Association of poorly controlled HbA1c with increased risk of progression to end-stage kidney disease and all-cause mortality in patients with diabetes and chronic kidney disease

Sheng-Jen Chen; Hsiu-Yin Chiang; Pei-Shan Chen; Shih-Ni Chang; Sheng-Hsuan Chen; Min-Yen Wu; Hung-Chieh Yeh; I-Wen Ting; Hsiu-Chen Tsai; Pei-Chun Chen; Chin-Chi Kuo

doi:10.1371/journal.pone.0274605

Abstract

Glycosylated hemoglobin (HbA1c) targets for patients with chronic kidney disease (CKD) and type 2 diabetes remain controversial. To evaluate whether baseline HbA_1c and HbA_1c trajectories are associated with the risk of end-stage kidney disease (ESKD) and all-cause mortality, we recruited adult patients with CKD and type 2 diabetes from a “Pre-ESKD Program” at a medical center in Taiwan from 2003 to 2017. Group-based trajectory modeling was performed to identify distinct patient groups that contained patients with similar longitudinal HbA1c patterns. Cox proportional hazard models were used to estimate hazard ratios (HRs) of ESKD and mortality associated with baseline HbA_1c levels and HbA_1c trajectories. In the analysis related to baseline HbA_1c (n = 4543), the adjusted HRs [95% confidence interval (CI)] of all-cause mortality were 1.06 (0.95–1.18) and 1.25 (95% CI, 1.07–1.46) in patients with an HbA_1c level of 7%–9% (53–75 mmol/mol) and >9% (>75 mmol/mol), respectively, as compared with those with an HbA1c level < 7% (<53 mmol/mol). In the trajectory analysis (n = 2692), three distinct longitudinal HbA_1c trajectories were identified: nearly optimal (55.9%), moderate to stable (34.2%), and poor control (9.9%). Compared with the “nearly optimal” HbA_1c trajectory group, the “moderate-to-stable” group did not have significantly higher mortality, but the “poorly controlled” group had 35% higher risk of mortality (adjusted HR = 1.35, 95% CI = 1.06–1.71). Neither baseline levels of HbA_1c nor trajectories were associated with ESKD risk. In conclusion, in patients with CKD and type 2 diabetes, poor glycemic control was associated with an elevated risk of mortality but not associated with a risk of progression to ESKD.

Citation: Chen S-J, Chiang H-Y, Chen P-S, Chang S-N, Chen S-H, Wu M-Y, et al. (2022) Association of poorly controlled HbA_1c with increased risk of progression to end-stage kidney disease and all-cause mortality in patients with diabetes and chronic kidney disease. PLoS ONE 17(9): e0274605. https://doi.org/10.1371/journal.pone.0274605

Editor: Mabel Aoun, Faculty of Medicine, Saint-Joseph University, LEBANON

Received: November 22, 2021; Accepted: August 31, 2022; Published: September 26, 2022

Copyright: © 2022 Chen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: Data cannot be shared publicly in order to protect patient confidentiality. Data are available from the China Medical University Hospital Institutional Data Access / Ethics Committee (contact via Ms. Maggie Shih; Email: a6034@mail.cmuh.org.tw) to researchers who meet the criteria for access to confidential data.

Funding: This study was supported by Health and Welfare Data Science Center, Ministry of Health Welfare, and Health Data Science Center, China Medical University Hospital, and grants from the Ministry of Science and Technology, Taiwan (Grant number: grant number: 108-2314-B-039-038-MY3, and MOST 111-2321-B-468-001). This study was not sponsored by industry. There was no additional external funding received for this study. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Diabetic nephropathy is a leading cause of end-stage kidney disease (ESKD) worldwide and accounts for a considerable proportion of the global ESKD incidence, including in Singapore (66.4%), the United States (46.9%), Taiwan (46.2%), Japan (42.5%), Canada (37.7%), and the United Kingdom (26.5%) [1]. For patients with coexisting type 2 diabetes and chronic kidney disease (CKD), optimal glycemic control targets have been explored in diverse populations. Currie et al. [2] reported a U-shaped association between all-cause mortality and glycosylated hemoglobin (HbA_1c) levels in patients with diabetes. In the Action to Control Cardiovascular Risk in Diabetes (ACCORD) [3] trial, the risk of all-cause mortality among patients with CKD stage 1–3 was higher in the intensive therapy group (median HbA_1c of approximately 6.5%, 48 mmol/mol in the 12th month of follow-up) than in the standard therapy group (median HbA_1c of approximately 7.6%, 60 mmol/mol in the 12th month of follow-up). Shurraw et al. [4] revealed a U-shaped association between the risk of all-cause mortality and a baseline HbA_1c level of <6.5% (48 mmol/mol) or > 8.0% (64 mmol/mol). Another study on baseline HbA_1c revealed that compared with patients who had CKD stage 3 or 4 and a baseline HbA_1c level < 6.0%, patients with CKD stage 3 or 4 and a baseline HbA_1c level > 9.0% (75 mmol/mol) had higher risk of ESKD (rather than all-cause mortality). However, the ESKD risk was lower in patients with CKD stage 5 [5]. The latest KDIGO (Kidney Disease: Improving Global Outcomes) guidelines suggest that the acceptable HbA_1c target ranges from 6.5%–8.0% (48–64 mmol/mol) [6]; however, this consensus on a glycemic control target was mainly based on clinical trials in which patients with preserved kidney function (i.e., those with an estimated glomerular filtration rate [eGFR] of ≧60 mL/min/1.73 m²) were selected [7–9]. Whether the study findings can be generalized to patients with coexisting diabetes and advanced CKD in real-world settings deserves attention [3, 10, 11].

None of the aforementioned studies have explored the prognostic role of the longitudinal trajectory of HbA_1c level in patients with type 2 diabetes and CKD; such an exploration could aid in optimal glycemic control threshold estimation. Although a single value of HbA_1c could reflect the average blood glucose level over a period of up to 3 months, its representativeness of longer-term glycemic control is insufficient, and thus, up to four annual HbA_1c measurements have been suggested [6]. Although the KDIGO Work Group noted the potential for more stringent glycemic control to improve clinical outcomes in terms of all-cause mortality, cardiovascular death, and CKD progression [6], more robust evidence is required to verify whether stringent glycemic control can modify the disease course of patients with type 2 diabetes and CKD. In this study, we used a 15-year single-center longitudinal database to systematically investigate the association of both baseline HbA_1c levels and HbA_1c trajectories with the risk of progression to ESKD and all-cause mortality in patients with type 2 diabetes and CKD.

