Inverted U-Shaped relationship Between Systemic Immune-Inflammation Index and Pulmonary Function: A Large Population-Based Study in US Adults

Introduction

Chronic respiratory diseases (CRD), including chronic obstructive pulmonary disease (COPD), asthma, and lung cancer, are a group of diseases that affect the health of the lungs and trachea.1,2 CRD is one of the leading causes of death worldwide and a major global public health problem.3

Pulmonary function test is of great value for the diagnosis of respiratory diseases and the evaluation of therapeutic effect. International guidelines have forced vital capacity (FVC), and forced expiratory volume in 1 second (FEV1) as parameters, such as chronic obstructive pulmonary disease (COPD), interstitial lung disease, asthma, and respiratory system disease diagnostic criteria.4 Early and dynamic pulmonary function assessment is an effective means of early screening for chronic respiratory diseases, which helps to take targeted intervention measures before the progression of lung disease.5 However, pulmonary function examination is not used as routine screening at all levels of medical and health care institutions worldwide, and most patients will undergo pulmonary function examination only when they have severe symptoms. At the same time, there are also some contraindications for pulmonary function testing,6 such as myocardial infarction, massive hemoptysis, shock patients, etc. Pulmonary function testing plays an important role in the perioperative assessment, treatment, and prognosis of thoracic surgery patients. However, within 3 weeks after thoracic surgery, pulmonary function testing is also contraindicated. There is a lack of effective means of dynamic detection of lung function in patients with these conditions.7

To sum up, finding a simple and effective biomarker to evaluate lung function is particularly important for early detection of lung function decline and timely prevention of further progression of impaired lung function.

COPD is characterized by chronic bronchitis, chronic airway obstruction, airway remodeling, and emphysema, resulting in a progressive, irreversible decline in lung function.8 A large number of studies9 have shown that chronic respiratory diseases such as COPD are related to chronic airway inflammation, which is the core of chronic respiratory diseases. Inflammatory mediators and destructive enzymes released by inflammatory cells, especially infiltrating immune cells, play an irreplaceable role in the destruction of lung tissue and the progressive decline of lung function. Accelerated decline in lung function increases the risk of chronic obstructive pulmonary disease (COPD), and smoking contributes to accelerated decline in lung function and the formation of COPD by increasing lung inflammation.10 Previous studies have shown that nutrients such as omega-3 fatty acids, which have anti-inflammatory properties, have a certain protective effect on lung function decline.11

Previous studies have also shown that single inflammatory cells such as white blood cells and neutrophils are associated with impaired lung function.12 However, at present, there is a lack of large-scale population studies on the relationship between comprehensive inflammatory indicators and lung function.

The Systemic Immunoinflammatory Index (SII) is a new index based on lymphocyte, neutrophil, and platelet counts. More and more evidences show that SII is a comprehensive indicator reflecting the immune and inflammatory state of the whole body system.13 At present, a large number of studies have confirmed that SII has a high prognostic value in tumors and chronic diseases caused by chronic inflammation or immune dysfunction, such as lung cancer,14,15 gastroesophageal adenocarcinoma,16,17 liver cancer,18 and cardiovascular and cerebrovascular diseases.19 SII can be analyzed with readily available data in blood routine, which is economical, convenient, and widely applicable, and may have potential predictive value for lung function decline. However, the relationship between SII and lung function in the population is still unclear.

The National Health and Nutrition Examination Survey (NHANES) is a population-based, cross-sectional survey designed to gather information about the health and nutrition of American households.20 The database uses a complex hierarchical, multistage probabilistic cluster sampling design to represent the entire US population. Therefore, we conducted a cross-sectional study based on large sample data from the NHANES to explore the relationship between SII and various indicators of lung function. The aim is to provide new strategies and recommendations for lung function assessment and protection.

