Exploring the relationship between remnant cholesterol and diabetic kidney disease in Chinese type 2 diabetes patients

Abstract

Background

Remnant cholesterol (RC) has been established as an independent risk factor for atherosclerotic cardiovascular disease. While the association between RC and diabetic kidney disease (DKD) remains unclear.

Methods

This cross-sectional study included 1,893 patients with T2D hospitalized across multiple centers from 2019 to 2024. The correlation of RC and DKD was analyzed by multiple logistic regression and restricted cubic spline (RCS) models. The subgroup analysis was to assess the stability of the correlation between RC and DKD.

Results

The participants comprise 1,340 without DKD and 553 with DKD. RC was significantly higher in DKD patients compared non-DKD patients (P = 0.012). In multiple logistic regression analysis, the results revealed a 43% increased DKD risk per 0.1 mmol/L RC increment (adjusted OR = 1.43, 95% CI = 1.10–1.86). Additionally, when analyzed as quartiles, participants in the highest RC quartile (Q4: >0.700 mmol/L) demonstrated 1.77-fold higher DKD risk compared to the lowest quartile (95% CI = 1.28–2.45, P = 0.001), with significant linear trend across quartiles (P < 0.001). Furthermore, RCS model demonstrated a biphasic relationship: risk increased linearly with RC levels below 0.96 mmol/L (β = 2.25 per 0.1 mmol/L, P = 0.001), transitioning to a plateau (β = 0.86, P = 0.588) with RC levels exceeded 0.96, suggesting lipid-mediated renal injury pathways may reach saturation. Subgroup analyses confirmed stability across demographic or clinical strata (all P > 0.05).

Conclusions

Our study establishes RC as an independent DKD biomarker in Chinese T2D patient, suggesting its dual utility as a pathophysiological indicator and preventive therapeutic target.

Introduction

The global prevalence of T2D has risen dramatically, making it one of the most pressing public health challenges. Among the microvascular complications of T2D, diabetic kidney disease (DKD) affects nearly 40% of patients, representing a significant cause of chronic kidney disease (CKD) and the leading contributor to end-stage renal disease (ESRD) (1). The progression to ESRD necessitates renal replacement therapy, which severely compromises patients' quality of life and imposes a substantial economic burden on healthcare systems. Despite its clinical importance, the pathogenesis and progression of DKD remain complex and multifactorial (2).

Dyslipidemia, characterized by elevated triglycerides and reduced high-density lipoprotein cholesterol (HDL-C), is a common metabolic abnormality in DKD patients. Previous studies have demonstrated a strong correlation between dyslipidemia and renal dysfunction in DKD, with persistent dyslipidemia not only exacerbating renal impairment but also increasing cardiovascular mortality risk (3). Among lipid parameters, remnant cholesterol (RC), a novel lipid marker representing the cholesterol content of triglyceride-rich lipoproteins (TRLs), including very low-density lipoproteins (VLDLs), intermediate-density lipoproteins (IDLs), and chylomicron remnants, has garnered increasing attention (4).

Emerging evidence has identified RC as a significant risk factor for atherosclerotic cardiovascular disease (ASCVD), hypertension, and stroke (5–7). Furthermore, RC has been shown to be positively associated with diabetes and serves as an independent predictor of T2D (4, 8). Recent studies have also suggested that RC is an independent risk factor for CKD in patients with prediabetes and T2D (9). However, studies on the relationship between RC and DKD in T2D patients are limited, particularly in the Chinese population. Zhu et al. (9). reported that RC is a risk factor for CKD in prediabetes and T2D patients based on a U.S. population, but their study lacked detailed data on diabetes-related complications and the use of medications such as angiotensin converting enzyme inhibitor (ACEI)/angiotensin II receptor blocker (ARB) or sodium-glucose cotransporter 2 inhibitors (SGLT-2i), which could influence DKD progression. Moreover, their findings suggested a linear relationship between RC and DKD in the NHANES database, but whether this relationship holds true in the Chinese population, with its distinct metabolic and genetic characteristics, remains unclear and warrants further investigation.

This study aims to address these gaps by evaluating the correlation between RC and DKD in a large cohort of Chinese T2D patients. By employing multivariable regression models to adjust for confounding factors, including diabetes-related complications and medication history, this study seeks to provide novel insights into the role of RC as a potential biomarker for early detection and progression monitoring of DKD in Chinese T2D patients.

