Abstract
Objective
To investigate the relationship between the uric acid to high-density lipoprotein cholesterol ratio (UHR) and ALL-cause and cardiovascular mortality among diabetic patients.
Methods
This study utilized health data from diabetic patients included in the National Health and Nutrition Examination Survey (NHANES) from 1999 to 2018. The Kaplan-Meier curves was employed to preliminarily explore the association between UHR, its components, and all-cause and cardiovascular mortality in diabetic patients, as well as to analyze UHR levels and mortality across different genders. Subsequently, the Cox proportional hazards model was used to further investigate the relationship between UHR, its components, and mortality in diabetic patients. Restricted cubic spline (RCS) curves were applied to examine the nonlinear relationship between UHR, its components, and mortality, with a particular focus on the association between UHR and mortality across different genders.
Results
This longitudinal cohort study included a total of 6,370 participants, comprising 3,268 males and 3,102 females. Kaplan-Meier analysis revealed a positive correlation between UHR, UA, and mortality in diabetic patients, while the association between HDL and mortality was negligible. The Cox proportional hazards model demonstrated a positive association between UHR and mortality in the diabetic population, while the statistical effects of UA and HDL on mortality were less pronounced compared to UHR. When analyzed by gender, no significant linear relationship was observed between UHR and mortality in either males or females. Subsequently, RCS analysis indicated a U-shaped nonlinear relationship between UHR and mortality in the overall diabetic population and among female patients, with a similar trend observed in males. Furthermore, stratified RCS analysis confirmed the persistence of the U-shaped relationship between UHR and prognosis across most subgroups.
Conclusion
This study found a U-shaped relationship between UHR and both ALL-cause and cardiovascular mortality in diabetic population. This suggests that clinicians should control UHR around 9–10 to improve the long-term prognosis of diabetic patients.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12872-024-04436-3.
Keywords: High density lipoprotein, Uric acid, Mortality rate, Metabolism, Inflammation, Diabetic patients, Longitudinal cohort study, NHANES database
Introduction
Diabetes is a global metabolic disease affecting hundreds of millions worldwide. According to the International Diabetes Federation (IDF), around 537 million adults (aged 20–79) had diabetes in 2021, with this number projected to soar to 1.3 billion by 2050 due to sharp increases in recent decades [1]. This surge in prevalence not only diminishes patients’ quality of life but also significantly raises the risk of all-cause and cardiovascular mortality [2]. Diabetic patients have a 1.5 to 2 times higher all-cause mortality rate compared to non-diabetic individuals [3], with cardiovascular disease responsible for over 50% of deaths. This creates a significant global health burden [4]. Thus, identifying a simple indicator to predict long-term health risks in diabetes is essential.
UHR (the ratio of uric acid to HDL cholesterol), a novel biomarker linked to inflammation and metabolism, was first introduced by Kocak [5]. Recently, UHR has shown potential in predicting diabetes onset [6], prediabetes [7], diabetic complications [8, 9], insulin resistance [10], hypertension [11], endocrine and metabolic disorders [12], metabolic syndrome, and cardiovascular diseases [13]. Given the inflammatory and metabolic nature of diabetes and cardiovascular diseases [14, 15], exploring the relationship between UHR and all-cause mortality and cardiovascular disease cause mortality in diabetic populations is necessary. Although no direct clinical data on UHR in diabetic patients exist, indirect evidence suggests significant differences between diabetic and non-diabetic populations. Diabetic patients’ elevated insulin levels increase uric acid reabsorption via renal transporters (URAT1, GLUT9), raising serum uric acid levels [16]. Higher BMI and fat accumulation in diabetics overactivate xanthine oxidase, leading to uric acid buildup, which exacerbates insulin resistance and accelerates the onset of complications. [17, 18]. Additionally, dyslipidemia, particularly lower HDL levels, is common, with diabetes disrupting lipoprotein metabolism and reducing HDL via various mechanisms [19]. As a result, insulin resistance, chronic inflammation, and dyslipidemia contribute to a higher UHR in diabetic patients, reflecting their metabolic disorder and indicating increased cardiovascular risk. Thus, UHR monitoring may serve as a valuable reference for evaluating the health status of diabetic patients.
To explore the relationship between UHR and the long-term prognosis of diabetic patients and to identify a simple and convenient prognostic indicator for diabetic patients’ health, we analyzed data provided by the National Health and Nutrition Examination Survey (NHANES) in the United States. This study aims to provide evidence for the clinical application of this indicator.
Materials and methods
This longitudinal cohort study utilized data from NHANES, a survey conducted by the National Center for Health Statistics (NCHS) in the United States. NHANES collects comprehensive data on the health and nutrition status of the U.S. civilian population, covering demographics, socio-economic status, dietary habits, and health-related issues. To ensure a representative sample, NHANES employs a stratified, multi-stage sampling method to select participants from various regions. The study protocol was reviewed and approved by the Research Ethics Review Committee of the NCHS at the Centers for Disease Control and Prevention (CDC). All NHANES procedures were approved by the NCHS Research Ethics Review Board (ERB), and participants provided written informed consent. Since this study involves secondary analysis of existing NHANES data, does not include participant privacy or personal information, and does not report or involve the use of any animal or human data or tissue, no additional informed consent or ethical approval from the ERB was necessary. More detailed information is available on the NHANES official website.
Selection criteria for study population
The study collected public data from 1999 to 2018. The inclusion and exclusion criteria for this study are as follows: Inclusion criteria: (1) Adult patients diagnosed with diabetes. Exclusion criteria: (1) Participants with gestational diabetes (2) Participants taking lipid-lowering medications or uric acid-lowering drugs (3) Participants missing elements required to calculate UHR (UA, HDL) (4) Participants with missing mortality data (5) Participants missing any required covariates.
The diagnostic criteria for diabetes in this study are as follows: (1) Physician-diagnosed diabetes, (2) Glycated hemoglobin (HbA1c) (%) ≥ 6.5, (3) Fasting blood glucose (mmol/l) ≥ 7.0, (4) Random blood glucose (mmol/l) ≥ 11.1, (5) Two-hour oral glucose tolerance test (OGTT) blood glucose (mmol/l) ≥ 11.1, and (6) Use of diabetes medication or insulin. Meeting any one of the above six criteria qualifies for a diabetes diagnosis. Ultimately, this study included 6,370 participants. The detailed screening process is shown in Fig. 1.
Fig. 1.
Flow diagram illustrating the participant selection process
Assessment of UHR and mortality
HDL and uric acid were measured from blood samples drawn from participants after an overnight fast. The measurement steps for HDL are as follows: a magnesium sulfate/dextran solution is added to the sample to form a water-soluble complex with non-HDL cholesterol, which does not react with the measurement reagent in subsequent steps. Then, polyethylene glycol esterase is added to convert HDL cholesterol esters into HDL cholesterol. The hydrogen peroxide generated by this reaction reacts with 4-aminoantipyrine and HSDA to form a purple or blue pigment. Finally, the HDL content is determined by photometric measurement at 600 nm. The measurement steps for UA are as follows: serum uric acid concentration is measured using the timed endpoint method with the DxC800 automated chemical analyzer. Uric acid is oxidized by uricase to produce allantoin and hydrogen peroxide. Hydrogen peroxide reacts with 4-aminoantipyrine (4-AAP) and 3,5 dichloro-2-hydroxybenzenesulfonate (DCHBS) in a peroxidase-catalyzed reaction to form a colored product, which is then measured photometrically at 520 nm to determine the uric acid content. The UHR (%) is then obtained by dividing UA (mg/dl) by HDL (mg/dl) and multiplying by 100 [20].
Mortality was assessed using the publicly available NHANES mortality dataset updated through December 31, 2019. Causes of death from specific diseases were determined based on the International Classification of Diseases, 10th Revision (ICD-10). Cardiovascular disease mortality (including ischemic heart disease, acute myocardial infarction, heart failure, hypertensive heart disease, rheumatic heart disease, acute myocarditis, pericardial disease, stroke, and cerebral hemorrhage) corresponds to ICD-10 codes I20-I51, I11, I00-I09, I13, I60-I69.
Covariates
This study considered multiple variables that could influence UHR and mortality risk. These variables encompass various demographic characteristics of the study population, including age, race, marital status, education level, and poverty-income ratio. Additionally, lifestyle factors such as smoking and alcohol consumption, BMI, and metabolic diseases like hypertension, kidney disease, and cancer were included. Hypertension was defined based on a doctor’s diagnosis, the use of hypertension medication, and abnormal average blood pressure. Kidney disease was defined as eGFR < 60mL/min/1.73 m² and UACR > 30 mg/g. Specific details can be found on the NHANES website.
