Abstract

Background

The prevalence of type 2 diabetes mellitus (T2DM) remains high and biomarkers related to T2DM are needed to be investigated. This study aimed to investigate the association between neutrophil percentage-to-albumin ratio (NPAR) with all-cause mortality and the risk of cardiovascular disease (CVD) mortality in community-dwelling individuals with T2DM.

Methods

This prospective cohort study included 3602 adults aged 20 or above who were diagnosed with diabetes by the American Diabetes Association criteria in the US National Health and Nutrition Examination Survey (NHANES) database from 1999 to 2016, and followed up until 2019. Multivariable Cox proportional hazards regression models were utilized to determine the relationship between NPAR with all-cause mortality and cardiovascular disease (CVD) mortality. Restricted cubic spline regression analyses were employed to explore the nonlinear relationship between NPAR and cardiovascular disease (CVD) mortality. The assessment of nonlinearity was conducted using the likelihood ratio test.

Results

After adjustment, the risk ratio for all-cause mortality in the highest NPAR group (鈮モ��15.40) compared to the lowest serum NPAR reference group (<鈥�13.30) was 1.62 (95% CI, 1.36, 1.94) with P鈥塿alues鈥�<鈥�0.001. Nevertheless, the risk ratio for cardiovascular disease (CVD) mortality in the highest NPAR group (鈮モ��15.40) versus the lowest serum NPAR reference group (<鈥�13.30) was 1.41 (95% CI, 0.99, 2.00) with a P鈥塿alue of 0.06. Among patients with T2DM, serum NPAR levels exhibited a nonlinear correlation with both CVD mortality risk and all-cause mortality risk, with P鈥塿alues鈥�<鈥�0.001 for both.

Conclusions

This study identified a significant association between elevated levels of the neutrophil percentage-to-albumin ratio (NPAR) and an increased risk of all-cause mortality among individuals with type 2 diabetes mellitus (T2DM). Conversely, no correlation was found between NPAR levels and CVD mortality.

Introduction

Diabetes mellitus (DM) is a serious chronic disease characterized by elevated levels of blood glucose and is one of the leading causes of death and disability worldwide, impacting people regardless of country, age, or gender [1]. At the same time, diabetes also imposes a significant burden on healthcare systems, with the International Diabetes Federation (IDF) estimating that global medical expenditures for diabetes are expected to reach $1054听billion by 2045 [2]. The prevalence of diabetes remains consistently high, with a global prevalence rate of 8.5%, reaching up to 15.7% in the highest age group (65鈥�80 years old) [3]. Type 2 diabetes mellitus (T2DM) was correlated with a twofold to threefold increased adjusted risk for cardiovascular disease (CVD) events, which included myocardial infarction (MI), hemorrhagic stroke, coronary death, and other vascular fatalities [4]. Therefore, limiting the prevalence and mortality rates of diabetes and improving the prognosis of patients with diabetes is an urgent issue to be addressed.

Previous studies have identified several biomarkers that are significantly indicative of diabetes, such as elevated levels of leucine, alanine, and oleic acid, which can increase the risk of type 2 diabetes by 1.89, 1.63, and 1.87 times, respectively [5]. Diabetes is also associated with certain inflammatory markers. Research has found that patients with type 2 diabetes have higher levels of systemic inflammatory markers, such as high-sensitivity C-reactive protein (hs-CRP) and 滨苍迟别谤濒别耻办颈苍鈥�1尾.

(IL-1尾), as well as higher levels of oxysterols and triglycerides [6]. The study of various biomarkers related to diabetes also aids in assessing the prognosis of patients with diabetes.

In recent years, the Neutrophil to Albumin Ratio (NPAR) has been extensively studied. This index is a composite biomarker calculated by dividing the percentage of neutrophil molecules by serum albumin level [7] and has been found to be closely associated with the prognosis of many diseases. NPAR has been found to be associated with the prognosis of various cardiovascular diseases, such as coronary artery disease [8], acute myocardial infarction [9], cardiogenic shock [10], heart failure [11, 12], atrial fibrillation [13, 14] and Cerebrovascular disease like stroke [15,16,17]. As an inflammatory index, NPAR is also linked with other diseases, such as sepsis [18], chronic obstructive pulmonary disease.

