The correlations of insomnia symptoms with psychopathology and hematological parameters as well as gender differences in patients with schizophrenia

Abstract

Patients with schizophrenia are at high risk for insomnia, but the associated influencing factors in this population have not been fully elucidated. Therefore, this study aimed to investigate the associations between insomnia symptoms and psychopathological indicators and hematological parameters in patients with schizophrenia, as well as to analysis the gender differences. From October 2022 to December 2024, this study recruited 184 patients with schizophrenia. The Insomnia Severity Index Scale (ISI), Brief Psychiatric Rating Scale (BPRS), Calgary Depression Scale (CDSS), and Modified Overt Aggression Scale (MOAS) were used to assess insomnia symptoms, psychotic symptoms, depressive symptoms, and aggressive behaviors, respectively. In addition, this study examined a range of blood parameters including leukocytes, neutrophils, lymphocytes, monocytes, platelets, total cholesterol, triglycerides, high density lipoprotein cholesterol, low density lipoprotein cholesterol, triiodothyronine (T3), thyroxine (T4), and thyroid-stimulating hormone (TSH); and calculated the neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR) and monocyte to lymphocyte ratio (MLR). Finally, independent factors of insomnia symptoms were identified using stepwise logistic regression, and the predictive value of each factor was evaluated via ROC curve analysis. The prevalence of insomnia symptoms in patients with schizophrenia was 34.8%, with 29.1% in males and 42.0% in females. Regression analysis showed that the independent influences of insomnia symptoms in the total sample were BPRS total score, CDSS total score, NLR (Ln) and T3. In male patients, BPRS total score, CDSS total score, and NLR (Ln) remained significant independent predictors. In contrast, only CDSS total score emerged as a statistically significant independent predictor in female patients. Furthermore, ROC curve analysis further demonstrated that the four-item combination of BPRS total score, CDSS total score, NLR (Ln), and T3 (AUC = 0.825, 95% CI = 0.766–0.885, P < 0.001) had a better ability to identify symptoms of insomnia in the total sample. Meanwhile, the three-item combination of BPRS total score, CDSS total score and NLR (Ln) (AUC = 0.850, 95% CI = 0.773–0.926, P < 0.001) had a better ability to identify insomnia symptoms in male patients. Patients with schizophrenia exhibited a relatively high risk of insomnia (29.1% in males vs. 42.0% in females); however, this difference was not statistically significant. Insomnia symptoms may be associated with psychotic symptoms, depressive symptoms, NLR, and T3. And there may be some gender differences in these correlations, with psychotic symptoms and NLR being independently linked to insomnia only in males.

Introduction

Schizophrenia is a highly prevalent and severe mental disorder that typically onset during late adolescence to early adulthood. Its primary manifestations include disturbances in thinking, perceptual abnormalities (e.g., auditory and visual hallucinations), delusions, and behavioral disorganization, often accompanied by significant cognitive impairment¹. In addition to core psychotic symptoms, sleep disturbances, particularly insomnia, are extremely prevalent among individuals with schizophrenia. Multiple epidemiological studies have reported that approximately 8% to 80% of patients with schizophrenia experience insomnia of varying severity; this wide range likely reflects differences in assessment tools, sample characteristics, and diagnostic criteria^2–4. Notably, in the clinical management of schizophrenia, prominent psychotic symptoms such as hallucinations and delusions often dominate diagnostic and therapeutic priorities, leading to inadequate recognition and delayed intervention for sleep disturbances. However, accumulating evidence indicates that comorbid insomnia not only exacerbates the severity of both positive and negative symptoms but also significantly impairs attention, working memory, and executive function, thereby markedly reducing social functioning and subjective quality of life^5,6. Furthermore, a systematic review has identified insomnia symptoms as an important early warning sign of schizophrenia relapse⁷. Given the substantial impact of insomnia on prognosis and quality of life in individuals with schizophrenia, a comprehensive understanding of its epidemiological characteristics and modifiable risk factors is critically important.

Previous research has robustly demonstrated that a range of psychopathological symptoms is associated with insomnia in individuals with schizophrenia. First, psychotic symptoms themselves may directly disrupt sleep homeostasis. For instance, auditory hallucinations occurring at night can fragment sleep continuity, while persecutory delusions may sustain a state of heightened vigilance. A cross-sectional study conducted in China further confirmed that the severity of psychotic symptoms is an independent risk factor for insomnia in patients with schizophrenia⁸. Moreover, depressive symptoms also play a significant role in insomnia among patients with schizophrenia. A study of outpatients with schizophrenia found that insomnia is significantly associated with depressive symptoms and serves as an independent predictor of reduced quality of life⁹. Aggressive behaviors, a highly disabling comorbid feature in schizophrenia, may also be associated with insomnia. Razki et al. reported that patients with schizophrenia who exhibit aggressive behaviors have a significantly elevated risk of sleep disturbances¹⁰. Further studies suggest that psychosocial factors including emotional dysregulation and cognitive distortions as well as dysfunction of the prefrontal glutamatergic system may underlie this comorbidity^11,12. Therefore, in clinical assessment and intervention, it is essential to recognize the critical role that psychopathological symptoms play in the onset and progression of insomnia in schizophrenia.

