Visual Abstract
Keywords: diabetes mellitus, type 2, diabetic kidney disease, GFR, cardiovascular diseases
Abstract
Key Points
Patients with type 2 diabetes mellitus who had glomerular hyperfiltration (GHF) were younger and had fewer comorbid metabolic disorders, despite poor glycemic control.
The relationship between eGFR and incident cardiovascular disease showed an inverted J-shaped pattern, which was highest among low filtration, followed by GHF.
GHF was associated with a higher risk of cardiovascular disease, particularly myocardial infarction and heart failure.
Background
The effects of glomerular hyperfiltration (GHF) on cardiovascular disease (CVD) risk in patients with type 2 diabetes mellitus were explored.
Methods
This retrospective cohort study enrolled 1,952,053 patients with type 2 diabetes mellitus from the Korean National Health Insurance Service database between 2015 and 2016. On the basis of age- and sex-specific eGFR percentiles, patients were classified into five groups: <5 (low filtration), 5–40, 40–60, 60–95, and >95 (GHF). Patients with incident CVD (myocardial infarction [MI], stroke, and hospitalization for heart failure) were followed up until December 2022.
Results
CVD occurred in 214,111 patients (11%). The incidence rates were 36.1, 20.8, 18.3, 18.7, and 19.3 per 1000 person-years for the eGFR groups, respectively. Low filtration (hazard ratio [HR], 1.56; 95% confidence interval [CI], 1.53 to 1.59) and GHF (HR, 1.13; 95% CI, 1.10 to 1.15) were associated with higher CVD risk adjusted for covariates than that of the eGFR 40–60 percentile, showing an inverted J-shaped relationship. GHF was associated with a higher risk of MI (HR, 1.06; 95% CI, 1.01 to 1.11) and heart failure (HR, 1.17; 95% CI, 1.14 to 1.20) and with a higher risk of stroke. eGFR was associated with CVD risk across subgroups of age, sex, obesity, hypertension, and dyslipidemia. The effect of GHF on CVD may be greater in younger patients (HR=1.30, 1.17, and 1.05 in <40, 40–60, and ≥65 years old, respectively).
Conclusions
GHF was associated with CVD, particularly MI and heart failure. Screening for GHF in the early stages of type 2 diabetes mellitus may be beneficial.
Introduction
Diabetic kidney disease is a global health problem affecting 25% of Korean patients with diabetes.1 It is well established that diabetic kidney disease is an independent risk factor of cardiovascular disease (CVD) and its associated mortality, strengthening evidence for annual checkups for albuminuria and eGFR with optimal glycemic and BP control.2 However, the effect of glomerular hyperfiltration (GHF), which is an abnormally elevated GFR, on the risk of CVD in patients with diabetes is not well understood.
GHF is a preclinical stage of CKD caused by systemic metabolic disorders, such as diabetes, obesity, and hypertension.3 The prevalence of GHF in patients with type 2 diabetes mellitus ranges from 6% to 73%, with the eGFR threshold ranging from 120 to 140 ml/min per 1.72 m2.4 GHF predisposes patients to nephron damage through an imbalance in vasoactive humoral factors and tubular signals, causing a net reduction in afferent arteriolar resistance and an increase in efferent arteriolar resistance.4 It irreversibly precedes albuminuria and a decline in renal function and associates with higher risk of CVD and mortality.5,6 Indeed, there is growing evidence of an inverted J-shaped or U-shaped association between GFR and cardiovascular outcomes among healthy individuals7,8 and those with a history of CVD.9 In addition, GHF is a characteristic finding in the early stages of diabetes mellitus (DM), and its prevalence is higher with suboptimal control of metabolic factors.10 Therefore, it is important to screen patients with DM for GHF. However, to the best of our knowledge, research on whether GHF can predict the risk of CVD in patients with type 2 diabetes mellitus is lacking.6,11
For this purpose, we aimed to analyze the effects of age- and sex-specific GHF on the incidence of CVD in patients with type 2 diabetes mellitus over a 7-year period on the basis of the National Health Insurance Database from 2015 to 2022 in Korea. Subgroup analyses according to age, sex, and the presence of metabolic disorders were also performed.
Research Design and Methods
Data Source and Study Population
This retrospective cohort study used data from the National Health Insurance Service (NHIS) of Korea. The NHIS operates a mandatory public insurance program for all citizens and supports public health policies and research activities by maintaining the National Health Insurance Database.12 Before the study commenced, approval was obtained from the Institutional Review Board of Soongsil University (SSU-202202-HR-409-1). The study was conducted in accordance with the Helsinki Declaration of 1975. The requirement for informed consent was waived.
A total of 2,616,828 Korean patients with type 2 diabetes mellitus (International Classification of Diseases (ICD) codes E11–E14 with prescription of antidiabetic drugs or fasting blood glucose ≥126 mg/dl) who underwent at least one general health checkup between January 1, 2015, and December 31, 2016, were enrolled in this study (Figure 1). The following individuals were excluded from the study: (1) participants younger than 20 years (N=323); (2) participants previously diagnosed with myocardial infarction (MI) (ICD codes I21 and I22+admission; N=138,054); (3) participants previously diagnosed with stroke (ICD codes I63 and I64+admission or confirmed by brain computed tomography or magnetic resonance imaging; N=281,552); (4) participants previously diagnosed with heart failure (ICD codes I50+admission; N=121,569); (5) participants with missing data from general health checkups (N=71,731); and (6) participants who developed CVD or died within 1 year after their health checkup (N=51,546). In total, 1,952,053 participants were included in the final analysis. The participants were followed up until the date of CVD occurrence or December 31, 2022.
Figure 1.
Patient selection. CVD, cardiovascular disease; GHF, glomerular hyperfiltration; LF, low filtration; MI, myocardial infarction.