Materials and MethodsData Sources

The data used in this study were obtained from the NHANES database. The NHANES database, a major program of the National Center for Health Statistics (NCHS) that provides vital health statistics, is unique in that it combines interviews and physical exams. The NHANES database includes demographic information, physical examinations, laboratory tests, and questionnaires to determine the prevalence of major diseases and risk factors for diseases.21 The data in this study were collected from the NHANES dataset for 3 periods from 2007 to 2012. All data in this study were approved by the Research Ethics Review Committee of the National Center for Health Statistics. All participants signed informed consent forms. Details can be downloaded from the CDC website (www.cdc.gov/nchs/nhanes/tutorials/default.asp).22

Study Subjects

The study was based on data from the 2007 to 2012 NHANES survey, as lung function data for these three cycles is complete in the NHANES database. A total of 30442 subjects were included in this study. According to the purpose of the study, the exclusion criteria were: (1) missing SII data (5463). (2) Lack of lung function data (6310). (3) Under 18 years of age (5127). (4) Participants with a pulmonary function data quality rating of C-F (n=463), cancer patients (n=911), heart failure patients (n=191), and demographic data missing (n=1852). Finally, 10,125 participants were included in the study. The process of research population screening is shown in Figure 1.

Figure 1 Flow chart of research population selection.

Pulmonary Function

The 2007–2012 NHANES provided spirometry for participants aged 6 to 79 years. The Ohio 822/827 dryroll volume spirometer followed the recommendations from the American Thoracic Society (ATS)/European Respiratory Society (ERS). Ruled out the following participants: chest pain; Physical problems with forced exhalation; Supplemental oxygen; Recent eye surgery, the chest or abdomen; Recent heart attack, stroke, tuberculosis, or coughing up blood. Expert reviewers from the NIOSH Quality Control Center reviewed all the spirometry data collected. To ensure the reliability of the analysis results, we only collected data that were rated as grades A and B in the quality of the spirometric data according to the American Thoracic Society (ATS) standard.23 The data we mainly used for analysis in this study included FVC, FEV1, FEV1/FVC, and PEF. Higher values for these measures of lung function represent better lung function. We calculated the percentage of predicted FEV1 (FEV1% or FEV1PRE) using the Hankinson equation.24

Systemic Immune-Inflammation Index(SII)

SII is calculated according to the complete blood count, and blood sample collection, processing, and quality control are carried out in accordance with the standardization process, The specific experimental methods can be obtained from the NHANES website (https://www.cdc.gov/nchs/nhanes). Lymphocyte, neutrophil, and platelet counts were measured by automated hematology analyzing devices (Coulter® DxH 800 analyzer) and present as ×103 cells/ mL. According to previous studies,25,26 the calculation of SII is based on an accurate formula: SII = platelet count×neutrophil count/lymphocyte count, which has high clinical application for evaluating the systemic immune and inflammatory status of individuals. In addition, the normality test suggested that the SII data were right-skewed distribution, and in order to ensure the normality of the data, log2 transformation of SII was performed during regression analysis (Figure S1, Table S1).

Covariates

Based on previous studies,27,28 covariates associated with SII and lung function were included in this study. Age (year), gender, race (Non-Hispanic White, Non-Hispanic Black, Mexican American, Others), BMI (kg/m2), Cotinine (ng/mL), Educational level (less than high school, high school, more than high school), poverty to income ratio (PIR)(<1.5,1.5–3.49,≥3.5), Hypertension history, Diabetes history, Respiratory diseases (asthma, emphysema, and chronic bronchitis). The concentration of cotinine was determined in thirds (Q1, 0.01–0.02ng/mL, Q2, 0.04–0.13 ng/mL, Q3, 27.17–280 ng/mL). Cotinine is an indicator of active and passive smoking and can better reflect an individual’s smoking status. The diagnosis of COPD depends on FEV1/FVC<0.7, and the questionnaire: ”Have you been told you have chronic obstructive pulmonary disease?

