Investigating the Patterns of Renal Function Variability in Early-Stage Chronic Kidney Disease by Cluster Analysis

Investigating the Patterns of Renal Function Variability in Early-Stage Chronic Kidney Disease by Cluster Analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Investigating the Patterns of Renal Function Variability in Early-Stage Chronic Kidney Disease by Cluster Analysis Arisa Kobayashi, Tsuyoshi Ohnishi, Tadahisa Okuda, Keita Hirano, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5921794/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 05 Aug, 2025 Read the published version in BMC Nephrology → Version 1 posted 4 You are reading this latest preprint version Abstract Background: Chronic kidney disease (CKD) is a significant global health concern, with increasing focus on predicting renal prognosis. While renal prognosis is often studied in advanced CKD, variability in renal function and its implications for long-term outcomes in early-stage CKD remain insufficiently examined. This study aimed to investigate renal prognosis in early-stage CKD within the general population, focusing on patterns of renal function variability and factors associated with high variability. Methods: This retrospective nationwide cohort study included participants from various geographical regions across Japan, representing a diverse general population. A total of 1,765 adults with early-stage CKD (eGFR 45–59 mL/min/1.73 m 2 ), based on two initial screening results, were analyzed. The primary outcome was the pattern of eGFR variability identified by cluster analysis using three parameters: mean residual (difference between linear prediction and observed value), maximum residual, and range. In addition, we used a logistic regression model in order to assess associations between clinical factors and the high-risk cluster. Results: We identified three distinct clusters based on eGFR variability using cluster analysis. Among these clusters, one exhibited significantly high variability with a high residual (median of mean residuals of 10.9 mL/min/1.73 m 2 and median of maximum residuals of 22.6 mL/min/1.73 m 2 ) and a wide range (median of range of 25.1 mL/min/1.73 m 2 ) (referred to as the "high variability cluster"). This cluster, comprising 4.6% of patients with early-stage CKD, demonstrated a more pronounced decline in eGFR over time. Factors such as younger age, proteinuria, antihypertensive drug use, and hyperglycemia were associated with the high variability cluster. Conclusions: This study highlights the presence of distinct eGFR variability patterns in early-stage CKD and identifies a subgroup at high risk for rapid renal decline. Monitoring eGFR variability provides critical insights into long-term prognosis and may inform targeted interventions. Considering these findings, early detection and management of patients with early CKD may improve disease progression and reduce the risk of adverse outcomes. Trial registration: This study is an observational study using a database and does not involve a health care intervention on human participants. Therefore, trial registration is not applicable. Chronic renal disease Early-Stage chronic kidney disease Renal prognosis Health screening Cluster analysis Figures Figure 1 Figure 2 Figure 3 Introduction Chronic kidney disease (CKD) is a major public health concern, affecting a significant proportion of the general population. Despite its high prevalence, CKD often remains undiagnosed due to the lack of specific symptoms during its early stages, and may lead to end-stage renal disease [ 1 ]. Progression of CKD may result in serious complications, including cardiovascular disease, and increased mortality risk [ 2 – 4 ]. Early detection and appropriate management of CKD are essential to prevent these complications. Staging CKD based on estimated glomerular filtration rate (eGFR) and urinary protein levels has been one of the important assessment methods for the management of CKD risk [ 5 ]. On the other hand, renal function ofte exhibit variability over time, and the implications of this variability remain insufficiently explored. Although changes in renal function and their associations with end-stage renal disease, dialysis initiation, and mortality have been reported [ 6 – 9 ], there is a lack of studies that comprehensively evaluate renal function variability and its prognostic implications using multiple parameters over sufficiently long observation periods. In addition, variability in renal function among patients with early-stage CKD, who are often identified through health screening, has not been thoroughly investigated. This gap highlights the need for stronger evidence to guide appropriate follow-up strategies after screening and to establish clearer recommendations for long-term monitoring. In Japan, a nationwide health screening system facilitates the early detection of CKD in asymptomatic individuals, offering an opportunity to examine the early stage of CKD in the general population. However, the clinical presentation of early-stage CKD is highly heterogeneous. Not all individuals with early-stage CKD progress to more advanced stages or require the same level of medical intervention [ 10 ]. Effective management strategies for early-stage CKD must therefore consider not only baseline measures of renal function but also mid- to long-term changes, including fluctuations and trends in eGFR [ 11 ]. Most studies focus on the eGFR slope—a measure of the linear rate of decline in renal function over time—as a prognostic indicator. However, by taking into account the variability of eGFR as well as the slope, it may be possible to reflect unstable periods that have clinical significance. Moreover, identifying patterns of eGFR variability using advanced techniques such as clustering analysis could provide valuable insights into disease trajectories and help stratify patients by risk. Nevertheless, research using this approach is not sufficient, and further research is needed to establish whether eGFR variability can offer unique prognostic value beyond traditional measures like the eGFR slope. To address these gaps, we leveraged a nationwide health screening cohort in Japan to investigate long-term changes in renal function among individuals with early-stage CKD. This study aimed to identify distinct patterns of renal function variability using clustering analysis, and explore the clinical and demographic factors associated with high-risk patterns of variability. By focusing on eGFR variability, this study seeks to advance our understanding of early-stage CKD progression and inform more personalized approaches to its management. Materials and Methods Data source and setting Data from a nationwide health-screening cohort between April 2011 and March 2022, provided by one of Japan’s largest employment-based health insurers (a national sample of employees of civil engineering and construction companies), were analyzed. We obtained health screening data from annual health assessments conducted by the insurer among insured adults [12]. The dataset included information on demographic characteristics (age, sex, and body mass index [BMI]), clinical factors (systolic blood pressure [SBP], diastolic blood pressure [DBP], and glycated hemoglobin (HbA1c) level), medication use (antihypertensive and antidiabetic drugs), and smoking status. Additionally, eGFR values were calculated using the Chronic Kidney Disease Epidemiology Collaboration equation for Japanese individuals [13]. Urinary protein levels were measured using the dip-stick test, and proteinuria was defined as a score of 1+ or higher. The basic clinical information was evaluated on the first health checkup date. This study received approval from the Institutional Review Board (IRB) of Kyoto University (IRB number: R0817). As we analyzed only anonymized data, the IRB waived the need for informed consent. This study complied with the principles of the Declaration of Helsinki. Participant selection We excluded individuals who had less than two eGFR measurements from all participants. Subsequently, the sample was further narrowed by excluding individuals with eGFR values, either first or second eGFR of ≥60 or <45 mL/min/1.73 m 2 . This selection criterion allowed us to focus on the early-stage CKD (CKD stage G3a) population. This is because many patients with CKD detected during health checkups are at an early stage and clinical decisions regarding the need for renal function monitoring in this population may be difficult to make. K-means clustering by eGFR variability To identify patterns of eGFR variability, we employed K-means clustering, focusing on the following three parameters: mean residual (average of the difference between linear predictions and observed values), maximum residual (highest difference between linear predictions and observed values), and range of values (difference between maximum and minimum observed values) (Additional file 1). Linear predictions were estimated using a linear regression model for repeated eGFR measurements. K-means clustering is a widely used method for grouping data and is effective in revealing significant insights from large datasets through exploratory analysis. This statistical method groups data based on similarities in their characteristics by focusing on the differences in the characteristics of the data. While the yearly eGFR decline slope is usually used as a surrogate endpoint for evaluating renal prognosis [14, 15], the slope assumes a linear decline in eGFR annually. Although this slope is useful for approximating linear trends in population-wide changes in renal function, it may not accurately reflect individual variations. In clinical practice, deviations from linear eGFR changes due to various factors that influence renal function variability are frequently observed [16]. Therefore, we used the residual and range for classification instead of the slope to assess the renal function variability more accurately. We initially calculated the residuals between the linear predictions and actual eGFR values for each individual. Subsequently, we computed the mean and maximum of these residuals and determined the range of eGFR values during the observation period. These parameters were standardized by dividing each of them by their standard deviations to ensure consistent scaling. The participants were then classified into three clusters using K-means clustering. The number of clusters was determined to be three based on clinical interpretability and evaluated using the general thresholds for the sum of squared errors (SSE) and the elbow method (Additional file 2). In addition, we conducted a sensitivity analysis with varying numbers of clusters to validate our findings. The cluster definitions were aligned with the observed patterns