Demographic Factors and Variables
Data on age; sex; smoking, drinking, and exercise status; presence of obesity, hypertension, and dyslipidemia; duration of diabetes; use of oral hypoglycemic agents (OHAs), insulin, sodium-glucose cotransporter 2 (SGLT2) inhibitors, glucagon-like peptide-1 receptor agonists, renin-angiotensin-aldosterone system (RAAS) blockers, antiplatelet agents, and statins; anthropometric measurements; and laboratory parameters at the time of enrollment were collected.
Smoking status was classified as nonsmoker, ex-smoker, or current smoker. Drinking status was classified as non-, mild, and heavy drinker (those consuming alcohol ≥30 g/d). Regular exercise was defined as vigorous exercise >3 d/wk or moderate-intensity activity>5 d/wk. Obesity was defined as body mass index (BMI) ≥25 kg/m2. Hypertension was defined by an ICD-10 code for hypertension (I10–I13, I15) with antihypertensive medications or systolic BP ≥140 mm Hg and/or diastolic BP ≥90 mm Hg at a general health checkup. Dyslipidemia was defined using the ICD-10 code for dyslipidemia (E78), with lipid-lowering agents or total cholesterol levels of ≥240 mg/dl at a general health checkup. Diabetes duration was classified as new onset, <5, 5–10, and ≥10 years. BMI was calculated as body weight in kilograms divided by height in meters squared (kg/m2). Waist circumference was measured in the horizontal plane midway between the lowest rib and iliac crest. BP was measured twice in the sitting position, and the mean value was calculated.
Urine and venous samples were collected after an overnight fast. Proteinuria was determined by a single dipstick urinalysis and semiquantified as negative, trace, 1+, 2+, 3+, or 4+ on the basis of the color scale. According to the Kidney Disease Improving Global Outcomes guidelines, in the absence of albuminuria measurements, dipstick urinalysis may serve as a substitute, with “trace to 1+” corresponding to microalbuminuria and “≥2+” indicating macroalbuminuria.13,14 Although dipstick urinalysis is considered less accurate for assessing albuminuria, trace proteinuria has been identified as a significant predictor of all-cause and cardiovascular mortality across various categories of eGFR.15 This suggests that dipstick urinalysis may be useful for risk stratification. Serum glucose, eGFR, total cholesterol, triglyceride, high density lipoprotein cholesterol, and low density lipoprotein cholesterol levels were also measured.
Definition of GHF
eGFR was calculated using the CKD Epidemiology Collaboration equations for creatinine, without adjusting for race.16 Sex- and age-specific eGFR percentiles were assessed (Supplemental Data). Previous studies have used eGFR percentiles to identify GHF, with findings published elsewhere.8,17,18 GHF was defined as an age- and sex-specific eGFR above the 95th percentile.19 The eGFR percentiles were categorized into <5 (low filtration [LF]), 5–40, 40–60, 60–95, and ≥95 (GHF) percentile groups for further analysis.
Outcomes
The primary outcome of this study was the incidence of CVD. CVD included MI (ICD codes I21 and I22+admission), stroke (ICD codes I63 and I64+admission or confirmed by brain computed tomography or magnetic resonance imaging), and hospitalization because of heart failure (ICD codes I50+admission).20 Those who were previously diagnosed with MI, stroke, or heart failure; or those diagnosed with CVD; or those who died within 1 year of their health checkup were excluded from the study. Incident CVD was followed up until December 31, 2022.
Statistical Analyses
Differences in demographics according to eGFR percentile groups were evaluated using ANOVA test for continuous variables and chi-square test for categorical variables. The incidence of CVD was expressed as the number of events per 1000 person-years. The follow-up duration was calculated on the basis of the time at which CVD was diagnosed. For patients who experienced two or more CVDs, the follow-up duration was determined on the basis of the first event. The hazard ratios (HRs) for CVD were analyzed using a Cox proportional hazards model, which was adjusted as follows: model 1, unadjusted; model 2, age, sex, smoking, drinking, regular exercise, BMI, hypertension, and dyslipidemia; and model 3 included all variables in model 2 plus the use of three or more OHAs, insulin, SGLT2 inhibitors, GLP1-RAs, RAAS blockers, antiplatelet agents, statins, serum glucose levels, and the presence of proteinuria. Subdistribution HRs for CVD were analyzed using the Fine–Gray model to account for the competing risk of death. All analyses were performed using statistical analysis system software (version 9.4; statistical analysis system Institute, Cary, NC), and a P value of 0.05 was considered significant.
Results
Baseline Characteristics
In total, 1,952,053 participants were included in this study. The mean age was 57.7±11.6 years (range, 20–104), and the male-to-female ratio was 1.7:1. Baseline characteristics of participants according to eGFR groups (<5 [LF], 5–40, 40–60, 60–95, and ≥95 [GHF] percentiles) are presented in Table 1. The mean eGFR levels of each eGFR group were 49.4, 73.6, 89.1, 98.9, and 110.5 ml/min per 1.73 m2, respectively. The mean age and DM duration tended to be lower according to the sequential order of the eGFR groups. In the GHF group, despite poor lifestyle habits, such as current smoking, heavy drinking, and less exercise, the prevalence and associated parameters of obesity, hypertension, and dyslipidemia were better, which is considered an advantage of younger age. However, the fasting glucose level was the highest, and the usage of OHA ≥3 classes and insulin was second highest after LF. The prevalence of proteinuria was highest in the LF group and did not differ in the other eGFR groups. Taken together, patients with GHF were young and had fewer comorbid metabolic disorders, despite poor glycemic control.
Table 1.