Statistical analyses
The statistical analysis in this study was conducted in strict accordance with the recommended NHANES design, utilizing appropriate weights for each analysis. For continuous variables following a normal distribution, data are presented as mean ± standard deviation; for continuous variables not following a normal distribution, data are presented as median. Baseline differences in continuous variables were assessed using analysis of variance, and baseline differences in categorical variables were assessed by the χ² test, with results displayed as percentage counts. To assess the associations between different levels of UHR in mortality, we categorized these indicators into quartiles: Quartile 1 (Q1 ≤ 25th percentile), Quartile 2 (Q2 > 25th and ≤ 50th percentile), Quartile 3 (Q3 > 50th and ≤ 75th percentile), and Quartile 4 (Q4 > 75th percentile). Kaplan-Meier analysis was used to preliminarily examine the relationship between UHR levels and ALL-cause as well as CVD-cause in diabetic patients. After adjusting for multiple covariates, Cox Proportional Hazards models were applied to further investigate the impact of UHR levels on ALL-cause, CVD mortality in diabetic patients, with results presented as hazard ratios (HR) and 95% confidence intervals (CI). The study used three models for analysis: crude model (no adjustment for confounding factors), Model 1 (adjusted for age, race, education level, PIR, and marital status to control for demographic factors), and Model 2 (further adjusted for BMI, smoking status, alcohol consumption, hypertension, CVD, cancer, and chronic kidney disease to control for the influence of medical history). Additionally, restricted cubic spline (RCS) analysis combined with multivariable-adjusted Cox Proportional Hazards models were used to evaluate the nonlinear relationship between UHR levels and ALL-cause as well as CVD mortality in diabetic patients. When a nonlinear relationship was present, a recursive algorithm was employed to determine the inflection point and further examine the threshold effect. Meanwhile, as UA and HDL are the primary components of UHR, we also conducted the aforementioned statistical analyses on UA and HDL individually. In this study, missing data on covariates were primarily handled through direct deletion, with multiple imputation used as a supplementary method. All analyses were performed using R software (version 4.3.1). Statistical significance was defined as a two-sided P < 0.05.
Results
Characteristics of participants
Based on the screening criteria, the study ultimately included 6,370 participants from the NHANES 1999–2018 data, comprising 3,268 men and 3,102 women. Considering the differing reference values for UA and HDL between men and women, participants were divided into four groups according to the quartiles of UHR for males and females separately.
It was observed that, compared to the Q1 group, the Q2, Q3, and Q4 groups exhibited poorer lifestyle habits (smoking and alcohol consumption), higher risk factors (higher uric acid, higher BMI, and lower HDL), and higher prevalence of metabolic and chronic diseases (chronic kidney disease, hypertension, cardiovascular disease, and cancer). Additionally, two important points were noted: firstly, the highest quartile values of UHR differed significantly between male and female participants, suggesting different thresholds and implications of UHR in diabetic men and women. Secondly, in male patients, the Q1 quartile had higher cardiovascular and ALL-cause mortality rates compared to Q2 or Q3, while in female patients, the Q1 quartile had higher CVD-cause mortality rates compared to Q2 and Q3 quartiles. These results indicate that the relationship between UHR and mortality might be nonlinear. Detailed information is provided in Table 1.
Table 1.
Characteristics of participants
| Male | Female | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Total | Q1 | Q2 | Q3 | Q4 | P | Total | Q1 | Q2 | Q3 | Q4 | P | |
| Range | <10.4 | >=10.4,<13.8 | >=13.8,<17.5 | >=17.5 | <8.04 | >=8.04,<10.6 | >=10.6,<13.8 | >=13.8 | ||||
| UA, mg/dL | 5.995±0.036 | 4.680±0.051 | 5.452±0.047 | 6.313±0.057 | 7.359 ± 0.059 | < 0.0001 | 5.498±0.036 | 4.099±0.043 | 5.051±0.049 | 5.728±0.045 | 6.979±0.057 | < 0.0001 |
| HDL, mg/dL | 43.660±0.305 | 56.959±0.620 | 45.126±0.401 | 40.881±0.396 | 33.613±0.288 | < 0.0001 | 51.933±0.421 | 67.529±1.191 | 54.342±0.490 | 47.731±0.395 | 39.473±0.348 | < 0.0001 |
| UHR(%) | 14.802±0.135 | 8.434±0.076 | 12.119±0.049 | 15.486±0.050 | 22.387±0.220 | < 0.0001 | 11.553±0.136 | 6.265±0.071 | 9.310±0.035 | 12.042±0.044 | 18.051±0.220 | < 0.0001 |
| Age, year | 57.996±0.325 | 58.010±0.614 | 58.490±0.544 | 58.151±0.602 | 57.301±0.621 | < 0.0001 | 59.105±0.372 | 58.681±0.669 | 59.738±0.748 | 58.490±0.693 | 59.522±0.624 | 0.464 |
| PIR | 3.004±0.046 | 3.008±0.077 | 3.141±0.085 | 2.984±0.056 | 2.903±0.076 | 0.21 | 2.502±0.044 | 2.653±0.088 | 2.453±0.080 | 2.576±0.076 | 2.336±0.082 | 0.043 |
| BMI | 32.038±0.200 | 28.828±0.288 | 30.992±0.281 | 33.309±0.415 | 34.535±0.379 | < 0.0001 | 33.787±0.189 | 30.635±0.387 | 32.426±0.286 | 35.489±0.481 | 36.268±0.380 | < 0.0001 |
| Race | 0.01 | < 0.001 | ||||||||||
| Non-Hispanic Black | 733 (11.450) | 222 (15.709) | 170 (10.165) | 168 ( 9.464) | 173 (11.260) | 786 (15.719) | 203 (17.184) | 188 (15.426) | 190 (14.716) | 205 (15.644) | ||
| Non-Hispanic White | 1311 (66.384) | 261 (58.748) | 335 (69.183) | 341 (68.343) | 374 (67.945) | 1107 (61.827) | 226 (56.200) | 256(59.159) | 293 (64.348) | 332 (66.997) | ||
| Other Race | 1224 (22.166) | 337 (25.543) | 309 (20.651) | 319 (22.193) | 259 (20.796) | 1209 (22.455) | 348 (26.616) | 331 (25.415) | 292 (20.936) | 238 (17.359) | ||
| Marital Status | 0.955 | 0.315 | ||||||||||
| non-single | 2334 (73.866) | 576 (72.913) | 582 (73.678) | 600 (74.669) | 576 (74.019) | 1515 (54.599) | 402 (57.721) | 373 (53.904) | 382 (55.500) | 358 (51.460) | ||
| single | 934 (26.134) | 244 (27.087) | 232 (26.322) | 228 (25.331) | 230 (25.981) | 1587 (45.401) | 375 (42.279) | 402 (46.096) | 393 (44.500) | 417 (48.540) | ||
| Education | 0.074 | 0.014 | ||||||||||
| <high school | 608 (9.704) | 161 (10.705) | 163 (9.607) | 159 (9.447) | 125 (9.219) | 589 (10.593) | 166 (11.173) | 158 (11.045) | 136 (10.206) | 129 (10.021) | ||
| ≥high school | 2660 (90.295) | 659 (89.296) | 651 (90.393) | 669 (90.554) | 681 (90.780) | 2513 (89.407) | 611 (88.827) | 617 (88.955) | 639 (89.794) | 646 (89.979) | ||
| Smoke | 0.059 | < 0.001 | ||||||||||
| No | 1261 (40.339) | 310 (39.106) | 334 (40.740) | 331 (44.389) | 286 (36.594) | 1890 (57.712) | 501 (59.906) | 503 (63.411) | 473 (59.397) | 413 (48.662) | ||
| Yes | 2007 (59.661) | 510 (60.894) | 480 (59.261) | 497 (55.611) | 520 (63.407) | 1212 (42.288) | 276 (40.094) | 272 (36.589) | 302 (40.604) | 362 (51.338) | ||
| Drinking | 0.05 | 0.011 | ||||||||||
| No | 280 ( 8.086) | 77 (8.161) | 67 (6.832) | 69 (9.570) | 67 (7.723) | 885 (23.899) | 202(20.016) | 249 (28.371) | 201 (21.952) | 233 (25.256) | ||
| Yes | 2988 (91.916) | 743 (91.839) | 747 (93.169) | 759 (90.430) | 739 (92.277) | 2217 (76.101) | 575 (79.984) | 526 (71.630) | 574 (78.048) | 542 (74.744) | ||
