Metabolic Syndrome, and Not Obesity, Is Associated with Chronic Kidney Disease

Chronic kidney disease (CKD) is a common, progressive condition with a clinical spectrum ranging from mild disease to an end-stage, debilitating state [1]. Its prognostic impact is not only related to the risk of needing renal replacement therapy or renal transplantation but also related to its ability to predict future occurrence of cardiovascular morbidity and mortality [2]. Clinically significant CKD is defined by an increased urinary albumin to creatinine ratio (UACR) and/or an estimated glomerular filtration rate (eGFR) lower than 60 mL/min/1.73 m², as patients with these features are at much higher risk of disease progression [3]. Its prevalence is increasing worldwide together with the growing prevalence of obesity and related cardio-metabolic abnormalities, including hypertension, type 2 diabetes, insulin resistance and their various combinations that are synthesized by the metabolic syndrome (MS) [4]. While metabolic derangements are well-known risk factors for the incidence of CKD, it is still debated whether obesity per se has an influence on kidney function and whether obese (OB) individuals without MS should be considered at higher risk compared with their non-OB counterparts. Previous studies that focused on this research question were conducted mainly in Asian populations and gave conflicting results, with some studies identifying an independent role of body mass index (BMI)-defined obesity on kidney function decline [5‒8], and others a neutral role after accounting for the different MS components [9‒12]. Moreover, most of them did not include an altered UACR as part of the primary endpoint. The present study, based on the 2005–2016 cycles of the National Health and Nutrition Examination Survey (NHANES), was therefore conceived to investigate the contribution of obesity and the different MS components on the prevalence of both reduced estimated glomerular filtration rate (eGFR) and increased UACR in the general US population.

For the present study, we analyzed data from 2005 to 2016 cycles of NHANES, which is conducted in the USA by the National Center for Health Statistics (NCHS) of the Centers for Disease Control and Prevention. NHANES is a cross-sectional survey program aimed to represent the general, noninstitutionalized US population of all ages. It uses a stratified, multistage, clustered probability sampling design with oversampling of several minorities, including non-Hispanic black, Hispanic and Asian persons, people with low income, and older adults. The survey consists of a structured interview conducted in the participants’ home, followed by a standardized physical examination and laboratory tests performed at a mobile examination center (MEC). Full methodology of data collection is available elsewhere [13]. The original survey was approved by the Centers for Disease Control and Prevention Research Ethics Review Board and written informed consent was obtained from all adult participants. The present analysis was deemed exempt by the Institutional Review Board at our institution, as the dataset used in the analysis was completely de-identified.

Body measurements, including height (cm), weight (kg), and waist circumference (cm), were ascertained during the MEC visit; BMI was calculated as weight in kilograms divided by height in meters squared and obesity was defined as a BMI ≥30 kg/m². Blood pressure was measured 3 times with a mercury sphygmomanometer by certified physicians. We used the mean of the 3 measurement as the representative value for the current study. Diabetes was defined in accordance with the American Diabetes Association criteria if any of the following conditions were met: (1) A self-reported diagnosis of diabetes; (2) use of antidiabetic drugs; (3) a hemoglobin A1c level ≥6.5% (48 mmol/mol); (4) a fasting plasma glucose ≥126 mg/dL; and (5) a random plasma glucose ≥200 mg/dL [14]. Laboratory methods for measurements of serum creatinine, UACR, hemoglobin A1c, lipid profile, alanine aminotransferase, aspartate aminotransferase, γ-glutamyltranspeptidase, insulin, and creatinine are reported in detail elsewhere [15].

The homeostatic model assessment of insulin resistance was computed as originally described [16]. MS was diagnosed according to the National Cholesterol Education Program Adult Treatment Panel criteria [17] if at least 3 of the following criteria were met: (1) fasting plasma glucose ≥100 mg/dL or drug treatment for elevated blood glucose; (2) high-density lipoprotein (HDL) cholesterol <40 mg/dL in men and <50 mg/dL in women or drug treatment for low HDL cholesterol; (3) triglycerides ≥150 mg/dL or drug treatment for elevated triglycerides; (4) waist circumference ≥102 cm in men and ≥88 cm in women; and (5) blood pressure values ≥130/85 mm Hg or drug treatment for hypertension.

Based on the presence of obesity and MS, patients were divided into 4 groups: non-OB individuals without MS (OB− MS−, reference group), non-OB individuals with MS (OB− MS+), OB individuals without MS (OB+ MS−), and OB individuals with MS (OB+ MS+). eGFR was computed according to the CKD-Epidemiology Collaboration equation and was considered to be reduced if <60 mL/min/1.73 m². UACR was considered abnormal if ≥30 mg/g.

