Document resource
WHAT IS ALREADY KNOWN ON THIS TOPIC Low high-density lipoprotein cholesterol (HDL-C) has been associated with the incidence of diabetic nephropathy or diabetic kidney disease in patients with type 2 diabetes mellitus (T2DM). However, evidence regarding its relationship with hard kidney outcomes, such as a ≥50% decline in estimated glomerular filtration rate or kidney failure requiring replacement therapy, remains limited.WHAT THIS STUDY ADDS This prospective analysis of over 1,000 patients with T2DM demonstrated that low HDL-C levels were independently associated with kidney events and all-cause mortality, whereas other lipid parameters, including triglycerides (TG), low-density lipoprotein cholesterol, non-HDL-C, and TG/HDL-C ratio, were not. The study highlights HDL-C as a distinct predictor of adverse outcomes beyond conventional lipid measures.HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY Our findings underscore the potential role of HDL-C as a prognostic marker in risk stratification of patients with T2DM. Clinicians may need to pay closer attention to low HDL-C levels, and future interventional research should investigate whether targeting HDL-C can improve kidney outcomes in this high-risk population.Introduction The global prevalence of type 2 diabetes mellitus (T2DM) is increasing, making the prevention and management of its long-term complications critically important for both patient quality of life and healthcare economics. T2DM is a leading cause of chronic kidney disease (CKD), and declining kidney function leading to end-stage kidney disease (ESKD) is strongly linked to cardiovascular disease and mortality. 1 In Japan, approximately 40% of patients who start dialysis have diabetic nephropathy as the underlying cause. Diabetic nephropathy has remained the most common cause of dialysis initiation for over a decade.2 Therefore, preventing the progression of kidney disease in patients with T2DM is essential to curb the growing number of new dialysis patients.Dyslipidemia is an established risk factor for cerebrovascular and cardiovascular events, as well as for mortality, and it is one of the most important risk factors in patients with T2DM.3 Low-density lipoprotein cholesterol (LDL-C), in particular, is closely associated with cardiovascular events and mortality. Recent meta-analyses have also demonstrated that elevated LDL-C levels are associated with the progression of CKD and the development of ESKD.4 Several studies have reported that low high-density lipoprotein cholesterol (HDL-C) levels are associated with worsening kidney function and an increased risk of kidney events in the general population, among US veterans, and in patients with CKD, independent of LDL-C levels.5–9 Furthermore, although multiple studies have shown that low HDL levels are associated with a higher risk of developing diabetic nephropathy or diabetic kidney disease (DKD) in patients with T2DM,10–17 their association with hard kidney outcomes, such as kidney failure or a 50% decrease in the estimated glomerular filtration rate (eGFR), has not been thoroughly investigated. Thus, the clinical importance of HDL in kidney disease progression and its management targets remains unclear.Therefore, the associations between lipid profiles, particularly HDL-C levels, and kidney disease progression were evaluated by comparing HDL-C with other lipid parameters in patients with T2DM receiving standard diabetes care, using longitudinal data from the Fukushima Cohort Study to clarify the clinical importance of HDL-C for kidney outcomes.Materials and methods Study population This study used longitudinal data from the Fukushima Cohort Study, a prospective, observational study of outpatients receiving care from specialists in nephrology or diabetology at Fukushima Medical University Hospital, designed to identify factors associated with adverse clinical outcomes including kidney disease progression, cardiovascular events, and death. Participants were outpatients who had at least one cardiovascular risk factor, such as hypertension, diabetes mellitus, dyslipidemia, or CKD not requiring dialysis. Details of the cohort have been described previously. 