Combined Effects of Dyslipidemia and Obesity Parameters on The Estimated Glomerular Filtration Rate in a Middle-Aged Population

Some studies have reported that chronic kidney disease (CKD) or the estimated glomerular ltration rate (eGFR) is signicantly associated with metabolic-related parameters, such as dyslipidemia and obesity. However, whether eGFR will change under the superposition of multiple specic metabolic indicators remains unclear. in this was by The and to


Background
In the past twenty years, chronic kidney disease (CKD), a public health burden with an increasing prevalence, has received increasing attention. According to the results of a cross-sectional survey involving 47,204 subjects, the prevalence of CKD in China has reached 10.8% [1]. Additionally, increasing evidence has shown that CKD is an independent risk factor for cardiovascular disease, cognitive dysfunction, and all-cause mortality [2], which pose serious threats to the survival period and quality of life. Therefore, the potential risk factors for CKD and identifying possible prevention strategies to curb this growing public health problem must be explored. CKD is mainly diagnosed by a reduced estimated glomerular ltration rate (eGFR), a feasible indicator to assess renal function both in disease conditions and healthy subjects [3]. eGFR is a very sensitive clinical indicator; in subjects with an eGFR higher than 60 ml/min per 1.73 m 2 , eGFR can also be considered a signi cant predictor of adverse cardiovascular events (estimated by the Framingham risk score) [4].
Additionally, a cluster of metabolic diseases correlates with the various degrees of eGFR decline. A crosssectional study covering 33,300 Chinese adults concluded that abdominal obesity, hypertension, and dyslipidemia are independent risk factors for decreased eGFR, even for subjects with an eGFR greater than 60 ml/min per 1.73 m 2 [5]. Another cross-sectional study recruiting 75,468 urban workers also revealed that an elevated blood pressure, fasting blood glucose, and dyslipidemia are independent risk factors for reduced eGFR, but the study did not identify a signi cant association between obesity and reduced eGFR [6]. A previous meta-analysis including 11 studies (N = 30,146) found that metabolic syndrome (MetS) is also an independent risk factor for decreasing eGFR. This nding shows that, under the superimposed effect of different metabolic components, renal ltration function might be also affected [7]. Therefore, because the prevalence of cardiometabolic disorders and CKD has been increasing in recent years, the potential relationship between the two must be explored. Although the increase in CKD prevalence may be due to lifestyle changes, life expectancy extension and medical technology development, considering the extensive biological effects of metabolic disorders, we should further consider and explore the impact of different metabolic factors on CKD. However, previous studies had limitations. All the studies explored the independent impact of a single metabolic parameter or overall MetS on eGFR, rather than the impact of the superposition of two speci c metabolic abnormalities.
Additionally, some conventional adiposity indicators are insu cient to predict the risk of adverse outcomes or re ect the wellbeing of participants. The development of anthropometrics meets the needs of clinical and epidemiological investigation and research, can accurately re ect the degree of human lipid accumulation and visceral fat content, and has its own characteristics in predicting the risk of obesity-related diseases. For example, emerging studies have recommended measuring the area of visceral or subcutaneous fat to evaluate the association between abdominal obesity and multiple adverse outcomes [8][9][10]. This evidence has provided more practical suggestions to prevent and manage obesity.
This study aimed to explore the association between the combined effects of dyslipidemia and obesity on eGFR in subjects with relatively normal renal function (eGFR > 60 ml/min per 1.73 m 2 ). The distribution and accumulation of abdominal fat were expressed as the visceral fat area (VFA) and subcutaneous fat area (SFA), estimated using magnetic resonance imaging (MRI). Additionally, the body mass index (BMI) and waist-hip ratio (WHR) were evaluated. These obesity indices were investigated to determine whether they correlated with decreased eGFR independently or via the combined effects with dyslipidemia. Our ndings may have a certain reference effect on speci c individuals in clinical practice.

Study Participants
The data analyzed in the present study were derived from the baseline population of a community-based cohort study in Shenyang, Liaoning Province [11]. Seven hundred fty subjects aged 40-65 years were recruited for the health examination project [Nanzhan Community Survey of Metabolic Disorders, NOVEMBER Study]. The study was approved by the Ethics Committee of the First Hospital of China Medical University. The study was performed in accordance with the principles of the Declaration of Helsinki. All the participants signed an informed consent form. Participants who met one of the following criteria were preliminarily excluded: 1) pregnant women or within the rst year of the postpartum period; 2) participants with a personal history of thyroid dysfunction or currently using thyroid medications; 3) participants with a personal history of malignant tumor or other chronic wasting diseases; 4) participants without complete abdominal MRI information.

