Serum uromodulin is inversely associated with the metabolic syndrome in the KORA F4 study.

OBJECTIVE: Metabolic syndrome and obesity are risk factors for chronic kidney disease. However, early kidney alterations may escape diagnosis in these conditions due to glomerular hyperfiltration. Uromodulin, a glycoprotein exclusively synthesized in tubular cells of the thick ascending limb of Henle's loop, is a novel tissue-specific biomarker for kidney function. In contrast to the commonly used markers creatinine and cystatin C, serum uromodulin does not primarily depend on glomerular filtration. We hypothesized that serum uromodulin is a marker for metabolic syndrome and related components.
DESIGN: The analyses included 1088 participants of the population-based KORA F4 study aged 62-81 years. Metabolic syndrome was present in 554 participants. After a mean follow-up time of 6.5 years, 621 participants were reevaluated, of which 92 had developed incident metabolic syndrome.
METHODS: The association of serum uromodulin with metabolic syndrome and its components were assessed using multivariable logistic regression models.
RESULTS: Serum uromodulin was inversely associated with metabolic syndrome after adjustment for sex, age, estimated glomerular filtration rate, physical activity, smoking, alcohol consumption and high-sensitivity C-reactive protein (OR 0.65; 95% CI 0.56-0.76 per standard deviation uromodulin; P < 0.001). Serum uromodulin was inversely associated with all single components of metabolic syndrome. However, serum uromodulin was not associated with new-onset metabolic syndrome after the follow-up period of 6.5 ± 0.3 years (OR 1.18; 95% CI 0.86-1.60).
CONCLUSIONS: Serum uromodulin is independently associated with prevalent, but not with incident metabolic syndrome. Low serum uromodulin may indicate a decreased renal reserve in the metabolic syndrome.

Introduction

The metabolic syndrome is associated with an increased risk for cardiovascular and renal complications, and chronic renal failure in metabolic syndrome may progress to end-stage renal disease (1, 2). Kidney pathophysiological alterations reveal a type of ischemic nephropathy with predominant tubular atrophy in addition to interstitial fibrosis, microvascular sclerosis and segmental glomerulosclerosis (3, 4). However, kidney alterations in metabolic syndrome may escape early clinical diagnosis, since obesity and other components of the metabolic syndrome, such as elevated fasting glucose and elevated blood pressure, are associated with glomerular hyperfiltration. The currently available kidney function markers (creatinine, cystatin C and urea) all reflect the glomerular filtration rate (GFR) and may thus deliver falsely inconspicuous results in case of glomerular hyperfiltration. Additionally, obesity may be associated with variable degrees of skeletal muscle loss, including sarcopenic obesity, which is especially frequent in elderly individuals and can further influence the estimated GFR (eGFR) values based on creatinine (5). Therefore, additional kidney function markers not directly depending on glomerular filtration would be useful to unravel early kidney disease in elderly individuals with metabolic syndrome.

The currently most promising alternate marker of kidney function is uromodulin. Uromodulin is a tissue-specific glycoprotein synthesized in tubular cells of the thick ascending limb of Henle's loop. The larger proportion of uromodulin is secreted into the urinary tract, where it is the most abundant protein under physiological conditions (50–100 mg/24 h) and exerts anti-infective, anti-lithogen and immunomodulatory functions (6, 7, 8, 9, 10). In addition to the apical secretion, a small amount of uromodulin is secreted basolaterally into the interstitial space and transferred into the blood stream (11, 12, 13). The physiological function of serum uromodulin (sUmod) is elusive to date, but sUmod emerged as a highly interesting marker of kidney function (12, 13). As a tubular protein, sUmod does not directly depend on glomerular filtration, but mirrors tubular integrity and indirectly nephron mass and thus the renal reserve (14).

In the current study, we used sUmod as a novel marker of kidney integrity in elderly participants of the population-based KORA F4 cohort. We hypothesized that sUmod was inversely associated with metabolic syndrome independently of the eGFR, indicating a decreased renal tubular reserve in metabolic syndrome, obesity and related conditions.