18–21 Patient enrollment was conducted between June 2012 and July 2014, with a total of 2,724 patients registered for the study. For the present analysis, 1,033 patients with T2DM for whom baseline data on HDL-C levels were available were included. The study was registered in the University Hospital Medical Information Network Clinical Trials Registry (UMIN-CTR) under study ID UMIN000040848. The study was conducted in accordance with the Declaration of Helsinki.Data collection Baseline characteristics, including medication usage and comorbid conditions, were extracted from patients’ medical records and laboratory data at the time of enrollment. The body mass index (BMI) was calculated as weight/height 2. Cardiovascular disease was defined as a documented history of fatal or non-fatal myocardial infarction, angina pectoris, sudden death, congestive or acute heart failure, arrhythmia, cerebrovascular disorder, chronic arteriosclerosis obliterans, or aortic disease. Blood pressure was measured in the seated position by trained staff using either a manual sphygmomanometer or an automated device. eGFR was determined using the following equation: eGFR (mL/min/1.73 m²) = 194 × age–0.287 × serum creatinine–1.094 × 0.739 (if female),22 with serum creatinine levels measured enzymatically. Proteinuria was defined as a dipstick urine test result of ≥1+. Laboratory values, including HDL-C, triglycerides (TG), LDL-C, and uric acid, were measured using standardized automated methods at our institution’s clinical laboratory. Diabetes mellitus was diagnosed based on at least one of the following criteria: fasting plasma glucose ≥126 mg/dL, 2-hour plasma glucose ≥200 mg/dL during an oral glucose tolerance test, hemoglobin A1c ≥6.5%, or current treatment with insulin, a glucagon-like peptide-1 receptor agonist, or oral antihyperglycemic agents.Exposure The primary exposure variable in this study was baseline HDL-C. In addition, other lipid parameters, including TG, LDL-C, non-HDL-C, and the TG/HDL-C ratio, were examined. Non-HDL-C was calculated using the formula: non-HDL-C (mg/dL) = total cholesterol (mg/dL) − HDL-C (mg/dL). 23 All analyses used baseline lipid measurements. Each lipid parameter was divided into quartiles as follows:HDL-C, Q1<42 mg/dL, Q2=42–48 mg/dL, Q3=49–58 mg/dL, Q4≥59 mg/dL;TG, Q1<80 mg/dL, Q2=80–112 mg/dL, Q3=113–163 mg/dL, Q4≥164 mg/dL;LDL-C, Q1<81 mg/dL, Q2=81–98 mg/dL, Q3=99–119 mg/dL, Q4≥120 mg/dL;Non-HDL-C, Q1<104 mg/dL, Q2=104–122 mg/dL, Q3=123–147 mg/dL, Q4≥148 mg/dL;TG/HDL-C ratio, Q1<1.47, Q2=1.47–2.31, Q3=2.32–3.62, Q4≥3.63.Outcomes Follow-up data were obtained from patients’ medical records. The primary outcome was defined as a kidney event, a composite of either a ≥50% decrease in eGFR from baseline or the initiation of kidney replacement therapy due to kidney failure. A 50% decrease in eGFR is a well-established surrogate endpoint in nephrology research. 18 24–27 The secondary outcome was all-cause mortality. All patients received standard diabetes care in accordance with the Japanese Clinical Practice Guideline for Diabetes 2019.28Statistical analysis Baseline characteristics are summarized as proportions for categorical variables and as medians with IQRs for continuous variables exhibiting skewed distributions. The Kruskal-Wallis test and one-factor analysis of variance were used to compare median values, and Tukey’s test was used to evaluate differences in proportions. Event-free survival across lipid parameter quartiles was compared using Kaplan-Meier survival curves and log-rank tests. Cox proportional hazards models were used to evaluate the associations between each lipid profile and clinical outcomes. Univariate analyses were performed first, followed by multivariable models to adjust for potential confounders. Model 1 was adjusted for age and sex, and Model 2 included additional adjustments for smoking history, history of cardiovascular disease, BMI, systolic blood