Data Collection
All the participants were required to participate in the survey after an overnight fast for at least 10 hours.
Demographic information such as sex, date of birth, educational quali cation, smoking and drinking status, and personal and family history of multiple diseases or medications were acquired using a standardized questionnaire. Each subject was measured for waist and hip circumference by trained nurses, and WHR was directly calculated. Weight and height were measured when the participants wore underwear without shoes. The BMI was calculated using the following formula: BMI = weight (kg) / height squared (m 2 ). Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured twice using a mercury sphygmomanometer on the right arm after a rest period for over 30 minutes. The average of the two measurements was calculated and regarded as the nal BP value.
MRI scans were obtained at the abdominal level between the L4 and L5 vertebrae in the prone position (FOV: 42 cm*42 cm; thickness: 1 cm; 6 layers; GE, USA). SFA and VFA were calculated by two technicians using SLICE-O-MATIC software (version 4.2).
Fasting venous blood was collected from each subject. The serum and plasma were stored immediately at -20°C and sent to the Endocrine Laboratory at the First Hospital of China Medical University. Additionally, each participant was asked to complete a 75-g oral glucose tolerance test, and 2-hour plasma glucose levels were also measured in the same laboratory. Serum thyroid-stimulating hormone (TSH) was detected using the Cobas 601 Analyzer (Roche Diagnostics, Basel, Switzerland). Highperformance liquid chromatography (BioRad VARIANT II Hemoglobin Analyzer, California, US) was applied to detect glycosylated hemoglobin (HbA1c) in venous blood samples. Fasting and 2-hour plasma glucose, fasting serum total cholesterol (TC), low-density lipoprotein cholesterol (LDL-c), high-density lipoprotein cholesterol (HDL-c), triglycerides (TG), and serum creatinine were all measured using an autobiochemical analyzer (Mindray BS180, Shenzhen, China).
Renal ltration function was assessed by eGFR. Similar to the nationwide study conducted by Zhang et al., eGFR was calculated using the modi ed Modi cation of Diet in Renal Disease (MDRD) equation, derived from the data on patients with chronic kidney disease in China [1,12]. The formula for calculating eGFR was as follows (Scr in mg/dL and age in years), and participants with an eGFR ≤ 60 ml/min per 1.73 m 2 were all excluded.

Diagnostic Criteria
If the subjects admitted that they were taking antihypertensive drugs or the average level of SBP or DBP was higher than 140/90 mmHg, hypertension was diagnosed. According to the kit instructions, the reference interval of TSH is 0.27-4.20 mU/L. If TSH exceeds or falls below the range or if the subject admits to a personal history of thyroid disease or is currently taking thyroid medications, the subject would be diagnosed with thyroid dysfunction. Diabetes was diagnosed based on the 2018 American Diabetes Association (ADA) guidelines [13]. If the subject had self-reported diabetes or met one of the following items, diabetes was diagnosed: fasting plasma glucose ≥7 mmol/L; two-hour plasma glucose ≥11.1 mmol/L; HbA1c ≥6.5%.
The diagnostic criteria for dyslipidemia were extracted from the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel: 1.7 mmol/L, 5.2 mmol/L, and 3.4 mmol/L were regarded as the thresholds for TG, TC and LDL-c abnormalities, respectively (described as "borderline high" in the guidelines). Additionally, 1.0 mmol/L and 1.3 mmol/L were applied as the thresholds for HDL-c abnormalities (described as a "marginal risk factor" in the guidelines) among men and women, respectively. If the subject met one of the above items or was currently taking lipid-lowering medications, dyslipidemia was determined.
In the present study, the indicators for assessing obesity mainly included BMI, WHR, VFA and SFA. High BMI was de ned as BMI ≥28 kg/m 2 , regardless of sex [14]. The threshold value of WHR was 0.88 in men and 0.86 in women [15]. The optimal threshold value of VFA was set at 80 cm 2 [16]. Additionally, if the subject's SFA were higher than its 75 th percentile, high SFA would be diagnosed [17].