Materials and methods

Study population

In 2002, Taiwan’s National Health Insurance launched the Project of Integrated Care of CKD and, since 2007, the project’s focus has been CKD stages 3b–5 [12]. This pre-end-stage kidney disease (ESKD) program was a multidisciplinary approach to the design of individualized care plans for a wide range of patients with CKD. Patients with eGFR < 45 mL/min/1.73 m² (i.e., CKD stage 3b–5), or with eGFR ≥ 45 mL/min/1.73 m² (i.e., CKD stage 1–3a) with evident proteinuria (urine protein and creatinine ratio ≥ 1000 mg/gm) were eligible to participate in the Pre-ESKD Program. The objective was to meet the therapeutic goals listed in the guidelines of the National Kidney Foundation Kidney Disease Outcomes Quality Initiative [13]. China Medical University Hospital (CMUH), a tertiary medical center located in Central Taiwan, joined the Pre-ESKD program in 2003. Consecutive patients with CKD who were willing to participate were prospectively enrolled. The CMUH pre-ESKD program currently includes more than 11 000 participants and has an overall retention rate of 90%. CKD diagnoses are based on the working diagnoses of nephrologists or the criteria outlined in the aforementioned initiative’s guidelines [13]. Patients in CKD stages 3b, 4, and 5 were, respectively, followed up at 12, 8, and 4 weeks, or as necessary. Biochemical markers of renal injury including serum creatinine, eGFR, and the spot urine protein–creatinine ratio (PCR) were measured at intervals of no more than 12 weeks. Detailed information on the Pre-ESKD Program has been provided previously [14, 15]. Throughout the manuscript, we use the phrase Pre-ESKD (end-stage kidney disease) program to refer to this multidisciplinary care program.

The index date was defined as the date of first enrollment in the Pre-ESKD program. We first identified patients with a diagnosis of diabetes based on the International Classification of Diseases, 9th and 10th revision Clinical Modification (ICD-9-CM 250 or ICD-10-CM E08-E11, E13) codes and prescriptions of antidiabetic agents before the index date as well as during an additional 1-year inclusion window following the index date. The exclusion criteria included (1) age < 20 years or > 90 years, (2) having a history of dialysis or kidney transplant before the index date, (3) having type 1 diabetes, and (4) not having a recorded baseline HbA_1c value. Baseline HbA_1c was defined as the HbA_1c value recorded 1 year before or 3 months after the index date; the measurement closest to the index date was used. Patients with type 1 diabetes were identified from certificates of catastrophic illness issued by the National Health Insurance Administration, Ministry of Health and Welfare of Taiwan. Because we wanted to observe longitudinal HbA1c patterns, only patients with at least three measurements of HbA1c were included in the trajectory analysis. Patients included in the HbA_1c trajectory analysis was had to have had at least a 6-month follow-up and a last HbA1c measurement at least 6 months after the index date (Fig 1). Consequently, 4543 patients were included in the baseline HbA_1c analysis and 2692 patients were included in the trajectory analysis (Fig 1).

Download:

Fig 1. Selection of the study population.

https://doi.org/10.1371/journal.pone.0274605.g001

Measurement of HbA_1c

All HbA_1c levels were measured at the central laboratory of CMUH. Before September 2013, HbA_1c was measured using Tosoh’s Automated Glycohemoglobin Analyzer HLC-723G7 (Tosoh G7; Tosoh Corporation, Minato-Ku, Tokyo, Japan). Two-point calibration was performed using a standard HbA_1c sample after every device power-up. The analyzer could distinguish between labile A_1c and stable HbA_1c, indicating a minimal risk of measurement error. Calculation of HbA_1c levels was based on the ratio of the stable HbA_1c fraction chromatographic area to that of total glycosylated hemoglobin, and the HbA1_c ratio of each result was automatically adjusted using the calibration equation [16]. From September 2013 onward, the HbA_1c measurement protocol was switched; Premier Hb9210 (Trinity Biotech Plc., Wicklow, Ireland.) was used thereafter. Premier Hb9210 uses boronate-affinity high-performance liquid chromatography to detect all types of the presented glycosylated Hb species. The final HbA_1c results are determined from a simple peak area fraction.

According to prespecified HbA1c values from the literature and the latest American Diabetes Association practice guidelines, we divided the patients into three groups: those with a baseline HbA_1c level < 7% (<53 mmol/mol), 7%–9% (53–75 mmol/mol), and >9% (>75 mmol/mol); [4, 5, 17]. In the trajectory analysis, we used all available HbA_1c measurements collected during follow-up for each patient to determine the patient subgroups with similar patterns in longitudinal HbA_1c.

Other covariables

Sociodemographic variables, including age, sex, education level, smoking status, and alcohol consumption, were collected through a questionnaire during enrollment. Smoking status and alcohol consumption status were categorized as never, former, and current [18]. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared, and the latest measurements obtained within ±2 years of the index date were used in the analysis. Baseline levels of all biochemical variables were determined using the latest measurements obtained within 90 days to 1 year of the index date. eGFR was calculated using the CKD epidemiology collaboration equation: [eGFR = 141 × min (S-Cre/κ, 1)^α × max(S-Cre/κ, 1)^−1.209 × 0.993^age × 1.018 [if patient is female] × 1.159 [if Black], where S-Cre is serum creatinine, κ is 0.7 for female patients and 0.9 for male patients, and α is −0.329 for female patients and −0.411 for male patients] [19]. The baseline eGFR of each patient was determined using their serum creatinine level, and patients were assigned to corresponding CKD stages based on the following cutoff values: > 90, 60–89.9, 30–59.9, 15–29.9, and < 15 mL/min/1.73 m². CKD stage was then determined on the basis of the following cutoff values for eGFR: >90 (stage 1), 60–89.9 (stage 2), 30–59.9 (stage 3), 15–29.9 (stage 4), and <15 (stage 5) mL/min/1.73 m². Missing values of the pooled urine protein–creatinine ratio (uPCR) were estimated from the urine albumin–creatine ratio (uACR) by using the following formula: ln(uACR) = 1.32 × ln(uPCR) − 2.64 [20]. Data on comorbidities and medication use were collected by searching the electronic health records within 1 year before the index date. Hypertension was defined as the presence of related diagnosis codes (ICD-9 codes 401–405 and ICD-10 codes I10–I15) or the prescription of an antihypertensive agent. Cardiovascular disease (CVD) included coronary artery disease, myocardial infarction, stroke, or heart failure (ICD-9 codes 394.9, 396, 410–414, 422.9, 424.0–424.2, 428.0, 428.9, 429.2, 430–438, and ICD-10 codes G45-G46, I11.0, I13.0, I13.2, I20-I25, I50, I60-I63, I69).

