Glomerular hyperfiltration in dipstick-negative type 2 diabetes mellitus: a novel determinant of kidney function decline in type 2 diabetes mellitus
Article information
Abstract
Background
The independent impact of type 2 diabetes mellitus (T2DM) on kidney outcomes beyond albuminuria remains unclear. We evaluated whether T2DM affects kidney outcomes in individuals with normal kidney function without albuminuria.
Methods
Data from the National Health Insurance Service-National Sample Cohort of Korea (2009–2015) were analyzed. Individuals with normal kidney function were stratified by T2DM status. The primary outcome was a composite kidney outcome consisting of initiation of kidney replacement therapy and a sustained decline in estimated glomerular filtration rate (eGFR) of ≥40% from baseline.
Results
Among 77,267 individuals with normal kidney function without albuminuria, patients with T2DM (n = 11,957) showed significantly steeper annual decline in eGFR than non-T2DM individuals (–0.113 mL/min per 1.73 m2 per year; 95% confidence interval [CI], –0.222 to –0.003). T2DM was associated with a 57% higher risk of composite kidney outcome (adjusted hazard ratio, 1.57; 95% CI, 1.28–1.92), independent of traditional risk factors. This association was strongest in individuals with glomerular hyperfiltration and longer T2DM duration (≥6 years).
Conclusion
Normal kidney function T2DM was associated with accelerated kidney function decline and a 1.5-fold increased risk of adverse kidney outcomes compared with normal kidney function non-T2DM, particularly in individuals with glomerular hyperfiltration and longer duration.
Introduction
Diabetes mellitus (DM), along with hypertension (HTN), is the main cause of chronic kidney disease (CKD) worldwide. DM accounts for 30%–50% of total CKD worldwide. Its share is expected to continue to increase in the future [1]. As the prevalence of DM increases, diabetic kidney disease (DKD) is also becoming more common and remains the leading cause of end-stage kidney disease in many developed and developing countries [2,3]. Along with high cardiovascular risk in DM itself and concomitant CKD, DKD has been predicted to become a substantial global health burden [4].
Current DKD definitions rely heavily on albuminuria, which may delay recognition of high-risk patients with preserved estimated glomerular filtration rate (eGFR). The classic pattern of DKD is induced by glomerular hyperfiltration, followed by micro- and macro-albuminuria [5,6]. A longitudinal study regarding population-based district diabetes data in Salford, Greater Manchester, UK showed that the annual decline rate of eGFR for those with DM without albuminuria was 19 times slower than those with DM and macro-albuminuria [7]. However, an increasing number of studies show that many diabetic patients develop CKD without the classic pathway of preexisting albuminuria [8–11]. A study conducted in Joslin Diabetes Center reported that up to 9% of type 1 DM (T1DM) patients and 20% of type 2 DM (T2DM) patients showed significant early progressive kidney function decline without albuminuria [12]. On the other hand, another report showed that kidney function declines in diabetics and non-diabetics with comparable levels of albuminuria were similar [13]. These previous controversial results might be due to different study designs and enrolled subjects. In general, patients with T2DM without albuminuria are considered to have kidney function decline comparable to the general population [14]. However, no study has directly compared patients with T2DM without albuminuria to normal individuals. In particular, most studies targeted populations whose underlying proteinuria levels are available only at baseline. Thus, a well-designed, large-scale study is needed to better understand the impact of DM on kidney function in the absence of pathological albuminuria.
Therefore, this study aimed to evaluate the association between T2DM and adverse kidney outcomes among individuals with normal kidney function and without albuminuria by comparing patients with T2DM to non-T2DM individuals with comparable baseline kidney function. This study also aimed to explore whether baseline eGFR and diabetes duration were associated with kidney function decline among patients with T2DM and preserved kidney function.