Statistical Analysis
The above data were input into the Statistical Package for the Social Sciences version 25 (SPSS Inc., Chicago, US). All p values obtained were based on two-tailed tests, with signi cance levels set at 0.05. In the descriptive analysis, continuous variables were described as means and standard deviation (SD), and dichotomous variables were described as numbers and corresponding percentages. Single-sample t-test and chi-squared test were used to compare differences in continuous and dichotomous variables, respectively, between the groups with different characteristics. The Spearman correlation coe cient was applied to estimate the relationships between eGFR and variables, including age, smoking status, and drinking status. Monovariate and multivariate linear regression were applied to analyze the effects of different variables on eGFR.

Baseline Characteristics
Six hundred forty-six residents aged 45-60 years were enrolled in the nal data analysis and included 319 men and 327 women. Table 1 shows the baseline characteristics overall and according to sex.
Certain differences were found in the lifestyles of men and women. Among them, the proportion of smokers (78.4% vs. 4.6%; p < 0.05) and drinkers (75.2% vs. 13.5%; p < 0.05) was signi cantly higher in men, who had a relatively unstable dining habit (88.4% vs. 95.7%; p < 0.05) and preferred to drink tea (40.8% vs. 15  Regular tea consumption is de ned as drinking tea every day. Regular diet is de ned as having at least two meals per day and having a xed mealtime.
HTN is de ned according to the current blood pressure measurement or current administration of antihypertensive medications. Dyslipidemia is de ned according to the current lipid levels or current administration of lipid lowering medications. Diabetes is de ned as the sum-up of self-reported diabetes and newly-diagnosed diabetes.
* indicates that the corresponding value in males are signi cantly different from females (p value < 0.05).

Correlation Between The Egfr Levels And Variables
The correlation between the eGFR levels and various anthropometric and biochemical parameters according to sex is shown in Table 2. Except for dietary habits and diabetes, all the correlations were statistically signi cant at p < 0.001, regardless of sex. In each group with a single sex, some of the variables were also signi cantly correlated with eGFR. Age and VFA were both negatively correlated with eGFR in men; and age, WHR, and dyslipidemia were all signi cantly negatively correlated with eGFR in women.  Table 3 presents the mean eGFR levels of subjects with isolated dyslipidemia without obesity, isolated obesity without dyslipidemia, and subjects with or without both features. The above results were also presented according to sex. Subjects with neither high obesity indices nor dyslipidemia were set as the reference. Regarding men with either dyslipidemia or a high obesity index, the eGFR values were signi cantly lower in those with isolated high BMI (p = 0.009) and isolated dyslipidemia without high WHR (p = 0.046). Additionally, except for subjects with both dyslipidemia and high SFA, the eGFR levels were all signi cantly lower in the other three double-positive subgroups. In male subjects, the decreasing trend of eGFR reached signi cance with the superposition of dyslipidemia and high BMI (p = 0.002) or SFA (p = 0.045). Regarding women with either dyslipidemia or a high obesity index, the levels of eGFR were signi cantly lower in subjects with isolated high WHR (p = 0.002) or isolated high SFA (p = 0.007). Except for BMI, the other three double-positive subgroups all showed a signi cantly lower eGFR in women. In female subjects, the decreasing trend of eGFR reached signi cance with the superposition of dyslipidemia and high WHR (p = 0.014) or SFA (p = 0.043).

Multivariate Linear Regression Analysis between eGFR and Dyslipidemia and Obesity
The abovementioned four obesity indicators (i.e., BMI, WHR, VFA, SFA) and dyslipidemia were assessed to determine whether they were independently associated with the decrease in eGFR (Table 4). According to the results of Spearman's correlation analysis, age, sex, smoking status, drinking status, tea consumption, and hypertension were set as confounding factors in model 2. None of the confounding factors was adjusted in model 1. According to the results in model 1, high BMI, high WHR, high VFA, and dyslipidemia were all risk factors for decreased eGFR (p < 0.05) in the general population. Additionally, high BMI and high WHR were negatively associated with eGFR in men and women, respectively (p < 0.05).
However, most of the crude results in model 1 did not show signi cance after adjusting for multiple confounding factors. High BMI was an independent risk factor for eGFR reduction (p < 0.05) in men, while other obesity indicators or dyslipidemia were not independently associated with the eGFR level overall or speci cally in men or women.