32,920 participants aged ≥20 years attended an MEC visit (Fig. 1). We first excluded 18,158 individuals that were not assigned to a morning session because of unavailable fasting biochemical parameters. Furthermore, of the remaining 14,762 participants, 2,407 were excluded due to unavailable data on one or more of the MS components, yielding a final sample of 12,355 individuals for the current analysis.

All analyses were conducted using Stata version 13.0 (StataCorp, College Station, TX, USA), accounting for the complex survey design of NHANES. We used appropriate weighting for each analysis, as suggested by the NCHS. Data are expressed as weighted proportions ± standard error for categorical variables and as weighted means ± standard error for continuous variables. Participants’ characteristics according to the presence of obesity and MS were compared using linear regression for continuous variables and the design-adjusted Rao-Scott χ² test for categorical variables. Post hoc Wald test was used to compare specific groups. Multivariable logistic regression analysis was performed in order to evaluate the effect of obesity and each of the MS components on the presence of reduced eGFR, increased UACR, or a combination of the 2. Age, sex, and race ethnicity were included in the model as covariates. Trends in the proportion of US adults with MS and obesity across calendar periods were assessed by logistic regression analysis, modeling the 2 conditions as dependent variables and the 2-year calendar periods as a continuous independent variable. These models were adjusted for age, race ethnicity, and sex. A 2-tailed value of p < 0.05 was considered statistically significant.

Of the 12,355 participants, 6,039 were OB− MS−, 1,853 were OB− MS+ and 2,859 were OB+ MS+. There were a total of 1604 OB+ MS− participants, representing 13.3% (95% confidence interval (CI): 12.4–14.3) of all participants and 37.3% (95% CI: 35.1–39.6) of the OB population. The clinical and metabolic features of the 4 groups are shown in Table 1. Age differed among all the groups; in particular, it was significantly lower in participants without MS than their MS+ counterparts. OB+ MS− showed the highest proportion of women (57%) and of Hispanic and non-Hispanic black individuals. BMI and waist circumference were significantly higher in the OB−MS+ group than the OB− MS− and in the OB+ MS+ than the OB+ MS−. A similar finding was present for the degree of insulin resistance estimated through the HOMA-IR. As expected, based on the MS diagnostic criteria, OB and non-OB participants with MS showed higher systolic blood pressure, diastolic blood pressure and triglyceride levels, and lower HDL levels than their counterparts without MS. This was paralleled by a similar trend in aspartate aminotransferase, alanine aminotransferase, and γ-glutamyltranspeptidase.

Diabetes prevalence did not differ significantly between OB− MS− and OB+ MS− participants, but was significantly higher in the MS+ groups. This was also the case for the prevalence of reduced eGFR and increased UACR.

The proportion of US adults with MS increased over time, from 31.3% (95% CI: 27.4–35.4) in 2005–2006 to 40.2% (95% CI: 36.6–44.0) in 2015‒2016 (p_trend = 0.006). Similarly, the proportion of participants with obesity increased from 33.6% (95% CI: 29.9–37.5) in 2005‒2006 to 40.3% (95% CI: 36.5–44.3) in 2015‒2016 (p_trend = 0.002).

To dissect out the relative contribution of obesity and MS on the prevalence of reduced eGFR, increased UACR, or their combination, we run 2 separate logistic regression models. Both were adjusted for age, sex, and ethnicity. The first model divided participants in the aforementioned 4 groups using the OB− MS− as the reference. Results of this analysis are shown in Figure 2. The odds of all endpoints were not increased for OB+ MS− participants, whereas both OB− MS+ and OB+ MS+ showed a higher prevalence of all endpoints, with OB+ MS+ having the highest odds ratios (ORs). Moreover, compared with non-Hispanic white participants, non-Hispanic blacks were at higher risk of all renal endpoints, while Hispanics were at higher risk for increased UACR and the composite outcome, but not for reduced eGFR. Finally, women were at higher risk of increased UACR than men.

The second approach was based on the evaluation of each of the MS components. As shown in Figure 3, elevated BP and low HDL levels were independently associated with all endpoints, while elevated blood glucose concentrations and triglycerides were only associated with increased UACR and the composite endpoint. On the other hand, waist circumference was not associated with any endpoint. Finally, diagnosed diabetes was significantly associated with both reduced eGFR (OR 1.99, 95% CI: 1.60–2.48, p < 0.001) and increased UACR (OR 3.46, 95% CI: 2.85–4.21, p < 0.001) in the studied population, after adjustment for age, sex, and race ethnicity.

