Association between neutrophil-to-lymphocyte ratio and specific depressive symptoms: an analysis of a population-based cross-sectional survey

Abstract

Background

The exact mechanisms underlying depression are not well understood. Chronic, low-grade inflammation is believed to play an important role in its development. The present study investigates the potential association between depressive symptoms and the neutrophil-to-lymphocyte ratio (NLR).

Methods

Seven data cycles of the National Health and Nutrition Examination Survey were extracted. Multivariable logistic regression and a generalized additive model were employed to determine the association.

Results

Thirty thousand eight hundred ninety-six subjects were analyzed. The results indicated that anhedonia and fatigue were significantly associated with NLR. Additionally, the generalized additive model results indicated a non-linear relationship between anhedonia, sleep disturbance and NLR. Subgroup analyses demonstrated that the correlation between anhedonia and NLR was significant in the above-60-year-old group (OR: 1.63, 95% CI: 1.14–2.33) and the male group (OR: 1.50, 95% CI: 1.07–2.10). Sleep disturbance was associated with NLR in the female group (OR: 1.36, 95% CI: 1.04–1.77). Fatigue was associated with NLR (OR: 1.30, 95% CI: 1.02–1.67) in the female group, as was the case in the non-Hispanic White group (OR: 1.32, 95% CI: 1.02–1.70).

Conclusions

There were associations between NLR and specific symptoms, and these associations varied across demographic subgroups. There was a non-linear association between anhedonia, sleep disturbance and NLR. These findings could potentially contribute to the advancement of precision medicine within the field of mental health.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12888-024-06214-z.

Introduction

Depression is a widespread and debilitating mental disorder that affects millions globally with its multifaceted symptoms [1], and there are not yet fully elucidated mechanisms underlying its development [2, 3]. Recent advances have demonstrated the pivotal role of neuroinflammation in the pathophysiology of depression [4–6]. Theories propose that chronic low-grade inflammation can disrupt neurotransmitter systems [7], impair neuroplasticity [3, 8], and disrupt mood regulation [9], all of which could contribute to the onset of depressive symptoms.

Previous studies have examined the association between inflammatory markers such as cytokines and depression [10–13]. These studies, which primarily focused on DSM-based categorical definitions of depression, demonstrated increased levels of pro-inflammatory cytokines in people diagnosed with depression. Despite providing valuable insights, it primarily summarized depression as a homogeneous entity and ignored the heterogeneity of symptomatology that characterizes the patient experience. According to the diagnostic criteria for depression in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) [14], there were 227 possible symptom combinations [15]. However, inflammatory depression may represent only a tiny proportion or be restricted to specific symptoms or characteristic populations [10, 16–18]. Given the heterogeneity of depressive symptoms, there is an urgent need to shift the research perspective toward understanding inflammation in relation to specific symptoms rather than examining it solely through diagnostic categories [17, 19]. This shift in perspective recognizes how complicated depression is at a clinical level and aims to uncover how neuroinflammation and symptoms manifest in complex ways. This could lead to more personalized and targeted therapeutic interventions.

Preliminary studies of specific inflammatory markers such as C-reactive protein (CRP), interleukin-6 (IL-6), and tumor necrosis factor-alpha (TNF-α) have indicated potential associations with specific depressive symptoms. For example, Amy T. Peters found that IL-6 correlates with reported depressed mood and TNF-α correlates with somatic symptoms in adolescence [20]. Markus Jokela and Philipp Frank observed that CRP is associated with sickness behavior (e.g., fatigue, lack of energy, sleep disorders, and appetite changes) [21, 22]. Daniel P. Moriarity found that CRP was associated with both appetite and fatigue [23], while Yuri Milaneschi, in a study from NESDA, found that IL-6 was additionally associated with anhedonia, sleep, and reduced appetite [24].