| Hypertension | < 0.0001 | < 0.0001 | ||||||||||
| No | 1041 (34.174) | 304 (43.362) | 264 (34.245) | 278 (34.661) | 195 (25.608) | 812 (28.134) | 252 (35.940) | 218 (30.691) | 186 (27.111) | 156 (19.551) | ||
| Yes | 2227 (65.826) | 516 (56.638) | 550 (65.755) | 550 (65.339) | 611 (74.392) | 2290 (71.866) | 525 (64.060) | 557 (69.309) | 589 (72.889) | 619 (80.449) | ||
| CKD | < 0.0001 | < 0.0001 | ||||||||||
| No | 1968 (64.929) | 531 (69.047) | 522 (72.037) | 507 (65.229) | 408 (53.595) | 1889 (63.889) | 543 (72.250) | 492 (65.415) | 472 (65.109) | 382 (53.509) | ||
| Yes | 1300 (35.071) | 289 (30.953) | 292 (27.963) | 321 (34.771) | 398 (46.405) | 1213 (36.111) | 234 (27.750) | 283 (34.585) | 303 (34.891) | 393 (46.491) | ||
| CVD | 0.007 | < 0.0001 | ||||||||||
| No | 2406 (75.554) | 647 (79.660) | 612 (76.697) | 607 (76.546) | 540 (69.724) | 2428 (78.783) | 656 (85.717) | 629 (80.654) | 603 (78.670) | 540 (70.736) | ||
| Yes | 862 (24.446) | 173 (20.340) | 202 (23.303) | 221 (23.454) | 266 (30.276) | 674 (21.217) | 121 (14.283) | 146 (19.346) | 172 (21.330) | 235 (29.264) | ||
| Cancer | 0.033 | 0.349 | ||||||||||
| No | 2837 (85.038) | 709 (83.235) | 699 (82.732) | 734 (88.355) | 695 (85.416) | 2689 (84.965) | 687 (85.863) | 682 (87.174) | 667 (83.969) | 653 (83.069) | ||
| Yes | 431 (14.962) | 111 (16.765) | 115 (17.268) | 94 (11.645) | 111 (14.584) | 413 (15.035) | 90 (14.137) | 93 (12.826) | 108 (16.031) | 122 (16.931) | ||
| ALL-cause mortality | 0.335 | 0.002 | ||||||||||
| No | 2286 (75.765) | 582 (74.289) | 593 (78.430) | 569 (76.272) | 542 (73.703) | 2330 (76.876) | 604 (80.060) | 594 (78.563) | 596 (78.403) | 536 (70.824) | ||
| Yes | 982 (24.235) | 238 (25.711) | 221 (21.570) | 259 (23.728) | 264 (26.297) | 772 (23.124) | 173 (19.940) | 181 (21.437) | 179 (21.597) | 239 (29.176) | ||
| CVD-cause mortality | 0.09 | 0.005 | ||||||||||
| No | 2927 (91.632) | 747 (92.520) | 742 (93.071) | 740 (91.990) | 698 (88.966) | 2836 (91.835) | 714 (92.655) | 720 (93.405) | 718 (93.217) | 684 (88.224) | ||
| Yes | 341 (8.368) | 73 (7.480) | 72 (6.929) | 88 (8.010) | 108 (11.034) | 266 (8.165) | 63 (7.345) | 55( 6.595) | 57 (6.783) | 91 (11.776) | ||
UA uric acid, HDL high-density lipoprotein, PIR poverty income ratio, BMI body mass index, CKD Chronic Kidney Disease, CVD Cardiovascular Disease
Kaplan-Meier analysis of UHR and long-term prognosis in diabetic populations
To explore the clinical significance of UHR in diabetic patients, this study employed Kaplan-Meier analysis to preliminarily assess the association between UHR levels and ALL-cause as well as CVD-cause mortality in diabetic patients, with a particular focus on gender differences. Additionally, since UA and HDL are the primary components of UHR, we also generated Kaplan-Meier curves for these variables, though they were not stratified by gender.
The results showed that higher UHR levels were positively correlated with all-cause and CVD mortality in the overall diabetic population, indicating that higher UHR levels are associated with lower survival rates (PALL < 0.001, PCVD < 0.001). When analyzing the relationship between UHR and mortality in diabetic patients of different genders, it was found that higher UHR levels were also significantly associated with increased all-cause and CVD mortality in both male (PALL = 0.031, PCVD < 0.001) and female patients (PALL < 0.001, PCVD < 0.001). The specific results are shown in Fig. 2. When we analyzed the UA and HDL quartile groups, we found a statistically significant relationship between the UA quartile group and the risk of ALL-cause and CVD-cause death(PALL < 0.001, PCVD < 0.001), while the relationship between HDL and the risk of all-cause and CVD-cause death was not significant(PALL = 0.328, PCVD = 0.487). (See Figure S1 and Figure S2)
Fig. 2.
Kaplan-Meier Analysis of UHR’s Impact on Long-term Prognosis in Diabetic Patients by Gender. Note:a Association between UHR and ALL-cause mortality in diabetes patients (gender-neutral) (b) Association between UHR and CVD-cause mortality in diabetes patients (gender-neutral) (c) Association between UHR and ALL-cause mortality in male diabetes patients (d) Association between UHR and CVD-cause mortality in male diabetes patients (e) Association between UHR and ALL-cause mortality in female diabetes patients (f) Association between UHR and CVD-cause mortality in female diabetes patients
Cox regression analysis of UHR and long-term prognosis in diabetic populations
To further investigate the relationship between UHR and the prognosis of male and female diabetic patients, as well as to compare the statistical differences among UHR, UA, and HDL, we applied a Cox proportional hazards model to analyze the entire diabetic cohort. Following this comparison, we are going to examine the relationship between UHR and mortality risk separately in male and female patients.
The results showed that the linear relationship between UHR and both ALL-cause mortality and CVD-cause mortality in the overall diabetic population was statistically significant(PALL=0.012, PCVD=0.005). Similar results were obtained when comparing the Q4 with Q1 quartiles (PALLQ4=0.039, PCVDQ4=0.047), and its relationship showed a trend (PALL trend=0.024, PCVD trend=0.012), these details are showed in Tables 2 and 3). However, the statistical relationship between UA and HDL with mortality risk in the diabetic population was less significant compared to UHR (details show in Table S1 and Table S2). The relationship between UA and mortality showed significance (PALL=0.002, PCVD=0.005), but when analyzed by quartiles (PALLQ4=0.074, PCVDQ4=0.052), the differences between UA groups were not significant, and the upward trend was less pronounced (PALL trend = 0.044, PCVD trend = 0.042). Similarly, the statistical relationship between HDL and mortality risk in the diabetic population was also less significant than that of UHR. The association between HDL and mortality showed (PALL= 0.689, PCVD= 0.626), and in quartile analysis (PALLQ4 = 0.632, PCVDQ4 = 0.529), the trend was also not apparent. Therefore, we chose UHR as a more suitable index for analyzing mortality risk in the diabetic population and conducted subgroup analyses by gender specifically for UHR. Subsequently, when analyzing male and female diabetic patients separately, the linear relationship results were not significant for the relationship between UHR and both ALL-cause and CVD-cause mortality. Specifically, for male patients, the P-values were not significant (PALL=0.709 and PCVD=0.448) ,and there were also no significant differences between the Q4 and Q1 quartiles (PALL Q4=0.348, PCVD Q4=0.431). Similarly, for female diabetic patients, the results were not significant (PALL=0.065, PCVD=0.153), and there were no evident differences between the Q4 and Q1 quartiles (PALL Q4=0.593, PCVD Q4=0.396). These results suggest that the relationship between UHR and both cardiovascular and ALL-cause mortality in diabetic patients might not be a simple linear relationship and requires further investigation to determine the exact nature of this relationship. Detailed results are shown in Tables 2 and 3.
Table 2.
