Original Investigation Modifiable Factors Associated With Copeptin Concentration: A General Population Cohort Maatje D.A. van Gastel, BSc,1 Esther Meijer, MD, PhD,1 Lieneke E. Scheven, MD,1 Joachim Struck, PhD,2 Stephan J.L. Bakker, MD, PhD,1 and Ron T. Gansevoort, MD, PhD1 Background: Vasopressin plays an important role in maintaining volume homeostasis. However, recent studies suggest that vasopressin also may play a detrimental role in the progression of chronic kidney disease. It therefore is of interest to identify factors that influence vasopressin concentration, particularly modifiable ones. Study Design: Cross-sectional analyses. Setting & Participants: Data used are from participants in a large general-population cohort study (Prevention of Renal and Vascular Endstage Disease [PREVEND]). Patients with a missing copeptin value (n 5 888), nonfasting blood sample (n 5 495), missing or assumed incorrect 24-hour urine collection (n 5 388), or heart failure (n 5 20) were excluded, leaving 6,801 participants for analysis. Factor: Identification of lifestyle- and diet-related factors that are associated with copeptin concentration. Outcomes: Copeptin concentration as surrogate for vasopressin. Measurements: Copeptin was measured by an immunoluminometric assay as a surrogate for vasopressin. Associations were assessed in uni- and multivariable linear regression analyses. Results: Median copeptin concentration was 4.7 (IQR, 2.9-7.6) pmol/L. When copeptin was studied as a dependent variable, the final stepwise backward model revealed associations with higher copeptin concentrations for lower 24-hour urine volume (P , 0.001), higher sodium excretion (P , 0.001), higher systolic blood pressure (P , 0.001), current smoking (P , 0.001), higher alcohol use (P , 0.001), higher urea excretion (P 5 0.003), lower potassium excretion (P 5 0.002), use of glucose-lowering drugs (P 5 0.02), higher body mass index (P , 0.001), and higher plasma glucose level (P , 0.001). No associations with copeptin concentration were found for C-reactive protein or use of diuretics or nondiuretic antihypertensives. Limitations: The cross-sectional study design does not allow firm conclusions on cause-effect relationships. Conclusions: Important lifestyle- and diet-related factors associated with copeptin concentration are current smoking, alcohol use, protein and potassium intake, and particularly fluid and sodium intake. These data form a rationale to investigate whether intervening on these factors results in a lower vasopressin concentration with concomitant beneficial renal effects. Am J Kidney Dis. 65(5):719-727. ª 2015 by the National Kidney Foundation, Inc. INDEX WORDS: Copeptin; vasopressin; general population cohort; lifestyle; diet; fluid intake; sodium intake; modifiable factor; kidney disease progression.

V

asopressin plays an important role in the physiology of volume homeostasis. However, it recently became clear that vasopressin also may play a deleterious role in the progression of heart failure and particularly chronic kidney disease (CKD).1-5 In cross-sectional epidemiologic studies performed in the general population and patients with diabetes, it has been shown that copeptin concentration (as a surrogate marker for vasopressin) is associated with albuminuria and estimated glomerular filtration rate (eGFR).2,6-8 In addition to this epidemiologic evidence, it has been shown in intervention studies that synthetic vasopressin infusion causes an increase in albuminuria, whereas lowering vasopressin activity by increasing water intake or by vasopressin antagonists induced renoprotection in animal models.3,9,10 It therefore is of interest to determine which factors can influence plasma vasopressin concentrations, particularly modifiable factors, because by intervening Am J Kidney Dis. 2015;65(5):719-727

regarding these factors it may be possible to ameliorate the detrimental effects associated with increased vasopressin concentration. Vasopressin is secreted by the pituitary gland, primarily in response to increases in plasma osmolality. As a consequence, the kidney will retain From the 1Department of Nephrology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands; and 2ThermoFisher Scientific, Clinical Diagnostics, Hennigsdorf, Germany. Received April 13, 2014. Accepted in revised form October 1, 2014. Originally published online December 8, 2014. Address correspondence to Ron T. Gansevoort, MD, PhD, Department of Nephrology, University Medical Center Groningen, PO Box 30.001, 9700 RB Groningen, the Netherlands. E-mail: r.t. [email protected]  2015 by the National Kidney Foundation, Inc. 0272-6386 http://dx.doi.org/10.1053/j.ajkd.2014.10.009 719

van Gastel et al

water, which will result in a decrease in plasma osmolality toward preexisting values. Plasma osmolality is dependent predominantly on fluid and osmolar intake. In turn, the latter is determined by the intake of sodium and protein (as main determinant of plasma urea concentration). The acute effects of fluid, sodium, and protein intake on vasopressin concentration have been well studied.9,11-17 However, less is known about the effect of long-term stable sodium and protein intake on vasopressin concentrations. Besides these dietrelated factors, there are other lifestyle factors, such as obesity, smoking, alcohol use, and coffee use, which have been suggested to influence vasopressin concentration.18-22 The evidence supporting a role for these factors is scarce and generally has been obtained in relatively small-scale studies, with a focus on single factors without adjustment for other factors and often with conflicting results.18-21 Measurement of vasopressin has been problematic for a long time because vasopressin is unstable in isolated plasma and most vasopressin assays have fairly limited sensitivity. Copeptin consists of the carboxy-terminal portion of provasopressin, the precursor of vasopressin, and is produced in equimolar amounts as vasopressin during precursor processing.23 Copeptin has been shown to be a relatively easily measurable stable substitute for circulating vasopressin.24,25 Given these considerations, we investigated, in an integrated manner in a large-scale, observational, cross-sectional study, which modifiable patient characteristics are associated with copeptin concentration.