Pro-inflammatory cytokines are known for their ability to disrupt neurotransmitter synthesis, release, and receptor function within the brain, which can precipitate depressive symptoms including anhedonia, sleep disturbances, and fatigue. Alternatively, inflammation can alter the activity of the hypothalamic–pituitary–adrenal (HPA) axis, potentially inducing depressive symptoms. Neutrophil-to-lymphocyte ratio (NLR), a readily available hematological parameter, has attracted considerable attention as a surrogate marker for systemic low-grade inflammation [25–28]. NLR reflects the balance between pro-inflammatory neutrophils and anti-inflammatory or immunoregulatory lymphocytes. Despite its increasing recognition in various health conditions, including cardiovascular disease and certain cancers, its application in the field of mental health remains poorly understood, particularly in the context of the association between NLR and the heterogeneous symptomatology of depression.

In the present study, we postulated the existence of complicated relationships between NLR and specific symptoms. The purpose of this study is to examine the association between NLR and specific depressive symptoms and symptom dimensions in individuals diagnosed with depression using a symptom-focused approach. Understanding the connection between inflammation and the different manifestations of depression would open opportunities for precision medicine strategies that improve patient outcomes and a personalized medicine approach to mental health care.

Material and methods

Data source

The National Health and Nutrition Examination Survey (NHANES) is a cross-sectional program of studies designed to assess the health and nutritional status of the population in the United States [29]. NHANES was initiated in the early 1960s and is currently administered by the National Center for Health Statistics (NCHS) of the Centers for Disease Control and Prevention (CDC). NHANES combines interviews, physical examinations, and laboratory tests to provide a comprehensive perspective on public health. It is a nationwide survey with a complex and multistage probability sampling design. CDC conducted a continuous, annual survey, collecting and releasing data for public use at two-year intervals. Seven data cycles (from 2005 to 2018) in the study were extracted based on the availability of relevant variables for our research objectives. Data cleaning procedures were implemented to address missing values, outliers, and inconsistencies, using manual review and statistical techniques where appropriate. The National Center for Health Statistics Research Ethics Review Board approved the NHANES protocols, and all subjects signed informed consent (Protocol #2005–06, Protocol #2011–17) (https://www.cdc.gov/nchs/nhanes/irba98.htm).

Participants

The inclusion criteria were as follows: (1) participants must be over 18 years old; (2) participants must have a complete mental health screening and a complete blood count. The exclusion criteria were as follows: (1) those who were currently suffering from acute inflammation (a C-reactive protein level ≥ 10 mg/L); (2) a history of cancer or malignancy.

Depression measurement

The Patient Health Questionnaire (PHQ-9) was employed to assess depressive symptoms and dimensions. PHQ-9 is a self-administered questionnaire based on the nine signs and symptoms of depression in the DSM-V, namely “Have little interest in doing things?” (P1: anhedonia), “Feeling down, depressed, or hopeless?” (P2: depressed mood), “Trouble sleeping or sleeping too much?” (P3: sleep disturbance), “Feeling tired or having little energy?”(P4: fatigue), “Poor appetite or overeating” (P5: Appetite change), “Feeling bad about yourself”(P6: low self-esteem), “Trouble concentrating on things”(P7: concentration problems), “Moving or speaking slowly or too fast” (P8: psychomotor changes), and “ Thought you would be better off dead” (P9: suicidal ideation) [30]. Each symptom item of PHQ-9 represents a specific depressive symptom and is rated on a scale from “0” (not at all) to “3” (nearly every day). If the item was scored 2 or above, the symptom was considered to be present [22]. Each item or symptom dimension of PHQ-9 is considered a separate phenotype. The cluster of depression symptoms with similar characteristics is called the depression dimension, which reflects different aspects of depression, such as cognitive dimension and somatic symptom dimension.

Neutrophil to lymphocyte ratio (NLR)

Routine complete blood counts were performed in duplicate and averaged across all survey participants with the Coulter HMX on the NHANES MEC. A complete blood count with a 5-part differential was used to analyze neutrophil and lymphocyte counts. NLR was calculated as the ratio of the absolute neutrophil count to the absolute lymphocyte count.