Baseline characteristics of participants
| Variable | Total | eGFR Percentile | P Value | ||||
|---|---|---|---|---|---|---|---|
| <5 (LF) | 5–40 | 40–60 | 60–95 | ≥95 (GHF) | |||
| n | 1,952,053 | 96,658 (5%) | 691,052 (35%) | 384,131 (20%) | 681,380 (35%) | 98,832 (5%) | |
| eGFR, ml/min per 1.73 m2 | 86.15±17.96 | 49.36±14.72 | 73.64±11.28 | 89.13±10.69 | 98.85±9.15 | 110.5±9.54 | <0.001 |
| Age, yr | 57.7±11.6 | 58.9±12.1 | 58.3±11.7 | 57.5±11.3 | 57.2±11.6 | 55.9±11.3 | <0.001 |
| Age group, yr | <0.001 | ||||||
| <40 | 112,879 (6) | 5705 (6) | 38,379 (6) | 23,721 (6) | 39,277 (6) | 5797 (6) | |
| 40–64 | 1,301,142 (67) | 60,895 (63) | 437,870 (63) | 254,006 (66) | 475,215 (70) | 73,156 (74) | |
| ≥65 | 538,032 (28) | 30,058 (31) | 214,803 (31) | 106,404 (28) | 166,888 (25) | 19,879 (20) | |
| Sex, male | 1,218,556 (62) | 60,707 (63) | 432,817 (627) | 238,472 (62) | 426,086 (63) | 60,474 (61) | <0.001 |
| Smoking | <0.001 | ||||||
| Non | 1,027,683 (53) | 52,658 (55) | 368,498 (53) | 203,105 (53) | 352,738 (52) | 50,684 (51) | |
| Ex | 442,737 (23) | 23,331 (24) | 164,276 (24) | 86,831 (23) | 149,532 (22) | 18,767 (19) | |
| Current | 481,633 (25) | 20,669 (21) | 158,278 (23) | 94,195 (25) | 179,110 (26) | 29,381 (30) | |
| Drinking | <0.001 | ||||||
| Non | 1,048,954 (54) | 59,539 (62) | 381,906 (55) | 204,985 (53) | 352,911 (52) | 49,613 (50) | |
| Mild | 706,882 (36) | 30,184 (31) | 246,563 (36) | 140,771 (37) | 252,967 (37) | 36,397 (37) | |
| Heavy | 196,217 (10) | 6935 (7) | 62,583 (9) | 38,375 (10) | 75,502 (11) | 12,822 (13) | |
| Regular exercise | 436,693 (22) | 20,645 (21) | 157,472 (22) | 87,441 (23) | 150,651 (22) | 20,484 (21) | <0.001 |
| Obesity | 1,000,081 (51) | 53,558 (55) | 369,227 (53) | 196,200 (51) | 334,653 (49) | 46,443 (47) | <0.001 |
| Hypertension | 1,031,625 (53) | 67,776 (70) | 376,638 (55) | 196,082 (51) | 341,999 (50) | 49,130 (50) | <0.001 |
| Dyslipidemia | 1,063,029 (55) | 61,059 (63) | 383,486 (56) | 207,143 (54) | 359,864 (53) | 51,477 (52) | <0.001 |
| DM duration | <0.001 | ||||||
| New onset | 620,126 (32) | 23,531 (24) | 228,352 (33) | 126,032 (33) | 212,411 (31) | 29,800 (30) | |
| <5 yr | 543,888 (28) | 20,256 (21) | 183,699 (27) | 109,238 (28) | 200,500 (29) | 30,195 (31) | |
| 5–10 yr | 370,376 (19) | 18,164 (19) | 128,396 (19) | 72,439 (19) | 131,997 (19) | 19,380 (20) | |
| ≥10 yr | 417,663 (21) | 34,707 (36) | 150,605 (22) | 76,422 (20) | 136,472 (20) | 19,457 (20) | |
| OHA ≥3 classes | 424,758 (22) | 27,553 (29) | 149,238 (22) | 80,132 (21) | 145,611 (21) | 22,224 (23) | <0.001 |
| Insulin | 125,580 (6) | 14,450 (15) | 41,902 (6) | 21,245 (6) | 40,958 (6) | 7025 (7) | <0.001 |
| SGLT2 inhibitors | 39,194 (2) | 1867 (2) | 12,491 (2) | 7496 (2) | 14,657 (2) | 2683 (3) | <0.001 |
| GLP-1Ras | 671 (0.03) | 62 (0.06) | 200 (0.03) | 104 (0.03) | 260 (0.04) | 45 (0.05) | <0.001 |
| RAAS blockers | 732,591 (38) | 57,181 (59) | 275,735 (40) | 136,179 (36) | 231,196 (34) | 32,300 (33) | <0.001 |
| Antiplatelet agents | 391,089 (20) | 26,870 (28) | 145,866 (21) | 73,978 (19) | 127,079 (19) | 17,296 (18) | <0.001 |
| Statins | 759,427 (39) | 44,142 (46) | 268,103 (39) | 147,995 (39) | 261,659 (38) | 37,528 (38) | <0.001 |
| BMI, kg/m2 | 25.4±3.6 | 25.7±3.6 | 25.5±3.5 | 25.3±3.5 | 25.2±3.6 | 25.1±3.8 | <0.001 |
| Waist circumference, cm | 86±9.0 | 87.1±9.2 | 86.3±8.9 | 85.9±8.9 | 85.7±9.0 | 85.4±9.4 | <0.001 |
| Systolic BP, mm Hg | 128.1±14.8 | 128.9±16.1 | 128.0±14.8 | 128.0±14.7 | 128.2±14.7 | 128.5±14.9 | <0.001 |
| Diastolic BP, mm Hg | 78.3±9.9 | 77.9±10.4 | 78.3±9.9 | 78.4±9.8 | 78.4±9.9 | 78.5±10.1 | <0.001 |
| Proteinuria | 125,023 (6) | 19,806 (21) | 44,769 (7) | 19,807 (5) | 34,933 (5) | 5708 (6) | <0.001 |
| Urine protein | <0.001 | ||||||
| Negative | 1,746,630 (90) | 72,120 (75) | 619,606 (90) | 349,589 (91) | 616,993 (91) | 88,322 (89) | |
| Trace | 80,400 (4) | 4732 (5) | 26,677 (4) | 14,735 (4) | 29,454 (4) | 4802 (5) | |
| +1 | 72,073 (4) | 7740 (8) | 25,992 (4) | 12,443 (3) | 22,220 (3) | 3678 (4) | |
| +2 | 36,819 (2) | 6994 (7) | 13,487 (22) | 5390 (1) | 9458 (1) | 1490 (2) | |