Outcomes and follow-up

Survival status and date of death was ascertained through data linkage with the National Death Registry of Taiwan. To minimize bias, we created a proxy outcome for progression to ESKD—a doubling of serum creatinine (S-Cre) concentration—in the main analysis to balance the risk of dialysis among the three baseline HbA1c groups. Progression to ESKD was defined as the initiation of peritoneal dialysis, hemodialysis, kidney transplantation, and doubling of S-Cre compared with the baseline. For each study participant, the follow-up period was from the index date until the earliest occurrence of ESKD, loss to follow-up, death, or December 31, 2017 whichever occurred first.

Statistical analyses

Continuous variables were expressed as a median and interquartile range (IQR), and the differences in continuous variables among the groups were determined using the Wilcoxon rank sum test. Categorical variables were expressed as percentages, and the differences in categorical variables among the HbA_1c categories were examined using a chi-squared test. P values for trends were calculated using Spearman’s correlation for continuous variables and the Cochran–Armitage trend test for categorical variables.

A semiparametric group-based trajectory model (GBTM) was used to characterize the distinct trajectories of HbA_1c during the follow-up period. The PROC TRAJ macro, developed using the SAS software package, fits a semiparametric mixture model to longitudinal data by using the maximum likelihood method [21–23]. GBTM is a useful approach for trajectory characterization when the number of potential subgroups and trajectory shapes of each subgroup are still unclear, and the Bayesian information criterion was employed to assess model fit by balancing model complexity. We empirically processed 2- and 3-group solutions and focused on the 3-group solution eventually after considering the sample size and facilitation of meaningful clinical interpretations. Missing values of sociodemographic variables were imputed using multiple imputation under the “missing at random” assumption. Associations of baseline HbA_1c and HbA_1c trajectories with the risk for ESKD and all-cause mortality were assessed using Cox proportional hazard models with age as the time scale. The subdistribution hazard model developed by Fine and Gray was fitted for ESKD; it accounted for competing risks of death without ESKD. We constructed three models with increasing levels of covariate adjustment. Model 1 was adjusted for sex, body mass index, smoking status, alcohol consumption, and education. Model 2 was further adjusted for systolic blood pressure, cardiovascular disease, primary etiologies of CKD, baseline medication use (contrast, nonsteroidal anti-inflammatory drugs, oral antidiabetic agents, insulin, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, diuretics, and epoetin), triglyceride level, and low-density lipoprotein cholesterol level. Model 3 was adjusted for all variables in Model 2 and baseline hemoglobin, eGFR, and pooled uPCR.

The dose–response relationship of baseline HbA_1c levels with all-cause mortality and ESKD risk was characterized using a restricted cubic spline in the Cox regression analysis with knots at the 10th, 50th, and 90th percentiles of the overall distribution of HbA_1c levels. We further performed exploratory subgroup analyses to evaluate potential effect modifications in the fully adjusted model according to age (< 65 vs. ≥ 65 years), sex, BMI category (< 25 vs. ≥ 25 kg/m²), smoking status, alcohol consumption, CKD stage (1–2 vs. 3–5), hypertension, and CVD. All statistical analyses were performed using SAS (version 9.4; SAS Institute Inc., Cary, NC, USA) and R (version 3.6.0; R Foundation for Statistical Computing, Vienna, Austria).

Ethics statement

All methods used in this study were performed in accordance with the relevant guidelines and regulations. The study was approved by the Big Data Center of CMUH and the Research Ethical Committee/Institutional Review Board of China Medical University Hospital (CMUH105-REC3-068); the need to obtain written informed consent for the present study was waived by the Research Ethical Committee of CMUH.

Results

Characteristics of study subjects by baseline HbA_1c levels

Among the 4543 subjects included in the baseline HbA1c analysis, the median age at enrollment was 67.6 years (IQR: 59.2–75.7), the median HbA_1c level was 7.1% (IQR: 6.30–8.20), and the median eGFR was 26.5 mL/min/1.73 m² (IQR: 13.8–43.8; Table 1). The median follow-up duration was 1.6 (IQR: 0.7–3.0) years for the development of ESKD and 3.8 (IQR: 1.9–6.3) years for all-cause mortality. At baseline, of all patients, 89.74% had a urine PCR value of ≥150 mg/g and 85.7% had a urine ACR value of ≥30 mg/g. Patients with a higher baseline HbA_1c level were younger and tended to have longer follow-up durations of ESKD and higher BMI (Table 1). In addition, patients with a higher HbA_1c level had a higher eGFR. Correspondingly, the proportion of CKD stage 5 and phosphorus and albumin levels were significantly lower in patients with a higher HbA1c level than in those with a lower HbA_1c level. Overall, 25.7% and 51.2% of our population were treated with an angiotensin-converting enzyme inhibitor (ACEi) and angiotensin II receptor blocker (ARB), respectively. The proportion of patients with progression to ESKD was significantly lower in the group with baseline HbA_1c of 7–9% (53–75 mmol/mol) whereas all-cause mortality was comparable in the three groups “Table 1”.