Materials and methods

Study participants

The KORA (Cooperative Health Research in the Region of Augsburg) F4 (2006-2008) and FF4 (2013-2014) cohort studies are follow-up examinations of the population-based KORA S4 study (1999-2001). Recruitment and eligibility criteria for the KORA studies have been described previously (15). The study design, standardized sampling methods and data collection (medical history, medication, anthropometric measurements, blood pressure) have been described in detail elsewhere (16, 17). All study participants gave written informed consent. The study was approved by the Ethics Committees of the Bavarian Medical Association in adherence to the Declaration of Helsinki. sUmod was measured in the 1119 participants aged 62–81 years of the KORA F4 study with available serum samples (from a total of 1161 participants in this age group). All variables necessary for the cross-sectional analyses were available in 1088 participants. Of these 1088 participants, 116 died and 341 declined participation in the FF4 survey or could not be contacted. Thus, the study sample in the longitudinal F4/FF4 examination comprised 631 participants, of which ten were excluded due to missing covariables, leaving 621 participants for the longitudinal analyses (Supplementary Fig. 1, see section on supplementary data given at the end of this article). The follow-up time was 6.5 ± 0.3 years. Clinical characteristics of the participants included in the follow-up analysis are available in Supplementary Table 1. Metabolic syndrome was defined according to the International Diabetes Federation definition as presence of at least three of the following five criteria: (1) elevated waist circumference (waist circumference ≥94 cm in men and ≥80 cm in women); (2) fasting triglycerides ≥1.7 mmol/L and/or use of fibrates or nicotinic acid; (3) high-density lipoprotein (HDL) cholesterol <1.0 mmol/L in men or <1.3 mmol/L in women and/or use of fibrates or nicotinic acid; (4) fasting glucose ≥5.6 mmol/l and/or use of glucose-lowering medication; (5) elevated blood pressure (systolic blood pressure ≥130 mmHg and/or diastolic blood pressure ≥85 mmHg and/or use of antihypertensive medication, given that the participants were aware of being hypertensive) (18). No participant took omega 3 fatty acid in a dose ≥2 g/day. Leisure time physical activity was assessed with two separate questions concerning leisure time sport activity in winter and in summer (cycling included). Possible answers were (1) >2 h, (2) 1–2 h, (3) <1 h and (4) none per week. Participants who had a total score <5, obtained by summing the numbers (1)–(4) relating to winter and summer, were classified to be ‘physically active’. Alcohol consumption was categorized in three groups defined as no (0 g/day), moderate (men 0.1–39.9 g/day and women 0.1–19.9 g/day), and high (men ≥40 g/day and women ≥20 g/day) alcohol consumption. Smoking behavior was categorized into three groups (active smoker, former smoker, never smoker).

Laboratory measurements

Blood was collected after an overnight fast of at least 8 h. The samples were kept at room temperature until centrifugation. Plasma and serum samples were assayed immediately or stored at −80°C. Measurements of serum creatinine (fresh serum), high-sensitivity C-reactive protein (hsCRP) (frozen plasma), glucose (fresh serum), total cholesterol, low-density lipoprotein (LDL), HDL and triglycerides (all fresh serum) were performed as described previously (19, 20). sUmod (frozen serum) was measured with a commercially available enzyme-linked immunosorbent assay kit (Euroimmun AG, Lübeck, Germany) with a lower detection limit of 2 ng/mL, an intra-assay coefficient of variation of 2.3% and inter-assay coefficients of variation of 4.4% and 9.5% for sUmod target values of 24.9 and 142.2 ng/mL, respectively. The measurement procedure was performed as described by Steubl et al. (12). Estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation (2009) based on serum creatinine (21).

Statistical analyses

Characteristics of the study participants were compared between participants with and without metabolic syndrome using t-tests in case of approximately normally distributed variables. Mann–Whitney U tests were performed for variables with skewed distributions. Binomial proportions were compared with chi-square tests. The associations of sUmod with the outcomes of interest were assessed in logistic regression models in case of categorical dependent variables and in linear regression models in case of continuous dependent variables. sUmod was analyzed as log-transformed continuous independent variable per standard deviation. In multinomial logistic and linear regression analyses, the associations of sUmod with the respective dependent variables were adjusted for covariates in different models: Model 1: age (continuous) and sex; Model 2: Model 1 plus eGFR (log-transformed, continuous); Model 3: Model 2 plus physical activity (active/inactive), smoking status (never/former/current), and alcohol consumption (no/moderate/high); Model 4: Model 3 plus hsCRP (log-transformed, continuous); Model 5: Model 4 plus arterial hypertension (yes/no) and fasting glucose (continuous); Model 6: Model 5 plus BMI (continuous) and treatment with statins and fibrates (yes/no, respectively). Model 1 and Model 2 were calculated for all outcome variables; Model 3 and Model 4 for the metabolic syndrome and its components; Model 5 for anthropometric measures and Model 6 for lipid parameters as outcome variables. Preexisting cases were excluded from the longitudinal incidence analyses. The level of statistical significance was set at 5% (two-sided). All calculations were performed using the statistical environment R, version 3.5.2.