Other variables

Following previous literature, the present study selected demographic, medical, and lifestyle variables as covariates. Specifically, age, race/ethnicity, sex, education level (below high school, high school, and above high school), marital status, body mass index (BMI), medical conditions, smoking history, drinking history, and antidepressant use were included. We categorized Mexican American/other Hispanics (the reference group), non-Hispanic whites, non-Hispanic blacks, and other races based on race/ethnicity. The female group served as the reference. There were three categories for marital status: married/living with a partner (reference group), single, and divorced/separated/widowed. BMI was calculated by dividing weight in kilograms by height in meters squared. Medical conditions were computed as the sum of chronic diseases (Supplementary Table S1). According to the CDC definition, participants were divided into three groups: current smokers, former smokers, and never smokers. An individual classified as a never smoker is someone who has not engaged in smoking throughout their adulthood or has smoked fewer than 100 cigarettes cumulatively. Conversely, a former smoker is defined as an adult who has exceeded the 100 cigarettes threshold in their lifetime but had abstained from smoking at the time of the survey. A current smoker, on the other hand, is defined as an adult who has smoked at least 100 cigarettes in total and continues to smoke actively. Excessive alcohol consumption is defined as the average consumption of five drinks per day in the past year. Information on antidepressant use was collected during the household interview by trained interviewers using computer-assisted face-to-face interviews. Antidepressant use was defined as at least one prescription antidepressant.

Statistical methods

To address the oversampling of minorities, sample weights, strata, and primary sampling units (PSUs) were taken into account following the guidelines as NHANES employed a complex stratified multistage probability sampling design. Before analysis, all datasets were visualized using boxplots for outliers. All values at three standard deviations (SD) were identified and replaced with the mean plus 3 SD. The winsorization threshold for NLR was 5.58. For continuous variables, data was presented as a mean (standard error, SE) and/or median (interquartile range, IQR). Data were described as frequencies and percentages for categorical variables. To evaluate the factor structure of the PHQ-9, confirmatory factor analyses (CFAs) were implemented utilizing structural equation modeling (SEM) employing a weighted least squares (WLS) estimation method. The fit for a one-factor model and a two-factor model with “non-somatic” and “somatic” factors was compared using the comparative fit index (CFI) and root-mean-square error of approximation (RMSEA). An excellent fit is indicated by CFI values of 0.95 and RMSEA values of 0.05. And CFI > 0.90 and RMSEA = 0.05–0.10 are considered satisfactory fits. In addition to the above indicators, the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) were compared between different structure models. We implemented the variance inflation factors (VIF) to assess the multicollinearity of all variables. The VIF cut-off value was set at 10 and no multicollinearity was observed. Multivariable logistic regression was utilized to model the associations between NLR and specific symptoms. Four models were constructed. In the crude model, no covariate was adjusted. Age, sex, race, education, and marital status were adjusted in model I. BMI and medical conditions were further adjusted in model II, and all covariates were adjusted in model III. The generalized additive model (GAM) was used to examine the non-linear relationship between NLR and depressive symptoms. Statistical interactions and subgroup analyses were performed using stratified multivariable regression analysis to determine the relationship between NLR and specific symptoms.

All analyses were conducted using R Statistical Software Version 4.2.3 (R Foundation for Statistics, http://www.r-project.org). Statistical significance was determined at a threshold of 0.05 for all two-tailed statistical tests.