Methods
The cohort profile
The National Health Insurance Service-National Sample Cohort (NHIS-NSC) of the Republic of Korea consisted of 1,025,340 subjects, comprising 2.2% of the total population [15]. It was part of the National Health Information Database for reimbursement. It contained general health examination data of participants with 13 years of follow-up from 2002 to 2015. Among them, data between 2009 and 2015 were targeted, because serum creatinine levels were available from 2009.
Data from general health examinations included anthropometry data, such as height, weight, waist circumference, and body mass index (BMI). Blood pressure, laboratory results, including total cholesterol, high-density lipoprotein cholesterol (HDL-C), triglycerides, aspartate aminotransferase (AST), alanine aminotransferase (ALT), serum creatinine, and urinalysis using a dipstick test were also included. Also, data about prescriptions and the International Classification of Diseases, 10th Revision (ICD-10) codes were available for this cohort.
This study protocol was reviewed and approved by the Institutional Review Board (IRB) of Soonchunhyang University Cheonan Hospital in Cheonan, Korea (No. SCHCA 2019−04−030). The need for informed consent was waived by the IRB. This study was conducted in accordance with the principles of the Declaration of Helsinki.
Study population and covariates
The annual decline in eGFR (i.e., annual eGFR slope) was estimated among participants who had two or more measurements of both serum creatinine and urine dipstick tests. To evaluate individuals with preserved kidney function and no dipstick-detectable proteinuria, we included participants with an eGFR ≥60 mL/min per 1.73 m2 and negative urine dipstick results. In our previous work, transiently documented as trace in the urine dipstick test showed a significant effect on kidney and cardiovascular outcomes [16]. Thus, trace or greater degree of urine dipstick test results were considered as positive. The eGFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation [17]. Concerns about the CKD-EPI equation in the elderly and controversies on appropriate eGFR levels according to age led to a target population aged 40 to 65 years [18].
To remove confounding effects, participants with disabilities, died with an ICD-10 code of S or T (indicating injuries, poisoning, and certain other consequences of external causes), or had a previous history of kidney replacement therapy at baseline were excluded. Subjects with too low serum creatinine levels (<0.4 mg/dL in male and <0.3 mg/dL in female) were also excluded, because low serum creatinine levels might indicate another undetected medical history that could affect creatinine levels, such as a history of sarcopenia or amputation. Covariates, such as sex, BMI, systolic blood pressure (SBP), diastolic blood pressure (DBP), hemoglobin, serum fasting glucose, total cholesterol, HDL-C, triglycerides, AST, ALT, Charlson Comorbidity Index (CCI) [19], history of HTN, smoking status, alcohol consumption, residence (urban vs. non-urban), and usage of renin-angiotensin system (RAS) blockers or statins, were adjusted during analysis. Past medical history on HTN was acquired based on self-reported questionnaire and ICD-10 codes (I10–I15). History of smoking and alcohol consumption was acquired based on a self-reported questionnaire. Degree of alcohol consumption was defined as none, moderate (up to two standard drinks/day for male and one standard drink/day for female), or heavy (more than moderate). Smoking history was categorized into current smoker or not (i.e., non-smoker or ex-smoker). Participants with missing values of covariates were excluded from further analysis.
Study outcomes
We designated a composite kidney outcome as the primary outcome. Those who met the following criteria were analyzed: (1) initiation of kidney replacement therapy of more than 3 months, and (2) a decline of ≥40% in eGFR from baseline. Confirmation of a decline of ≥40% in eGFR from baseline through follow-up laboratory results defined the composite kidney outcome. In addition, we estimated the expected time point for the eGFR decline ≥40% from baseline using the annual eGFR decline. Annual eGFR decline from baseline was calculated employing a linear mixed model [20].
The secondary outcome consisted of all-cause death, cardiovascular death, and a composite total outcome. All-cause death was defined as death due to any cause. Cardiovascular death was defined as death with the primary cause of death being cardiovascular diseases, coded based on ICD-10 I-codes. The composite total outcome was defined as a combination of the composite kidney outcome and all-cause death.