Results

Study population characteristics

Table 1 displays the baseline characteristics of the total study population and stratified by metabolic syndrome. Metabolic syndrome was present in 43% of the participating women and in 56% of the men, corresponding to 51% of the total cohort. Participants with metabolic syndrome had significantly lower sUmod concentrations compared to participants without metabolic syndrome (P < 0.001).

Table 1

Characteristics of the study participants.a

	Total study cohort	No metabolic syndrome	Metabolic syndrome	P value
n	1088	534	554
Female sex, n (%)	537 (49)	301 (56)	236 (43)	<0.001k
Age (years)	70.3 ± 5.5	70.1 ± 5.6	70.5 ± 5.4	0.32i
Elevated waist circumference n (%)b	913 (84)	375 (70)	538 (97)	<0.001k
Elevated triglycerides n (%)c	308 (28)	31 (6)	277 (50)	<0.001k
Reduced HDL cholesterol n (%)d	206 (19)	17 (3)	189 (34)	<0.001k
Elevated fasting glucose n (%)e	528 (49)	80 (15)	448 (81)	<0.001k
Elevated blood pressure n (%)f	797 (73)	288 (54)	509 (92)	<0.001k
BMI (kg/m2)	28.7 ± 4.3	26.9 ± 3.9	30.4 ± 4.3	<0.001i
Waist circumference (cm)	98.1 ± 12.1	92.5 ± 10.9	103.4 ± 10.6	<0.001i
Waist-to-hip ratio	0.91 ± 0.08	0.88 ± 0.08	0.94 ± 0.07	<0.001i
Total cholesterol (mmo/L)	5.77 ± 1.04	5.85 ± 1.04	5.69 ± 1.07	0.007i
HDL cholesterol (mmol/L)	1.45 ± 0.37	1.61 ± 0.34	1.30 ± 0.31	<0.001i
Ratio total cholesterol/HDL cholesterol	4.18 ± 1.13	3.75 ± 0.87	4.59 ± 1.19	<0.001i
LDL cholesterol (mmol/L)	3.65 ± 0.93	3.67 ± 0.94	3.61 ± 0.93	0.24i
Triglycerides (mmol/L)	1.28 (0.93; 1.78)	1.06 (0.82; 1.35)	1.70 (1.20; 2.33)	<0.001j
hsCRP (mg/dL)	1.52 (0.78; 3.11)	1.29 (0.64; 2.48)	1.98 (0.98; 3.82)	<0.001j
eGFR (mL/min/1.73 m2)	77.9 (67.3; 87.7)	80.1 (67.3; 87.8)	75.4 (64.2; 86.2)	<0.001j
Physically inactive n (%)g	537 (49)	236 (44)	301 (54)	0.001k
Smoker never/former/current (%)	49/45/6	51/41/7	47/48/5	0.07k
Alcohol consumption no/moderate/high (%)h	32/56/12	30/59/11	33/54/13	0.23k
Serum uromodulin (ng/mL)	153 (111; 208)	176 (130; 227)	132 (99; 186)	<0.001j

aMean ± standard deviation, median (first quartile; third quartile), or number of participants (proportion in %). bDefined as ≥80 cm in women and ≥94 cm in men. cDefined as ≥1.7 mmol/L and/or intake of fibrates or nicotinic acid. dDefined as <1.0 mmol/L in men and <1.3 mmol/L in women and/or intake of fibrates or nicotinic acid. eDefined as ≥5.6 mmol/L and/or intake of anti-diabetic medication. fDefined as systolic blood pressure ≥130 mmHg and/or diastolic blood pressure ≥85 mmHg and/or use of antihypertensive medication, given that the participants were aware of being hypertensive. gPhysically inactive: <1-h sports/week in winter and summer. hAlcohol consumption: no (0 g/day), moderate (men 0.1–39.9 g/day and women 0.1–19.9 g/day), high (men ≥40 g/day and women ≥20 g/day). iT-test; jMann–Whitney U test; kchi-square test.