| +3 | 12,506 (0.6) | 3776 (4) | 4196 (1) | 1542 (0.4) | 2562 (0.4) | 430 (0.4) | |
| +4 | 3625 (0.2) | 1296 (1) | 1094 (0.2) | 432 (0.1) | 693 (0.1) | 110 (0.1) | |
| Fasting glucose, mg/dl | 146.2±45.7 | 144.2±53.4 | 144.7±44.3 | 145.9±44.3 | 147.5±46.1 | 151.3±49.7 | <0.001 |
| Total cholesterol, mg/dl | 188.0±43.6 | 185.6±46.2 | 189.2±43.8 | 188.6±43.2 | 187.0±43.1 | 186.3±43.9 | <0.001 |
| Triglyceride, mg/dl | 139.4 (139.3 to 139.5) | 151.1 (150.6 to 151.7) | 141.3 (141.2 to 141.5) | 138.1 (137.8 to 138.3) | 136.7 (136.5 to 136.9) | 138.2 (137.7 to 138.7) | <0.001 |
| HDL-C, mg/dl | 51.3±14.9 | 48.8±14.8 | 50.9±14.0 | 51.5±15.1 | 51.9±15.6 | 52.2±14.7 | <0.001 |
| LDL-C, mg/dl | 105.1±38.5 | 102.8±40.1 | 106.3±38.7 | 105.9±38.3 | 104.2±38.1 | 102.6±38.4 | <0.001 |
Glomerular hyperfiltration is defined as an eGFR percentile above 95. Low filtration is defined as an eGFR percentile below five. BMI, body mass index; DM, diabetes mellitus; GHF, glomerular hyperfiltration; GLP-1RAs, glucagon-like peptide-1 receptor agonists; HDL-C, high density lipoprotein cholesterol; LDL-C, low density lipoprotein cholesterol; LF, low filtration; OHA, oral hypoglycemic agent; RAAS, renin-angiotensin-aldosterone system; SGLT2, sodium-glucose cotransporter 2.
Medication data were based on the history from 1 year before the health checkup.
Continuous variables are presented as mean±SD, and categorical variables are presented as percentages. Triglyceride values are presented as geometric mean (95% confidence interval).
CVD Risk According to the GHF
During the study period (January 1, 2015, to December 31, 2022), 214,111 patients (11%) experienced CVD. The incidence rates of MI, stroke, and heart failure were 3% (n=49,153), 3% (n=54,872), and 8% (n=153,439), respectively.
The incidence rate and risk of CVD according to the eGFR percentile groups are presented in Table 2. The incidence rates of CVD were 36.1, 20.8, 18.3, 18.7, and 19.3 per 1000 person-years according to the sequential order of the eGFR groups. In the crude analysis (model 1), compared with the eGFR 40–60 percentile (reference), both lower and higher eGFRs were associated with a higher risk of CVD. After adjustments (model 3), the bidirectional increasing risk of CVD became more prominent, forming an inverted J-shaped pattern. The HRs were 1.56 (95% confidence interval [CI], 1.53 to 1.59), 1.07 (95% CI, 1.05 to 1.08), 1 (reference), 1.03 (95% CI, 1.01 to 1.04), and 1.13 (95% CI, 1.10 to 1.15) across the respective eGFR groups (P < 0.001). Competing risk analysis using the Fine–Gray model showed similar results.
Table 2.
Incidence rate and risk of cardiovascular disease according to the eGFR percentile
| Outcome | eGFR Percentile | Event/Total | Incidence Rate/1000 PY | Model 1 | Model 2 | Model 3 | Model 3, sHRa |
|---|---|---|---|---|---|---|---|
| CVD | <5 (LF) | 17,909/96,658 | 36.1 | 2.00 (1.96 to 2.02) | 1.79 (1.76 to 1.83) | 1.56 (1.53 to 1.59) | 1.51 (1.49 to 1.54) |
| 5–40 | 78,085/691,052 | 20.8 | 1.14 (1.13 to 1.16) | 1.08 (1.07 to 1.09) | 1.07 (1.05 to 1.08) | 1.06 (1.05 to 1.07) | |
| 40–60 | 38,378/384,131 | 18.3 | 1 (ref.) | 1 (ref.) | 1 (ref.) | 1 (ref.) | |
| 60–95 | 69,391/681,380 | 18.7 | 1.03 (1.01 to 1.04) | 1.03 (1.01 to 1.04) | 1.03 (1.01 to 1.04) | 1.02 (1.01 to 1.04) | |
| ≥95 (GHF) | 10,348/98,832 | 19.3 | 1.06 (1.04 to 1.09) | 1.14 (1.12 to 1.17) | 1.13 (1.10 to 1.15) | 1.12 (1.09 to 1.14) | |
| P value | <0.001 | <0.001 | <0.001 | <0.001 | |||
| MI | <5 (LF) | 4233/96,658 | 8.0 | 1.99 (1.92 to 2.06) | 1.80 (1.74 to 1.87) | 1.58 (1.52 to 1.64) | 1.52 (1.46 to 1.58) |
| 5–40 | 18,052/691,052 | 4.6 | 1.15 (1.12 to 1.18) | 1.10 (1.07 to 1.13) | 1.08 (1.05 to 1.11) | 1.08 (1.05 to 1.10) | |
| 40–60 | 8790/384,131 | 4.1 | 1 (ref.) | 1 (ref.) | 1 (ref.) | 1 (ref.) | |
| 60–95 | 15,809/681,380 | 4.1 | 1.02 (1.00 to 1.05) | 1.02 (1.00 to 1.05) | 1.02 (0.99 to 1.05) | 1.02 (0.99 to 1.04) | |
| ≥95 (GHF) | 2269/98,832 | 4.1 | 1.01 (0.97 to 1.06) | 1.07 (1.02 to 1.12) | 1.06 (1.01 to 1.11) | 1.05 (1.00 to 1.10) | |
| P value | <0.001 | <0.001 | <0.001 | <0.001 | |||
| Stroke | <5 (LF) | 4390/96,658 | 8.3 | 1.83 (1.77 to 1.90) | 1.62 (1.56 to 1.68) | 1.40 (1.35 to 1.46) | 1.34 (1.29 to 1.39) |
