Association of systemic immune-inflammation index with all-cause and cardiovascular mortality among adults with depression: evidence from NHANES 2005–2018

Abstract

Background

The relationship between the systemic immune-inflammatory index (SII) and the mortality of adults with depression is uncertain.

Methods

This study included adults with depression who were surveyed in the National Health and Nutrition Examination Survey (NHANES) from 2005 to 2018. Cox proportional hazards regression models to compute hazard ratios (HR) and 95% confidence intervals (CI) for mortality.The restricted cubic spline(RCS), Kaplan-Meier curve analysis, time-dependent ROC analysis, subgroup and sensitivity analyses were also used.

Results

A total of 2442 adults with depression were included in the final analysis(average age: 46.51 ± 0.44 years). During a median follow-up of 89 months, there were 302 all-cause deaths and 74 cardiovascular deaths. The fully adjusted model showed that an increment of 100 unit in SII corresponded to an increased HR of 1.05(95% CI,1.02,1.08, p = 0.003) for all-cause mortality and 1.06(95% CI,1.02,1.10, p = 0.004) for cardiovascular mortality, respectively. The RCS analysis indicated a J-shape relationship between SII and all-cause mortality and a positive linear association between SII and cardiovascular mortality.The time-dependent ROC analysis exhibited excellent efficacy in SII for predicting all-cause and cardiovascular mortality at 1, 3, 5 and 10 years.

Conclusions

Higher SII levels were associated with increased risk of all-cause and cardiovascular mortality in adults with depression.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12888-024-06463-y.

Introduction

Cardiovascular disease (CVD) is the predominant cause of death worldwide, resulting in increased morbidity and mortality. The Global Burden of Disease study reveals that the prevalence of cardiovascular diseases has increased from 271 million people in 1990 to 523 million in 2019 [1]. The incidence of CVD has witnessed a substantial increase, with the number of fatalities linked to CVD soaring from 12.1 million in 1990 to a staggering 18.6 million in 2019 [1]. Depression, an increasingly recognized mental health disorder, reaches approximately 350 million people globally, posing a rapidly escalating challenge to public heath on a worldwide scale [2]. The prevalence and burden of depression in the United States has continued to rise in recent decades, with at last one in five United States adults experiencing depression during their lifetime, indicating a significant burden of depression worldwide [3, 4]. Both depression and cardiovascular disease are widespread and impose a considerable disease burden globally. The co-occurrence of depression and cardiovascular disease is prevalent, highlighting an interdependent connection between the two disorders [5]. Depression exerts a detrimental effect on the development and outcomes of cardiovascular disease [6]. The coexistence of depression alongside cardiovascular disease results in a poorer overall prognosis for both conditions [7]. Hence, recognizing the co-morbid risk factors for depression and cardiovascular disease in a timely manner is essential for the prevention, postponement, or reduction of mortality associated with these conditions.

Despite a wealth of evidence indicating a comorbidity risk between depression and cardiovascular diseases, the mechanisms underlying this relationship are not yet fully understood. Inflammation is conjectured to be intricately intertwined with depression and cardiovascular diseases, potentially acting as a crucial bridge between these two conditions. Inflammation is integral to the evolution of atherosclerosis and atherosclerotic thrombosis. It is instrumental at every stage, from the initial endothelial injury and subendothelial cholesterol deposition to plaque formation and the eventual rupture of plaques, which can provoke thrombotic complications [8]. This process is marked by the engagement of interleukins, cytokines, and inflammatory cellular elements, which propel the formation and progression of plaques, culminating in the manifestation of coronary artery disease [8]. Numerous studies have illustrated that pro-inflammatory markers are related to a greater risk of cardiovascular disease and serve as predictors for the likelihood of subsequent cardiovascular events [9]. Similarly, emerging studies point to inflammation as a central player in onset and development depression.

Inflammation mediators, particularly cytokines, may lead to depression by affecting the levels of neurochemistry. For instance, inflammation can promote the expression of inflammatory cytokines, thereby facilitating the development of depressive symptoms [10]. In addition, inflammatory mediators may also affect the synthesis and reuptake of neurotransmitters by activating the kynurenine pathway of tryptophan metabolism, thereby affecting emotional regulation [11]. Clinical trial has demonstrated extensive activation of the immune inflammatory response in groups of individuals with depressive disorders [12]. Inflammation is a key factor in both depression and cardiovascular diseases. The shared inflammatory mechanisms between depression and CVD include the activation of the hypothalamic-pituitary-adrenal (HPA) axis, increased levels of pro-inflammatory cytokines, and autonomic dysfunction. These mechanisms may contribute to the development and exacerbation of both conditions [13]. These excessive immune-infammatory reactions not only hasten depression progression but also cause irreversible harm to cardiovascular endothelial, leading to increased mortality and adverse cardiovascular events [14].

The systemic immune inflammatory index (SII) is a composite biomarker obtained by multiplying neutrophils by platelets and dividing by lymphocyte count measured in peripheral blood. It is considered to be a key indicator of systemic inflammatory status. The SII index, initially used to assess prognosis in patients with solid cancers, is now recognized as a reliable marker of inflammatory status [15]. More and more studies are focusing on the important impact of SII in individuals with depression. Prior studies have confirmed that high levels of SII are associated with an enhanced risk of depression in patients with stroke [16], diabetes mellitus [17], hemodialysis patients [18], and tuberculosis [19]. Several clinical studies have reported that the SII is positively and significantly associated with depression [20, 21]. Nevertheless, research on the presence and extent of the association between SII and mortality in individuals with depression is still a relative under-researched area.

