyses were performed for longitudi-
(Table 1, Table 2, Figure 2, and Figure 3). Effect sizes were mod- nal data to avoid overinterpretation.
erate for neutrophils (g = 0.69; 95% CI, 0.49 to 0.89; Bonfer- Neutrophils decreased longitudinally (g = −0.30; 95% CI,
roni-adjusted P < .001; n = 40 951 [47 between-group com- −0.45 to −0.15; Bonferroni-adjusted P < .001; n = 896 [4 within-
parisons]) and monocytes (g = 0.49; 95% CI, 0.24 to 0.75; group comparisons]) and eosinophils increased longitudi-
Bonferroni-adjusted P < .001; n = 40 513 [44 between-group nally (g = 0.61; 95% CI, 0.52 to 0.71; Bonferroni-adjusted
comparisons]). Differences in eosinophils (g = 0.02; 95% CI, P < .001; n = 876 [3 within-group comparisons]) after success-
−0.16 to 0.20; Bonferroni-adjusted P > .99; n = 3277 [18 be- ful treatment of acute psychosis (Table 1 and eTable 4 and eFig-
tween-group comparisons]), basophils (g = 0.14; 95% CI, −0.06 ures 4 and 6 in Supplement 1). Table 2 and eTables 5 and 6 and
to 0.34; Bonferroni-adjusted P = .85; n = 2614 [13 between- eFigures 5, 7, and 8 in Supplement 1 contain data for mono-
group comparisons]), and lymphocytes (g = −0.08; 95% CI, cytes, basophils, and lymphocytes.
−0.21 to 0.06; Bonferroni-adjusted P > .99; n = 41 693 [59 be-
tween-group comparisons]) were not significant (eTables 4-6 Sensitivity Analysis
and eFigures 1-4 in Supplement 1). Heterogeneity was explored by comparing leukocyte differ-
ences across 54 clinical8,28-36,38-40,42-44,47-56,58,59,61-63,65-67,69-83,
85-89
Chronic vs First-Episode Schizophrenia and 10 population-based studies.19,37,41,45,46,57,60,64,68,84
Neutrophil counts were higher with large effect sizes in first- Neutrophil effect sizes were moderate in both settings, with
episode schizophrenia compared with healthy controls larger effects in clinical studies (Table 1). Monocyte effect size
(g = 0.85; 95% CI, 0.35-1.34; Bonferroni-adjusted P = .004; was moderate in clinical studies but not significant in popula-
n = 2602 [17 between-group comparisons]) but moderately el- tion-based settings (Table 2). No such differences were found
evated in chronic schizophrenia compared with healthy con- for eosinophils, basophils, and lymphocytes (eTables 4-6 in
trols (g = 0.61; 95% CI, 0.44-0.77; Bonferroni-adjusted P < .001; Supplement 1). Leave-1-out analysis confirmed significant over-
n = 38 349 [30 between-group comparisons]). Monocyte counts all neutrophil and monocyte effects, with heterogeneity re-
showed a large effect size in first-episode schizophrenia maining significant regardless of the study excluded (eTables 7-8
(g = 0.91; 95% CI, 0.25-1.57; Bonferroni-adjusted P = .03; in Supplement 1).
n = 2308 [16 between-group comparisons]) but only small ef-
fects in chronic schizophrenia (g = 0.28; 95% CI, 0.13-0.44; Publication Bias
Bonferroni-adjusted P = .002; n = 38 205 [28 between-group Funnel plots and Egger tests showed no significant bias for
comparisons]) (Table 1 and Table 2). Eosinophil, basophil, monocyte, eosinophil, basophil, or lymphocyte counts, but sig-
and lymphocyte effect sizes for patients with first-episode and nificant publication bias was detected for neutrophil differ-
chronic schizophrenia compared with healthy controls were ences (eFigures 9-13 in Supplement 1). The Orwin fail-safe test
similar (eTables 4-6 in Supplement 1). indicated 12 710 studies would be needed to nullify the neu-
trophil findings. No bias analyses were conducted for longi-
Effect of Antipsychotic Medication and Comorbidities tudinal data because of the small number of studies.
Neutrophil differences were large in patients who were
antipsychotic naive or free (Table 1), whereas effect size was Matching and Effect Sizes
moderate in patients taking medication. Monocyte differ- In 51 of 62 studies,8, 19, 28-30, 33-35, 38, 39, 41-43, 45-49, 51-60, 62-66, 68-81,
83, 85, 86, 88
ences followed a similar pattern, with the largest effect in an- patients and controls were age-matched. Five
tipsychotic-naive patients (Table 2). (Although there were studies31,32,40,50,82 were sex-matched only, and 6 were un-
only 2 pairwise comparisons for monocyte counts in antipsy- matched for age or sex.28,36,37,44,61,67,89 Effect sizes for age-
chotic-free individuals, we included these data because of the and sex-matched studies were similar to overall leukocyte
large effect size and for completeness.) Studies excluding subpopulation effects (Table 1, Table 2, and eTables 4-6 in
comorbidities8,28-32,35-39,41,43-66,68,70-86,88,89 showed a greater Supplement 1).
neutrophil and monocyte effect size (Table 1 and Table 2).

Meta-Regression and Covariate Analyses

Between-group meta-regression analyses found no signifi-
Discussion
cant relationship between neutrophil counts and moderators To our knowledge, this meta-analysis is the largest to date ex-
like sex, BMI, age, smoking, duration of illness, age at onset, amining absolute WBC subpopulation counts in patients with
PANSS scores, sample size, and NOS scores (Table 1). Mono- schizophrenia and healthy controls, focusing on disease- and

jamapsychiatry.com (Reprinted) JAMA Psychiatry Published online March 5, 2025 E5

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Research Original Investigation Differences in Blood Leukocyte Subpopulations in Schizophrenia

Table 1. Between- and Within-Group Meta-Analyses for Neutrophil Counts in Schizophrenia vs Healthy Controlsa