Statistical analysis
All statistical analyses were performed using R version 3.3 (The R Foundation for Statistical Computing). Continuous variables are expressed as the mean ± standard deviation. Categorical variables are expressed as counts and percentages. Differences between groups were compared using the Mann-Whitney test, chi-square test, or Fisher exact test, as appropriate.
Cox proportional hazard model was used to estimate the hazard ratio (HR) of T2DM subjects with normal kidney function (non-albuminuric and eGFR ≥60 mL/min per 1.73 m2), compared to the non-T2DM group with normal kidney function. Covariates, including baseline age, eGFR, sex, BMI, SBP, hemoglobin, serum fasting glucose, total cholesterol, triglycerides, HDL-C, AST, ALT, CCI, history of HTN, smoking status, alcohol consumption, residence, and usage of RAS blockers or statins, were included in the model to adjust for confounders. Continuous variables, including age, eGFR, BMI, SBP, hemoglobin, serum fasting glucose, total cholesterol, triglycerides, HDL-C, AST, and ALT, were normalized with scaling, and used as covariates. The CCI was categorized into (0, 1, and 2+). Whether patients benefited from Medical Aid was not included in the model, because the proportion of those receiving Medical Aid was too small (<0.1%).
To estimate the annual decline of eGFR, the β coefficient of interaction term between time (i.e., time point of serum creatinine follow-up) and group according to T2DM was calculated. To adjust covariates, variables used in the Cox proportional hazard model were included in a linear mixed model using the lme4 R package [21].
To balance possible confounders between normal kidney function T2DM and normal kidney function non-T2DM groups, inverse probability of treatment weighting (IPTW) and propensity score matching were used [22]. Propensity score was calculated using logistic model using variables that could affect the outcome (variables used to adjust confounders in the Cox proportional hazard model), including baseline age, eGFR, sex, BMI, SBP, hemoglobin, serum fasting glucose, total cholesterol, triglycerides, HDL-C, AST, ALT, CCI, history of HTN, smoking status, alcohol consumption, residence, and usage of RAS blockers or statins [23]. Stabilized IPTW was calculated based on propensity score, and then the values below 1 percentile and above 99 percentile were truncated.
Results
Study populations
The study population and participant selection process are summarized in Supplementary Fig. 1 (available online). NHIS-NSC health examination results of 587,339 participants were available between 2009 and 2015. Among them, 446,352 and 448,068 had at least two urine dipstick results and serum creatinine levels, respectively. The number of participants with at least two urine dipstick results and at least two serum creatinine results in the same period was 385,892. To have subjects who had a follow-up period longer than 4 years, participants who had their first health check-up between 2009 and 2011 were selected. The number of subjects who had negative findings across all urine dipstick screenings in the inclusion period was 326,974.
Of these 326,974 subjects, 152,937 were excluded based on pre-determined exclusion criteria. Subjects who were registered as handicapped before enrollment (n = 19,543) and those who died with an ICD-10 code of S or T (n = 693) were excluded. Subjects whose baseline serum creatinine levels were too low (<0.4 mg/dL in male and <0.3 mg/dL in female) to be measured accurately by the test (n = 390), those who were younger than 40 or older than 65 (n = 133,306), those whose eGFRs were already below 60 mL/min per 1.73 m2 (n = 17,350), those who had a previous history of kidney replacement therapy (n = 12), those who had T1DM (n = 420), and those who had missing values (waist circumference, n = 110; BMI, n = 101; SBP, n = 48; DBP, n = 47; hemoglobin, n = 24; total cholesterol, n = 4; triglycerides, n = 7; HDL-C, n = 31; alcoholic history, n = 4), were also excluded. As well, 96,770 non-T2DM subjects whose serum fasting glucose levels were higher than 100 mg/dL were excluded to minimize the effect of least hyperglycemia. Finally, 77,267 subjects, including 11,957 normal kidney function patients with T2DM and 65,310 normal kidney function non-T2DM subjects, were eligible for analysis. During the follow-up period, almost 70% of subjects in both groups had ≥3 times of measurements of serum creatinine and urine dipstick tests (Supplementary Fig. 2, available online).