Download:

Table 1. Demographic and clinical characteristics of the study population by baseline HbA_1c categories.

https://doi.org/10.1371/journal.pone.0274605.t001

Characteristics of participants by HbA_1c trajectories

Overall, 2692 patients were enrolled in the trajectory analysis, and the median number of HbA_1c measurements was 8 (IQR: 5–14) per patient during the study period. The median follow-up duration was 2.6 (IQR: 1.6–4.0) years for the development of ESKD and 4.4 (IQR: 2.7–6.5) years for all-cause mortality. Three distinct longitudinal HbA_1c trajectories were identified by the GBTM: nearly optimal (55.9%), moderate-to-stable (34.2%), and poorly controlled (9.9%) (Fig 2). The HbA_1c trajectory of the “nearly-optimal” group was stably below a HbA_1c level of 7% (53 mmol/mol), whereas the “moderate-to-stable” and “poorly controlled” groups had HbA_1c trajectories that fluctuated at approximately 8% (64 mmol/mol) and 10% (86 mmol/mol), respectively. Both the “moderate-to-stable” and “poorly controlled” groups had downward trends in HbA_1c levels during the follow-up, particularly in the case of the “poorly controlled” group. Compared with the “nearly optimal” and “moderate-to-stable” groups, patients in the “poorly controlled” group—similar to those with baseline HbA_1c levels of > 9% (> 75 mmol/mol)—were younger and tended to have longer follow-up for progression to ESKD and a higher BMI. Those patients were also less likely to have baseline CKD stage 4–5 with a corresponding higher eGFR at baseline (S1 Table). Deviating slightly from the observations made in the baseline analysis, the proportion of patients with progression to ESKD was slightly higher in the “poorly controlled” group than in the other groups, whereas all-cause mortality was comparable among the three groups (S1 Table).

Download:

Fig 2. HbA_1c trajectories by group-based trajectory modeling as per the three-trajectory solution.

The solid lines indicate the mean estimated trajectories; the points represent the mean observed trajectories.

https://doi.org/10.1371/journal.pone.0274605.g002

ESKD risk and all-cause mortality based on baseline HbA_1c and HbA_1c trajectory

The analysis of baseline HbA_1c, revealed that 2053 ESKD events and 1698 deaths occurred over a total 9888 and 19 253 person-years of follow-up, respectively. The incidence and HR of developing ESKD and all-cause mortality are revealed in Table 2. In the unadjusted model, a modest inverse association was found between the baseline HbA1c category and risk of ESKD (P for trend = 0.028); the HR (95% CI) for a baseline HbA1c level of 7%–9% (53–75 mmol/mol) and >9% (> 75 mmol/mol) versus an HbA_1c level of <7% (53 mmol/mol) was 0.92 (0.83–1.01) and 0.87 (0.76–1.002), respectively. The inverse association remained, although attenuated, after controlling for demographics, smoking, and alcohol consumption in model 1 and additionally after controlling for systolic blood pressure, cardiovascular disease, lipid levels, primary etiologies of CKD, and medication use in model 2. However, in model 3, in which baseline hemoglobin, eGFR, and pooled uPCR were additionally controlled, the inverse association between the baseline HbA_1c level and risk of progression to ESKD was not found. We did not discover a significant association between the baseline HbA_1c level and all-cause mortality in the unadjusted model and adjusted models 1 and 2. However, a positive association was found in model 3 (P for trend = 0.009); the HR (95% CI) for a baseline HbA_1c of >9% (>75 mmol/mol) versus <7% (53 mmol/mol) was 1.25 (1.07–1.46). The dose–response curve between baseline HbA_1c levels and the risk of all-cause mortality in model 3 showed a monotonic relationship (P = 0.02; Fig 3B), but such a relationship did not appear between baseline HbA_1c and risk of ESKD (Fig 3A). An exploratory subgroup analysis revealed that the associations between the baseline HbA_1c level and risk of all-cause mortality were consistent in patient subgroups stratified in accordance with in priori selected variables (Fig 4). Generally, the association between a high baseline HbA_1c level and all-cause mortality was stronger in patients free of advanced CKD, hypertension, and CVD.

Download:

Fig 3.

Dose-response plot of the baseline HbA1c and adjusted hazard ratios for (A) progression to end-stage kidney disease and (B) all-cause mortality according to baseline HbA_1c (%). Solid lines represent adjusted hazard ratios based on restricted cubic splines for baseline HbA_1c, with knots at the 10th, 50th, and 90th percentiles. Shaded areas represent the upper and lower 95% confidence intervals. The reference was set at the 10th percentile of HbA_1c levels. Variables adjusted are the same as that shown in Model 3 presented in Table 2. Missing values were imputed by multiple imputation.

https://doi.org/10.1371/journal.pone.0274605.g003

Download:

Fig 4. Subgroup analysis of the hazard ratios (95% confidence interval) of all-cause mortality associated with baseline HbA_1c groups.

BMI: body mass index, CKD: chronic kidney disease, CVD: cardiovascular disease.

https://doi.org/10.1371/journal.pone.0274605.g004

Download:

Table 2. Hazard ratios (95% confidence interval) of progression to end-stage kidney disease (ESKD) and all-cause mortality associated with baseline HbA_1c and HbA_1c trajectory groups.

https://doi.org/10.1371/journal.pone.0274605.t002

In analysis of HbA_1c trajectories, 1157 ESKD events and 799 deaths occurred over a total 8116 and 12 745 person-years of follow-up, respectively. No significant association was discovered between the HbA1c trajectory categories and risk of developing ESKD in the unadjusted model or any adjusted model (Table 2). In model 3, the adjusted HR (95% CI) of progression to ESKD was 1.03 (0.92–1.16) for the “moderate-to-stable” HbA_1c group and 1.13 (0.94–1.35) for the “poor control” group as compared with the “nearly optimal” group (Table 2). However, the HbA1c trajectory categories were associated with all-cause mortality. In model 3, the “poor control” group had 35% higher risk of mortality (6%–71%) than the “nearly optimal” group (Table 2).

