2.1 Data Source

The National Health and Nutrition Examination Survey (NHANES) database, a program run by the National Center for Health Statistics (NCHS) of the Centers for Disease Control and Prevention (CDC), provided the data for this study [14]. After obtaining the cases for this research from NHANES (1988–2018), we then selected the cases with strict inclusion and exclusion criteria (Figure 1). Moreover, diabetes was diagnosed in those who satisfied one or more of the following criteria: (1) 7.0 mmol/L for fasting plasma glucose or a 2-h oral glucose tolerance test level of 11.1 mmol/L; (2) 11.11 mmol/L for random blood glucose; (3) > 6.5% for glycohemoglobin (HbA1c); (4) use of insulin or diabetic medication; and (5) self-reported medical diagnosis of diabetes [15]. NPAR was calculated using the same blood sample and the following formula: Neutrophil percentage (in total WBC count) (%) × 100/Albumin (g/dL).

2.2 Definition of Clinical Covariates and Outcomes

The National Death Index database of the CDC served as the source of the mortality statistics. Every person's follow-up period was set to last from the time of participation until their death, whichever came first, or until December 31, 2019. Diabetes-cause and all-cause mortality were classified using the International Statistical Classification of Diseases, 10th Revision (ICD-10) codes [16]. The three categories of smoking status were never smoked (defined as fewer than 100 cigarettes in one's lifetime), former smoked (more than 100 cigarettes but no longer smokes), and current smoked (more than 100 cigarettes and smokes sometimes or daily presently). The drinking status was classified as follows: never drinking (defined as fewer than 12 drinks in a lifetime), former drinking (defined as more than 12 drinks in a single year and no drinking the previous year, or more than 12 drinks in a single year and no drinking the previous year), mild drinking (defined as fewer than 1 drink per day for females and fewer than 2 drinks per day for males), moderate drinking (defined as 1–3 drinks per day for females and 2–4 drinks per day for males on average), and heavy drinking (defined as fewer than 4 drinks per day for women or fewer than 5 drinks per day for men). A self-reported history of hypertension, the use of antihypertensive medication, and an average systolic blood pressure of at least 140 mmHg and/or average diastolic blood pressure of at least 90 mmHg were all considered indicators of hypertension. Nonalcoholic fatty liver disease (NAFLD) status was defined as serum ALT > 30 IU/L in males and > 19 IU/L in women without significant alcohol intake (> 3 drinks per day for men and > 2 drinks per day for women) and/or viral hepatitis (infections with the hepatitis B or hepatitis C viruses) [17].

2.3 Statistical Analysis

Weighted means were used to represent continuous data, whereas percentages were used to express categorical ones. To examine differences between two groups for continuous variables, the Student's t-test and the Mann–Whitney U test were used when applicable. The chi-squared test was used to assess group differences in categorical data.

Using optimally selected rank statistics, we determined the ideal NPAR cutoff point corresponding to the most significant correlation with the prognosis with the “maxstat” package. Participants were then divided into groups based on NPAR, higher and lower. To illustrate the possibly nonlinear relationship between the NPAR and both diabetes-cause and all-cause mortality in people with diabetes, a restricted cubic spline (RCS) with three knots was used.

The Kaplan–Meier method was utilized to compute the probabilities of survival outcomes, and the log-rank test was employed to compare them. Using the “timeROC” program, a time-dependent receiver–operator characteristic curve (ROC) analysis was performed to assess the NPAR's accuracy in predicting survival outcomes at various time points. R 4.4.0 and GraphPad Prism 10.0 were used for the statistical analyses. A two-tailed p < 0.05 indicated statistical significance.

3.1 Characteristics of the Study Population

Table 1 shows the characteristics of the study population. The current study included 3858 individuals with diabetes as volunteers. Participants were divided into two groups: the higher group (NPAR > 14.4, n = 1891) and the lower group (NPAR ≤ 14.4, n = 1967) based on the optimal NPAR cutoff value (14.4), which corresponded to the most significant connection with survival adopted based on optimally selected rank statistics. The participants in the higher NPAR group were older, had a shorter mean follow-up period, were more likely to have coronary heart disease, and were less likely to have NAFLD (Table 1). Moreover, diabetes patients with high NPAR had higher levels of serum creatinine, BUN, and HbA1c, and had lower levels of eGFR, TC, and HDL than those patients with low NPAR (Table 1).

3.2 Associations Between NPAR and All-Cause and Diabetes-Cause Mortality

Of the 3858 individuals with diabetes, 1198 (31.1%) died over a mean follow-up of 7.86 years; of these, 326 (8.4%) had diabetes-related deaths and 872 (22.6%) had deaths from other causes. In patients with diabetes, RCS analysis revealed a positive linear correlation (nonlinear p = 0.061) between the NPAR and all-cause mortality (Figure 2A). We found that the risk of all-cause mortality in the unadjusted model (Crude Model) increased considerably with an increase in the NPAR value (HR = 1.11, 95% CI: 1.08–1.12, p < 0.001) (Table 2). These impressive results still hold after multivariate correction for some or all of the clinical variables (Table 2). The higher NPAR group had a significantly higher risk of all-cause mortality in both the crude model (HR = 1.58, 95% CI: 1.41–1.77, p < 0.001), adjusted models 1 (HR = 1.44, 95% CI: 1.30–1.62, p < 0.001), and adjusted models 2 (HR = 1.37, 95% CI: 1.22–1.52, p < 0.001), according to Cox regression analysis (Table 2). We found similar results when analyzing diabetes-cause deaths (Figure 2B and Table 2).

Compared with low NPAR group, high NPAR group had a low survival rate of diabetes cases in all-cause death (Figure 3A, HR = 1.58, p < 0.001) with area under the curve (AUC) of the ROC curve being 0.725, 0.739, and 0.734 for 3-year, 5-year, and 10-year all-cause mortality (Figure 3B), respectively. Moreover, Kaplan–Meier survival rates for diabetes-cause mortality also suggested a lower survival rate in the high NPAR group, (Figure 3C) with the AUC of the 3-year, 5-year, and 10-year ROC curve being 0.754, 0.752, and 0.745 (Figure 3D), respectively. The concordance index of NPAR for the all-cause death (Figure 3E) and diabetes-cause death (Figure 3F) was higher than that of neutrophil percentage and albumin. According to these findings, the NPAR seems to have a reliable short- and long-term prognostic value for both all-cause and diabetes-cause death.