Baseline characteristics of a cohort
Baseline characteristics of the study population are summarized in Supplementary Table 1 (available online). Subjects in the normal kidney function T2DM group were older, had lower baseline eGFR and higher blood pressure, and were taking more RAS inhibitors and statins. The normal kidney function T2DM group consisted of more males and more obese subjects. The CCI was higher in the normal kidney function T2DM group. There were more smokers and heavy drinkers in the normal kidney function T2DM group. More subjects in the normal kidney function non-T2DM group tended to live in urban areas than those in the normal kidney function T2DM group. The mean follow-up period was 67.7 ± 11.1 months in the normal kidney function T2DM group, and 67.2 ± 10.3 months in the normal kidney function non-T2DM group. Since there were discrepancies in baseline characteristics between the two groups, we used IPTW and propensity score matching methods to balance covariates, resulting in much improvement in the balanced states of covariates between the two groups (Table 1).
Standardized mean differences showing balances between groups after IPTW and PSM
Composite kidney outcome and estimated glomerular filtration rate decline according to type 2 diabetes mellitus status
The Kaplan-Meier curve of composite kidney outcome showed higher outcomes in the T2DM group than in the non-T2DM group in crude, IPTW, and propensity score matching models (Fig. 1). In the overall cohort, 1,365 composite kidney outcome events occurred, including 361 events in the T2DM group and 1,004 events in the non-T2DM group. The incidence rate difference was 258.9 per 100,000 person-years (95% confidence interval [CI], 201.3–316.4). T2DM was associated with a higher risk of composite kidney outcome in the crude model (HR, 1.94; 95% CI, 1.72–2.19) and after multivariable adjustment (adjusted HR, 1.57; 95% CI, 1.28–1.92) (Supplementary Table 2, available online). In contrast, all-cause and cardiovascular mortality did not differ significantly between groups after multivariable adjustment.
Kaplan-Meier curves for composite kidney outcome in the crude model and after employment of inverse probability of treatment weighting and propensity score matching.
Kaplan-Meier curves for (A) crude model, after employment of (B) inverse probability of treatment weighting (IPTW), and (C) propensity score matching, are shown. No diabetes mellitus (DM) and DM denote the non-type 2 DM (T2DM) and T2DM groups with normal kidney function, respectively. Table 1 presents the baseline characteristics and balances between groups after IPTW and propensity score matching.
The annual decline of eGFR analyzed using a linear mixed model also showed that the normal kidney function T2DM group showed a steeper decline of eGFR than the normal kidney function non-T2DM group (Supplementary Fig. 3, available online). Impact of T2DM with normal kidney function on annual eGFR decline was –0.300 mL/min/1.73 m2 per year (95% CI, –0.368 to –0.231; p < 0.001) in the crude model. After adjusting for possible covariates, such as age, sex, BMI, SBP, hemoglobin, serum fasting glucose, total cholesterol, triglyceride, HDL-C, AST, ALT, CCI, HTN, smoking, alcohol, residence, and use of RAS inhibitors and statins, the impact of T2DM with normal kidney function on annual eGFR was –0.113 mL/min/1.73 m2 per year (95% CI, –0.222 to –0.003; p = 0.04).
Association of baseline estimated glomerular filtration rate and diabetes duration with composite kidney outcomes in patients with type 2 diabetes mellitus and normal kidney function
We examined the associations between T2DM and other baseline clinical variables and the composite kidney outcome in the overall cohort (Supplementary Fig. 4, available online). Older age (HR, 1.39; 95% CI, 1.31–1.49) and HTN (HR, 1.28; 95% CI, 1.05–1.55) were associated with a higher risk of the composite kidney outcome. In contrast, current smoking (HR, 0.81; 95% CI, 0.67–0.97) and heavy alcohol intake (HR, 0.81; 95% CI, 0.66–0.98) were associated with lower HRs.