| 5–40 | 20,532/691,052 | 5.3 | 1.16 (1.13 to 1.19) | 1.09 (1.06 to 1.11) | 1.07 (1.04 to 1.10) | 1.06 (1.04 to 1.09) | |
| 40–60 | 9891/384,131 | 4.6 | 1 (ref.) | 1 (ref.) | 1 (ref.) | 1 (ref.) | |
| 60–95 | 17,576/681,380 | 4.6 | 1.01 (0.98 to 1.03) | 1.00 (0.97 to 1.03) | 1.00 (0.97 to 1.02) | 1.00 (0.97 to 1.02) | |
| ≥95 (GHF) | 2483/98,832 | 4.5 | 0.99 (0.94 to 1.03) | 1.06 (1.02 to 1.11) | 1.05 (1.00 to 1.10) | 1.04 (1.00 to 1.09) | |
| P value | <0.001 | <0.001 | <0.001 | <0.001 | |||
| Heart failure | <5 (LF) | 13,898/96,658 | 27.4 | 2.17 (2.13 to 2.22) | 1.93 (1.89 to 1.97) | 1.65 (1.62 to 1.68) | 1.59 (1.56 to 1.62) |
| 5–40 | 55,612/691,052 | 14.6 | 1.15 (1.13 to 1.16) | 1.08 (1.07 to 1.10) | 1.06 (1.05 to 1.08) | 1.06 (1.04 to 1.07) | |
| 40–60 | 27,133/384,131 | 12.7 | 1 (ref.) | 1 (ref.) | 1 (ref.) | 1 (ref.) | |
| 60–95 | 49,270/681,380 | 13.1 | 1.03 (1.02 to 1.05) | 1.03 (1.02 to 1.05) | 1.03 (1.02 to 1.05) | 1.03 (1.02 to 1.05) | |
| ≥95 (GHF) | 7526/98,832 | 13.9 | 1.09 (1.07 to 1.12) | 1.18 (1.15 to 1.21) | 1.17 (1.14 to 1.20) | 1.16 (1.13 to 1.19) | |
| P value | <0.001 | <0.001 | <0.001 | <0.001 |
Glomerular hyperfiltration is defined as an eGFR percentile above 95. Low filtration is defined as an eGFR percentile below 5. CVD, cardiovascular disease; GHF, glomerular hyperfiltration; LF, low filtration; MI, myocardial infarction; PY, person-years; sHR, subdistribution hazard ratio.
Model 1, nonadjusted; model 2, adjusted for age, sex, smoking, drinking, regular exercise, body mass index, hypertension, and dyslipidemia; model 3, model 2+oral hypoglycemic agents ≥3, insulin use, sodium-glucose cotransporter 2 inhibitors, glucagon-like peptide-1 receptor agonists, renin-angiotensin–aldosterone system blockers, antiplatelet agents, statins, serum glucose levels, and presence of proteinuria.
Fine and Gray competing risk regression analysis was performed.
In particular, MI and heart failure were bidirectionally associated with higher risks after adjustments (Table 2), with both LF and GHF associating with higher risk. Conversely, the risk of stroke was higher in LF but GHF was only marginally associated with a higher risk.
Subgroup Analysis
The risk of GHF on CVD according to age (<40, 40–65, and ≥65); sex; and the presence of obesity, hypertension, and dyslipidemia was analyzed (Figure 2). The inverted J-shaped pattern remained consistent, with a P value for interaction of<0.001, indicating that eGFR affects the risk of CVD across different subgroups. Notably, eGFR differently affected the risk of CVD according to age: In patients younger than 40 years, the effect of LF and GHF on the risk of CVD was similar (HR, 1.34; 95% CI, 1.17 to 1.55; HR, 1.30; 95% CI, 1.12 to 1.50, respectively). GHF was associated with higher HRs for CVD outcomes in men compared with women and in those without obesity, hypertension, or dyslipidemia than in those with these comorbidities.
Figure 2.
Incidence of CVD according to eGFR percentile bullets and bars indicate HRs and 95% CIs, respectively. The Cox proportional hazards model was adjusted for age, sex, smoking, drinking, regular exercise, BMI, hypertension, dyslipidemia, OHA ≥3 agents, insulin use, SGLT2 inhibitors, GLP1-RAs, RAAS blockers, antiplatelet agents, statins, serum glucose levels, and presence of proteinuria. BMI, body mass index; CI, confidence interval; GLP-1RA, glucagon-like peptide-1 receptor agonist; HR, hazard ratio; OHA, oral hypoglycemic agent; RAAS: renin-angiotensin-aldosterone system; SGLT2, sodium-glucose cotransporter 2.
The effects of eGFR on the risk of MI, stroke, and heart failure according to subgroups are presented in Supplemental Figures 1–3. Notably, among those younger than 40 years, the effect of GHF on incident MI (HR, 1.36; 95% CI, 1.03 to 1.81) was higher than that of LF (HR, 1.02; 95% CI, 0.75 to 1.38), despite wide CIs (possibly owing to the small number of participants; Supplemental Figure 1). This pattern was similar for heart failure development (HR, 1.30; 95% CI, 1.10 to 1.55 for GHF versus HR, 1.42; 95% CI, 1.21 to 1.67 for LF; Supplemental Figure 3).