of eGFR variability, and the second and subsequent eGFR values were used to define the eGFR variability. At least three measurements were used to capture the variability over a certain period. The baseline eGFR values of the population with eGFR >60 or <45 mL/min/1.73 m 2 are, to some extent, defined by the baseline eGFR values for subsequent eGFR changes. Therefore, they were excluded from this study. Participant characteristics and eGFR change Participant characteristics were described according to eGFR variability clusters. We defined hypertension as SBP ≥140 mmHg or DBP ≥90 mmHg and diabetes as an HbA1c level ≥6.5%. After stratification by eGFR variability clusters, we modeled nonlinear relationships between time (follow-up years) and eGFR using a fractional polynomial model separately. The predicted and observed eGFR values over time were plotted. Association between clinical factors and the high variability cluster We examined the associations between clinical factors and being in the high-variability cluster using a logistic regression model. Clinical factors such as age, sex, BMI, urine protein, smoking status, blood pressure, blood glucose, and the use of medications for hypertension or diabetes were included as explaratory variables. In this analysis, we defined high blood pressure as SBP of 140 or higher or DBP of 90 or higher. Also we defined high blood glucose as HbA1c of 6.5% or higher. Statistical analysis Patient characteristics are presented as medians (interquartile ranges) for continuous variables and numerical values (percentages) for categorical variables. Differences in background characteristics between the clusters were assessed utilizing the chi-square test for binary variables and the t-test for continuous variables. A sensitivity analysis was conducted employing different numbers of clusters to examine the effect of cluster quantity on the results. All statistical analyses were performed using STATA MP (version 18.0; STATA Corporation, College Station, TX, US). For all analyses, a two-tailed p value < 0.05 was considered statistically significant. Results Patterns of eGFR variability This study initially included 282,412 participants. Among these, 96,651 participants were excluded because they had fewer than two eGFR measurements, leaving 185,761 participants who had at least three eGFR measurements. Among these remaining participants, an additional 183,996 were excluded because either their first or second eGFR measurement was ≥60 or <45 mL/min/1.73 m 2 . Consequently, the final study group consisted of 1,765 participants, all of whom had both their first and second eGFR measurements between 45 and 59 mL/min/1.73 m 2 . Fig. 1 illustrates the participant-selection process. Based on K-means clustering, we identified three patterns of eGFR variability over time. The median follow-up time was 4.1 (interquartile range (IQR), 2.9–6.4) years, with eGFR measurements primarily conducted on an annual or semiannual basis, aligning with participants’ health screening frequency. Additionally, the median number of eGFR measurements was 5 (IQR, 3–7) times per participant. Participants were categorized into the following three clusters: clusters 1 (4.6%), 2 (32.9%), and 3 (62.5%). Fig. 2 shows the distribution of eGFR variability parameters according to these clusters. For clusters 1, 2, and 3, the medians (IQR) of the mean residual were 10.9 (9.7–13.1), 6.0 (4.9–7.2), and 3.1 (2.4–3.9), respectively. Likewise, the medians (IQR) of the maximum residual were 22.6 (20.2–26.5), 11.4 (9.8–13.8), and 5.6 (4.4–7.0), while those of the range were 25.1 (20.9–29.8), 12.9 (9.5–16.4), and 6.6 (4.5–9.2) for clusters 1, 2, and 3, respectively. Notably, cluster 1 exhibited the highest residuals (both mean and maximum values) and widest range and was defined as the high variability cluster. Participant characteristics based on the patterns of eGFR variability Among the 1,765 participants with early-stage CKD, 1,073 (60.8%) were aged >60 years, 222 (12.6%) were female, and 202 (11.5%) had proteinuria (Table 1). Participants in the high variability cluster (cluster 1) were young (71.6% were aged <60 years), included many male participants (92.6%), and had a higher prevalence of proteinuria (38.3%), high blood pressure (35.8%), and diabetes mellitus (29.1%). Changes in eGFR values over time by clusters of eGFR variability Fig. 3 shows the change in eGFR values over time according to the cluster. The eGFR plots exhibited the widest dispersion in cluster 1 ( high-variability cluster). We observed an eGFR decline in cluster 1, characterized by a steeper slope compared to the other clusters. In cluster 1, renal function at 10 years was estimated to decline to an eGFR of nearly 40 mL/min/1.73 m2, which was lower than that observed in the other clusters. Risk factors for the high variability cluster According to the exploratory multivariate logistic regression model, younger age, proteinuria, antihypertensive drug use, and hyperglycemia were associated with being in the high-variability cluster (Table 2). However, no significant differences were found in sex, BMI, or smoking status. Sensitivity analysis Even when the number of clusters was set to four in K-means clustering, we found a similar high variability cluster (Additional file 3), which showed a wide distribution and significant eGFR decline (Additional file 4). Discussion A high-risk pattern of eGFR changes over time with high variability and greater decline was identified among patients with early-stage CKD in the general population using annual renal function screening results. Younger age, proteinuria, antihypertensive drug use, and hyperglycemia were associated with high variability. These results suggest that some patterns of eGFR variability exist among patients with early-stage CKD, which indicates an opportunity to design a segmented CKD management plan according to health screening results [ 3 ]. Furthermore, the risk factors shown in this study may lead to a decline in renal function through several possible mechanisms. The fact that younger age is associated with a higher risk of premature CKD exacerbation may be related to the eGFR estimation formula, which considers age when computing eGFR and calculates a lower value as age increases, reflecting age-related decline in renal function. In short, for the same eGFR value, the lower the age, the poorer the intrinsic kidney function. This is believed to be a risk factor for renal function decline over time, and the controversial clinical significance of renal function decline in older individuals [ 17 ] may also support a higher risk of renal exacerbation in younger patients. Additionally, a higher frequency of glomerulonephritis and other renal diseases may contribute to an increased risk of decline in renal function in young people. Although it was impossible to investigate the types of antihypertensive medications used in this study, the higher risk of renal function variability among individuals with antihypertensive medication adherence could be due to the effects of angiotensin-converting enzyme inhibitors and angiotensin receptor blockers. Proteinuria is a risk factor for the progression of renal failure [ 18 ] and a marker of kidney disease [ 19 ], which may be related to unstable kidney function. Hyperglycemia also increases renal fibrosis due to osmotic changes, even after short-term exposure [ 20 ]. It can cause damage to glomerular vessels and tubulointerstitial tubules, which may lead to transient changes in renal function. Lifestyle-related diseases, such as hypertension and diabetes, as well as smoking are risk factors for the development and progression of CKD [ 21 ]. The risk factors for CKD progression in this study, such as antihypertensive medication use and hyperglycemia, were aligned with those reported in previous studies. The strength of this study is the analysis of a nationwide health-screening cohort over time, which included a large number of individuals with early-stage CKD in the general population. As most individuals with early-stage CKD do not visit physicians, determining their long-term prognosis from medical facilities and registry data among patients diagnosed with CKD is difficult. Using nationwide health screening data, we analyzed the long-term renal function prognosis of patients with early-stage CKD. We identified risk factors for a population with significantly high renal function variability and poor long-term renal prognosis, which aligned with the established risk factors. To identify risk factors in the high-variability population with poor long-term renal prognosis, this study compared the other two clusters together with the combined population. No significant differences were found in the clinical backgrounds of clusters 2 and 3; the comparison with cluster 1 in the combined population allowed us to focus on the risk in the population with high variability. In the current Japanese medical practice, patients with mild renal impairment, which is the focus of our analysis, usually do not receive medical intervention [ 1 ]. Therefore, our study emphasizes the significance of early detection and follow-up of a frequently overlooked group of patients with mild renal dysfunction and informs the public about the importance of appropriate medical intervention. Another strength of this study is that multiple parameters were set and classified. Although previous reviews have shown the presence of renal function variability [ 22 ], this study examined the patterns of its variability by a data-driven approach using the following three parameters: maximum residual, mean residual, and range of renal function. This may have provided a more specific identification of the high-risk cluster, leading to a decline in renal function. This study had some limitations. First, we analyzed a male-focused working-age cohort in Japan; therefore, generalizability to other populations should be carefully considered. As clustering analysis is a data-driven approach that fits the data in this study, future studies examining the patterns of eGFR change in other populations are warranted. Additionally, differences in renal function variability based on sex could have led to an overestimation of residuals and ranges in renal function compared to the general population. However, this is a valuable study for examining the long-term renal prognosis of early-stage CKD in young working-age men for whom renal prognosis is important. Second, this study was conducted using a health-screening database and was based on data from annual or semiannual follow-ups. However, the period between each measurement of eGFR may be relatively long for assessing variability. While the Kidney Disease: Improving Global Outcome (KDIGO) guidelines [ 5 ] require multiple kidney function