pressure, eGFR, hemoglobin A1c, and presence of proteinuria. HRs and 95% CIs are reported, and p values were calculated using the Wald χ² test. The proportional hazards assumption was assessed using Schoenfeld residuals (global test p=0.181), indicating no evidence of violation. To evaluate potential non-linear relationships between HDL-C levels and clinical outcomes, sensitivity analyses were conducted using restricted cubic spline functions adjusted for covariates in Model 2. The spline models used four knots located at the 5th, 35th, 65th, and 95th percentiles of the HDL-C distribution. Subgroup analyses were also performed to examine the robustness of the associations between HDL-C and outcomes, stratified by age (<65 vs ≥65 years), sex, eGFR (<60 vs ≥60 mL/min/1.73 m²), presence of proteinuria, and BMI (<25 vs ≥25 kg/m²). All statistical analyses were performed using SPSS Statistics V.29.0 (IBM, Armonk, New York, USA) and Stata/MP V.15.1 (StataCorp LLC, College Station, Texas, USA). A two-sided p value of <0.05 was considered significant.Results Patients’ characteristics Of the 2,724 individuals enrolled in the Fukushima Cohort Study, a total of 1,033 participants were included in the final analysis ( online supplemental figure S1). Table 1 presents the baseline characteristics stratified by HDL-C quartiles. Detailed baseline characteristics are presented in online supplemental table S1. The median age of participants was 66 years, and 56.1% were male. The median BMI was 25.2 kg/m², median eGFR was 68.6 mL/min/1.73 m², median hemoglobin A1c was 6.7%, and median HDL-C was 49 mg/dL. HDL-C quartiles differed significantly in sex, BMI, smoking, cardiovascular disease, eGFR, uric acid, TG, LDL-C, TG/HDL-C ratio, and proteinuria. Patients in the lowest HDL-C quartile (Q1) were more likely to be male and have a history of smoking, cardiovascular disease, and proteinuria. These individuals also had a higher BMI, lower eGFR, elevated serum uric acid and TG levels, lower LDL-C levels, and a higher TG/HDL-C ratio. With respect to medication use, ACE inhibitors or angiotensin II receptor blockers were more frequently prescribed in the lowest HDL-C quartile, whereas statin use was less common. Use of other lipid-lowering agents was comparable across the HDL-C quartiles.SP110.1136/bmjdrc-2025-005581.supp1Supplementary dataTable 1Patients’ baseline characteristics by HDL-C quartileVariablesMissing dataTotalHDL-C (mg/dL)P for trendn%Q1: <42Q2: 42–48Q3: 49–58Q4: 58<N1033251237269276Age (years)0066 (58–74)68 (59–74)65 (55–73)66 (58–73)66 (60–75)0.102Male sex (%)0056.174.95748.745.7<0.001History of cardiovascular disease (%)0015.323.911.412.613.4<0.001eGFR (mL/min/1.73 m2)0068.6 (53.7–81.8)62.5 (44.3–76.0)67.9 (51.5–83.0)69.4 (55.8–82.7)71.9 (61.6–83.1)<0.001Hemoglobin A1c (%)141.46.7 (6.3–7.4)6.8 (6.2–7.4)6.8 (6.3–7.4)6.8 (6.4–7.4)6.7 (6.2–7.2)0.219HDL-C (mg/dL)0049 (42–59)37 (33–39)45 (43–47)53 (51–56)66 (62–74)<0.001Proteinuria (%)191.81926.721.917.111.2<0.001The values in the table indicate medians (25–75% percentile) or percentages, as appropriate. The Kruskal-Wallis test and one-factor analysis of variance were used to compare median values, and Tukey’s test was used to evaluate differences in proportions.eGFR, estimated glomerular filtration rate; HDL-C, high-density lipoprotein cholesterol.Lipid profiles and kidney events During a median follow-up period of 5.3 years, 91 of 1,033 participants experienced a kidney event, defined as either a ≥50% decrease in eGFR from baseline or the onset of kidney failure requiring dialysis or transplantation. Kaplan-Meier analyses showed significant differences in event-free survival among HDL-C quartiles (p<0.001; figure 1a) and TG/HDL-C ratio quartiles (p=0.008). In contrast, no significant associations were observed for TG, LDL-C, or non-HDL-C quartiles (online supplemental figure S2). On univariate Cox proportional hazards analysis, individuals in the lowest HDL-C quartile (Q1) had a significantly higher risk of kidney events than the reference group (Q3). This association remained robust after