Association between UHR and mortality across cox proportional hazards models
| Outcome | Crude Model | P | Model 1 | P | Model 2 | P |
|---|---|---|---|---|---|---|
| HR (95%CI) | HR (95%CI) | HR (95%CI) | ||||
| ALL-cause mortality (gender-neutral diabetic patients) | 1.018(1.006,1.031) | 0.005 | 1.028(1.017,1.040) | < 0.0001 | 1.015(1.003,1.027) | 0.012 |
| CVD-cause mortality (gender-neutral diabetic patients) | 1.026(1.010,1.042) | 0.001 | 1.040(1.024,1.056) | < 0.0001 | 1.023(1.007,1.040) | 0.005 |
| ALL-cause mortality (Male diabetic patients) | 1.004(0.988,1.020) | 0.626 | 1.009(0.993,1.025) | 0.27 | 0.997(0.980,1.013) | 0.709 |
| CVD-cause mortality (Male diabetic patients) | 1.019(0.999,1.040) | 0.064 | 1.026(1.003,1.049) | 0.024 | 1.009(0.985,1.034) | 0.448 |
| ALL-cause mortality (Female diabetic patients) | 1.044(1.025,1.063) | < 0.0001 | 1.034(1.015,1.054) | < 0.001 | 1.018(0.999,1.037) | 0.065 |
| CVD-cause mortality (Female diabetic patients) | 1.045(1.021,1.070) | < 0.001 | 1.037(1.011,1.063) | 0.005 | 1.017(0.994,1.041) | 0.153 |
Crude Model: No-adjust, Model 1: adjust for age, race, PIR, marital status, and educational level (no sex), Model 2: adjust for covariates in Model 1, BMI, smoke, drinking, Hypertension, CVD, Cancer, and CKD
Table 3.
Association between UHR quartile groups and mortality across cox proportional hazards models
| Outcome | Groups | Crude Model | P | Model 1 | P | Model 2 | P |
|---|---|---|---|---|---|---|---|
| HR (95%CI) | HR (95%CI) | HR (95%CI) | |||||
| ALL-cause mortality (gender-neutral diabetic patients) | Q1 | ref | ref | ref | ref | ref | ref |
| Q2 | 1.076(0.913,1.268) | 0.382 | 1.113(0.950,1.303) | 0.184 | 1.045(0.869,1.258) | 0.638 | |
| Q3 | 1.060(0.891,1.260) | 0.513 | 1.215(1.029,1.434) | 0.022 | 1.102(0.923,1.317) | 0.281 | |
| Q4 | 1.367(1.129,1.654) | 0.001 | 1.490(1.228,1.809) | <0.001 | 1.236(1.011,1.512) | 0.039 | |
| P trend | 0.002 | <0.001 | 0.024 | ||||
| CVD-cause mortality (gender-neutral diabetic patients) | Q1 | ref | ref | ref | ref | ref | ref |
| Q2 | 0.914(0.681,1.227) | 0.55 | 0.957(0.718,1.274) | 0.762 | 0.866(0.629,1.193) | 0.379 | |
| Q3 | 1.089(0.829,1.431) | 0.538 | 1.288(0.988,1.678) | 0.061 | 1.117(0.845,1.477) | 0.436 | |
| Q4 | 1.514(1.123,2.041) | 0.007 | 1.702(1.261,2.298) | <0.001 | 1.349(1.003,1.814) | 0.047 | |
| P trend | 0.002 | <0.001 | 0.012 | ||||
| ALL-cause mortality (Male diabetic patients) | Q1 | ref | ref | ref | ref | ref | ref |
| Q2 | 0.799(0.601,1.060) | 0.12 | 0.850(0.645,1.118) | 0.245 | 0.864(0.656,1.140) | 0.302 | |
| Q3 | 0.975(0.750,1.267) | 0.85 | 1.019(0.787,1.321) | 0.886 | 0.990(0.769,1.276) | 0.941 | |
| Q4 | 1.038(0.806,1.338) | 0.773 | 1.068(0.826,1.382) | 0.615 | 0.886(0.687,1.141) | 0.348 | |
| P trend | 0.382 | 0.317 | 0.575 | ||||
| CVD-cause mortality (Male diabetic patients) | Q1 | ref | ref | ref | ref | ref | ref |
| Q2 | 0.879(0.535,1.443) | 0.609 | 0.937(0.571,1.538) | 0.798 | 0.942(0.596,1.489) | 0.799 | |
| Q3 | 1.132(0.742,1.727) | 0.564 | 1.176(0.760,1.822) | 0.467 | 1.125(0.748,1.692) | 0.573 | |
| Q4 | 1.495(0.991,2.253) | 0.055 | 1.512(0.972,2.354) | 0.067 | 1.185(0.776,1.809) | 0.431 | |
| P trend | 0.019 | 0.029 | 0.298 | ||||
| ALL-cause mortality (Female diabetic patients) | Q1 | ref | ref | ref | ref | ref | ref |
| Q2 | 1.054(0.825,1.346) | 0.676 | 0.865(0.678,1.104) | 0.244 | 0.748(0.570,0.982) | 0.037 | |
| Q3 | 1.176(0.920,1.504) | 0.195 | 1.136(0.888,1.454) | 0.31 | 0.919(0.689,1.225) | 0.563 | |
| Q4 | 1.611(1.252,2.073) | <0.001 | 1.384(1.071,1.789) | 0.013 | 1.068(0.839,1.360) | 0.593 | |
| Ptrend | <0.001 | 0.002 | 0.17 | ||||
| CVD-cause mortality (Female diabetic patients) | Q1 | ref | ref | ref | ref | ref | ref |
| Q2 | 0.882(0.559,1.391) | 0.588 | 0.723(0.467,1.118) | 0.145 | 0.600(0.376,0.955) | 0.031 | |
| Q3 | 1.007(0.673,1.507) | 0.972 | 0.975(0.656,1.450) | 0.9 | 0.752(0.489,1.159) | 0.197 | |
| Q4 | 1.782(1.243,2.554) | 0.002 | 1.555(1.095,2.209) | 0.014 | 1.156(0.828,1.613) | 0.396 | |
| P trend | 0.003 | 0.007 | 0.099 |
Note: Crude Model: No-adjust, Model 1: adjust for age, race, PIR, marital status, and educational level. Model 2: adjust for covariates in Model 1, BMI, smoke, drinking, Hypertension, CVD, Cancer, and CKD
RCS curves of UHR and long-term prognosis in diabetic populations
The analysis revealed a nonlinear relationship between UHR, UA, HDL, and mortality risk in diabetic populations. To further explore the potential nonlinear associations between UHR and both CVD-cause and ALL-cause mortality in diabetic patients, we generate RCS curves for UHR and also plotted RCS curves for UA and HDL.Additionally, because we focused more on the relationship between UHR and mortality risk, we only generated gender-stratified RCS curves for UHR and calculated the effect analysis for different UHR thresholds.
Our results show a U-shaped relationship between UHR and both all-cause mortality and CVD mortality in the overall diabetic population. Among female diabetic patients, a similar U-shaped relationship between UHR and mortality was observed, while in male patients, a comparable trend was present but without significant nonlinear effects (Fig. 3). Additionally, the relationships between UA and HDL and mortality were also nonlinear. Beyond their respective inflection points, higher levels of UA and very high HDL were associated with increased mortality risk, resembling the pattern observed for UHR. As a combined index of UA and HDL, UHR successfully illustrates the synthetical cumulative effect of these two blood markers on survival outcomes through its graphical representation (Figures S3 and S4).
Fig. 3.
RCS Curves of UHR’s Impact on Long-term Prognosis in Diabetic Patients by Gender. Note: a Association between UHR and all-cause mortality in diabetes patients (gender-neutral) (b) Association between UHR and CVD-cause mortality in diabetes patients (gender-neutral) (c) Association between UHR and all-cause mortality in male diabetes patients (d) Association between UHR and CVD-cause mortality in male diabetes patients (e) Association between UHR and all-cause mortality in female diabetes patients (f) Association between UHR and CVD-cause mortality in female diabetes patients
Subsequently, we examined the threshold effects of UHR for mortality. For overall diabetic patients, the inflection point for the relationship between UHR and ALL-cause mortality is 10.29. To the left of this inflection point, each unit increase in UHR is associated with a 9.7% decrease in ALL-cause mortality; the inflection point for the relationship between UHR and CVD-cause mortality is 10.05; to the left of this point, each unit increase in UHR is associated with a 16.4% decrease in CVD-cause mortality. For female diabetic patients, the inflection point for the relationship between UHR and ALL-cause mortality is 9.51; to the left of this point, each unit increase in UHR is associated with a 17.1% decrease in ALL-cause mortality; the inflection point for the relationship between UHR and CVD-cause mortality is 9.23; to the left of this point, each unit increase in UHR is associated with a 24.1% decrease in CVD-cause mortality (see Table 4). It is important to note that while UHR and mortality do not show a significant statistical significance to the right of the inflection points for both the overall diabetic population and female diabetic patients, there is a noticeable upward trend. Additionally, although the trend between UHR and mortality is not clear for male patients, the overall pattern is similar.
