The association between heart rate variability and inflammatory markers in first-episode depressed patients with different sleep quality

Abstract

Objective

To investigate the associations between systemic inflammatory markers and heart rate variability (HRV) parameters across different levels of sleep disturbance in patients with first-episode depression.

Methods

A total of 76 patients with first-episode moderate-to-severe depression and 30 healthy controls were enrolled. Patients were stratified into three groups based on Pittsburgh Sleep Quality Index (PSQI) scores: low (LSD), moderate (MSD), and severe (SSD) sleep disturbance. Peripheral inflammatory markers and 24-hour HRV parameters were measured. Partial Spearman correlation analyses were performed adjusting for age, sex, BMI, smoking status, and alcohol consumption.

Results

Significant group differences were observed in inflammatory markers (IL-6, IL-1β, TNF-α, CRP, ESR) and parasympathetic-related HRV indices (SDANN, SDNN, HF), with the SSD group exhibiting the most pronounced alterations (all FDR-adjusted p < 0.05). Moreover, significant negative correlations between inflammatory markers and HRV parameters were detected only in the SSD group after covariate adjustment.

Conclusion

The interaction between systemic inflammation and autonomic dysfunction appears to vary with sleep disturbance severity in depression. These findings highlight the importance of considering sleep disturbance stratification when evaluating physiological alterations in depressive disorders. Further studies are needed to clarify the underlying mechanisms and potential clinical implications of these associations.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12888-025-07256-7.

Introduction

Depression is a prevalent psychiatric disorder characterized by emotional, cognitive, and physiological impairments, among which sleep disturbance is one of the most common and persistent symptoms [1, 2]. In addition to being a hallmark feature of depression, sleep difficulties, including problems with sleep initiation, maintenance, and early awakening, may contribute to its pathophysiology through interactions with key physiological systems, particularly the autonomic nervous system (ANS) and the immune-inflammatory network [3, 4].

Heart rate variability (HRV), a noninvasive biomarker of ANS function, reflects the dynamic balance between sympathetic and parasympathetic activity [5]. Reduced HRV has been frequently reported in individuals with depression, indicating impaired autonomic flexibility [6]. Elevated levels of pro-inflammatory cytokines, such as IL-6, IL-1β, and TNF-α, further indicate immune dysregulation commonly observed in depression [7–9]. Previous studies have identified negative associations between autonomic dysfunction and systemic inflammation in depressive populations [10]. Inverse relationships between HRV parameters and inflammatory markers have been reported in individuals with depression and other stress-related conditions [11]. Relevent studies treated sleep disturbance as a covariate or employed global Pittsburgh Sleep Quality Index (PSQI) scores without stratifying by severity. The potential moderating role of sleep disturbance severity in these interactions remains underexplored.

In addition, prior research often regarded sleep quality as a secondary variable or assessed it using a single global PSQI score analyzed as a continuous or dichotomous measure [12, 13]. Although this approach facilitates general classification, it may overlook physiological heterogeneity among individuals with varying levels of sleep disturbance. Whether more severe forms of sleep disturbance in depression are associated with distinct patterns of autonomic and immune dysfunction remains unclear. Stratification based on sleep disturbance severity may provide a more refined framework for identifying dose–response relationships between sleep dysfunction, autonomic regulation, and systemic inflammation.

To address this gap, the present study stratified patients with first-episode depression into three groups based on PSQI scores: low sleep disturbance (LSD), moderate sleep disturbance (MSD), and severe sleep disturbance (SSD) [14, 15]. This stratified approach moves beyond traditional global scoring by modeling gradations of sleep disturbance and provides new perspectives on the physiological heterogeneity of depression. This stratification-based design addresses the limitations of previous methods and provides new insights into physiological heterogeneity that would otherwise be masked by global scoring approaches. The findings suggest that sleep disturbance severity may influence the strength of associations between HRV and inflammatory markers, with stronger correlations observed in patients with severe sleep disturbance. These results extend prior HRV and inflammation research and offer further insight into the interplay between sleep dysfunction, autonomic imbalance, and immune activation in depression. Such insights may contribute to the development of more personalized approaches for managing physiological comorbidities in this population.