measurements, confirmed over at least 90 days, for CKD diagnosis, many epidemiological studies have usually defined CKD based on a single kidney function measurement. The need for information and evidence on how to treat individuals with early-stage CKD, such as those with stage G3a CKD, is warranted; therefore, our study was conducted to contribute to this evidence. In the future, using data from shorter follow-up intervals when conducting studies that are more in line with the CKD diagnostic criteria would be desirable. Third, the results of the clustering analysis were significantly affected by the number of clusters set in our decision. However, we found a similar high-variability cluster by sensitivity analysis with a four-cluster set. Fourth, the risk factors for poor renal prognosis identified in this study are already well established. Nonetheless, the consistency of these risk factors with those associated with high variability in renal function supports the idea that high variability contributes to poor renal prognosis. In conclusion, a pattern of high variability and significant decline in eGFR was identified among patients with early-stage CKD in the general population. Younger age, proteinuria, antihypertensive drug use, and hyperglycemia were associated with high variability in eGFR. While these risk factors are well known to indicate poor prognostic outcomes, our study goes beyond their identification. We also uncovered variability in early-stage CKD outcomes by concentrating on patterns of renal function variability. The clinical significance of monitoring renal function variability should be examined in future studies. Abbreviations CKD Chronic kidney disease DBP Diastolic blood pressure IRB Institutional Review Board LA Levey AS SBP Systolic blood pressure SSE Sum of squared errors Declarations Ethical approval and consent to participate This study protocol was approved by the Institutional Review Board (IRB) of Kyoto University (IRB number: R0817). As we analyzed only anonymized data, the IRB waived the need for informed consent. Consent for publication Not applicable. Availability of Data and Materials No data are available. The data underlying this article are not shared due to the privacy policy of data providers. Competing interests We have received grant support from Kyowa Kirin Co., Ltd. No other financial disclosures have been reported. Funding Shingo Fukuma was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Number 22H03314 and the Japan Science and Technology (JST) CREST Grant Number JPMJCR22D2. These funds had no role in this research design. Authors’ contributions AK was responsible for the study conceptions, study design, data curation, data analysis, data interpretation, statistical analysis and writing of the paper. Tsuyoshi O was responsible for study conceptions and data interpretation. TO advised on the study design and data analysis. KH and TI provided advice regarding the idea, study design and data analysis. TY was a supervisor. SF managed the overall conduct of the study and wrote the manuscript. All authors contributed intellectual content during the drafting or revision of the manuscript and approved the final manuscript. Acknowledgments We would like to thank the Health Insurance Association for Architecture and Civil Engineering companies for their support in developing the database. References Yamada Y, Ikenoue T, Saito Y, Fukuma S. Undiagnosed and untreated chronic kidney disease and its impact on renal outcomes in the Japanese middle-aged general population. J Epidemiol Community Health. 2019;73:1122–27. doi: 10.1136/jech-2019-212858. Jha V, Garcia-Garcia G, Iseki K, Li Z, Naicker S, Plattner B , et al . Chronic kidney disease: global dimension and perspectives. Lancet. 2013;382:260–72. doi: 10.1016/S0140-6736(13)60687-X. Chen TK, Knicely DH, Grams ME. Chronic Kidney Disease Diagnosis and Management: A Review. JAMA. 2019;322:1294–304. doi: 10.1001/jama.2019.14745. Collaboration GBDCKD. Global, regional, and national burden of chronic kidney disease, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2020;395:709–33. doi: 10.1016/S0140-6736(20)30045-3. Inker LA, Astor BC, Fox CH, Isakova T, Lash JP, Peralta CA , et al . KDOQI US commentary on the 2012 KDIGO clinical practice guideline for the evaluation and management of CKD. Am J Kidney Dis. 2014;63:713–35. doi: 10.1053/j.ajkd.2014.01.416. Al-Aly Z, Balasubramanian S, McDonald JR, Scherrer JF, O'Hare AM. Greater variability in kidney function is associated with an increased risk of death. Kidney Int. 2012;82:1208–14. doi: 10.1038/ki.2012.276. Chen SC, Lin MY, Huang TH, Hung CC, Chiu YW, Chang JM , et al . Variability in estimated glomerular filtration rate by area under the curve predicts renal outcomes in chronic kidney disease. ScientificWorldJournal. 2014;2014:802037. doi: 10.1155/2014/802037. Tsai C-W, Huang H-C, Chiang H-Y, Chung C-W, Chiu H-T, Liang C-C, 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;34:2066–78. Okada S, Nishioka Y, Kanaoka K, Koizumi M, Kamitani F, Nakajima H, et al. Annual variation of estimated glomerular filtration rate in health check-ups associated with end-stage kidney disease. Sci Rep. 2024;14:21065. Kiberd B. Screening for chronic kidney disease. BMJ. 2010;341:c5734. doi: 10.1136/bmj.c5734. Wouters OJ, O'Donoghue DJ, Ritchie J, Kanavos PG, Narva AS. Early chronic kidney disease: diagnosis, management and models of care. Nat Rev Nephrol. 2015;11:491–502. doi: 10.1038/nrneph.2015.85. Ikegami N, Yoo BK, Hashimoto H, Matsumoto M, Ogata H, Babazono A , et al . Japanese universal health coverage: evolution, achievements, and challenges. Lancet. 2011;378:1106–15. doi: 10.1016/S0140-6736(11)60828-3. Horio M, Imai E, Yasuda Y, Watanabe T, Matsuo S. Modification of the CKD epidemiology collaboration (CKD-EPI) equation for Japanese: accuracy and use for population estimates. Am J Kidney Dis. 2010;56:32–8. doi: 10.1053/j.ajkd.2010.02.344. Grams ME, Sang Y, Ballew SH, Matsushita K, Astor BC, Carrero JJ , et al . Evaluating Glomerular Filtration Rate Slope as a Surrogate End Point for ESKD in Clinical Trials: An Individual Participant Meta-Analysis of Observational Data. J Am Soc Nephrol. 2019;30:1746–55. doi: 10.1681/ASN.2019010008. Inker LA, Heerspink HJL, Tighiouart H, Levey AS, Coresh J, Gansevoort RT , et al . GFR Slope as a Surrogate End Point for Kidney Disease Progression in Clinical Trials: A Meta-Analysis of Treatment Effects of Randomized Controlled Trials. J Am Soc Nephrol. 2019;30:1735–45. doi: 10.1681/ASN.2019010007. Li L, Astor BC, Lewis J, Hu B, Appel LJ, Lipkowitz MS , et al . Longitudinal progression trajectory of GFR among patients with CKD. Am J Kidney Dis. 2012;59:504–12. doi: 10.1053/j.ajkd.2011.12.009. O'Hare AM, Choi AI, Bertenthal D, Bacchetti P, Garg AX, Kaufman JS , et al . Age affects outcomes in chronic kidney disease. J Am Soc Nephrol. 2007;18:2758–65. doi: 10.1681/ASN.2007040422. Inker LA, Levey AS, Pandya K, Stoycheff N, Okparavero A, Greene T , et al . Early change in proteinuria as a surrogate end point for kidney disease progression: an individual patient meta-analysis. Am J Kidney Dis. 2014;64:74–85. doi: 10.1053/j.ajkd.2014.02.020. Snyder S, John JS. Workup for proteinuria. Prim Care. 2014;41:719–35. doi: 10.1016/j.pop.2014.08.010. Polhill TS, Saad S, Poronnik P, Fulcher GR, Pollock CA. Short-term peaks in glucose promote renal fibrogenesis independently of total glucose exposure. Am J Physiol Renal Physiol. 2004;287:F268–73. doi: 10.1152/ajprenal.00084.2004. Kalantar-Zadeh K, Jafar TH, Nitsch D, Neuen BL, Perkovic V. Chronic kidney disease. Lancet. 2021;398:786–802. doi: 10.1016/S0140-6736(21)00519-5. Thöni S, Keller F, Denicolò S, Buchwinkler L, Mayer G. Biological variation and reference change value of the estimated glomerular filtration rate in humans: A systematic review and meta-analysis. Front Med (Lausanne). 2022;9:1009358. doi: 10.3389/fmed.2022.1009358. Tables Table 1. Characteristics by clusters Cluster 1 Cluster 2 Cluster 3 Total p-value n=81 n=581 n=1,103 n=1,765 Age, years [n(%)] <45 13 (16.0) 33 (5.7) 25 (2.3) 71 (4.0) <0.001 45–59 45 (55.6) 241 (41.5) 335 (30.4) 621 (35.2) ≥60 23 (28.4) 307 (52.8) 743 (67.4) 1,073 (60.8) Female [n(%)] 6 (7.4) 67 (11.5) 149 (13.5) 222 (12.6) 0.18 Basal eGFR [mL/min/1.73 m 2 (IQR)] 55.4 (51.2–57.9) 55.1 (50.3–58.0) 55.0 (52.2–57.5) 55.0 (51.8–57.7) 0.93 Body mass index [n(%)] <20 kg/m 2 3 (3.7) 20 (3.4) 72 (6.5) 95 (5.4) <0.001 20–24.9 kg/m 2 23 (28.4) 264 (45.4) 530 (48.1) 817 (46.3) ≥25 kg/m 2 55 (67.9) 297 (51.1) 500 (45.4) 852 (48.3) Urine protein [n(%)] 31 (38.3) 91 (15.7) 80 (7.3) 202 (11.5) <0.001 Hypertension [n(%)] 29 (35.8) 188 (32.4) 318 (28.8) 535 (30.3) 0.18 Antihypertensive drugs [n(%)] 57 (70.4) 336 (57.9) 517 (46.9) 910 (51.6) <0.001 High blood pressure and drugs [n(%)] High(-)×Drug(-) 16 (19.8) 179 (30.9) 443 (40.2) 638 (36.2) <0.001 High(-)×Drug(+) 36 (44.4) 213 (36.7) 342 (31.0) 591 (33.5) High(+)×Drug(-) 8 (9.9) 65 (11.2) 143 (13.0) 216 (12.2) High(+)×Drug(+) 21 (25.9) 123 (21.2) 175 (15.9) 319 (18.1) Current smoking [n(%)] 24 (29.6) 104 (17.9) 166 (15.0) 294 (16.7) 0.002 High blood glucose [n(%)] 23 (29.1) 92 (16.0) 121 (11.1) 236 (13.6) <0.001 Anti-diabetic drugs [n(%)] 18 (22.2) 94 (16.2) 109 (9.9) 221 (12.5) <0.001 High blood glucose and drugs [n(%)] High(-)×Drug(-) 52 (65.8) 455 (79.3) 937 (86.3) 1,444 (83.0) <0.001 High(-)×Drug(+) 4 (5.1) 27 (4.7) 28 (2.6) 59 (3.4) High(+)×Drug(-) 9 (11.4) 27 (4.7) 40 (3.7) 76 (4.4) High(+)×Drug(+) 14 (17.7) 65 (11.3) 81 (7.5) 160 (9.2) eGFR, estimated glomerular filtration rate; IQR, interquartile range. High blood pressure: Systolic blood pressure of ≥140 mmHg or diastolic blood pressure of ≥90 mmHg. High blood glucose: glycated hemoglobin level of ≥6.5 %. Table 2. Risk factors for the clusters with high variability Odds ratio p-value 95% confidence Interval Age, year <45 9.86 0.000 4.39–22.1 45–59 2.95 0.000 1.72–5.04 ≥60 reference Female 1.10 0.852 0.41–2.92 Body mass index <20 kg/m 2 1.27 0.731 0.33–4.94 20–24.9 kg/m 2 reference ≥25 kg/m 2 1.26 0.408 0.73–2.17 Urine protein 1+ or more 3.18 0.000 1.88–5.39 Smoke 1.65 0.069 0.96–2.82 High blood pressure and drugs High(-)×Drug(-) reference High(-)×Drug(+) 2.17 0.022 1.12–4.20 High(+)×Drug(-) 1.00 0.996 0.37–2.69 High(+)×Drug(+) 1.94 0.082 0.91–4.10 High blood glucose and drugs High(-)×Drug(-) reference High(-)×Drug(+) 1.73 0.336 0.57–5.25 High(+)×Drug(-) 3.15 0.005 1.41–7.07 High(+)×Drug(+) 1.87 0.068 0.95–3.66 Blood Pressure high：Systolic blood pressure ≥140 mmHg or Diastolic blood pressure ≥90 mmHg, Blood Glucose high：HbA1c ≥6.5% Additional Declarations Competing interest reported. We have received grant support from Kyowa Kirin Co., Ltd. No other financial disclosures have been reported. Supplementary Files Supplementarymaterial.docx Cite Share Download PDF