Inverse association of sUmod with the metabolic syndrome

sUmod displayed an inverse association with the metabolic syndrome (odds ratio (OR) 0.58; 95% confidence interval (CI) 0.50–0.67) after adjustment for sex and age (Table 2). Additional correction for eGFR modestly attenuated the association (OR 0.61; 95% CI 0.53–0.70). Further adjustment for physical activity, smoking and alcohol consumption had hardly any influence (OR 0.62; 95% CI 0.53–0.72) and additional adjustment for hsCRP did not substantially attenuate the associations of sUmod with the metabolic syndrome (OR 0.65; 95% CI 0.56–0.76).

Table 2

Odds ratios (95% confidence interval) for metabolic syndrome and single components of the metabolic syndrome as dependent variables and sUmod as independent variable (per standard deviation): results of logistic regression models.

Metabolic syndrome (yes: n = 554; no: n = 534)	Elevated waist circumferencea (yes: n = 913; no: n = 175)	Elevated triglyceridesb (yes: n = 308; no: n = 780)	Reduced HDL cholesterolc (yes: n = 206; no: n = 882)	Elevated fasting glucosed (yes: n = 528; no: n = 560)	Elevated blood pressuree (yes: n = 797; no: n = 291)
Adjustment for sex and age
0.58 (0.50–0.67)f	0.71 (0.59–0.86)f	0.69 (0.60–0.80)f	0.70 (0.60–0.81)f	0.69 (0.60–0.79)f	0.64 (0.54–0.75)f
Adjustment for sex, age and eGFR
0.61 (0.53–0.70)f	0.71 (0.59–0.87)f	0.74 (0.64–0.86)f	0.76 (0.64–0.90)f	0.71 (0.62–0.81)f	0.67 (0.56–0.79)f
Adjustment for sex, age, eGFR, physical activity, smoking and alcohol consumption
0.62 (0.53–0.72)f	0.73 (0.60–0.89)g	0.75 (0.64–0.87)f	0.78 (0.66–0.93)g	0.72 (0.62–0.82)f	0.67 (0.56–0.79)f
Adjustment for sex, age, eGFR, physical activity, smoking, alcohol consumption and hsCRP
0.65 (0.56–0.76)f	0.82 (0.67–1.01)	0.75 (0.65–0.88)f	0.81 (0.68–0.97)h	0.74 (0.64–0.86)f	0.70 (0.59–0.83)f

fP < 0.001; gP < 0.01; hP < 0.05.

aDefined as ≥80 cm in women and ≥94 cm in men. bDefined as ≥1.7 mmol/l and/or intake of fibrates or nicotinic acid. cDefined as <1.0 mmol/l in men and <1.3 mmol/l in women and/or intake of fibrates or nicotinic acid. dDefined as ≥5.6 mmol/l and/or intake of anti-diabetic medication. eDefined as systolic blood pressure ≥130 mmHg and/or diastolic blood pressure ≥85 mmHg and/or use of antihypertensive medication, given that the participants were aware of being hypertensive.

Inverse association of sUmod with single components of the metabolic syndrome

sUmod was inversely associated with each single component of the metabolic syndrome (Table 2). After adjustment for age, sex, eGFR, physical activity, smoking and alcohol consumption, sUmod was significantly inversely associated with an elevated waist circumference, elevated triglycerides, reduced HDL cholesterol, elevated fasting glucose and elevated blood pressure. As for the metabolic syndrome itself, the major influencing factor for the associations of sUmod with the single metabolic syndrome components was eGFR, except for the association of sUmod with elevated waist circumference, which was hardly influenced by adjustment for eGFR. Additional adjustment for hsCRP attenuated most of the associations of sUmod with the metabolic syndrome and its components moderately, which still remained significant, except for the association of sUmod with elevated waist circumference that lost significance.

In addition to waist circumference, BMI and waist-to-hip ratio were also inversely associated with sUmod after adjustment for sex and age (P < 0.001; Table 3). The association of sUmod with BMI remained significant in the multivariable regression models (P < 0.01), whereas its association with the waist-to-hip ratio lost significance after multivariable adjustment.