METHODS Study Design and Population This investigation was conducted using data obtained in individuals who participate in the Prevention of Renal and Vascular Endstage Disease (PREVEND) prospective cohort study that started in 1997. Details of the study protocol have been published elsewhere.26 In summary, all inhabitants of the city of Groningen aged 28 to 75 years were sent a questionnaire on demographics, disease history, smoking habits, and medication use and a vial to collect a first-morning-void urine sample. Of these individuals, 40,856 (47.8%) responded. From these individuals, the PREVEND cohort was selected with the aim to create a cohort enriched for persons with higher albuminuria. After exclusion of patients with type 1 diabetes mellitus (defined as requiring the use of insulin) and pregnant women (defined by self-report), all individuals with urinary albumin concentration . 10 mg/L (n 5 7,768) were invited, of whom 6,000 participated. Furthermore, a randomly selected control group with urinary albumin concentration , 10 mg/L (n 5 3,394) was invited, of whom 2,592 participated. These 8,592 participants constitute the PREVEND cohort and visited a research clinic for in-detail assessment of clinical characteristics, including blood and urine collection. For the present study, individuals from the PREVEND cohort with a missing copeptin value (n 5 888), 720

nonfasting blood sample (n 5 495), missing or assumed incorrect 24-hour urine collection (n 5 388, see below), and heart failure (n 5 20) were excluded, leaving 6,801 participants for the present analyses. Of these participants, at the initial screening, 4,786 had an albumin concentration . 10 mg/L (70.4%), and 2,015 (29.6%), ,10 mg/L. The PREVEND Study was approved by the medical ethics committee of our institution, was conducted in accordance with the International Conference of Harmonization Good Clinical Practice Guidelines, and adheres to the ethical principles that have their origin in the Declaration of Helsinki.

Measurements and Definitions At baseline, participants completed 2 visits at a screening facility for an extensive general workup. They filled in a questionnaire on disease history, medication use, and lifestyle factors such as smoking and coffee and alcohol use. Height and weight were measured and body mass index (BMI) was calculated. Body surface area was calculated according to the Mosteller equation.27 Blood pressure was measured on the right arm in a supine position, every minute for 10 and 8 minutes on the first and second visits, respectively, to our outpatient unit, with an automatic device (Dinamap XL model 9300; Johnson-Johnson Medical). Systolic blood pressure (SBP) was calculated as the mean of the last 2 measurements of each of the 2 visits. Two 24-hour urine samples were collected, after thorough oral and written instructions on how to perform such a urine collection, in which creatinine, sodium, urea, and potassium were measured. Participants were fasting from midnight onward when they visited the PREVEND research clinic between 8:00 and 11:00 AM, where blood was drawn for measurement of copeptin, creatinine, cholesterol, glucose, albumin, and high-sensitivity Creactive protein (CRP). Thus, participants had not used coffee, alcohol, or cigarettes for at least 8 hours. We measured copeptin in baseline samples of the PREVEND cohort using a sandwich immunoassay (B.R.A.H.M.S. AG/ThermoFisher), with a lower limit of detection of 0.4 pmol/L and functional assay sensitivity (defined as when the assay has a 20% interassay coefficient of variation) , 1 pmol.28 History of cardiovascular disease was defined as self-reported myocardial infarction, cardiac surgery, percutaneous transluminal coronary angioplasty, or cerebrovascular accident. Information for specific drug use was obtained from the Inter-Action Data-Base, which comprises pharmacy-dispensing data from community pharmacists located in the northern regions of the Netherlands.29 Included in our definition of diuretic use were all agents mentioned under C03 of the Anatomical Therapeutic Chemical classification system. Use of other antihypertensives was defined as all agents mentioned under C02 of the same classification system, with the exception of agents mentioned under C02L (antihypertensives and diuretics in combination). Individuals who used these latter agents were included in the diuretics group. Serum creatinine concentration was used to calculate eGFR using the CKD-EPI (CKD Epidemiology Collaboration) creatinine equation.30 For 24-hour urine volume and sodium, urea, and potassium excretion, we used the mean volume or excretion of the mentioned analytes in the two 24-hour urine samples. We deliberately did not study overall urine osmolality because urine osmolality is determined by intake of fluid and osmoles, with intake of osmoles in turn being determined by intake of sodium, potassium, and protein as a source of urea. We studied these variables separately to allow assessing their individual influence. For these food parameters, the average of the 2 values at baseline was used. If the difference between expected and measured 24-hour urine volume was outside the 95% distribution range, the 24-hour urine collection was assumed to be incorrect. The expected 24-hour urine volume was calculated by comparing Am J Kidney Dis. 2015;65(5):719-727

Copeptin in the General Population creatinine clearance estimated by the Cockcroft-Gault formula and actual creatinine clearance.