Multivariate Regression Analysis between eGFR and the Combined Effects of Dyslipidemia and Obesity
Multivariate linear regression was performed to determine the association between isolated and combined effects of obesity and dyslipidemia and eGFR (Table 5). Similarly, all the regression coe cients were adjusted for age, sex, smoking status, drinking status, tea consumption, and hypertension, and the subgroup with neither obesity nor dyslipidemia was regarded as the reference. In the general population, none of the isolated dyslipidemia or isolated high obesity indicators showed a signi cant association with the variation in eGFR. However, the combined effect of dyslipidemia and high WHR is an independent risk factor for eGFR reduction (p = 0.033). Notes: All the listed regression coe cients are adjusted for age, gender, current smoking, current drinking, tea consumption, and hypertension.
Hypertension is de ned according to the current blood pressure measurement or current administration of antihypertensive medications. Dyslipidemia is de ned according to the current lipid levels or current administration of lipid lowering medications. High BMI was de ned as BMI ≥ 28 kg/m 2 . High WHR was de ned as WHR ≥ 0.88 for males, and WHR ≥ 0.86 for females. VFA ≥ 80 cm 2 was de ned as high, and SFA ≥ 75th percentile was considered high.
Compared with female subjects, male subjects showed a more profound association. High BMI was an independent risk factor for decreased eGFR in men, both isolated (p = 0.004) and combined with dyslipidemia (p = 0.016). Additionally, double positivity for dyslipidemia and high VFA was also negatively associated with the eGFR in men (p = 0.039). However, none of the regression coe cients showed signi cance in women.

Discussion
To our best knowledge, this study is the rst concerning the combined effects of dyslipidemia and obesity parameters in relation to eGFR variation. Our study not only con rms some of the previous results but also supplements them. High BMI is an independent risk factor for eGFR reduction, and eGFR declines more signi cantly under the combined effect of high BMI and dyslipidemia, particularly in men. Regarding other obesity indicators (such as WHR, VFA, and SFA), we found no signi cant independent association with eGFR. However, several of the abdominal adiposity indicators can be regarded as risk factors for eGFR reduction if superimposed with dyslipidemia, overall or in men. This study provides a reference value for clinical practice. For patients with both dyslipidemia and high obesity indicators, glomerular dysfunction should be monitored for timely prevention or intervention.
Although previous studies have provided numerous conclusions regarding cardiometabolic disorders and CKD or eGFR reduction, additive interactions were not often estimated. Additionally, we could only conclude the negative impact of a fuzzy cluster of metabolic parameters on eGFR considering the de nition of MetS [18], rather than obtaining the impact of a speci c combination of several parameters. Previous evidence is of limited help in clinical practice. Our study provides a novel nding that, although dyslipidemia is not an independent risk factor for decreased eGFR, the eGFR of dyslipidemic men with high BMI is also signi cantly lower than that of double-negative men, and double positivity for dyslipidemia and high BMI is also an independent risk factor for decreased eGFR in men. Furthermore, the double-positive group for high WHR and dyslipidemia showed a signi cantly lower eGFR in each sex group, and double positivity was also a signi cant risk factor for reduced eGFR in the general population.
Additionally, the level of eGFR in the double-positive group for high VFA and dyslipidemia were signi cantly lower in both sexes. Double positivity for dyslipidemia and high VFA was also a signi cant risk factor for reduced eGFR, particularly in men.
Recently, with the continued increase in longevity and lifestyle changes in China, the prevalence of various metabolic abnormalities remains high [19][20][21][22]. The burden of CKD and end-stage renal disease (ESRD) caused by this condition is a concern. According to previous evidence, several cardiometabolic abnormalities are independent risk factors for CKD or eGFR reduction. Considering the previous evidence, a potential impact of the combined effect of dyslipidemia and various obesity phenotypes may exist. Based on most of the previous conclusions, high BMI or obesity may be an independent risk factor for decreased eGFR or CKD. For example, a recent cohort study involving 15,229 middle-aged Chinese individuals revealed that overweight/obesity (BMI ≥ 24 kg/m 2 ) is independently associated with the risk of CKD [23]. Additionally, several studies found a negative relationship between BMI per se and eGFR values, a nding similar to the present results [24][25][26]. However, some scholars have different opinions. Ji et al. investigated the CKD prevalence and related risk factors in Qingdao, China. The study revealed that neither general nor central obesity shows a signi cant association with the risk of reduced eGFR [27]. In another Taiwanese cross-sectional study, the researchers divided 14,983 subjects into two groupsmetabolically healthy and unhealthy groups. The CKD risk did not increase signi cantly with increasing weight in the metabolically healthy group, while the CKD risk showed an upward trend with weight in the other group. The study suggested that BMI per se is not an independent risk factor for CKD [28]. In our study, the signi cant association between BMI per se and eGFR con rms some of the previous evidence.
Based on the previous studies, we further obtained the subjects' WHR, VFA and SFA. The above three adiposity indicators per se did not affect eGFR. However, under the combined effect with dyslipidemia, eGFR showed signi cant differences and a signi cant decrease between different subgroups, and several of the combined effects could be considered independent risk factors for eGFR decline. The above conclusion is a supplement to the previous evidence.
Regarding the impact of dyslipidemia on eGFR or CKD risk, the current evidence remains inconsistent.