Materials and methods

Participants

Participants who met the following criteria were included in the study: (1) aged 18–65 years; (2) diagnosed with first-episode moderate-to-severe depression based on DSM-5 criteria; (3) no prior use of antidepressants or sedative medications within the past two weeks; and (4) ability to provide informed consent.

Exclusion criteria included: (1) comorbid psychiatric disorders such as bipolar disorder or anxiety; (2) serious physical illnesses including cardiovascular, hepatic, renal, or endocrine diseases (e.g., diabetes); (3) recent use (within the past three months) of medications that could affect HRV or inflammatory markers, such as β-blockers or anti-inflammatory agents; (4) cognitive impairment (IQ ≤ 80); and (5) pregnancy or lactation.

To account for potential confounders, participants’ body mass index (BMI), smoking and alcohol use history were recorded and considered during analysis. A full description of the inclusion and exclusion criteria is provided in Supplementary material 1–1.

Clinical assessment

The 24-item Hamilton Depression Rating Scale (HAMD-24) was used to evaluate the severity of depressive symptoms. This clinician-administered scale assesses various symptom dimensions, including mood, somatic complaints, and cognitive function, based on both patient self-report and clinical interview.

The PSQI was used to assess sleep quality over the past month. The PSQI consists of seven components, including subjective sleep quality, sleep latency, and daytime dysfunction, with a total score ranging from 0 to 21. Higher scores indicate poorer sleep quality.

All assessments were conducted by trained psychiatrists following standardized scoring procedures to ensure consistency and accuracy. Detailed descriptions of the HAMD-24 and PSQI, including item composition, scoring criteria, and interpretation guidelines, are provided in Supplementary material 1–2 and 1–3, respectively. Higher scores indicate poorer sleep quality and were used to stratify participants into sleep disturbance groups, as detailed in the Statistical Analysis section.

Grouping based on PSQI

This study focused on first-episode, moderate-to-severe depressed patients (HAMD-24 score ≥ 20), a population known to frequently experience sleep disturbances with physiological implications. A healthy negative control (NC) group was included for comparison. Participant demographic characteristics are provided in Supplementary material 1–4.

To explore the associations between sleep quality, HRV, and inflammatory markers, depressed patients were classified into three groups based on PSQI scores: LSD (0–5), MSD (6–15), and SSD (16–21). This stratification was based on the initial PSQI validation study, which defined a score above 5 as indicative of clinically significant sleep disturbance [16]. The further subdivisions into low, moderate, and severe levels were informed by thresholds that have been frequently applied in prior clinical and population-based studies, supporting the clinical relevance and comparability of these categories [17]. A data-driven clustering approach was initially considered; however, the limited sample size precluded stable solutions. Therefore, we adopted a literature-supported and clinically interpretable stratification scheme to enhance analytical sensitivity and facilitate cross-study comparability. Based on PSQI scores, 20 patients were classified as LSD, 35 as MSD, and 23 as SSD.

HRV measurement

HRV data were collected using a 24-hour dynamic 18-lead electrocardiogram (ECG) system (Pengyang Manufacturer, model PENGYANG-K18, sampling frequency 500 Hz). Each participant was fitted with nine Ag/AgCl electrodes placed on the chest and limbs, and ECG signals were continuously recorded during both daily activities and rest periods.

Raw signals were processed using dedicated ECG analysis software to remove artifacts and irregularities. The following HRV parameters were extracted to evaluate ANS function: low-frequency power (LF), high-frequency power (HF), LF/HF ratio, standard deviation of NN intervals (SDNN), standard deviation of the average NN intervals (SDANN), SDNN index, root mean square of successive differences (rMSSD), percentage of successive NN intervals > 50 ms (pNN50), triangular index (TRIIDX), minimum heart rate (Min HR), maximum heart rate (Max HR), and mean heart rate (Avg HR).

Inflammatory marker analysis

Peripheral venous blood samples were collected after a 12-hour overnight fast. Trained nurses conducted blood collection using standard procedures. For each participant, 5 mL of blood was drawn from the antecubital vein, serum was separated, and samples were stored at − 20℃ until analysis.