was adjusted for age, sex, eGFR, and 24-hour urine volume. Variables that were additionally entered in this model were cardiovascular disease history, current smoking, alcohol use, coffee use, BMI, body surface area, SBP, use of diuretics, use of antihypertensives without diuretic use, high-sensitivity CRP level, plasma glucose level, diabetes mellitus, use of glucose-lowering drugs, serum cholesterol level, use of lipid-lowering drugs, and sodium, urea, and potassium excretion (as surrogate measures for sodium, protein, and potassium intake, respectively). P . 0.05 was used to exclude factors from the model. Before factors were definitively excluded, it was first tested whether nonlinear associations with copeptin concentration existed by adding the square of the variable under study to the model. Variables that were maintained in the final model also were tested for nonlinear associations and interactions. Exponential b’s and confidence intervals are used because the dependent (copeptin) was ln-transformed. For the ln-transformed predictor variable highsensitivity CRP, the effect per doubling is used. Figure 1 was generated from results of the final model and graphically shows the strength of the association of each significant variable with copeptin concentration, after adjustment for all other variables in the model. Only the modifiable variables that are associated significantly with copeptin concentration are shown. Data with a continuous distribution are shown stratified into quintiles, and categorical data, stratified into clinical classes. As sensitivity analyses, we repeated all the mentioned analyses excluding individuals who used diuretics or had diabetes mellitus (both are well-known factors associated with vasopressin concentration) and in a subcohort of the PREVEND Study that is representative of the general population to determine whether the cohort selection design with enrichment for individuals with higher albuminuria influenced our results.32 For this purpose, all participants with albumin concentrations , 10 mg/L (n 5 2,592)

Statistical Analyses Continuous data are reported as mean 6 standard deviation (SD), whereas non-normally distributed variables are reported as median with interquartile range (IQR). Baseline characteristics of the study population are stratified according to sex-specific quintiles of copeptin concentration because copeptin concentration is known to be higher in men than in women.24,31 Differences between the 5 strata were tested by analysis of variance (F test) for continuous data, Kruskal-Wallis test for non-normally distributed variables, and c2 test for categorical variables. To test for a significant trend between quintiles, we used linear regression analysis for continuous variables and analysis of linear-by-linear associations for categorical variables. For all linear regression analyses, non-normally distributed variables (ie, copeptin, high-sensitivity CRP, and albuminuria values) were natural log (ln)-transformed to meet the assumptions for linear regression analyses. For all continuous variables, the effect per 1-SD increment is shown. To corroborate earlier findings that copeptin may be linked to CKD, we first performed linear regression analyses testing associations of copeptin concentration as an independent variable with eGFR and albuminuria as dependent variables. Second, we performed multivariable linear regression analyses with copeptin concentration as a dependent variable. In these analyses, we adjusted for 3 nonmodifiable factors (age, sex, and eGFR) and 24-hour urine volume (as a surrogate for fluid intake) as a well-known and acknowledged determinant of vasopressin concentration.9,11,12 Next, we performed stepwise backward multivariable linear regression analysis to investigate in an integrated manner the association between copeptin concentration and modifiable factors. This model again 7 P < 0.001

P < 0.001

P < 0.001

P < 0.001

P < 0.001

24-h urine volume

24-h sodium excretion

Current smoking (no. of cigarettes)

Systolic blood pressure

Use of alcohol (no. of servings)

P = 0.003

P = 0.002

P = 0.02

P < 0.001

P < 0.001

24-h urea excretion

24-h potassium excretion

Use of GLD

Body mass index

Plasma glucose

Copeptin (pmol/L)

6 5 4 3 0 2

Copeptin (pmol/L)

7 6 5 4 3 0 2

Figure 1. Modifiable participant characteristics associated independently with copeptin concentration. Variables are shown adjusted for all other variables that were contained in the final stepwise backward regression model, allowing comparison of the strength of associations of all individual variables with copeptin concentration. Variables are shown in decreasing order of the strength of their association with copeptin concentration. Continuous variables are subdivided in quintiles numbered 1 to 5, representing the lowest to highest quintile, respectively, whereas categorical variables are subdivided in clinical classes. For categorical variables, number of participants per category are as follows: current smoking: no (n 5 2,013), ,6 cigarettes/day (n 5 979), 6 to 20/day (n 5 2,674), .20/day (n 5 861); use of alcohol: no (n 5 1,675), 1 to 4 servings/month (n 5 1,028), 2 to 7/week (n 5 2,228), 1 to 3/day (n 5 1,269), .4/day (n 5 308); use of glucose-lowering drugs (GLD): no (n 5 5,420), yes (n 5 80). Am J Kidney Dis. 2015;65(5):719-727

721

van Gastel et al were added to a subset of individuals whose albumin concentrations were $10 mg/L that were selected by proportionally taking a computer-generated random subset (n 5 840). The 18 individuals with either proteinuria or kidney disease were excluded, resulting in a subcohort with 3,414 participants. All analyses were performed using IBM SPSS Statistics, version 20.0.0.1. P , 0.05 was considered to indicate statistical significance.

RESULTS A total of 6,801 participants were analyzed, of whom 3,191 were men, with a mean age of 48.7 6 12.5 years and median copeptin concentration of 4.7 (IQR, 2.9-7.6) pmol/L. There was a significant

difference in copeptin concentrations between men and women, with the former having higher median values (6.3 [IQR, 4.1-9.5]) than the latter (3.6 [IQR, 2.4-5.6] pmol/L; P , 0.001). Baseline characteristics of the study population therefore are given according to sex-stratified quintiles of copeptin concentration (Table 1). Nearly all characteristics were associated significantly with copeptin concentrations, including a positive association for albuminuria and an inverse one for eGFR. The associations of copeptin concentration with albuminuria and eGFR remained significant after adjustment for age and sex (P , 0.001 for

Table 1. Baseline Characteristics of Study Population According To Sex-Stratified Quintiles of Plasma Copeptin Concentration Quintile 1a (n 5 1,368)

Quintile 2b (n 5 1,364)