(COPD) [7]and pneumonia and infection subsequent to stroke [19,20,21].

Meanwhile, NPAR has also been found to be related to certain metabolic diseases, such as non-alcoholic fatty liver disease [22, 23] and diabetic retinopathy [24] However, the relationship between NPAR and all-cause mortality, as well as cardiovascular disease (CVD) mortality in patients with diabetes, remains to be studied.

Cardiac conditions [25] and survival [26] are characterized by a high burden of inflammation. Inflammatory markers based on neutrophils or albumin have been associated with various inflammatory conditions. Specifically, neutrophil-based markers have been found to be elevated in type 2 diabetes mellitus [27], gastrointestinal diseases [28], thyroiditis [29], and Covid-19 infection [30]. Similarly, serum albumin-based markers have been linked to inflammatory conditions such as diabetic nephropathy [31] and diabetic neuropathy [32]. Given these associations, it is rational to investigate the relationship between the neutrophil-to-albumin ratio and mortality.

This study aims to investigate the correlation between NPAR and all-cause mortality, as well as mortality caused by CVD among patients with diabetes by including patients over 20 years old with T2DM from the National Health and Nutrition Examination Survey (NHANES) database from 1999 to 2016.

Method

Data source

This study was designed to investigate the relationship between NPAR with long-term mortality in adults among community-dwelling individuals with diabetes by analyzing data from the National Health and Nutrition Examination Survey (NHANES) conducted between 1999 and 2016.

The NHANES data was collected by National Center for Health Statistics (NCHS), a part of the Centers for Disease Control and Prevention (CDC) of the United States.

During the process of data collection and analysis, NHANES employed a complex multistage design and invited participants to complete various tests, including home interviews, physical examinations, professional measurements, and laboratory tests. Due to this rigorous methodology, the NHANES database is considered reliable and multidimensional, allowing for participant assessments that can be generalized to the population level [33].

Study population selection

The data was collected from the NHANES database between 1999 and 2016, including individuals aged鈥夆墺鈥�20 years with diabetes. Participants with diabetes mellitus(DM) were identified by answering positively to questions such as 鈥淎re you taking insulin?鈥� or 鈥淗as a doctor ever told you that you have diabetes or are you taking medicine for high blood sugar?鈥� or having an HbA1c鈥夆墺鈥�6.5%, a fasting glucose鈥夆墺鈥�126听mg/dL, or blood glucose level鈥夆墺鈥�200听mg/dL during oral glucose tolerance test (OGTT) recorded in the NHANES laboratory data [7].

Study variables

Study endpoints

In this study, we observed all-cause mortality, defined as death from any cause, as the study endpoint. In addition to this, we also focused on mortality caused by cardiovascular disease (CVD) with identification based on the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes ranging from I00 to I09, I11, I13, I20 to I51, and I60 to I69.

NPAR

The data for calculating NPAR were obtained from the complete blood count (CBC) Profile available in the NHANES database. CBC parameters were obtained using the Beckman Coulter method. Detailed information on laboratory methods, quality control, and data processing can be found at the following website: . For the computation of NPAR, the percentage of neutrophils (the proportion of neutrophils in WBC counts) needed to be calculated first. The specific formula is as follows: Neutrophil percentage(%)脳100/Albumin(g/dL).

