This study included a population sourced from the NHANES database spanning 1999 to 2018. The NHANES is a comprehensive health survey conducted nationwide by the Centers for Disease Control and Prevention (CDC).

The inclusion and exclusion criteria for this study were as follows: 1) Inclusion of CVDs, which encompasses coronary heart disease, congestive heart failure, myocardial infarction, stroke, and angina pectoris. The definition of CVD mortality aligns with the relevant guidelines outlined in the International Classification of Diseases, Tenth Revision (ICD-10); 2) Participants with missing data on high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), total cholesterol (TC) and triglycerides (TG) (n=34,166); 3) Exclusion of patients with missing urine creatinine or albumin data (n=6,220); 4) Exclusion of participants under 20 years of age (n=29,951); 5) Exclusion of participants with incomplete cardiovascular data (n=1,562); 6) Exclusion of individuals with missing diabetes status or non-diabetic patients (n=20,661); 7) Deletion of participants with incomplete CVD mortality data (n=1,319); 8) Exclusion of patients with missing educational attainment information (n=1,342). Ultimately, out of 100321 participants screened, we identified 5,136 individuals diagnosed with diabetes and 1,606 cases suffering from DKD ( Figure 1 ). Detailed information can be found at https://www.cdc.gov/nchs/nhanes/.

Plasma biochemicals measured in this study included fasting blood glucose (FBG), HDL-C, LDL-C, TC, TG, serum creatinine and albumin. Variables such as age, sex, education level, smoking status, waist circumference, and body mass index (BMI) were recorded by trained researchers. The categories of race include Mexican American, Other Hispanic, Non-Hispanic White, Non-Hispanic Black, and Other Race. The calculation formula for NHHR index is derived by subtracting HDL-C from TC, followed by dividing the result by HDL-C (14). In this study, the diagnosis of T2DM was established based on a fasting blood glucose value equal to or greater than 7.0 mmol/L. The patient had been diagnosed by a doctor with a CVDs. The diagnosis of DKD in T2DM patients was determined using an all-age spectral correction equation for creatinine to estimate the glomerular filtration rate (eGFR). Specifically, a bilateral eGFR of less than 60 mL/min/1.73 m² (15) and a urinary albumin-to-creatinine ratio (UACR) of ≥30 mg/g were used as criteria for determination.

This study employed sample weighting methods in accordance with NHANES guidelines to mitigate potential biases arising from the complex multi-stage sampling design of NHANES. Categorical variables were weighted by percentage, while continuous variables were adjusted based on their average values and standard deviations. The linear relationship between the NHHR index and mortality from CVDs in T2DM as well as DKD was analyzed using a restricted cubic spline approach, followed by adjustments for multiple confounding factors. Regressions models, including univariate and multivariate Cox regressions, were used to evaluate the NHHR index’s hazard ratios in connection with T2DM and DKD. Model 1 was uncorrected; Model 2 adjusted for sex, age, and BMI, whereas Model 3 included additional factors such asFBG, smoking status, race, TG, TC, hypertension, serum creatinine, albumin, eGFR, waist circumference, education level and UACR. Data analyses were performed using R language (version 4.4.1) and SPSS software (version 27.0), with a significance threshold set at p < 0.05.

The study population comprised 5,136 patients with T2DM from the NHANES database. The NHHR index was categorized into four quartiles, revealing significant differences among most variables between groups. Patients in the first quartile had the highest average age of 62.27 years. In the fourth quartile of NHHR, various indicators were generally elevated compared to other groups, including hypertension (21.7%), TG (3.2 mmol/l), LDL-C (3.05 mmol/l), TC (5.92 mmol/l), FBG (157.52 mg/dl), and a higher proportion of individuals with education levels below high school (39.0%). Additionally, urine albumin levels averaged at 140.14 mg/l, while urine creatinine clearance rate (UACR) was recorded at 157.17 mg/g ( Table 1 ). Among the cohort of 1,606 patients with DKD, further analysis of the fourth quartile revealed that triglyceride levels reached an average of 3.59 mmol/l; this group also exhibited the highest mean age at 69.60 years old. Similarly, within this fourth quartile for NHHR, notable increases were observed across several metrics including TC (5.97 mmol/l), FBG (182.61 mg/dl), and a rise in individuals with education levels below high school to approximately 43.3%. Furthermore, urine albumin levels escalated to an average of 482.31 mg/l while UACR increased significantly to reach values of up to 551.22 mg/g ( Table 2 ).