Routine hematological parameters, including white blood cell (WBC) count, neutrophil count (NEYT), lymphocyte count (LYMPH), C-reactive protein (CRP), and erythrocyte sedimentation rate (ESR), were measured using automated analyzers (Mindray 7500, Beckman ACCESS). Serum cytokines such as IL-1β, interleukin-2 receptor (IL-2R), IL-6, interleukin-8 (IL-8), interleukin-10 (IL-10), and TNF-α were quantified using enzyme-linked immunosorbent assay kits (Wovent, China) in accordance with the manufacturer’s instructions.

Statistical analysis

Given the exploratory nature of this study, no formal a priori power analysis was performed. The sample size was determined by the number of eligible patients enrolled during the study period, based on feasibility considerations, and was comparable to previous studies investigating HRV and inflammatory markers in depression [18, 19].

However, we acknowledge that some subgroup comparisons—particularly those involving stratified sleep disturbance severity—may be underpowered due to sample size constraints. Thus, the interpretation of marginally significant results should be made cautiously, and future studies with larger samples are warranted to validate these findings.

All statistical analyses were performed using GraphPad Prism (version 10.2.3) and SPSS. The Shapiro–Wilk test was used to assess the normality of continuous variables, and Levene’s test was used to evaluate the homogeneity of variances.Continuous variables were summarized as mean ± standard deviation (SD) or median (interquartile range), as appropriate.

Categorical variables were compared using the chi-squared (χ²) test. Differences in HRV parameters and inflammatory markers across sleep disturbance groups were evaluated using analysis of covariance (ANCOVA), adjusting for age, sex, BMI, smoking status, and alcohol consumption. Adjusted means and 95% confidence intervals (CIs) were reported, and post hoc pairwise comparisons were conducted using Bonferroni correction.

Partial Spearman correlation analysis was conducted to examine associations between inflammatory markers and HRV parameters, with adjustment for the same covariates. In total, 108 pairwise correlations between HRV parameters (n = 12) and inflammatory markers (n = 11) were tested. The Benjamini–Hochberg procedure was applied to control the false discovery rate (FDR) at 5%, following recommendations for exploratory physiological studies involving correlated outcomes. An effect size of |r| ≥ 0.5 and FDR-adjusted p < 0.05 were considered statistically significant. Bootstrap resampling (1,000 iterations) was performed to evaluate the robustness of the correlation results.

Results

Baseline characteristics and covariate comparisons

Partial Spearman correlation analyses between HRV parameters and inflammatory markers were conducted for all sleep disturbance groups (LSD, MSD, SSD), adjusting for age, sex, BMI, smoking status, and alcohol consumption. The objective was to evaluate whether the associations between HRV and inflammatory markers remained significant after controlling for potential confounding factors.

Following covariate adjustment, the strength and patterns of associations varied across groups. In the LSD and MSD groups, only a limited number of significant correlations were observed, and no consistent trends were identified across HRV or inflammatory markers (Supplementary material 2 − 1).

In contrast, the SSD group demonstrated a greater number of significant associations (Fig. 1). Notably, IL-6 was negatively correlated with SDANN (r = −0.52) and TRIIDX (r = −0.46), while IL-1β and IL-8 showed negative correlations with LF (r = −0.51 and r = −0.57, respectively). Positive correlations were observed between CRP and TRIIDX (r = 0.51), and between IL-1β and TRIIDX (r = 0.48). These associations persisted after covariate adjustment, suggesting that the observed relationships were not attributable to confounding by age, body composition, or lifestyle factors.

Fig. 1.

Covariate-adjusted partial correlations between HRV parameters and inflammatory markers in the SSD group. Partial Spearman correlation analyses were performed adjusting for age, sex, BMI, smoking status, and alcohol consumption. The color scale indicates the strength and direction of correlations (blue for negative, red for positive). Significant correlations are marked with (FDR-adjusted p < 0.05)

Sensitivity analysis

To evaluate the robustness of the observed associations, a sensitivity analysis was conducted using multiple PSQI grouping strategies, including Original, WideGroup1, WideGroup2, Tertile, and Dichotomy schemes (Fig. 2). The number of significant HRV–inflammation pairs, sample balance, and structural stability were compared across grouping methods (Supplementary material 2–2).