Results

Weighted populations characteristics

A total of 70,190 people participated in the cross-sectional survey from 2005 to 2018. Of those, 42,143 participants were over 18 years old. 1511 were excluded without a complete blood count, and 5884 were excluded without a mental health screening. 3852 were excluded because of acute inflammation (n = 843) or cancer (n = 3009). Finally, 30,896 subjects were included in the analysis (Fig. 1). The median age was 44.00 years, with an interquartile range of 31.00 to 57.00 years. 15,497 (50.09%) were female, 8390 (15.04%) were Mexican American/Other Hispanic, 12,363 (66.11%) were non-Hispanic white, 6717 (11.24%) were non-Hispanic black, and 3426 (7.61%) were other races. Medical conditions were 1.26 (0.02). Table 1 lists the characteristics of the participants. Supplementary Table S2 lists the frequencies of symptom occurrences.

Fig.1.

Flow chart of study

Table 1.

Weighted Baseline characteristics of participants in NHANES 2005–2018

Characteristic	No. (%) of Participantsa
Age, median (IQR), years	44.00(31.00, 57.00)
18 to ≤ 39	11,909(39.38)
40 to ≤ 59	9877(37.31)
60 to ≤ 79	7615(19.99)
≥ 80	1495(3.32)
Sex
Male	15,399(49.91)
Female	15,497(50.09)
Race/ethnicity
Mexican American/ Other Hispanic	8390(15.04)
Non-Hispanic White	12,363(66.11)
Non-Hispanic Black	6717(11.24)
Other race	3426(7.61)
Education
Below high school	7148(15.91)
High school	6668(23.54)
Above high school	15,124(60.55)
Marital
Married/Living with partner	17,409(63.39)
Never married	5926(19.27)
Divorced/Separated/ Widowed	6083(17.34)
BMI, mean (SE), kg/m**2
< 18.5	538(1.68)
18.5 to 25	8781(29.27)
25 to 30	10,034(32.87)
> 30	11,259(36.18)
Medical conditions	1.26(0.02)
Excessive alcohol use
Yes	4248(15.77)
No	21,203(84.23)
Smoking status
Never	16,791(56.17)
Former	6696(23.32)
Current	6155(20.51)
Antidepressant
Yes	2956(11.88)
No	27,940(88.12)
Anti-inflammatory drugs
Yes	1527(4.36)
No	29,369(95.64)
NLR, mean (SE)	2.14(0.01)

Abbreviation: IQR Interquartile range, SE Standard error, BMI Body mass index, NLR Neutrophil to lymphocyte ratio

^aData are presented as unweighted number of participants and weighted percentages, or weighted mean and weighted standard error, weighted to be nationally representative

Confirmatory factor analyses

CFAs of the factor structure showed that both models gave a satisfactory fit. Based on the CFA results, a two-factor model with “non-somatic” and “somatic” showed the best fit (Supplementary Table S3) for model comparison [31, 32].

Association between NLR and specific depressive symptoms

Table 2 presents the associations between NLR and specific depressive symptoms (dichotomous variables). In the crude model, anhedonia, depressed mood, sleep disturbance, and fatigue were significantly associated with NLR. After adjustment for age, sex, race, education, and marital status, anhedonia, depressed mood, sleep disturbance, fatigue, low self-esteem, and concentration problems were associated with NLR in model I. Similarly, anhedonia, depressed mood, sleep disturbance, and fatigue were associated with NLR after adjustment for age, sex, race, education, marital status, BMI, and medical conditions. However, only anhedonia and fatigue were associated with NLR after full adjustment.

Table 2.

Multivariable logistic regression analyses for NLR and specific depression symptoms, Weighted