Discussion

Our findings revealed that a high HbA_1c level at the time of Pre-ESKD Program enrollment and a poorly controlled HbA_1c trajectory over the follow-up period were associated with increased risk of all-cause mortality in patients with type 2 diabetes and CKD. Despite emerging evidence endorsing the relaxation of HbA_1c as a goal for older patients with multiple comorbidities including CKD, maintaining the longitudinal HbA_1c level at <9% (75 mmol/mol) remains vital for improving patients’ overall survival. The null associations of progression to ESKD with a high baseline HbA_1c level and a “poorly controlled” HbA_1c trajectory should be interpreted cautiously, because the risk of ESKD associated with HbA_1c levels may have been modified by the differential erythrocyte lifespan between early and advanced of CKD.

The relatively linear dose–response relationship between the baseline HbA_1c level and risk of all-cause mortality was inconsistent with the findings of a study by Shurraw et al. [4], who found a U-shaped relationship between baseline HbA_1c and all-cause mortality in patients with CKD stage 3 and 4. In further analyses of Shurraw et al. in which CKD stages 3 and 4 were stratified separately, the magnitude of the increased risk of ESKD associated with poor glycemic control—single baseline HbA_1c level >9% (>75 mmol/mol) as opposed to <7% (<53 mmol/mol)—was greater among patients with CKD stage 3 than among those with CKD stage 4 [4]. This minor discrepancy between our study and that of Shurraw et al. is likely due to differences in the study population and statistical approaches. First, the present study further included patients with CKD stage 5 and patients with CKD stages 1–3a with evident proteinuria. Second, our study was based on a well-interoperated dataset incorporating a single institution’s electronic medical records and the national Pre-ESKD Program, meaning that our confounding control was better because variables such as smoking status, alcohol consumption, hemoglobin level, proteinuria, lipid profile, and medication use were available [4]. Conversely, most studies have demonstrated a positive association between baseline HbA_1c and the risk of progression to ESKD [4, 5]. On the basis of overarching findings across studies including our own, we can conclude that avoiding an HbA_1c level of >9% (>75 mmol/mol) is likely to benefit patients with type 2 diabetes and CKD, even when the disease stage is advanced. Notably, we did not observe better kidney prognosis and mortality outcomes in patients with an HbA_1c level of <7% (<53 mmol/mol) compared with those having an HbA_1c level of 7%–9% (53–75 mmol/mol). However, whether the therapeutic goal of an HbA_1c level >7%–7.5% should be relaxed is beyond the scope of this study and requires clinical and research consensus concerning the definition of intensive glycemic control for patients with diabetes and CKD (ADA 2022) [24].

Few empirical studies have explored the prognostic role of longitudinal trends in HbA_1c in patients with type 2 diabetes and CKD. A study of 770 patients with type 2 diabetes and CKD demonstrated that a “moderate increase” HbA_1c trajectory was associated with increased risk of CKD progression compared with a “near-optimal stable” trajectory [25]. The kidney function of that study population was relatively well preserved (median eGFR = 84.8 mL/min/1.73 m²). In addition, instead of the development of ESKD, CKD progression was defined by a decline in CKD stage with a ≥25% reduction of baseline eGFR [25]. The consistently observed association between a poor glycemic control trajectory and increased risks of ESKD and mortality in patients with diabetes and CKD highlights the importance of taking proactive measures to prevent hyperglycemic states over the course of CKD care. An integrated CKD care program should target diabetic patients with poor long-term glycemic control, particularly those with early CKD.

This study has several limitations. First, this was a retrospective cohort study and we could not derive causal inferences from its results. Second, selection bias due to correlation between the HbA_1c level and CKD stage should be considered. Briefly, patients with more advanced CKD would have shorter red blood cell survival, leading to a relatively low HbA_1c level even in a similar glycemic milieu. Patients with CKD in the lower stratum of baseline HbA_1c were more likely to have a more advanced CKD stage and therefore progress more rapidly to ESKD. To minimize this bias, we created a proxy outcome for progression to ESKD—a doubling of the S-Cre concentration—in the main analysis; this balanced the risk of dialysis among the three baseline HbA_1c groups. We also restricted our analysis of patients with CKD stage 3 and found that a baseline HbA_1c level of >9% (>75 mmol/mol) was significantly associated with increased risk of progression to ESKD [aHR 1.35 (95% CI, 1.04–1.75)] but not increased risk of mortality [aHR 1.05 (95% CI, 0.78–1.41)], as compared with the <7% (53 mmol/mol) group. This observation provides a complimentary perspective to our main findings and supports the hypothesis that an HbA_1c level > 9%, is associated with increased risk of both ESKD and mortality in patients with CKD. The effect of poorly controlled HbA_1c may be modified by an inherited propensity toward outcomes of interest, which was introduced by the differential erythrocyte lifespan across CKD stages. Third, the possibility of residual confounding—such as a lack of access to detailed dietary information and compliance with medication—could not be completely excluded. Fourth, the follow-up duration may have been insufficient to observe the development of ESKD in patients with CKD stages 1–3. To minimize the impact of the potentially insufficient follow-up, we also used doubling of S-Cre as a surrogate endpoint to define the progression of CKD to ESKD.

Conclusion

In individuals with CKD and type 2 diabetes, maintaining the HbA_1c level < 9% (<75 mmol/mol) remains crucial for halting CKD progression and reducing the mortality risk. Patients in the early stages of CKD were particularly vulnerable to the negative effects of chronic hyperglycemia and accelerated progression to ESKD. Whether the development and integration of a glycemic optimization protocol into the existing CKD program for patients with diabetes and CKD can help lower the CKD-related healthcare burden requires clinical trial validation and thus warrants further study.

Supporting information

S1 Table. Demographic and clinical characteristics of the study population by the longitudinal HbA_1c trajectories.

https://doi.org/10.1371/journal.pone.0274605.s001

(DOCX)

S2 Table. Hazard ratios (95% confidence interval) of 30% decline of estimated glomerular filtration rate (eGFR), doubling serum creatinine, progression to end-stage kidney disease (ESKD), and all-cause mortality associated with baseline HbA_1c groups.

https://doi.org/10.1371/journal.pone.0274605.s002

(DOCX)

Acknowledgments

We appreciate the support from the Big Data Center of CMUH for data exploration, statistical analysis, manuscript preparation, and the support of the iHi Clinical Research Platform. We would like to thank the Health and Welfare Data Science Center, Ministry of Health Welfare, and Health Data Science Center, China Medical University Hospital for providing administrative, technical support.