Method

Study population

The study selected 1,893 T2D patients hospitalized in Endocrinology Department of the Second Affiliated Hospital of Guangzhou Medical University, Endocrinology Department of the Fourth Affiliated Hospital of Guangzhou Medical University, and Endocrinology Department of the Affiliated Hospital of Guangdong Medical University from 2019 to 2024. The inclusion criteria were adults with newly diagnosed and treated T2D based on the diagnostic criteria recommended by the Chinese Diabetes Society. The exclusion criteria included: (1) Comorbid kidney diseases: non-diabetic nephropathies (e.g., chronic nephritis, glomerulonephritis, polycystic kidney disease); Acute kidney injury (AKI) or renal transplantation history; (2) Severe systemic diseases: Malignancy (currently undergoing chemotherapy or radiotherapy); Decompensated liver cirrhosis; Severe cardiac insufficiency; Active infections or autoimmune diseases (e.g., systemic lupus erythematosus); (3) Pregnancy or lactation; (4) Long-term use of glucocorticoids, immunosuppressants, or non-steroidal anti-inflammatory drugs (NSAIDs); (5) Missing data: Incomplete or absent key variables essential for residual cholesterol (RC) calculation (e.g., total cholesterol, LDL-C, HDL-C). The patient screening flowchart is shown in Figure 1. This retrospective study was approved by the Ethics Committee of The Second Affiliated Hospital of Guangzhou Medical University (Approval number 2020-hs-29), following the Declaration of Helsinki.

Figure 1.

Patient screening process flowchart.

Data collection, measurement and definition

Basic clinical data of the participants were retrospectively collected as follows: gender, age, body mass index (BMI), smoking status, alcohol consumption, medical history, past history including hypertension, coronary atherosclerotic heart disease (CAHD), cerebral infarction (CI) and diabetic complication such as diabetic ketoacidosis (DKA), hyperosmolar hyperglycemic syndrome (HHS), diabetic retinopathy (DR), diabetic peripheral neuropathy (DPN), peripheral vascular disease (PVD), lower extremity arterial occlusive disease (LEAOD) and diabetic foot (DF). BMI was calculated as the ratio of weight in kilograms/height in meters squared. We collected the laboratory tests: fasting blood glucose (FBG), Hemoglobin A1c (HbA1c), fasting C peptide (F-C peptide), cholesterol, triglyceride, low-density lipoprotein cholesterol (LDL-C), low-density lipoprotein cholesterol (HDL-C), serum creatinine (Cr), estimated glomerular filtration rate (eGFR), and urea albumin creatinine ratio (UACR) levels. RC was calculated as TC minus LDL-C minus HDL-C. This estimation method has been recognized as the most common calculation method (10). A systolic blood pressure (SBP) over 140 mmHg or a diastolic blood pressure (DBP) over 90 mmHg, and the use of any anti-hypertensive medication were all considered to be hypertension. Those participants with a UACR ≥ 30 mg/g or an eGFR < 60 ml/min/1.73 m² are considered to have a Diabetic Kidney Disease (DKD), who diagnosed with diabetes based on the above criteria (11).

Statistical analysis

Continuous variables were expressed as means with standard deviations (SD) for normally distributed data or medians with interquartile ranges (IQR) for non-normally distributed data. Categorical variables were presented as frequencies and percentages. To compare differences between groups, independent t-tests or one-way analysis of variance (ANOVA) were applied for normally distributed continuous variables, while the Mann–Whitney U test or Kruskal–Wallis test was used for non-normally distributed variables. Chi-square tests were employed for categorical variables. Logistic regression models were constructed to evaluate the association between RC and the risk of DKD. Three models were developed with progressive adjustments: Non-adjusted model included no covariates, Adjust I model adjusted for gender, age and Adjust II model further adjusted for potential confounders selected a priori based on clinical relevance and prior literature, including smoking history, diabetic microvascular complications (DR, DPN, DF), macrovascular comorbidities (CAHD, hypertension, CI, PAD), and medication use (ACEI/ARB, SGLT-2i, lipid-lowering agents), because these factors are associated with DKD risk and may also correlate with lipid metabolism. To assess multicollinearity among covariates in the adjusted models, collinearity diagnostics were performed using variance inflation factors (VIFs), and no evidence of problematic multicollinearity was observed. To explore the dose-response relationship between RC and DKD, RC was analyzed both as a continuous variable and in quartiles. Generalized additive models (GAM) with penalized s Adjust II model further adjusted for potential confounders pline functions were used to visualize the potential non-linear relationship between RC and DKD. GAM smoothing was fitted using the mgcv framework with thin plate regression splines (default maximum degrees of freedom = 10), and the smoothing parameter was selected by generalized cross-validation (GCV). A restricted cubic spline (RCS) was additionally used to present the dose–response curve, with 4 knots placed at the 5th, 35th, 65th, and 95th percentiles of RC. Interaction terms were tested using log-likelihood ratio tests to assess heterogeneity across subgroups.