At the biological level, the neuroimmune and metabolic perspectives are progressively reshaping our understanding of schizophrenia and its comorbid conditions. Growing evidence suggests that chronic low-grade inflammation may serve as a potential pathogenic driver in the onset and progression of schizophrenia^13,14. For instance, Barlattani et al. posited that glymphatic clearance dysfunction, leading to the accumulation of neuroinflammation, constitutes a critical pathway in the pathogenesis of psychiatric disorders such as schizophrenia, a process that is intricately linked to sleep-wake rhythms¹⁵. Under inflammatory or other physiological stressors, leukocytes undergo characteristic shifts including neutrophilia and lymphopenia. Based on these changes, ratios such as the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and monocyte-to-lymphocyte ratio (MLR) have emerged as widely adopted, clinically accessible, and biologically informative peripheral markers of systemic inflammation¹⁶. A cross-sectional study demonstrated that insomnia severity is positively correlated with NLR, PLR, and MLR¹⁷. Furthermore, circulating metabolic biomarkers also appear to play a non-negligible role in the manifestation of insomnia. A case-control study found that individuals with schizophrenia and dyslipidemia exhibited significantly more severe sleep disturbances particularly in subjective sleep quality, sleep latency, and total sleep time¹⁸. Green et al. further reported a robust association between thyroid dysfunction and sleep disturbances¹⁹. Nonetheless, current research on insomnia in schizophrenia faces several key limitations. First, although the volume of relevant studies is substantial, findings regarding the associations between insomnia and psychopathology remain inconsistent across reports. Second, most existing studies focus predominantly on conventional clinical variables, whereas investigations into inflammatory biomarkers (e.g., NLR, MLR, PLR) and metabolic parameters (e.g., lipid profile, thyroid hormones) remain limited in scope and scale. Third, few studies have examined whether these clinical and biological associations differ by gender.

Therefore, this study aimed to systematically investigate the associations between insomnia symptoms and psychopathological indicators (psychotic symptoms, depressive symptoms, and aggression), inflammatory markers (NLR, PLR, and MLR), and metabolic parameters (blood lipids and thyroid function indices) in patients with schizophrenia, and further examine whether these associations differ significantly by gender.

Methods

Study design and participants

This cross-sectional study was conducted at the Fourth Affiliated Hospital of Anhui Medical University between October 2022 and December 2024. Inclusion criteria: (1) Age 18–75 years; (2) Clinical diagnosis of schizophrenia according to the International Classification of Diseases, 10th Revision (ICD-10); (3) Ability and willingness to fully comply with all study procedures. Exclusion criteria: (1) Primary diagnosis of another psychiatric disorder (e.g., intellectual developmental disorder); (2) Presence of severe active infection, organic brain disease, autoimmune disorders, or other major systemic illnesses, including respiratory or gastrointestinal diseases; (3) Substance or alcohol abuse (excluding tobacco smoking); (4) Use of glucocorticoids, immunomodulatory agents, or other medications known to significantly influence inflammatory biomarkers within the past 12 months; (5) Pregnancy or lactation. A total of 196 individuals were initially enrolled. Of these, 12 were excluded due to missing critical clinical or laboratory data or inadequate blood sample quality. The final analytical sample comprised 184 participants (Fig. 1).

Fig. 1.

Case screening flowchart.

The study protocol was approved by the Ethics Committee of the Fourth Affiliated Hospital of Anhui Medical University (approval number: KYXM-202210-015). All procedures were conducted in accordance with the ethical standards of the 2013 Helsinki Declaration (https://www.wma.net/policies-post/wma-declaration-of-helsinki/). Written informed consent was obtained from all participants and, where applicable, their legally authorized representatives.

Measuring instruments

General data

This study employed a self-developed questionnaire to collect participants’ general data, including sex, age, body mass index (BMI), smoking history, age at first episode, illness duration, and the type and dosage of currently prescribed antipsychotic medications. To enable standardized cross-drug dose comparisons, all antipsychotic dosages were converted to chlorpromazine-equivalent doses (mg/day), using the World Health Organization (WHO)-recommended Defined Daily Dose (DDD) methodology²⁰.

Insomnia symptoms

Insomnia symptoms over the preceding two weeks were assessed using the Insomnia Severity Index (ISI), a 7-item self-report questionnaire²¹. Each item is rated on a 5-point Likert scale (0–4), yielding a total score ranging from 0 to 28; and higher scores indicate greater subjective insomnia severity. Consistent with widely adopted clinical and research cutoff criteria, an ISI total score > 7 was defined in this study as indicative of clinically significant insomnia²².

Psychotic symptoms

The severity of current psychotic symptoms was assessed using the Brief Psychiatric Rating Scale (BPRS)²³. The BPRS comprises 18 items assessing multiple domains, including conceptual disorganization, hallucinations, delusions, emotional blunting, tension (agitation), anxiety, orientation, and insight. Each item is rated on a 7-point Likert scale (1–7) based on clinical observation and patient self-report, yielding a total score ranging from 18 to 126. Higher total scores reflect greater overall severity and greater clinical activity of psychotic symptoms.

Depressive symptoms

Depressive symptoms in individuals with schizophrenia were assessed using the Calgary Depression Scale for Schizophrenia (CDSS), a 9-item clinician-administered scale specifically developed for this population²⁴. Each item is rated on a 4-point scale (0–3), yielding a total score ranging from 0 to 27; and higher scores indicate greater severity of depressive symptoms.

Aggressive behaviors

Aggressive behaviors were assessed using the Modified Overt Aggression Scale (MOAS)²⁵. The MOAS comprises four behavioral domains: verbal aggression, aggression against objects, auto-aggression (e.g., self-injury), and physical aggression against others. Each item is rated on a 5-point severity scale (0–4). To reflect differential clinical impact, ratings are weighted by domain: verbal aggression ×1, aggression against objects ×2, auto-aggression ×3, and physical aggression against others ×4. The weighted total score ranges from 0 to 60; and higher scores indicate greater frequency, intensity, and clinical severity of overt aggressive behaviors.

Hematological parameter assays

Fasting venous blood samples were collected from all participants between 6:00 and 7:00 a.m. following an overnight fast of at least 8 h. Samples were processed immediately and analyzed at the Department of Clinical Laboratory, the Fourth Affiliated Hospital of Anhui Medical University. White blood cell count, neutrophil count, lymphocyte count, monocyte count, and platelet count were measured using the Sysmex XN-9000 automated hematology analyzer. From these, NLR, PLR, and MLR were calculated. Serum total cholesterol, triglycerides, high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol were quantified using the Siemens Advia Chemistry XPT automated clinical chemistry analyzer. Serum concentrations of triiodothyronine (T3), thyroxine (T4), and thyroid-stimulating hormone (TSH) were determined using the Roche cobas 8000 e 801 fully automated chemiluminescent immunoassay system. To meet the normality assumption required for parametric statistical analyses, natural logarithmic transformation (ln, base e) was applied to variables exhibiting marked right-skewed distributions—specifically NLR, PLR, MLR, total cholesterol, and TSH.