Prior presentation

Part of this study was presented as abstract at Joint Conference on Medical Informatics in Taiwan 2020 on Air, 24–30 August 2020.

References

1. System USRD. 2020 USRDS Annual Data Report: Epidemiology of kidney disease in the United States. National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD: 2020.
2. Currie CJ, Peters JR, Tynan A, Evans M, Heine RJ, Bracco OL, et al. Survival as a function of HbA1c in people with type 2 diabetes: a retrospective cohort study. The Lancet. 2010;375(9713):481–9. pmid:20110121
- View Article
- PubMed/NCBI
- Google Scholar
3. Papademetriou V, Lovato L, Doumas M, Nylen E, Mottl A, Cohen RM, et al. Chronic kidney disease and intensive glycemic control increase cardiovascular risk in patients with type 2 diabetes. 2015;87(3):649–59. pmid:25229335
- View Article
- PubMed/NCBI
- Google Scholar
4. Shurraw S, Hemmelgarn B, Lin M, Majumdar SR, Klarenbach S, Manns B, et al. Association between glycemic control and adverse outcomes in people with diabetes mellitus and chronic kidney disease: a population-based cohort study. Archives of Internal Medicine. 2011;171(21):1920–7. pmid:22123800
- View Article
- PubMed/NCBI
- Google Scholar
5. Kuo I-C, Lin HY-H, Niu S-W, Hwang D-Y, Lee J-J, Tsai J-C, et al. Glycated hemoglobin and outcomes in patients with advanced diabetic chronic kidney disease. Scientific reports. 2016;6:20028. pmid:26818011
- View Article
- PubMed/NCBI
- Google Scholar
6. Navaneethan SD, Zoungas S, Caramori ML, Chan JC, Heerspink HJ, Hurst C, et al. Diabetes Management in Chronic Kidney Disease: Synopsis of the 2020 KDIGO Clinical Practice Guideline. Annals of Internal Medicine. 2020. pmid:33166222
- View Article
- PubMed/NCBI
- Google Scholar
7. Group AC. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. New England journal of medicine. 2008;358(24):2560–72. pmid:18539916
- View Article
- PubMed/NCBI
- Google Scholar
8. Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HAW. 10-year follow-up of intensive glucose control in type 2 diabetes. New England journal of medicine. 2008;359(15):1577–89. pmid:18784090
- View Article
- PubMed/NCBI
- Google Scholar
9. Ruospo M, Saglimbene VM, Palmer SC, De Cosmo S, Pacilli A, Lamacchia O, et al. Glucose targets for preventing diabetic kidney disease and its progression. Cochrane Database of Systematic Reviews. 2017;(6). pmid:28594069
- View Article
- PubMed/NCBI
- Google Scholar
10. Zoungas S, Chalmers J, Ninomiya T, Li Q, Cooper M, Colagiuri S, et al. Association of HbA 1c levels with vascular complications and death in patients with type 2 diabetes: evidence of glycaemic thresholds. Diabetologia. 2012;55(3):636–43. pmid:22186981
- View Article
- PubMed/NCBI
- Google Scholar
11. Zoungas S, Arima H, Gerstein HC, Holman RR, Woodward M, Reaven P, et al. Effects of intensive glucose control on microvascular outcomes in patients with type 2 diabetes: a meta-analysis of individual participant data from randomised controlled trials. The lancet Diabetes & endocrinology. 2017;5(6):431–7. pmid:28365411
- View Article
- PubMed/NCBI
- Google Scholar
12. Lin CM, Yang MC, Hwang SJ, Sung JM. Progression of stages 3b-5 chronic kidney disease—preliminary results of Taiwan national pre-ESRD disease management program in Southern Taiwan. J Formos Med Assoc. 2013;112(12):773–82. Epub 2013/12/07. pmid:24309170.
- View Article
- PubMed/NCBI
- Google Scholar
13. National Kidney Foundation. KDOQI Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification and Stratification. Am J Kidney Dis. 2002;39:S1–S000.
- View Article
- Google Scholar
14. King EK, Hsieh MH, Chang DR, Lu CT, Ting IW, Wang CCN, et al. Prediction of non-responsiveness to pre-dialysis care program in patients with chronic kidney disease: a retrospective cohort analysis. Sci Rep. 2021;11(1):13938. Epub 2021/07/08. pmid:34230524.
- View Article
- PubMed/NCBI
- Google Scholar
15. Tsai CW, Huang HC, Chiang HY, Chung CW, Chiu HT, Liang CC, et al. First-year estimated glomerular filtration rate variability after pre-end-stage renal disease program enrollment and adverse outcomes of chronic kidney disease. Nephrol Dial Transplant. 2019;34(12):2066–78. Epub 2018/07/10. pmid:29982714.
- View Article
- PubMed/NCBI
- Google Scholar
16. Chapelle J-P, Teixeira J, Maisin D, Assink H, Barla G, Stroobants AK, et al. Multicentre evaluation of the Tosoh HbA1c G8 analyser. Clinical chemistry and laboratory medicine. 2010;48(3):365–71. pmid:20020820
- View Article
- PubMed/NCBI
- Google Scholar
17. Association AD. 6. Glycemic Targets: Standards of Medical Care in Diabetes—2020. Diabetes Care. 2020;43(Supplement 1):S66–S76. pmid:31862749
- View Article
- PubMed/NCBI
- Google Scholar
18. Tsai C-W, Chiu H-T, Huang H-C, Ting I-W, Yeh H-C, Kuo C-C. Uric acid predicts adverse outcomes in chronic kidney disease: a novel insight from trajectory analyses. Nephrology Dialysis Transplantation. 2018;33(2):231–41. pmid:29140472
- View Article
- PubMed/NCBI
- Google Scholar
19. Levey AS, Stevens LA, Schmid CH, Zhang Y, Castro AF III, Feldman HI, et al. A new equation to estimate glomerular filtration rate. Annals of internal medicine. 2009;150(9):604–12. pmid:19414839
- View Article
- PubMed/NCBI
- Google Scholar
20. Yamamoto K, Yamamoto H, Yoshida K, Niwa K, Nishi Y, Mizuno A, et al. The total urine protein-to-creatinine ratio can predict the presence of microalbuminuria. PLoS One. 2014;9(3):e91067. pmid:24614247
- View Article
- PubMed/NCBI
- Google Scholar
21. Nagin DS. Analyzing developmental trajectories: a semiparametric, group-based approach. Psychological methods. 1999;4(2):139.
- View Article
- Google Scholar
22. Nagin DS, Odgers CL. Group-based trajectory modeling in clinical research. Annual review of clinical psychology. 2010;6:109–38. pmid:20192788
- View Article
- PubMed/NCBI
- Google Scholar
23. Jones BL, Nagin DS, Roeder K. A SAS procedure based on mixture models for estimating developmental trajectories. Sociological methods & research. 2001;29(3):374–93.
- View Article
- Google Scholar
24. Schernthaner G, Ritz E, Schernthaner G-H. Strict glycaemic control in diabetic patients with CKD or ESRD: beneficial or deadly? Nephrology Dialysis Transplantation. 2010;25(7):2044–7. pmid:20388630
- View Article
- PubMed/NCBI
- Google Scholar
25. Low S, Zhang X, Wang J, Yeoh LY, Liu YL, Ang SF, et al. The impact of HbA1c trajectories on chronic kidney disease progression in type 2 diabetes. Nephrology. 2018. https://doi.org/0.1093/ndt/gfy104.SP428.
- View Article
- Google Scholar