Higher baseline eGFR was also associated with a higher HR for the composite kidney outcome (Supplementary Fig. 4, available online). Given the potential role of glomerular hyperfiltration in kidney function decline, we further evaluated annual eGFR decline according to baseline eGFR and diabetes duration. Fig. 2 visualizes the overall distribution of annual eGFR decline according to baseline eGFR and diabetes duration. To further evaluate these patterns, we performed stratified analyses according to diabetes duration, baseline age, and baseline eGFR (Tables 2, 3). In Table 2, annual eGFR decline differed significantly according to diabetes duration (overall p < 0.001). A T2DM duration of ≥6 years was significantly associated with a more rapid decline in kidney function compared with shorter durations (1–2 years vs. 3–5 years, p = 0.09; 1–2 years vs. ≥6 years, p < 0.001; 3–5 years vs. ≥6 years, p = 0.04).
Annual decline of eGFR in patients with normal kidney T2DM or normal kidney non-T2DM.
Annual decline in eGFR was calculated using a linear mixed model. The x-axis and y-axis represent the baseline eGFR of subjects and the annual decline of eGFR, respectively. Red dots denote patients with normal kidney T2DM, while blue dots represent those with normal kidney non-T2DM. Each regression line represents a LOWESS smoothing line according to T2DM duration in the normal kidney T2DM group.
DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; T2DM, type 2 DM.
Annual eGFR decline in T2DM patients with normal kidney function, stratified by risk factors
HRs for composite kidney outcome after stratification according to T2DM duration, age, and baseline eGFR
For the composite kidney outcome, we further compared patients with T2DM and normal kidney function with non-T2DM individuals within the same strata of age and baseline eGFR (Table 3). In younger patients (age, 40–54 years), lower eGFR strata showed higher HRs irrespective of T2DM duration (HR, 5.53, 8.08, and 5.36 for T2DM durations of 1–2, 3–5, and ≥6 years, respectively). In the eGFR 90–104 mL/min per 1.73 m2 stratum, a short duration of T2DM was not associated with a higher risk of the composite kidney outcome (HR, 0.81 for T2DM duration of 1–2 years), whereas higher risks were observed among those with longer T2DM durations (HR, 1.93 and 1.95 for T2DM durations of 3–5 and ≥6 years, respectively). In younger patients (age, 40–54 years), the group with shorter T2DM duration and glomerular hyperfiltration had a higher risk of the composite kidney outcome than the non-T2DM group (HR, 1.63 for T2DM duration of 1–2 years). In contrast, in older patients (age ≥55 years), higher HRs were observed in those with longer T2DM duration and lower eGFR strata.
Discussion
This study sought to determine the impact of T2DM itself on kidney outcome by comparing a normal kidney function T2DM population to a normal kidney function non-T2DM population. In this national health examination cohort study, the normal kidney function T2DM group showed a steeper decline of eGFR slope than the normal kidney function non-T2DM group. This result was consistent after adjusting for possible variables. The hazard risk of normal kidney function in T2DM on composite kidney outcome was 1.57 times higher. This implies that T2DM itself without albuminuria is also a risk factor of kidney impairment, which strengthens the concept of non-albuminuric DKD.