Discussion
This study explored the effect of the GHF on incident CVD in Korean patients with type 2 diabetes mellitus using NHIS data from 2015 to 2022. The GHF was defined as age- and sex-specific eGFR percentiles ≥95. Although the participants with GHF were young, had a short duration of diabetes, and had a low prevalence of comorbidities, they showed poor glycemic control. After adjustment, the effect of eGFR on incident CVD showed an inverted J-shaped pattern, which was highest among LF (HR, 1.56), followed by GHF (HR, 1.13). GHF was associated with higher risk of MI and heart failure by 1.06 and 1.17 times, respectively, compared with that of the eGFR 40–60 percentile group and was marginally associated with a higher risk of stroke by 1.05 times. In particular, the effect of the GHF was prominent in patients younger than 40 years.
It is assumed that GHF is not a kidney pathophysiology alone but a marker for the presence and severity of type 2 diabetes mellitus, obesity, and hypertension. In a middle-aged nondiabetic population in Norway, the presence of metabolic syndrome was associated with GHF.21 According to a cohort of approximately 200,000 patients with type 2 diabetes mellitus in Italy, the prevalence of GHF was 0.6%, which was more frequent in men; those with lower age and BMI; and those with suboptimal control of glucose, BP, and lipids.10 Similar to previous studies, we showed that participants with GHF were young and had few comorbid metabolic syndrome factors (47%–52%) but had poor glucose control.
In this study, LF can be considered CKD stage 3a (mean eGFR 49±15 ml/min per 1.73 m2). It is noteworthy that CKD stage 3a was associated with the highest CVD risk, followed by GHF (HR, 1.56 and 1.12, respectively). Previous studies among healthy middle-aged individuals in Canada and Italy have consistently reported that GHF was associated with a higher CVD risk in participants with CKD stage 3a.7,8 After an average follow-up of 12 years in the general Korean population, GHF was associated with a higher risk of CVD and associated mortality by 1.2 and 1.7 times, respectively.22 Moreover, among participants with a history of CVD, those with GHF were younger, had a higher prevalence of diabetes, and were at a higher risk of CVD and mortality, followed by those with CKD stage 3.9 Another 21-year longitudinal study of Italian patients with type 1 and type 2 diabetes reported that GHF and CKD stage 3 were similarly associated with a higher risk of worsening kidney function and cardiorenal mortality.18 GHF was also an independent factor for all-cause mortality comparable with CKD 3b in patients with type 2 diabetes mellitus after an average follow-up period of 7 years.23
To the best of our knowledge, no prior studies in patients with type 2 diabetes mellitus have explored the effect of GHF on the development of CVD, including its individual components, such as MI, stroke, and heart failure. Therefore, in this nationwide population-based study, we found that GHF was associated with a higher risk of CVD, particularly MI and heart failure. Sensitivity analyses evaluating the longitudinal trajectory of eGFR for incident CVD indicated that progression to GHF from the 60th to 95th percentile of eGFR over 2 years, or persistent GHF, was associated with a higher risk of developing CVD, particularly heart failure (Supplemental Table 1). In addition, subgroup analyses suggested that the effect of GHF on CVD risk may be higher at a younger age; accordingly, the effect of CKD3a and GHF on CVD risk was similar (HR, 1.34 and 1.30, respectively) among those younger than 40 years. This trend was consistent for incident heart failure, and GHF had a greater effect on incident MI compared with CKD stage 3a. Overall, GHF may serve as an independent risk factor of CVD, especially MI and heart failure, in patients with type 2 diabetes mellitus younger than 40 years.
Multifactorial mechanisms leading to GHF development in patients with diabetes have been suggested.4,24 The vascular theory states that diabetes leads to a greater reduction in resistance in afferent arterioles than in efferent arterioles, altering renal plasma flow and the filtration fraction.25 The tubular theory states that hyperglycemia stimulates the sodium-glucose cotransporter in the proximal tubule, inhibiting tubuloglomerular feedback, which in turn vasodilates the afferent arteriole.26–28 Maladaptive activation of the RAAS and sympathetic nervous systems also leads to GHF,29 which may result in arterial stiffness, impaired BP variation, and higher CVD and associated mortality.6–8,30 On the basis of these potential mechanisms, further studies on the treatment of GHF with SGLT2 inhibitors and RAAS blockade are warranted.
This study has some limitations. First, we determined the presence of GHF by calculating the eGFR. However, higher eGFR may result from decreased creatinine levels due to factors such as sarcopenia, obesity, inflammation, or high protein intake, rather than true GHF.31–34 However, considering that GFR can be overestimated in older individuals,35 the fact that the effect of GHF was more pronounced in younger participants in this study is encouraging. All analyses were adjusted for BMI and physical activity to reduce potential bias. Second, potential selection bias may be present owing to the requirement of one general health checkup between 2015 and 2016. However, general health checkups are mandatory for Koreans every 2 years starting at the age of 40 years. In fact, the participation rate for these checkups exceeded 75% among the 35 million population during 2015 and 2016,36 making significant selection bias unlikely. Finally, baseline assessments were not repeated during follow-ups, preventing us from evaluating changes in renal function, medication use, or other risk factors over time. Future longitudinal studies are required to investigate the effect of GHF trajectory progression on CVD outcomes.
Despite these limitations, this study has several strengths. To the best of our knowledge, this is the first study to show that GHF is associated with higher CVD risk in patients with type 2 diabetes mellitus. Second, compared with the majority of previous studies that used a single threshold for GHF,19 we used age- and sex-adapted GHF thresholds. Third, we collected and adjusted for the use of medications with renoprotective effects, such as RAAS blockers, SGLT2 inhibitors, glucagon-like peptide-1 receptor agonists, antiplatelet agents, and statins, in all our analyses. Finally, this study used the most representative and high-quality health data sources in Korea.
In conclusion, GHF is a risk factor of CVD, particularly MI and heart failure, with young adults aged younger than 40 years may be at a higher risk compared with older individuals. Early identification of GHF in patients with type 2 diabetes mellitus may offer an opportunity for preventive strategies to lower the risk of CVD.
Supplementary Material
Disclosures
Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/CJN/C142.