Table 4.
Threshold effect analysis of UHR on ALL-cause and CVD-cause mortality in diabetic patients
| Outcome | Inflexion point | HR (95% CI) | P | P for interaction |
|---|---|---|---|---|
|
ALL-cause mortality (Overall diabetic patients) |
<10.29 | 0.903(0.823,0.992) | 0.033 | 0.006 |
| ≥ 10.29 | 1.014(1.000,1.028) | 0.054 | ||
|
CVD-cause mortality (Overall diabetic patients) |
<10.05 | 0.834(0.740,0.940) | 0.003 | < 0.0001 |
| ≥ 10.05 | 1.017(0.998,1.038) | 0.084 | ||
|
ALL-cause mortality (Female diabetic patients) |
<9.51 | 0.829(0.726,0.946) | 0.005 | 0.006 |
| ≥ 9.51 | 1.019(0.997,1.042) | 0.09 | ||
|
CVD-cause mortality (Female diabetic patients) |
<9.23 | 0.759(0.619,0.930) | 0.008 | < 0.001 |
| ≥ 9.23 | 1.012(0.980,1.045) | 0.466 |
Stratified RCS curves of UHR and long-term prognosis in different diabetic populations
Based on the aforementioned RCS curves, the relationship between UHR and long-term prognosis in both male and female diabetic populations shows a U-shaped curve, with no evident linear association. To further assess the stability of this U-shaped association across various populations, we generated stratified RCS curves. The results revealed that, although there may be differences in effect values, after plotting for groups of different ages, BMIs, races, CVD conditions, hypertension statuses, and cancer statuses, the U-shaped trend remained consistent (see Fig. 4). This confirms that the U-shaped relationship between UHR and long-term prognosis in diabetic populations persists across different groups.
Fig. 4.
RCS Curves of UHR’s Impact on Long-term Prognosis in Diabetic Patients from Different Populations. Note: a Association between UHR and all-cause mortality in diabetes patients (stratified by age) (b) Association between UHR and CVD-cause mortality in diabetes patients (stratified by age) (c) Association between UHR and all-cause mortality in diabetes patients (stratified by BMI) (d) Association between UHR and CVD-cause mortality in diabetes patients (stratified by BMI) (e) Association between UHR and all-cause mortality in diabetes patients (stratified by races) (f) Association between UHR and CVD-cause mortality in diabetes patients (stratified by races) (g) Association between UHR and all-cause mortality in diabetes patients (stratified by cardiovascular conditions) (h) Association between UHR and CVD-cause mortality in diabetes patients (stratified by cardiovascular conditions) (i) Association between UHR and all-cause mortality in diabetes patients (stratified by hypertension status) (j) Association between UHR and CVD-cause mortality in diabetes patients (stratified by hypertension status) (k) Association between UHR and all-cause mortality in diabetes patients (stratified by cancer status) (l) Association between UHR and CVD-cause mortality in diabetes patients (stratified by cancer status)
Discussion
This longitudinal cohort study included a total of 6,370 participants, comprising 3,268 men and 3,102 women. In our preliminary exploration of the relationship between UHR and mortality using Kaplan-Meier analysis, we found that UHR was positively correlated with all-cause mortality and CVD-cause mortality in the overall diabetic population and in diabetic patients of different sexes; higher UHR values were associated with lower survival rates. UA showed similar results, with higher UA levels associated with increased mortality risk. However, the statistical effect of HDL on mortality risk in the diabetic population was not significant. To further compare the relationships between UHR, its components, and mortality, as well as the gender-specific relationships between UHR and mortality, we conducted a Cox Proportional Hazards model. After adjusting for other covariates, a positive relationship between UHR and mortality was observed in the diabetic population, whereas the statistical effects of UA and HDL were less significant than that of UHR. Furthermore, when the Cox proportional hazards model was stratified by gender, no significant linear relationship between UHR and mortality was found. This led us to hypothesize that the relationship between UHR, its components, and mortality might be nonlinear. The subsequent RCS analysis confirmed our hypothesis, showing that the relationship between UHR and mortality in the diabetic population represents the combined impact of UA and HDL on mortality, exhibiting a U-shaped curve. It also revealed a U-shaped nonlinear relationship between UHR and mortality in the general population and among female diabetic patients, with a similar pattern observed in male patients. In addition, the relationship between UHR and prognosis in different diabetic populations was also mapped by stratified RCS curves, which showed the aforementioned U-shaped relationship. Specifically, an increase in UHR before the inflection point was associated with a decrease in mortality, whereas an increase in UHR after the inflection point was associated with an increase in mortality. Therefore, our findings suggest that UHR has two opposing predictive effects on the outcomes of diabetic patients before and after the inflection point. Thus, clinicians may aim to maintain a UHR around 9–10 to improve the long-term prognosis of diabetic patients.
Currently, the role of the UHR as an inflammatory and metabolic marker in predicting various diseases and adverse events has gained increasing recognition. Aktas G identified UHR as a strong predictor of glycemic control in male diabetic patients [21]. This phenomenon may be attributed to the chronic inflammation and metabolic disturbances commonly observed in diabetes. Elevated UA levels contribute to oxidative stress and inflammatory responses, while low HDL levels are associated with diminished anti-inflammatory capacity [22]. The combined assessment of UHR may reflect the overall metabolic burden and inflammatory status of patients. Xu J investigated the association between UHR and abdominal aortic disease, revealing a significant positive correlation with the incidence of aortic dissection and aneurysm [23].This association may stem from elevated UA levels stimulating smooth muscle cell proliferation and inflammatory responses, promoting arterial wall remodeling and weakening. Simultaneously, low HDL levels reduce antioxidant protection, and together, through UHR, these factors may more sensitively indicate pathological changes in arterial structure. Similarly, Deng F highlighted that higher UHR values can predict plaque rupture in acute coronary syndrome patients, associating it with plaque erosion and ulceration [24]. Plaque rupture involves increased inflammatory factors and oxidative stress, where elevated UA exacerbates plaque instability by promoting endothelial dysfunction and releasing inflammatory cytokines [25], while reduced HDL levels diminish protection against inflammation and lipid accumulation. Elevated UHR encapsulates the imbalance of these opposing effects, making it a sensitive marker for plaque status. Yang Y confirmed the prognostic value of UHR in predicting major adverse cardiovascular events (MACE) in patients with chronic total occlusion (CTO) [26], while Aydın C demonstrated an association between higher UHR and reduced collateral circulation in CTO patients [27]. Elevated UA in CTO patients may inhibit collateral vessel formation through localized oxidative stress and endothelial cell dysfunction, while low HDL levels further weaken anti-inflammatory and vascular repair capabilities. UHR thus reflects the severity of these pathological processes.
Despite these findings, studies on the association of UHR with mortality and long-term prognosis remain limited. A few reports have linked UHR to mortality risk. Liu R demonstrated the significant clinical value of UHR in predicting all-cause and cardiovascular mortality in dialysis patients [28]. In such patients, impaired UA excretion leads to UA accumulation, inducing inflammation and vascular calcification. Simultaneously, compromised renal function impairs liver synthesis of HDL while increasing low-density lipoprotein and triglycerides, weakening HDL’s anti-inflammatory and antioxidant effects and increasing mortality risk. Hence, high UHR values may signify the cumulative impact of these adverse factors, indicating a higher mortality risk. However, it is important to note that although higher UA levels and lower HDL levels are generally associated with increased mortality [29, 30], highlighting the contribution of high UHR to mortality, it is equally crucial to recognize the risks linked to low UHR, which may be associated with insufficient UA levels or excessively high HDL levels. This is because lower UA levels may fail to provide sufficient antioxidant protection, increasing vulnerability to oxidative stress and impairing vascular integrity. As a natural antioxidant, UA plays a critical role in neutralizing free radicals, and its deficiency may exacerbate endothelial dysfunction and inflammation. Meanwhile, excessively high HDL levels in chronic inflammatory environments may undergo structural and functional abnormalities, losing their anti-inflammatory and antioxidant properties, and even exhibiting pro-inflammatory effects [31–33]. The URRAH study by Palatini P, involving 18,072 participants, revealed that hyperuricemia increased cardiovascular mortality risk in individuals with high HDL levels, supporting the association of low UHR with elevated cardiovascular risk [34]. In this study, participants with extremely high HDL levels had lower UHR values, where the combined effects of high HDL and high UA heightened mortality risk, indicating that UHR could uncover potential HDL dysfunction in diverse settings [35]. Additionally, Liu C reported increased cardiovascular mortality in participants with HDL levels ≥ 80 mg/dL [36], while Madsen CM confirmed that elevated HDL levels correlated with higher overall mortality [37]. This suggests that HDL is not always protective. Extremely high HDL levels may signify “dysfunctional HDL,” which loses its anti-inflammatory and antioxidant properties under inflammatory or oxidative stress environments, explaining these paradoxical findings [38].Interestingly, elevated UA not universally harmful. Ma P observed that moderate UA concentrations exert antioxidant effects on hypoxic myocardial cells [39], Waring WS found that intravenous UA could protect endothelial function in diabetic smokers [40], and Amaro S reported improved stroke outcomes with elevated UA during reperfusion therapy [41]. These findings indicate that UA may have protective effects within a lower range due to its antioxidant properties but may trigger oxidative stress and inflammation beyond this threshold.