	Crude model		Model I		Model II		Model III
	OR	p	OR	p	OR	p	OR	p
P1	1.34(1.11, 1.63)	0.003	1.46(1.19, 1.79)	< 0.001	1.35(1.10,1.65)	0.004	1.29(1.04,1.61)	0.02
P2	1.33(1.07, 1.66)	0.01	1.43(1.13, 1.80)	0.003	1.31(1.04,1.66)	0.03	1.22(0.94,1.59)	0.14
P3	1.36(1.15, 1.59)	< 0.001	1.34(1.12, 1.61)	0.002	1.23(1.03,1.46)	0.02	1.20(0.99,1.44)	0.06
P4	1.41(1.22,1.63)	< 0.001	1.42(1.21, 1.67)	< 0.001	1.30(1.10,1.52)	0.002	1.25(1.03,1.51)	0.02
P5	1.07(0.90,1.27)	0.46	1.20(0.99, 1.45)	0.07	1.05(0.86,1.28)	0.63	1.04(0.84,1.29)	0.72
P6	1.21(0.96,1.51)	0.10	1.27(1.01, 1.61)	0.04	1.15(0.90,1.47)	0.26	1.08(0.83,1.40)	0.57
P7	1.22(1.00,1.50)	0.051	1.30(1.05, 1.60)	0.02	1.21(0.98,1.48)	0.07	1.14(0.91,1.43)	0.25
P8	1.01(0.76,1.35)	0.93	1.07(0.79, 1.45)	0.64	0.97(0.73,1.30)	0.85	0.93(0.67,1.28)	0.65
P9	1.13(0.63,2.04)	0.69	1.17(0.66, 2.07)	0.59	1.09(0.60,1.95)	0.79	1.09(0.57,2.08)	0.81

Crude model: unadjusted. Adjust model I: adjusted for age, sex, race, education, and marital status. Adjust model II: adjusted for age, sex, race, education, marital status, BMI, and medical conditions. Adjusted model III: age, sex, race, education, marital status, BMI, medical conditions, smoking status, alcohol use, antidepressant and anti-inflammatory drug use

Abbreviation: NLR Neutrophil to lymphocyte ratio, OR odds Ratio, CI Confidence interval

Association between NLR and symptom dimensions

In the crude model, NLR was associated with both symptom dimensions. The same trends were found in models I and II. However, after controlling for all confounders, neither dimension was associated with NLR. Table 3 displays the results.

Table 3.

Multivariable logistic regression analyses for NLR and symptom dimensions

	Crude model		Model I		Model II		Model III
	β	p	β	p	β	p	β	p
Nonsomatic dimension	0.20(0.06,0.33)	< 0.001	0.26(0.12,0.39)	< 0.001	0.18(0.04,0.31)	0.01	0.12(-0.04,0.27)	0.14
Somatic dimension	0.24(0.13,0.35)	< 0.001	0.27(0.14,0.39)	< 0.001	0.16(0.04,0.28)	0.01	0.13(-0.01,0.26)	0.06

Crude model: unadjusted. Adjust model I: adjusted for age, sex, race, education, marital status. Adjust model II: adjusted for age, sex, race, education, marital status, BMI, medical conditions. Adjusted model III: age, sex, race, education, marital status, BMI, medical conditions, smoking status, alcohol use, antidepressant and anti-inflammatory drug use

Non-linear relationship between NLR and depressive symptoms

The generalized additive model was employed to graphically examine the potential associations between NLR and specific depressive symptoms. The results of GAM estimation indicated that a non-linear relationship was observed between NLR and anhedonia and sleep disturbance after all confounders were adjusted (p < 0.05, edf > 2) (Fig. 2A-I).

Fig. 2.

The non-linear relationship between NLR and depressive symptom by smooth curve fitting. A The non-linear relationship between NLR and P1 (anhedonia) B The non-linear relationship between NLR and P2 (depressed mood) C The non-linear relationship between NLR and P3 (sleep disturbance) D The non-linear relationship between NLR and P4 (fatigue) E The non-linear relationship between NLR and P5 (appetite change) F The non-linear relationship between NLR and P6 (low self-esteem) G The non-linear relationship between NLR and P7 (concentration problems) H The non-linear relationship between NLR and P8 (psychomotor changes) I The non-linear relationship between NLR and P9 (suicidal ideation)