Fig. 2.

Sensitivity analysis of significant HRV–inflammation pairs across different PSQI grouping strategies. The number of significant HRV–inflammation pairs (|r| ≥ 0.5, FDR-adjusted p < 0.05) was compared across five PSQI grouping strategies: Original, WideGroup1, WideGroup2, Tertile, and Dichotomy

The Original grouping strategy yielded 17 significant HRV–inflammation pairs, with relatively balanced sample distribution and stable structural characteristics(Table 1). In comparison, WideGroup1 and WideGroup2 produced a larger number of significant pairs (30 and 26, respectively), but exhibited poorer sample balance and greater variability in structural stability. The Tertile and Dichotomy groupings resulted in fewer significant pairs (12 and 18, respectively), with inconsistent structural properties. These observations indicate that the Original grouping strategy provided a more stable and balanced framework for analyzing the associations of interest, thereby supporting the reliability of the main results under this grouping scheme.

Table 1.

Summary of sensitivity analysis across five sleep quality grouping strategies

Grouping Strategy	Significant Pairs	Sample Balance	Structure Stability	Recommendation
Original	17	Good	Stable	Recommend
WideGroup1	30	Poor	Unstable	Not Recommend
WideGroup2	26	Poor	Unstable	Not Recommend
Tertile	12	Good	Stable	Recommend
Dichotomy	18	Poor	Variable	Conditional

Inflammatory markers assessment

Inflammatory marker levels were compared across four groups: NC, LSD, MSD, and SSD. ANCOVA was performed for each inflammatory marker, adjusting for age, sex, BMI, smoking status, and alcohol consumption.

Significant group differences were observed for IL-6, IL-1β, TNF-α, CRP, ESR, NEYT, LYMPH, and WBC (FDR-adjusted p < 0.05). Compared with the NC group, patients with SSD exhibited elevated levels of IL-6, IL-1β, TNF-α, CRP, and ESR (all p < 0.01), indicating heightened systemic inflammatory activity. The LSD and MSD groups showed intermediate levels, with a trend toward increasing inflammatory markers with greater severity of sleep disturbance.

In contrast, no significant group differences were found for IL-2R and IL-10 after covariate adjustment. Six representative inflammatory markers (NEY, LYMPH, TNF-α, IL-1β, WBC, IL-6), which best reflected the differences across groups, are presented in Fig. 3. Detailed results for the remaining markers are provided in Supplementary material 2–3.

Fig. 3.

Differential analysis of inflammatory markers across four groups: NC, LSD, MSD, and SSD. Bar plots show levels of (A) NEYT, (B) LYMPH, (C) TNF-α, (D) IL-1β, (E) WBC, and (F) IL-6. Data are shown as mean ± standard error of the mean (SEM). Statistical comparisons were performed using ANCOVA with Bonferroni correction. (*p < 0.05 was considered statistically significant; **p < 0.01 indicated a significant difference; ***p < 0.001 denoted a highly significant difference; ****p < 0.0001 indicated an extremely significant difference)

Heart rate variability assessment

HRV parameters were compared across four groups (NC, LSD, MSD, SSD). ANCOVA was performed for each HRV index, adjusting for age, sex, BMI, smoking status, and alcohol consumption. Post hoc pairwise comparisons were performed with Bonferroni correction.

Significant group differences were observed in SDANN, SDNN, SDNN Index, LF, LF/HF ratio, and HF (FDR-adjusted p < 0.05). Compared with the NC group, patients with SSD exhibited markedly reduced SDANN, SDNN, SDNN Index, and HF (all p < 0.001), alongside elevated LF and LF/HF ratio, indicating a shift toward sympathetic predominance and diminished parasympathetic modulation.

Patients in the LSD and MSD groups showed intermediate alterations: both groups displayed significant reductions in SDANN, SDNN, SDNN Index, and HF (p < 0.001 or p < 0.01 vs. NC), with elevated LF and LF/HF ratio particularly evident in the SSD group. These findings suggest a graded disruption of autonomic balance with increasing sleep disturbance severity.