The aim of this study was to investigate the relationship between SII and all-cause and cardiovascular mortality in a cohort using data from the the National Health and Nutrition Examination Survey(NHANES) database. We explored the dose-response correlation between SII and all-cause and cardiovascular mortality and determined the predictive validity of SII for all-cause and cardiovascular mortality.

Methods

Study design and data source

This investigation utilized data extracted from the National Health and Nutrition Examination Survey (NHANES) 2005–2018 and linked mortality information from the National Death Index (NDI). The survey comprises interviews, physical examinations and laboratory tests which is conducted every two years.The primary aim of NHANES is to assess the health and nutritional condition of adults and children in the United States. The NHANES protocol was approved by the National Center for Health Statistics (NCHS) Research Ethics Review Board approval. This study was conducted in accordance with the Declaration of Helsinki. All participants provided written informed consent prior to their involvement in the study.

Participants

This study extracted raw data from the NHANES database from 2005 to 2018, with a total of 70,190 individuals. The specific data for each period are as follows: 2005–2006 (N = 10348), 2007–2008 (N = 10149), 2009–2010 (N = 10537), 2011–2012 (N = 9756), 2013–2014 (N = 0175), 2015–2016 (N = 9971), 2017–2018 (N = 9254). Consistent with the criteria established in prior NHANES studies, depression in this study is identified by a PHQ-9 score of 10 or higher [22]. The inclusion and exclusion criteria for our study were as follows: Initially, we excluded participants < 20 years of age (N = 30441), leaving 39,749 participants. Subsequently, we excluded individuals with PHQ-9 < 10 (N = 31163) and those with missing PHQ-9 data (N = 5544), as well as individuals with missing SII data (N = 132) and mortality data (N = 4), leaving 2906 participants. Finally, we excluded individuals with missing demographic data, including marriage (N = 3), education (N = 3), BMI (N = 43), PIR (N = 251), and drinking status (N = 100); pre-existing conditions data, including cancer (N = 3), coronary heart disease (N = 18), and diabetes (N = 42); and medication use data, including antidepressant use(N = 1).The final number of participants included in the sample for this research was 2442 (Supplementary Fig. 1).

Definition of depression

The Patient Health Questionnaire (PHQ-9) is a nine-item questionnaire that is used to evaluate the severity of depressive symptoms experienced by individuals over a two-week period. Extensive validation studies have confirmed the reliability of the PHQ-9 as a diagnostic tool for depression. Each of the nine items is scored from 0 to 3 according to how the patient responds. The categories of response are “not at all,” “a few days,” “more than half the days,” and “almost every day.” The PHQ-9 total score range from 0 to 27. Participants with a score ≥ 10 were considered to have depression [22]. Furthermore, depression is classified into subcategories, including moderate depression(scores 10–14), moderately severe depression(scores 15–19), and severe depression(scores 20–27) [23].

Exposure measurement

The complete blood count is conducted on the Coulter DxH 800 analyzer under the supervision of trained medical personnel. This instrument is a fully automated, quantitative haematology analyser used for in vitro diagnostics in clinical laboratories, particularly for mass screening of patients. The SII was calculated as platelet count × neutrophil count/lymphocyte count. It was expressed as × 10^®9 cells/µl in accordance with previous studies. Details of procedures, quality control measures and data handling are available on the NHANES website.

Outcome ascertainment

Mortality data obtained up to 31 December 2019 were extracted from the CDC’s National Death Index database. In this study, causes of death were categorized according to the International Statistical Classification of Diseases, Tenth Revision (ICD-10). The classification was applied to determine deaths attributed to cardiovascular causes, which comprise rheumatic heart diseases (codes I00-I09), hypertensive heart and renal disease (code I11), ischaemic heart disease (code I13), heart failure (codes I20-I51) and cerebrovascular diseases (codes I60-I69).

Covariates

Potential confounders were included as covariates in the analysis, including demographic variables, physical examination data and comorbidities. Participants were categorized based on their ethnic identity, with the categories being Mexican American, Black, White and Other Race.Educational attainment was grouped into three levels: below high school, high school and above high school.

The study divided participants’ smoking status into three categories: Never smoke, Former smoke and Current smoke. Alcohol consumption was classified into three groups: Never drink, Former drink, Current drink. Family income was defined using the poverty income ratio (PIR). It was divided into three levels: below 1.30 (Low), between 1.30 and 3.50 (Medium) and above 3.50 (High). Body mass index (BMI) is calculated by dividing a person’s weight in kilograms by the square of their height in metres. It is then divided into three categories: under 25.0, between 25.0 and to < 30.0, and ≥ 30.0 kg/m². Hypertension was defined by self-reported current use of antihypertensive medication or a physician’s diagnosis. Diabetes criteria included: Fasting Plasma Glucose ≥ 126 mg/dL, Glycosylated Hemoglobin ≥ 6.5%, 2-h plasma glucose from an oral glucose tolerance test ≥ 200 mg/dL, current use of insulin or diabetes medications, and self-reported physician-diagnosed diabetes. Coronary heart disease and cancer were defined on the basis of self-reported physician-diagnosed conditions. Antidepressant use was classified as ‘no’ for those not taking any medication, ‘other’ for those taking medication other than antidepressants and ‘yes’ for those taking antidepressants.