adjustment for potential confounders in multivariable Models 1 and 2. In Model 2, adjusted HRs were 2.61 (95% CI 1.32 to 5.14) for Q1, 1.87 (95% CI 0.90 to 3.90) for Q2, and 2.03 (95% CI 0.93 to 4.42) for Q4 compared with the reference value of Q3 (table 2). No significant associations were found between quartiles of TG, LDL-C, or non-HDL-C and the risk of kidney events (online supplemental tables S2–S4). Though a higher TG/HDL-C ratio was associated with increased kidney risk on univariate analysis, this association was attenuated and was non-significant after multivariable adjustment (online supplemental table S5). Restricted cubic spline analysis treating HDL-C as a continuous variable and adjusting for relevant covariates showed that lower HDL-C levels were significantly associated with increased kidney event risk (p for overall association=0.003; p for non-linearity=0.018, figure 2a).Figure 1Kaplan-Meier curves for the incidence of kidney event (a) and all-cause death (b) by HDL-C quartile at baseline. HDL-C, high-density lipoprotein cholesterol.Figure 2Distributions and model-adjusted restricted cubic splines assessing the relationships between HDL-C and kidney events (a) and all-cause death (b). The solid lines represent adjusted HR estimates, and the dashed lines represent 95% CIs. Model adjusted for age, sex, smoking history, history of cardiovascular disease, body mass index, systolic blood pressure, eGFR, hemoglobin A1c, and proteinuria. HDL-C, high-density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate.Table 2HRs and 95% CIs for the associations of HDL-C levels with kidney eventsHR (95% CI)HDL-C (mg/dL)<4242–4849–5859≤Crude2.97 (1.64 to 5.38)1.68 (0.87 to 3.23)1.00 (ref.)0.98 (0.48 to 2.01)Model 13.13 (1.70 to 5.76)1.73 (0.90 to 3.35)1.00 (ref.)0.96 (0.47 to 1.96)Model 22.61 (1.32 to 5.14)1.87 (0.90 to 3.90)1.00 (ref.)2.03 (0.93 to 4.42)Model 1, adjusted for age and sex. Model 2, adjusted for Model 1 covariates plus smoking history, history of cardiovascular disease, body mass index, systolic blood pressure, eGFR, hemoglobin A1c, and proteinuria.eGFR, estimated glomerular filtration rate; HDL-C, high-density lipoprotein cholesterol.Lipid profile and all-cause mortality During the follow-up period, 78 participants died. As with kidney events, Kaplan-Meier analysis showed that individuals in the lowest HDL-C quartile (Q1) had a significantly higher risk of all-cause mortality than the other quartiles ( figure 1b). Multivariable Cox regression analysis further confirmed that low HDL-C levels were independently associated with increased risk of death. Compared with the reference group (Q3), adjusted HRs for all-cause mortality were 2.27 (95% CI 1.16 to 4.42) for Q1, 1.33 (95% CI 0.62 to 2.85) for Q2, and 1.15 (95% CI 0.55 to 2.40) for Q4 compared with Q3 (online supplemental table S6). In contrast, no significant associations were observed between all-cause mortality and quartiles of TG, LDL-C, non-HDL-C, or the TG/HDL-C ratio (online supplemental tables S7–S10). Restricted cubic spline analysis treating HDL-C as a continuous variable showed a U-shaped relationship between HDL-C levels and all-cause mortality (p for overall association=0.004; p for non-linearity=0.019), with the lowest mortality risk occurring at an HDL-C level of approximately 60 mg/dL (figure 2b).Subgroup analyses To further investigate the relationships between HDL-C levels and the risk of kidney events and all-cause mortality, subgroup analyses were conducted. These analyses did not identify any significant interactions between HDL-C and baseline characteristics, including age (<65 vs ≥65 years), sex, eGFR (<60 vs ≥60 mL/min/1.73 m²), presence of proteinuria, or BMI (<25 vs ≥25 kg/m²) ( table 3 and online supplemental table S11). For kidney events, the association between low HDL-C and increased risk appeared more pronounced in older adults (≥65 years), females, individuals with preserved kidney function (eGFR ≥45 mL/min/1.73 m²), and those without proteinuria at baseline. For all-cause mortality, the elevated risk associated with low HDL-C