Missingness was quantified for all variables (Supplementary Table S1). Primary analyses were conducted using complete-case data in the final analytic cohort (N = 1,893). Because UACR has been routinely implemented in China mainly in the past ~3 years, it was not consistently available earlier; thus, we performed a sensitivity analysis restricted to participants with measured UACR.

All statistical analyses were performed using R software and EmpowerStats (X&Y Solutions, Inc., Boston, MA, USA). A two-tailed p-value of < 0.05 was considered statistically significant.

Results

Baseline characteristics of participants

A total of 1,893 T2D patients were included in this study, comprising 1,340 without DKD and 553 with DKD. As shown in Table 1, participants with DKD were older and had significantly higher levels of F-C peptide, Cr, UACR, and RC, while HbA1c, FBG, and eGFR were significantly lower compared to those without DKD (P < 0.05). Compared with the non-DKD group, the DKD group exhibited significantly higher levels of RC [0.51 (0.36–0.75) mmol/L vs. 0.46 (0.31–0.67) mmol/L, P < 0.001]. Additionally, the proportion of newly diagnosed diabetes was higher in the non-DKD group. Regarding complications, the prevalence of DR, PVD, LEAOD, DPN and DF was significantly higher in the DKD group, while DKA and HHS were more frequent in the non-DKD group. The incidence of comorbidities such as CAHD, hypertension, and CI was also significantly higher in the DKD group. Moreover, the use of ACEI/ARB and lipid-lowering drugs was more common in DKD patients, whereas BMI, cholesterol, triglycerides, HDL-C, LDL-C, smoking, drinking, and SGLT-2i use showed no significant differences between the two groups.

Table 1.

Baseline characteristics of participants according to DKD.

Items	DKD		P value
	Without	With
N	1,340	553
Age (year)	61 (51–69)	65 (57–72)	< 0.001
Gender: Male(%)	705 (52.61)	298 (53.89)	0.613
BMI (kg/m2)	24.10 (21.92–26.75)	24.51 (22.03–26.73)	0.305
HbA1c (%)	9.6 (7.4–11.8)	9.0 (7.3–11.3)	0.009
FBG (mmol/L)	7.35 (5.37–10.19)	6.74 (5.10–9.88)	0.030
F-C peptide (μg/L)	2.03 (1.30–2.87)	2.16 (1.40–3.33)	0.002
Cholesterol (mmol/L)	4.58 (3.84–5.36)	4.56 (3.78–5.46)	0.863
Triglyceride (mmol/L)	1.40 (1.02–2.00)	1.47 (1.02–2.13)	0.157
HDL-C (mmol/L)	1.02 (0.86–1.19)	0.99 (0.84–1.16)	0.052
LDL-C (mmol/L)	3.00 (2.25–3.72)	2.91 (2.18–3.75)	0.753
Cr (μmol/L)	71.00 (59.00–84.00)	82.00 (65.00–114.00)	< 0.001
eGFR (ml/min/1.73 m2)	93.26 (76.99–110.31)	78.49 (51.73–100.91)	< 0.001
UACR (mg/gCr)	16.90 (6.95–76.50)	65.40 (20.35–169.40)	< 0.001
RC (mmol/L)	0.46 (0.31–0.67)	0.51 (0.36–0.75)	< 0.001
Newly diagnosed T2D: Yes (%)	473 (35.30)	158 (28.57)	0.005
Drinking: with (%)	109 (8.13)	42 (7.59)	0.694
Smoking: with (%)	269 (20.07)	93 (16.82)	0.101
DKA/HHS: with (%)	138 (10.30)	34 (6.15)	0.004
DR: with (%)	61 (4.55)	89 (16.09)	< 0.001
PVD: with (%)	127 (9.48%)	99 (17.90%)	< 0.001
LEAOD: with (%)	255 (19.03)	181 (32.73)	< 0.001
DPN: with (%)	465 (34.70)	271 (49.01)	< 0.001
DF: with (%)	28 (2.09)	22 (3.98)	0.020
CAHD: with (%)	135 (10.07)	111 (20.07)	< 0.001
Hypertension: with (%)	590 (44.03)	392 (70.89)	< 0.001
CI: with (%)	210 (15.67)	154 (27.85)	< 0.001
ACEI/ARB: with (%)	424 (31.64)	259 (46.84)	< 0.001
SGLT-2i: with (%)	298 (22.24)	138 (24.95)	0.202
Lipid-lowering drug: with (%)	550 (41.04)	281 (50.81)	< 0.001