Table 3

Association estimates between sUmod and anthropometric outcome variables: β coefficient ± standard error from linear regression models are given per standard deviation sUmod (n = 1088).

BMI (continuous)	Waist circumference (continuous)	Waist-to-hip ratio (continuous)
Adjustment for sex and age
−0.21 ± 0.03a	−0.18 ± 0.02a	−0.10 ± 0.02a
Adjustment for sex, age and eGFR
−0.19 ± 0.03a	−0.17 ± 0.03a	−0.09 ± 0.02a
Adjustment for sex, age, eGFR, physical activity, smoking, alcohol consumption, hsCRP, arterial hypertension, fasting glucose
−0.08 ± 0.03b	−0.08 ± 0.02b	−0.04 ± 0.02

aP < 0.001; bP < 0.01.

Analysis of further serum lipid parameters revealed an association between sUmod and increased total cholesterol, LDL and HDL cholesterol, and the ratio of total cholesterol to HDL cholesterol after adjustment for sex, age and eGFR. However, these associations did not remain significant in the fully adjusted model (Table 4). Thus, among the serum lipid parameters, only two major components of the metabolic syndrome, HDL cholesterol (β: 0.10 ± 0.03) and triglycerides (β: −0.10 ± 0.03), displayed a significant association with sUmod after multivariable adjustment.

Table 4

Association estimates between sUmod and parameters of lipid metabolism: β coefficient ± standard error from linear regression models are given per standard deviation sUmod (n = 1088).

Total cholesterol	LDL cholesterol	HDL cholesterol	Ratio total cholesterol to HDL cholesterol	Triglycerides
Adjustment for sex and age
0.12 ± 0.03a	0.10 ± 0.03b	0.21 ± 0.03a	−0.12 ± 0.03a	−0.18 ± 0.03a
Adjustment for sex, age and eGFR
0.11 ± 0.03b	0.09 ± 0.03c	0.19 ± 0.03a	−0.10 ± 0.03b	−0.16 ± 0.03a
Adjustment for sex, age, eGFR, physical activity, smoking, alcohol consumption, hsCRP, arterial hypertension, fasting glucose, BMI, treatment with statins and fibrates
0.06 ± 0.03	0.04 ± 0.03	0.10 ± 0.03a	−0.54 ± 0.03	−0.10 ± 0.03a

aP < 0.001; bP < 0.01; cP < 0.05.

Lack of an association of sUmod with incident metabolic syndrome in the longitudinal analysis

Baseline sUmod was not associated with new-onset metabolic syndrome or incidence of any component of the metabolic syndrome after the follow-up time of 6.5 ± 0.3 years (Table 5). Furthermore, sUmod was not associated with changes of BMI or waist-to-hip ratio after adjustment for sex, age, eGFR and the respective baseline parameter (data not shown). After multivariable adjustment including sex, age, eGFR, physical activity, smoking, alcohol consumption, hsCRP, arterial hypertension, fasting glucose, BMI, the respective baseline parameter and intake of statins and fibrates in the follow-up examination, there was no association of sUmod with total cholesterol, LDL, HDL, the ratio of total cholesterol to HDL cholesterol and triglycerides in the follow-up examination (data not shown).

Table 5

Odds ratios (95% confidence interval) for incident metabolic syndrome and incident single components of the metabolic syndrome as dependent variables and sUmod as independent variable (per standard deviation): results of logistic regression models (n = 621). The numbers of participants with and without event are given in parentheses.

Metabolic syndrome (yes: n = 92; no: n = 233)	Elevated waist circumferencea (yes: n = 60; no: n = 56)	Elevated triglyceridesb (yes: n = 46; no: n = 409)	Reduced HDL cholesterolc (yes: n = 9; no: n = 508)	Elevated fasting glucosed (yes: n = 121; no: n = 209)	Elevated blood pressuree (yes: n = 62; no: n = 126)
Adjustment for sex and age
1.05 (0.79–1.39)	0.69 (0.45–1.07)	1.05 (0.74–1.50)	0.77 (0.40–1.50)	0.92 (0.71–1.19)	0.99 (0.70–1.40)
Adjustment for age, sex and eGFR
1.13 (0.84–1.52)	0.69 (0.45–1.08)	1.03 (0.72–1.48)	0.79 (0.40–1.56)	0.96 (0.74–1.26)	0.99 (0.70–1.40)
Adjustment for sex, age, eGFR, physical activity, smoking, alcohol consumption and hsCRP
1.18 (0.86–1.60)	0.75 (0.47–1.21)	1.03 (0.71–1.48)	0.68 (0.34–1.36)	1.00 (0.76–1.32)	1.07 (0.74–1.55)