No. of participants Meta-analysis Heterogeneity

No. of Bonferroni P Bonferroni
pairwise SCZ HC Hedges g (95% CI) P value P I2 Q value P
Group wise
Between-group SCZ vs HC
All 47 24 814 16 137 0.69 (0.49 to 0.89) <.001b <.001b 98.395 735.383 <.001b <.001b
b b
Without antipsychotic 12 856 1052 1.16 (0.51 to 1.82) <.001 .003 97.301 229.120 <.001b <.001b
treatment
Antipsychotic naive 10 665 702 1.17 (0.37 to 1.97) .004b .02b 97.384 225.599 <.001b <.001b
b b
Antipsychotic free 2 191 350 1.14 (0.94 to 1.33) <.001 <.001 0.000 0.007 .93 >.99
With antipsychotic 34 23 911 15 055 0.52 (0.38 to 0.65) <.001b <.001b 95.618 299.625 <.001b <.001b
treatment
First-episode SCZ 17 1354 1248 0.85 (0.35 to 1.34) <.001b .004b 96.921 272.612 <.001b <.001b
b b
Chronic SCZ 30 23 460 14 889 0.61 (0.44 to 0.77) <.001 <.001 97.055 396.190 <.001b <.001b
b b
Age and sex matched 32 9913 9003 0.65 (0.46 to 0.84) <.001 <.001 94.627 434.463 <.001b <.001b
b b
Clinical samples 39 24 366 15 779 0.72 (0.49 to 0.94) <.001 <.001 98.671 684.469 <.001b <.001b
b b
Populational 8 448 358 0.58 (0.14 to 1.02) .01 .048 87.674 41.381 <.001b <.001b
Without clinical 43 24 207 15 726 0.72 (0.51 to 0.94) <.001b <.001b 98.585 730.444 <.001b <.001b
conditions
With clinical conditions 4 607 411 0.32 (0.20 to 0.45) <.001b <.001b 0.024 3.228 .36 >.99
Within-group SCZ baseline 4 896 NA −0.30 (−0.45 to <.001b <.001b 31.544 6.408 .09 .47
and follow up (all) −0.15)
Moderator in meta-regression
No. of participants Meta-regression Meta-regression
No. of
pairwise SCZ HC Slope (95% CI) P value Intercept z P value
Between-group SCZ vs HC
% Males (all) 46 24 764 16 087 −0.01 (−0.02 to 0.01) .40 1.047 2.263 .02b
Without antipsychotic 11 806 1002 −0.03 (−0.07 to 0.02) .22 2.700 2.006 .045b
treatment
With antipsychotic 34 23 911 15 055 0.00 (−0.01 to 0.01) .92 0.486 1.477 .14
treatment
BMI (all)c 23 1501 1518 −0.05 (−0.18 to 0.08) .45 2.132 1.287 .20
Without antipsychotic 7 510 850 0.35 (−0.06 to 0.76) .09 −6.963 −1.368 .17
treatment
With antipsychotic 15 944 638 −0.05 (−0.12 to 0.03) .21 1.789 1.829 .07
treatment
Age (all) 47 24 814 16 137 −0.02 (−0.05 to 0.01) .22 1.342 2.518 .01b
Age (without 12 856 1052 −0.03 (−0.17 to 0.10) .62 2.286 1.001 .32
antipsychotic
treatment)
Age (with antipsychotic 34 23 911 15 055 −0.01 (−0.03 to 0.01) .35 0.836 2.420 .02b
treatment)
% Smokers (all) 19 1553 1616 −0.02 (−0.04 to 0.00) .08 1.432 3.738 <.001b
Duration of illness (all) 29 3087 2661 −0.01 (−0.03 to 0.02) .57 0.727 5.576 <.001b
Age at onset (all) 30 3112 2705 0.03 (−0.01 to 0.06 .16 0.025 0.053 .96
PANSS total scores (all) 23 1975 1896 0.01 (−0.01 to 0.03) .26 0.126 0.180 .86
Positive scores (all) 25 2064 1949 0.03 (−0.01 to 0.06) .16 0.268 0.677 .50
Negative scores (all) 25 2064 1949 0.00 (−0.05 to 0.05) .93 0.752 1.577 .12
General scores (all) 20 1862 1788 0.01 (−0.03 to 0.04) .76 0.592 0.910 .36
Sample size (all)d 47 24 814 16 137 −0.00004 (−0.00009 .23 0.726 6.953 <.001b
to 0.00002)
NOS (all) 47 24 814 16 137 −0.02 (−0.20 to 0.16) .86 0.809 1.208 .23
Abbreviations: BMI, body mass index; HC, healthy controls; NA, not applicable; between “without antipsychotic treatment” and “with antipsychotic
NOS, Newcastle-Ottawa Scale; PANSS, Positive and Negative Syndrome Scale; treatment” only account for 46 of 47 total studies.
SCZ, schizophrenia. b
Values significant at P and Bonferroni-adjusted P <.05.
a
In the within-group analysis, PANSS total scores refer to baseline. There was c
Calculated as weight in kilograms divided by height in meters squared.
1 study with an unclear mix of patients, some receiving antipsychotics and d
The slope and the 95% CI are given up to the fifth decimal place, as this is
others not, without providing subgroup data. As a result, these data were
the point at which they are clearly distinguishable.
omitted from the table above. Consequently, the pairwise comparisons

E6 JAMA Psychiatry Published online March 5, 2025 (Reprinted) jamapsychiatry.com

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Differences in Blood Leukocyte Subpopulations in Schizophrenia Original Investigation Research

Table 2. Between- and Within-Group Meta-Analyses for Monocyte Counts in Schizophrenia vs Healthy Controlsa

No. of participants Meta-analysis Heterogeneity

No. of Bonferroni P Bonferroni
pairwise SCZ HC Hedges g (95% CI) P value P I2 Q value P
Group wise
Between-group SCZ vs HC
All 44 24 562 15 951 0.49 (0.24 to 0.75) <.001b <.001b 99.006 558.098 <.001b <.001b
b b
Without antipsychotic 10 671 942 1.02 (0.28 to 1.76) .007 .04 97.550 164.026 <.001b <.001b
treatment
Antipsychotic naive 8 480 592 1.06 (0.11 to 2.00) .03b .14 97.593 160.703 <.001b <.001b
b b
Antipsychotic free 2 191 350 0.82 (0.63 to 1.01) <.001 <.001 0.001 1.110 .29 >.99
With antipsychotic 32 23 752 14 885 0.32 (0.10 to 0.55) .004b .02b 98.546 261.716 <.001b <.001b
treatment
First episode SCZ 16 1178 1130 0.91 (0.25 to 1.57) .007b .03b 98.006 315.673 <.001b <.001b
b b
Chronic SCZ 28 23 384 14 821 0.28 (0.13 to 0.44) <.001 .002 96.746 205.954 <.001b <.001b
b b
Age and sex-matched 32 9914 9037 0.42 (0.19 to 0.65) <.001 .002 96.625 322.819 <.001b <.001b
b b
Clinical samples 35 24 085 15 566 0.44 (0.19 to 0.70) <.001 .003 98.954 475.637 <.001b <.001b
Populational 9 477 385 0.72 (−0.10 to 1.53) .08 .42 96.755 78.975 <.001b <.001b
b b
Without clinical 42 24 115 15 655 0.51 (0.25 to 0.78) <.001 <.001 99.066 554.082 <.001b <.001b
conditions
With clinical conditions 2 447 296 0.15 (0.01 to 0.30) .04b .21 0.000 0.117 .73 >.99
Within-group SCZ baseline 3 876 NA −0.09 (−0.40 to 0.21) .56 >.99 83.179 22.167 <.001b <.001b
and follow up
Moderator in meta-regression
No. of participants Meta-regression Meta-regression
No. of
pairwise SCZ HC Slope (95% CI) P value Intercept z P value
Between-group SCZ vs HC
% Males (all) 43 24 531 15 920 −0.01 (−0.03 to 0.01) .49 0.891 1.541 .12
Without antipsychotic 10 671 942 0.00 (−0.05 to 0.04) .85 1.254 0.985 .32
treatment
With antipsychotic 31 23 721 14 854 −0.01 (−0.03 to 0.01) .40 0.782 1.404 .16
treatment
BMI (all)c 22 1498 1521 −0.06 (−0.24 to 0.12) .52 2.250 0.967 .33
Without antipsychotic 7 550 843 0.48 (0.07 to 0.90) .02 −10.551 −2.072 .04
treatment
With antipsychotic 14 901 648 −0.12 (−0.30 to 0.05) .17 3.761 1.610 .11
treatment
Age (all) 43 24 531 15 920 −0.04 (−0.08 to 0.00) .05 1.836 2.631 .009
Without antipsychotic 10 671 942 −0.11 (−0.28 to 0.05) .18 4.676 1.691 .09
treatment
With antipsychotic 31 23 721 14 854 −0.02 (−0.06 to 0.01) .18 1.128 1.870 .06
treatment
% Smokers (all) 17 1430 1525 −0.03 (−0.06 to 0.01) .10 1.598 2.841 .005
Duration of illness (all) 28 2914 2665 −0.02 (−0.06 to 0.02) .25 0.682 3.196 .001
Age at onset (all) 29 2939 2709 0.01 (−0.05 to 0.08) .64 0.135 0.164 .87
PANSS total scores (all) 24 2050 1905 0.01 (−0.02 to 0.03) .43 −0.008 −0.009 .99
Positive scores (all) 25 2073 1893 0.02 (−0.03 to 0.08) .41 0.242 0.392 .70
Negative scores (all) 25 2073 1893 −0.01 (−0.09 to 0.06) .71 0.981 1.325 .19
General scores (all) 21 1898 1766 0.01 (−0.03 to 0.06) .60 0.304 0.324 .75
Sample size (all)d 44 24 562 15 951 −0.00001 (−0.00009 .71 0.508 3.777 <.001
to 0.00006)
NOS (all) 44 24 562 15 951 0.20 (−0.07 to 0.46) .15 −0.965 −0.957 .34
Abbreviations: BMI, body mass index; HC, healthy controls; NA, not applicable; treatment” and “with antipsychotic treatment” only account for 42 of 44 total
NOS, Newcastle-Ottawa Scale; PANSS, Positive and Negative Syndrome Scale; studies.
SCZ, schizophrenia. b
Values significant at P and Bonferroni-adjusted P <.05.
a
In the within-group analysis, PANSS total scores refer to baseline. There were c
Calculated as weight in kilograms divided by height in meters squared.
2 studies with an unclear mix of patients, some receiving antipsychotics and d
The slope and the 95% CI are given up to the fifth decimal place, as this is the
others not, without providing subgroup data. These studies are not shown
point at which they are clearly distinguishable.
in the table above because we needed at least 3 for a proper calculation.
Consequently, the pairwise comparisons between “without antipsychotic

jamapsychiatry.com (Reprinted) JAMA Psychiatry Published online March 5, 2025 E7

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Research Original Investigation Differences in Blood Leukocyte Subpopulations in Schizophrenia