Statistical analysis

Statistical analyses were performed using SPSS version 23.0 (IBM Corp., Armonk, NY, USA). Normality of continuous variables was assessed using the Kolmogorov-Smirnov (K-S) test. Normally distributed continuous variables are presented as mean ± standard deviation (SD), and group comparisons were conducted using independent-samples t-tests. Non-normally distributed continuous variables are reported as median (interquartile range) [M (P₂₅, P₇₅)], with between-group comparisons performed using the Mann-Whitney U test. Categorical variables are expressed as frequency (percentage) [n (%)], and group differences were evaluated using the chi-square (χ²) test. To identify independent predictors of clinically significant insomnia (ISI > 7) in individuals with schizophrenia, a binary logistic regression model was constructed. Variable selection employed the “Forward: Likelihood Ratio” method. All variables with P < 0.10 in the univariate analyses were included as candidate variables in the initial models. Specifically, for the total sample, the included variables were: sex, use of atypical antipsychotics, BPRS total score, CDSS total score, MOAS total score, white blood cell count, neutrophil count, NLR (Ln), MLR (Ln), T3, and T4. For the male subgroup, the included variables were: BPRS total score, CDSS total score, MOAS total score, white blood cell count, neutrophil count, monocyte count, NLR (Ln), MLR (Ln), and T4. For the female subgroup, the included variables were: BPRS total score, CDSS total score, and T3. The final models retained only independent predictors that reached statistical significance (P < 0.05). To evaluate potential multicollinearity among the predictor variables, variance inflation factor (VIF) was calculated for each covariate. All VIF values were below 5.0, which is a commonly accepted threshold indicating that multicollinearity is not a material concern in this model. Furthermore, receiver operating characteristic (ROC) curve analysis was performed to evaluate the discriminative performance of statistically significant independent predictors for insomnia. The area under the ROC curve (AUC), sensitivity, and specificity were calculated. All statistical tests were two-tailed, and a P-value < 0.05 was considered statistically significant.

Results

Comparison of clinical characteristics between insomnia and non-insomnia groups in the total sample

A total of 184 individuals with schizophrenia were included in this study, comprising 103 males (56.0%). The mean age was 44.27 ± 12.77 years. Mean age at first episode was 26.93 ± 9.34 years, mean illness duration was 201.85 ± 133.08 months, and mean chlorpromazine-equivalent antipsychotic dose was 405.71 ± 378.27 mg/day (Table 1). The most commonly prescribed antipsychotics were clozapine (N = 62), aripiprazole (N = 37), quetiapine (N = 29), risperidone (N = 22), paliperidone (N = 8), perospirone (N = 19), ziprasidone (N = 16), amisulpride (N = 12), sulpiride (N = 2), and penfluridol (N = 4).

Table 1.

Comparison of sociodemographic, psychopathological, and hematological parameters between insomnia and non-insomnia groups in the total sample.

Variables	Total sample (N = 184)	Insomnia (N = 64)	Non-insomnia (N = 120)	t/Z/χ2	P
General demographic data
Gender, n (%)				3.300	0.069
Males	103 (55.98)	30 (46.88)	73 (60.83)
Females	81 (44.02)	34 (53.12)	47 (39.17)
Age (years), mean (SD)	44.27 (12.77)	43.44 (13.74)	43.91 (12.51)	-0.235	0.815
BMI (kg/m2), mean (SD)	24.66 (4.21)	24.67 (4.46)	24.74 (4.04)	-0.110	0.912
Smoking, n (%)	51 (27.72)	15 (23.44)	36 (30.00)	0.897	0.344
Age at first episode (years), mean (SD)	26.93 (9.34)	26.45 (9.98)	27.19 (9.00)	-0.510	0.611
Illness duration (months), median (P25, P75)	192.00 (96.00, 300.00)	186.00 (87.00, 339.00)	192.00 (96.00, 300.00)	-0.209 a	0.834
Use of atypical antipsychotics, n (%)	137 (74.46)	42 (65.63)	95 (79.17)	4.024	0.045
Chlorpromazine equivalents (mg/day), mean (SD)	405.71 (378.27)	390.33 (405.35)	413.91 (364.51)	-0.402	0.688
Psychopathology
BPRS total score, mean (SD)	42.04 (9.03)	46.38 (8.14)	39.73 (8.65)	5.069	< 0.001
CDSS total score, median (P25, P75)	3.00 (1.00, 6.00)	6.00 (3.00, 8.00)	2.00 (0.00, 4.00)	-5.734 a	< 0.001
MOAS total score, mean (SD)	3.23 (4.12)	4.14 (4.97)	2.74 (3.52)	2.002	0.048
Hematological parameters
White blood cell count (×109/L), mean (SD)	6.52 (1.89)	6.92 (2.25)	6.30 (1.63)	1.958	0.053
Neutrophil count (×109/L), mean (SD)	4.07 (1.52)	4.50 (1.75)	3.84 (1.34)	2.862	0.005
Lymphocyte count (×109/L), mean (SD)	1.93 (0.61)	1.90 (0.63)	1.94 (0.60)	-0.465	0.642
Platelet count (×109/L), mean (SD)	218.89 (63.62)	211.72 (60.25)	222.72 (65.27)	-1.118	0.265
Monocyte count (×109/L), mean (SD)	0.38 (0.14)	0.40 (0.17)	0.36 (0.11)	1.511	0.134
NLR (Ln), mean (SD) b	0.73 (0.42)	0.85 (0.41)	0.67 (0.41)	2.700	0.008
PLR (Ln), mean (SD) b	4.74 (0.37)	4.73 (0.39)	4.75 (0.37)	-0.351	0.726
MLR (Ln), mean (SD) b	-1.64 (0.35)	-1.58 (0.37)	-1.67 (0.34)	1.760	0.080
Total cholesterol (Ln) (mmol/L), mean (SD) b	1.45 (0.36)	1.45 (0.35)	1.45 (0.37)	0.062	0.951
Triglyceride (mmol/L), mean (SD)	1.51 (0.91)	1.55 (1.05)	1.49 (0.83)	0.447	0.656
High-density lipoprotein cholesterol (mmol/L), mean (SD)	0.98 (0.26)	1.00 (0.28)	0.97 (0.25)	0.732	0.465
Low-density lipoprotein cholesterol (mmol/L), mean (SD)	2.41 (0.76)	2.43 (0.79)	2.40 (0.75)	0.207	0.836
T3 (ng/ml), mean (SD)	0.92 (0.19)	0.86 (0.18)	0.94 (0.19)	-2.833	0.005
T4 (µg/L), mean (SD)	74.33 (15.99)	70.20 (15.08)	76.54 (16.09)	-2.602	0.010
TSH (Ln) (uIU/ml), mean (SD) b	0.52 (0.77)	0.51 (0.73)	0.52 (0.79)	-0.072	0.943