In a linear mixed model to ascertain risk factors for rapid kidney impairment in the non-albuminuric DKD, longer T2DM duration, older age, higher baseline eGFR, higher BMI, higher CCI, HTN, and the use of RAS inhibitors or statins showed a tendency for a steeper eGFR slope. Duration of T2DM showed controversial results in previous studies based on renal biopsies remaking non-DKD, with some studies showing shorter duration of T2DM as a risk factor of kidney impairment [24]. Since our study enrolled a large amount of data accounting for relatively healthy people with baseline eGFR higher than 60 mL/min per 1.73 m2, it could be more focused on the decline of kidney function of general non-albuminuric patients with T2DM. Some variables showed unexpected results. For example, the usage of RAS inhibitors and statins was related to a steeper decline in annual eGFR. Considering that the current study was based on national health data, it could be interpreted that the usage of those medications known to give beneficial effects of kidney protection [25] might be expressed as each disease condition of HTN and dyslipidemia, respectively. On the other hand, current smoking and heavy alcohol consumption showed a relatively less steep decline in annual eGFR. While some previous studies have reported similar results on smoking and preserved kidney function [26], the results are still uncertain to confirm the relationship between smoking and kidney function of those with non-albuminuric T2DM, since causality is not revealed. One possible hypothesis for this is that healthier, younger people might be more likely to be smokers and alcohol drinkers.
According to our study findings, individuals with T2DM without albuminuria might experience declining kidney function due to glomerular hyperfiltration even in the absence of albuminuria, as well as a longer duration of T2DM. The possible mechanism behind our results is that hyperfiltration itself, even without albuminuria, is a risk factor for the decline of kidney function, considering the sequential progression steps of classic DKD [27,28].
Sustained kidney function in a person is a fundamental issue affecting both the quality of life and the quantity of life in terms of life span. The prevalence of DM, a chronic disease that is already the leading cause of CKD in many developed and developing countries, is increasing faster than predicted [29,30]. Understanding the mechanism of developing CKD in patients with DM is crucial for its management and the prevention of DKD.
Since the progression of DKD by the pathway of glomerular hyperfiltration followed by albuminuria and proteinuria is a known important mechanism of kidney function loss, treatment focusing on this mechanism has been studied, and medications that could ameliorate glomerular hyperfiltration, like RAS inhibitors and sodium-glucose cotransporter 2 (SGLT2) inhibitors, are recommended as first-line therapy for this population [31,32]. More recently, finerenone, a nonsteroidal mineralocorticoid receptor antagonist, has also shown kidney-protective benefits in patients with T2DM and albuminuric CKD [33]. In this context, the treatment era of the present cohort should be considered when interpreting our findings. SGLT2 inhibitors are now central kidney-protective therapies in DKD management, and emerging evidence suggests that they may help preserve kidney health earlier in the course of T2DM [34]. However, the present cohort was established before SGLT2 inhibitors became routinely used for kidney protection in clinical practice, and before the widespread use of other contemporary kidney-protective therapies. Therefore, our findings may reflect the natural course of kidney function decline in patients with T2DM and preserved kidney function before the modifying effects of contemporary kidney-protective therapies.
Emerging evidence shows that patients with DM can develop CKD without the known classic pathway [35,36]. Some studies have investigated the nature of non-albuminuric DKD. A cross-sectional study by MacIsaac et al. [37] in 2004 showed that those with non-albuminuric DM were more likely to be older and female. Declining rates of glomerular filtration rate were not significantly different from those of albuminuric DM. The Japan Diabetes Clinical Data Management study (JDDM15) in 2009 reported that those with normoalbuminuric DM and low eGFR were more likely to be older, female, and fewer smokers than those with normoalbuminuric DM and preserved eGFR [26]. A study published in 2012 based on the National Health and Nutrition Examination Survey enrolling 2,798 DM and 15,743 non-DM demonstrated that DM and women were more common in those with normoalbuminuric CKD [38]. On the other hand, a nationally representative cohort study in Australia in 2009 reported that non-albuminuric kidney impairment was less common in DM than in the general population, independent of sex or DM duration [14]. These inconsistent results of previous studies on the characteristics of non-albuminuric DKD are probably due to different study populations and designs. This suggests that more studies are needed to confirm the nature of non-albuminuric DKD and the impact of DM itself on kidney outcome without the classic pathway of glomerular hyperfiltration or concomitant albuminuria.