Subgroups analyses

Anhedonia was associated with NLR among the above 60 years group (OR: 1.63, 95% CI: 1.14–2.33), and the male group (OR: 1.50, 95% CI: 1.07–2.10) (Fig. 3A). There was no significant association between depressed mood and NLR in all subgroups (Fig. 3B). Sleep disturbance was associated with NLR in the female group (OR: 1.36, 95% CI: 1.04–1.77) (Fig. 3C). Fatigue was associated with NLR in the female group (OR: 1.30, 95% CI: 1.02–1.67) and the non-Hispanic White group (OR: 1.32, 95% CI: 1.02–1.70) (Fig. 3D). No association was observed between appetite change, low self-esteem, concentration problems and psychomotor changes with NLR (Fig. 3E-H). Because of the inadequate positive sample size, no analysis was conducted for suicidal ideation in subgroups.

Fig. 3.

Subgroup analysis for NLR and specific depressive symptoms. A Subgroup analysis between NLR and P1 (anhedonia) B Subgroup analysis between NLR and P2 (depressed mood) C Subgroup analysis between NLR and P3 (sleep disturbance) D Subgroup analysis between NLR and P4 (fatigue) E Subgroup analysis between NLR and P5 (appetite change) F Subgroup analysis between NLR and P6 (low self-esteem) G Subgroup analysis between NLR and P7 (concentration problems) H Subgroup analysis between NLR and P8 (psychomotor changes)

Discussion

The present study aimed to investigate the complicated relationship between specific depressive symptoms and NLR. Our results showed that each 1-unit increment in NLR was linked to a 29% heightened risk of anhedonia (95% CI: 1.04- 1.61) and a 25% increased risk of fatigue (95% CI: 1.03- 1.51). A J-shaped relationship between anhedonia and NLR was observed, this implied that NLR values exceeding a certain threshold was linked to an increased incidence of anhedonia. Additionally, a U-shaped relationship between sleep disturbance and NLR was observed, this indicated that both very high and very low NLR levels may be associated with an increased incidence of sleep disturbance. This is of great significance for the future direction of clinical precision medicine. Subgroup analyses showed different patterns of association. Specifically, the associations were only observed for a few symptoms (anhedonia, sleep disturbance, and fatigue) among specific subgroups. These findings underscore the importance of considering demographic variations when examining associations between depression and inflammation.

In previous studies, the investigation of inflammation’s role in depression conventionally focused on establishing links at the diagnostic level [22, 33, 34]. However, this approach has inherent limitations [23]. Firstly, it tends to overlook the heterogeneity within the MDD population, where different symptom profiles may be associated with varying degrees of inflammatory activity [17, 19]. Secondly, diagnostic categorization may not fully capture the complex and nuanced spectrum of depressive symptoms experienced by individuals. Symptom-level analysis offers scientific rigor by allowing for the identification of distinct symptom clusters that may be more or less inflammation-driven [35]. This precision can facilitate targeted interventions and personalized medicine approaches, potentially improving treatment outcomes for subsets of patients with elevated inflammatory markers [36]. By focusing on NLR and its relationship to each of the nine symptoms and two symptom dimensions of the PHQ-9, we sought to contribute to a growing body of literature that moves beyond a one-size-fits-all diagnostic framework [22].

Our results are consistent with previous studies that have found correlations between specific depressive symptoms and elevated inflammatory markers, supporting the inflammatory theory of depression. However, there is disagreement about which specific symptoms are most closely associated with inflammation. Some studies emphasized the relationship between somatic symptoms and inflammatory cytokines [34, 37–39]. There was no significant association between either dimension of depressive symptoms and NLR in our findings. These differences might stem from variations in study populations, methodologies, or the differential sensitivity and specificity of biomarkers [33]. The difference in the definition of the symptom dimension may affect the results. Unlike CRP and IL-6, NLR is an indirect measure that reflects the balance between pro-inflammatory (neutrophils) and anti-inflammatory (lymphocytes) responses. As such, NLR might capture unique aspects of immune dysregulation that are not fully represented by single cytokine measurements. Moreover, the complexity of the immune system and the multifactorial nature of depression suggest that different inflammatory pathways may be involved, leading to different associations with specific depressive symptoms [40].