[ref1] 1. System USRD. 2020 USRDS Annual Data Report: Epidemiology of kidney disease in the United States. National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD: 2020.

[ref2] 2. Currie CJ, Peters JR, Tynan A, Evans M, Heine RJ, Bracco OL, et al. Survival as a function of HbA1c in people with type 2 diabetes: a retrospective cohort study. The Lancet. 2010;375(9713):481–9. pmid:20110121
View Article
PubMed/NCBI
Google Scholar

[3] View Article

[4] PubMed/NCBI

[5] Google Scholar

[ref3] 3. Papademetriou V, Lovato L, Doumas M, Nylen E, Mottl A, Cohen RM, et al. Chronic kidney disease and intensive glycemic control increase cardiovascular risk in patients with type 2 diabetes. 2015;87(3):649–59. pmid:25229335
View Article
PubMed/NCBI
Google Scholar

[7] View Article

[8] PubMed/NCBI

[9] Google Scholar

[ref4] 4. Shurraw S, Hemmelgarn B, Lin M, Majumdar SR, Klarenbach S, Manns B, et al. Association between glycemic control and adverse outcomes in people with diabetes mellitus and chronic kidney disease: a population-based cohort study. Archives of Internal Medicine. 2011;171(21):1920–7. pmid:22123800
View Article
PubMed/NCBI
Google Scholar

[11] View Article

[12] PubMed/NCBI

[13] Google Scholar

[ref5] 5. Kuo I-C, Lin HY-H, Niu S-W, Hwang D-Y, Lee J-J, Tsai J-C, et al. Glycated hemoglobin and outcomes in patients with advanced diabetic chronic kidney disease. Scientific reports. 2016;6:20028. pmid:26818011
View Article
PubMed/NCBI
Google Scholar

[15] View Article

[16] PubMed/NCBI

[17] Google Scholar

[ref6] 6. Navaneethan SD, Zoungas S, Caramori ML, Chan JC, Heerspink HJ, Hurst C, et al. Diabetes Management in Chronic Kidney Disease: Synopsis of the 2020 KDIGO Clinical Practice Guideline. Annals of Internal Medicine. 2020. pmid:33166222
View Article
PubMed/NCBI
Google Scholar

[19] View Article

[20] PubMed/NCBI

[21] Google Scholar

[ref7] 7. Group AC. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. New England journal of medicine. 2008;358(24):2560–72. pmid:18539916
View Article
PubMed/NCBI
Google Scholar

[23] View Article

[24] PubMed/NCBI

[25] Google Scholar

[ref8] 8. Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HAW. 10-year follow-up of intensive glucose control in type 2 diabetes. New England journal of medicine. 2008;359(15):1577–89. pmid:18784090
View Article
PubMed/NCBI
Google Scholar

[27] View Article

[28] PubMed/NCBI

[29] Google Scholar

[ref9] 9. Ruospo M, Saglimbene VM, Palmer SC, De Cosmo S, Pacilli A, Lamacchia O, et al. Glucose targets for preventing diabetic kidney disease and its progression. Cochrane Database of Systematic Reviews. 2017;(6). pmid:28594069
View Article
PubMed/NCBI
Google Scholar

[31] View Article

[32] PubMed/NCBI

[33] Google Scholar

[ref10] 10. Zoungas S, Chalmers J, Ninomiya T, Li Q, Cooper M, Colagiuri S, et al. Association of HbA 1c levels with vascular complications and death in patients with type 2 diabetes: evidence of glycaemic thresholds. Diabetologia. 2012;55(3):636–43. pmid:22186981
View Article
PubMed/NCBI
Google Scholar

[35] View Article

[36] PubMed/NCBI

[37] Google Scholar

[ref11] 11. Zoungas S, Arima H, Gerstein HC, Holman RR, Woodward M, Reaven P, et al. Effects of intensive glucose control on microvascular outcomes in patients with type 2 diabetes: a meta-analysis of individual participant data from randomised controlled trials. The lancet Diabetes & endocrinology. 2017;5(6):431–7. pmid:28365411
View Article
PubMed/NCBI
Google Scholar

[39] View Article

[40] PubMed/NCBI

[41] Google Scholar

[ref12] 12. Lin CM, Yang MC, Hwang SJ, Sung JM. Progression of stages 3b-5 chronic kidney disease—preliminary results of Taiwan national pre-ESRD disease management program in Southern Taiwan. J Formos Med Assoc. 2013;112(12):773–82. Epub 2013/12/07. pmid:24309170.
View Article
PubMed/NCBI
Google Scholar