Note: BMI, Body Mass Index; BPRS, Brief Psychiatric Rating Scale; CDSS, Calgary Depression Scale; MOAS, Modified Overt Aggression Scale; NLR, Neutrophil-to-Lymphocyte Ratio; PLR, Platelet-to-Lymphocyte Ratio; MLR, Monocyte-to-Lymphocyte Ratio; T3, Triiodothyronine; T4, Thyroxine; TSH, Thyroid-Stimulating Hormone; ^a, Mann-Whitney U test; ^b, The logarithm with a natural number as the base; SD, Standard Deviation; Bolded P value: < 0.05.

Regarding insomnia symptoms, the prevalence of insomnia in the total sample was 34.8%. Compared with the non-insomnia group, participants in the insomnia group exhibited significantly higher total scores on the BPRS, the CDSS, and the MOAS; higher neutrophil count and NLR (Ln); and lower T3 and T4 concentrations (all P < 0.05) (Table 1; Fig. 2).

Fig. 2.

Comparison of hematological parameters between the insomnia and non-insomnia groups in the total sample. Note: Error bars represent standard deviation.

Comparison of clinical characteristics between insomnia and non-insomnia groups in male patients

In male patients with schizophrenia in this study, the prevalence of insomnia symptoms was 29.1%. Compared with the non-insomnia group, male patients in the insomnia group exhibited significantly higher total scores on the BPRS, the CDSS, and the MOAS; higher white blood cell count, neutrophil count, monocyte count, and NLR (Ln) and MLR (Ln); and lower T4 concentration (all P < 0.05) (Table 2; Fig. 3).

Table 2.

Comparison of sociodemographic, psychopathological, and hematological parameters between insomnia and non-insomnia groups in male patients.

Variables	Insomnia (N = 30)	Non-insomnia (N = 73)	t/Z/χ2	P
General demographic data
Age (years), mean (SD)	44.03 (14.61)	45.10 (12.68)	-0.369	0.713
BMI (kg/m2), mean (SD)	23.69 (3.50)	24.89 (3.81)	-1.484	0.141
Smoking, n (%)	14 (46.67)	34 (46.58)	0.000	0.993
Age at first episode (years), mean (SD)	25.77 (10.41)	27.03 (8.41)	-0.644	0.521
Illness duration (months), median (P25, P75)	234.00 (129.00, 360.00)	204.00 (102.00, 300.00)	-0.581 a	0.561
Use of atypical antipsychotics, n (%)	19 (63.33)	57 (78.08)	2.391	0.122
Chlorpromazine equivalents (mg/day), mean (SD)	305.95 (328.94)	395.62 (338.64)	-1.231	0.221
Psychopathology
BPRS total score, mean (SD)	46.43 (8.03)	39.40 (8.96)	3.729	< 0.001
CDSS total score, median (P25, P75)	5.00 (3.00, 8.25)	2.00 (0.00, 4.00)	-4.154 a	< 0.001
MOAS total score, mean (SD)	5.83 (5.87)	3.03 (3.84)	2.415	0.020
Hematological parameters
White blood cell count (×109/L), mean (SD)	7.41 (2.20)	6.52 (1.72)	2.119	0.030
Neutrophil count (×109/L), mean (SD)	4.96 (1.80)	3.98 (1.49)	2.824	0.006
Lymphocyte count (×109/L), mean (SD)	1.87 (0.60)	2.00 (0.55)	-1.041	0.300
Platelet count (×109/L), mean (SD)	219.87 (64.57)	220.95 (66.52)	-0.075	0.940
Monocyte count (×109/L), mean (SD)	0.45 (0.20)	0.37 (0.11)	2.743	0.007
NLR (Ln), mean (SD) b	0.97 (0.43)	0.67 (0.42)	3.241	0.002
PLR (Ln), mean (SD) b	4.78 (0.44)	4.70 (0.35)	0.879	0.382
MLR (Ln), mean (SD) b	-1.45 (0.39)	-1.69 (0.33)	3.209	0.002
Total cholesterol (Ln) (mmol/L), mean (SD) b	1.41 (0.21)	1.48 (0.44)	-0.824	0.412
Triglyceride (mmol/L), mean (SD)	1.64 (0.96)	1.60 (0.90)	0.204	0.839
High-density lipoprotein cholesterol (mmol/L), mean (SD)	0.91 (0.23)	0.92 (0.20)	-0.126	0.900
Low-density lipoprotein cholesterol (mmol/L), mean (SD)	2.49 (0.84)	2.48 (0.75)	0.084	0.933
T3 (ng/ml), mean (SD)	0.92 (0.17)	0.98 (0.18)	-1.454	0.149
T4 (µg/L), mean (SD)	71.64 (13.51)	78.95 (17.19)	-2.078	0.040
TSH (Ln) (uIU/ml), mean (SD) b	0.30 (0.80)	0.53 (0.85)	-1.289	0.200

Note: BMI, Body Mass Index; BPRS, Brief Psychiatric Rating Scale; CDSS, Calgary Depression Scale; MOAS, Modified Overt Aggression Scale; NLR, Neutrophil-to-Lymphocyte Ratio; PLR, Platelet-to-Lymphocyte Ratio; MLR, Monocyte-to-Lymphocyte Ratio; T3, Triiodothyronine; T4, Thyroxine; TSH, Thyroid-Stimulating Hormone; ^a, Mann-Whitney U test; ^b, The logarithm with a natural number as the base; SD, Standard Deviation; Bolded P value: < 0.05.