Therefore, understanding the intricate relationship between UA and HDL becomes crucial in assessing health outcomes. As a combined marker, UHR reflects the dynamic balance between UA and HDL, revealing the interplay between anti-inflammatory and pro-inflammatory mechanisms, as well as antioxidant, pro-oxidant, and metabolic processes, in disease progression. Our research demonstrates a U-shaped relationship between UHR and mortality, highlighting the critical role of maintaining an appropriate balance between UA and HDL for metabolic and cardiovascular health. Fluctuations in UHR on either side of the inflection point may increase mortality risk. Moreover, our study suggests that individual measurements of UA or HDL may not fully capture their prognostic significance, whereas UHR as a composite biomarker could provide deeper clinical insights. Future research should explore the broad application of UHR across different populations to develop more personalized disease management strategies.
With the aforementioned objective in mind, and given the well-known differences between genders in UA and HDL levels, we first conducted a study on diabetic populations of different sexes. However, the differences in UA and HDL ranges between male and female diabetic patients may have contributed to the inconsistent effect values observed in the analysis of UHR and mortality. This disparity is driven by hormonal influences on HDL fluctuations [42, 43], higher UA levels linked to men’s poorer lifestyle choices and protein-rich diets [44, 45], as well as men’s greater susceptibility to abdominal obesity and lipid metabolism disorders [46, 47]. These factors lead to uneven UHR distribution between genders and different mortality risks among diabetic populations [48–50]. Although our Cox and RCS analyses did not find a significant UHR-mortality relationship in men, the overall RCS curve for males was very similar to that of the general diabetic population and female participants, suggesting that heterogeneity among the study subjects may account for these differences.
At the same time, to investigate the U-shaped relationship between UHR and long-term prognosis in different diabetic populations, we plotted stratified RCS curves. The results showed that while different age groups, BMIs, ethnicities, cardiovascular conditions, hypertension statuses, and cancer statuses displayed varying effect values, they all maintained a similar U-shaped relationship. Among them, there were significant differences in the curves of different ethnicities, BMI, and cardiovascular conditions regarding long-term prognosis. The relationship between UHR and prognosis in diabetic populations was relatively consistent between White and Black individuals, while the relationship was less evident in other ethnic groups (Asian populations). This discrepancy is not only attributed to genetic differences in HDL and UA expression among different ethnic groups [51], but also to variations in dietary and lifestyle habits [52]. Additionally, we found that an elevated UHR in diabetic patients with a low BMI may pose a greater risk compared to those with medium or high BMI. This is partly because low BMI diabetic patients may belong to the type 1 diabetes category, where the severe deficiency in glucose metabolism often leads to poorer long-term prognosis and higher mortality risks [53]. Furthermore, individuals with low BMI have less resistance to disease progression compared to those with normal or high BMI [54]. Moreover, diabetic patients with cardiovascular disease are at a higher risk of mortality than those without cardiovascular disease, which is expected, as UA and HDL are closely associated with prognosis in coronary heart disease patients [55]. Therefore, the stratified RCS analysis reveals that the U-shaped relationship between UHR and long-term prognosis remains stable across different diabetic populations.
In conclusion, this study uncovered a U-shaped relationship between UHR and long-term prognosis in diabetic patients, suggesting that clinicians and patients should consciously manage UHR levels through lifestyle adjustments, such as quitting smoking, reducing alcohol consumption, exercising, increasing dietary fiber intake, and the use of necessary medications, such as uric acid-lowering agents or statins. Keeping UHR within an appropriate range could potentially improve the long-term prognosis of diabetic patients. Besides, this study possesses several notable strengths. To our knowledge, our research is the first to explore and establish the relationship between UHR and both ALL-cause and CVD-cause mortality in diabetic patients, making it innovative. Furthermore, taking into account the differences in UA and HDL between genders, we separately calculated the impact of UHR on outcomes for each gender, as well as its effect on mortality in the overall population. This method helps reduce the risk of overlooking key insights and is particularly well-suited for indices with gender differences.
However, our study has limitations. While UHR reflects both inflammation and metabolic status, potentially offering broader applications than uric acid or HDL alone, it may obscure the independent roles of these markers. Future studies should explore their individual effects on diabetic prognosis. Additionally, various unaccounted factors, such as environmental influences, diet, genetics, physical activity, stress, and medications, may impact UHR and mortality. Though we adjusted for general population characteristics, our findings require further validation. Besides, Longitudinal cohort studies, while valuable for exploring hypotheses, have limitations in establishing real causality due to extended follow-up and external factors. Future research should focus on randomized controlled trials (RCTs) and the biological mechanisms behind UHR’s effects on cardiovascular health and mortality for deeper insights. Lastly, due to genetic, lifestyle, healthcare, and socioeconomic differences across regions, our findings may not fully apply globally. Caution is needed when generalizing these results, and future studies should include more diverse populations to improve applicability.
Conclusion
This study found a U-shaped relationship between UHR and both ALL-cause mortality and CVD-cause mortality in the overall diabetic population, a pattern that remained evident across different subgroups. Through calculations, we identified the inflection point of the relationship between UHR and both ALL-cause and CVD-specific mortality to be around 9–10. Maintaining UHR near this inflection point is more beneficial for the long-term prognosis of diabetic patients.
Supplementary Information
Supplementary Material 1: Table S1. Relationship between UA and HDL levels with ALL-cause mortality risk in diabetic populations.
Supplementary Material 2. Table S2. Relationship between UA and HDL levels with CVD-cause mortality risk in diabetic populations.
Supplementary Material 3: Table S3. Original data information.
Supplementary Material 4: Figure S1. Kaplan-Meier curves illustrating the relationship between UA levels and mortality risk in diabetic populations:(A) Association between UA levels and ALL-cause mortality (B) Association between UA levels and CVD-cause mortality.
Supplementary Material 5: Figure S2. Kaplan-Meier curves illustrating the relationship between HDL levels and mortality risk in diabetic populations:(A) Association between HDL levels and ALL-cause mortality (B) Association between HDL levels and CVD-cause mortality
Supplementary Material 6: Figure S3. Restricted Cubic Spline curves showing the relationship between UA levels and mortality risk in diabetic populations: (A) Association between UA levels and ALL-cause mortality (B) Association between UA levels and CVD-cause mortality.
Supplementary Material 7: Figure S4. Restricted Cubic Spline curves showing the relationship between HDL levels and mortality risk in diabetic populations: (A) Association between HDL levels and ALL-cause mortality (B) Association between HDL levels and CVD-cause mortality.
Acknowledgements
Not applicable.
Abbreviations
- UHR
Uric acid to high-density lipoprotein cholesterol ratio
- NCHS
National Center for Health Statistics
- NHANES
National Health and Nutrition Examination Survey
- UA
Uric Acid
- HDL
High-Density Lipoprotein
- BMI
Body Mass Index
- PIR
Poverty Income Ratio
- CVD
Cardiovascular Disease
- CKD
Chronic Kidney Disease
- HR
Hazard ratios
- CI
Confidence intervals
- RCS
Restricted cubic spline
Authors’ contributions
Writing the first draft of the manuscript, statistical analyses, data organization, writing review, and editing : XC Huang and LS Hu.Research, statistical analyses, and editing : XJ Wang. Design research, supervision, editing, review, revision of the manuscript, and funding:J Li. All authors approved the final version of the manuscript.