[43] View Article

[44] PubMed/NCBI

[45] Google Scholar

[ref13] 13. National Kidney Foundation. KDOQI Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification and Stratification. Am J Kidney Dis. 2002;39:S1–S000.
View Article
Google Scholar

[47] View Article

[48] Google Scholar

[ref14] 14. King EK, Hsieh MH, Chang DR, Lu CT, Ting IW, Wang CCN, et al. Prediction of non-responsiveness to pre-dialysis care program in patients with chronic kidney disease: a retrospective cohort analysis. Sci Rep. 2021;11(1):13938. Epub 2021/07/08. pmid:34230524.
View Article
PubMed/NCBI
Google Scholar

[50] View Article

[51] PubMed/NCBI

[52] Google Scholar

[ref15] 15. Tsai CW, Huang HC, Chiang HY, Chung CW, Chiu HT, Liang CC, et al. First-year estimated glomerular filtration rate variability after pre-end-stage renal disease program enrollment and adverse outcomes of chronic kidney disease. Nephrol Dial Transplant. 2019;34(12):2066–78. Epub 2018/07/10. pmid:29982714.
View Article
PubMed/NCBI
Google Scholar

[54] View Article

[55] PubMed/NCBI

[56] Google Scholar

[ref16] 16. Chapelle J-P, Teixeira J, Maisin D, Assink H, Barla G, Stroobants AK, et al. Multicentre evaluation of the Tosoh HbA1c G8 analyser. Clinical chemistry and laboratory medicine. 2010;48(3):365–71. pmid:20020820
View Article
PubMed/NCBI
Google Scholar

[58] View Article

[59] PubMed/NCBI

[60] Google Scholar

[ref17] 17. Association AD. 6. Glycemic Targets: Standards of Medical Care in Diabetes—2020. Diabetes Care. 2020;43(Supplement 1):S66–S76. pmid:31862749
View Article
PubMed/NCBI
Google Scholar

[62] View Article

[63] PubMed/NCBI

[64] Google Scholar

[ref18] 18. Tsai C-W, Chiu H-T, Huang H-C, Ting I-W, Yeh H-C, Kuo C-C. Uric acid predicts adverse outcomes in chronic kidney disease: a novel insight from trajectory analyses. Nephrology Dialysis Transplantation. 2018;33(2):231–41. pmid:29140472
View Article
PubMed/NCBI
Google Scholar

[66] View Article

[67] PubMed/NCBI

[68] Google Scholar

[ref19] 19. Levey AS, Stevens LA, Schmid CH, Zhang Y, Castro AF III, Feldman HI, et al. A new equation to estimate glomerular filtration rate. Annals of internal medicine. 2009;150(9):604–12. pmid:19414839
View Article
PubMed/NCBI
Google Scholar

[70] View Article

[71] PubMed/NCBI

[72] Google Scholar

[ref20] 20. Yamamoto K, Yamamoto H, Yoshida K, Niwa K, Nishi Y, Mizuno A, et al. The total urine protein-to-creatinine ratio can predict the presence of microalbuminuria. PLoS One. 2014;9(3):e91067. pmid:24614247
View Article
PubMed/NCBI
Google Scholar

[74] View Article

[75] PubMed/NCBI

[76] Google Scholar

[ref21] 21. Nagin DS. Analyzing developmental trajectories: a semiparametric, group-based approach. Psychological methods. 1999;4(2):139.
View Article
Google Scholar

[78] View Article

[79] Google Scholar

[ref22] 22. Nagin DS, Odgers CL. Group-based trajectory modeling in clinical research. Annual review of clinical psychology. 2010;6:109–38. pmid:20192788
View Article
PubMed/NCBI
Google Scholar

[81] View Article

[82] PubMed/NCBI

[83] Google Scholar

[ref23] 23. Jones BL, Nagin DS, Roeder K. A SAS procedure based on mixture models for estimating developmental trajectories. Sociological methods & research. 2001;29(3):374–93.
View Article
Google Scholar

[85] View Article

[86] Google Scholar

[ref24] 24. Schernthaner G, Ritz E, Schernthaner G-H. Strict glycaemic control in diabetic patients with CKD or ESRD: beneficial or deadly? Nephrology Dialysis Transplantation. 2010;25(7):2044–7. pmid:20388630
View Article
PubMed/NCBI
Google Scholar

[88] View Article

[89] PubMed/NCBI

[90] Google Scholar

[ref25] 25. Low S, Zhang X, Wang J, Yeoh LY, Liu YL, Ang SF, et al. The impact of HbA1c trajectories on chronic kidney disease progression in type 2 diabetes. Nephrology. 2018. https://doi.org/0.1093/ndt/gfy104.SP428.
View Article
Google Scholar

[92] View Article

[93] Google Scholar

Figures

Abstract

Introduction

Materials and methods

Study population

Measurement of HbA1c

Other covariables

Outcomes and follow-up

Statistical analyses

Ethics statement

Results

Characteristics of study subjects by baseline HbA1c levels

Characteristics of participants by HbA1c trajectories

ESKD risk and all-cause mortality based on baseline HbA1c and HbA1c trajectory

Discussion

Conclusion

Supporting information

S1 Table. Demographic and clinical characteristics of the study population by the longitudinal HbA1c trajectories.

S2 Table. Hazard ratios (95% confidence interval) of 30% decline of estimated glomerular filtration rate (eGFR), doubling serum creatinine, progression to end-stage kidney disease (ESKD), and all-cause mortality associated with baseline HbA1c groups.

Acknowledgments

Prior presentation

References

Measurement of HbA_1c

Characteristics of study subjects by baseline HbA_1c levels

Characteristics of participants by HbA_1c trajectories

ESKD risk and all-cause mortality based on baseline HbA_1c and HbA_1c trajectory

S1 Table. Demographic and clinical characteristics of the study population by the longitudinal HbA_1c trajectories.

S2 Table. Hazard ratios (95% confidence interval) of 30% decline of estimated glomerular filtration rate (eGFR), doubling serum creatinine, progression to end-stage kidney disease (ESKD), and all-cause mortality associated with baseline HbA_1c groups.