Fig. 3.

Comparison of hematological parameters between the insomnia and non-insomnia groups in male patients. Note: Error bars represent standard deviation.

Comparison of clinical characteristics between insomnia and non-insomnia groups in female patients

In female patients with schizophrenia in this study, the prevalence of insomnia symptoms was 42.0%. Compared with the non-insomnia group, female patients in the insomnia group exhibited significantly higher total scores on the BPRS and the CDSS (both P < 0.05) (Table 3). No other between-group differences reached statistical significance.

Table 3.

Comparison of sociodemographic, psychopathological, and hematological parameters between insomnia and non-insomnia groups in female patients.

Variables	Insomnia (N = 34)	Non-insomnia (N = 47)	t/Z/χ2	P
General demographic data
Age (years), mean (SD)	42.91 (13.12)	42.06 (12.15)	0.300	0.765
BMI (kg/m2), mean (SD)	25.54 (5.06)	24.51 (4.41)	0.970	0.335
Smoking, n (%)	1 (2.94)	2 (4.26)	0.096	0.757
Age at first episode (years), mean (SD)	27.06 (9.71)	27.45 (9.95)	-0.175	0.862
Illness duration (months), median (P25, P75)	144.00 (78.00, 267.00)	180.00 (60.00, 252.00)	-0.010 a	0.992
Use of atypical antipsychotics, n (%)	23 (67.65)	38 (80.85)	1.850	0.174
Chlorpromazine equivalents (mg/day), mean (SD)	464.78 (454.32)	442.32 (403.57)	0.234	0.815
Psychopathology
BPRS total score, mean (SD)	46.32 (8.35)	40.23 (8.22)	3.269	0.002
CDSS total score, median (P25, P75)	6.00 (3.00, 8.25)	2.00 (1.00, 4.00)	-3.953 a	< 0.001
MOAS total score, mean (SD)	2.65 (3.45)	2.30 (2.93)	0.491	0.625
Hematological parameters
White blood cell count (×109/L), mean (SD)	6.49 (2.23)	5.96 (1.44)	1.217	0.229
Neutrophil count (×109/L), mean (SD)	4.11 (1.63)	3.62 (1.04)	1.522	0.134
Lymphocyte count (×109/L), mean (SD)	1.92 (0.67)	1.85 (0.66)	0.461	0.646
Platelet count (×109/L), mean (SD)	204.53 (56.15)	225.47 (63.90)	-1.530	0.130
Monocyte count (×109/L), mean (SD)	0.35 (0.12)	0.36 (0.13)	-0.150	0.881
NLR (Ln), mean (SD) b	0.74 (0.36)	0.68 (0.41)	0.658	0.513
PLR (Ln), mean (SD) b	4.69 (0.34)	4.82 (0.39)	-1.599	0.114
MLR (Ln), mean (SD) b	-1.69 (0.32)	-1.64 (0.36)	-0.599	0.551
Total cholesterol (Ln) (mmol/L), mean (SD) b	1.49 (0.44)	1.40 (0.20)	1.241	0.218
Triglyceride (mmol/L), mean (SD)	1.48 (1.12)	1.32 (0.68)	0.773	0.442
High-density lipoprotein cholesterol (mmol/L), mean (SD)	1.08 (0.29)	1.06 (0.29)	0.331	0.742
Low-density lipoprotein cholesterol (mmol/L), mean (SD)	2.37 (0.76)	2.29 (0.73)	0.492	0.624
T3 (ng/ml), mean (SD)	0.81 (0.17)	0.89 (0.18)	-1.980	0.051
T4 (µg/L), mean (SD)	68.92 (16.44)	72.79 (13.54)	-1.160	0.249
TSH (Ln) (uIU/ml), mean (SD) b	0.70 (0.61)	0.50 (0.69)	1.342	0.183

Note: BMI, Body Mass Index; BPRS, Brief Psychiatric Rating Scale; CDSS, Calgary Depression Scale; MOAS, Modified Overt Aggression Scale; NLR, Neutrophil-to-Lymphocyte Ratio; PLR, Platelet-to-Lymphocyte Ratio; MLR, Monocyte-to-Lymphocyte Ratio; T3, Triiodothyronine; T4, Thyroxine; TSH, Thyroid-Stimulating Hormone; ^a, Mann-Whitney U test; ^b, The logarithm with a natural number as the base; SD, Standard Deviation; Bolded P value: < 0.05.

Logistic stepwise regression analysis

The logistic stepwise regression analysis in the total sample revealed that BPRS total score (OR = 1.072, 95% CI = 1.023–1.124, P = 0.004), CDSS total score (OR = 1.293, 95% CI = 1.158–1.444, P < 0.001), NLR (Ln) (OR = 3.119, 95% CI = 1.224–7.948, P = 0.017), and T3 (OR = 0.102, 95% CI = 0.014–0.729, P = 0.023) were all independently associated with insomnia symptoms in patients with schizophrenia (Table 4). In the gender-stratified analyses, among male patients, BPRS total score (OR = 1.083, 95% CI = 1.017–1.153, P = 0.013), CDSS total score (OR = 1.269, 95% CI = 1.101–1.462, P = 0.001), and NLR (Ln) (OR = 5.632, 95% CI = 1.660–19.110, P = 0.006) were identified as independent risk factors for insomnia. In contrast, among female patients, only CDSS total score (OR = 1.376, 95% CI = 1.164–1.627, P < 0.001) remained significantly and independently associated with insomnia.