Funding
This study was supported by the National Natural Science Foundation of China (No. 82474494), National Key Research and Development Program of China (No. 2022YFC3500102), the Beijing Municipal Science and Technology Development Funding Program of Traditional Chinese Medicine (No. JJ-2020-69), and High Level Chinese Medical Hospital Promotion Project (No. HLCMHPP2023065). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Data availability
All data generated or analysed during this study are included in this published article and its supplementary information files.
Declarations
Ethics approval and consent to participate
The study protocol was reviewed and approved by the Research Ethics Review Committee of the NCHS at the Centers for Disease Control and Prevention (CDC). All NHANES procedures were approved by the NCHS Research Ethics Review Board (ERB), and participants provided written informed consent. Since this study involves secondary analysis of existing NHANES data, does not include participant privacy or personal information, and does not report or involve the use of any animal or human data or tissue, no additional informed consent or ethical approval from the ERB was necessary. More detailed information is available on the NHANES official website.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Xuanchun Huang and Lanshuo Hu contributed equally to this work and share first authorship.
Contributor Information
Jun Li, Email: gamyylj@163.com.
Xuejiao Wang, Email: wxuejiao@bucm.edu.cn.
References
- 1.Gregg EW, Buckley J, Ali MK, Davies J, Flood D, Mehta R, et al. Improving health outcomes of people with diabetes: target setting for the WHO Global Diabetes Compact. Lancet. 2023;401(10384):1302–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Tang O, Matsushita K, Coresh J, Sharrett AR, McEvoy JW, Windham BG, et al. Mortality implications of Prediabetes and Diabetes in older adults. Diabetes Care. 2020;43(2):382–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Rao Kondapally Seshasai S, Kaptoge S, Thompson A, Di Angelantonio E, Gao P, Sarwar N, et al. Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med. 2011;364(9):829–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kim SM, Lee G, Choi S, Kim K, Jeong SM, Son JS, et al. Association of early-onset diabetes, prediabetes and early glycaemic recovery with the risk of all-cause and cardiovascular mortality. Diabetologia. 2020;63(11):2305–14. [DOI] [PubMed] [Google Scholar]
- 5.Kocak MZ, Aktas G, Erkus E, Sincer I, Atak B, Duman T. Serum uric acid to HDL-cholesterol ratio is a strong predictor of metabolic syndrome in type 2 diabetes mellitus. Rev Assoc Med Bras (1992). 2019;65(1):9–15. [DOI] [PubMed] [Google Scholar]
- 6.Kosekli MA, Aktas G. Serum uric acid to HDL cholesterol ratio is Associated with Diabetic Control in New Onset Type 2 Diabetic Population. Acta Clin Croat. 2023;62(2):277–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Balci SB, Atak BM, Duman T, Ozkul FN, Aktas G. A novel marker for prediabetic conditions: uric acid-to-HDL cholesterol ratio. Bratisl Lek Listy. 2024;125(3):145–8. [DOI] [PubMed] [Google Scholar]
- 8.Xuan Y, Zhang W, Wang Y, Wang B, Xia F, Zhang K, et al. Association between Uric acid to HDL cholesterol ratio and Diabetic complications in men and Postmenopausal women. Diabetes Metab Syndr Obes. 2023;16:167–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Aktas G, Yilmaz S, Kantarci DB, Duman TT, Bilgin S, Balci SB, et al. Is serum uric acid-to-HDL cholesterol ratio elevation associated with diabetic kidney injury? Postgrad Med. 2023;135(5):519–23. [DOI] [PubMed] [Google Scholar]
- 10.Zhou X, Xu J. Association between serum uric acid-to-high-density lipoprotein cholesterol ratio and insulin resistance in patients with type 2 diabetes mellitus. J Diabetes Investig. 2024;15(1):113–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Aktas G, Khalid A, Kurtkulagi O, Duman TT, Bilgin S, Kahveci G, et al. Poorly controlled hypertension is associated with elevated serum uric acid to HDL-cholesterol ratio: a cross-sectional cohort study. Postgrad Med. 2022;134(3):297–302. [DOI] [PubMed] [Google Scholar]
- 12.Kosekli MA, Kurtkulagii O, Kahveci G, Duman TT, Tel BMA, Bilgin S et al. The association between serum uric acid to high density lipoprotein-cholesterol ratio and non-alcoholic fatty liver disease: the abund study. Rev Assoc Med Bras (1992). 2021;67(4):549–554. [DOI] [PubMed]
- 13.Han X, Tan X, Liu M, Wei Y, He A, Pan Y, et al. Association between serum uric acid/high-density lipoprotein cholesterol ratio and hypertension among reproductive-aged women. J Health Popul Nutr. 2023;42(1):123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Aktas G. Association between the Prognostic Nutritional Index and Chronic Microvascular complications in patients with type 2 diabetes Mellitus. J Clin Med. 2023;12(18):5952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Sincer I, Gunes Y, Mansiroglu AK, Aktas G. Differential value of eosinophil count in acute coronary syndrome among elderly patients. Aging Male. 2020;23(5):958–61. [DOI] [PubMed] [Google Scholar]
- 16.Singh SK, Singh R, Singh SK, Iquebal MA, Jaiswal S, Rai PK. Uric acid and diabetes mellitus: an update. Postgrad Med J. 2023;99(1178):1220–5. [DOI] [PubMed] [Google Scholar]
- 17.Hernandez-Hernandez ME, Torres-Rasgado E, Pulido-Perez P, Nicolás-Toledo L, Martínez-Gómez M, Rodríguez-Antolín J, et al. Disordered glucose levels are Associated with Xanthine oxidase activity in overweight type 2 Diabetic women. Int J Mol Sci. 2022;23(19):11177. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Martínez-Sánchez FD, Vargas-Abonce VP, Guerrero-Castillo AP, Santos-Villavicencio ML, Eseiza-Acevedo J, Meza-Arana CE, et al. Serum uric acid concentration is associated with insulin resistance and impaired insulin secretion in adults at risk for type 2 diabetes. Prim Care Diabetes. 2021;15(2):293–9. [DOI] [PubMed] [Google Scholar]
- 19.Denimal D, Monier S, Bouillet B, Vergès B, Duvillard L. High-density lipoprotein alterations in type 2 diabetes and obesity. Metabolites. 2023;13(2):253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Bazmandegan G, Dehghani MH, Karimifard M, Kahnooji M, Balaee P, Zakeri MA, et al. Uric acid to HDL ratio: a marker for predicting incidence of metabolic syndrome in patients with type II diabetes. Nutr Metab Cardiovasc Dis. 2024;34(4):1014–20. [DOI] [PubMed] [Google Scholar]
- 21.Aktas G, Kocak MZ, Bilgin S, Atak BM, Duman TT, Kurtkulagi O. Uric acid to HDL cholesterol ratio is a strong predictor of diabetic control in men with type 2 diabetes mellitus. Aging Male. 2020;23(5):1098–102. [DOI] [PubMed] [Google Scholar]
- 22.Thacker SG, Zarzour A, Chen Y, Alcicek MS, Freeman LA, Sviridov DO, et al. High-density lipoprotein reduces inflammation from cholesterol crystals by inhibiting inflammasome activation. Immunology. 2016;149(3):306–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Xu J, Du X, Zhang S, Zang X, Xiao Z, Su R, et al. Diagnostic value of uric acid to high-density lipoprotein cholesterol ratio in abdominal aortic aneurysms. Ann Med. 2024;56(1):2357224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Deng F, Jia F, Sun Y, Zhang L, Han J, Li D, et al. Predictive value of the serum uric acid to high-density lipoprotein cholesterol ratio for culprit plaques in patients with acute coronary syndrome. BMC Cardiovasc Disord. 2024;24(1):155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Tan M, Yang S, Xu X. High-density lipoprotein cholesterol and carcinogenesis. Trends Endocrinol Metab. 2023;34(5):303–13. [DOI] [PubMed] [Google Scholar]
- 26.Yang Y, Zhang J, Jia L, Su J, Ma M, Lin X. Uric acid to high-density lipoprotein cholesterol ratio predicts adverse cardiovascular events in patients with coronary chronic total occlusion. Nutr Metab Cardiovasc Dis. 2023;33(12):2471–8. [DOI] [PubMed] [Google Scholar]