Duan et al. investigated the prevalence and associated factors of CKD and reduced eGFR in Henan
Province and found that dyslipidemia is not an independent risk factor for eGFR < 60 ml/min per 1.73 m 2 [29]. However, two-sample Mendelian randomization studies found that reduced serum HDL-c and elevated TGs lead to signi cant worsening of the eGFR or CKD risk [30,31]. A meta-analysis covering 47 trials also found that a common lipid-lowering medication, statins, signi cantly delays the decline of eGFR in patients with renal disease [32]. The present study also explored the potential relationship between serum lipids and eGFR from one aspect. Although some positive conclusions on the relationship between dyslipidemia and eGFR were reported, our study found no signi cant association between dyslipidemia per se and eGFR decline. The difference between the previous and present results may be due to sampling bias because we excluded subjects with an eGFR of less than 60 ml/min per 1.73 m 2 at the beginning of the study. In the present study, only when dyslipidemia and several obesity indicators were superimposed did eGFR show a signi cant decrease or association with the double-positive group.
Overall, the combined effects of dyslipidemia and obesity indicators also show certain differences between different sexes. When we compare the trend of eGFR decline among subjects of different genders, there are certain differences in the results. When we conducted multivariate regression analysis in different sexes, statistical signi cance was only found in men. From the present results, the combined effects of dyslipidemia and obesity indicators are closely related to the eGFR reduction in men, and more restudies are warranted to con rm whether the above conclusion is valid in female subjects. Previous similar studies have also found that men who are obese or with other metabolic abnormalities may be more likely to have renal dysfunction. In a cross-sectional survey by Xiao et al., researchers found that another novel lipid accumulation index, visceral adiposity index (VAI), is more likely to be associated with CKD, especially in men [33]. Another meta-analysis suggested that compared with women, hypertensive men are generally at higher risk of CKD or ESRD. This disparity is unlikely to be explained by biological differences alone [34]. We speculate that male subjects have signi cantly worse lifestyles in the present study, and the prevalence of underlying diseases is also signi cantly higher, such as hypertension, dyslipidemia, and obesity. The above factors may be related to the signi cantly lower baseline eGFR in men, and the signi cant discrepancies in the abovementioned baseline indicators might also contribute to the sex difference of the present results.
This study has several limitations. First, this is a cross-sectional study, so the causal relationship between metabolic indicators and eGFR has not been con rmed. Second, eGFR is a variable determined by multiple factors, and its value is determined by the genetic background or environmental factors. It is somewhat one-sided to analyze the variation in eGFR only from the perspectives of dyslipidemia and obesity. Finally, the sampling scope of this study is relatively limited, only including the middle-aged urban population. Therefore, the present conclusions must be supported by more large-sample studies.

Conclusions
BMI is independently and negatively associated with a decrease in eGFR in a middle-aged Chinese population. Most importantly, the combined effects of dyslipidemia and obesity indices such as BMI and VFA are synergistically associated with the risk of eGFR reduction in men, and the combination of dyslipidemia and WHR is also associated with eGFR reduction in the general population. Accordingly, a better understanding of the combined effects of these modi able risk factors can help promote primary prevention in susceptible subgroups.

Declarations Ethics Approval and Consent to Participate
The study was approved by the Ethics Committee of the First Hospital of China Medical University. The study was performed in accordance with the principles of the Declaration of Helsinki. All the participants signed an informed consent form.

Consent for Publication
Not applicable.

Availability of Data and Materials
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request (Yaxin Lai, Email: laiyaxin811005@126.com).

Con icts of Interests
All authors declare that there is no con ict of interest regarding the publication of this paper.