was more evident in males, individuals with preserved kidney function (eGFR ≥45 mL/min/1.73 m²), those without proteinuria, and those with a lower BMI (<25 kg/m²). However, none of the interaction terms between HDL-C and these covariates was significant for either kidney events or all-cause mortality, indicating no effect modification by these baseline factors.Table 3Subgroup analyses of the associations of HDL-C levels with kidney eventsHR (95% CI)P for interactionHDL-C (mg/dL)<4242–4849–5859≤Overall2.28 (1.18 to 4.38)1.53 (0.76 to 3.10)1.00 (ref.)1.78 (0.84 to 3.79)Age (y) ≥65 y2.37 (1.06 to 5.30)1.43 (0.56 to 3.65)1.00 (ref.)1.37 (0.47 to 3.96)0.812 <65 y2.07 (0.64 to 6.67)1.53 (0.48 to 4.89)1.00 (ref.)2.03 (0.62 to 6.63)Sex distribution Male1.79 (0.73 to 4.41)1.10 (0.40 to 3.02)1.00 (ref.)1.78 (0.59 to 5.39)0.541 Female2.85 (1.10 to 7.38)2.33 (0.88 to 6.18)1.00 (ref.)2.00 (0.70 to 5.76)eGFR (mL/min/1.73 m2) ≥453.40 (1.03 to 11.2)2.64 (0.81 to 8.63)1.00 (ref.)2.18 (0.67 to 7.10)0.918 <452.15 (0.97 to 4.79)1.18 (0.48 to 2.86)1.00 (ref.)1.43 (0.50 to 4.08)Proteinuria Positive2.02 (0.92 to 4.42)1.26 (0.54 to 2.95)1.00 (ref.)1.91 (0.70 to 5.18)0.863 Negative3.61 (1.10 to 11.8)1.55 (0.42 to 5.80)1.00 (ref.)1.58 (0.47 to 5.27)Body mass index (kg/m2) <252.44 (0.99 to 6.00)1.75 (0.62 to 4.90)1.00 (ref.)1.55 (0.55 to 4.42)0.992 ≥251.91 (0.72 to 5.12)1.15 (0.41 to 3.19)1.00 (ref.)1.73 (0.58 to 5.23)Adjusted for age, sex, eGFR, hemoglobin A1c, and proteinuria.eGFR, estimated glomerular filtration rate; HDL-C, high density lipoprotein cholesterol; y, years.Discussion In this study, longitudinal data from the Fukushima CKD cohort were analyzed to investigate the associations between lipid profiles, particularly HDL-C levels, and the progression of kidney disease in patients with T2DM. While TG, LDL-C, non-HDL-C, and TG/HDL-C were not associated with outcomes, lower HDL-C was linked to higher risks of kidney events and death. These associations remained robust after adjustment for multiple confounding factors, indicating an independent relationship. To the best of our knowledge, this is the first study to demonstrate a significant association between lower HDL-C levels and kidney events, defined as a composite outcome of either a ≥50% decrease in eGFR from baseline or the initiation of kidney replacement therapy due to kidney failure, in patients with T2DM.Several previous studies have primarily examined the association between HDL-C levels and kidney outcomes in the general population as well as in patients with CKD. Bowe et al reported that lower HDL-C levels were associated with increased risks of incident CKD, doubling of serum creatinine, ≥30% decrease in eGFR, and ESKD requiring kidney replacement therapy in a large cohort of US veterans.5 Similarly, a large population-based cohort study from the UK demonstrated that lower HDL-C levels were independently linked to the development of advanced CKD,8 suggesting a consistent risk pattern across different ethnic and geographic populations. Consistent findings have also been observed in Japanese populations. The Japan Specific Health Checkups study showed that HDL-C levels <40 mg/dL were associated with an increased risk of a ≥40% decrease in eGFR in the general Japanese population.7 Moreover, Tsuruya et al reported that the TG/HDL-C ratio was associated with CKD progression in Japanese adults,29 and Nam et al demonstrated that both low and high HDL-C levels were linked to CKD progression in Korean patients with non-dialysis dependent CKD,9 further highlighting the clinical significance of HDL-C as a marker of dyslipidemia contributing to adverse kidney outcomes.In patients with T2DM, several studies have reported an association between low HDL-C levels and the development or progression of diabetic nephropathy or DKD, defined as the new onset or worsening of albuminuria and/or a decrease in eGFR to <60 mL/min/1.73 m². Zoppini et al found that higher HDL-C levels were associated with lower risk of eGFR decrease to <60 mL/min/1.73 m² in Italian patients with T2DM.16 Low HDL-C levels