Association of immune-inflammatory markers with chronic kidney disease in older adults with obesity: a large-scale cross-sectional study

Abstract

Obesity is a major independent risk factor for chronic kidney disease (CKD). Population-level evidence linking immune dysregulation to CKD in obese individuals remains limited. In this cross-sectional study, we analyzed data from community-dwelling adults aged 65–85 years in Shanghai, China. The immune-inflammatory markers were compared between obese and non-obese groups. The associations between routinely measured immune-inflammatory markers (neutrophils, lymphocytes, platelet-to-lymphocyte ratio [PLR]) and CKD prevalence were evaluated using multivariable logistic regression, restricted cubic splines (RCSs). Obese individuals had higher neutrophil, monocyte, and lymphocyte counts and lower PLR than the non-obese group. Among participants with obese, elevated neutrophil and lymphocyte counts were independently associated with greater odds of CKD (adjusted odds ratio [aOR] for highest tertile: 1.42, 95% confidence interval [CI] 1.07–1.90; and 1.53, 95% CI 1.16–2.01, respectively), whereas higher PLR was associated with reduced odds (aOR 0.70, 95% CI 0.53–0.92). Adjusted RCS plots revealed nonlinear associations. Elevated neutrophils and monocytes together with lower PLR emerged as key immunoinflammatory markers that might be associated with obesity-associated CKD. These findings highlight the clinical potential of routine immune-inflammatory biomarkers for early and precise risk stratification of obesity-related CKD.

Graphical abstract

LAY SUMMARY

Did you know that obesity may silently harm your kidneys through your body’s own immune system? We analyzed health data from over 15,000 older adults (aged 65–85) in Shanghai and found a clear pattern: obese individuals with higher levels of certain immune cells (neutrophils and lymphocytes) and lower platelet-to-lymphocyte ratios had a 40%–50% higher risk of chronic kidney disease. The good news? These markers come from routine, inexpensive blood tests already available in primary care. This means your doctor may be able to spot early kidney risk using the very same blood work you already get during checkups. Our findings also open the door to future immune-based strategies for preventing or treating obesity-related kidney damage, bringing us one step closer to personalized intervention.

KEY MESSAGES

What is known Obesity promotes chronic kidney disease (CKD) via metabolic and hemodynamic mechanisms, but population-level evidence linking systemic immune dysregulation to CKD in obese individuals is limited.

What this study adds In obese older adults, higher neutrophils and lymphocytes and lower platelet-to-lymphocyte ratio were independently associated with ∼40%–50% higher CKD odds, with the strongest effects in women, those aged ≥75 years, and non-hypertensive participants.

Potential impact Routine, low-cost blood counts could enable immediate CKD risk stratification in obese older adults and support future trials of immunomodulatory therapies.

Introduction

Chronic kidney disease (CKD) poses a critical global public health challenge affecting approximately 673.7 million individuals worldwide, with a prevalence exceeding 10% in the general adult population [1]. In China, the disease burden is especially high, with an estimated 150.5 million patients, and 1.96 million deaths in 2019 [2]. Obesity has emerged as an important independent risk factor for CKD, driven largely by obesity-related glomerulopathy and metabolic dysfunction [3,4]. Identifying potential targets for early intervention in obesity-associated CKD is therefore imperative.

Wang et al. have systematically synthesized the interplay between traditional risk factors (dyslipidemia, hypertension, hyperglycemia, obesity) and emerging inflammatory and metabolic biomarkers, emphasizing that a significant proportion of adverse events occur in individuals deemed low or intermediate risk by conventional models. This paradigm shift underscores the clinical relevance of identifying accessible biomarkers, such as systemic immune-inflammatory indices, that may enhance risk stratification beyond traditional cardiometabolic parameters [5]. This shift highlights the clinical importance of identifying accessible biomarkers, such as systemic immune, inflammatory indices, that could improve risk stratification beyond standard cardiometabolic measures [5]. Dysregulated immune responses, especially those involving autoimmunity and adaptive immunity, have been shown to contribute to CKD pathogenesis [6,7]. Epidemiological studies further indicate that accumulation of innate immune cells correlates with CKD progression [8]. CKD patients have reduced numbers of naive and conventional T cells [9,10]. For example, they exhibit shortened telomeres in both CD4+ and CD8+ T cells [11]. Zhao MH et al. report that altered proliferation of CD4 T cells from patients with membranous nephropathy (MN) compared with cells from healthy individuals [12]. In obesity, inflammation is systemic and low-grade, driven by immune signaling that involves both T lymphocytes and innate immune components [13,14]. Therefore, obesity and CKD may share common immune-inflammatory mechanisms. Routine peripheral blood leukocyte differentials provide a simple, clinically accessible assessment of immune status [15,16]. Neutrophils, monocytes, systemic immune-inflammatory index (SII), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR) primarily reflect innate immunity, whereas lymphocytes and the lymphocyte-to-monocyte ratio (LMR) represent adaptive immune activity and the overall balance of the inflammatory response [17]. A growing body of cross-sectional and prospective cohort studies has linked systemic immune-inflammatory indices to the pathogenesis of heart failure, moyamoya disease, depressive symptoms, prostate cancer, and gouty arthritis [18–21]. Although systemic immune dysfunction in CKD is well recognized, the profile of immune-inflammatory biomarkers in CKD has not been fully characterized. In aging populations, the coexistence of obesity, metabolic dysfunction, and declining renal function is a growing concern. Zhou et al. recently document an increasing burden of chronic disease among middle-aged and older Chinese adults and highlighted the strong influence of socioeconomic factors on health disparities in this group [22]. These epidemiological findings emphasize the need for practical, low-cost methods to detect CKD early among community-dwelling older adults in China. To address this gap, this large-scale cross-sectional community-based study investigated the differential associations of innate and adaptive immune markers with prevalent CKD among older adults in Shanghai, China.