BMI, body mass index; HbA1c, hemoglobin A1c; FBG, fasting blood glucose; F-C peptide, fasting C peptide; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; Cr, creatinine; eGFR, estimated glomerular filtration rate; UACR, urea albumin creatinine ratio; RC, remnant cholesterol; DKA, diabetic ketoacidosis; HHS, hyperosmolar hyperglycemic syndrome; DR, diabetic retinopathy; PVD, peripheral vascular disease; LEAOD, lower extremity arterial occlusive disease; DPN, diabetic peripheral neuropathy; DF, diabetic foot; CAHD, coronary atherosclerotic heart disease; CI, cerebral infarction; ACEI, angiotensin converting-enzyme inhibitor; ARB, angiotensin II receptor blocker; SGLT-2i, sodium-glucose cotransporter protein-2 inhibitors.

Association analysis between RC and DKD

Logistic regression models were constructed to evaluate the association between RC and DKD. In the non-adjusted model, RC as a continuous variable was significantly associated with an increased risk of DKD (OR= 1.39, 95% CI = 1.10–1.75). This association remained significant in adjusted I (OR = 1.52, 95% CI = 1.20–1.93) and adjusted II models (OR =1.43, 95% CI =1.10–1.86) (Table 2).

Table 2.

Correlation of RC and DKD in T2D patients using multiple models.

Exposure	Non-adjusted	Adjust I	Adjust II
Continuous RC	1.39 (1.10, 1.75)	1.52 (1.20, 1.93)	1.43 (1.10, 1.86)
RC quartile
Q1 ( ≤ 0.32)	1.0	1.0	1.0
Q2 (0.330, 0.470)	1.42 (1.06, 1.91)	1.36 (1.01, 1.82)	1.22 (0.89, 1.69)
Q3 (0.480, 0.690)	1.54 (1.15, 2.07)	1.53 (1.14, 2.06)	1.42 (1.03, 1.96)
Q4 (≥0.700)	1.85 (1.38, 2.47)	1.97 (1.47, 2.65)	1.77 (1.28, 2.45)
P for trend	< 0.001	< 0.001	< 0.001

OR, odds ratio; 95%CI, 95% confidence interval.

Non-adjusted model: adjust for none Adjust I model: adjusted for gender, age Adjust II model: adjusted for gender, age, smoking history, DR, PAD, DPN, DF, CAHD, hypertension, CI, ACEI/ARB, SGLT-2i and lipid-lowering agents.

When RC was treated as a categorical variable, the highest quartile of RC was associated with a significantly higher DKD risk compared to the lowest quartile (OR = 1.85, 95% CI = 1.38–2.47, P < 0.001). Adjusted II models showed a similar trend, with ORs for the 2nd, 3rd, and 4th quartiles of 1.22 (95% CI = 0.89–1.69, P = 0.21), 1.42 (95% CI = 1.03–1.96, P = 0.03), and 1.77 (95% CI = 1.28–2.45, P = 0.001), respectively (Table 2). Furthermore, a significant trend was observed across quartiles in the adjusted models (P for trend < 0.001), indicating a dose-response relationship between RC levels and DKD risk. These findings suggest that higher RC levels are consistently associated with an increased risk of DKD, even after adjusting for potential confounders.