were also reportedly associated with the risk of new onset or worsening of albuminuria.14 15 Russo et al reported that low HDL-C is associated with the risk of the development of DKD, defined as either low eGFR <60 mL/min/1.73 m² or an eGFR reduction >30% and/or albuminuria using a database of large population of outpatients with T2DM.11 A recent meta-analysis also found a relationship between low HDL-C and the risk of eGFR decrease to <60 mL/min/1.73 m² and/or albuminuria in patients with T2DM.10 Several studies have consistently demonstrated an association between low HDL-C levels and the onset of diabetic nephropathy or DKD. In addition, a subgroup analysis of the aforementioned Korean CKD cohort revealed a consistent, though borderline, relationship between low HDL-C and CKD progression in patients with and without diabetes.9 However, the relationship between HDL-C levels and hard kidney outcomes has been insufficiently investigated in patients with T2DM. Therefore, the present study is the first to prospectively evaluate the association between HDL-C levels and clinically meaningful kidney outcomes, defined as a ≥50% decrease in eGFR or the onset of kidney failure requiring kidney replacement therapy, in a Japanese cohort of patients with T2DM. Comparing HDL-C with other lipids for kidney events and mortality provided a more comprehensive view of its clinical relevance.The observed association between low HDL-C levels and an increased risk of kidney events may be attributed to several biological mechanisms. Whereas HDL is traditionally recognized for its role in reverse cholesterol transport, recent studies have emphasized its pleiotropic functions, including anti-inflammatory, antioxidative, and endothelial-protective properties.30–32 Reduced HDL-C levels may impair reverse cholesterol transport, leading to cholesterol accumulation within renal tissues, which in turn contributes to glomerulosclerosis and atherosclerosis of the renal vasculature. In addition, HDL is known to suppress proinflammatory cytokines such as tumor necrosis factor alpha, interleukin-6, and C-reactive protein. A reduction in HDL-C may weaken this anti-inflammatory effect, thereby promoting chronic inflammation and interstitial fibrosis. HDL also carries antioxidant enzymes, such as paraoxonase-1, that mitigate oxidative stress. Consequently, low HDL-C may exacerbate oxidative damage and endothelial dysfunction through increased accumulation of oxidized LDL. Furthermore, under conditions of persistent inflammation and oxidative stress, commonly seen in T2DM, HDL particles may undergo structural and functional modifications, resulting in the formation of dysfunctional HDL.33 34 These altered HDL particles may not only lose their protective functions, but they may actively contribute to vascular injury and inflammation, accelerating microvascular damage and fibrosis in the kidney. Supporting this hypothesis, experimental studies have demonstrated that administration of reconstituted HDL ameliorates diabetic nephropathy in animal models,35 suggesting that restoring HDL functionality may offer renoprotective effects. Although HDL function was not directly assessed in the present study, the finding that low HDL-C levels were associated with kidney events even in lower-risk subgroups (eg, those with preserved eGFR or without proteinuria) suggests that functional impairment of HDL may play a role in the early pathogenesis of kidney disease, preceding overt clinical deterioration. Pemafibrate, a selective peroxisome proliferator-activated receptor alpha modulator, has been reported to increase HDL-C levels in patients with T2DM.36 Future studies are warranted to determine whether therapies that increase HDL-C levels or enhance HDL function can translate into improved kidney outcomes.Although previous observational studies and meta-analyses have suggested that elevated LDL-C levels may contribute to CKD progression and that statin therapy can slow the decline in kidney function,4 37 38 there was no significant association between LDL-C levels and kidney