Table 4.

Independent factors associated with insomnia.

Variables	B	SE	Wald χ2	OR	95% CI	P
The total sample
BPRS total score	0.069	0.024	8.332	1.072	1.023–1.124	0.004
CDSS total score	0.257	0.056	20.931	1.293	1.158–1.444	< 0.001
NLR (Ln)	1.137	0.477	5.679	3.119	1.224–7.948	0.017
T3	-2.280	1.002	5.175	0.102	0.014–0.729	0.023
Male patients
BPRS total score	0.079	0.032	6.150	1.083	1.017–1.153	0.013
CDSS total score	0.238	0.072	10.866	1.269	1.101–1.462	0.001
NLR (Ln)	1.729	0.623	7.690	5.632	1.660–19.110	0.006
Female patients
CDSS total score	0.319	0.085	13.985	1.376	1.164–1.627	< 0.001

Note: BPRS, Brief Psychiatric Rating Scale; CDSS, Calgary Depression Scale; NLR, Neutrophil-to-Lymphocyte Ratio; T3, Triiodothyronine; SE, Standard Error; OR, Odds Ratio; CI, Confidence Interval; Bolded P value: < 0.05.

ROC curve analysis

Table 5; Fig. 4 present the predictive performance of each independent factor for insomnia symptoms in patients with schizophrenia. In the total sample, ROC curve analysis showed the following: BPRS total score (AUC = 0.707, 95% CI = 0.631–0.784, P < 0.001); CDSS total score (AUC = 0.755, 95% CI = 0.681–0.828, P < 0.001); NLR (Ln) (AUC = 0.617, 95% CI = 0.533-0.700, P = 0.009); and T3 (AUC = 0.626, 95% CI = 0.543–0.709, P = 0.005). Notably, when a multi-marker model was constructed by combining BPRS total score, CDSS total score, NLR (Ln), and T3, the AUC increased to 0.825 (95% CI: 0.766–0.885, P < 0.001), which was significantly higher than that of any individual marker. This combined model also demonstrated high sensitivity and specificity, indicating superior discriminative ability for identifying insomnia symptoms.

Table 5.

Predictive value of each independent risk factor for insomnia symptoms.

Variables	Sensitivity	Specificity	AUC	95% CI	P
The total sample
BPRS total score	0.797	0.592	0.707	0.631–0.784	< 0.001
CDSS total score	0.656	0.775	0.755	0.681–0.828	< 0.001
NLR (Ln)	0.734	0.517	0.617	0.533-0.700	0.009
T3	0.766	0.500	0.626	0.543–0.709	0.005
Combined a	0.797	0.725	0.825	0.766–0.885	< 0.001
Male patients
BPRS total score	0.667	0.740	0.713	0.605–0.821	0.001
CDSS total score	0.833	0.630	0.758	0.660–0.857	< 0.001
NLR (Ln)	0.800	0.562	0.692	0.588–0.796	0.002
Combined b	0.800	0.822	0.850	0.773–0.926	< 0.001
Female patients
CDSS total score	0.676	0.787	0.757	0.647–0.867	< 0.001

Note: BPRS, Brief Psychiatric Rating Scale; CDSS, Calgary Depression Scale; NLR, Neutrophil-to-Lymphocyte Ratio; T3, Triiodothyronine; ^a, Combination of BPRS total score, CDSS total score, NLR (Ln) and T3; ^b, Combination of BPRS total score, CDSS total score and NLR (Ln); AUC, Area Under the Curve; CI, Confidence Interval; Bolded P value: < 0.05.

Predictive models prone to overfitting due to lack of external validation.

Fig. 4.

ROC curves illustrating the predictive performance of each independent risk factor for insomnia symptoms. Note: Predictive models prone to overfitting due to lack of external validation.

In male patients, ROC analysis revealed that BPRS total score (AUC = 0.713, 95% CI = 0.605–0.821, P = 0.001), CDSS total score (AUC = 0.758, 95% CI = 0.660–0.857, P < 0.001), and NLR (Ln) (AUC = 0.692, 95% CI = 0.588–0.796, P = 0.002) all exhibited moderate predictive capacity. The combined model incorporating these three markers achieved an AUC of 0.850 (95% CI: 0.773–0.926, P < 0.001), significantly outperforming each individual parameter and demonstrating enhanced discriminatory performance and potential clinical utility.

In female patients, only CDSS total score was retained as a significant predictor, with an AUC of 0.757 (95% CI: 0.647–0.867, P < 0.001), indicating that depression severity has good predictive validity for insomnia symptoms in this subgroup.

Discussion

This study conducted a cross-sectional survey in patients with schizophrenia, revealing an insomnia symptom prevalence of 34.8%. Similarly, a previous cross-sectional study among Chinese patients with schizophrenia reported that approximately 36.0% of participants experienced comorbid insomnia²⁶. However, existing literature shows considerable heterogeneity in the reported prevalence of insomnia among patients with schizophrenia. For instance, a systematic review by Waite et al. found that up to 50% of individuals with schizophrenia reported insomnia symptoms²⁷. This variation may stem from several methodological inconsistencies. First, differences in sample sources may affect sample representativeness and consequently influence prevalence estimates. Second, the lack of standardization in assessment instruments represents another major source of variability. Batalla-Martín and colleagues systematically reviewed the literature and demonstrated that depending on the measurement tool used, reported insomnia rates among individuals with schizophrenia can vary substantially, ranging from 7.9% to 41.2%³. In terms of gender, this study found that the prevalence of insomnia was 42.0% among female patients with schizophrenia and 29.1% among male patients; however, this difference did not reach statistical significance. This finding was consistent with previous reports by Ayers et al., who also found no significant sex differences in sleep problems among individuals with schizophrenia²⁸. In contrast, another survey of Chinese patients with chronic schizophrenia demonstrated a significantly higher prevalence of insomnia in female patients compared with males, a disparity closely associated with female-specific clinical symptoms²⁹. Overall, whether sex differences exist in the prevalence of insomnia among patients with schizophrenia remains inconclusive. Future large-scale, multicenter longitudinal studies are warranted to further investigate this issue.