Material and methods

This cross-sectional study was performed in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. All participants or their legal representatives provided written informed consent. The study was approved by the Ethics Committee of Tongren Hospital, Shanghai Jiao Tong University (Authorization number: No. K2024-024-01). The study protocol was approved by the Institutional Ethics Committee and conformed to the principles of the Declaration of Helsinki.

Study participants

A total of 38,772 participants aged 65–85 years underwent health examinations in the Changning Strict of Shanghai, China, since February 2024. Data were collected using self-administered questionnaires, computer-assisted personal interviews, physical and functional examinations, and laboratory biochemical tests. All participants provided informed consent to participate in the study. Participants lacking data on Scr, neutrophils, monocytes, platelets, lymphocytes, NLR, PLR, LMR (N = 22,015), those with cancer or haematological disorders (N = 1,055), individuals with severe liver impairment (>5× upper limit of normal (ULN)) (N = 28), missing or aberrant BMI values (N = 1,257) were excluded from this study. Finally, a total of 14,417 participants were included (Figure 1). No formal sample size calculation was performed, as this real-world study utilized an existing large-scale data set with a sample size sufficient to support the planned analyses and ensure statistical stability.

Figure 1.

Flowchart.

Immunity markers

Blood cell counts and their ratios were used as immune biomarkers. Peripheral blood samples were collected from each participant, and after minimal local processing, shipped to the central laboratory for cryopreservation and analysis. Standard hematology assays were performed within 24 h of collection. Neutrophil, monocyte, platelet, and lymphocyte counts from the complete set of 31 parameters reported by the instrument were extracted. Using complete blood cell counts, four novel immunologic indices including SII, NLR, PLR, and LMR were derived. The SII was calculated as follows, (neutrophil count × platelet count)/lymphocyte count. NLR, PLR, and LMR were computed as the ratios of neutrophil count to lymphocyte count, platelet count to lymphocyte count, and lymphocyte count to monocyte count, respectively. To enhance the robustness of our findings, immune markers were treated as tertile-based categorical variables and continuous variables.

CKD presence

Estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equation [23]. According to the Kidney Disease: Improving Global Outcomes (KDIGO) 2012 guideline, CKD is defined as eGFR < 60 mL/min/1.73 m² and/or markers of kidney damage (e.g., albuminuria) persisting for ≥3 months [24]. Due to the cross-sectional design of this study and the absence of serial eGFR measurements and urinary albumin-to-creatinine ratio (UACR) data, we operationally defined CKD as a single eGFR < 60 mL/min/1.73 m².

Assessment of covariates

Age was categorized as 65–69 years, 70–74 years, 75–79 years, and 80–85 years. Marital status was dichotomized as married/cohabitating or separated/divorced/widowed. Based on the widely recognized evidence that metabolic risks and obesity-related comorbidities begin to escalate at a BMI threshold of 25 kg/m² in Asian populations [25,26], and considering that adherence to the Chinese clinical guideline criterion of BMI ≥ 28 kg/m² would potentially delay early clinical intervention, we adopted BMI > 25 kg/m² as the cutoff for defining obesity in the present study. Accordingly, participants were classified into two groups: the obese group (BMI > 25 kg/m²) and the non-obese group (BMI ≤ 25 kg/m²). The cutoff values for abdominal obesity were defined as waist circumference greater than 90 cm for males and greater than 85 cm for females. History of hypertension, diabetes, nonalcoholic fatty liver disease (NAFLD), coronary heart disease (CHD), and stroke was extracted from the participants’ self-reported medical histories. Laboratory indices included aspartate aminotransferase (AST), alanine aminotransferase (ALT), creatinine, urea, eGFR, hemoglobin A1c (HbA1c), high-density lipoprotein (HDL) cholesterol, high-density lipoprotein (LDL) cholesterol, triglycerides (TG), and total cholesterol (TC). Behavioral variables were categorized as smoking status (never, former [quit ≥ 12 months], current), alcohol use (never, occasional [1–3 drinks/week], frequent [≥4 drinks/week]), and physical activity (never, occasional [1–2 sessions/week], daily [≥3 sessions/week]).

Statistical analysis

According to the obese diagnosis, participants were classified into two groups: obese (BMI > 25 kg/m²) and non-obese (BMI ≤ 25 kg/m²). Subsequently, participants were classified into the CKD and non-CKD groups in the obese and non-obese groups, respectively. Characteristics were compared using the Mann-Whitney test or t-test for continuous variables based on data distribution, and the χ² test for categorical variables, expressing continuous variables as mean ± SD or median (IQR), and categorical variables as counts (percentages). The ORs for CKD according to immune markers were evaluated using univariate and multivariate logistic regression analyses. To ensure that ORs represented the evaluated contribution of an SD increment in immunity markers, lymphocytes, SII, NLR, PLR, and LMR were log-transformed before analysis. Model 1 is not adjusted. Model 2 was adjusted for demographic characteristics, including age, sex, marital status, and BMI (as continuous). Model 3 was further adjusted for lifestyle factors including smoking status, alcohol use, physical activity, and clinical and laboratory indices including baseline diabetes, TG, TC, HTN, and AST/ALT. To control the false discovery rate (FDR) at 5%, the p-values were further adjusted using the Benjamini-Hochberg method (labeled as adj-P (FDR 5%)) [27]. Restricted cubic spline (RCSs) analysis was used to further investigate the nonlinear relationships between the eight immunity markers and CKD presence. Subgroup analyses were conducted by age (65–74 years or 75–85 years), sex (male or female), BP status (HTN and non-HTN), and diabetes conditions (diabetes and non-diabetes). In the sensitivity analyses, were conducted to test the robustness the associations of neutrophils, lymphocytes, and PLR with CKD presence. The thresholds values for neutrophils, lymphocytes, and PLR were determined using RCS analysis. These thresholds were used to categorize the variables into binary groups to assess their association with the presence of CKD. Statistical significance was set at p < 0.05 (two-sided) was considered statistically significant. The analysis was performed using only available data. The data were analyzed using SAS version 9.3 (SAS Institute Inc.).