Figure 2. Forest Plot for Random-Effects Between-Group Meta-Analysis of Neutrophil Granulocyte Counts
in Patients With Schizophrenia (n = 24 814) and Healthy Controls (n = 16 137)

Decreased Increased P value after

neutrophils neutrophils Bonferroni
Source Hedges g (95% CI) in SCZ in SCZ P value correction
Caldwell et al,37 1991 –0.79 (–1.81 to 0.22) .12 .58
Wilke et al,80 1996 1.26 (0.79 to 1.72) .001c .001c
Delieu et al,44 2006 0.52 (0.18 to 0.85) .003b .01a
Chow et al,42 2015 0.68 (0.26 to 1.11) .001b .007b
Miller et al,59 2015 0.17 (–0.19 to 0.52) .35 >.99
Pavlović et al,66 2016 0.50 (0.22 to 0.78) <.001c 002b
Özdin et al,65 2017 0.47 (0.25 to 0.70) <.001c .001c
Yüksel et al,83 2018 1.02 (0.61 to 1.43) <.001c <.001c
Moody et al,60 2018 0.93 (0.42 to 1.45) <.001c .002b
Catak et al,38 2018 0.16 (–0.11 to 0.44) .24 >.99
Kelly et al,55 2018 0.35 (–0.27 to 0.96) .26 >.99
Orhan et al,64 2018 1.94 (1.32 to 2.56) <.001c .001c
Garcia-Rizo et al,45 2019 0.26 (–0.06 to 0.58) .10 .52
Núñez et al,19 2019 0.29 (0.01 to 0.57) .04a .20
Tunç et al,76 2019 0.53 (0.02 to 1.04) .04a .19
Olcina Rodríguez et al,70 2019 0.64 (0.35 to 0.92) .001c <.001c
Garcia-Ruiz et al,46 2020 0.66 (0.25 to 1.07) .001b .007b
Steiner et al,8 2020 (SCZ) 1.14 (0.92 to 1.37) <.001c <.001c
Steiner et al,8 2020 (FEP) 1.17 (0.95 to 1.39) <.001c <.001c
Balcioglu et al,32 2020 (SCZ) 0.61 (0.49 to 0.74) <.001c <.001c
Bulut et al,35 2021 0.67 (0.37 to 0.98) <.001c <.001c
Keleş Altun et al,54 2021 0.49 (0.08 to 0.89) .02a .09
Onur et al,63 2021 1.18 (0.72 to 1.64) <.001c <.001c
Zhu et al,85 2022 –0.05 (–0.19 to 0.09) .46 >.99
Xu et al,81 2022 0.09 (0.06 to 0.13) <.001c <.001c
Obeagu et al,62 2022 2.05 (1.56 to 2.53) <.001c <.001c
Garés-Caballer et al,47 2022 0.05 (–0.46 to 0.57) .84 >.99
Gokulakrishnan et al,49 2022 0.96 (0.40 to 1.53) <.001c .004b
Huang et al,50 2022 0.40 (0.02 to 0.77) .04a .19
Leung et al,57 2022 0.98 (0.41 to 1.54) <.001c .003b
Ali et al,29 2022 1.57 (1.10 to 2.05) <.001c .001c
Wei et al,79 2022 0.22 (0.19 to 0.25) <.001c <.001c
Pujol et al,69 2022 0.34 (0.02 to 0.66) .04a .18
Bioque et al,34 2022 0.27 (0.09 to 0.44) .002b .01a
Sahpolat et al,74 2022 2.71 (2.10 to 3.31) <.001c <.001c
Kılıç et al,56 2023 0.72 (0.36 to 1.08) <.001c <.001c
Balcioglu et al,31 2023 (A, SCZ) 1.12 (0.74 to 1.50) <.001c <.001c
Balcioglu et al,31 2023 (A, FEP) 3.86 (3.22 to 4.49) <.001c <.001c
Balcioglu et al,30 2023 (B, SCZ) 0.47 (0.09 to 0.84) .01a .07
Inaltekin et al,52 2023 0.31 (–0.01 to 0.64) .05 .27
Šagud et al,72 2023 0.46 (0.24 to 0.69) <.001c <.001c
Cabello-Rangel et al,36 2023 –0.18 (–0.49 to 0.13) .26 >.99
Çelik et al,39 2023 –0.12 (–0.48 to 0.24) .52 >.99
Imre et al,51 2023 0.28 (–0.10 to 0.66) .14 .70
Juchnowicz et al,53 2023 0.59 (0.12 to 1.06) .01a .06
After Bonferroni correction for the
leukocyte subpopulations, the overall
Zhou et al,84 2024 0.13 (–0.22 to 0.47) .46 >.99
P value for neutrophils remained
Wang et al,78 2024 1.13 (0.78 to 1.49) <.001c .001c
significant at P < .001. Square sizes
Overall 0.69 (0.49 to 0.89) <.001c <.001c
are proportional to sample sizes.
a
P < .05.
–2 –1 0 1 2 3 4 5
b
Hedges g (95% CI) P < .01.
c
P < .001.

treatment-related differences and potential confounders. Neu- monocyte counts in first-episode vs chronic schizophrenia, but
trophil and monocyte levels were notably higher in first- contrast with findings from Jackson et al.10 Increased neutro-
episode schizophrenia compared with chronic cases, suggest- phil and monocyte counts support the immune hypothesis
ing differences in immune cell counts at various disease stages. of schizophrenia, emphasizing innate immune activation.
Furthermore, neutrophils and monocytes were largely el- These cells produce proinflammatory cytokines, which are
evated in patients who were antipsychotic-naive or -free but elevated in acute schizophrenia.4 Elevated neutrophils, key in
moderately elevated in those taking antipsychotics. bacterial defense,10 could suggest infections trigger acute
These findings from the cross-sectional comparisons align psychosis in vulnerable individuals.3 Altered immune cell
with our previous research8 showing higher neutrophil and counts might also stem from congenital immune changes,

E8 JAMA Psychiatry Published online March 5, 2025 (Reprinted) jamapsychiatry.com

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Differences in Blood Leukocyte Subpopulations in Schizophrenia Original Investigation Research

Figure 3. Forest Plot for Random-Effects Between-Group Meta-Analysis of Monocyte Counts in Patients
With Schizophrenia (n = 24 562) and Healthy Controls (n = 15 951)