The RCS model revealed a significant linear increase in DKD risk with RC levels below 0.96, indicating a steady rise in risk of DKD as RC levels increased. However, when RC levels exceeded 0.96, the DKD risk exhibited a non-linear pattern, with the risk increasing at a slower rate compared to the linear phase (Figure 2 and Table 3).

Figure 2.

Restricted cubic spline (RCS) plots to demonstrate the level of RC in T2DM patients in relation to the occurrence of DKD. The red solid line indicates the smooth curve fit between the variables. The blue bands represent the 95% of confidence interval from the fit. The covariates that were adjusted to the model were the same as described above.

Table 3.

Threshold effects analysis of RC and DKD using liner regression models.

RC	DKD (β, 95%CI, P value)
Total
The standard linear model	1.43 (1.10, 1.86) 0.007
Model
Infection point (K)	0.96
RC < 0.96	2.25 (1.37, 3.68) 0.001
RC >0.96	0.86 (0.49, 1.49) 0.588
P for Log-likelihood ratio	0.036

Th model was adjusted for gender, age, smoking history, DR, PAD, DPN, DF, CAHD, hypertension, CI, ACEI/ARB, SGLT-2i and lipid-lowering drug.

Subgroup and sensitivity analyses

Subgroup analysis showed no significant interactions between RC and DKD risk across subgroups stratified by gender, age, BMI, RA, PVD, and LEAOD (all P for interaction > 0.05), indicating the robustness of the association (Figure 3). Notably, several subgroup-specific associations reached statistical significance (P < 0.05), which reflects within-subgroup effects rather than evidence of effect modification.

Figure 3.

The association between RC and the occurrence of DKD by subgroup analysis in T2D patients. The RC was analyzed using continuous variables, and interaction tests were also conducted to derive the interaction P-value. The black line segments represent the 95% CI for each group, and the ends of the lines represent the upper and lower 95% CI limits. The black diamond reflects the midpoint of the line segment and illustrates the effect value.

In sensitivity analyses restricted to participants with measured UACR, the association between RC and DKD remained consistent with the main analysis (Supplementary Table S2).

Discussion

This cross-sectional study was aimed to explore the underlying relationship of RC to DKD in Chinese patients with T2D. Our findings demonstrate that in T2D patients with DKD the RC level was significantly higher that those without DKD [median (IQR): 0.51 (0.36–0.75) vs. 0.46 (0.31–0.67) mmol/L, P < 0.001]. Multivariable logistic regression revealed a 43% increased DKD risk per 0.1 mmol/L RC increment (adjusted OR = 1.43, 95%CI = 1.10–1.86). When analyzed as quartiles, participants in the highest RC quartile (Q4) demonstrated 1.77-fold higher DKD risk compared to the lowest quartile (Q1) (95%CI = 1.28–2.45, P = 0.001), with significant linear trend across quartiles (P < 0.001). Furthermore, RCS plots demonstrated a biphasic relationship: below 0.96 mmol/L, risk increased linearly with RC (β = 2.25 per 0.1 mmol/L, P = 0.001), transitioning to a plateau (β = 0.86, P = 0.588) above this threshold, suggesting lipid-mediated renal injury pathways may reach saturation. The biphasic relationship in our study may was similar to previous study reported a steep-rise–plateau pattern between RC and DKD risk (12). Dai et al. (12) speculated that lower to moderate RC levels may markedly amplify endothelial dysfunction, and lipid deposition, triggering inflammatory and fibrotic pathways to accelerate renal injury. However, once these pathogenic pathways are maximally activated, further higher RC level may confer limited additional risk, contributing to a saturation-like plateau. Additionally, the subgroup analysis confirmed the correlation between RC and the risk of DKD in T2D patients. These results suggested that elevated RC levels represent a potential risk factor for DKD in T2D patients.