Quintile 3c (n 5 1,357)

Quintile 4d (n 5 1,373)

Copeptin (pmol/L)

2.0 [1.6-2.7]

3.0 [2.6-4.5]

4.2 [3.6-6.2]

6.1 [5.0-8.7]

Age (y)

47.8 6 12.5

47.6 6 12.3

48.6 6 12.7

48.7 6 12.2

51.1 6 12.6

641 (46.9) 65 (4.8)

636 (46.6) 51 (3.7)

639 (47.3) 67 (5.0)

653 (47.3) 77 (5.6)

622 (46.5) 76 (5.7)

Male sex CVD history

Quintile 5e (n 5 1,339)

P

P Trend

11.6 [8.3-14.7] ,0.001 ,0.001 ,0.001 ,0.001 0.9 0.1

0.9 0.05

,0.001 ,0.001

Current smoking

406 (29.8)

410 (30.2)

454 (33.8)

485 (35.2)

521 (39.0)

Alcohol use

979 (71.8)

1,054 (77.6)

1,016 (75.7)

1,030 (75.0)

984 (73.8)

0.01

0.7

Coffee use

1,295 (95.1)

1,290 (95.1)

1,280 (95.3)

1,313 (95.4)

1,256 (94.1)

0.6

0.4

BMI (kg/m2)

25.4 6 3.6

25.8 6 3.9

26.1 6 4.1

26.4 6 4.3

26.4 6 4.3

BSA (m2)

1.91 6 0.20

1.93 6 0.21

1.93 6 0.21

1.94 6 0.21

1.93 6 0.22

SBP (mm Hg)

126.5 6 19.6

127.2 6 19.1

127.7 6 18.9

128.5 6 19.6

131.6 6 21.6

46 (3.4) 128 (9.4)

44 (3.2) 125 (9.2)

39 (2.9) 124 (9.2)

43 (3.1) 121 (8.8)

68 (5.1) 165 (12.3)

Use of diuretics Use of other antihypertensives

0.001

0.004

,0.001 ,0.001 0.02 0.01

0.04 0.03

86.7 6 14.3

86.2 6 15.0

83.9 6 14.8

83.1 6 14.8

Albuminuria (mg/24 h)

8.1 [6.0-13.8]

8.8 [6.3-15.4]

9.2 [6.3-16.0]

9.9 [6.6-16.6]

hs-CRP (mg/L)

1.1 [0.5-2.7]

1.1 [0.5-2.6]

1.3 [0.6-2.8]

1.3 [0.6-3.1]

1.5 [0.6-3.3]

,0.001 ,0.001

4.8 6 0.8

4.8 6 1.1

4.8 6 0.9

4.9 6 1.1

5.0 6 1.5

,0.001 ,0.001

34 (2.5) 16 (1.2)

49 (3.6) 19 (1.5)

36 (2.7) 21 (1.6)

36 (2.6) 9 (0.7)

66 (4.9) 17 (1.3)

5.6 6 1.1

5.5 6 1.1

5.6 6 1.1

5.6 6 1.1

5.8 6 1.2

eGFR (mL/min/1.73 m2)

Plasma glucose (mmol/L) Diabetes mellitus Use of glucose-lowering drugs Serum cholesterol (mmol/L) Use of lipid-lowering drugs

80.1 6 17.0

,0.001 ,0.001

,0.001 ,0.001

11.4 [7.1-23.4] ,0.001 ,0.001

0.001 0.2

0.01 0.5

,0.001 ,0.001

57 (4.3)

49 (3.7)

57 (4.4)

52 (3.9)

64 (4.9)

1.8 6 0.6 74.3 6 21.4

1.6 6 0.5 72.5 6 20.3

1.6 6 0.5 71.8 6 20.1

1.5 6 0.5 69.6 6 20.6

1.4 6 0.5 66.8 6 20.7

Urea excretion (mmol/24 h)

353 6 99

360 6 102

355 6 99

353 6 104

346 6 107

0.008

0.01

Sodium excretion (mmol/24 h)

137 6 49

143 6 49

144 6 50

142 6 51

140 6 51

0.002

0.2

Urine volume (L/24 h) Potassium excretion (mmol/24 h)

0.7

0.4

,0.001 ,0.001 ,0.001 ,0.001

Note: Values for categorical variables are given as number (percentage); values for continuous variables are given as mean 6 standard deviation or median [interquartile range]. The P value of differences between strata for parametric variables is obtained using analysis of variance, Kruskal-Wallis test in case of a non-normally distributed variable, or c2 test in case of a categorical variable. P for trend for continuous variables is obtained using linear regression analysis, natural log–transformed in case of a non-normally distributed variable or linear-by-linear associations in case of a categorical variable. Abbreviations: BMI, body mass index; BSA, body surface area; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; hs-CRP, high-sensitivity C-reactive protein; SBP, systolic blood pressure. a Corresponding copeptin values: #3.76 pmol/L (men); #2.20 pmol/L (women). b Corresponding copeptin ranges: .3.76-5.35 pmol/L (men); .2.20-3.09 pmol/L (women). c Corresponding copeptin ranges: .5.35-7.40 pmol/L (men); .3.09-4.27 pmol/L (women). d Corresponding copeptin ranges: .7.40-10.50 pmol/L (men); .4.27-6.27 pmol/L (women). e Corresponding copeptin values: .10.50 pmol/L (men); .6.27 pmol/L (women). 722