Decreased Increased P value after

monocytes monocytes Bonferroni
Source Hedges g (95% CI) in SCZ in SCZ P value correction
Caldwell et al,37 1991 –0.47 (–1.47 to 0.52) .33 >.99
Wilke et al,80 1996 0.76 (0.32 to 1.19) <.001c .003b
Zorrilla et al,86 1996 0.43 (0.14 to 0.72) .004b .02a
Rudolf et al,71 2004 0.26 (–0.24 to 0.76) .30 >.99
Miller et al,59 2015 0.05 (–0.30 to 0.41) .76 >.99
Pavlović et al,66 2016 0.47 (0.19 to 0.75) <.001c .005b
Özdin et al,65 2017 0.00 (–0.22 to 0.22) >.99 >.99
Yüksel et al,83 2018 0.71 (0.31 to 1.10) <.001c .002b
Moody et al,60 2018 0.71 (0.20 to 1.21) .006b .03a
Catak et al,38 2018 0.08 (–0.20 to 0.35) .59 >.99
Kelly et al,55 2018 0.37 (–0.25 to 0.99) .23 >.99
Orhan et al,64 2018 4.48 (3.54 to 5.43) <.001c <.001c
Garcia-Rizo et al,45 2019 0.44 (0.13 to 0.76) .006b .03a
Núñez et al,19 2019 0.19 (–0.08 to 0.47) .17 .84
Olcina Rodríguez et al,70 2019 0.00 (–0.28 to 0.28) >.99 >.99
Garcia-Ruiz et al,46 2020 0.33 (–0.08 to 0.73) .11 .56
Steiner et al,8 2020 (SCZ) 0.76 (0.54 to 0.98) <.001c <.001c
Steiner et al,8 2020 (FEP) 1.41 (1.18 to 1.64) <.001c <.001c
Balcioglu et al,32 2020 (SCZ) 0.09 (–0.04 to 0.21) .16 .82
Sahpolat et al,73 2021 0.10 (–0.22 to 0.42) .53 >.99
Keleş Altun et al,54 2021 0.58 (0.17 to 0.98) .005b .03a
Onur et al,63 2021 1.73 (1.23 to 2.22) <.001c <.001c
Yan et al,82 2022 0.07 (–0.32 to 0.47) .72 >.99
Zhu et al,85 2022 0.10 (–0.04 to 0.23) .18 .91
Xu et al,81 2022 0.23 (0.20 to 0.26) <.001c <.001c
Garés-Caballer et al,47 2022 0.00 (–0.52 to 0.52) >.99 >.99
Gokulakrishnan et al,49 2022 0.09 (–0.45 to 0.62) .75 >.99
Huang et al,50 2022 0.65 (0.27 to 1.03) <.001c .004b
Leung et al,57 2022 0.28 (–0.26 to 0.82) .30 >.99
Ali et al,29 2022 –0.50 (–0.92 to –0.08) .02a .10
Wei et al,79 2022 0.35 (0.32 to 0.38) <.001c <.001c
Chen et al,41 2022 0.36 (–0.17 to 0.89) .18 .88
Bioque et al,34 2022 0.14 (–0.04 to 0.31) .12 .62
Kılıç et al,56 2023 0.56 (0.21 to 0.92) .002b .009b
Balcioglu et al,31 2023 (A, SCZ) 0.99 (0.62 to 1.37) <.001c <.001c
Balcioglu et al,31 2023 (A, FEP) 4.30 (3.62 to 4.99) <.001c <.001c
Balcioglu et al,30 2023 (B, SCZ) –0.04 (–0.41 to 0.33) .82 >.99
Inaltekin et al,52 2023 –0.28 (–0.60 to 0.04) .08 .41
Šagud et al,72 2023 0.00 (–0.22 to 0.22) >.99 >.99
Çelik et al,39 2023 –0.21 (–0.57 to 0.15) .25 >.99
Imre et al,51 2023 –0.26 (–0.64 to 0.12) .18 .92
Juchnowicz et al,53 2023 0.49 (0.02 to 0.95) .04a .20 After Bonferroni correction for the
leukocyte subpopulations, the overall
Zhou et al,84 2024 0.40 (0.02 to 0.75) .02a .12
P value for monocytes remained
Wang et al,78 2024 1.69 (1.30 to 2.07) <.001c <.001c
significant at P < .001. Square sizes
Overall 0.49 (0.24 to 0.75) <.001c <.001c
are proportional to sample sizes.
a
P < .05.
–2 –1 0 1 2 3 4 5 6
b
Hedges g (95% CI) P < .01.
c
P < .001.

supported by genetic studies linking leukocyte counts to in patients who were antipsychotic-naive or -free than in
schizophrenia.90 treated individuals with schizophrenia.
Larger leukocyte differences in patients who were anti- The more prominent findings in patients with first-
psychotic-naive or -free compared with patients receiving episode vs relapsed disease suggest an early peak of immune
medication may result from neutropenic and monocytope- activation that may trigger disease onset and progression. This
nic effects of antipsychotics like clozapine and olanzapine, aligns with inflammatory blood markers being associated with
which reduce inflammation and improve symptoms.91-96 transition to psychosis in individuals at high clinical risk.99,100
Alternatively, antipsychotic medication may dampen the im- However, the persistent neutrophil elevation in schizophrenia
mune system by reducing inflammatory cytokines.97,98 These suggests a broader role of chronic low-grade inflammation,
findings, together with the current literature, support the hy- which probably also contributes to the increased cardiovascu-
pothesis that immune cell counts and inflammation are higher lar and metabolic risk linked to this disease.101 Inflammation

jamapsychiatry.com (Reprinted) JAMA Psychiatry Published online March 5, 2025 E9

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Research Original Investigation Differences in Blood Leukocyte Subpopulations in Schizophrenia

may underlie both metabolic dysfunction and schizophrenia. population counts (64 studies, 26 349 patients with schizo-
Our finding that BMI is associated with monocyte counts in pa- phrenia, 16 379 healthy controls), providing high statistical
tients who were antipsychotic naive or free aligns with obesity- power and enabling subgroup analyses to identify effects of
related low-grade inflammation,102,103 highlighting the need to disease state and medication (eTable 9 in Supplement 1). With
consider BMI and metabolic syndrome as covariates in future a wealth of data from contacting original study authors and
studies. Importantly, our results show that elevated neutro- its longitudinal component, our meta-analysis expands
phils and monocytes are not solely explained by smoking, and solidifies knowledge in this key area. We assessed dis-
suggesting an intrinsic link to schizophrenia itself. These find- ease severity using PANSS criteria, whereas previous studies
ings highlight the need for longitudinal studies to investigate relied on inpatient vs outpatient status or did not address this
immune mechanisms across disease stages, considering both point.10,14 We also conducted meta-regression to explore as-
systemic and brain-specific effects of chronic inflammation. sociations with various potential moderators (eTable 9 in
Longitudinal comparisons also support a decrease in neu- Supplement 1).
trophil counts and an increase in eosinophil counts after antipsy- One limitation is that subgroup sample size imbalances
chotic treatment, possibly due to normalization of the inflam- and missing BMI and PANSS data in some studies reduced
matory state, symptom improvement, or immunomodulating the certainty of meta-regression results. Other moderators
medication effects. No associations between longitudinal neu- such as race and ethnicity, genetic and epigenetic data, sub-
trophil differences and disease modifiers were found, likely be- stance use, and environmental stressors were not included
cause of the limited number of studies. because of insufficient data. Sensitivity analyses suggested
In contrast to previous studies linking neutrophil7,8 and that heterogeneity could not be explained by excluding spe-
monocyte counts 104,105 with disease severity, our meta- cific studies. Potential publication bias was indicated for
analysis found no such associations except for eosinophils. This neutrophil counts, but the Orwin fail-safe test suggested
discrepancy may stem from variations in the number and char- minimal bias. Lastly, effects of specific antipsychotics could
acteristics of studies that reported PANSS scores, which could not be determined because of a lack of detailed data in most
influence overall findings. studies.
Clinical studies typically show larger effect sizes than popu-
lation-based studies because of smaller, more homogeneous
groups with severe symptoms and controlled settings.106
We confirmed this for neutrophils but not for monocytes. The
Conclusions
significant neutrophil and monocyte increases in population- This meta-analysis revealed elevated neutrophil and mono-
based studies suggest our findings apply broadly, despite po- cyte counts in schizophrenia, particularly in individuals with
tential confounding variables. first-episode schizophrenia and those who were antipsychotic-
Leave-1-out analyses confirmed the robustness of our naive or -free. The observed decrease in neutrophil counts
results, with effect sizes unaffected by any single study. The and increase in eosinophil counts after antipsychotic treat-
observed heterogeneity likely reflects individual differences ment may reflect treatment-related immune modulation rather
in WBC counts influenced by genetic, lifestyle, and environ- than acute illness itself. Notably, the longitudinal increase in
mental factors, as well as study-specific variables like demo- eosinophils is consistent with the concept of the “eosinophil-
graphics, antipsychotic treatment status, and disease stage. driven dawn of recovery,” a phenomenon observed during the
While subgroup analyses mitigated some heterogeneity, per- resolution of bacterial inflammation.107 However, the signifi-
sistent variability underscores the complex relationship be- cance of this finding in schizophrenia remains unclear due to
tween immune responses and schizophrenia. This heteroge- the lack of data on microbial exposure in the original studies
neity adds real-world relevance. It highlights the variability of analyzed in our meta-analysis.
immune dysregulation in schizophrenia. Our study underscores the importance of considering po-
tential disease modifiers and underlying conditions, such as
Strengths and Limitations BMI, when examining immune-related changes in schizophre-
The main strength of this meta-analysis is its inclusion of the nia, paving the way for new therapeutic approaches grounded
largest number of studies to date on absolute leukocyte sub- in the etiopathology of the disease.