Results

Characteristics of study participants

A total of 14,417 participants were included in the initial analysis, comprising 9874 (68·5%) in the non-obese group (BMI ≤25 kg/m²) and 4543 (31·5%) in the obese group (BMI >25 kg/m²). Participants in the obese group were slightly older (72.2 [SD 4.9] vs 71.9 [SD 4.9] years, p = 0.002), more likely to be male (48.1% vs 41.2%, p < 0.001), and had a higher prevalence of hypertension (67.7% vs 53.0%, p < 0.001), diabetes (27.1% vs 21.7%, p < 0.001), and NAFLD (64.0% vs 32.7%, p < 0.001). They also had a higher level of cardiometabolic profile, including higher HbA1c, TG, and lower HDL cholesterol (all p < 0.001). Among the 3313 obese and 3313 propensity score matching (PSM)-matched non-obese individuals, the demographic variables including age, sex, and marital status, and health-related behaviors were also balanced. Similar trends were observed for the high prevalence of obesity-related comorbidities and adverse cardiometabolic profiles in the unmatched data (Table 1). Furthermore, obese individuals with CKD had a higher prevalence of HTN, diabetes, coronary heart disease, and stroke; higher levels of AST/ALT, creatine, urea, HbA1c, neutrophils, and monocytes; and lower levels of HDL cholesterol, total cholesterol, LDL cholesterol, eGFR, platelets, and PLR (all p < 0.001) (Table 2).

Table 1.

Characteristics of participants with and without obese among all participants.

	unmatched			1:1 PSM matched
	Non-obese (BMI ≤ 25 kg/m2)	Obese (BMI > 25 kg/m2)	P value	Non-obese (BMI ≤ 25 kg/m2)	Obese (BMI >25 kg/m2)	P value
Participants (n)	9874 (68.5)	4543 (31.5)		3313 (50.0)	3313 (50.0)
Demographics
Age, years	71.9 ± 5.0	72.2 ± 4.9	0.002	72.4 ± 4.9	72.4 ± 4.8	0.859
65–69	3802 (38.5)	1589 (35.0)	<0.001	1049 (31.7)	1049 (31.7)	1.000
70–74	3169 (32.1)	1590 (35.0)		1229 (37.1)	1229 (37.1)
75–79	2001 (20.3)	929 (20.4)		716 (21.6)	716 (21.6)
80–85	902 (9.1)	435 (9.6)		319 (9.6)	319 (9.6)
Male	4066 (41.2)	2185 (48.1)	<0.001	1647 (49.7)	1647 (49.7)	1.000
Married/cohabitating	8246 (83.5)	3815 (84.0)	0.506	3231 (97.5)	3231 (97.5)	1.000
Anthropometry parameters
BMI, kg/m2	22.2 ± 1.9	27.4 ± 2.7	<0.001	22.3 ± 1.9	27.3 ± 2.6	<0.001
Waist circumference, cm	80.5 ± 19.6	90.5 ± 29.1	<0.001	80.5 ± 8.0	90.1 ± 27.3	<0.001
≥90 for male or 85 for female	1584 (16.0)	2739 (60.3)	<0.001	526 (15.9)	1930 (58.3)	<0.001
History of disease
HTN	5238 (53.0)	3077 (67.7)	<0.001	1860 (56.1)	2288 (69.1)	<0.001
Diabetes	2142 (21.7)	1233 (27.1)	<0.001	788 (23.8)	928 (28.0)	<0.001
NAFLD	3232 (32.7)	2907 (64.0)	<0.001	1080 (32.6)	2092 (63.1)	<0.001
Coronary Heart Disease	1429 (14.5)	754 (16.6)	<0.001	512 (15.5)	560 (16.9)	0.109
Strokes	1533 (15.5)	815 (17.9)	<0.001	489 (14.8)	586 (17.7)	0.001
Blood tests
AST/ALT	1.336 ± 0.470	1.184 ± 0.408	<0.001	1.318 ± 0.464	1.185 ± 0.409	<0.001
Creatine, mg/dL	0.847 ± 0.200	0.877 ± 0.210	<0.001	0.868 ± 0.210	0.879 ± 0.207	0.008
Urea, mg/dL	5.683 ± 1.482	5.705 ± 1.487	0.563	5.713 ± 1.506	5.707 ± 1.429	0.960
eGFR, mL/min/l.73 m2	77.938 ± 12.615	76.505 ± 13.254	<0.001	77.401 ± 12.773	76.443 ± 13.061	0.002
HbA1c, %	6.146 ± 0.914	6.347 ± 1.028	<0.001	6.201 ± 0.979	6.322 ± 0.993	<0.001
HDL cholesterol, mmol/L	1.447 ± 0.387	1.298 ± 0.320	<0.001	1.410 ± 0.376	1.291 ± 0.322	<0.001
LDL cholesterol, mmol/L	3.007 ± 0.910	2.986 ± 0.932	0.115	2.970 ± 0.900	2.971 ± 0.926	0.776
Triglycerides (mg/dL)	1.496 ± 0.886	1.710 ± 0.964	<0.001	1.512 ± 0.906	1.695 ± 0.926	<0.001
Total cholesterol, mmol/L	5.144 ± 1.071	5.010 ± 1.097	<0.001	5.070 ± 1.067	4.986 ± 1.093	<0.001
Neutrophils (×109/L)	3.294 ± 1.109	3.482 ± 1.094	<0.001	3.350 ± 1.122	3.471 ± 1.082	<0.001
Monocytes (×109/L)	0.423 ± 0.149	0.446 ± 0.142	<0.001	0.433 ± 0.148	0.451 ± 0.143	<0.001
Platelets (×109/L)	227.875 ± 60.582	226.322 ± 60.095	0.151	227.212 ± 62.026	225.433 ± 60.193	0.352
Lymphocytes (×109/L)	2.030 (1.650,2.460)	2.110 (1.730,2.560)	<0.001	2.050 (1.660,2.490)	2.120 (1.730,2.560)	<0.001
SII	342.239 (248.391,467.608)	344.556 (254.633,471.716)	0.144	343.533 (249.329,466.480)	341.521 (253.357,470.298)	0.813
NLR	1.542 (1.185,2.007)	1.574 (1.217,2.042)	0.006	1.550 (1.191,2.016)	1.567 (1.212,2.034)	0.323
PLR	109.550 (86.667,138.462)	104.721 (84.021,129.605)	<0.001	107.730 (85.092,137.468)	103.759 (83.484,128.647)	<0.001
LMR	0.198 (0.157,0.251)	0.201 (0.160,0.254)	0.024	0.201 (0.159,0.254)	0.203 (0.163,0.257)	0.117
Health-related behavior
Smoke status			<0.001			1.000
Never	7691 (77.9)	3319 (73.1)		2774 (83.7)	2774 (83.7)
Previous	523 (5.3)	353 (7.8)		292 (8.8)	292 (8.8)
Current	558 (5.7)	314 (6.9)		247 (7.5)	247 (7.5)
Alcohol use			<0.001			1.000
Never	7234 (73.3)	3173 (69.8)		2669 (80.6)	2669 (80.6)
Occasional	1302 (13.2)	680 (15.0)		565 (17.1)	565 (17.1)
Daily	205 (2.1)	113 (2.5)		79 (2.4)	79 (2.4)
Exercise			0.177			1.000
Never	3534 (35.8)	1626 (35.8)		1374 (41.5)	1374 (41.5)
Occasional	1862 (18.9)	883 (19.4)		712 (21.5)	712 (21.5)
Daily	3421 (34.6)	1484 (32.7)		1227 (37)	1227 (37)