Statistical analysis

Given the complex and multi-stage sampling strategy of NHANES, we used appropriate sample weights to adjust for selection and non-response bias to assure that the findings are representative of the United States population. Baseline characteristics of participants are described by weighted means and standard errors for continuous variables, and by numbers and weighted percentages for categorical variables. In order to determine the differences in the basic characteristics between the groups, one-way ANOVA and the Chi-square test were used to analyse the continuous variables and the categorical variables, respectively.

The participants were divided into quartiles according to SII. Multivariable weighted cox proportional hazards regression models were utilized to assess hazard ratios (HRs) for all-cause and cardiovascular disease(CVD) mortality. Three different adjustment models were used. Model 1 had not been adjusted. Model 2 adjusted for age and sex. Model 3 included further adjustment for educational level, marital status, BMI, PIR, alcohol consumption, smoking status, diabetes, hypertension, cancer, coronary heart disease and antidepressant use. The goodness-of-fit test with Schoenfeld residuals to verify the proportional hazards (PH) assumption. Restricted cubic spline (RCS) regression models were used to investigate the dose-response relationship between the SII and mortality risk. The Kaplan-Meier method was applied to analyse the probabilities of survival outcomes in terms of all-cause and CVD mortality during follow-up. The significance of the observed differences was tested by log-rank tests. Time-dependent ROC analysis was used to explore the predictive value of SII for predicting 1 year, 3 year, 5 year, 10 year all-cause and cardiovascular mortality among adults with depression. In stratified analyses, the association between baseline SII and mortality was examined in the subgroups by age, sex, race, BMI, smoking, drinking status, history of hypertension and diabetes, with the full adjusted model excluding the stratification factor. The survey-weighted Wald test was employed to evaluate the potential interaction. To ascertain the robustness of the findings, sensitivity analyses were conducted. Several sensitivity analyses were conducted to test the robustness of our findings including: (1) exclusion of death cases that occurred during the first year of follow up; (2) excluding data from patients with history of cancer; (3) excluding data from patients with history of coronary heart disease; (4) further adjusting for the level of depression.

All data analyses were performed with R software (version R-4.4.1). A two-tailed p-value of lower than 0.05 regarded as statistically significant.

Results

Patient characteristics

The final analysis utilized data from 2442 adults with depression within the NHANES 2005–2018 cohort. According to the quartiles of SII, participants were equally classified into four groups: lower SII (Q1, ≤ 343.059), low middle SII (Q2, 343.060–495.622), middle SII (Q3, 495.623–723.552) and high SII (Q4, > 723.752).

Baseline characteristics of patients are shown in Table 1, the mean age of the participants was 46.51 ± 0.44 years old. The subjects in the high SII group were more likely to be older, white people, widowed/separated/divorced, former drinking status; and tend to have a history of hypertension and diabetes. There were no significant differences between the quartiles SII groups in terms of sex, educational level, smoking status, PIR, BMI, coronary heart disease and cancer(all P > 0.05).

Table 1.

Characteristics of the study participants by quartiles of SII in NHANES 2005–2018 cohort