ARTICLE INFORMATION Steiner); German Center for Mental Health (DZPG), Clinical Immunology, Hannover Medical School
Accepted for Publication: December 1, 2024. Partner Site Halle-Jena-Magdeburg, Magdeburg, (MHH), Hannover, Germany (Jacobs);
Germany (Nickl-Jockschat, Steiner); Department of Department of Psychiatry and Psychotherapy,
Published Online: March 5, 2025. Psychiatry, Iowa Neuroscience Institute, Ludwig-Maximilian-University, Munich, Germany
doi:10.1001/jamapsychiatry.2024.4941 Department of Neuroscience and Pharmacology, (Schiltz).
Author Affiliations: Department of Psychiatry, University of Iowa, Iowa City (Nickl-Jockschat); Author Contributions: Drs Dobrowolny and
Otto-von-Guericke-University Magdeburg, Laboratory of Neuroproteomics, Department of Steiner had full access to all of the data in the study
Magdeburg, Germany (Dudeck, Nussbaumer, Biochemistry and Tissue Biology, University of and take responsibility for the integrity of the data
Nickl-Jockschat, Guest, Dobrowolny, Meyer-Lotz, Campinas (UNICAMP), Campinas, Brazil (Guest); and the accuracy of the data analysis. Joint first
Schiltz, Steiner); Laboratory of Translational Center for Precision Health, McWilliams School of authorship: Mr Dudeck and Ms Nussbaumer
Psychiatry, Otto-von-Guericke-University Biomedical Informatics, The University of Texas contributed equally as co–first authors.
Magdeburg, Magdeburg, Germany (Dudeck, Health Science Center at Houston, Houston (Zhao, Concept and design: Fernandes, Steiner.
Nussbaumer, Guest, Dobrowolny, Meyer-Lotz, Fernandes); Department of Rheumatology and

E10 JAMA Psychiatry Published online March 5, 2025 (Reprinted) jamapsychiatry.com

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Differences in Blood Leukocyte Subpopulations in Schizophrenia Original Investigation Research

Acquisition, analysis, or interpretation of data: comparisons between schizophrenia, bipolar 20. Cochrane Training. Cochrane Handbook for
All authors. disorder, and depression. Schizophr Bull. 2018;44 Systematic Reviews of Interventions version 6.4.
Drafting of the manuscript: Dudeck, Nussbaumer, (1):75-83. doi:10.1093/schbul/sbx035 Updated August 2023. Accessed January 21, 2025.
Guest, Schiltz, Fernandes, Steiner. 6. Steiner J, Schmitt A, Falkai P. Publishing negative http://www.training.cochrane.org/handbook
Critical review of the manuscript for important results of neurobiological studies in mental 21. Peterson J, Welch V, Losos M, Tugwell P.
intellectual content: Nickl-Jockschat, Guest, disorders will advance knowledge in The Newcastle-Ottawa Scale (NOS) for Assessing
Dobrowolny, Meyer-Lotz, Zhao, Jacobs, Fernandes, pathophysiology. Schizophr Bull. 2023;49(3):539- the Quality of Nonrandomised Studies in Meta-
Steiner. 541. doi:10.1093/schbul/sbad011 Analyses. Ottawa Hospital Research Institute. 2011.
Statistical analysis: Dobrowolny, Zhao, Schiltz, 7. Singh D, Guest PC, Dobrowolny H, et al. Changes 22. Soeken KL, Sripusanapan A. Assessing
Fernandes. in leukocytes and CRP in different stages of major publication bias in meta-analysis. Nurs Res. 2003;
Obtained funding: Steiner. depression. J Neuroinflammation. 2022;19(1):74. 52(1):57-60. doi:10.1097/00006199-200301000-
Administrative, technical, or material support: doi:10.1186/s12974-022-02429-7 00009
Nickl-Jockschat, Dobrowolny, Meyer-Lotz, Jacobs, 8. Steiner J, Frodl T, Schiltz K, et al. Innate immune 23. Bowden J, Tierney JF, Copas AJ, Burdett S.
Steiner. cells and C-reactive protein in acute first-episode Quantifying, displaying and accounting for
Supervision: Guest, Zhao, Fernandes, Steiner. psychosis and schizophrenia: relationship to heterogeneity in the meta-analysis of RCTs using
Conflict of Interest Disclosures: psychopathology and treatment. Schizophr Bull. standard and generalised Q statistics. BMC Med Res
Dr Nickl-Jockschat reported personal fees from 2020;46(2):363-373. doi:10.1093/schbul/sbz068 Methodol. 2011;11:41. doi:10.1186/1471-2288-11-41
Johnson and Johnson and nonfinancial support 9. Liew PX, Kubes P. The neutrophil’s role during 24. Higgins JP, Thompson SG, Deeks JJ, Altman DG.
from Boehringer Ingelheim outside the submitted health and disease. Physiol Rev. 2019;99(2):1223- Measuring inconsistency in meta-analyses. BMJ.
work. No other disclosures were reported. 1248. doi:10.1152/physrev.00012.2018 2003;327(7414):557-560. doi:10.1136/bmj.327.
10. Jackson AJ, Miller BJ. Meta-analysis of total 7414.557
Funding/Support: This project received funding
from the German Federal Ministry of Education and and differential white blood cell counts in 25. Lau J, Ioannidis JP, Schmid CH. Quantitative
Research (German Center for Mental Health [DZPG] schizophrenia. Acta Psychiatr Scand. 2020;142(1): synthesis in systematic reviews. Ann Intern Med.
start phase, site Halle-Jena-Magdeburg, location 18-26. doi:10.1111/acps.13140 1997;127(9):820-826. doi:10.7326/0003-4819-
Magdeburg, funding code 01EE2305D) and the 11. Karageorgiou V, Milas GP, Michopoulos I. 127-9-199711010-00008
Saxony-Anhalt Ministry of Research (start-up Neutrophil-to-lymphocyte ratio in schizophrenia: 26. Leucht S, Rothe P, Davis JM, Engel RR.
funding DZPG, Magdeburg, funding code I 212) A systematic review and meta-analysis. Schizophr Res. Equipercentile linking of the BPRS and the PANSS.
to Dr Steiner. 2019;206:4-12. doi:10.1016/j.schres.2018.12.017 Eur Neuropsychopharmacol. 2013;23(8):956-959.
12. Mazza MG, Capellazzi M, Lucchi S, Tagliabue I, doi:10.1016/j.euroneuro.2012.11.004
Role of the Funder/Sponsor: The funders had no