In recent years, RC has garnered significant attention as a novel atherosclerotic lipid indicator. Its pathophysiological significance stems from its heterogeneous characteristics—RC represents cholesterol content in a group of atherogenic remnant lipoprotein particles (including very-low-density lipoprotein remnants and chylomicron remnants) with distinct densities, volumes, and apolipoprotein compositions (13, 14). Notably, RC's metabolic disruption extends beyond the cardiovascular system. Large-scale population studies reveal its significant association with the risk of T2D development (15), where elevated baseline RC levels can increase diabetes onset risk (16). Furthermore, the association between RC and macrovascular complications in diabetes has been validated in multiple studies (17, 18). While RC has been well-established as a pathological factor in macrovascular complications of diabetes (17, 18), its role in microvascular complications remains under-explored. Zhu et al. (9) demonstrated RC's association with CKD risk in prediabetes and T2D populations but failed to distinguish DKD from non-DKD. Notably, their U.S.-based cohort lacks documentation of ACEI/ARB and SGLT-2i usage—critical medications influencing DKD progression. Cross-sectional studies associate RC with DR in T2D (19), while Jansson Sigfrids et al. (20) extended these findings to DR and DKD in type 1 diabetes mellitus (T1D) populations, suggesting RC's renal effects may transcend diabetes subtypes. Although Wu et al. (21). Firstly reported the RC-DKD association in T2D, their Martin-Hopkins equation for RC calculation (requiring dynamic adjustment based on triglycerides and non-HDL cholesterol) has proven clinically impractical. Furthermore, their model omitted adjustments for nephroprotective agents. To address these gaps, our study adopted the consensus formula (RC = TC – LDL-C – HDL-C) (4, 10), which balances scientific validity with clinical utility, and incorporated ACEI/ARB/SGLT-2i usage into multivariate models. Additionally, we established a dose-response RC-DKD relationship in Chinese T2D participants through RCS analysis.

This study confirmed through logistic regression models that the positive association between RC and DKD in T2D patients remains statistically significant, regardless of whether RC was treated as a continuous or categorical variable. The association persisted even after controlling for confounding factors including gender, age, smoking history, DR, PAD, DPN, DF, CAHD, hypertension, CI, and the use of ACEI/ARBs, SGLT-2 inhibitors, and lipid-lowering agents. Notably, while prior studies inadequately controlled for the potential influence of ACEI/ARBs, SGLT-2i, and lipid-lowering agents on DKD progression or RC levels, this study addressed these methodological gaps through rigorous adjustments. RCS modeling further revealed a biphasic dose-response relationship between RC and DKD risk: a linear increase in DKD risk was observed with rising RC levels at RC concentrations < 0.96 mmol/L, whereas the risk trajectory transitioned to a non-linear pattern when RC levels exceeded this threshold. This non-linear relationship suggests potential biological saturation of RC-mediated renal injury mechanisms beyond specific concentration thresholds. In subgroup analyses, the positive association between RC and DKD was directionally consistent across strata of sex, age, BMI, RA, PVD, and LEAOD, and no statistically significant effect modification was detected (all P for interaction > 0.05). Although several strata showed P < 0.05 for the within-subgroup association, these P values reflect statistical evidence within that subgroup and do not imply heterogeneity of effects between subgroups. The apparent differences in significance across strata are likely related to sample size and event distribution, and multiple subgroup comparisons may also yield nominally significant findings. Overall, the subgroup results support the robustness of the RC–DKD association rather than identifying a specific high-risk subgroup with a clearly different effect size. Conventional lipid parameters (e.g., triglycerides, HDL-C) have been previously validated in epidemiological and genetic studies as correlates of DKD (22–24). Our findings extend this evidence by identifying RC as an independent novel lipid biomarker associated with DKD, thereby augmenting the dyslipidemia theory of diabetic nephropathy. However, the precise molecular mechanisms underlying RC-induced renal injury remain unclear. Current hypotheses focus on three primary mechanisms: (1) the Low-Grade Inflammatory Pathway, where remnant lipoproteins activate nuclear factor κB (NF-κB) signaling through Toll-like receptor 4 (TLR4), promoting inflammatory cytokine secretion in glomerular mesangial cells (25). Inflammation is one of dominant factors that promote the progression of DKD (26). Genome-wide transcriptome analysis studies reported that inflammatory signaling pathway including NF-κB/TLR4 are closely associated with DKD development (27). Previous study reported that TLR4/NF-κB signaling could induce pyroptosis in tubular cells in DKD (28). Previous evidence has identified that RC was associated that NF-κB signaling (29, 30). (2) mTOR is a central regulator of metabolism and cell growth in response to nutrient signaling (31, 32). Activation of mTOR signaling is a key pathological node in DKD progression, driving renal injury through multiple pathways (33, 34). RC is essentially the cholesterol in the remnants of TRL. Excessive nutrients or energy can lead to dyslipidemia including the increase of RC level. And RC increase also induced energy metabolism disorder. Therefore, the mTOR signaling pathway potentially may serve as a mediator linking RC to the occurrence and development of DKD. However, the specific molecular mechanism of the effect of RC on DKD still requires further investigation. (3) The insulin resistance-mediated pathway, in which elevated RC levels create a lipotoxic microenvironment that exacerbates insulin signaling dysfunction in renal tubular epithelial cells. Higher RC level was associated with increased prevalence of IR and T2D (35). IR drives the progression of DKD through multiple pathways including metabolic disorder, inflammation and oxidative stress. IR also contributed to increase the free fatty acid (FFA) levels, lipid deposition in the kidney. We speculated that elevated RC levels create a lipotoxic microenvironment for kidney mediated by insulin resistance. However, the specific mechanism still awaits further researches to validate (36). Future mechanistic studies utilizing animal models with temporal and spatial resolution, combined with cellular/molecular approaches, will be essential to validate these mechanistic hypotheses and explore the therapeutic potential of RC-targeted interventions for diabetic nephropathy.