Am J Kidney Dis. 2015;65(5):719-727

Copeptin in the General Population

both). In general, characteristics of the 1,791 individuals who were excluded from the present analyses did not differ substantially from included individuals (Table S1, available as online supplementary material). Table 2 shows associations between the participant-related variables of interest and copeptin concentrations. As expected, 24-hour urine volume was associated strongly and negatively with copeptin concentration. Adjusted for age, sex, eGFR, and 24-hour urine volume, most other variables also were associated significantly with copeptin concentration. Second, a stepwise backward linear regression model was made. The overall r2 of the final model was 0.265. Associations with higher copeptin concentrations were found with higher sodium excretion

(P , 0.001), current smoking (P , 0.001), higher SBP (P , 0.001), alcohol use (P , 0.001), higher urea excretion (P 5 0.003), lower potassium excretion (P 5 0.002), and use of glucose-lowering drugs (negatively, P 5 0.02). For some of the continuous distributed variables, there was significantly better fit for nonlinear relationships, that is, eGFR, 24-hour urine volume, BMI, and plasma glucose level (all P , 0.001). No associations with copeptin concentrations were found for the characteristics of coffee use, diabetes mellitus, serum cholesterol level, CRP level, and use of diuretics or nondiuretic antihypertensives. In Fig 1, associations between copeptin concentrations and all modifiable variables that were retained in the final stepwise backward regression model are

Table 2. Multivariable Linear Regression Analyses With Copeptin Concentration as Dependent Variable Adjusted for Sex, Age, eGFR, and 24-h Urine Volumea eb,b

e95%

CI,b

Stepwise Backward Model

P

eb,b

e95%

CI,b

P

Age, per 12.5-y oldera

0.986

0.969-1.004

0.1

0.990

0.970-1.010

0.3

Sex, female vs malea

0.589

0.572-0.607

,0.001

0.646

0.624-0.668

,0.001

eGFR, per 15.4–mL/min/1.73 m2 greatera eGFR2

0.887

0.870-0.903

,0.001

0.553 1.584

0.496-0.617 1.420-1.766

,0.001 ,0.001

24-h urine volume, per 0.5–L/24 h greatera 24-h urine volume2

0.853

0.840-0.865

,0.001

0.700 1.192

0.670-0.732 1.143-1.244

,0.001 ,0.001

CVD history, yes vs no

0.980

0.914-1.050

0.6

Current smoking, yes vs no

1.101

1.067-1.135

,0.001

1.112

1.078-1.147

,0.001

Alcohol use, yes vs no

1.060

1.024-1.097

0.001

1.089

1.052-1.126

,0.001

Coffee use, yes vs no BMI, per 4.2-kg/m2 greater BMI2

0.956 1.038

0.894-1.023 1.022-1.053

0.2 ,0.001

0.727 1.381

0.647-0.816 1.231-1.545

,0.001 ,0.001

1.041

1.023-1.060

,0.001

0.899 1.150

0.854-0.947 1.095-1.209

,0.001 ,0.001

0.847

0.737-0.972

0.02

0.002

BSA, per 0.2-m2 greater

1.036

1.017-1.053

,0.001

SBP, per 19.9–mm Hg greater

1.044

1.026-1.062

,0.001

Use of diuretics, yes vs no

1.082

0.998-1.174

0.06

Use of other antihypertensives, yes vs no hs-CRP, per 1.2-mg/L greater

0.998 1.007

0.948-1.050 1.002-1.012

0.9 0.003

Plasma glucose, per 1.1-mmol/L greater Glucose2

1.043

1.027-1.059

,0.001

Diabetes mellitus, yes vs no

1.153

1.060-1.252

0.001

Use of antidiabetics, yes vs no Serum cholesterol, per 1.1-mmol/L greater

0.950 1.021

0.830-1.087 1.006-1.038

0.5 0.007

Use of lipid lowering drugs, yes vs no

0.935

0.868-1.008

0.08

Potassium excretion, per 20.8–mmol/24 h greater

0.995

0.978-1.012

0.6

0.970

0.952-0.990

1.046

1.029-1.064

,0.001

1.034

1.011-1.055

0.003

1.093

1.076-1.111

,0.001

1.100

1.078-1.121

,0.001

Urea excretion, per 102.1–mmol/24 h greater

Sodium excretion, per 50.2–mmol/24 h greater

Note: For continuous variables, the effect per 1–standard deviation increment is used. Abbreviations: BMI, body mass index; BSA, body surface area; CI, confidence interval; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; hs-CRP, high sensitivity C-reactive protein; ln, natural log; SBP, systolic blood pressure. a Adjusted for the 3 other variables. b Values were back ln-transformed for ease of comparison. For hs-CRP, an ln-transformed response variable, the effect per doubling is used. Am J Kidney Dis. 2015;65(5):719-727

723

van Gastel et al

shown. As shown in Fig 1, the 24-hour urine volume has a 2-fold difference in copeptin concentrations between the lowest and highest quintiles of volume intake. Of note, this figure also shows that associations of BMI and plasma glucose level with copeptin concentrations were J-shaped, with both low and high values being associated with higher copeptin concentration. As sensitivity analyses, we repeated these analyses excluding the 240 individuals using diuretics and/or the 221 individuals with diabetes mellitus. Essentially similar results were obtained. The same holds true when analyses were performed using data from a subcohort of the PREVEND cohort (n 5 2,676) that is representative of the general population (Tables S2 and S3; Fig S1).