Abbreviations: BMI, body mass index; HTN, hypertension; NAFLD, nonalcoholic fatty liver disease; AST, aspartate aminotransferase; ALT, alanine aminotransferase; eGFR, estimated glomerular filtration rate; HbA1c, hemoglobin A1c; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SII, systemic immune-inflammation index; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; LMR, lymphocyte-to-monocyte ratio. Values are mean (standard deviation), median (inter-quartile range) or percent.

Table 2.

Characteristics of participants with and without CKD among obese and non-obese group.

	Non-obese (BMI ≤ 25)			Obese (BMI > 25)
	non-CKD	CKD	P value	non-CKD	CKD	P value
Participants (n)	2986	327		2928	385
Demographics
Age, years	72.1 ± 4.7	75.9 ± 5.1	<0.001	72.0 ± 4.6	75.6 ± 5.2	<0.001
65–69	1007 (33.7)	42 (12.8)	<0.001	994 (33.9)	55 (14.3)	<0.001
70–74	1140 (38.2)	89 (27.2)		1117 (38.1)	112 (29.1)
75–79	613 (20.5)	103 (31.5)		597 (20.4)	119 (30.9)
80–85	226 (7.6)	93 (28.4)		220 (7.5)	99 (25.7)
Male	1466 (49.1)	181 (55.4)	0.032	1439 (49.1)	208 (54.0)	0.072
Married/cohabitation	2912 (97.5)	319 (97.6)	0.972	2860 (97.7)	371 (96.4)	0.119
Anthropometry parameters
BMI, kg/m2	22.2 ± 1.9	22.4 ± 2.0	0.036	27.3 ± 2.6	27.5 ± 3.0	0.079
Waist circumference, cm	80.4 ± 7.8	81.3 ± 9.4	0.028	90.2 ± 28.8	89.6 ± 9.2	0.388
≥90 for male or 85 for female	472 (15.8)	54 (16.5)	0.660	1706 (58.3)	224 (58.2)	0.989
History of disease
HTN	1628 (54.5)	232 (70.9)	<0.001	1977 (67.5)	311 (80.8)	<0.001
Diabetes	685 (22.9)	103 (31.5)	<0.001	803 (27.4)	125 (32.5)	0.038
NAFLD	981 (32.9)	99 (30.3)	0.345	1860 (63.5)	232 (60.3)	0.212
Coronary Heart Disease	450 (15.1)	62 (19.0)	0.065	464 (15.8)	96 (24.9)	<0.001
Strokes	435 (14.6)	54 (16.5)	0.346	495 (16.9)	91 (23.6)	0.001
Blood tests
AST/ALT	1.307 ± 0.457	1.421 ± 0.510	<0.001	1.177 ± 0.411	1.246 ± 0.390	<0.001
Creatine, mg/dL	0.826 ± 0.145	1.249 ± 0.306	<0.001	0.834 ± 0.150	1.222 ± 0.255	<0.001
Urea, mg/dL	5.535 ± 1.244	7.335 ± 2.436	<0.001	5.531 ± 1.246	7.048 ± 1.930	<0.001
eGFR, mL/min/l.73m2	80.295 ± 9.415	50.976 ± 8.285	<0.001	79.703 ± 9.714	51.655 ± 7.424	<0.001
HbA1c, %	6.190 ± 0.958	6.309 ± 1.165	0.408	6.305 ± 0.992	6.455 ± 0.997	0.029
HDL cholesterol, mmol/L	1.416 ± 0.374	1.362 ± 0.385	0.005	1.297 ± 0.322	1.248 ± 0.313	0.002
LDL cholesterol, mmol/L	2.986 ± 0.896	2.826 ± 0.921	<0.001	2.996 ± 0.918	2.777 ± 0.962	<0.001
Triglycerides (mg/dL)	1.502 ± 0.903	1.607 ± 0.932	0.020	1.686 ± 0.919	1.762 ± 0.972	0.230
Total cholesterol, mmol/L	5.091 ± 1.065	4.877 ± 1.074	<0.001	5.016 ± 1.086	4.756 ± 1.120	<0.001
Neutrophils (×109/L)	3.317 ± 1.121	3.646 ± 1.081	<0.001	3.444 ± 1.070	3.677 ± 1.153	<0.001
Monocytes (×109/L)	0.428 ± 0.142	0.476 ± 0.184	<0.001	0.448 ± 0.143	0.474 ± 0.145	<0.001
Platelets (×109/L)	228.320 ± 61.780	217.098 ± 63.434	<0.001	226.027 ± 59.111	220.917 ± 67.779	0.017
Lymphocytes (×109/L)	2.050 (1.670,2.490)	2.090 (1.600,2.520)	0.855	2.110 (1.730,2.550)	2.170 (1.720,2.675)	0.095
SII	340.755 (248.572,465.055)	361.661 (258.009,480.067)	0.057	340.382 (253.597,471.636)	348.000 (253.267,458.449)	0.741
NLR	1.528 (1.179,1.992)	1.744 (1.294,2.231)	<0.001	1.556 (1.207,2.029)	1.628 (1.255,2.071)	0.073