Funding
N.H. Kim: National Research Foundation (NRF-2019M3E5D3073102 and NRF-2023R1A2C200347), National IT Industry Promotion Agency (No. S0252-21-1001), and Korea Health Industry Development Institute (H123C0679).
Author Contributions
Conceptualization: Nan Hee Kim.
Data curation: Kyung-Do Han.
Formal analysis: Kyung-Do Han.
Funding acquisition: Nan Hee Kim.
Methodology: Kyung-Do Han.
Supervision: Nan Hee Kim.
Writing – original draft: Seung Min Chung.
Writing – review & editing: Kyung-Do Han, Inha Jung, Nan Hee Kim, Da Young Lee, Jun Sung Moon, So Young Park, Ji A. Seo, Ji Hee Yu.
Data Sharing Statement
Partial restrictions to the data and/or materials apply. Raw data are available to researchers upon relevant request, with approval by the Korean National Health Insurance Sharing Service.
Supplemental Material
This article contains the following supplemental material online at http://links.lww.com/CJN/C141.
Supplemental Figure 1. Incidence of MI according to eGFR percentile.
Supplemental Figure 2. Incidence of stroke according to eGFR percentile.
Supplemental Figure 3. Incidence of heart failure according to eGFR percentile.
Supplemental Table 1. Sensitivity analysis: Incidence rate and risk of CVD according to the repeated measures of eGFR percentile.
Supplemental Data. Sex- and age-specific eGFR cutoffs.
References
- 1.Kim NH, Seo MH, Jung JH, Han KD, Kim MK, Kim NH. 2023 diabetic kidney disease fact sheet in Korea. Diabetes Metab J. 2024;48(3):463–472. doi: 10.4093/dmj.2023.0310 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Choi JH Lee KA Moon JH, et al. 2023 clinical practice guidelines for diabetes mellitus of the Korean diabetes association. Diabetes Metab J. 2023;47(5):575–594. doi: 10.4093/dmj.2023.0282 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Alicic RZ, Rooney MT, Tuttle KR. Diabetic kidney disease: challenges, progress, and possibilities. Clin J Am Soc Nephrol. 2017;12(12):2032–2045. doi: 10.2215/CJN.11491116 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tonneijck L Muskiet MH Smits MM, et al. Glomerular hyperfiltration in diabetes: mechanisms, clinical significance, and treatment. J Am Soc Nephrol. 2017;28(4):1023–1039. doi: 10.1681/ASN.2016060666 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Magee GM, Bilous RW, Cardwell CR, Hunter SJ, Kee F, Fogarty DG. Is hyperfiltration associated with the future risk of developing diabetic nephropathy? A meta-analysis. Diabetologia. 2009;52(4):691–697. doi: 10.1007/s00125-009-1268-0 [DOI] [PubMed] [Google Scholar]
- 6.Kanbay M Ertuglu LA Afsar B, et al. Renal hyperfiltration defined by high estimated glomerular filtration rate: a risk factor for cardiovascular disease and mortality. Diabetes Obes Metab. 2019;21(11):2368–2383. doi: 10.1111/dom.13831 [DOI] [PubMed] [Google Scholar]
- 7.Reboldi G Verdecchia P Fiorucci G, et al. Glomerular hyperfiltration is a predictor of adverse cardiovascular outcomes. Kidney Int. 2018;93(1):195–203. doi: 10.1016/j.kint.2017.07.013 [DOI] [PubMed] [Google Scholar]
- 8.Dupuis ME, Nadeau-Fredette AC, Madore F, Agharazii M, Goupil R. Association of glomerular hyperfiltration and cardiovascular risk in middle-aged healthy individuals. JAMA Netw Open. 2020;3(4):e202377. doi: 10.1001/jamanetworkopen.2020.2377 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Inrig JK Gillespie BS Patel UD, et al. Risk for cardiovascular outcomes among subjects with atherosclerotic cardiovascular disease and greater-than-normal estimated glomerular filtration rate. Clin J Am Soc Nephrol. 2007;2(6):1215–1222. doi: 10.2215/CJN.00930207 [DOI] [PubMed] [Google Scholar]
- 10.Trico D Fadini GP Morieri ML, et al. Prevalence and risk factors of glomerular hyperfiltration in adults with type 2 diabetes: a population-based study. Diabetes Metab Res Rev. 2024;40(3):e3782. doi: 10.1002/dmrr.3782 [DOI] [PubMed] [Google Scholar]
- 11.Adeva-Andany MM, Fernandez-Fernandez C, Carneiro-Freire N, Castro-Quintela E, Pedre-Pineiro A, Seco-Filgueira M. Insulin resistance underlies the elevated cardiovascular risk associated with kidney disease and glomerular hyperfiltration. Rev Cardiovasc Med. 2020;21(1):41–56. doi: 10.31083/j.rcm.2020.01.5102 [DOI] [PubMed] [Google Scholar]
- 12.Cheol Seong S Kim YY Khang YH, et al. Data resource profile: the national health information database of the national health insurance Service in South Korea. Int J Epidemiol. 2017;46(3):799–800. doi: 10.1093/ije/dyw253 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.KDIGO. Chapter 1: definition and classification of CKD. Kidney Int Suppl (2011). 2013;3(1):19–62. doi: 10.1038/kisup.2012.64 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sumida K Nadkarni GN Grams ME, et al. Conversion of urine protein-creatinine ratio or urine dipstick protein to urine albumin-creatinine ratio for use in chronic kidney disease screening and prognosis: an individual participant-based meta-analysis. Ann Intern Med. 2020;173(6):426–435. doi: 10.7326/M20-0529 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Matsushita K van der Velde M Astor BC, et al.; Chronic Kidney Disease Prognosis Consortium, Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet. 2010;375(9731):2073–2081. doi: 10.1016/S0140-6736(10)60674-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Inker LA Eneanya ND Coresh J, et al. New creatinine- and cystatin C-based equations to estimate GFR without race. N Engl J Med. 2021;385(19):1737–1749. doi: 10.1056/NEJMoa2102953 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Palatini P. Glomerular hyperfiltration: a marker of early renal damage in pre-diabetes and pre-hypertension. Nephrol Dial Transplant. 2012;27(5):1708–1714. doi: 10.1093/ndt/gfs037 [DOI] [PubMed] [Google Scholar]