Six representative HRV parameters (SDANN, SDNN, SDNN Index, LF, LF/HF ratio, and HF), reflecting both sympathetic-parasympathetic modulation and overall variability, were selected for presentation in the main text (Fig. 4). Detailed results of remaining HRV indices are provided in Supplementary material 2–4.

Fig. 4.

Comparison of HRV parameters across four groups (NC, LSD, MSD, SSD). Bar plots display (A) SDANN, (B) SDNN, (C) SDNN Index, (D) LF, (E) LF/HF ratio, and (F) HF. Data are presented as mean ± SEM. Statistical comparisons were performed using ANCOVA with Bonferroni correction

Correlation analysis

Partial Spearman correlation analysis was performed between inflammatory markers and HRV parameters, adjusting for age, sex, BMI, smoking status, and alcohol consumption.

In the SSD group, several significant negative correlations were observed (FDR-adjusted p < 0.05, |r| ≥ 0.5), primarily between elevated inflammatory markers (IL-6, IL-1β, TNF-α, CRP, ESR) and reduced parasympathetic-related HRV indices (SDANN, SDNN, HF) (Fig. 5). Notably, higher IL-6 and IL-1β levels were associated with lower SDANN and SDNN, while increased TNF-α and CRP levels correlated with diminished HF. These findings indicate a statistical association between heightened systemic inflammation and impaired autonomic regulation in patients with severe sleep disturbance, which may warrant further investigation. In contrast, no significant correlations were detected in the LSD and MSD groups after covariate adjustment (Supplementary material 2–5).

Fig. 5.

Partial Spearman correlation heatmap between inflammatory markers and HRV parameters in the SSD group (n = 23). Only significant correlations (FDR-adjusted p < 0.05, |r| ≥ 0.5) are shown. Color scale indicates direction and strength of correlation (red = positive, blue = negative)

Discussion

This study demonstrated that systemic inflammatory activity and ANS function exhibit graded alterations in patients with first-episode depression according to the severity of sleep disturbance. The most pronounced changes, including elevated inflammatory markers (IL-6, IL-1β, TNF-α, CRP, ESR) and reduced parasympathetic-related HRV parameters, were observed in patients with SSD. Notably, significant negative correlations between inflammatory markers and HRV parameters were predominantly detected in the SSD group, suggesting that immune–autonomic dysregulation may be more prominent when sleep disturbance becomes more severe. These findings indicate that stratifying depressive patients by sleep quality may help identify patterns of physiological variability, although further studies are required to determine the clinical relevance of this approach.

Our results are consistent with prior studies reporting autonomic imbalance and heightened inflammation in major depressive disorder [20, 21]. Reduced heart rate variability (HRV), particularly in relation to diminished parasympathetic modulation as indicated by lower HF and SDNN values, has been linked to poor prognosis in depression [22]. In parallel, elevated levels of IL-6, TNF-α, and CRP have been established as peripheral biomarkers of immune dysregulation [23, 24]. However, most previous research treated sleep disturbance as a secondary factor. In contrast, our stratified approach based on sleep severity revealed distinct patterns of HRV–inflammation associations across different levels of sleep dysfunction. This aligns with emerging evidence suggesting that sleep disturbances are associated with neuroimmune alterations in depression [25].

The mechanisms driving these associations are likely multifactorial. Previous studies suggest that elevated levels of pro-inflammatory cytokines (e.g., IL-6, TNF-α) may be associated with reduced vagal tone and parasympathetic activity, possibly through pathways such as the cholinergic anti-inflammatory reflex [26, 27]. Conversely, reduced vagal signaling has also been associated with elevated inflammatory markers, suggesting a potential bidirectional relationship. In addition, poor sleep quality has been linked to increased activity of the hypothalamic–pituitary–adrenal axis and elevated sympathetic arousal [28], which may contribute to immune-autonomic imbalance. These observations may help explain the stronger immune-autonomic associations found in the SSD group; however, the cross-sectional design of our study precludes conclusions about causality. Instead, the findings suggest that sleep disturbance severity may serve as an indicator of physiological dysregulation in depression.