role in the design and conduct of the study; Rossetti A, Clerici M. Monocyte count in 27. van Erp TG, Preda A, Nguyen D, et al.
collection, management, analysis, and schizophrenia and related disorders: a systematic Converting positive and negative symptom scores
interpretation of the data; preparation, review, review and meta-analysis. Acta Neuropsychiatr. between PANSS and SAPS/SANS. Schizophr Res.
2020;32(5):229-236. doi:10.1017/neu.2020.12 2014;152(1):289-294. doi:10.1016/j.schres.2013.11.013
or approval of the manuscript; and decision to
submit the manuscript for publication. 13. Mazza MG, Lucchi S, Rossetti A, Clerici M. 28. Achiron A, Noy S, Pras E, Lereya J, Hermesh H,
Neutrophil-lymphocyte ratio, monocyte- Laor N. T-cell subsets in acute psychotic
Data Sharing Statement: See Supplement 2.
lymphocyte ratio and platelet-lymphocyte ratio schizophrenic patients. Biol Psychiatry. 1994;35(1):
Additional Contributions: We thank Alexandra in non-affective psychosis: a meta-analysis and 27-31. doi:10.1016/0006-3223(94)91163-0
Neyazi, MD, Department of Psychiatry, systematic review. World J Biol Psychiatry. 2020;21 29. Ali E, Embaby A, Ibrahim D. Mean platelet
Otto-von-Guericke-University Magdeburg, for (5):326-338. doi:10.1080/15622975.2019.1583371 volume, red cell distribution width, and lymphocyte
helpful feedback on the first manuscript draft. We 14. Clausen M, Christensen RHB, da Re M, ratios as surrogate predictors of subclinical
are grateful to all the authors of the original studies Benros ME. Immune cell alterations in psychotic inflammation in schizophrenia: a cross-sectional
who provided us with additional information and disorders: a comprehensive systematic review and study. EuroMediterranean Biomedical J. 2022;17
data for the meta-analysis. None of the above meta-analysis. Biol Psychiatry. 2024;96(5):331-341. (39):181-185. doi:10.3269/1970-5492.2022.17.39
received financial compensation for their doi:10.1016/j.biopsych.2023.11.029 30. Balcioglu YH, Gokcay H, Yesilkaya UH,
contributions. 15. Goldsmith DR, Rapaport MH, Miller BJ. Balcioglu SSK. Blood viscosity and inflammatory
A meta-analysis of blood cytokine network indices in treatment-resistant schizophrenia:
REFERENCES alterations in psychiatric patients: comparisons a retrospective cross-sectional study. Dusunen
1. Benros ME, Nielsen PR, Nordentoft M, Eaton WW, between schizophrenia, bipolar disorder and Adam-Journal of Psychiatry and Neurological
Dalton SO, Mortensen PB. Autoimmune diseases and depression. Mol Psychiatry. 2016;21(12):1696-1709. Sciences. 2023;36(2):81-89. doi:10.14744/DAJPNS.
severe infections as risk factors for schizophrenia: doi:10.1038/mp.2016.3 2023.00210
a 30-year population-based register study. Am J 16. Marcinowicz P, Więdłocha M, Zborowska N, 31. Balcıoğlu YH, Gökçay H, Yeşilkaya ÜH,
Psychiatry. 2011;168(12):1303-1310. doi:10.1176/appi. et al. A Meta-analysis of the influence of Namlı MN. Blood viscosity and inflammation
ajp.2011.11030516 antipsychotics on cytokines levels in first episode in first-episode and acute exacerbations of
2. Benros ME, Waltoft BL, Nordentoft M, et al. psychosis. J Clin Med. 2021;10(11):2488. schizophrenia: a case-control study with healthy
Autoimmune diseases and severe infections as risk doi:10.3390/jcm10112488 controls. Noro Psikiyatr Ars. 2023;60(3):265-270.
factors for mood disorders: a nationwide study. 17. Miller BJ, Gassama B, Sebastian D, Buckley P, doi:10.29399/npa.28394
JAMA Psychiatry. 2013;70(8):812-820. doi:10.1001/ Mellor A. Meta-analysis of lymphocytes in 32. Balcioglu YH, Kirlioglu SS. C-reactive
jamapsychiatry.2013.1111 schizophrenia: clinical status and antipsychotic protein/albumin and neutrophil/albumin ratios
3. Nielsen PR, Benros ME, Mortensen PB. Hospital effects. Biol Psychiatry. 2013;73(10):993-999. as novel inflammatory markers in patients with
contacts with infection and risk of schizophrenia: doi:10.1016/j.biopsych.2012.09.007 schizophrenia. Psychiatry Investig. 2020;17(9):
a population-based cohort study with linkage of 18. Adamowicz DH, Shilling PD, Palmer BW, et al. 902-910. doi:10.30773/pi.2020.0185
Danish national registers. Schizophr Bull. 2014;40 Associations between inflammatory marker profiles 33. Bessler H, Eviatar J, Meshulam M, Tyano S,
(6):1526-1532. doi:10.1093/schbul/sbt200 and neurocognitive functioning in people with Djaldetti M, Sirota P. Theophylline-sensitive
4. Halstead S, Siskind D, Amft M, et al. Alteration schizophrenia and non-psychiatric comparison T-lymphocyte subpopulation in schizophrenic
patterns of peripheral concentrations of cytokines subjects. J Psychiatr Res. 2022;149:106-113. patients. Biol Psychiatry. 1987;22(8):1025-1028.
and associated inflammatory proteins in acute doi:10.1016/j.jpsychires.2022.02.029 doi:10.1016/0006-3223(87)90012-6
and chronic stages of schizophrenia: a systematic 19. Núñez C, Stephan-Otto C, Usall J, et al; 34. Bioque M, Catarina Matias-Martins A,
review and network meta-analysis. Lancet Psychiatry. PEPs Group. Neutrophil count is associated with Llorca-Bofí V, et al; 2EPs GROUP. Neutrophil to
2023;10(4):260-271. doi:10.1016/S2215-0366(23) reduced gray matter and enlarged ventricles in lymphocyte ratio in patients with a first episode
00025-1 first-episode psychosis. Schizophr Bull. 2019;45(4): of psychosis: a two-year longitudinal follow-up
5. Wang AK, Miller BJ. Meta-analysis of 846-858. doi:10.1093/schbul/sby113 study. Schizophr Bull. 2022;48(6):1327-1335.
cerebrospinal fluid cytokine and tryptophan doi:10.1093/schbul/sbac089
catabolite alterations in psychiatric patients:

jamapsychiatry.com (Reprinted) JAMA Psychiatry Published online March 5, 2025 E11

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Research Original Investigation Differences in Blood Leukocyte Subpopulations in Schizophrenia

35. Bulut NS, Yorguner N, Çarkaxhiu Bulut G. a validated mass spectrometric method. J Pharm with schizophrenia in Southeast, Nigeria. Intl J Res
The severity of inflammation in major Biomed Anal. 2020;180:113056. doi:10.1016/j.jpba. Rep Hematol. 2022;5(2):107-112.
neuropsychiatric disorders: comparison of 2019.113056 63. Onur D, Neslihan AK, Samet K. A comparative
neutrophil-lymphocyte and platelet-lymphocyte 49. Gokulakrishnan K, Nikhil J, Vs S, et al. Altered study of complete blood count inflammatory
ratios between schizophrenia, bipolar mania, intestinal permeability biomarkers in schizophrenia: markers in substance-free acute psychotic disorder
bipolar depression, major depressive disorder, and a possible link with subclinical inflammation. Ann and substance-induced psychosis. Early Interv
obsessive compulsive disorder. Nord J Psychiatry. Neurosci. 2022;29(2-3):151-158. doi:10.1177/ Psychiatry. 2021;15(6):1522-1530. doi:10.1111/eip.
2021;75(8):624-632. doi:10.1080/08039488. 09727531221108849 13089
2021.1919201
50. Huang K, Tang Y, Chen Z, et al. Comparison of 64. Orhan F, Schwieler L, Fatouros-Bergman H,
36. Cabello-Rangel H, Basurto-Morales M, hematological parameters between first-episode et al; Karolinska Schizophrenia Project (KaSP)
Botello-Aceves E, Pazarán-Galicia O. Mean platelet schizophrenia and anti-NMDAR encephalitis. Front Consortium. Increased number of monocytes and
volume, platelet count, and neutrophil/lymphocyte Cell Dev Biol. 2022;10:895178. doi:10.3389/fcell. plasma levels of MCP-1 and YKL-40 in first-episode
ratio in drug-naïve patients with schizophrenia: 2022.895178 psychosis. Acta Psychiatr Scand. 2018;138(5):
a cross-sectional study. Front Psychiatry. 2023;14: 432-440. doi:10.1111/acps.12944
1150235. doi:10.3389/fpsyt.2023.1150235 51. İmre O, Caglayan C, Muştu M. The relationship
of cognitive dysfunction with inflammatory 65. Özdin S, Sarisoy G, Böke Ö. A comparison of
37. Caldwell CL, Irwin M, Lohr J. Reduced natural markers and carotid intima media thickness in the neutrophil-lymphocyte, platelet-lymphocyte
killer cell cytotoxicity in depression but not in schizophrenia. J Pers Med. 2023;13(9):1342. and monocyte-lymphocyte ratios in schizophrenia
schizophrenia. Biol Psychiatry. 1991;30(11):1131-1138. doi:10.3390/jpm13091342 and bipolar disorder patients: a retrospective file
doi:10.1016/0006-3223(91)90183-M review. Nord J Psychiatry. 2017;71(7):509-512.
52. İnaltekin A, Yağci İ. Evaluation of simple
38. Catak Z, Uzmez E, Ozturk N, Ugur K. Research markers of inflammation and systemic immune doi:10.1080/08039488.2017.1340517
article comparison of neutrophil-to-lymphocyte, inflammation index in schizophrenia, bipolar 66. Pavlović M, Babić D, Rastović P, Babić R,
platelet-to-lymphocyte, and monocyte-to- disorder patients and healthy controls [in Turkish]. Vasilj M. Metabolic syndrome, total and
lymphocyte ratios in patients with schizophrenia, Turk Psikiyatri Derg. 2023;34(1):11-15. differential white blood cell counts in patients with
bipolar disorder, and major depressive disorder. doi:10.5080/u26248 schizophrenia. Psychiatr Danub. 2016;28(suppl 2):
Int J Med Biochem. 2018;1(3):106-110. 216-222.
53. Juchnowicz D, Dzikowski M, Rog J, et al.
39. Çelik HEA. Evaluation of substance induced The usefulness of a complete blood count in the 67. Pfeiffer CC, Smyrl EG, Iliev V. Extreme basophil
and substance free first-episode psychosis in terms prediction of the first episode of schizophrenia counts and blood histamine levels in schizophrenic
of inflammatory whole blood count parameters. diagnosis and its relationship with oxidative stress. outpatients as compared to normals. Res Commun
Klinik Psikiyatri Dergisi-Turkish Journal of Clinical PLoS One. 2023;18(10):e0292756. doi:10.1371/ Chem Pathol Pharmacol. 1972;4(1):51-59.
Psychiatry. 2023;26(3):170-176. doi:10.5505/kpd. journal.pone.0292756
2023.14564 68. Printz DJ, Strauss DH, Goetz R, et al.
54. Keleş Altun İ, Atagün Mİ, Erdoğan A, et al. Elevation of CD5+ B lymphocytes in schizophrenia.
40. Chang SH, Chiang SY, Chiu CC, et al. Expression Serum hepcidin / ferroportin levels in bipolar Biol Psychiatry. 1999;46(1):110-118. doi:10.1016/
of anti-cardiolipin antibodies and inflammatory disorder and schizophrenia. J Trace Elem Med Biol. S0006-3223(98)00307-2
associated factors in patients with schizophrenia. 2021;68:126843. doi:10.1016/j.jtemb.2021.126843
Psychiatry Res. 2011;187(3):341-346. doi:10.1016/j. 69. Pujol N, Bergé D, Mané A, et al; PEPs GROUP.
psychres.2010.04.049 55. Kelly DL, Li X, Kilday C, et al. Increased The influence of modifiable cardiovascular risk
circulating regulatory T cells in medicated people factors on cognition, functioning, and inflammatory
41. Chen S, Fan F, Xuan FL, et al. Monocytic subsets with schizophrenia. Psychiatry Res. 2018;269: markers in first-episode psychosis: results from a
impact cerebral cortex and cognition: differences 517-523. doi:10.1016/j.psychres.2018.09.006 2-year follow-up study. Psychiatry Res. 2022;316:
between healthy subjects and patients with 114760. doi:10.1016/j.psychres.2022.114760
first-episode schizophrenia. Front Immunol. 2022; 56. Kılıç N, Tasci G, Yılmaz S, Öner P, Korkmaz S.
13:900284. doi:10.3389/fimmu.2022.900284 Monocyte/HDL cholesterol ratios as a new 70. Olcina Rodríguez JF, Díaz Fernández J,
inflammatory marker in patients with Orta Mira N, López-Briz E, Gómez-Pajares F.
42. Chow V, Reddel C, Pennings G, et al. Global schizophrenia. J Pers Med. 2023;13(2):13. Neutrophil/lymphocyte ratio: can a non-specific
hypercoagulability in patients with schizophrenia doi:10.3390/jpm13020276 marker of inflammation helps to confirm the
receiving long-term antipsychotic therapy. inflammatory hypothesis of the serious mental
Schizophr Res. 2015;162(1-3):175-182. doi:10.1016/j. 57. Leung KK, Wong YC, Shea KS, et al. Altered
neutrophil-to-lymphocyte ratio in patients with diseases? a case-control study. Med Hypotheses.
schres.2014.12.042 2019;130:109279. doi:10.1016/j.mehy.2019.
non-affective first episode psychosis and its
43. Coffey CE, Sullivan JL, Rice JR. T lymphocytes relationship with symptom severity and cognitive 109279
in schizophrenia. Biol Psychiatry. 1983;18(1):113-119. impairment. Sci Rep. 2022; 13(1):11453. 71. Rudolf S, Schlenke P, Broocks A, et al. Search for
44. Delieu JM, Horobin RW, Duguid JK. Formation doi:10.1038/s41598-023-37846-y atypical lymphocytes in schizophrenia. World J Biol
of immature neutrophil leucocytes in schizophrenic 58. Masserini C, Vita A, Basile R, et al. Lymphocyte Psychiatry. 2004;5(1):33-37. doi:10.1080/
patients treated with various antipsychotic drugs: subsets in schizophrenic disorders: relationship 15622970410029905
comparisons and predictions. J Psychopharmacol. with clinical, neuromorphological and treatment 72. Šagud M, Madžarac Z, Nedic Erjavec G, et al.
2006;20(6):824-828. doi:10.1177/ variables. Schizophr Res. 1990;3(4):269-275. The associations of neutrophil-lymphocyte,
0269881106061112 doi:10.1016/0920-9964(90)90008-U platelet-lymphocyte, monocyte-lymphocyte ratios
45. Garcia-Rizo C, Casanovas M, Fernandez-Egea E, 59. Miller BJ, Kandhal P, Rapaport MH, Mellor A, and immune-inflammation index with negative
et al. Blood cell count in antipsychotic-naive Buckley P. Total and differential white blood cell symptoms in patients with schizophrenia.
patients with non-affective psychosis. Early Interv counts, high-sensitivity C-reactive protein, and Biomolecules. 2023;13(2):297. doi:10.3390/
Psychiatry. 2019;13(1):95-100. doi:10.1111/eip.12456 cardiovascular risk in non-affective psychoses. biom13020297
46. Garcia-Ruiz B, Moreno L, Muntané G, et al. Brain Behav Immun. 2015;45:28-35. doi:10.1016/j. 73. Sahpolat M, Ayar D, Ari M, Karaman MA.
Leukocyte and brain DDR1 hypermethylation is bbi.2014.12.005 Elevated monocyte to high-density lipoprotein
altered in psychosis and is correlated with stress 60. Moody G, Miller BJ. Total and differential white ratios as an inflammation markers for schizophrenia
and inflammatory markers. Epigenomics. 2020;12 blood cell counts and hemodynamic parameters in patients. Clin Psychopharmacol Neurosci. 2021;19
(3):251-265. doi:10.2217/epi-2019-0191 first-episode psychosis. Psychiatry Res. 2018;260: (1):112-116. doi:10.9758/cpn.2021.19.1.112
47. Garés-Caballer M, Sánchez-Ortí JV, 307-312. doi:10.1016/j.psychres.2017.11.086 74. Sahpolat M, Karaman MA, Copur EO, Ayar D,
Correa-Ghisays P, et al. Immune-inflammatory 61. Nyland H, Naess A, Lunde H. Lymphocyte Sesliokuyucu C. Increased neutrophil to
biomarkers predict cognition and social functioning subpopulations in peripheral blood from lymphocyte and platelet to lymphocyte ratios in
in patients with type 2 diabetes mellitus, major schizophrenic patients. Acta Psychiatr Scand. patients with first episode psychosis. Med J Bakirkoy.
depressive disorder, bipolar disorder, and 1980;61(4):313-318. doi:10.1111/j.1600-0447.1980. 2022;18(4):483-483. doi:10.4274/BMJ.galenos.2022.
schizophrenia: a 1-year follow-up study. Front Neurol. tb00584.x 2021.11-9
2022;13:883927. doi:10.3389/fneur.2022.883927 75. Schleifer SJ, Keller SE, Siris SG, Davis KL,
62. Obeagu EI, Johnson AD, Arinze-Anyiam OC,
48. Kamil Gharab KM, Onmaz DE, Abusoglu S, et al. Anyiam AF, Ramos GF, Esimai BN. Neutrophils to Stein M. Depression and immunity: lymphocyte
The relationship between serum clozapine lymphocytes ratio and some cytokines in patients function in ambulatory depressed patients,
concentrations and hematological parameters by hospitalized schizophrenic patients, and patients

E12 JAMA Psychiatry Published online March 5, 2025 (Reprinted) jamapsychiatry.com