Statistical Analysis

According to the NHANES analysis guidelines and tutorials, sample weights were adjusted in this study (original 2-year sample weights /2).29 Therefore, we used a Mobile Examination Center (MEC) exam weight (WTMEC2YR) for our analysis, as some of the variables included in this study were collected in MEC. Furthermore, since we combined 3 NHANES survey cycles, the sample weight used in the final analysis is equal to 1/3 of the “WTMEC2YR” value. The influence of weight was fully considered in the subsequent population characteristics analysis and regression analysis. All analyses were based on participants with complete data, and individuals with missing covariate data were excluded from the final analysis. First, the participants were divided into four groups according to the Log2-SII quartile for analysis. Weighted linear regression analysis was used for continuous variables and a design-adjusted Chi-square test was used for categorical variables. In this study, mean ±standard deviation was used to report population characteristics for continuous variables and percentages were used to report population characteristics for categorical variables. Since the influence of sample weights has been adjusted, the specific number of cases was not listed. Multiple linear regression model was used to analyze the relationship between SII and pulmonary function indicators, and the quartile of SII was used as a categorical variable for the trend test (p for trend). Three models were used for the analysis: Model 1 did not take into account any confounding variables, Model 2 adjusted for age, sex, and race, and Model 3 took into account the effects of BMI, Cotinine, education, income, history of hypertension, history of diabetes and respiratory diseases on the basis of the model 2. After fully adjusting for confounding factors, we converted all continuous covariates into categorical variables for hierarchical multiple regression analysis and used the likelihood ratio test to evaluate the interaction effect. Regression coefficient beta values and 95% confidence intervals (CI) were used to describe these results. In this study, a generalized additive model and smooth curve fitting method were used to evaluate the nonlinear relationship between SII and pulmonary function indicators, and the inflection points in the curve were calculated by the non-linear model recursive algorithm to further explore the threshold effect. P<0.05 was considered statistically significant. R (version 2, R Foundation for Statistical Computing, Vienna, Austria) and Empower Stats (version 2.0, Boston, Massachusetts, USA) were used for all statistical analysis.

ResultsBaseline Characteristics

A total of 10,125 samples were included in this study. Due to the effect of sample weight, the specific number of cases for categorical variables was not listed in the baseline table The interquartile range of the Log-SII was used for weighted baseline characteristic analysis. The results showed that all variables except cotinine showed significant differences among the four groups. Compared with participants in the lowest quartile, those with higher Log-SII were more likely to be Female, Non-Hispanic White, and have higher BMI, education level, and income level. At the same time, with the increase of Log-SII, the incidence of diabetes, hypertension, and respiratory diseases gradually increased. In particular, FEV1/FVC and PEF decreased gradually with the increase of Log-SII level (Table 1).

Table 1 Weighted baseline characteristics of all participants stratified by Quartile of Log2-SII

The Association Between the SII and Lung Function

Table 2 showed the relationship between SII and lung function in three weighted generalized linear regression models. Since SII was a non-normal distribution, Log2 transformation was performed in the analysis. In the fully adjusted model (Model 3), each one unit increase in Log2-SII was significantly associated with a decrease in FVC(β, −23.4061; 95% CI, −42.2805- −4.5317), FEV1(β, −46.7730; 95% CI, −63.3371- −30.2089), FEV1%(β, −0.7923; 95% CI, −1.1635- −0.4211), FEV1/FVC(β, −0.6366; 95% CI, −0.8328- −0.4404) and PEF(β, −121.4468; 95% CI, −164.1939- −78.6998).

Table 2 Relationship between SII and lung functions

We further converted log2-SII into quartiles for sensitivity analysis. The results showed that the negative correlation between SII and lung function indicators remained stable With the increase of SII, the lung function indexes in each adjusted model showed a gradual downward trend (all p for trend<0.001). Compared with participants in the lowest SII quartile, those in the highest quartile had 45.08 decreases in FVC(β, −45.0815; 95% CI, −84.858- −5.305), 92.26 decreases in FEV1(β, −92.2563; 95% CI, −127.164- −57.349), 1.56 decreases in FEV1%(β, −1.5623; 95% CI, −2.3448- −0.7798), 1.23 decreases in FEV1/FVC(β, −1.2348; 95% CI, −1.6487- −0.8209), and 271.45 decreases in PEF(β, −271.4547; 95% CI, −361.489- −181.420) (Table 3).

Table 3 Effect of Log2-SII quartiles on lung function indices

Subgroup Analysis

To determine the potential interaction of covariates on the association between SII and lung function and the stability of the association between SII and lung function, The study participants were also stratified by age, gender, race, BMI, Cotinine, Educational level, PIR, Hypertension history, Diabetes history and Respiratory diseases for analysis and interaction test. All subgroup analyses were fully adjusted for confounding factors. All covariables except the grouping variable were adjusted. Detailed results of the subgroup analysis are shown in Figures 2–6. SII was negatively correlated with lung function parameters (FVC, FEV1, FEV1%, FEV1/FEVC, PEF) in all subgroup analyses. Although we observed positive β-effect values for FVC in the Non-Hispanic Black population, FEV1 in the population with college degrees or above, and FEV1/FVC in the BMI≥30 population, none of the above results were statistically significant (p>0.05), suggesting the robustness of our findings.