- 18.Moriconi D Sacchetta L Chiriaco M, et al. Glomerular hyperfiltration predicts kidney function decline and mortality in type 1 and type 2 diabetes: a 21-year longitudinal study. Diabetes Care. 2023;46(4):845–853. doi: 10.2337/dc22-2003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Cachat F, Combescure C, Cauderay M, Girardin E, Chehade H. A systematic review of glomerular hyperfiltration assessment and definition in the medical literature. Clin J Am Soc Nephrol. 2015;10(3):382–389. doi: 10.2215/CJN.03080314 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hicks KA Mahaffey KW Mehran R, et al. 2017 cardiovascular and stroke endpoint definitions for clinical trials. Circulation. 2018;137(9):961–972. doi: 10.1161/CIRCULATIONAHA.117.033502 [DOI] [PubMed] [Google Scholar]
- 21.Bystad EW, Stefansson VTN, Eriksen BO, Melsom T. The association between metabolic syndrome, hyperfiltration, and long-term GFR decline in the general population. Kidney Int Rep. 2023;8(9):1831–1840. doi: 10.1016/j.ekir.2023.06.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Park M, Yoon E, Lim YH, Kim H, Choi J, Yoon HJ. Renal hyperfiltration as a novel marker of all-cause mortality. J Am Soc Nephrol. 2015;26(6):1426–1433. doi: 10.1681/ASN.2014010115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Penno G Orsi E Solini A, et al. Renal hyperfiltration is independently associated with increased all-cause mortality in individuals with type 2 diabetes: a prospective cohort study. BMJ Open Diabetes Res Care. 2020;8(1):e001481. doi: 10.1136/bmjdrc-2020-001481 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Helal I, Fick-Brosnahan GM, Reed-Gitomer B, Schrier RW. Glomerular hyperfiltration: definitions, mechanisms and clinical implications. Nat Rev Nephrol. 2012;8(5):293–300. doi: 10.1038/nrneph.2012.19 [DOI] [PubMed] [Google Scholar]
- 25.Vora JP Dolben J Dean JD, et al. Renal hemodynamics in newly presenting non-insulin dependent diabetes mellitus. Kidney Int. 1992;41(4):829–835. doi: 10.1038/ki.1992.127 [DOI] [PubMed] [Google Scholar]
- 26.Vallon V, Blantz RC, Thomson S. Glomerular hyperfiltration and the salt paradox in early [corrected] type 1 diabetes mellitus: a tubulo-centric view. J Am Soc Nephrol. 2003;14(2):530–537. doi: 10.1097/01.ASN.0000051700.07403.27 [DOI] [PubMed] [Google Scholar]
- 27.Pruijm M Wuerzner G Maillard M, et al. Glomerular hyperfiltration and increased proximal sodium reabsorption in subjects with type 2 diabetes or impaired fasting glucose in a population of the African region. Nephrol Dial Transplant. 2010;25(7):2225–2231. doi: 10.1093/ndt/gfq008 [DOI] [PubMed] [Google Scholar]
- 28.Vervoort G, Veldman B, Berden JH, Smits P, Wetzels JF. Glomerular hyperfiltration in type 1 diabetes mellitus results from primary changes in proximal tubular sodium handling without changes in volume expansion. Eur J Clin Invest. 2005;35(5):330–336. doi: 10.1111/j.1365-2362.2005.01497.x [DOI] [PubMed] [Google Scholar]
- 29.Schmieder RE, Veelken R, Schobel H, Dominiak P, Mann JF, Luft FC. Glomerular hyperfiltration during sympathetic nervous system activation in early essential hypertension. J Am Soc Nephrol. 1997;8(6):893–900. doi: 10.1681/ASN.V86893 [DOI] [PubMed] [Google Scholar]
- 30.Cherney DZ Sochett EB Lai V, et al. Renal hyperfiltration and arterial stiffness in humans with uncomplicated type 1 diabetes. Diabetes Care. 2010;33(9):2068–2070. doi: 10.2337/dc10-0767 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Melsom T Fuskevag OM Mathisen UD, et al. Estimated GFR is biased by non-traditional cardiovascular risk factors. Am J Nephrol. 2015;41(1):7–15. doi: 10.1159/000371557 [DOI] [PubMed] [Google Scholar]
- 32.Chang AR, Zafar W, Grams ME. Kidney function in obesity-challenges in indexing and estimation. Adv Chronic Kidney Dis. 2018;25(1):31–40. doi: 10.1053/j.ackd.2017.10.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Schei J Stefansson VT Mathisen UD, et al. Residual associations of inflammatory markers with eGFR after accounting for measured GFR in a community-based cohort without CKD. Clin J Am Soc Nephrol. 2016;11(2):280–286. doi: 10.2215/CJN.07360715 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Cirillo M, Lombardi C, Chiricone D, De Santo NG, Zanchetti A, Bilancio G. Protein intake and kidney function in the middle-age population: contrast between cross-sectional and longitudinal data. Nephrol Dial Transplant. 2014;29(9):1733–1740. doi: 10.1093/ndt/gfu056 [DOI] [PubMed] [Google Scholar]
- 35.Lees JS, Shlipak MG. Age-adapted eGFR thresholds underestimate the risks beyond kidney failure associated with CKD in older populations. J Am Geriatr Soc. 2022;70(12):3660–3664. doi: 10.1111/jgs.18046 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.National Health Insurance Service. A Frustrating Year in 2020, Have You Had a Health Checkup?, 2018. Accessed January 23, 2025. https://www.nhis.or.kr/magazin/164/html/sub1.html [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Partial restrictions to the data and/or materials apply. Raw data are available to researchers upon relevant request, with approval by the Korean National Health Insurance Sharing Service.