DISCUSSION This study was performed to identify modifiable factors that are associated with plasma vasopressin concentrations in community-dwelling individuals. The final stepwise backward regression model revealed associations with higher copeptin concentrations for lower 24-hour urine volume, higher sodium excretion, higher blood pressure, current smoking, higher alcohol use, higher urea excretion, lower potassium excretion, use of glucose-lowering drugs, higher BMI, and higher plasma glucose level. As has been described in previous studies, we found a negative association between copeptin concentration and eGFR (P , 0.001) and a positive association with albuminuria (P , 0.001).2,33 These findings corroborate the observations that suggest that vasopressin may be related to CKD and confirm the relevance of finding modifiable determinants of higher copeptin concentrations because intervening regarding these factors may have a beneficial effect on CKD. Our final regression model revealed that sex was associated strongly with copeptin concentration. Although the exact mechanism behind this phenomenon is not known, it is well described and may be caused by higher osmolar intake in men.2,24,31,34 More of interest for this study are the modifiable factors that are associated with copeptin concentration, of which the most important are 24-hour urine volume, which is a surrogate for fluid intake, and sodium excretion.35 Both were associated with copeptin concentrations. For 24-hour urine volume (ie, fluid intake), this finding is in line with the physiologic role of vasopressin and is widely acknowledged.9,11,12 In a steady-state situation, 24-hour sodium excretion can be used as a surrogate for sodium intake.36 It is known that short-term changes in sodium intake have effects on vasopressin concentrations.13-15 However, 724

the literature is limited about whether long-term ad libitum sodium intake has an effect on vasopressin concentration, with contradictory results.37 We found a strong positive association of sodium intake with copeptin concentration.24,25 We hypothesize that higher sodium intake will result in higher plasma osmolality and with that, higher vasopressin concentration.38 We also found associations between 24-hour excretions of osmoles of urea and potassium and copeptin concentrations. Urea excretion is the main component of protein intake, which can be calculated from 24-hour urea excretion using the Maroni formula.39,40 We expected that higher protein intake would result in higher copeptin concentrations because protein intake determines osmolar load, together with sodium intake. The considerably weaker association of urea excretion with copeptin concentration compared to sodium excretion perhaps may be explained by vasopressin’s effect on urea excretion, with a high vasopressin concentration resulting in low fractional urea excretion to enable the kidney to reabsorb water.41 Because the formula to calculate protein intake is dependent on urea excretion, these 2 phenomena may counteract. Another, and in our opinion more likely, explanation for the weaker association of urea versus sodium excretion might be that urea, although in urine, a constituent of osmolar load, can freely cross cell membranes and as such, in plasma is not a determinant of effective osmolality and consequently is not a determinant of vasopressin (and copeptin) concentration.42 Potassium excretion is known to be influenced by vasopressin concentration because vasopressin can influence potassium secretion in the late distal nephron.43,44 However, in steady-state situations such as in our communitydwelling participants, 24-hour potassium excretion reflects daily potassium intake. In our opinion, a reverse cause-effect association therefore is more likely being that potassium intake, by its blood pressure2lowering effect, increases copeptin concentration. We furthermore found associations between copeptin concentration and smoking, alcohol use, SBP, use of glucose-lowering drugs, BMI, and plasma glucose level. In general, these findings are in line with the scant literature that has investigated these issues, albeit that these studies did not investigate whether such associations were independent of other factors, as was done in the present study.18-21 Worth discussing in more depth are the associations with SBP, BMI, and glucose level. Both an increase in plasma osmolality and a decrease in blood volume are known to be strong physiologic stimuli for vasopressin release.45-47 It also is known that vasopressin is a vasoconstrictor that causes blood pressure to Am J Kidney Dis. 2015;65(5):719-727

Copeptin in the General Population

increase acutely and transiently. An opposite causal relationship therefore also is possible. We hypothesize that given the fact that we found a positive association between copeptin concentration and blood pressure, copeptin (ie, vasopressin) was a cause and not the result of higher blood pressure. Associations of copeptin concentrations with use of glucoselowering drugs, BMI, and plasma glucose level were relatively weak and J-shaped (Fig 1) and therefore of limited clinical relevance. Especially higher, but also lower, values were associated with higher plasma copeptin concentrations. When interpreting our results, it should be taken into consideration that statistical significance of the mentioned associations does not necessarily imply a strong biological influence of the variables under study on vasopressin concentrations. We therefore visualized these associations in Fig 1, in which the associations for each individual variable were adjusted for all other variables in the final regression model. This figure shows that strong associations were found especially for 24-hour urine volume and sodium excretion. All other variables in general had ,1 pmol/L difference in copeptin concentration between the 2 extreme groups. We acknowledge that our study has limitations. First, the PREVEND Study cohort is enriched for individuals with higher albuminuria, which may affect the results obtained. However, our sensitivity analyses that were performed in a subcohort representative of the general population showed essentially similar results, making bias caused by our cohort selection design unlikely. Second, due to the crosssectional study design, no firm conclusions can be drawn on the cause-effect relationship of the associations that we found. Third, we had to exclude 1,791 individuals from the present analyses because of missing values or incorrect 24-hour urinary collection. Baseline characteristics of these individuals in general did not differ substantially from those of individuals who were included, although due to the large number of individuals included in the present analyses, sometimes statistically significant differences were found. Therefore, and because the reason of exclusion was random, it is unlikely that this selection may have resulted in significant bias. Last, nearly all included individuals were of European ancestry, and our results therefore may not be valid for other populations. Strengths of the present study are that we have data from a large cohort of community-dwelling individuals with extensive information for a large number of covariates and the availability of two 24-hour urine collections. In conclusion, we found that higher copeptin concentration, as a surrogate for vasopressin, is Am J Kidney Dis. 2015;65(5):719-727

associated with lower kidney function and higher albuminuria, compatible with the alleged deleterious renal effects of vasopressin. Important modifiable participant-related factors associated with copeptin concentrations are current smoking, alcohol use, and particularly fluid and sodium intake. These data form a rationale to investigate whether intervening regarding these factors results in lower vasopressin concentrations with concomitant beneficial renal effects.