PLR	108.271 (85.511,137.567)	102.075 (80.617,136.191)	0.016	104.725 (84.021,129.641)	98.235 (78.419,121.628)	<0.001
LMR	0.199 (0.157,0.252)	0.221 (0.172,0.275)	<0.001	0.202 (0.162,0.256)	0.211 (0.167,0.268)	0.071
Health-related behavior
Smoke status			0.448			0.280
Never	2508 (84.0)	266 (81.3)		2461 (84.1)	313 (81.3)
Previous	260 (8.7)	32 (9.8)		250 (8.5)	42 (10.9)
Current	218 (7.3)	29 (8.9)		217 (7.4)	30 (7.8)
Alcohol use			0.086			0.439
Never	2420 (81)	249 (76.1)		2352 (80.3)	317 (82.3)
Occasional	495 (16.6)	70 (21.4)		503 (17.2)	62 (16.1)
Daily	71 (2.4)	8 (2.4)		73 (2.5)	6 (1.6)
Exercise			0.020			0.540
Never	1257 (42.1)	117 (35.8)		1217 (41.6)	157 (40.8)
Occasional	624 (20.9)	88 (26.9)		621 (21.2)	91 (23.6)
Daily	1105 (37)	122 (37.3)		1090 (37.2)	137 (35.6)

Abbreviations: BMI, body mass index; HTN, hypertension; NAFLD, nonalcoholic fatty liver disease; AST, aspartate aminotransferase; ALT, alanine aminotransferase; eGFR, estimated glomerular filtration rate; HbA1c, hemoglobin A1c; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SII, systemic immune-inflammation index; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; LMR, lymphocyte-to-monocyte ratio. Values are mean (standard deviation), median (inter-quartile range) or percent.

Manifestations of immune and derived ratios markers classified by BMI and CKD

Among the 3313 obese and 3313 PSM matched non-obese individuals, obese individuals had higher levels of neutrophils, monocytes, and lymphocytes and lower levels of PLR, compared with the non-obese individuals. Among the obese individuals, individuals with CKD had higher levels of neutrophils and monocytes and lower levels of platelets and PLR. Non-obese individuals with CKD had higher levels of neutrophils, monocytes, NLR, and LMR and lower levels of platelet and PLR (Table 2, Figure 2).

Figure 2.

Manifestations of immune and derived ratios markers classified by BMI and CKD presence.

Associations of immune and immune-derived ratios markers with CKD presence among the obese individuals

In multivariable models (Table 3, model 3), continuous neutrophils and lymphocytes (log transformed) were positively associated with CKD presence, and continuous PLR (log transformed) was negatively associated with CKD presence (neutrophils: adjusted odds ratio [aOR] 1.141; 95% CI 1.031–1.262); log-lymphocytes: aOR = 3.384; 95% CI 1.465–7.821; log-PLR: aOR = 0.695; 95% CI 0.526–0.919). Notably, individuals in the highest tertile of both neutrophil and log-lymphocyte counts exhibited an around 1.5-fold times risk of CKD, whereas those in the highest tertile of the PLR showed a 0.30-fold reduced risk (neutrophils: aOR = 1.423; 95% CI 1.069–1.895; lymphocytes: aOR = 1.530; 95% CI 1.163–2.013; log PLR: aOR = 0.695; 95% CI 0.526–0.919). The p-values in log-lymphocytes and log-PLR found in logistic analysis did not significantly change in FDR adjusting (Table 3). The adjusted RCSs plots revealed nonlinear associations between the immune and derived ratio markers and CKD presence, with the immune and derived ratio markers being kept as continuous variables (Figure 3).

Table 3.

ORs (95% Cls) for CKD according to the blood cell counts and derived ratios among participants with obese.