In this large cross-sectional study of US adults from the general population, obesity per se was not independently associated with the presence of reduced eGFR and/or increased UACR. This was suggested by 2 complementary analyses: (1) OB+ MS− participants did not show an increased prevalence of either reduced eGFR or increased UACR compared with OB− MS−subjects and (2) after accounting for age, sex, ethnicity, and the other MS criteria, increased waist circumference was not independently associated with any considered kidney endpoint.

These results suggest that increased body fat alone is not an important determinant of CKD independently of well-known risk factors, such as diabetes, hypertension, or other proposed mechanisms in a healthy population. It is also conceivable that our finding of the lack of a statistically significant association between obesity and reduced eGFR may in part reflect the fact that some participants with obesity might be in a transient state of hyperfiltration (in which GFR increases due to nephron overload [18]), which could precede a subsequent progressive decline in eGFR. On the other hand, MS was independently associated with both endpoints in OB and non-OB individuals. In particular, the BP and HDL criteria were associated with both endpoints, while elevated blood glucose and triglycerides were only linked with increased UACR.

Several recent analyses using a cross-sectional or longitudinal study design have examined differences in the risk of CKD between OB people with and without MS, who were usually labeled “metabolically unhealthy obesity” and “metabolically healthy obesity,” respectively [5‒8, 11]. Results were discordant, as some did not find any increased risk of CKD in subjects with metabolically healthy obesity compared with metabolically healthy non-OB people, while others found that a residual increased risk of CKD remains. A possible explanation of these inconsistencies is related to different definitions of the primary outcome, as many studies did not include albuminuria as an indicator. Another possible source of heterogeneity is related to the specific criteria for defining MS as some studies applied the NCEP-ATP-III criteria, while others used the IDF criteria. Our results confirm previous reports from the NHANES III cycle (1988‒1994), which consistently showed a detrimental effect of MS on kidney outcomes, including both altered UACR and low eGFR [19, 20]; it should be noted that prevalence of both MS and obesity increased significantly from these previous studies and that this trend continued in the following years, with higher rates in 2015‒2016 than 2005‒2006, raising concerns for the future incidence of CKD.

Concerning the role of the different components of MS on CKD, it is well-known that diabetes and hypertension represent the most common drivers of reduced kidney function in the general population. Moreover, triglyceride levels are known to be associated with an increased risk of CKD [21, 22], and may contribute to kidney dysfunction through their pro-inflammatory and atherogenic effects or acting as a marker for insulin resistance [23, 24]. Finally, several epidemiological studies found an association between low HDL-C and poor kidney function or progression of CKD [25, 26]. A large study by Bowe et al. [25] performed in a cohort of almost 2 million male veterans from the USA followed for a median of 9 years, showed that individuals with HDL-C concentrations below 30 mg/dL had a 10–20% higher risk for CKD and/or progression of CKD than individuals with concentration above 40 mg/dL. Our results confirm these findings in a healthier population and in both sexes.

The major strengths of the present study are the following: (1) this is a population-based study aimed at being representative of the US population and it includes both sexes and different ethnic origins; (2) acquisition of the anthropometric parameters of interest was homogenous; (3) there were a high number of subjects in each group, leading to a high statistical power. The present study also has some limitations that need to be acknowledged: (1) its cross-sectional nature does not allow causal inference and the evaluation of incidence of disease; (2) eGFR was calculated based on the CKD-Epidemiology Collaboration equation and not by more accurate techniques, such as creatinine clearance or the gold standard inulin clearance; nevertheless, equations based on serum creatinine are well suited for large epidemiological studies; (3) UACR was measured once, while repeated testing might be necessary to exclude spurious results.

This large, cross-sectional, population-based study shows that MS, and not obesity per se, is associated with kidney disease in US adults. In particular, OB patients without MS do not seem to have an increased prevalence of reduced eGFR or increased UACR compared with non-OB individuals. Moreover, high blood pressure and low HDL levels are associated with both renal endpoints, while high blood glucose and increased triglycerides are only associated with increased UACR.

NHANES was approved by the National Center for Health Statistics (NCHS) Institutional Board, and written informed consent was obtained from adult participants. The present analysis was deemed exempt by the Review Board at our institution, as the dataset used in the analysis was completely de-identified.

All authors have no conflicts of interest to declare.

The authors received no specific funding for the present study.

S.C. and G.P. designed the study, wrote, and edited the manuscript, S.C. researched and analyzed data. C.B. and R.T. wrote and reviewed the manuscript. All the authors approved the final version of the manuscript to be published. G.P. is the guarantor of this work.

All data used in the current analysis are publicly available through the National Center for Health Statistics and can be accessed at: https://wwwn.cdc.gov/nchs/nhanes/default.aspx.