Characteristic	Total	Q1	Q2	Q3	Q4	P value
N	2442	612	609	610	611
Age, years	46.51 ± 0.44	45.29 ± 1.00	45.26 ± 0.77	47.53 ± 0.75	47.79 ± 0.68	0.01
Gender, n(%)						0.06
Female	1542(63.62)	387(61.25)	367(59.14)	395(67.37)	393(66.42)
Male	900(36.38)	225(38.75)	242(40.86)	215(32.63)	218(33.58)
Race, n(%)						< 0.001
Black	526(12.85)	201(21.53)	123(12.22)	109(10.07)	93( 8.57)
White	1085(65.66)	206(55.80)	264(66.70)	298(67.54)	317(71.44)
Mexican American	358( 7.57)	82(6.53)	100(8.49)	83(7.18)	93(7.96)
Other	473(13.92)	123(16.14)	122(12.59)	120(15.21)	108(12.02)
Marital status, n(%)						0.02
Married/Living with partner	1102(48.36)	274(45.97)	273(48.30)	280(50.29)	275(48.65)
Widowed/Separated/Divorced	840(30.82)	196(28.29)	203(27.88)	216(32.72)	225(34.02)
Never married	500(20.82)	142(25.73)	133(23.82)	114(16.98)	111(17.33)
Educational level, n(%)						0.17
Below High school	833(24.27)	214(25.33)	196(21.52)	219(26.62)	204(23.75)
High school	268(11.04)	56( 8.82)	70( 9.98)	74(12.07)	68(13.01)
Above High school	1341(64.69)	342(65.86)	343(68.50)	317(61.31)	339(63.24)
PIR, n(%)						0.74
Low	1306(41.83)	332(45.20)	329(43.12)	316(39.25)	329(40.12)
Medium	706(31.22)	176(29.97)	176(31.16)	177(31.87)	177(31.74)
High	430(26.95)	104(24.83)	104(25.72)	117(28.88)	105(28.14)
BMI, kg/m2						0.06
<25.0	583(25.87)	177(30.18)	125(24.92)	150(25.02)	131(23.85)
25.0 to < 30.0	637(26.05)	153(24.28)	178(28.56)	165(29.70)	141(21.67)
≥30.0	1222(48.08)	282(45.55)	306(46.52)	295(45.28)	339(54.48)
Smoking status, n(%)						0.67
Never	969(38.05)	258(41.52)	256(39.20)	228(35.03)	227(36.82)
Former	551(22.43)	137(21.48)	130(22.80)	144(23.08)	140(22.26)
Current	922(39.52)	217(36.99)	223(38.00)	238(41.89)	244(40.92)
Drinking status, n(%)						0.002
Never	288( 8.77)	89(11.81)	63( 6.49)	76(10.17)	60( 6.98)
Former	534(19.66)	115(16.64)	116(16.15)	142(22.04)	161(23.41)
Current	1620(71.56)	408(71.55)	430(77.36)	392(67.79)	390(69.61)
Hypertension, n(%)						0.002
No	1203(52.61)	306(53.80)	335(59.35)	298(52.58)	264(45.09)
Yes	1239(47.39)	306(46.20)	274(40.65)	312(47.42)	347(54.91)
Diabetes, n(%)						0.04
No	1811(79.90)	468(83.26)	462(82.42)	455(78.31)	426(76.05)
Yes	631(20.10)	144(16.74)	147(17.58)	155(21.69)	185(23.95)
Coronary heart disease, n(%)						0.18
No	2292(94.59)	573(93.31)	580(96.62)	572(94.76)	567(93.57)
Yes	150( 5.41)	39(6.69)	29(3.38)	38(5.24)	44(6.43)
Cancer, n(%)						0.98
No	2168(88.60)	543(88.05)	539(88.58)	548(88.93)	538(88.77)
Yes	274(11.40)	69(11.95)	70(11.42)	62(11.07)	73(11.23)
Antidepressant use, n(%)						0.04
No	675(26.30)	172(29.75)	177(27.06)	177(28.80)	149(20.17)
Other	977(36.71)	265(38.09)	246(36.87)	220(31.95)	246(39.93)
Yes	790(36.98)	175(32.16)	186(36.06)	213(39.25)	216(39.90)
Depression severity, n(%)						0.11
Moderate	1531(63.23)	410(69.15)	382(65.37)	369(58.98)	370(60.07)
Moderately severe	650(26.53)	145(20.86)	160(25.10)	175(30.42)	170(29.09)
Severe	261(10.25)	57( 9.99)	67( 9.53)	66(10.60)	71(10.83)

Abbreviations SII, Systemic Immune-Inflammation; PIR, poverty income ratio; BMI, body mass index; NHANES, National Health and Nutrition Examination Survey

All estimates accounted for complex survey designs, and all percentages were weighted

Moreover, Supplementary Table 1 presents the characteristics of the participants stratified according to their survival state. During a median follow-up period of 89 months, our study recorded 302 all-cause deaths and 74 cardiovascular deaths among participants. Compared with those who were still alive, participants who had died were more likely to be older, male, white people, former drinking, widowed/separated/divorced, have a low PIR level and educational level, have a high SII level; and tend to have a history of hypertension, diabetes, cancer and coronary heart disease, which could potentially affect the survival outcome.

Association of the SII with all-cause mortality

Table 2 displayed the unadjusted and multivariable-adjusted hazard ratios along with the 95% confidence intervals. The relationship between baseline SII and mortality was investigated using both a continuous and a categorical variable. When analyzing SII as a continuous variable, per 100 unit increased in SII corresponded to an increased HR of 1.05(95% CI,1.02,1.08) for all-cause mortality in the fully adjusted model. When analyzed as categorical variables and compared with the Q2 SII group (reference), individuals in the Q1 group had a HR of 1.94 (95% CI,1.18, 3.19), the Q3 group have a HR of 1.88(95% CI,1.16,3.07) and the Q4 group have a HR of 2.07(95% CI,1.30,3.28) for all-cause mortality, respectively. The PH assumption was verified by the Schonfeld goodness-of-fit test(GLOBAL p = 0.942 for all-cause mortality). The RCS analysis revealed that the association between the SII and all-cause mortality was nonlinear and J-shaped(nonlinear P = 0.008) (Fig. 1A). The KM analysis for all-cause mortality revealed a significant disparity between the quartiles of SII groups (P = 0.002) (Fig. 2A). The area under the curve (AUC) for all-cause mortality was 0.829 at 1 year, 0.836 at 3 years, 0.841 at 5 years, and 0.852 at 10 years through time-dependent ROC analysis (Fig. 3A).

Table 2.