From a psychopathological perspective, the present study confirmed an independent association between the severity of psychotic symptoms and insomnia, a finding consistent with numerous previous studies^22,28,30. A domestic cross-sectional study demonstrated that schizophrenia patients with insomnia tend to exhibit more severe psychotic symptoms³⁰. First, within the schizophrenia spectrum disorders, positive symptoms appear to be closely associated with sleep disturbances. For instance, persecutory delusions may induce a state of persistent hypervigilance, while auditory hallucinations, especially commanding or commentating types, often become more salient during nighttime when external stimuli are reduced. These disturbances may be associated with difficulties in sleep onset and impaired sleep continuity, thereby contributing to the development of insomnia^31,32. This symptom-driven cognitive-emotional hyperarousal represents the most direct pathway through which psychosis precipitates sleep disruption. Second, psychotic symptoms may act as chronic stressors that significantly elevate anxiety and fear levels. This sustained psychological burden can lead to sympathetic nervous system activation and dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis, including abnormal cortisol rhythms, ultimately impairing sleep homeostasis³³. Notably, the present study found that the association between insomnia and psychotic symptom severity was stronger in male patients. A similar pattern was observed by Chen et al., who reported in a comparative study that although female patients with schizophrenia were more likely to report sleep disturbances, a significant correlation between sleep problems and greater psychotic symptom severity was observed only in males³⁴. Furthermore, Moniem et al. found that male patients with schizophrenia tended to have more severe psychotic symptoms and poorer social functioning, factors that may render them more vulnerable to insomnia³⁵. Together, these findings suggest potential gender-specific pathways linking psychopathology and sleep disturbance in schizophrenia. Of course, the gender differences in correlations observed in this study may also be attributable to insufficient statistical power due to the sample size; therefore, the interpretation of the findings should be approached with caution.

This study further highlighted a strong link between depressive symptoms and sleep disturbances in individuals with schizophrenia. This finding was consistent with numerous earlier investigations^29,36. For example, a survey by Abu Khait et al. demonstrated a significant positive correlation between the severity of depression and insomnia risk in patients with schizophrenia³⁶. This association can be interpreted from several interrelated perspectives. On one hand, shared neurobiological underpinnings may represent a core pathway for their co-occurrence. Both depressive symptoms and insomnia are closely linked to hyperactivity of the HPA axis, dysregulated functional connectivity between the prefrontal cortex and limbic system, and imbalances in key neurotransmitter systems, including serotonin, norepinephrine, and gamma-aminobutyric acid (GABA)³⁷. Given that patients with schizophrenia already exhibit baseline abnormalities in these systems, the emergence of depressive episodes may further exacerbate the vulnerability of the sleep-wake regulatory network. On the other hand, the psychological burden associated with depression, including feelings of hopelessness, self-devaluation, and emotional distress, can directly impair the ability to initiate and maintain sleep through mechanisms such as cognitive and emotional rumination³⁸. Notably, depressive symptoms serve not only as an indicator of overall illness severity but are also associated with increased psychotic symptom burden, higher relapse rates, and poorer functional outcomes³⁹. The resulting consequences, such as treatment nonadherence, social withdrawal, and heightened healthcare utilization, may act as persistent secondary contributors to chronic insomnia. Therefore, in clinical practice, sleep disturbances in schizophrenia patients with comorbid depressive symptoms should be carefully assessed and actively managed. Targeted interventions addressing both mood and sleep are warranted to improve overall quality of life and long-term prognosis.

In hematological parameters, elevated NLR was independently and positively associated with insomnia symptoms in patients with schizophrenia. Indeed, multiple studies have consistently demonstrated that sleep disturbances particularly insomnia are frequently accompanied by systemic low-grade inflammation and immune dysregulation^40,41. However, research investigating the associations between inflammatory biomarkers such as NLR, PLR, and MLR and insomnia remains in its infancy. For instance, a cross-sectional study reported significant correlations between sleep disturbances and leukocyte count, neutrophil count, NLR, and MLR in individuals with schizophrenia⁴¹. The potential links between insomnia symptoms and inflammatory markers such as NLR in schizophrenia may operate across multiple biological levels. However, it is important to note that current evidence supports only a bidirectional association between the two, rather than a unidirectional causal relationship. First, neuroinflammation may increase blood-brain barrier (BBB) permeability and directly disrupt key sleep-wake regulatory nuclei such as the suprachiasmatic nucleus (SCN) and the ventrolateral preoptic area (VLPO), either via vagal afferent signaling or through direct cytokine action, thereby contributing to insomnia⁴². Second, Capuron et al. proposed that inflammatory mediators may disrupt sleep homeostasis by altering the metabolism and availability of critical neurotransmitters, notably serotonin⁴³. Third, chronic inflammation may also promote hyperactivation of the HPA axis and downregulates the expression of core circadian clock genes (e.g., Clock), which is associated with sleep fragmentation and alterations in circadian phase⁴⁴. Notably, a complex bidirectional interplay exists between sleep disturbances and systemic inflammation. Insomnia is not only a potential consequence of inflammatory activation but can also exacerbate immune dysregulation by disrupting circadian rhythms, activating the HPA axis, and altering cytokine signaling⁴⁵. Multiple population-based studies have demonstrated that even partial nocturnal sleep deprivation can activate inflammatory signaling pathways, including those involving NF-κB and members of the signal transducer and activator of transcription (STAT) family^46,47. Furthermore, our study revealed that the association between NLR and insomnia was statistically significant only among male patients with schizophrenia, suggesting potential sex-specific differences in the inflammation-insomnia relationship. However, given the paucity of gender-stratified investigations in existing literature on schizophrenia populations, this observation remains exploratory and warrants rigorous validation in larger, prospective, multicenter cohort studies. Meanwhile, ROC curve analysis demonstrated that NLR combined with clinical symptoms had moderate predictive efficacy for insomnia in patients with schizophrenia. However, due to the lack of external validation and the potential risk of overfitting, this model should currently be regarded as exploratory and requires further verification. Collectively, these findings not only advance our understanding of the immunological underpinnings of insomnia comorbidity in schizophrenia but also provide novel theoretical foundations for early risk identification, personalized intervention strategies, and the clinical translation of targeted anti-inflammatory therapeutics. Nevertheless, future longitudinal studies are required to elucidate the specific causal pathways.