Overall, the NHHR index exhibited a nonlinear association with CVD mortality risk in both T2DM and DKD patients, characterized by an “L” curve indicating a decreasing risk followed by an increasing risk. The confounding factors adjusted for in Figure 2 and Table 3 include age, sex, BMI, FBG, smoking status, race, TC, TG, hypertension, serum creatinine levels, albumin levels, eGFR, waist circumference, education level, and UACR. In T2DM patients, a higher NHHR index was associated with a decreased probability of mortality [HR, 0.925; 95%CI (0.888-0.964), P<0.001]. Two-segmented linear regression analysis identified the inflection point of the NHHR index at 1.68. Below the specified threshold, an increase in the NHHR index was linked to a 72.8% decrease in the risk of death [HR, 0.372; 95%CI (0.254-0.547), P<0.001]. On the other hand, for values above 1.68, the NHHR index did not demonstrate a substantial association with the likelihood of death [HR, 0.959; 95%CI (0.919-1.001), P=0.054] ( Figure 2A , Table 3 ).

A similar nonlinear relationship was noted among DKD patients as well; here too, a higher NHHR index corresponded to reduced probability of cardiovascular death [HR, 0.921; 95%CI(0.872-0.972), P<0.001], with an inflection point established at 1.82—higher than that found in T2DM patients’ cohort. When the NHHR index fell below this value of 1.82, there was a notable decrease of approximately 48.2% per unit increase in probability of death [HR, 0.502;95%CI (0.347-0.727), P<0.001]; however, when it exceeded this threshold, no significant correlation emerged between the NHHR index and probability of death [HR, 0.0961;95%CI (0.0907-1.02), P=0.19] ( Figure 2B , Table 3 ).

In univariate COX regression analyses, the NHHR index in the fourth quartiles demonstrated statistically significant correlations with cardiovascular mortality risk among diabetic patients. It is noteworthy that Model 2 adjusted for several confounding variables, including age, sex, and BMI, revealing a strong association. Furthermore, in Model 3, which accounted for additional variables such as FBG, smoking status, race, TC, TG, hypertension, serum creatinine levels, albumin levels, eGFR, waist circumference, education level and UACR alongside age and sex and BMI; the NHHR index demonstrated a more pronounced relationship with the mortality risk among patients with T2DM [HR: 0.82; 95% CI (0.69-0.97), P=0.019] ( Table 4 ). In Model 3, a correlation was observed between the risk ratio of DKD leading to death and the third quartile of the NHHR index [HR: 1.51; 95%CI (1.06-2.14); P = 0.022]. In addition, the fourth quartile of the NHHR index and the DKD mortality risk ratio were also significant in Model 3 [HR, 2.04; 95% CI (1.28-3.26); P = 0.003] ( Table 4 ).

In the analysis of NHHR index and cardiovascular mortality risk ratio for T2DM, notable correlations were observed with elevated BMI, sex, tobacco use, education level, and hypertension ( Figure 3 ). However, no meaningful interactions were observed in these subgroups. For DKD mortality risk ratios related to the NHHR index, significant findings emerged among different genders, smokers, individuals with lower education levels, and non-hypertensive patients; yet their interactions showed no statistical significance ( Figure 4 ).

The research discovered a correlation that is not linear between the NHHR index and individuals suffering from both T2DM and DKD with CVDs, offering potential insights into the underlying pathological mechanisms of these conditions. A higher NHHR index is linked to a reduced mortality risk, indicating its potential as a valuable predictive marker. Furthermore, the study also identified a turning point in the NHHR index for patients with T2DM and DKD, which may help us determine the future treatment targets for these diseases.

While this study used a relatively large sample size, there is still a chance of selection bias impacting the generalizability of its findings. The study focused on the trends that existed between the NHHR index and the risk of death and did not explore other potential predictors. Additionally, because this study had a retrospective design, there is an inherent risk of bias influencing its results. It is possible that researchers did not adequately consider confounding factors affecting mortality risks in T2DM and DKD, which could lead to biased outcomes.

The NHHR index exhibits a non-linear relationship with the risk of CVD mortality in patients with DKD and T2DM. Specifically, when the NHHR is below 1.68, an increase in the NHHR index is associated with a reduction in CVD mortality risk among individuals with T2DM. Similarly, when the NHHR falls below 1.82, an elevation in the NHHR index correlates with a decreased risk of CVD mortality in patients suffering from DKD. These findings may hold potential value for assessing prognosis using the NHHR index within these two populations; however, further prospective studies are necessary to validate these results.