Association of SII with all-cause and cardiovascular mortality in adults with depression from NHANES 2005–2018

	Model 1		Model 2			Model 3
	HR(95%CI)	P value		HR(95%CI)	P value		HR(95%CI)	P value
All-cause mortality
Per 100 increment	1.05(1.02,1.09)	0.001		1.05(1.02,1.08)	0.003		1.05(1.02,1.08)		0.003
Quartiles
Q1	2.16(1.30,3.60)	0.003		2.05(1.25,3.37)	0.004		1.94(1.18,3.19)		0.01
Q2	ref			ref			ref
Q3	2.09(1.27,3.44)	0.004		1.86(1.12,3.10)	0.02		1.88(1.16,3.07)		0.01
Q4	2.28(1.41,3.68)	< 0.001		2.16(1.34,3.47)	0.002		2.07(1.30,3.28)		0.002
P for trend		0.003			0.01				0.01
CVD mortality
Per 100 increment	1.08(1.05,1.12)	< 0.001		1.06(1.02,1.10)	0.003		1.06(1.02,1.10)		0.004
Quartiles
Q1	2.32(0.89,5.99)	0.08		2.09(0.87,5.02)	0.10		1.82(0.76,4.34)		0.18
Q2	ref			ref			ref
Q3	4.02(1.81,8.91)	< 0.001		3.46(1.55,7.73)	0.002		3.11(1.36,7.13)		0.01
Q4	3.10(1.30,7.41)	0.01		3.14(1.30,7.60)	0.01		2.79(1.14,6.85)		0.03
P for trend		0.01			0.004				0.01

Abbreviations SII, Systemic Immune-Inflammation; CVD, cardiovascular disease; PIR, poverty income ratio; BMI, body mass index; NHANES, National Health and Nutrition Examination Survey; HR, hazard ratio; CI, confidence interval

Model 1: no adjusted

Model 2: adjusted for age and sex

Model 3: further adjusted for race, education level, marital status, PIR, BMI, smoking status, drinking status, diabetes, hypertension, coronary heart disease, cancer and antidepressant use

Fig. 1.

The adjusted restricted cubic splines depicting the relationships between SII and mortality based on data from NHANES 2005–2018. (A) Association of SII with all-cause mortality. P for non-linearity = 0.008. (B) Association between SII and cardiovascular mortality. P for non-linearity = 0.52. Adjusted for age, gender, race, educational level, marital status, PIR, BMI, smoking status, drinking status, diabetes, hypertension, coronary heart disease, cancer, and antidepressant use. Abbreviations: SII, Systemic immune-inflammatory index; NHANES, National Health and Nutritional Examination Survey

Fig. 2.

Kaplan–Meier curves of the survival rate with quartiles of SII group. (A) All-cause mortality; (B) Cardiovascular mortality. Abbreviations: SII, Systemic immune-inflammatory index

Fig. 3.

Time-dependent ROC curves and time-dependent AUC values of the SII for predicting mortality. All-cause mortality (A and B) and Cardiovascular mortality (C and D). Adjusted for age, gender, race, educational level, marital status, PIR, BMI, smoking status, drinking status, diabetes, hypertension, coronary heart disease, cancer and antidepressant use. Abbreviations: SII, Systemic immune-inflammatory index; ROC, receiver operating characteristics curve; AUC, area under the curve

The results of the two-piecewise Cox regression analysis for all-cause mortality are shown in Table 3. A threshold value of 4.418 was determined for SII. On the left of the inflection point, per 100 unit change in SII resulted in a 6% decrease (HR = 1.06, 95% CI 1.03–1.09). While on the right of the inflection point, per 100 unit change in SII had no significant effect on all-cause mortality.

Table 3.

Threshold effect analysis of SII on all-cause mortality using the two-piecewise regression model

	Model 1			Model 2		Model 3
	HR(95 CI%)	P value		HR(95 CI%)	P value	HR(95 CI%)	P value
< threshold value	1.07(1.04,1.09)		< 0.001	1.06(1.03,1.09)	< 0.001	1.06(1.03,1.09)	< 0.001
≥ threshold value	0.77(0.57,1.05)		0.10	0.81(0.60,1.10)	0.17	0.85(0.65,1.13)	0.26

Abbreviations SII, Systemic Immune-Inflammation; CVD, cardiovascular disease; PIR, poverty income ratio; BMI, body mass index; NHANES, National Health and Nutrition Examination Survey; HR, hazard ratio; CI, confidence interval

Model 1: no adjusted

Model 2: adjusted for age and sex

Model 3: further adjusted for race, education level, marital status, PIR, BMI, smoking status, drinking status, diabetes, hypertension, coronary heart disease, cancer and antidepressant use

Association of the SII with cardiovascular mortality

When analyzing SII as a continuous variable, per 100 unit in SII corresponded to an increased HR of 1.06(95% CI,1.02,1.10) for cardiovascular mortality in the fully adjusted model. When analyzed as categorical variables and compared with the Q2 SII group (reference), individuals in the Q1 group had a HR of 1.82(95% CI,0.76,4.34), the Q3 group have a HR of 3.11(95% CI,1.36,7.13) and the Q4 group have a HR of 2.79(95% CI,1.14,6.85) for CVD mortality, respectively. The PH assumption was verified by the Schonfeld goodness-of-fit test(GLOBAL p = 0.598 for CVD mortality). The RCS analysis indicated a positive linear(nonlinear P = 0.52) relationship between SII and cardiovascular mortality, with the HR for mortality increasing significantly as SII values rose (Fig. 1B). The results of the KM survival rates for cardiovascular mortality showed that survival was reduced with the higher SII (P = 0.02) (Fig. 2B). For cardiovascular mortality, the AUC was higher as 0.921 at 1 year,0.898 at 3 years, 0.900 at 5 years, and 0.896 at 10 years through time-dependent ROC analysis (Fig. 3B).