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025
 Differences in Blood Leukocyte Subpopulations in Schizophrenia Original Investigation Research

hospitalized for herniorrhaphy. Arch Gen Psychiatry. monocyte/lymphocyte ratios in schizophrenia. 382(9896):951-962. doi:10.1016/S0140-6736(13)
1985;42(2):129-133. doi:10.1001/archpsyc.1985. Australas Psychiatry. 2022;30(1):95-99. 60733-3
01790250023003 doi:10.1177/10398562211022753 97. Dutheil S, Watson LS, Davis RE, Snyder GL.
76. Tunç S, Atagün Mİ, Neşelioğlu S, Bilgin YY, 86. Zorrilla EP, Cannon TD, Gur RE, Kessler J. Lumateperone normalizes pathological levels of
Başbuğ HS, Erel Ö. Ischemia-modified albumin: Leukocytes and organ-nonspecific autoantibodies acute inflammation through important pathways
a unique marker of global metabolic risk in in schizophrenics and their siblings: markers of known to be involved in mood regulation. J Neurosci.
schizophrenia and mood disorders. Psychiatry vulnerability or disease? Biol Psychiatry. 1996;40 2023;43(5):863-877. doi:10.1523/JNEUROSCI.0984-
Clin Psychopharmacol. 2019;29(2):123-129. (9):825-833. doi:10.1016/0006-3223(95)00598-6 22.2022
doi:10.1080/24750573.2018.1517466 87. Mauri MC, Volonteri LS, Dell’Osso B, et al. 98. Feng Z, Zhang Y, You X, et al. Effects of
77. Varsak N, Aydin M, Eren I. The evaluation Predictors of clinical outcome in schizophrenic risperidone on blood levels of interleukin-6 in
of neutrophil-lymphocyte ratio in patients with patients responding to clozapine. J Clin schizophrenia: a meta-analysis. Medicine (Baltimore).
first episode psychosis. Fam Pract Palliative Care. Psychopharmacol. 2003;23(6):660-664. 2020;99(15):e19694. doi:10.1097/MD.
2016;1(3):65-69. doi:10.22391/920.287411 doi:10.1097/01.jcp.0000095351.32154.3a 0000000000019694
78. Wang Y, Wang G, Gong M, et al. Systemic 88. Xiong Z, Cheng M, Zhu P, et al. Association 99. Khoury R, Nasrallah HA. Inflammatory
inflammatory biomarkers in schizophrenia are of blood cell counts with the risk of olanzapine- biomarkers in individuals at clinical high risk for
changed by ECT administration and related to or clozapine-induced dyslipidemia in Chinese psychosis (CHR-P): state or trait? Schizophr Res.
the treatment efficacy. BMC Psychiatry. 2024; schizophrenia patients. Hum Psychopharmacol. 2018;199:31-38. doi:10.1016/j.schres.2018.04.017
24(1):53. doi:10.1186/s12888-023-05469-2 2019;34(4):e2699. doi:10.1002/hup.2699 100. Mondelli V, Blackman G, Kempton MJ, et al;
79. Wei Y, Wang T, Li G, et al. Investigation 89. Sperner-Unterweger B, Whitworth A, EU GEI High Risk Study. Serum immune markers
of systemic immune-inflammation index, Kemmler G, et al. T-cell subsets in schizophrenia: and transition to psychosis in individuals at clinical
neutrophil/high-density lipoprotein ratio, a comparison between drug-naive first episode high risk. Brain Behav Immun. 2023;110:290-296.
lymphocyte/high-density lipoprotein ratio, patients and chronic schizophrenic patients. doi:10.1016/j.bbi.2023.03.014
and monocyte/high-density lipoprotein ratio Schizophr Res. 1999;38(1):61-70. doi:10.1016/ 101. Khandaker GM. Commentary: causal
as indicators of inflammation in patients with S0920-9964(98)00175-3 associations between inflammation,
schizophrenia and bipolar disorder. Front 90. Steen NE, Rahman Z, Szabo A, et al. Shared cardiometabolic markers and schizophrenia:
Psychiatry. 2022;13:941728. doi:10.3389/fpsyt. genetic loci between schizophrenia and white the known unknowns. Int J Epidemiol. 2019;48(5):
2022.941728 blood cell counts suggest genetically determined 1516-1518. doi:10.1093/ije/dyz201
80. Wilke I, Arolt V, Rothermundt M, Weitzsch C, systemic immune abnormalities. Schizophr Bull. 102. Brooks GC, Blaha MJ, Blumenthal RS.
Hornberg M, Kirchner H. Investigations of cytokine 2023;49(5):1345-1354. doi:10.1093/schbul/ Relation of C-reactive protein to abdominal
production in whole blood cultures of paranoid sbad082 adiposity. Am J Cardiol. 2010;106(1):56-61.
and residual schizophrenic patients. Eur Arch 91. Myles N, Myles H, Xia S, et al. Meta-analysis doi:10.1016/j.amjcard.2010.02.017
Psychiatry Clin Neurosci. 1996;246(5):279-284. examining the epidemiology of clozapine-
doi:10.1007/BF02190280 103. Trayhurn P, Wood IS. Adipokines:
associated neutropenia. Acta Psychiatr Scand. inflammation and the pleiotropic role of white
81. Xu H, Wei Y, Zheng L, et al. Relation between 2018;138(2):101-109. doi:10.1111/acps.12898 adipose tissue. Br J Nutr. 2004;92(3):347-355.
unconjugated bilirubin and peripheral biomarkers 92. Glocker C, Grohmann R, Burkhardt G, et al. doi:10.1079/BJN20041213
of inflammation derived from complete blood Antipsychotic drug-induced neutropenia: results
counts in patients with acute stage of 104. Dimitrov DH. Correlation or coincidence
from the AMSP drug surveillance program between monocytosis and worsening of psychotic
schizophrenia. Front Psychiatry. 2022;13:843985. between 1993 and 2016. J Neural Transm (Vienna).
doi:10.3389/fpsyt.2022.843985 symptoms in veterans with schizophrenia?
2023;130(2):153-163. doi:10.1007/s00702-023- Schizophr Res. 2011;126(1-3):306-307.
82. Yan J, Chen Y, Ju P, et al. Network association 02589-7 doi:10.1016/j.schres.2010.06.003
of biochemical and inflammatory abnormalities 93. Patel R, Lima A, Burke C, Hoffman M.
with psychiatric symptoms in first-episode 105. Miller BJ, Buckley PF, McEvoy JP.
Monocytopenia in clozapine-induced Inflammation, substance use, psychopathology,
schizophrenia patients. Front Psychiatry. 2022;13: agranulocytosis: insights into pathophysiology
834539. doi:10.3389/fpsyt.2022.834539 and cognition in phase 1 of the clinical antipsychotic
and treatment. BMJ Case Rep. 2019;12(1):bcr- trials of intervention effectiveness study. Schizophr
83. Yüksel RN, Ertek IE, Dikmen AU, Göka E. 2018-226016. doi:10.1136/bcr-2018-226016 Res. 2018;195:275-282. doi:10.1016/j.schres.2017.
High neutrophil-lymphocyte ratio in schizophrenia 94. Alvir JM, Lieberman JA, Safferman AZ, 08.027
independent of infectious and metabolic Schwimmer JL, Schaaf JA. Clozapine-induced
parameters. Nord J Psychiatry. 2018;72(5):336-340. 106. Fernandes BS, Steiner J, Bernstein HG, et al.
agranulocytosis: incidence and risk factors in C-reactive protein is increased in schizophrenia but
doi:10.1080/08039488.2018.1458899 the United States. N Engl J Med. 1993;329(3): is not altered by antipsychotics: meta-analysis and
84. Zhou L, Ma X, Wang W. Immune dysregulation 162-167. doi:10.1056/NEJM199307153290303 implications. Mol Psychiatry. 2016;21(4):554-564.
is associated with symptom dimensions and 95. Malhotra K, Vu P, Wang DH, Lai H, Faziola LR. doi:10.1038/mp.2015.87
cognitive deficits in schizophrenia: accessible Olanzapine-induced neutropenia. Ment Illn. 2015;7
evidence from complete blood count. BMC 107. Gleich GJ. Historical overview and perspective
(1):5871. doi:10.4081/mi.2015.5871 on the role of the eosinophil in health and disease.
Psychiatry. 2024;24(1):48. doi:10.1186/s12888-
023-05430-3 96. Leucht S, Cipriani A, Spineli L, et al. In: Lee JJ, Rosenberg HF, eds. Eosinophils in Health
Comparative efficacy and tolerability of and Disease. Academic Press; 2013:1-11. doi:10.1016/
85. Zhu X, Zhou J, Zhu Y, et al. Neutrophil/ 15 antipsychotic drugs in schizophrenia: B978-0-12-394385-9.00001-8
lymphocyte, platelet/lymphocyte and a multiple-treatments meta-analysis. Lancet. 2013;

jamapsychiatry.com (Reprinted) JAMA Psychiatry Published online March 5, 2025 E13

© 2025 American Medical Association. All rights reserved, including those for text and data mining, AI training, and similar technologies.
Downloaded from jamanetwork.com by guest on 03/08/2025