Status: Published Journal Publication published 05 Aug, 2025 Read the published version in BMC Nephrology → Version 1 posted Editorial decision: Revision requested 31 Jan, 2025 Editor assigned by journal 30 Jan, 2025 Submission checks completed at journal 30 Jan, 2025 First submitted to journal 28 Jan, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5921794","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":409544717,"identity":"3e97e5c0-8d6b-4b9f-9fe7-d651ff4007b1","order_by":0,"name":"Arisa Kobayashi","email":"","orcid":"","institution":"Kyoto University Graduate School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Arisa","middleName":"","lastName":"Kobayashi","suffix":""},{"id":409544718,"identity":"115296af-4c17-434a-aa2e-7d71b98f0a82","order_by":1,"name":"Tsuyoshi Ohnishi","email":"","orcid":"","institution":"Kyoto University Graduate School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Tsuyoshi","middleName":"","lastName":"Ohnishi","suffix":""},{"id":409544720,"identity":"7241d353-c7e0-431f-89ca-4c0611de3a92","order_by":2,"name":"Tadahisa Okuda","email":"","orcid":"","institution":"Kyoto University Graduate School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Tadahisa","middleName":"","lastName":"Okuda","suffix":""},{"id":409544722,"identity":"6515ed38-8937-43b6-951c-45037ddd9c4d","order_by":3,"name":"Keita Hirano","email":"","orcid":"","institution":"Kyoto University Graduate School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Keita","middleName":"","lastName":"Hirano","suffix":""},{"id":409544724,"identity":"c4515b01-6645-42a7-9981-41260e989c18","order_by":4,"name":"Tatsuyoshi Ikenoue","email":"","orcid":"","institution":"Kyoto University Graduate School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Tatsuyoshi","middleName":"","lastName":"Ikenoue","suffix":""},{"id":409544725,"identity":"d4d65da7-33ab-4814-996b-a27d6d72892e","order_by":5,"name":"Takashi Yokoo","email":"","orcid":"","institution":"The Jikei University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Takashi","middleName":"","lastName":"Yokoo","suffix":""},{"id":409544726,"identity":"0b08a6a0-e782-45aa-bbc8-2b8e91b25de5","order_by":6,"name":"Shingo Fukuma","email":"data:image/png;base64,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","orcid":"","institution":"Kyoto University Graduate School of Medicine","correspondingAuthor":true,"prefix":"","firstName":"Shingo","middleName":"","lastName":"Fukuma","suffix":""}],"badges":[],"createdAt":"2025-01-29 04:53:06","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5921794/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5921794/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12882-025-04344-4","type":"published","date":"2025-08-05T15:57:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":75406047,"identity":"bfcdc202-f90f-47fa-8a1c-bb8f4640db51","added_by":"auto","created_at":"2025-02-04 08:49:43","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":155310,"visible":true,"origin":"","legend":"\u003cp\u003eSelection process of study participants.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5921794/v1/b644ea02660bc6dc90127ff1.png"},{"id":75406045,"identity":"af6cc9f8-42b2-4cd1-aba0-94aa0773499b","added_by":"auto","created_at":"2025-02-04 08:49:42","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":64473,"visible":true,"origin":"","legend":"\u003cp\u003eParameters of estimated glomerular filtration rate (eGFR) variability by clusters. This figure shows the mean residual (mean value of residual between observed eGFR values and linear predictions during the follow-up period), maximum residual (maximum value of residual between observed eGFR values and linear predictions during the follow-up period), and range of eGFR (difference between maximum and minimum values during the follow-up period) for the three clusters. Cluster 1 showed high variability.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5921794/v1/04e44ea932837b8e243f3c9e.png"},{"id":75408039,"identity":"3bcb42a5-4b3e-4ccd-9046-57fa7df6a377","added_by":"auto","created_at":"2025-02-04 08:57:43","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":192701,"visible":true,"origin":"","legend":"\u003cp\u003eEstimated glomerular filtration rate (eGFR) changes over time in the three clusters. This figure shows plots of eGFR and non-linear predictions with 95% confidence intervals for the three clusters. Cluster 1 indicated high variability with a large decline in eGFR.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5921794/v1/9394a03ecf0006a27043628a.png"},{"id":88814240,"identity":"6c07e352-f848-46df-9848-809077a778e9","added_by":"auto","created_at":"2025-08-11 16:08:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1115467,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5921794/v1/defe5073-22d1-4638-b693-dd43bd427083.pdf"},{"id":75406043,"identity":"bcd543cc-a999-46c6-b3bb-d5eeabf75c76","added_by":"auto","created_at":"2025-02-04 08:49:42","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":328685,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-5921794/v1/5a442ab54d9a0996a6315aa2.docx"}],"financialInterests":"Competing interest reported. We have received grant support from Kyowa Kirin Co., Ltd. No other financial disclosures have been reported.","formattedTitle":"Investigating the Patterns of Renal Function Variability in Early-Stage Chronic Kidney Disease by Cluster Analysis","fulltext":[{"header":"Introduction","content":"\u003cp\u003eChronic kidney disease (CKD) is a major public health concern, affecting a significant proportion of the general population. Despite its high prevalence, CKD often remains undiagnosed due to the lack of specific symptoms during its early stages, and may lead to end-stage renal disease [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Progression of CKD may result in serious complications, including cardiovascular disease, and increased mortality risk [\u003cspan additionalcitationids=\"CR3\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Early detection and appropriate management of CKD are essential to prevent these complications. Staging CKD based on estimated glomerular filtration rate (eGFR) and urinary protein levels has been one of the important assessment methods for the management of CKD risk [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. On the other hand, renal function ofte exhibit variability over time, and the implications of this variability remain insufficiently explored.\u003c/p\u003e \u003cp\u003eAlthough changes in renal function and their associations with end-stage renal disease, dialysis initiation, and mortality have been reported [\u003cspan additionalcitationids=\"CR7 CR8\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e], there is a lack of studies that comprehensively evaluate renal function variability and its prognostic implications using multiple parameters over sufficiently long observation periods. In addition, variability in renal function among patients with early-stage CKD, who are often identified through health screening, has not been thoroughly investigated. This gap highlights the need for stronger evidence to guide appropriate follow-up strategies after screening and to establish clearer recommendations for long-term monitoring.\u003c/p\u003e \u003cp\u003eIn Japan, a nationwide health screening system facilitates the early detection of CKD in asymptomatic individuals, offering an opportunity to examine the early stage of CKD in the general population. However, the clinical presentation of early-stage CKD is highly heterogeneous. Not all individuals with early-stage CKD progress to more advanced stages or require the same level of medical intervention [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Effective management strategies for early-stage CKD must therefore consider not only baseline measures of renal function but also mid- to long-term changes, including fluctuations and trends in eGFR [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eMost studies focus on the eGFR slope\u0026mdash;a measure of the linear rate of decline in renal function over time\u0026mdash;as a prognostic indicator. However, by taking into account the variability of eGFR as well as the slope, it may be possible to reflect unstable periods that have clinical significance. Moreover, identifying patterns of eGFR variability using advanced techniques such as clustering analysis could provide valuable insights into disease trajectories and help stratify patients by risk. Nevertheless, research using this approach is not sufficient, and further research is needed to establish whether eGFR variability can offer unique prognostic value beyond traditional measures like the eGFR slope.\u003c/p\u003e \u003cp\u003eTo address these gaps, we leveraged a nationwide health screening cohort in Japan to investigate long-term changes in renal function among individuals with early-stage CKD. This study aimed to identify distinct patterns of renal function variability using clustering analysis, and explore the clinical and demographic factors associated with high-risk patterns of variability. By focusing on eGFR variability, this study seeks to advance our understanding of early-stage CKD progression and inform more personalized approaches to its management.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eData source and setting\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData from a nationwide health-screening cohort between April 2011 and March 2022, provided by one of Japan\u0026rsquo;s largest employment-based health insurers (a national sample of employees of civil engineering and construction companies), were analyzed. We obtained health screening data from annual health assessments conducted by the insurer among insured adults [12]. The dataset included information on demographic characteristics (age, sex, and body mass index [BMI]), clinical factors (systolic blood pressure [SBP], diastolic blood pressure [DBP], and glycated hemoglobin (HbA1c) level), medication use (antihypertensive and antidiabetic drugs), and smoking status. Additionally, eGFR values were calculated using the Chronic Kidney Disease Epidemiology Collaboration equation for Japanese individuals [13]. Urinary protein levels were measured using the dip-stick test, and proteinuria was defined as a score of 1+ or higher. The basic clinical information was evaluated on the first health checkup date.\u003c/p\u003e\n\u003cp\u003eThis study received approval from the Institutional Review Board (IRB) of Kyoto University (IRB number: R0817). As we analyzed only anonymized data, the IRB waived the need for informed consent. This study complied with the principles of the Declaration of Helsinki.