The non-linear relationship between anhedonia, sleep disturbance, and NLR suggested a complex interaction. This curvature could indicate threshold effects relationships, where both very high and very low NLR levels might be associated with increased risk for these depressive symptoms. The observed potential non-linearity could be attributed to several mechanisms, such as the effects of pro-inflammatory cytokines on dopamine function, modulation of the brain’s reward system, activation of the HPA axis, along with genetic predispositions, environmental stressors, and lifestyle. The diverse associations of NLR with depressive symptoms across specific subgroups imply a combination of biological and sociocultural influences. Although biologically plausible, NLR’s role in depression is likely modified by sociocultural factors, including stress, environment, and lifestyle, which may account for the observed differences. Unraveling this complexity is essential for advancing precision medicine in mental health, underscoring the necessity for research that integrates biological markers with sociocultural contexts. Further research should examine the impact of NLR thresholds on treatment responses to facilitate more tailored intervention strategies. Moreover, elucidating the relationship between NLR and depressive symptoms may pave the way for innovative anti-inflammatory treatments, particularly for those with treatment-resistant depression. Incorporating NLR assessments into standard clinical practice could improve diagnostic accuracy and prognostic capabilities, leading to enhanced mental health outcomes via targeted interventions.

There are a few limitations to this study. Firstly, it is difficult to draw causal conclusions from the inherent nature of its cross-sectional design. Longitudinal studies are needed to elucidate the causality and directionality of the depression-inflammation relationship. Second, although we adjusted for numerous confounders, unmeasured variables may still have influenced our results. The unmeasured confounders include certain diseases (such as autoimmune diseases), infections, diet, and chronic stress. Third, the generalizability of our findings is limited as they pertain specifically to an American population and may not extend to other ethnicities. Our study reveals that demographic factors significantly modulate the relationship between depressive symptoms and inflammation. Future research should be conducted to validate these findings in diverse populations. The study possesses several notable strengths. Firstly, our study benefits from a large, diverse sample and rigorous statistical adjustments, enhancing the generalizability and robustness of our findings. Secondly, we investigated the relationship between depressive symptoms, symptom dimensions and inflammation at the symptom level. Thirdly, we examined the relationship between NLR and symptoms in specific subgroups.

Conclusion

In summary, there were associations between NLR and specific depressive symptoms, particularly those related to psychological symptoms. These associations varied across demographic subgroups. There was a non-linear association between anhedonia, sleep disturbance, and NLR. However, the cross-sectional design of the study precluded determination of causality. Future longitudinal studies, particularly studies that could further investigate the mechanisms underlying the observed associations, are warranted to replicate these findings and establish the directionality and causality of the association. Such studies will facilitate understanding of the biological mechanisms underlying symptom-specific inflammation and assess the efficacy of targeted anti-inflammatory interventions. Ultimately, elucidating the symptom-level depression-inflammation nexus could advance precision medicine in mental health.

Authors’ contributions

MS Shan: conceptualization, formal analysis, Writing – original draft. SH Wang: Data curation, Writing – original draft. ZH Sun: Writing – original draft. Y Yang: Data curation, Writing - original draft. Y Pan:Conceptualization, Writing - review & editing. All authors contributed to manuscript revision, read, and approved the submitted version.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets supporting the conclusions of this article are publicly available for download from the National Center for Health Statistics at the Centers for Disease Control. https://www.cdc.gov/nchs/nhanes/index.htm.

Declarations

Ethics approval and consent to participate

The study was approved by the US National Center for Health Statistics Research Ethics Review Board (Protocol #2005–06, Protocol #2011–17, Protocol #2018–01) (https://www.cdc.gov/nchs/nhanes/irba98.htm). All methods were performed in accordance with the Declaration of Helsinki and the relevant guidelines. And all participants provided written informed consent. All methods were carried out in accordance with relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.