aDefined as ≥80 cm in women and ≥94 cm in men. bDefined as ≥1.7 mmol/L and/or intake of fibrates. bDefined as ≥1.7 mmol/L and/or intake of fibrates. cDefined as <1.0 mmol/L in men and <1.3 mmol/l in women and/or intake of fibrates. dDefined as ≥5.6 mmol/L and/or intake of anti-diabetic medication. eDefined as systolic blood pressure ≥130 mmHg and/or diastolic blood pressure ≥85 mmHg and/or use of antihypertensive medication, given that the participants were aware of being hypertensive. fNumber of participants with/without event.

Discussion

The key finding of the current study is the strong and independent inverse relationship between sUmod and the metabolic syndrome in a large population-based elderly cohort. Low sUmod levels were associated with single components of the metabolic syndrome (elevated triglycerides, reduced HDL cholesterol, elevated fasting glucose and elevated blood pressure), indicating that sUmod is widely connected with metabolic alterations. To our knowledge, a detailed examination of the association of uromodulin with the metabolic syndrome has not been published before.

In the urinary tract, uromodulin exerts immunomodulatory functions (9, 10). Whether immunomodulatory uromodulin properties also play a role in the circulation is not yet known. Nevertheless, uromodulin is capable of binding immunoglobulins and complement and to activate immune cells in experimental settings, indicating the presence of relevant systemic effects (22, 23, 24). Since obesity and metabolic syndrome are associated with low-grade inflammation, immunomodulatory sUmod properties might positively influence metabolism in this condition. This may serve as one explanation of an inverse connection of uromodulin and metabolic disturbances. In our cohort, sUmod was inversely associated with hsCRP (25). However, adjustment for hsCRP only moderately attenuated the association of sUmod with the metabolic syndrome and its components, indicating that chronic subclinical inflammation as assessed by hsCRP is not the only factor explaining the interplay of the metabolic syndrome and sUmod. Only the association of sUmod with elevated waist circumference was more strongly attenuated and lost significance after additional adjustment for hsCRP, possibly reflecting a stronger impact of chronic subclinical inflammation related to increased visceral adipose tissue on this association.

Despite the strong inverse associations in the cross-sectional analysis, there was no evidence for an association between sUmod and the development of metabolic syndrome, nor any of its components. Thus, metabolic factors may negatively influence sUmod levels, indicating that the metabolic syndrome may impair tubular integrity and protein synthesis. Recently, Scheurlen et al. described an increase of sUmod following Roux-en-Y-gastric bypass in patients with severe obesity, indicating improved ‘kidney health’ or renal reserve after bariatric surgery, which was not detectable by traditional kidney function markers (26). We (25) and others (27) previously showed that sUmod was inversely associated with type 2 diabetes independently of the eGFR. Thus, sUmod possibly represents a tissue-specific marker for early diabetic nephropathy that escapes GFR estimates due to diabetes-related glomerular hyperfiltration. The inverse association of sUmod with BMI and waist circumference independently of fasting glucose in the current study indicates that sUmod may also unravel other early kidney alterations related to hyperfiltration independently of hyperglycemia-related hyperfiltration. An inverse relation of sUmod with BMI was also described by Steubl et al. (12) and is possibly due to a relatively lower nephron mass in obesity, which is characteristically not detected by eGFR due to obesity-related hyperfiltration, but represents a risk factor for chronic kidney disease (2). However, in another cohort comprising 200 participants, of which 170 suffered from chronic kidney disease, sUmod was not associated with BMI (28). These discrepancies supposedly originate from differences in the study populations, since in overt preexisting chronic kidney disease, uromodulin values are substantially reduced as a sequelae of tubular atrophy, interstitial fibrosis and an overall loss of nephrons unrelated to metabolic alterations. The discrepancies between the studies emphasize the need to further elucidate the impact of sUmod on human health and disease in the general community.