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eParticipant selection\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe excluded individuals who had less than two eGFR measurements from all participants. Subsequently, the sample was further narrowed by excluding individuals with eGFR values, either first or second eGFR of \u0026ge;60 or \u0026lt;45 mL/min/1.73 m\u003csup\u003e2\u003c/sup\u003e. This selection criterion allowed us to focus on the early-stage CKD (CKD stage G3a) population. This is because many patients with CKD detected during health checkups are at an early stage and clinical decisions regarding the need for renal function monitoring in this population may be difficult to make.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eK-means clustering by eGFR variability\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify patterns of eGFR variability, we employed K-means clustering, focusing on the following three parameters: mean residual (average of the difference between linear predictions and observed values), maximum residual (highest difference between linear predictions and observed values), and range of values (difference between maximum and minimum observed values) (Additional file 1). Linear predictions were estimated using a linear regression model for repeated eGFR measurements.\u003c/p\u003e\n\u003cp\u003eK-means clustering is a widely used method for grouping data and is effective in revealing significant insights from large datasets through exploratory analysis. This statistical method groups data based on similarities in their characteristics by focusing on the differences in the characteristics of the data.\u003c/p\u003e\n\u003cp\u003eWhile the yearly eGFR decline slope is usually used as a surrogate endpoint for evaluating renal prognosis [14, 15], the slope assumes a linear decline in eGFR annually. Although this slope is useful for approximating linear trends in population-wide changes in renal function, it may not accurately reflect individual variations. In clinical practice, deviations from linear eGFR changes due to various factors that influence renal function variability are frequently observed [16]. Therefore, we used the residual and range for classification instead of the slope to assess the renal function variability more accurately.\u003c/p\u003e\n\u003cp\u003eWe initially calculated the residuals between the linear predictions and actual eGFR values for each individual. Subsequently, we computed the mean and maximum of these residuals and determined the range of eGFR values during the observation period. These parameters were standardized by dividing each of them by their standard deviations to ensure consistent scaling. The participants were then classified into three clusters using K-means clustering. The number of clusters was determined to be three based on clinical interpretability and evaluated using the general thresholds for the sum of squared errors (SSE) and the elbow method (Additional file 2). In addition, we conducted a sensitivity analysis with varying numbers of clusters to validate our findings. The cluster definitions were aligned with the observed patterns of eGFR variability, and the second and subsequent eGFR values were used to define the eGFR variability. At least three measurements were used to capture the variability over a certain period. The baseline eGFR values of the population with eGFR \u0026gt;60 or \u0026lt;45 mL/min/1.73 m\u003csup\u003e2\u003c/sup\u003e are, to some extent, defined by the baseline eGFR values for subsequent eGFR changes. Therefore, they were excluded from this study.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eParticipant characteristics and eGFR change\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipant characteristics were described according to eGFR variability clusters. We defined hypertension as SBP \u0026ge;140 mmHg or DBP \u0026ge;90 mmHg and diabetes as an HbA1c level \u0026ge;6.5%. After stratification by eGFR variability clusters, we modeled nonlinear relationships between time (follow-up years) and eGFR using a fractional polynomial model separately. The predicted and observed eGFR values over time were plotted.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAssociation\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003ebetween clinical factors and the high variability cluster\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe examined the associations between clinical factors and being in the high-variability cluster using a logistic regression model. Clinical factors such as age, sex, BMI, urine protein, smoking status, blood pressure, blood glucose, and the use of medications for hypertension or diabetes were included as explaratory variables. In this analysis, we defined high blood pressure as SBP of 140 or higher or DBP of 90 or higher. Also we defined high blood glucose as HbA1c of 6.5% or higher.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eStatistical analysis\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePatient characteristics are presented as medians (interquartile ranges) for continuous variables and numerical values (percentages) for categorical variables. Differences in background characteristics between the clusters were assessed utilizing the chi-square test for binary variables and the t-test for continuous variables. A sensitivity analysis was conducted employing different numbers of clusters to examine the effect of cluster quantity on the results. All statistical analyses were performed using STATA MP (version 18.0; STATA Corporation, College Station, TX, US). For all analyses, a two-tailed p value \u0026lt; 0.05 was considered statistically significant.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003ePatterns of eGFR variability\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study initially included 282,412 participants. Among these, 96,651 participants were excluded because they had fewer than two eGFR measurements, leaving 185,761 participants who had at least three eGFR measurements. Among these remaining participants, an additional 183,996 were excluded because either their first or second eGFR measurement was \u0026ge;60 or \u0026lt;45 mL/min/1.73 m\u003csup\u003e2\u003c/sup\u003e. Consequently, the final study group consisted of 1,765 participants, all of whom had both their first and second eGFR measurements between 45 and 59 mL/min/1.73 m\u003csup\u003e2\u003c/sup\u003e. Fig. 1 illustrates the participant-selection process.\u003c/p\u003e\n\u003cp\u003eBased on K-means clustering, we identified three patterns of eGFR variability over time. The median follow-up time was 4.1 (interquartile range (IQR), 2.9\u0026ndash;6.4) years, with eGFR measurements primarily conducted on an annual or semiannual basis, aligning with participants\u0026rsquo; health screening frequency. Additionally, the median number of eGFR measurements was 5 (IQR, 3\u0026ndash;7) times per participant. Participants were categorized into the following three clusters: clusters 1 (4.6%), 2 (32.9%), and 3 (62.5%). Fig. 2 shows the distribution of eGFR variability parameters according to these clusters. For clusters 1, 2, and 3, the medians (IQR) of the mean residual were 10.9 (9.7\u0026ndash;13.1), 6.0 (4.9\u0026ndash;7.2), and 3.1 (2.4\u0026ndash;3.9), respectively. Likewise, the medians (IQR) of the maximum residual were 22.6 (20.2\u0026ndash;26.5), 11.4 (9.8\u0026ndash;13.8), and 5.6 (4.4\u0026ndash;7.0), while those of the range were 25.1 (20.9\u0026ndash;29.8), 12.9 (9.5\u0026ndash;16.4), and 6.6 (4.5\u0026ndash;9.2) for clusters 1, 2, and 3, respectively. Notably, cluster 1 exhibited the highest residuals (both mean and maximum values) and widest range and was defined as the high variability cluster.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eParticipant characteristics based on the patterns of eGFR variability\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAmong the 1,765 participants with early-stage CKD, 1,073 (60.8%) were aged \u0026gt;60 years, 222 (12.6%) were female, and 202 (11.5%) had proteinuria (Table 1). Participants in the high variability cluster (cluster 1) were young (71.6% were aged \u0026lt;60 years), included many male participants (92.6%), and had a higher prevalence of proteinuria (38.3%), high blood pressure (35.8%), and diabetes mellitus (29.1%).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eChanges in eGFR values over time by clusters of eGFR variability\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFig. 3 shows the change in eGFR values over time according to the cluster. The eGFR plots exhibited the widest dispersion in cluster 1 ( high-variability cluster). We observed an eGFR decline in cluster 1, characterized by a steeper slope compared to the other clusters. In cluster 1, renal function at 10 years was estimated to decline to an eGFR of nearly 40 mL/min/1.73 m2, which was lower than that observed in the other clusters.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eRisk factors for the high variability cluster\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAccording to the exploratory multivariate logistic regression model, younger age, proteinuria, antihypertensive drug use, and hyperglycemia were associated with being in the high-variability cluster (Table 2). However, no significant differences were found in sex, BMI, or smoking status.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eSensitivity analysis\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEven when the number of clusters was set to four in K-means clustering, we found a similar high variability cluster (Additional file 3), which showed a wide distribution and significant eGFR decline (Additional file 4).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eA high-risk pattern of eGFR changes over time with high variability and greater decline was identified among patients with early-stage CKD in the general population using annual renal function screening results. Younger age, proteinuria, antihypertensive drug use, and hyperglycemia were associated with high variability. These results suggest that some patterns of eGFR variability exist among patients with early-stage CKD, which indicates an opportunity to design a segmented CKD management plan according to health screening results [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Furthermore, the risk factors shown in this study may lead to a decline in renal function through several possible mechanisms.