Subgroup analysis

Subgroup analyses were undertaken by stratification of the major covariates to validate the reliability of the findings in the presence of potential confounders. As depicted in Table 4, the interaction analyses were performed between SII and all-cause and CVD mortality within age(≤ 65 vs. > 65 years), sex(female vs. male), race(White vs. Non white), BMI(< 30 vs. ≥ 30 kg/m²), smoking status (never, former, or current), drinking status (never, former, or current), hypertension (yes or no) and diabetes (yes or no) stratifications after with the fully adjusted model except for the stratifying variable. In the race (White) subgroup, per 100 unit SII increased corresponded to an increased HR of 1.13 (95% CI,1.06–1.20) in CVD mortality. In the race (Non white) subgroup, per 100 unit SII increased was not significant associated with an enhanced risk of CVD mortality, as indicated by HR of 0.99(95% CI, 0.92–1.06). No significant interactions were observed between SII levels and other stratified variables (P for interaction > 0.05).

Table 4.

Subgroup analyses of SII and mortality risk in adults with depression

Variable	All-cause mortality		Cardiovascular mortality
	HR (95 CI%)	P for interaction	HR (95 CI%)	P for interaction
Age
≤ 65 y	1.04(1.00,1.09)	0.37	1.11(1.05, 1.16)	0.37
‌> 65 y	1.07(0.98,1.17)		1.04(0.96, 1.12)
Gender
Female	1.04(0.99,1.10)	0.42	1.07(1.01,1.13)	0.19
Male	1.08(1.02,1.14)		1.18(1.06, 1.31)
Race
White	1.07(1.02,1.13)	0.27	1.13(1.06, 1.20)	0.02
Non white	1.02(0.98,1.07)		0.99(0.92, 1.06)
BMI
<30 kg/m2	1.05(1.00,1.11)	0.81	1.10(0.97,1.25)	0.89
≥30 kg/m2	1.06(1.00,1.12)		1.09(1.04, 1.14)
Smoking
Never	1.04(0.98,1.11)	0.61	1.20( 0.96, 1.49)	0.22
Former	1.10(1.02, 1.19)		1.06(0.98,1.14)
Current	1.06(1.01,1.12)		1.17(1.07, 1.29)
Drinking
Never	1.05(0.96, 1.16)	0.63	1.09(1.01, 1.18)	0.74
Former	1.06(1.00,1.13)		1.12( 0.97, 1.29)
Current	1.08(1.02,1.14)		1.12(1.02, 1.23)
Hypertension
No	1.03(0.97,1.09)	0.44	1.06(0.96, 1.16)	0.94
Yes	1.06(1.02,1.11)		1.09(1.03, 1.16)
Diabetes
No	1.03(0.98,1.08)	0.07	1.06(0.96, 1.17)	0.64
Yes	1.11(1.04,1.18)		1.10(1.05, 1.16)

In stratified analyses, the correlation between baseline SII and mortality was determined in subgroups by age, sex, race, BMI, smoking status, drinking status, history of hypertension and diabetes, with the fully adjusted model excluding stratification factors. The squares indicate hazard ratios, with horizontal lines indicating 95% CI. Abbreviations: SII, Systemic Immune-Inflammation; BMI, body mass index; HR, hazard ratio; CI, confidence interval

Sensitivity analysis

In order to guarantee the reliability of the research findings, several sensitivity analysis were conducted to ascertain the robustness of the model. Following the exclusion of participants who died within 12 months of follow-up (Supplementary Table 2), the results were not significantly altered. After excluding participants with a history of cancer (Supplementary Table 3), the results were basically unchanged. After excluding participants with a history of coronary heart disease (Supplementary Table 4), the results were remain stable. After further adjusting for the level of depression (Supplementary Table 5), the results were still firm.

Discussion

In this large-sample nationally representative cohort investigation, we explored the relationship between SII and mortality among adults with depression utilizing the NHANES data collected from 2005 to 2018. In the current study, elevated SII levels were a significantly positive associated with the risk of all-cause and CVD mortality. The RCS analysis indicated a J-shape relationship between SII and all-cause mortality and a positive linear association between SII and CVD mortality.The time-dependent ROC analysis suggested that the SII exhibited superior predictive power for in both short-term and long-term mortality. In addition, stratified analyses and sensitivity analyses demonstrated the robustness of our findings.

The association between SII levels and mortality risk has been investigated in many previous studies. The results of the multi-adjusted models indicated that individuals in the highest tertile of SII exhibited an elevated risk of mortality from all-cause (HR = 1.48, 95%CI,1.48–1.48) and CVD mortality (HR = 1.60, 95% CI,1.60–1.61) in comparison to those in the lowest tertile within the general population [24].These findings have been further confirmed in mortality studies focusing on hypertension [25], type 2 diabetes [26], osteoarthritis [27] and nonalcoholic fatty liver disease [28]. However, the existing evidence base on the relationship between SII levels and all-cause and cardiovascular mortality in adults with depression is limited. In our study, we found a 5% increase in all-cause mortality and a 6% increase in cardiovascular mortality for per 100 units increase in SII after adjustment for confounders.