events or all-cause mortality in the present study. This discrepancy may be explained by the characteristics of the present study cohort, which had a high prevalence of statin use and well-controlled LDL-C levels under specialist care. In patients with T2DM, the recommended LDL-C management targets for the prevention of atherosclerotic disease are <120 mg/dL for primary prevention and <100 mg/dL for secondary prevention.39 In the participants of the present study, 75% achieved LDL-C levels <120 mg/dL, and 51% achieved levels <100 mg/dL. These factors may have mitigated any potential adverse effects of LDL-C on kidney outcomes in this population. In contrast, because HDL-C levels are generally difficult to modify through standard interventions, their association with kidney events may have been more readily detectable in the present analysis.A major strength of the present study lies in the use of a relatively large, population-based cohort of patients receiving standard diabetes care, along with the application of appropriate statistical methods to ensure reliable results. However, several limitations should be acknowledged. First, lipid profiles were assessed only at baseline, and changes over time or the effects of subsequent treatments were not considered. Second, due to the observational design of the study, causal relationships cannot be established, and residual confounding from unmeasured variables—such as inflammatory markers, dietary and alcohol intake, physical activity, or the use of medications initiated during follow-up—may still exist. At baseline, the use of sodium–glucose cotransporter 2 inhibitors, mineralocorticoid receptor antagonists, and glucagon-like peptide-1 receptor agonists was minimal, and data on their initiation during follow-up were unavailable, despite their known effects on kidney disease progression. This limitation should be taken into account in future studies. In addition, urinary albumin, a key predictor of kidney disease progression in T2DM, was not included due to substantial missing data; instead, dipstick proteinuria results were used as a confounder, which may have influenced the results. Third, because of the limited sample size in the present study, the number of all-cause mortality events was relatively small, potentially limiting the statistical power for detecting associations. To more precisely confirm the association between HDL-C levels and adverse outcomes in patients with T2DM, further validation in larger cohorts using more appropriate statistical methods is warranted. Fourth, the study population consisted exclusively of Japanese individuals residing in Japan, which may limit the generalizability of our findings to other ethnic groups or healthcare settings. Therefore, caution should be exercised when extrapolating these results to broader populations. Finally, the observed association between extremely high HDL-C levels and mortality should be interpreted with caution, since causal inference cannot be drawn from observational data alone. Despite these limitations, the present findings suggest that HDL-C may serve as an independent predictor of adverse kidney outcomes and all-cause mortality in Japanese patients with T2DM. Future studies assessing HDL function and interventions are needed to refine risk stratification and identify new therapies. Moreover, because HDL-C is routinely measured in daily practice and health checkups, it may represent a simple and accessible marker for risk stratification and early intervention to prevent kidney disease progression.Conclusion In this study, low HDL-C levels were independently associated with clinically significant kidney outcomes, including a ≥50% decline in eGFR and the onset of kidney failure, as well as with all-cause mortality in Japanese patients with T2DM. As HDL-C may be both predictive and modifiable, clinicians should pay closer attention to low HDL-C. Future prospective studies are warranted to determine whether interventions aimed at increasing HDL-C levels can lead to improved kidney outcomes in patients with T2DM.