Regarding metabolic parameters, our study further revealed that lower T3 levels might be associated with a higher risk of insomnia. Song et al. demonstrated that individuals with overt or subclinical hypothyroidism exhibit significantly prolonged sleep latency, reduced total sleep time, and diminished subjective sleep quality compared with euthyroid controls⁴⁸. Indeed, as key endogenous regulators, thyroid hormones not only govern neurodevelopment and modulate higher-order brain functions⁴⁹, but also serve as core upstream controllers of the molecular mechanisms underlying circadian rhythms⁵⁰. Thyroid hormone receptors can directly bind to the promoter regions of core clock genes (such as Clock and Per), precisely regulating their transcriptional activity⁵⁰. Low levels of T3 may induce phase shifts in the expression rhythms of these clock genes, thereby impairing the function of SCN as the master pacemaker and ultimately being associated with disruptions in the sleep-wake cycle⁵¹. Of course, these associations require longitudinal validation. Furthermore, this circadian disruption engages in a bidirectional vicious cycle with monoaminergic neurotransmitter dysregulation. On one hand, intact circadian rhythms are a prerequisite for maintaining the optimal release of serotonin and norepinephrine within specific time windows; on the other hand, serotonin and norepinephrine themselves act as critical non-photic zeitgebers, responsible for transmitting arousal signals to the SCN to reset the biological clock^52–54. When thyroid hormone signaling is attenuated, it not only directly suppresses the efficiency of neurotransmitter synthesis but may also weakens the feedback regulatory capacity of neurotransmitters on the rhythmic system by disrupting the oscillatory properties of clock genes, thereby being associated with insomnia. Moreover, diminished thyroid hormone signaling has been implicated in the emergence of affective symptoms such as depression and anxiety, which are themselves robust predictors and perpetuating factors of insomnia⁵⁵. However, current research on the relationship between thyroid hormones and sleep disorders remains in its preliminary exploratory stage, and more in-depth mechanistic studies are urgently needed to fully elucidate the intrinsic connections between them. Additionally, the association was not statistically significant in the gender-stratified analysis, likely due to insufficient statistical power caused by small subgroup sample sizes. Future research with expanded cohorts is needed to verify this observation.

However, this study also had several limitations: (1) Due to its cross-sectional design, the study can only identify statistical associations among variables and cannot establish causal directionality between insomnia symptoms and inflammatory markers (e.g., NLR) or other clinical factors. Therefore, future longitudinal studies are urgently needed to further elucidate the temporal relationships and potential causal mechanisms among these variables. (2) Insomnia assessment relied on a standardized self-report questionnaire, which was inherently susceptible to subjective perception bias, recall error, and confounding effects of emotional states (e.g., comorbid anxiety or depression). (3) The sample was drawn exclusively from a single tertiary hospital in Anhui Province, and the overall sample size was relatively limited, thereby constraining both geographic representativeness and statistical power. In particular, smaller female sample (n = 81) may limit detection of additional associations, contributing to observed differences. (4) Given the absence of external validation, the predictive models in this study (e.g., those evaluated via ROC curves) may be susceptible to overfitting. (5) Although this study adjusted for chlorpromazine equivalent doses, different classes of antipsychotic medications may still exert significantly heterogeneous effects on sleep architecture, metabolic parameters, and inflammatory markers. Specifically, agents with potent histamine H1 receptor antagonistic activity (such as olanzapine and clozapine) are often associated with more pronounced sedative effects and more significant metabolic adverse events⁵⁶. Furthermore, regarding immunomodulation, emerging evidence suggests that certain second-generation antipsychotics may exert potential anti-inflammatory effects by modulating cytokine levels⁵⁷. In addition, our study also did not comprehensively document patients’ use of hypnotics, sedatives, or benzodiazepines, which may act as potential confounding factors. These medications not only directly alter sleep architecture, thereby interfering with insomnia assessment, but also modulate systemic inflammation levels, potentially introducing confounding effects on the study results⁵⁸. Therefore, future high-quality clinical studies are needed to address these limitations and further validate the findings of this study.

Conclusion

Patients with schizophrenia exhibited a relatively high risk of insomnia (29.1% in males vs. 42.0% in females); however, this difference was not statistically significant. This study suggested that insomnia symptoms were not isolated phenomena but were intricately linked to multiple domains, including psychotic symptoms, depressive mood, systemic inflammation (e.g., NLR), and thyroid function parameters (e.g., T3). Moreover, these associations may exhibit sex-specific patterns; for instance, psychotic symptoms and NLR were independently associated with insomnia only in male patients. Based on these findings, clinicians should prioritize routine monitoring and comprehensive assessment of sleep quality in individuals with schizophrenia. Furthermore, early intervention strategies should be proactively implemented for patients identified as having risk factors for insomnia.

Author contributions

HL and KZ: the study’s design. YS, LL, XZ, XS: collection, analyses and interpretation of the data. YS: drafting of the manuscript. HL and KZ: revision of the manuscript. The final version of the publication was approved by all authors.

Funding

Anhui Provincial Natural Science Foundation Project (No. 2108085MH275).

Data availability

The data supporting this study are available upon reasonable request from the corresponding author.

Competing interests

The authors declare no competing interests.

Ethics approval

The study protocol was approved by the Ethics Committee of the Fourth Affiliated Hospital of Anhui Medical University (approval number: KYXM-202210-015). All procedures were conducted in accordance with the ethical standards of the 2013 Helsinki Declaration (https://www.wma.net/policies-post/wma-declaration-of-helsinki/). Written informed consent was obtained from all participants and, where applicable, their legally authorized representatives.