Additionally, our findings emphasize the potential heterogeneity in neuroimmune dysregulation within the depressed population. The lack of consistent associations in the LSD and MSD groups, contrasted with the pronounced immune-autonomic correlations in the SSD group, suggests that sleep disturbance severity may delineate a biologically distinct subtype of depression. This supports the growing interest in biomarker-informed subtyping and precision psychiatry, wherein sleep-related physiological signatures could aid in identifying more vulnerable individuals.

From a clinical perspective, incorporating sleep quality assessments into routine psychiatric evaluation may not only aid in symptom management but also serve as a proxy for underlying physiological dysregulation. Patients with severe sleep disturbance may benefit from tailored therapeutic strategies targeting both sleep and inflammation, such as cognitive behavioral therapy for insomnia, vagus nerve stimulation, or anti-inflammatory adjunct treatments. These targeted interventions could potentially improve both mood and somatic outcomes, although this remains to be confirmed in interventional studies.

This study has several limitations. First, its cross-sectional design precludes causal inference regarding the directionality of HRV–inflammation interactions. Second, although PSQI stratification was based on validated thresholds [29, 30], the use of self-reported sleep measures may introduce subjectivity and recall bias. Third, although the overall sample size was comparable to prior studies, some subgroup comparisons—especially in stratified analyses—may be underpowered, potentially limiting the ability to detect more subtle associations. Finally, unmeasured confounding factors (e.g., physical activity, dietary patterns, subclinical anxiety) could not be entirely ruled out.

Future research should adopt longitudinal and interventional designs to examine whether improving sleep quality can normalize autonomic and immune function in depression. In addition, mechanistic studies integrating neuroimaging, metabolomics, and gut microbiome assessments may help elucidate the bidirectional pathways linking sleep, ANS dysfunction, and immune dysregulation. Clinically, integrating sleep-focused and inflammation-informed approaches may contribute to improved management of physiological comorbidities in depression, pending confirmation in longitudinal and interventional studies.

Conclusion

The severity of sleep disturbance appears to influence the relationship between systemic inflammation and autonomic dysfunction in first-episode depression. Incorporating sleep quality stratification into clinical assessments may facilitate a more precise characterization of physiological heterogeneity in depressive disorders and support the development of targeted treatment approaches.

Abbreviations

ANS

Autonomic Nervous System

Avg HR

Average Heart Rate

ANCOVA

Analysis of Covariance

BMI

Body Mass Index

CRP

C-Reactive Protein

ECG

Electrocardiogram

ESR

Erythrocyte Sedimentation Rate

FDR

False Discovery Rate

HRV

Heart Rate Variability

HAMD-24

24-item Hamilton Depression Rating Scale

HF

High-Frequency Power

IL

Interleukin

IL-1β

Interleukin-1 beta

IL-2R

Interleukin-2 Receptor

IL-6

Interleukin-6

IL-8

Interleukin-8

IL-10

Interleukin-10

LSD

Low Sleep Disturbance

LF

Low-Frequency Power

LYMPH

Lymphocyte Count

MSD

Moderate Sleep Disturbance

Min HR

Minimum Heart Rate

Max HR

Maximum Heart Rate

NEYT

Neutrophil Count

NC

Negative Control

PSQI

Pittsburgh Sleep Quality Index

pNN50

Percentage of Successive NN Intervals > 50 ms

rMSSD

Root Mean Square of Successive Differences

SSD

Severe Sleep Disturbance

SDNN

Standard Deviation of NN Intervals

SDANN

Standard Deviation of the Average NN Intervals

TRIIDX

Triangular Index

TNF-α

Tumor Necrosis Factor-alpha

WBC

White Blood Cell

Authors’ contributions

Author contribution statement Study Concept and Design: A.A., Y.Z.Data Collection and Clinical Assessments: J.Y., R.M.Data Analysis and Interpretation: A.A., R.M.Statistical Analysis: A.A., M.M.Manuscript Drafting: A.A., R.M.Manuscript Revision and Critical Review: Y.Z.Final approval of the manuscript: all authors.

Funding

This study did not receive any specific funding.

Data availability

Data is provided within the manuscript or supplementary information files.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of the People’s Hospital of Xinjiang Uygur Autonomous Region (Approval number: KY2024013032). All participants provided written informed consent prior to enrollment. The study was conducted in accordance with the ethical standards of the institutional research committee and the principles of the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.