\u003c/p\u003e \u003cp\u003eThe fact that younger age is associated with a higher risk of premature CKD exacerbation may be related to the eGFR estimation formula, which considers age when computing eGFR and calculates a lower value as age increases, reflecting age-related decline in renal function. In short, for the same eGFR value, the lower the age, the poorer the intrinsic kidney function. This is believed to be a risk factor for renal function decline over time, and the controversial clinical significance of renal function decline in older individuals [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] may also support a higher risk of renal exacerbation in younger patients. Additionally, a higher frequency of glomerulonephritis and other renal diseases may contribute to an increased risk of decline in renal function in young people. Although it was impossible to investigate the types of antihypertensive medications used in this study, the higher risk of renal function variability among individuals with antihypertensive medication adherence could be due to the effects of angiotensin-converting enzyme inhibitors and angiotensin receptor blockers. Proteinuria is a risk factor for the progression of renal failure [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e] and a marker of kidney disease [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e], which may be related to unstable kidney function. Hyperglycemia also increases renal fibrosis due to osmotic changes, even after short-term exposure [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. It can cause damage to glomerular vessels and tubulointerstitial tubules, which may lead to transient changes in renal function. Lifestyle-related diseases, such as hypertension and diabetes, as well as smoking are risk factors for the development and progression of CKD [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. The risk factors for CKD progression in this study, such as antihypertensive medication use and hyperglycemia, were aligned with those reported in previous studies.\u003c/p\u003e \u003cp\u003eThe strength of this study is the analysis of a nationwide health-screening cohort over time, which included a large number of individuals with early-stage CKD in the general population. As most individuals with early-stage CKD do not visit physicians, determining their long-term prognosis from medical facilities and registry data among patients diagnosed with CKD is difficult. Using nationwide health screening data, we analyzed the long-term renal function prognosis of patients with early-stage CKD. We identified risk factors for a population with significantly high renal function variability and poor long-term renal prognosis, which aligned with the established risk factors. To identify risk factors in the high-variability population with poor long-term renal prognosis, this study compared the other two clusters together with the combined population. No significant differences were found in the clinical backgrounds of clusters 2 and 3; the comparison with cluster 1 in the combined population allowed us to focus on the risk in the population with high variability. In the current Japanese medical practice, patients with mild renal impairment, which is the focus of our analysis, usually do not receive medical intervention [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Therefore, our study emphasizes the significance of early detection and follow-up of a frequently overlooked group of patients with mild renal dysfunction and informs the public about the importance of appropriate medical intervention. Another strength of this study is that multiple parameters were set and classified. Although previous reviews have shown the presence of renal function variability [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], this study examined the patterns of its variability by a data-driven approach using the following three parameters: maximum residual, mean residual, and range of renal function. This may have provided a more specific identification of the high-risk cluster, leading to a decline in renal function.\u003c/p\u003e \u003cp\u003eThis study had some limitations. First, we analyzed a male-focused working-age cohort in Japan; therefore, generalizability to other populations should be carefully considered. As clustering analysis is a data-driven approach that fits the data in this study, future studies examining the patterns of eGFR change in other populations are warranted. Additionally, differences in renal function variability based on sex could have led to an overestimation of residuals and ranges in renal function compared to the general population. However, this is a valuable study for examining the long-term renal prognosis of early-stage CKD in young working-age men for whom renal prognosis is important. Second, this study was conducted using a health-screening database and was based on data from annual or semiannual follow-ups. However, the period between each measurement of eGFR may be relatively long for assessing variability. While the Kidney Disease: Improving Global Outcome (KDIGO) guidelines [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e] require multiple kidney function measurements, confirmed over at least 90 days, for CKD diagnosis, many epidemiological studies have usually defined CKD based on a single kidney function measurement. The need for information and evidence on how to treat individuals with early-stage CKD, such as those with stage G3a CKD, is warranted; therefore, our study was conducted to contribute to this evidence. In the future, using data from shorter follow-up intervals when conducting studies that are more in line with the CKD diagnostic criteria would be desirable. Third, the results of the clustering analysis were significantly affected by the number of clusters set in our decision. However, we found a similar high-variability cluster by sensitivity analysis with a four-cluster set. Fourth, the risk factors for poor renal prognosis identified in this study are already well established. Nonetheless, the consistency of these risk factors with those associated with high variability in renal function supports the idea that high variability contributes to poor renal prognosis.\u003c/p\u003e \u003cp\u003eIn conclusion, a pattern of high variability and significant decline in eGFR was identified among patients with early-stage CKD in the general population. Younger age, proteinuria, antihypertensive drug use, and hyperglycemia were associated with high variability in eGFR. While these risk factors are well known to indicate poor prognostic outcomes, our study goes beyond their identification. We also uncovered variability in early-stage CKD outcomes by concentrating on patterns of renal function variability. The clinical significance of monitoring renal function variability should be examined in future studies.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eCKD Chronic kidney disease\u003c/p\u003e\n\u003cp\u003eDBP Diastolic blood pressure\u003c/p\u003e\n\u003cp\u003eIRB Institutional Review Board\u003c/p\u003e\n\u003cp\u003eLA Levey AS\u003c/p\u003e\n\u003cp\u003eSBP Systolic blood pressure\u003c/p\u003e\n\u003cp\u003eSSE Sum of squared errors\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthical approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study protocol was approved by the Institutional Review Board (IRB) of Kyoto University (IRB number: R0817). As we analyzed only anonymized data, the IRB waived the need for informed consent.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of Data and Materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo data are available. The data underlying this article are not shared due to the privacy policy of data providers.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe have received grant support from Kyowa Kirin Co., Ltd. No other financial disclosures have been reported.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eShingo Fukuma was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Number 22H03314 and the Japan Science and Technology (JST) CREST Grant Number JPMJCR22D2. These funds had no role in this research design.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAK was responsible for the study conceptions, study design, data curation, data analysis, data interpretation, statistical analysis and writing of the paper. Tsuyoshi O was responsible for study conceptions and data interpretation. TO advised on the study design and data analysis. KH and TI provided advice regarding the idea, study design and data analysis. TY was a supervisor. SF managed the overall conduct of the study and wrote the manuscript.\u003c/p\u003e\n\u003cp\u003eAll authors contributed intellectual content during the drafting or revision of the manuscript and approved the final manuscript.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to thank the Health Insurance Association for Architecture and Civil Engineering companies for their support in developing the database.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eYamada Y, Ikenoue T, Saito Y, Fukuma S. Undiagnosed and untreated chronic kidney disease and its impact on renal outcomes in the Japanese middle-aged general population. J Epidemiol Community Health. 2019;73:1122\u0026ndash;27. doi: 10.1136/jech-2019-212858.\u003c/li\u003e\n \u003cli\u003eJha V, Garcia-Garcia G, Iseki K, Li Z, Naicker S, Plattner B\u003cem\u003e, et al\u003c/em\u003e. Chronic kidney disease: global dimension and perspectives. Lancet. 2013;382:260\u0026ndash;72. doi: 10.1016/S0140-6736(13)60687-X.\u003c/li\u003e\n \u003cli\u003eChen TK, Knicely DH, Grams ME. Chronic Kidney Disease Diagnosis and Management: A Review. JAMA. 2019;322:1294\u0026ndash;304. doi: 10.1001/jama.2019.14745.\u003c/li\u003e\n \u003cli\u003eCollaboration GBDCKD. Global, regional, and national burden of chronic kidney disease, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2020;395:709\u0026ndash;33. doi: 10.1016/S0140-6736(20)30045-3.\u003c/li\u003e\n \u003cli\u003eInker LA, Astor BC, Fox CH, Isakova T, Lash JP, Peralta CA\u003cem\u003e, et al\u003c/em\u003e. KDOQI US commentary on the 2012 KDIGO clinical practice guideline for the evaluation and management of CKD. Am J Kidney Dis. 2014;63:713\u0026ndash;35. doi: 10.1053/j.ajkd.2014.01.416.\u003c/li\u003e\n \u003cli\u003eAl-Aly Z, Balasubramanian S, McDonald JR, Scherrer JF, O\u0026apos;Hare AM. Greater variability in kidney function is associated with an increased risk of death. Kidney Int. 2012;82:1208\u0026ndash;14. doi: 10.1038/ki.2012.276.\u003c/li\u003e\n \u003cli\u003eChen SC, Lin MY, Huang TH, Hung CC, Chiu YW, Chang JM\u003cem\u003e, et al\u003c/em\u003e. Variability in estimated glomerular filtration rate by area under the curve predicts renal outcomes in chronic kidney disease. ScientificWorldJournal. 2014;2014:802037. doi: 10.1155/2014/802037.\u003c/li\u003e\n \u003cli\u003eTsai C-W, Huang H-C, Chiang H-Y, Chung C-W, Chiu H-T, Liang C-C, 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;34:2066\u0026ndash;78.