In the current study, we showed that non-albuminuric T2DM was associated with kidney outcomes. Our study had a strength of being based on a reliable national health cohort that enrolled a large number of subjects with objective health data, compared to previous studies regarding non-albuminuric T2DM. By comparing patients with non-albuminuric T2DM with non-T2DM individuals with comparable baseline kidney function, our study provided an opportunity to evaluate the association between T2DM and kidney outcomes while minimizing baseline kidney-related differences.
However, this study also has some limitations. Due to the study design, we could not establish the exact mechanism of causality between non-albuminuric T2DM and kidney outcomes. Urine albumin-to-creatinine ratio was not available for this analysis; therefore, albuminuria was assessed using urine dipstick results. Previous population-based studies have reported high negative predictive values of negative urine dipstick results for detecting albuminuria, supporting the use of urine dipstick testing as a practical screening tool in large epidemiologic datasets [39,40]. In addition, our study included only participants who had negative urine dipstick results on at least two examinations. Therefore, although undetected microalbuminuria cannot be completely excluded, the likelihood of clinically relevant albuminuria was considered to be low. Antidiabetic medication use was not incorporated into the present analysis. The study period preceded the routine clinical use of SGLT2 inhibitors and glucagon-like peptide-1 receptor agonists for kidney protection; nevertheless, other antidiabetic medications, including insulin, may reflect diabetes severity and could have influenced kidney outcomes. In addition, the exact onset of diabetes could not be determined because T2DM status and duration were defined using claims and health examination data. Counterintuitive associations observed for RAS inhibitor use, statin use, current smoking, and heavy alcohol intake should also be interpreted cautiously. The associations with RAS inhibitor and statin use may reflect confounding by indication, whereas the lower HRs associated with current smoking and heavy alcohol intake may reflect residual confounding or selection bias rather than protective effects. Despite the use of IPTW and propensity score matching, residual or unmeasured confounding may remain because factors such as diet, medication adherence, and other lifestyle-related variables were not captured in the NHIS-NSC database. Our findings suggest that non-albuminuric T2DM is associated with a greater annual decline in kidney function, but future studies are needed to clarify the mechanisms underlying kidney function decline in T2DM beyond the classic hyperfiltration–albuminuria pathway. Although our study used relatively long-term data of 6 years from 2009 to 2015, more reliable results might have been obtained if the study had been conducted for an additional period, considering the long course of continuous kidney function decline in DM. Lastly, our study excluded individuals with T1DM. Therefore, we could not evaluate the association between non-albuminuric T1DM and kidney outcomes.
In conclusion, this nationwide cohort study showed that T2DM was associated with a greater annual decline in kidney function and a 1.5-fold increased risk of adverse kidney outcomes among individuals with preserved kidney function and no albuminuria.
Supplementary Materials
Supplementary data are available at Kidney Research and Clinical Practice online (https://doi.org/10.23876/j.krcp.26.030).
Notes
Conflicts of interest
The authors declare no conflicts of interest.
Funding
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (Ministry of Science and ICT [MSIT]) (No. RS-2025-00513642; to EYL), and the Soonchunhyang University Research Fund (to EYL). This work was also supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (No. NRF-2021R1G1A1009254; to DC).
Data sharing statement
Detailed data are available from the National Health Insurance Service. While these data are not publicly accessible due to restrictions and licensing agreements specific to this study, they can be made available upon a reasonable request to the Korea National Health Insurance Service, following their review process.
Authors’ contributions
Conceptualization: EYL
Data curation, Formal analysis, Methodology, Software: SP
Funding acquisition: DC, EYL
Investigation: DC, SP
Project administration: DC, SP, EYL
Resources: NJC
Supervision: YYC, HWG
Validation: DSK, DJL
Visualization: DC
Writing–original draft: DC
Writing–review & editing: All authors
All authors read and approved the final manuscript.