Immunity markers	Model 1			Model 2			Model 3
	OR (95% CI)	P value	adi-P (FDR 5%)	OR (95% CI)	P value	adi-P (FDR 5%)	OR(95% CI)	P value	adi-P (FDR 5%)
Neutrophils (×109/L)
Continous (per 1 SD)	1.205 (1.099,1.321)	<0.001	<0.001	1.172 (1.062,1.293)	0.002	0.016	1.141 (1.031,1.262)	0.010	0.080
T1 (≤2.92; N = 1116)	Ref			Ref			Ref
T2 (>2.92–3.80; N = 1096)	1.422 (1.080,1.874)	0.012	0.096	1.284 (0.965,1.708)	0.086	0.688	1.255 (0.940,1.676)	0.123	0.984
T3 (>3.80; N = 1101)	1.702 (1.302,2.225)	<0.001	<0.001	1.508 (1.140,1.994)	0.004	0.032	1.423 (1.069,1.895)	0.016	0.128
Monocytes (×109/L)
Continous (per 1 SD)	3.282 (1.635,6.589)	<0.001	<0.001	2.122 (0.992,4.538)	0.053	0.106	2.011 (0.924,4.378)	0.078	0.156
T1(≤0.38; N = 1189)	Ref			Ref			Ref
T2(>0.38–0.49; N = 1064)	1.340 (1.024,1.752)	0.033	0.264	1.284 (0.972,1.697)	0.079	0.158	1.220 (0.920,1.618)	0.167	0.334
T3(>0.49; N = 1060)	1.539 (1.184,2.001)	0.001	0.008	1.315 (0.993,1.741)	0.056	0.112	1.241 (0.933,1.651)	0.138	0.276
Platelets (×109/L)
Continous (per 1 SD)	0.999 (0.997,1.000)	0.116	0.155	1.000 (0.998,1.002)	0.794	0.907	1.000 (0.998,1.002)	0.831	0.831
T1(≤200; N = 1116)	Ref			Ref			Ref
T2(>200–244; N = 1094)	0.785 (0.607,1.015)	0.065	0.087	0.903 (0.691,1.181)	0.456	0.521	0.948 (0.722,1.245)	0.699	0.699
T3(>244; N = 1103)	0.777 (0.601,1.006)	0.055	0.073	0.940 (0.717,1.234)	0.657	0.751	0.940 (0.711,1.241)	0.660	0.660
Lymphocytes (×109/L, Lg transformed)
Continous (per 1 SD)	2.156 (0.955,4.868)	0.064	0.128	2.940 (1.292,6.690)	0.010	0.040	3.384 (1.465,7.821)	0.004	0.016
T1(≤0.27; N = 1117)	Ref			Ref			Ref
T2(>0.27–0.38; N = 1102)	1.006 (0.769,1.315)	0.966	1.932	1.138 (0.862,1.502)	0.360	1.440	1.196 (0.901,1.587)	0.216	0.864
T3(>0.38; N = 1094)	1.280 (0.989,1.655)	0.060	0.120	1.432 (1.097,1.870)	0.008	0.032	1.530 (1.163,2.013)	0.002	0.008
SII (Lg transformed)
Continous (per 1 SD)	1.088 (0.651,1.816)	0.748	0.748	1.033 (0.614,1.739)	0.902	0.902	0.932 (0.552,1.572)	0.791	0.904
T1(≤2.45; N = 1104)	Ref			Ref			Ref
T2(>2.45–2.62; N = 1104)	1.162 (0.895,1.508)	0.260	0.260	1.240 (0.947,1.625)	0.118	0.118	1.207 (0.918,1.587)	0.177	0.202
T3(>2.62; N = 1105)	1.074 (0.825,1.399)	0.595	0.595	1.018 (0.775,1.337)	0.899	0.899	0.959 (0.726,1.266)	0.767	0.877
NLR (Lg transformed)
Continous (per 1 SD)	1.699 (0.917,3.148)	0.092	0.147	1.125 (0.593,2.134)	0.718	0.957	0.913 (0.478,1.746)	0.784	1.045
T1(≤0.12; N = 1105)	Ref			Ref			Ref
T2(>0.12–0.26; N = 1106)	1.086 (0.832,1.418)	0.546	0.874	1.022 (0.776,1.347)	0.876	1.168	0.989 (0.748,1.306)	0.936	1.248
T3(>0.26; N = 1102)	1.249 (0.963,1.621)	0.094	0.150	1.030 (0.785,1.352)	0.832	1.109	0.944 (0.715,1.246)	0.684	0.912
PLR (Lg transformed)
Continous (per 1 SD)	0.335 (0.167,0.673)	0.002	0.005	0.413 (0.204,0.837)	0.014	0.037	0.404 (0.1988,0.824)	0.013	0.035
T1(≤1.95; N = 1104)	Ref			Ref			Ref
T2(>1.95–2.07; N = 1104)	0.842 (0.655,1.083)	0.180	0.480	0.928 (0.715,1.203)	0.571	1.523	0.927 (0.712,1.208)	0.576	1.536
T3(>2.07; N = 1105)	0.656 (0.503,0.855)	0.002	0.005	0.705 (0.535,0.928)	0.013	0.035	0.695 (0.526,0.919)	0.011	0.029
LMR (Lg transformed)
Continous (per 1 SD)	1.723 (0.866,3.430)	0.121	0.138	0.855 (0.412,1.776)	0.675	1.080	0.726 (0.347,1.519)	0.396	0.634
T1(≤–0.75; N = 1100)	Ref			Ref			Ref
T2(>–0.75 to –0.63; N = 1108)	1.097 (0.841,1.432)	0.493	0.563	0.979 (0.742,1.291)	0.879	1.406	0.956 (0.722,1.264)	0.750	1.200
T3(>–0.63; N = 1105)	1.219 (0.939,1.583)	0.137	0.157	0.955 (0.723,1.261)	0.744	1.190	0.901 (0.679,1.195)	0.470	0.752

Abbreviations: SII, systemic immune-inflammation index; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; LMR, lymphocyte-to-monocyte ratio. Model 1 was unadjusted. Mode 2 was adjusted for age, sex, marital status and BMI (as continuous). Model 3 was adjusted for age, sex, marital status, BMI (as continuous), smoke status, alcohol use, physitcal activity, baseine diabetes, TG, TC, HTN and AST/ALT. Lymphocytes, SII, NLR, PLR and LMR were log transformed make OR relect the evaluated contribution of one SD incremen of immune markers. The p values were further adusted to control the false discovery rate at 5% using the Beniamini-Hochberg method labeled as adiP (FDR). Bold value indicates statistical significance (p < 0.05).