ACKNOWLEDGEMENTS Support: The PREVEND Study was supported financially by grant E.013 of the Dutch Kidney Foundation. Copeptin was measured by B.R.A.H.M.S. AG/ThermoFisher. The supporting agencies had no role in the design or conduct of the study; collection, analysis, or interpretation of the data; or preparation and approval of the manuscript. Financial Disclosure: The authors declare that they have no other relevant financial interests. Contributions: Research idea and study design: MDAvG, SJLB, RTG; data acquisition: SJLB, RTG, JS; data analysis/ interpretation: MDAvG, EM, LES, RTG; statistical analysis: MDAvG, LES, SJLB, RTG; supervision or mentorship: EM, LES, RTG. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. RTG takes responsibility that this study has been reported honestly, accurately, and transparently; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.

SUPPLEMENTARY MATERIAL Table S1: Characteristics of PREVEND Study participants included and excluded in present analyses. Table S2: Baseline characteristics of the PREVEND subcohort representative of the general population, by sex-stratified copeptin quintiles. Table S3: Multivariable linear regression with copeptin concentration as the dependent variable in the PREVEND subcohort representative of the general population. Figure S1: Modifiable participant characteristics independently associated with copeptin concentration in the PREVEND subcohort representative of the general population. Note: The supplementary material accompanying this article (http://dx.doi.org/10.1053/j.ajkd.2014.10.009) is available at www.ajkd.org

REFERENCES 1. Maisel A, Xue Y, Shah K, et al. Increased 90-day mortality in patients with acute heart failure with elevated copeptin: secondary results from the Biomarkers in Acute Heart Failure (BACH) Study. Circ Heart Fail. 2011;4(5):613-620. 2. Meijer E, Bakker SJ, Halbesma N, de Jong PE, Struck J, Gansevoort RT. Copeptin, a surrogate marker of vasopressin, is associated with microalbuminuria in a large population cohort. Kidney Int. 2010;77(1):29-36. 3. Meijer E, Boertien W, Zietse R, Gansevoort R. Potential deleterious effects of vasopressin in chronic kidney disease and particularly autosomal dominant polycystic kidney disease. Kidney Blood Press Res. 2011;34(4):235-244. 725

van Gastel et al 4. Yalta K, Yalta T, Sivri N, Yetkin E. Copeptin and cardiovascular disease: a review of a novel neurohormone. Int J Cardiol. 2013;167(5):1750-1759. 5. Bankir L, Bouby N, Ritz E. Vasopressin: a novel target for the prevention and retardation of kidney disease? Nat Rev Nephrol. 2013;9(4):223-239. 6. Nishikawa T, Omura M, Iizuka T, Saito I, Yoshida S. Short-term clinical trial of 1-(1-[4-(3-acetylaminopropoxy)benzoyl]-4-piperidyl)-3, 4-dihydro-2(1H)-quinolinone in patients with diabetic nephropathy. Possible effectiveness of the specific vasopressin V1 receptor antagonist for reducing albuminuria in patients with non-insulin dependent diabetes mellitus. Arzneimittelforschung. 1996;46(9):875-878. 7. Boertien W, Riphagen I, Drion I, et al. Copeptin, a surrogate marker for arginine vasopressin, is associated with declining glomerular filtration in patients with diabetes mellitus (ZODIAC33). Diabetologia. 2013;56(8):1680-1688. 8. Velho G, Bouby N, Hadjadj S, et al. Plasma copeptin and renal outcomes in patients with type 2 diabetes and albuminuria. Diabetes Care. 2013;36(11):3639-3645. 9. Bouby N, Bachmann S, Bichet D, Bankir L. Effect of water intake on the progression of chronic renal failure in the 5/6 nephrectomized rat. Am J Physiol. 1990;258(4, pt 2):F973-F979. 10. Bardoux P, Bichet D, Martin H, et al. Vasopressin increases urinary albumin excretion in rats and humans: involvement of V2 receptors and the renin-angiotensin system. Nephrol Dial Transplant. 2003;18(3):497-506. 11. Fyhrquist F, Wallenius M, Hollemans H. Radioimmunoassay of vasopressin in unextracted plasma. Scand J Clin Lab Invest. 1976;36:841-847. 12. Khokhar A, Ramage C, Slater J. Radioimmunoassay of arginine-vasopressin in human urine and its use in physiological and pathological states. J Endocrinol. 1978;79(3):375-389. 13. Cowley AJ, Skelton M, Merrill D, Quillen EJ, Switzer S. Influence of daily sodium intake on vasopressin secretion and drinking in dogs. Am J Physiol. 1983;245:R860-R872. 14. Kjeldsen S, Os I, Forsberg G, et al. Dietary sodium intake increases vasopressin secretion in man. J Clin Hypertens. 1985;1: 123-131. 15. Crofton J, Dustan H, Share L, Brooks D. Vasopressin secretion in normotensive black and white men and women on normal and low sodium diets. J Endocrinol. 1986;108(2):191-199. 16. Daniels B, Hostetter T. Effects of dietary protein intake on vasoactive hormones. Am J Physiol. 1990;258(5, pt 2): R1095-R1100. 17. Little J, Radford EJ. Circulating antidiuretic hormone in rats: effects of dietary electrolytes and protein. Am J Physiol. 1964;207:821-825. 18. Chiodera P, Capretti L, Marchesi M, et al. Abnormal arginine vasopressin response to cigarette smoking and metoclopramide (but not to insulin-induced hypoglycemia) in elderly subjects. J Gerontol. 1991;46:M6-M10. 19. Chiodera P, Volpi R, Capretti L, et al. Gammaaminobutyric acid mediation of the inhibitory effect of endogenous opioids on the arginine vasopressin and oxytocin responses to nicotine from cigarette smoking. Metabolism. 1993;42:762-765. 20. Linkola J, Ylikahri R, Fyhrquist F, Wallenius M. Plasma vasopressin in ethanol intoxication and hangover. Acta Physiol Scand. 1978;104:180-187. 21. Izzo JJ, Ghosal A, Kwong T, Freeman R, Jaenike J. Age and prior caffeine use alter the cardiovascular and adrenomedullary responses to oral caffeine. Am J Cardiol. 1983;52:769-773. 22. Enhörning S, Bankir L, Bouby N, et al. Copeptin, a marker of vasopressin, in abdominal obesity, diabetes and microalbuminuria: 726