This study still has certain limitations. First of all, because of its cross-sectional observation and inadequate sample sizes, it was not feasible to establish a clear causal link between RC and DKD. Secondly, since RC was calculated by the formula and not directly measured, the results are subject to error which would affect its accuracy. However, in practice, it is relatively simple and easy to calculate RC in large-scale population studies. Finally, we didn't specifically categorize lipid-lowering drugs when discussing whether they could be a biasing factor for RC because of the insufficient sample size. Future studies should collect larger sample sizes and conduct prospective studies to further investigate the relationship between RC and DKD.

In conclusion, this study establishes RC as an independent risk indicator of diabetic kidney disease in Chinese T2D patients, revealing a novel biphasic threshold effect with distinct linear and nonlinear risk trajectories. These findings position RC as both a pathophysiological marker for DKD prevention.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the University Research Project of Guangzhou Education Bureau (Grant Number: 2024312097 to Yanli Li), the Guangzhou Medical University 2025-2026 Academic Year University-Level Undergraduate Innovation Training Program Project (Project Number: 20261216 to Chenfan Liu), the Doctor Initiation Program of the Second Affiliated Hospital of Guangzhou Medical University (Grant No. 010G271116 to Zhuoqing Hu), the Guangzhou Science and Technology Planning Project (Grant No. 202201010889 to Zhuoqing Hu), and the Joint Provincial and Municipal Fund of Basic and Applied Basic Research Fund of Guangdong Province in 2022 (Grant No. 2022A1515110807 to Zhuoqing Hu).

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

The studies involving humans were approved by Ethics Committee of the Second Affiliated Hospital of Guangzhou Medical University (Ethical approval number is 2020-hs-29). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

YL: Project administration, Writing – original draft, Funding acquisition, Writing – review & editing. ZX: Data curation, Conceptualization, Writing – original draft. DX: Writing – original draft, Methodology, Investigation. XY: Validation, Writing – original draft, Supervision. CL: Data curation, Writing – review & editing, Formal analysis. QX: Visualization, Conceptualization, Investigation, Writing – review & editing. CL: Data curation, Validation, Resources, Formal analysis, Writing – review & editing. YL: Software, Resources, Writing – review & editing, Validation. DC: Writing – review & editing, Visualization, Data curation, Resources. TD: Visualization, Resources, Writing – review & editing, Validation, Supervision. WL: Validation, Methodology, Supervision, Writing – review & editing, Investigation. AL: Visualization, Validation, Conceptualization, Supervision, Writing – review & editing, Writing – original draft. ZH: Project administration, Writing – review & editing, Funding acquisition, Writing – original draft.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnut.2026.1793007/full#supplementary-material