Figure 3.

Restricted cubic spline regression analysis of immunity marker indices for CKD presence among participants with obesity A Neutrophils. B Monocytes. C Platelets. D Lymphocytes. E SII. F NLR. G PLR. H LMR. Lymphocytes, SII, NLR, PLR and LMR were log transformed.

Subgroup analyses

Subgroup analyses were conducted by age (65–74 years or 75–85 years), sex (male or female), HTN conditions (HTN and non-HTN), and diabetes conditions (diabetes and non-diabetes). After FDR adjustment, the positive associations between neutrophil counts and CKD prevalence were significantly enhanced in participants aged 75–85 years and those with HTN. Similarly, log-lymphocyte exhibited stronger positive associations with CKD in females, the 75–85-year age group, and non-HTN individuals. Conversely, inverse associations between log-PLR and CKD were observed in females, the 75–85-year group, and non-HTN participants (Supplementary Tables 1–4).

Sensitivity and explanatory analysis

In RCS analysis, the thresholds values were identified as 0.33 × 10⁹/L for log-lymphocytes and 2.02 for the log-PLR. Elevated log-lymphocytes (>0.33 × 10⁹/L) or low log-PLR (<2.02 × 10⁹/L) were independently associated with the presence of CKD presence. (log-lymphocytes > 0.33 × 10⁹/L: OR = 1.28 (95% CI 1.02–1.60); p = 0.034; log-PLR < 2.02 × 10⁹/L: OR = 1.39 (95% CI 1.11–1.74); p = 0.005) (Supplementary Table 5). Given the exploratory nature of these analyses and the lack of external validation, these findings should be interpreted with caution and are not yet ready for clinical application. The decision tree detected 7 paths associated with CKD presence. The concomitant presence of HTN, high levels of neutrophils (≥3.035 × 10⁹/L), and CHD defined the class at the highest risk of CKD, unlike the reference class represented by individuals with the concomitant presence of non-HTN, and low levels of Lg-PLR (≤1.905) (aOR = 4.33; 95% CI 2.60–7.19; p < 0.001) (Supplementary Figure 1). In addition, we performed an additional 1:1 PSM analysis fully adjusting for comprehensive conventional risk factors, including age, sex, marital status, hypertension, diabetes, NAFLD, coronary heart disease, stroke, and multiple health-related behaviors (Supplementary Tables 6 and 7). After rigorous matching, the core associations between immune-inflammatory markers and CKD presence remained stable in the obese population. Specifically, elevated neutrophils, monocytes, lymphocytes, and LMR were positively associated with CKD occurrence. After Benjamini-Hochberg FDR correction for multiple testing, these associations remained statistically significant or marginally significant, which was highly consistent with the primary results of our study. When participants were stratified by BMI ≥ 28 kg/m², no significant associations between immune-inflammatory markers (including neutrophils, monocytes, platelets, lymphocytes, SII, NLR, PLR, and LMR) and CKD presence were observed in obese individuals. (Supplementary Tables 8–10).

Discussion

In this large-scale cross-sectional study of community-dwelling residents of Shanghai, we demonstrated for the first time that dysregulated immune-inflammatory markers, specifically elevated neutrophils, lymphocytes, and reduced PLR, are independently associated with the presence of CKD among obese individuals. These findings demonstrating a correlation between circulating immune dysregulation and CKD in obesity, are consistent the emerging paradigm that structural cells function as environment-supporting innate immune cells [28].

Neutrophils are central to innate immunity and indicate acute inflammatory states. Elevated neutrophil counts also suppress T-lymphocyte activity [29]. The positive association between lymphocyte count and CKD observed in our study remains contentious. Guo et al. reported that higher systemic immune-inflammation index (SII) scores were significantly associated with increased CKD risk, which indirectly suggests an inverse relationship between lymphocyte levels and CKD risk [30]. Mechanistically, these observations reinforce the view of obesity as a chronic inflammatory condition. Persistent elevation of neutrophils may reflect sustained low-grade inflammation and oxidative stress, processes that are linked to endothelial dysfunction and glomerular damage [31]. Obesity-related metabolic and inflammatory cues can activate endothelial signaling pathways, notably the C-X-C motif chemokine receptor family. In particular, activation of endothelial CXCR2 has been implicated in amplifying renal inflammation via the NF-κB pathway, promoting immune cell infiltration and causing endothelial glycocalyx shedding; together, these effects contribute to glomerular injury and impaired function. This endothelial-centered mechanism plausibly links the observed rise in circulating neutrophils (which can be recruited by CXCR2 ligands) to the resulting kidney damage [32]. The independent association between higher lymphocyte counts and CKD suggests a role for adaptive immune activity, possibly directed against modified renal antigens in the context of metabolic dysregulation [33]. The observed protective effect of a lower PLR was driven mainly by a relative reduction in platelet counts, which could reflect reduced thrombotic microangiopathy or altered platelet–endothelial interactions in obese individuals with preserved renal function [34]. These mechanistic interpretations remain speculative; given our cross-sectional design, we cannot confirm the specific molecular pathways.