\u003c/li\u003e\n \u003cli\u003eOkada S, Nishioka Y, Kanaoka K, Koizumi M, Kamitani F, Nakajima H, et al. Annual variation of estimated glomerular filtration rate in health check-ups associated with end-stage kidney disease. Sci Rep. 2024;14:21065.\u003c/li\u003e\n \u003cli\u003eKiberd B. Screening for chronic kidney disease. BMJ. 2010;341:c5734. doi: 10.1136/bmj.c5734.\u003c/li\u003e\n \u003cli\u003eWouters OJ, O\u0026apos;Donoghue DJ, Ritchie J, Kanavos PG, Narva AS. Early chronic kidney disease: diagnosis, management and models of care. Nat Rev Nephrol. 2015;11:491\u0026ndash;502. doi: 10.1038/nrneph.2015.85.\u003c/li\u003e\n \u003cli\u003eIkegami N, Yoo BK, Hashimoto H, Matsumoto M, Ogata H, Babazono A\u003cem\u003e, et al\u003c/em\u003e. Japanese universal health coverage: evolution, achievements, and challenges. Lancet. 2011;378:1106\u0026ndash;15. doi: 10.1016/S0140-6736(11)60828-3.\u003c/li\u003e\n \u003cli\u003eHorio M, Imai E, Yasuda Y, Watanabe T, Matsuo S. Modification of the CKD epidemiology collaboration (CKD-EPI) equation for Japanese: accuracy and use for population estimates. Am J Kidney Dis. 2010;56:32\u0026ndash;8. doi: 10.1053/j.ajkd.2010.02.344.\u003c/li\u003e\n \u003cli\u003eGrams ME, Sang Y, Ballew SH, Matsushita K, Astor BC, Carrero JJ\u003cem\u003e, et al\u003c/em\u003e. Evaluating Glomerular Filtration Rate Slope as a Surrogate End Point for ESKD in Clinical Trials: An Individual Participant Meta-Analysis of Observational Data. J Am Soc Nephrol. 2019;30:1746\u0026ndash;55. doi: 10.1681/ASN.2019010008.\u003c/li\u003e\n \u003cli\u003eInker LA, Heerspink HJL, Tighiouart H, Levey AS, Coresh J, Gansevoort RT\u003cem\u003e, et al\u003c/em\u003e. GFR Slope as a Surrogate End Point for Kidney Disease Progression in Clinical Trials: A Meta-Analysis of Treatment Effects of Randomized Controlled Trials. J Am Soc Nephrol. 2019;30:1735\u0026ndash;45. doi: 10.1681/ASN.2019010007.\u003c/li\u003e\n \u003cli\u003eLi L, Astor BC, Lewis J, Hu B, Appel LJ, Lipkowitz MS\u003cem\u003e, et al\u003c/em\u003e. Longitudinal progression trajectory of GFR among patients with CKD. Am J Kidney Dis. 2012;59:504\u0026ndash;12. doi: 10.1053/j.ajkd.2011.12.009.\u003c/li\u003e\n \u003cli\u003eO\u0026apos;Hare AM, Choi AI, Bertenthal D, Bacchetti P, Garg AX, Kaufman JS\u003cem\u003e, et al\u003c/em\u003e. Age affects outcomes in chronic kidney disease. J Am Soc Nephrol. 2007;18:2758\u0026ndash;65. doi: 10.1681/ASN.2007040422.\u003c/li\u003e\n \u003cli\u003eInker LA, Levey AS, Pandya K, Stoycheff N, Okparavero A, Greene T\u003cem\u003e, et al\u003c/em\u003e. Early change in proteinuria as a surrogate end point for kidney disease progression: an individual patient meta-analysis. Am J Kidney Dis. 2014;64:74\u0026ndash;85. doi: 10.1053/j.ajkd.2014.02.020.\u003c/li\u003e\n \u003cli\u003eSnyder S, John JS. Workup for proteinuria. Prim Care. 2014;41:719\u0026ndash;35. doi: 10.1016/j.pop.2014.08.010.\u003c/li\u003e\n \u003cli\u003ePolhill TS, Saad S, Poronnik P, Fulcher GR, Pollock CA. Short-term peaks in glucose promote renal fibrogenesis independently of total glucose exposure. Am J Physiol Renal Physiol. 2004;287:F268\u0026ndash;73. doi: 10.1152/ajprenal.00084.2004.\u003c/li\u003e\n \u003cli\u003eKalantar-Zadeh K, Jafar TH, Nitsch D, Neuen BL, Perkovic V. Chronic kidney disease. Lancet. 2021;398:786\u0026ndash;802. doi: 10.1016/S0140-6736(21)00519-5.\u003c/li\u003e\n \u003cli\u003eTh\u0026ouml;ni S, Keller F, Denicol\u0026ograve; S, Buchwinkler L, Mayer G. Biological variation and reference change value of the estimated glomerular filtration rate in humans: A systematic review and meta-analysis. Front Med (Lausanne). 2022;9:1009358. doi: 10.3389/fmed.2022.1009358.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTable 1. Characteristics by clusters\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"661\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003eCluster 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eCluster 2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003eCluster 3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" style=\"width: 95px;\"\u003e\n \u003cp\u003ep-value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003en=81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003en=581\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003en=1,103\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003en=1,765\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eAge, years [n(%)]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003e\u0026lt;45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e13 (16.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e33 (5.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e25 (2.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e71 (4.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003e45\u0026ndash;59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e45 (55.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e241 (41.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e335 (30.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e621 (35.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003e\u0026ge;60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e23 (28.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e307 (52.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e743 (67.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e1,073 (60.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eFemale [n(%)]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e6 (7.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e67 (11.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e149 (13.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e222 (12.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.18\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eBasal eGFR\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e[mL/min/1.73 m\u003csup\u003e2\u003c/sup\u003e (IQR)]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e55.4\u003c/p\u003e\n \u003cp\u003e(51.2\u0026ndash;57.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e55.1\u003c/p\u003e\n \u003cp\u003e(50.3\u0026ndash;58.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e55.0\u003c/p\u003e\n \u003cp\u003e(52.2\u0026ndash;57.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e55.0\u003c/p\u003e\n \u003cp\u003e(51.8\u0026ndash;57.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.93\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eBody mass index [n(%)]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u0026lt;20 kg/m\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e3 (3.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e20 (3.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e72 (6.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e95 (5.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003e20\u0026ndash;24.9 kg/m\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e23 (28.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e264 (45.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e530 (48.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e817 (46.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003e\u0026ge;25 kg/m\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e55 (67.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e297 (51.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e500 (45.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e852 (48.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eUrine protein [n(%)]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e31 (38.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e91 (15.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e80 (7.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e202 (11.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHypertension [n(%)]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e29 (35.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e188 (32.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e318 (28.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e535 (30.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.18\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eAntihypertensive drugs [n(%)]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e57 (70.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e336 (57.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e517 (46.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e910 (51.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHigh blood pressure and drugs [n(%)]\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHigh(-)\u0026times;Drug(-)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e16 (19.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e179 (30.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e443 (40.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e638 (36.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHigh(-)\u0026times;Drug(+)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e36 (44.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e213 (36.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e342 (31.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e591 (33.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHigh(+)\u0026times;Drug(-)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e8 (9.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e65 (11.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e143 (13.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e216 (12.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHigh(+)\u0026times;Drug(+)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e21 (25.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e123 (21.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e175 (15.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e319 (18.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eCurrent smoking [n(%)]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e24 (29.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e104 (17.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e166 (15.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e294 (16.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHigh blood glucose [n(%)]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e23 (29.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e92 (16.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e121 (11.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e236 (13.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eAnti-diabetic drugs [n(%)]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e18 (22.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e94 (16.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e109 (9.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e221 (12.