In clinical research, SII has been extensively demonstrated to be a risk indicator for the progression of depression. A cross-sectional study conducted in America pointed out that the systemic immune inflammatory index significantly affects the risk of depression, with a 2% increase in depression risk for every 100 units increase in SII [20]. Another study further explored the dose-response association between SII and depression and demonstrated a J-shaped non-linear relationship between log2-SII and depression incidence [29]. A large sample observational study conducted in China discovered that patients with high SII scores were 3.614 times more likely to have moderate/severe depression than those with low SII scores [30]. A retrospective cohort study performed in Thailand found that baseline SII was significantly higher in Major depressive disorder(MDD) patients comparison to healthy controls, and SII had area under the ROC curve (AUC) values greater than 0.7 [31]. These results are in support of our findings that the SII is an independent factor for adverse outcomes in patients with depression.

Multiple studies have shown that depression is more common in people with cardiovascular disease. It is considered a strong risk factor for the occurrence of cardiovascular disease in healthy people and is predictive of adverse events in patients with a history of cardiovascular disease [32]. A study assessing the 10-year risk of coronary heart disease in patients with major depression found that the 10-year risk of coronary heart disease was significantly higher in patients with major depression than in healthy controls [33]. A retrospective cross-sectional study of 84 patients with coronary heart disease (CHD) discovered that the SII was higher in patients with CHD with MDD than in those without MDD [34]. A meta-analysis including nineteen prospective cohort studies found that depression was associated with a 36% increase in the risk of coronary death compared with non-depressed persons [35]. A cohort study based on the UK Biobank indicated that depression was related to an increased the likelihood of premature coronary heart disease and SII might play a mediatory role in the connection between depression and premature coronary heart disease [36]. These results support our finding that SII may contribute to adverse cardiovascular outcomes in patients with depression.

There is increasing evidence that the pathophysiological processes of depression are involved in systemic immune activation. Studies on immune inflammation and depression animal models have indicated that systemic inflammatory response triggered by lipopolysaccharide, tumour necrosis factor α and other pro-inflammatory factors can cause depression-like behaviour in mice. Inflammation impacts depression by interfering with the synthesis of key monoamine transmitters, activating the Hypothalamic-pituitary-adrenal axis and inducing oxidative stress responses, leading to abnormal glutamic acid function and malnutrition. Several meta-analyses have confirmed that pro-inflammatory cytokines and acute-phase proteins are elevated in people with depression [37]. Patients with depression have higher levels of interleukin-6(IL-6), tumor necrosis factor(TNF) and the soluble interleukin-2 receptor in the blood compared to healthy individuals [38]. Compared with healthy controls, the levels of IL-4, IL-6, and IL-10 changed significantly in major depressive disorder patients [39].The biological connection between inflammation, depression and CVD may be linked to the high levels of proinflammatory cytokines released by macrophages, such as IL-6, TNF-α and IL-1β, which play a pivotal role in the pathogenesis of both CVD and depression [40].

Chronic low-grade inflammation is gaining recognition as a crucial factor that can intensify both depression and cardiovascular risks, possibly through interconnected biological mechanisms. Inflammatory play a crucial role in regulating the composition and stability of atherosclerotic plaque in the vascular system. Inflammation’s role in atherosclerosis is multifaceted, involving the attraction and activation of immune cells, the occurrence of hypoxia and the deposition of oxidized LDL, and the enhancement of angiogenesis [41]. These factors collectively exacerbate an inflammatory response. Prolonged exposure to chronic inflammation can inflict damage on the endothelium of blood vessel, there by precipitating atherosclerosis [42]. The inflammation within blood vessels, coupled with accumulation of fatty plaques, can result in obstruction and thrombosis, subsequently, increasing the risk of cardiovascular incidents [8]. Within in the central nervous system, cytokines are pivotal in modulating neuronal development, survival, and plasticity [43]. In cases of chronic low-grade inflammation, there is a sustained release of pro-inflammatory cytokines and a reduction in anti-inflammatory cytokines.This heightened cytokine production in the bloodstream can breach the blood-brain barrier, triggering further the production of inflammatory cytokine by local immune cells within the brain [44]. Pro-inflammatory cytokines in the brain have been demonstrated to elicit “sickness behaviors”, which encompass diminished appetite, fatigue, and a decline in mood—symptoms that are indicative of depression [14]. In essence, the initiation of inflammatory responses is a shared characteristic between depression and cardiovascular disease.

The overlapping mechanisms between cardiovascular disease and depression hold significant clinical implications, potentially paving the way for therapeutic strategies that target inflammation to prevent and intervene in these conditions early on. At present, depression and cardiovascular disease are typically addressed separately with psychiatrists focusing on depression and cardiologists on heart-related issues. Chronic low-grade inflammation seems to be a shared pathway that can aggravate depression both and cardiovascular risks. This realization underscores the need for integrated care approaches, where managing inflammation could be a key strategy in treating depression and mitigating cardiovascular risks. Identifying individuals with both conditions simultaneously can be clinically advantageous for risk stratification and determining appropriate treatments. Considering these important insights, our research strongly advocates for the adoption of a comprehensive and integrated care model. Through an integrated, interdisciplinary management model for both depression and cardiovascular diseases, we can grasp the intrinsic connections between these conditions, thereby reducing their occurrence and death rates.