The tonsil–glomerular axis has been implicated in the pathogenesis of IgA nephropathy, and T-nodule formation in tonsillar tissue correlates with active glomerular crescents [35]. This observation suggests that some lymphocyte elevation in our CKD cohort may reflect shared mucosal immune dysfunction rather than simple confounding. Data on conditions known to affect lymphocyte counts, chronic pharyngitis, chronic tonsillitis, autoimmune diseases, acute infections, and use of immunomodulatory medications, were not available for this study [36]. Although our multivariable models adjusted for major cardiometabolic comorbidities, residual confounding from these unmeasured factors cannot be excluded.

The role of insulin resistance as a key pathogenic link between obesity, immune dysregulation, and CKD warrants consideration. Little et al. have elucidated the molecular pathways through which insulin resistance, exacerbated by adipose tissue inflammation and adipokine dysregulation, contributes to glomerular injury and progressive renal dysfunction [37]. The decision tree analysis identified concurrent hypertension, elevated neutrophil count, and established CHD as the features conferring the highest risk of CKD. This may be explained by the pathophysiological primacy of hemodynamic stress and inflammatory cascades over direct metabolic injury in advanced renal decline. Glycemic parameters, though mechanistically linked to CKD initiation through advanced glycation end-products, may become less discriminative in late-stage disease where non-metabolic pathways dominate progression.

The RCS analyses revealed potential inflection points for log-lymphocytes (0.33 × 10⁹/L) and log-PLR (2.02). While these numeric values might offer preliminary insights for future etiological research, we emphasize that they were derived from an exploratory analysis of a single dataset without external validation. Several factors limit their current clinical applicability.We explicitly call for their replication in large, diverse, and prospectively designed cohorts before any consideration is given to clinical translation. Subgroup analyses revealed that lymphocytes are a robust risk indicator in females, older adults (≥75 years) with a 3- to 14-fold increase in CKD risk, and non-HTN individuals. PLR might serve as a consistent protective factor across multiple high-risk subgroups, with a maximal effect size in non-HTN individuals. This study suggests that age or sex modulates immune-renal crosstalk, potentially via hormonal influences on macrophage polarization or age-related immunosenescence [30–32]. HTN-and/or diabetes-driven renal injury may dominate pathological processes, obscuring independent immune contributions. In non-HTN/non-diabetic individuals, immune dysregulation (e.g., monocyte infiltration or lymphocyte activation) may serve as a primary driver of early CKD. These findings advocate stratified approaches to CKD risk assessment, where immune profiling could enhance early detection in specific populations (e.g., females with normotension).

These findings have several important clinical implications. They suggested that inflammatory indices warrant inclusion in existing CKD risk stratification frameworks. Furthermore, targeted immunomodulatory strategies, such as neutrophil extracellular trap inhibition or lymphocyte subset-specific therapies, hold promise for personalized interventions in high-risk obese populations. However, this study had some limitations. First, it was a cross-sectional analysis, meaning that causality could not be established definitively and reverse causality cannot be excluded. In addition, we acknowledge the potential for residual confounding due to the absence of several clinically relevant variables that may influence both immune markers and CKD risk. Specifically, data on albuminuria/proteinuria, use of renin-angiotensin system inhibitors, sodium-glucose cotransporter 2 inhibitors, glucagon-like peptide-1 receptor agonists, non-steroidal anti-inflammatory drugs, corticosteroid therapy, acute infection, and autoimmune diseases were not available in this study. Second, the absence of flow cytometry and ELISA data in this study limits the investigation of the association between immune cell subsets and CKD risk, particularly lymphocyte subsets. Third, owing to the lack of standardization and the absence of reference intervals and cutoffs for immune markers, it was difficult to implement the results in clinical practice. Fourth, we defined CKD by a single eGFR measurement <60 mL/min/1.73 m² without UACR data or confirmatory testing after three months, which does not fully meet KDIGO diagnostic criteria. This approach may have excluded participants with normal eGFR but significant albuminuria and thus introduced misclassification bias. Such non-differential misclassification would likely bias effect estimates and cause underestimation of true associations. The absence of UACR data also prevents complete CKD staging and the distinction between albuminuric and non-albuminuric phenotypes. Fifth, the proposed cutoffs lack external validation, and the lymphocyte threshold (0.33 × 10⁹/L) falls below the normal range, limiting interpretability. These exploratory findings require independent prospective replication before clinical translation.

In conclusion, the present study demonstrated that higher neutrophil and lymphocyte counts and lower PLR were independently associated with prevalent CKD in obese older adults. These findings suggest a potential link between systemic immune-inflammatory dysregulation and obesity-related renal impairment. The nonlinear associations observed in this cohort further enrich the epidemiological evidence regarding immune-related renal injury. Notably, as a cross-sectional exploratory analysis, this study cannot establish causal relationships. Further prospective cohort studies and mechanistic investigations are warranted to validate our findings.

Funding Statement

This work was supported by the National Natural Science Foundation of China (Grant No. 82170745; Grant No. 82502104) , the Research Fund of Shanghai Tongren Hospital, Shanghai Jiao Tong University School of Medicine (Grant No. TRYJ2022LC01, TR2024RC12) and Changning District Public Hospital Reform and High-Quality Development Enhancement Project (No. 2026GY008).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statements

Due to privacy restrictions, the data supporting this study are available only to qualified researchers upon reasonable request. Requests for data access should be directed to the corresponding author at ouyang1985@sjtu.edu.cn. The dataset DOI is https://doi.org/10.4121/526614ae-f023-4e30-ae26-615f6656af0f.