Covariates

In this study, covariates were broadly classified into sociological factors and clinical characteristics. Sociological factors included ethnicity and race (self-reported), education level (categorized as less than 9th grade, 9-11th grade, high school/GED or equivalent, some college or AA degree, and college graduate or above), poverty-income ratio (PIR, calculated as the ratio of family income to poverty threshold), and marital status (comprised of married, widowed, divorced, separated, never married, living with a partner, refused, and do not know [34]). Clinical characteristics encompassed BMI (calculated as weight in kilograms divided by height in meters squared), smoking status (based on whether an individual had smoked more than 100 cigarettes in their lifetime [35]), physician-diagnosed history of diabetes, hypertension, and hypercholesterolemia (self-reported), kidney disease (assessed by calculating eGFR [36]), history of hypoglycemic medication use(including insulin, biguanides, sulfonylureas, thiazolidinediones, meglitinides, antidiabetic combinations and dipeptidyl peptidase 4 inhibitors) and various clinical laboratory indices. The laboratory indices examined included C-reactive protein (CRP), total cholesterol, high-density lipoprotein cholesterol, LDL-C level, glycohemoglobin level and liver function indicators (consisting of levels of aspartate aminotransferase, alanine transaminase, gamma-glutamyl transpeptidase, lactate dehydrogenase, and circulating homocysteine).

Statistical analysis

NHANES is a large sample of US adults, which is cross-sectional, nationally representative, with its sampling method being quite complex. Therefore, utilizing sample weights, clustering, and stratification is inevitable in this study, with each participant鈥檚 follow-up time calculated from the date of enrollment to the date of the participant鈥檚 demise or the end of the follow-up period (December 31, 2019).

To calculate hazard ratios (HRs) and 95% confidence intervals (CIs) for the association of neutrophil-to-albumin ratio (NPAR) with risks of CVD and all-cause mortality, we used the classic statistical method of multivariable Cox proportional hazard regression models. Schoenfeld residuals were employed to test the proportional hazard assumption, with two multivariable models being constructed simultaneously.

Furthermore, we generated weighted Kaplan鈥揗eier survival curves and conducted log-rank tests to assess cumulative survival differences across varying NPAR outcomes.

Unlike the initial model, which only included the ratio of neutrophil percentage to albumin, we adjusted for numerous sociological factors such as age, gender, race, ethnicity, education level, poverty income ratio (PIR), and marital status in one of the models in this research. In another model, we further adjusted for clinical characteristics of subjects like smoking status, duration of diabetes, hypertension or hypercholesterolemia, hypertensive medication use, lipid-lowering medications, history of hypoglycemic medication use (including insulin, biguanides, sulfonylureas, thiazolidinediones, meglitinides, antidiabetic combinations and dipeptidyl peptidase 4 inhibitors), BMI, systolic blood pressure (mmHg), diastolic blood pressure (mmHg), LDL-C level and glycohemoglobin level.

We analyzed the non-linear association between neutrophil-to-albumin ratio (NPAR) and all-cause mortality as well as CVD mortality, using the restricted tertile method (5th, 35th, and 65th percentiles), with the lowest percentile group serving as the control. This approach helps reduce the impact of outliers. The likelihood ratio test was employed to examine nonlinearity.

Stratification was done by age (<鈥�65 or 65 years), sex (men or women), ethnicity and race (self-reported Mexican American, non-Hispanic black, non-Hispanic white, or other), education level (high school, less than high school, or more than high school), smoking status (never, former, or current), diabetes (yes or no), hypertension (yes or no), hypercholesterolemia (yes or no), BMI (<鈥�28 or 鈮モ��28听kg/m²), systolic blood pressure (<鈥�140 or 鈮モ��140 mmHg), and diastolic blood pressure (<鈥�90 or 鈮モ��90 mmHg), with the P鈥塿alues being used to estimate the significance of interactions between the neutrophil percentage to albumin ratio (NPAR) and the stratified factors.

Additionally, we performed a series of sensitivity analyses to strengthen the robustness of our findings: First, to mitigate the potential for reverse causation bias, we excluded participants who had died within the initial 2 years of follow-up. Furthermore, we replicated the primary analyses using quintiles of serum NPAR levels to explore dose-response relationships. Concurrently, participants with a documented history of cardiovascular disease (CVD) were further omitted from the main analyses. Finally, to investigate the potential role of inflammation, lipid levels, and liver and kidney function in any observed associations, we further adjusted for C-reactive protein (CRP) levels, lipid profile (including total cholesterol high-density lipoprotein cholesterol and low-density lipoprotein cholesterol), indicators of kidney function (estimated glomerular filtration rate), and indicators of liver function (aspartate aminotransferase levels, alanine aminotransferase, 纬-glutamyl transferase, lactate dehydrogenase, and circulating homocysteine).