The strong association of sUmod with HDL and triglycerides is a remarkable finding. In order to determine the independence of the correlation of sUmod with HDL and triglycerides from obesity and other metabolic traits, we performed a thorough adjustment for possible confounders. The associations still remained significant, indicating an intrinsic interaction of sUmod with HDL and triglycerides. In the urinary tract, uromodulin is able to bind various components including leukocytes, fimbriated bacteria and calcium phosphate nanocrystals (6). It is not yet clarified, whether sUmod also has similar properties in the circulation and can bind, modulate, neutralize or eliminate serum components as well. Intrinsic systemic effects of sUmod are very probable. For example, sUmod is inversely associated with incident cardiovascular events independently of the eGFR and well-known traditional cardiovascular risk factors, indicating an intrinsic vasoprotective effect of the glycoprotein that has not yet been specified (29, 30). However, the weak or missing associations of sUmod with total cholesterol and LDL, and the lack of an association of sUmod with lipid levels in the longitudinal analysis argue against a broad implication of sUmod in lipid metabolism.

Study strengths and limitations

Strengths of our study are the population-based design with a large, well-characterized community-based cohort, and the follow-up time of 6.5 years. Unlike most previous studies using urinary uromodulin, we measured sUmod with a sensitive and robust ELISA. In contrast to uromodulin of urine origin, which forms various polymers with changing episodes and different antigenic sites (31), sUmod is a stable antigen lacking such important pre-analytical disadvantages (13). To our knowledge, this is the first study elucidating the association of sUmod with the metabolic syndrome. However, only participants aged 62–81 years were included, so that the relation of sUmod with metabolic disturbances remains to be confirmed in a younger population. Further, the incidence of some single components of the metabolic syndrome, namely elevated triglycerides and reduced HDL cholesterol was relatively low, limiting the power of the study in these longitudinal analyses.

Conclusions

We demonstrate an inverse association of prevalent metabolic syndrome with sUmod in elderly study participants from the general community. Further studies are required to assess whether a decrease of sUmod levels in metabolic syndrome, indicating tubular injury, reflects a reduced renal reserve due to a relatively lower nephron mass in this condition or whether obesity and other risk factors in concert with the metabolic syndrome influence uromodulin biosynthesis.

Declaration of interest

W Koenig reports personal fees from AstraZeneca, Novartis, Pfizer, The Medicines Company, DalCor, Kowa, Amgen, Sanofi, Berlin-Chemie, grants and non-financial support from Roche Diagnostics, Beckmann, Singulex and Abbott. M Roden reports personal fees from Eli Lilly, Poxel S.A., Boehringer Ingelheim, Sanofi US, Terra Firma, Servier Labatories, Novo Nordisk, ProSciento Inc., Fishawack Group, Novartis Pharma, TARGET Pharmasolutions, Gilead Sciences, Kenes Group and Nutricia/ Danone. J Scherberich has a patent at the University Charite Berlin pending. The reported disclosures are not directly related to this manuscript. All other authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

The KORA study was initiated and financed by the Helmholtz Zentrum München – German Research Center for Environmental Health, which is funded by the German Federal Ministry of Education and Research (BMBF) and by the State of Bavaria. The KORA Study Group consists of A Peters (speaker), H Schulz, L Schwettmann, R Leidl, M Heier, K Strauch, and their co-workers, who are responsible for the design and conduct of the KORA studies. The authors gratefully acknowledge the contribution by all of the field staff members conducting the KORA F4 study and thank all study participants. The study was supported by a research grant from the Virtual Diabetes Institute (Helmholtz Zentrum München) and the Clinical Cooperation Group Diabetes, Ludwig-Maximilians-University München and Helmholtz Zentrum München, and by the German Diabetes Center. The German Diabetes Center was supported by the Federal Ministry of Health (Berlin, Germany) and the Ministry of Culture and Science of the state North Rhine Westphalia (Düsseldorf, Germany). Further support was obtained from the Deutsche Diabetes Gesellschaft (DDG). The KORA F4 study was partly funded by a grant from the German Research Foundation (DFG) (RA-45913/3-1).

Author contribution statement

Conception and design of the study: C M, C Huth, C Herder, M R, M H, A P, W K, W R, J Scherberich and J Seissler; collection of data: C T, C M, C Huth, C Herder, M R, M H, A P, W K, W R, A L, J Scherberich and J Seissler; data analysis, interpretation of results, writing of the manuscript: C T, H T, A L, J Seissler and J Scherberich; all authors revised the manuscript critically for important intellectual content and approved the final version.