The interplay between depression, inflammation, and cardiovascular diseases is complex and multifaceted. Probing into the potential causes of inflammation might shed light on the connections between depression and cardiovascular disease. Behavioral, genetic, and environmental factors are intricately linked and may contribute to the underlying mechanisms of these comorbidities. Unhealthy lifestyle behaviors, such as smoking, poor diet, sedentary lifestyle and lack of exercise, contribute to chronic low-grade inflammation and the development of both depression and cardiovascular diseases [8]. Depression and cardiovascular disease have overlapping genetic risk factors, and part of this genetic overlap may be due to shared inflammation. Mendelian randomization analysis has indicated that inflammatory markers like IL-6 and C-reactive protein(CRP) may have a causal link with depression among cardiovascular risk factors [45]. This research highlights the significant role of chronic low-level inflammation in both depression and cardiovascular disease, underlining the genetic basis of inflammation in these comorbidities. The living environmental influences, particularly psychosocial stress, are crucial determinants in the occurrence of depression. Chronic stress can trigger inflammatory response, leading to atherosclerosis, which is a precursor to cardiovascular diseases and can worsen depressive symptoms [46]. Inflammatory mediators may promote the co-occurrence of these two diseases under stress conditions. The above studies indicate that inflammation can exacerbate depression and cardiovascular disease, and may be influenced by some common environmental, behavioral, and genetic driving factors. Given the role of inflammatory mediators in depression and cardiovascular diseases, interventions targeting these factors have the potential to improve psychological and physiological outcomes. For instance, improving dietary habits, boosting physical exercise, and alleviating stress might contribute to reducing the levels of inflammatory mediators. In addition, genetic screening can identify susceptible individuals, allowing for a focus on preventive measures. By employing the aforementioned strategies, it is possible to achieve beneficial effects on depression and cardiovascular diseases.

The significance of this study is to find that systemic inflammatory biomarkers are risk factors for all-cause mortality and cardiovascular death in patients with depression. It highlights the potential value of measuring systemic inflammatory biomarkers in identifying the risk of death in patients with depression. The SII provides an insight into the systemic inflammation and activation of the immune system of the individual. Our findings support that SII can serve as a candidate predictor of depression prognosis and has high accuracy. The study provides a better understanding of the relationship between systemic inflammation and depression and has the potential to inform new diagnosis and treatment options for depression. Consequently, it is advisable to combine SII with other biomarkers in clinical practice to enhance predictive accuracy.

This study boasts several strengths. Firstly, this is the first prospective cohort study to confirm that high SII levels increase all-cause mortality and cardiovascular mortality in adults with depression. Further more, we conducted several sensitivity analyses and found that our study is robust. Third, the time-dependent ROC analysis indicated that the SII exhibited superior predictive power for in both the short- and long-term in all-cause and CVD mortality.

However, there are several limitations to this study. First, the study population was restricted to adults with depression in the United States, which may constrain the model’s applicability to other countries and reduce generalizability. Future research should be conducted in more diverse populations to ensure the generalizability and applicability of the findings. Second, despite adjusting for several potential confounders in the analysis, we cannot exclude the possibility that SII may be influenced by other unidentified factors. Future research should focus on controlling for these potential confounders more rigorously or employ more precise methods to measure them, thereby allowing for a more accurate assessment of the relationship between SII and depression. Third, our study is based on observational study, which limits our ability to establish causality. Future studies should consider employing randomized controlled trials or Mendelian randomization analyses to further explore causality.

Conclusions

In a nationally representative sample in the United States, we observed a significantly positive association between SII and all-cause mortality and cardiovascular mortality among adults with depression. This finding has important clinical implications, as it suggests that SII could serve as a valuable prognostic biomarker in this patient population. By monitoring SII levels, clinicians may be able to identify patients with depression who are at higher risk of adverse outcomes, allowing for more targeted and intensive treatment strategies. Our findings highlight the importance of SII into routine clinical practice as a biomarker for predicting all-cause and cardiovascular mortality in adults with depression.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

SII

Systemic immune-inflammatory index

BMI

Body mass index

PIR

Poverty income ratio

CI

Confidence interval

HR

Hazard ratio

ICD-10

International Classification of Diseases 10th Revision

NDI

National Death Index

NHANES

National Health and Nutrition Examination Survey

PHQ-9

Patient Health Questionnaire systolic blood pressure

CDC

Centers for Disease Control

PH

Proportional hazards

KM

Kaplan-Meier

AUC

Area Under Curve

RCS

Restricted cubic spline

IL-6

Interleukin-6

TNF

Tumor necrosis factor

CRP

C-reactive protein

CHD

Coronary heart disease

MDD

Major depressive disorder

Author contributions

W. W contributed to study design and revision of the manuscript. X. Y. contributed to data analysis and writing of the manuscript. S. T. and L. W. contributed to data collection. H. Z. and M. L. contributed to data interpretation. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Grant numbers: 82160227), National Natural Science Foundation of China (Grant numbers: 82360234), Natural Science Foundation of Jiangxi Province (Grant numbers: 20224BAB206036), Jiangxi Provincial Department of Education Science and Technology Program Project (Grant numbers: GJJ210125), and National Nature Incubation Project of the Second Affiliated Hospital of Nanchang University (Grant numbers:2023YNFY12019).

Data availability

All data generated or analyzed during this study is included at this URL. https://www.cdc.gov/nchs/nhanes/index.htm.

Declarations

Ethics approval and consent to participate

Study protocols for NHANES were approved by the NCHS ethnics review board (Protocol #2005-06 Protocol, Protocol#2011-17,https://www.cdc.gov/nchs/nhanes/irba98.htm). All the participants signed the informed consent before participating in the study. 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.