All statistical analyses were conducted utilizing R software, version 4.3.2, with a two-tailed P-value鈥�<鈥�0.05 established as the threshold for determining statistical significance.

Result

Study cohort selection

A total of 62,160 participants from NHANES study were extracted, followed up from 1999 to 2016. Among them, 5294 patients aged 20 years and older diagnosed with diabetes by a physician were included in the study. After excluding patients who did not have complete primary variables and mortality information during follow-up, 3602 subjects were included in the study cohort, which is detailed in Fig.听1.

Fig. 1

Steps of study population selection

Basic characteristics of community-dwelling individuals with DM

Among 3602 adults with diabetes aged鈥夆墺鈥�20 years, we found that women, older age, never married, non-Hispanic whites, longer duration of diabetes (鈮モ��10 years), taking antihypertensive medications, higher BMI, and lower diastolic blood pressure were associated with higher NPAR levels. Educational level, poverty status, smoking status, presence of hypertension or dyslipidemia, use of lipid-lowering medications, and systolic blood pressure were not significantly associated with NPAR levels. The details mentioned above are illustrated in Table听1.

Table 1 Baseline characteristics of participants

Association analysis

As shown in Figs.听2 and 3, there was a nonlinear association between serum NPAR levels and all-cause mortality, as well as cardiovascular disease (CVD) mortality (P鈥�<鈥�0.001 for both). After multivariate adjustment, compared to the reference group (first tertile), the risk ratio for all-cause mortality in the third tertile was 1.76 (95% CI, 1.47, 2.11) under Model 2; whereas under Model 3, the risk ratio for all-cause mortality in the third tertile was 1.62 (95% CI, 1.36, 1.94). For CVD mortality, under Model 2, the risk ratio for the third tertile was 1.52 (95% CI, 1.07, 2.16); and under Model 3, the risk ratio for CVD mortality in the third tertile was 1.41 (95% CI, 0.99, 2.00), with specific data presented in Table听2. From the K-M curves in Figs.听4 and 5, it can be observed that as the serum NPAR levels increase compared to the control group (first tertile), there is a decrease in survival rates, with P鈥塿alues for all-cause mortality and CVD mortality being p鈥�<鈥�0.001 and p鈥�=鈥�0.06, respectively.

Fig. 2

Association between serum NPAR levels and all-cause mortality

Fig. 3

Association between serum NPAR levels and CVD mortality

Table 2 Cox regression of the association between NPAR and all-cause mortality and CVD-cause mortality among patients with diabetes

Fig. 4

Kaplan-Meier curves for survival probability, with follow-up in years of all-cause mortality

Fig. 5

Kaplan-Meier curves for survival probability, with follow-up in years of CVD mortality

Stratified and sensitivity analyses

In stratified analysis, we observed that the correlation between all-cause mortality and cardiovascular disease mortality with neutrophil-percentage-to-albumin ratio (NPAR) exhibited significant differences only under stratification by body mass index (BMI). We found compared to the reference group (first quartile), the risk ratio for all-cause mortality in the highest NPAR level group was 2.02 (95% CI, 1.53, 2.67) for people with a BMI less than 28, and 1.42 (95% CI, 1.10, 1.84) for those with a BMI of 28 or greater, P鈥�=鈥�0.002. The specific results are shown in tables S1 and S2 in the appendix.

In sensitivity analyses, when excluding participants who died within the first 2 years of follow-up, the associations did not materially change overall. Moreover, when further excluding participants with a history of cardiovascular disease at baseline, or after adjustment for CRP level, lipid level, or liver function鈥搑elated or kidney function鈥搑elated indexes additionally, or when repeating the main analysis by quintiles of serum NPAR levels, the correlation between NPAR and all-cause mortality remained largely unchanged, whereas the association with cardiovascular disease mortality was attenuated. The results of the sensitivity analyses are detailed in tables S3 to S6 in the appendix.