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHigh blood glucose and drugs [n(%)]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHigh(-)\u0026times;Drug(-)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e52 (65.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e455 (79.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e937 (86.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e1,444 (83.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHigh(-)\u0026times;Drug(+)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e4 (5.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e27 (4.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e28 (2.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e59 (3.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHigh(+)\u0026times;Drug(-)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e9 (11.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e27 (4.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e40 (3.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e76 (4.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eHigh(+)\u0026times;Drug(+)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e14 (17.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e65 (11.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e81 (7.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e160 (9.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eeGFR, estimated glomerular filtration rate; IQR, interquartile range. High blood pressure: Systolic blood pressure of \u0026ge;140 mmHg or diastolic blood pressure of \u0026ge;90 mmHg. High blood glucose: glycated hemoglobin level of \u0026ge;6.5 %.\u003c/p\u003e\n\u003cp\u003eTable 2. Risk factors for the clusters with high variability\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" align=\"\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eOdds ratio\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ep-value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e95%\u003c/p\u003e\n \u003cp\u003econfidence Interval\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;Age, year\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026lt;45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e9.86\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e4.39\u0026ndash;22.1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;45\u0026ndash;59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.72\u0026ndash;5.04\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026ge;60\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\"\u003e\n \u003cp\u003ereference\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;Female\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.852\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.41\u0026ndash;2.92\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003eBody mass index\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026lt;20 kg/m\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.731\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.33\u0026ndash;4.94\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; 20\u0026ndash;24.9 kg/m\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\"\u003e\n \u003cp\u003ereference\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026ge;25 kg/m\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.408\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.73\u0026ndash;2.17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;Urine protein\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; 1+ or more\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.88\u0026ndash;5.39\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;Smoke\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.069\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.96\u0026ndash;2.82\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;High blood pressure and drugs\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; High(-)\u0026times;Drug(-)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\"\u003e\n \u003cp\u003ereference\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; High(-)\u0026times;Drug(+)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.12\u0026ndash;4.20\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; High(+)\u0026times;Drug(-)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.996\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.37\u0026ndash;2.69\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; High(+)\u0026times;Drug(+)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.94\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.082\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.91\u0026ndash;4.10\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;High blood glucose and drugs\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; High(-)\u0026times;Drug(-)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\"\u003e\n \u003cp\u003ereference\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; High(-)\u0026times;Drug(+)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.73\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.336\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.57\u0026ndash;5.25\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; High(+)\u0026times;Drug(-)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3.15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.41\u0026ndash;7.07\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; High(+)\u0026times;Drug(+)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.87\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.068\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.95\u0026ndash;3.66\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eBlood Pressure high：Systolic blood pressure \u0026ge;140 mmHg or Diastolic blood pressure \u0026ge;90 mmHg, Blood Glucose high：HbA1c \u0026ge;6.5%\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-nephrology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bnep","sideBox":"Learn more about [BMC Nephrology](http://bmcnephrol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bnep/default.aspx","title":"BMC Nephrology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Chronic renal disease, Early-Stage chronic kidney disease, Renal prognosis, Health screening, Cluster analysis","lastPublishedDoi":"10.21203/rs.3.rs-5921794/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5921794/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003e Chronic kidney disease (CKD) is a significant global health concern, with increasing focus on predicting renal prognosis. While renal prognosis is often studied in advanced CKD, variability in renal function and its implications for long-term outcomes in early-stage CKD remain insufficiently examined. This study aimed to investigate renal prognosis in early-stage CKD within the general population, focusing on patterns of renal function variability and factors associated with high variability.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003e This retrospective nationwide cohort study included participants from various geographical regions across Japan, representing a diverse general population. A total of 1,765 adults with early-stage CKD (eGFR 45–59 mL/min/1.73 m\u003csup\u003e2\u003c/sup\u003e), based on two initial screening results, were analyzed. The primary outcome was the pattern of eGFR variability identified by cluster analysis using three parameters: mean residual (difference between linear prediction and observed value), maximum residual, and range. In addition, we used a logistic regression model in order to assess associations between clinical factors and the high-risk cluster.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e We identified three distinct clusters based on eGFR variability using cluster analysis. Among these clusters, one exhibited significantly high variability with a high residual (median of mean residuals of 10.9 mL/min/1.73 m\u003csup\u003e2\u003c/sup\u003e and median of maximum residuals of 22.6 mL/min/1.73 m\u003csup\u003e2\u003c/sup\u003e) and a wide range (median of range of 25.1 mL/min/1.73 m\u003csup\u003e2\u003c/sup\u003e) (referred to as the \"high variability cluster\"). This cluster, comprising 4.6% of patients with early-stage CKD, demonstrated a more pronounced decline in eGFR over time. Factors such as younger age, proteinuria, antihypertensive drug use, and hyperglycemia were associated with the high variability cluster.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions:\u003c/strong\u003e This study highlights the presence of distinct eGFR variability patterns in early-stage CKD and identifies a subgroup at high risk for rapid renal decline. Monitoring eGFR variability provides critical insights into long-term prognosis and may inform targeted interventions. Considering these findings, early detection and management of patients with early CKD may improve disease progression and reduce the risk of adverse outcomes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTrial registration: \u003c/strong\u003eThis study is an observational study using a database and does not involve a health care intervention on human participants. Therefore, trial registration is not applicable.\u003c/p\u003e","manuscriptTitle":"Investigating the Patterns of Renal Function Variability in Early-Stage Chronic Kidney Disease by Cluster Analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-02-04 08:49:37","doi":"10.21203/rs.3.rs-5921794/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-01-31T12:32:57+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-01-30T10:43:15+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-01-30T10:41:05+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Nephrology","date":"2025-01-29T04:39:48+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-nephrology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bnep","sideBox":"Learn more about [BMC Nephrology](http://bmcnephrol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bnep/default.aspx","title":"BMC Nephrology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"139987c6-b263-48f1-b3f5-a5f7b5424ff5","owner":[],"postedDate":"February 4th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-08-11T16:05:18+00:00","versionOfRecord":{"articleIdentity":"rs-5921794","link":"https://doi.org/10.1186/s12882-025-04344-4","journal":{"identity":"bmc-nephrology","isVorOnly":false,"title":"BMC Nephrology"},"publishedOn":"2025-08-05 15:57:00","publishedOnDateReadable":"August 5th, 2025"},"versionCreatedAt":"2025-02-04 08:49:37","video":"","vorDoi":"10.1186/s12882-025-04344-4","vorDoiUrl":"https://doi.org/10.1186/s12882-025-04344-4","workflowStages":[]},"version":"v1","identity":"rs-5921794","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5921794","identity":"rs-5921794","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}