- 27.Aydın C, Emlek N. The relationship between uric acid to high-density lipoprotein cholesterol ratio and collateral index in patients with chronic total occlusion. Kardiologiia. 2021;61(9):61–5. [DOI] [PubMed] [Google Scholar]
- 28.Liu R, Peng Y, Wu H, Diao X, Ye H, Huang X, et al. Uric acid to high-density lipoprotein cholesterol ratio predicts cardiovascular mortality in patients on peritoneal dialysis. Nutr Metab Cardiovasc Dis. 2021;31(2):561–9. [DOI] [PubMed] [Google Scholar]
- 29.Zoppini G, Targher G, Negri C, Stoico V, Perrone F, Muggeo M, et al. Elevated serum uric acid concentrations independently predict cardiovascular mortality in type 2 diabetic patients. Diabetes Care. 2009;32(9):1716–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Xu Y, Zhu J, Gao L, Liu Y, Shen J, Shen C, et al. Hyperuricemia as an independent predictor of vascular complications and mortality in type 2 diabetes patients: a meta-analysis. PLoS ONE. 2013;8(10):e78206. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Özpamuk-Karadeniz F, Karadeniz Y, Kaya A, Altay S, Can G, Onat A. Low serum uric acid predicts risk of a Composite Disease Endpoint. Med (Kaunas). 2021;57(4):361. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Baker JF, Weber DR, Neogi T, George MD, Long J, Helget LN, et al. Associations between low serum urate, body composition, and Mortality. Arthritis Rheumatol. 2023;75(1):133–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Khera AV, Cuchel M, de la Llera-Moya M, Rodrigues A, Burke MF, Jafri K, et al. Cholesterol efflux capacity, high-density lipoprotein function, and atherosclerosis. N Engl J Med. 2011;364(2):127–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Palatini P, Virdis A, Masi S, Mengozzi A, Casiglia E, Tikhonoff V, et al. Hyperuricemia increases the risk of cardiovascular mortality associated with very high HDL-cholesterol level. Nutr Metab Cardiovasc Dis. 2023;33(2):323–30. [DOI] [PubMed] [Google Scholar]
- 35.Li D, Yuan S, Deng Y, Wang X, Wu S, Chen X, et al. The dysregulation of immune cells induced by uric acid: mechanisms of inflammation associated with hyperuricemia and its complications. Front Immunol. 2023;14:1282890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Liu C, Dhindsa D, Almuwaqqat Z, Ko YA, Mehta A, Alkhoder AA, et al. Association between High-Density Lipoprotein Cholesterol Levels and adverse Cardiovascular outcomes in high-risk populations. JAMA Cardiol. 2022;7(7):672–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Madsen CM, Varbo A, Nordestgaard BG. Extreme high high-density lipoprotein cholesterol is paradoxically associated with high mortality in men and women: two prospective cohort studies. Eur Heart J. 2017;38(32):2478–86. [DOI] [PubMed] [Google Scholar]
- 38.Pirillo A, Catapano AL. Increasing high-density lipoprotein cholesterol levels for cardiovascular benefit: the end of a dream? Eur J Prev Cardiol. 2019;26(5):531–2. [DOI] [PubMed] [Google Scholar]
- 39.Ma P, Zhao M, Li Y, Zhang G, Ma Y, Shi Y, et al. The protective effects of uric acid against myocardial ischemia via the Nrf2 pathway. Eur J Pharmacol. 2023;959:176062. [DOI] [PubMed] [Google Scholar]
- 40.Waring WS, McKnight JA, Webb DJ, Maxwell SR. Uric acid restores endothelial function in patients with type 1 diabetes and regular smokers. Diabetes. 2006;55(11):3127–32. [DOI] [PubMed] [Google Scholar]
- 41.Amaro S, Urra X, Gómez-Choco M, Obach V, Cervera A, Vargas M, et al. Uric acid levels are relevant in patients with stroke treated with thrombolysis. Stroke. 2011;42(1 Suppl):S28–32. [DOI] [PubMed] [Google Scholar]
- 42.Lamon-Fava S. High-density lipoproteins: effects of alcohol, estrogen, and phytoestrogens. Handb Exp Pharmacol. 2005;170:297–323. [DOI] [PubMed] [Google Scholar]
- 43.Rubinow KB, Vaisar T, Tang C, Matsumoto AM, Heinecke JW, Page ST. Testosterone replacement in hypogonadal men alters the HDL proteome but not HDL cholesterol efflux capacity. J Lipid Res. 2012;53(7):1376–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Haj Mouhamed D, Ezzaher A, Neffati F, Douki W, Gaha L, Najjar MF. Effect of cigarette smoking on plasma uric acid concentrations. Environ Health Prev Med. 2011;16(5):307–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Nieradko-Iwanicka B. The role of alcohol consumption in pathogenesis of gout. Crit Rev Food Sci Nutr. 2022;62(25):7129–37. [DOI] [PubMed] [Google Scholar]
- 46.Stadler JT, Lackner S, Mörkl S, Trakaki A, Scharnagl H, Borenich A, et al. Obesity affects HDL metabolism, composition, and subclass distribution. Biomedicines. 2021;9(3):242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Richette P, Clerson P, Périssin L, Flipo RM, Bardin T. Revisiting comorbidities in gout: a cluster analysis. Ann Rheum Dis. 2015;74(1):142–7. [DOI] [PubMed] [Google Scholar]
- 48.Norhammar A. Diabetes and cardiovascular mortality: the impact of sex. Lancet Diabetes Endocrinol. 2018;6(7):517–9. [DOI] [PubMed] [Google Scholar]
- 49.Gustafsson I, Brendorp B, Seibaek M, Burchardt H, Hildebrandt P, Køber L, et al. Influence of diabetes and diabetes-gender interaction on the risk of death in patients hospitalized with congestive heart failure. J Am Coll Cardiol. 2004;43(5):771–7. [DOI] [PubMed] [Google Scholar]
- 50.Lopez-de-Andres A, Jimenez-Garcia R, Hernández-Barrera V, de Miguel-Yanes JM, Albaladejo-Vicente R, Villanueva-Orbaiz R, et al. Are there sex differences in the effect of type 2 diabetes in the incidence and outcomes of myocardial infarction? A matched-pair analysis using hospital discharge data. Cardiovasc Diabetol. 2021;20(1):81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Lukkunaprasit T, Rattanasiri S, Turongkaravee S, Suvannang N, Ingsathit A, Attia J, et al. The association between genetic polymorphisms in ABCG2 and SLC2A9 and urate: an updated systematic review and meta-analysis. BMC Med Genet. 2020;21(1):210. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Heitman K, Fanelli SM, Taylor CA. Contribution of snacks to Dietary Intakes stratified by Race/Ethnicity in the United States. J Am Nutr Assoc. 2023;42(1):85–93. [DOI] [PubMed] [Google Scholar]
- 53.Schofield J, Ho J, Soran H. Cardiovascular Risk in Type 1 diabetes Mellitus. Diabetes Ther. 2019;10(3):773–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Antonopoulos AS, Oikonomou EK, Antoniades C, Tousoulis D. From the BMI paradox to the obesity paradox: the obesity-mortality association in coronary heart disease. Obes Rev. 2016;17(10):989–1000. [DOI] [PubMed] [Google Scholar]
- 55.Chen J, Pan Y, Gao Q, Zhuang R, Ma L. Association of hyperuricemia with coronary heart disease: protocol for an updated systematic review and dose-response meta-analysis. PLoS ONE. 2024;19(8):e0308719. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Material 1: Table S1. Relationship between UA and HDL levels with ALL-cause mortality risk in diabetic populations.
Supplementary Material 2. Table S2. Relationship between UA and HDL levels with CVD-cause mortality risk in diabetic populations.
Supplementary Material 3: Table S3. Original data information.
Supplementary Material 4: Figure S1. Kaplan-Meier curves illustrating the relationship between UA levels and mortality risk in diabetic populations:(A) Association between UA levels and ALL-cause mortality (B) Association between UA levels and CVD-cause mortality.
Supplementary Material 5: Figure S2. Kaplan-Meier curves illustrating the relationship between HDL levels and mortality risk in diabetic populations:(A) Association between HDL levels and ALL-cause mortality (B) Association between HDL levels and CVD-cause mortality
Supplementary Material 6: Figure S3. Restricted Cubic Spline curves showing the relationship between UA levels and mortality risk in diabetic populations: (A) Association between UA levels and ALL-cause mortality (B) Association between UA levels and CVD-cause mortality.
Data Availability Statement
All data generated or analysed during this study are included in this published article and its supplementary information files.