Discussion

Within the context of this extensive, longitudinal cohort study involving adults with diabetes within the United States, we have discerned higher levels of NPAR (鈮モ��1540) were linked to a greater risk of death from all causes. Comprehensive stratified and sensitivity analyses corroborated the stability and reliability of our observed associations.

Previously, numerous studies have reported on the correlation between NPAR levels and mortality rates due to various causes in different populations. In two other studies involving patients undergoing peritoneal dialysis, it was consistently found that higher NPAR levels were independently related to an increased risk of all-cause and cardiovascular death [37, 38]. As for patients in the intensive care unit, studies have also shown that higher NPAR levels upon admission were independently related to increased all-cause mortality and hospital stay [39]. While among adults with T2DM, there is limited evidence regarding the levels of NPAR and their long-term health outcomes. Previous explorations into NPAR levels among diabetes populations found that elevated NPAR was related to an increased risk of diabetic retinopathy [24]. However, the association between NPAR and all-cause mortality, along with CVD mortality in individuals with diabetes, has yet to be investigated. This article addresses the gap in this field.

The mechanism behind this may be that NPAR itself is an indicator of inflammatory status. It is well known that low serum albumin levels are an indicator of the severity of inflammation and susceptibility to infectious complications [40], and an increase in NPAR levels generally indicates an exacerbation of infections or inflammatory levels. Diabetes is also closely related to many inflammatory responses. Therefore, the increased risk of death caused by elevated NPAR levels in diabetes populations could also be due to an aggravation of inflammatory states.

The findings regarding NPAR levels in this study have clinical implications, as they can aid in the early identification of patients with diabetes who are at an increased risk of all-cause and cardiovascular mortality. This early detection allows for the implementation of preventative measures, thereby offering a potential strategy to improve the overall prognosis of patients with diabetes mellitus.

The current investigation possesses several methodological strengths. To the extent of our awareness, this study represents the most extensive examination to date regarding the correlations between NPAR levels and both cardiovascular disease and all-cause mortality within individuals with diabetes, taking into account an array of potential confounding variables. Furthermore, the analytical framework is predicated upon a nationally representative sample of adult diabetes patients in the United States, thereby enhancing the potential for the extrapolation of results to a broader population.

The current study is constrained by several limitations. Initially, the determination of circulating NPAR levels was contingent upon a solitary serum assessment, which may not adequately mirror the long-term condition. Secondly, covariates documented at baseline are susceptible to temporal fluctuations, potentially diminishing the true correlation between NPAR levels and mortality rates. Thirdly, due to insufficient data, the severity of diabetes could not be meticulously accounted for in the present analysis, despite adjustments made for the duration of the diabetes condition. Fourthly, as the serum NPAR levels were categorized based on the tertiles of the specific study population, the outcomes may lack comparability with other investigations employing different cut points. Furthermore, it is impossible to entirely dismiss the potential for residual or unidentified confounding factors. Moreover, given the observational nature of the research design, the results gleaned should not be interpreted as causative inferences. Lastly, we did not include Glucagon-Like Peptide-1 Receptor Agonists (GLP-1 RA) and sodium-dependent glucose transporters 2 inhibitor (SGLT2i) in the models due to the minimal representation of SGLT2i and GLP-1 RA in our dataset, thus we conclude that their exclusion from the models is justified.

Conclusion

In a nationally representative sample of adults with diabetes in the United States, we observed non-linear associations between neutrophil percentage-to-albumin ratio (NPAR) and all-cause mortality. Elevated levels of NPAR were associated with an increased risk of all-cause mortality. However, no correlation was observed between NPAR levels and CVD mortality. These findings suggest that NPAR is a valuable hematological inflammatory marker for assessing the prognosis of patients with diabetes.