the prospective Malmö Diet and Cancer Study cardiovascular cohort. Int J Obes (Lond). 2013;37(4):598-603. 23. de Bree F, Burbach J. Structure-function relationships of the vasopressin prohormone domains. Cell Mol Neurobiol. 1998;18(2):173-191. 24. Morgenthaler NG, Struck J, Alonso C, Bergmann A. Assay for the measurement of copeptin, a stable peptide derived from the precursor of vasopressin. Clin Chem. 2006;52:112-119. 25. Morgenthaler N, Struck J, Jochberger S, Dunser M. Copeptin. Clinical use of a new surrogate. Trends Endocrinol Metab. 2008;19:43-49. 26. Pinto-Sietsma S, Janssen W, Hillege H, Navis G, de Zeeuw D, de Jong P. Urinary albumin excretion is associated with renal functional abnormalities in a nondiabetic population. J Am Soc Nephrol. 2000;11:1882-1888. 27. Mosteller R. Simplified calculation of body surface area. N Engl J Med. 1987;317:1098. 28. Fenske W, Störk S, Blechschmidt A, Maier S, Morgenthaler N, Allolio B. Copeptin in the differential diagnosis of hyponatremia. J Clin Endocrinol Metab. 2009;94:123-129. 29. Schirm E, Monster T, de Vries R, van den Berg P, de Jongvan den Berg LT. How to estimate the population that is covered by community pharmacies? An evaluation of two methods using drug utilisation information. Pharmacoepidemiol Drug Saf. 2004;13:173-179. 30. Levey A, Stevens L, Schmid C, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150: 604-612. 31. Bhandari S, Loke I, Davies J, Squire I, Struck J, Ng L. Gender and renal function influence plasma levels of copeptin in healthy individuals. Clin Sci (Lond). 2009;116:257-263. 32. Lambers Heerspink H, Brantsma A, de Zeeuw D, et al. Albuminuria assessed from first-morning-void urine samples versus 24-hour urine collections as a predictor of cardiovascular morbidity and mortality. Am J Epidemiol. 2008;168(8):897-905. 33. Meijer E, Bakker S, de Jong P, et al. Copeptin, a surrogate marker of vasopressin, is associated with accelerated renal function decline in renal transplant recipients. Transplantation. 2009;88:561-567. 34. Perucca J, Bouby N, Valeix P, Bankir L. Sex difference in urine concentration across differing ages, sodium intake, and level of kidney disease. Am J Physiol Regul Integr Comp Physiol. 2007;292(2):R700-R705. 35. Perrier E, Rondeau P, Poupin M, et al. Relation between urinary hydration biomarkers and total fluid intake in healthy adults. Eur J Clin Nutr. 2013;67(9):939-943. 36. Brown I, Tzoulaki I, Candeias V, Elliott P. Salt intakes around the world: implications for public health. Int J Epidemiol. 2009;38(3):791-813. 37. Graffe C, Bech J, Pedersen E. Effect of high and low sodium intake on urinary aquaporin-2 excretion in healthy humans. Am J Physiol Renal Physiol. 2012;302:F264-F275. 38. He F, Markandu N, Sagnella G, de Wardener H, MacGregor G. Plasma sodium: ignored and underestimated. Hypertension. 2005;45(1):98-102. 39. Maroni B, Steinman T, Mitch W. A method for estimating nitrogen intake of patients with chronic renal failure. Kidney Int. 1985;27:58-65. 40. Masud T, Manatunga A, Cotsonis G, Mitch WE. The precision of estimating protein intake of patients with chronic renal failure. Kidney Int. 2002;62:1750-1756. 41. Bankir L. Antidiuretic action of vasopressin: quantitative aspects and interaction between V1a and V2 receptor-mediated effects. Cardiovasc Res. 2001;51(3):372-390. Am J Kidney Dis. 2015;65(5):719-727

Copeptin in the General Population 42. Zerbe R, Robertson G. Osmoregulation of thirst and vasopressin secretion in human subjects: effect of various solutes. Am J Physiol. 1983;244(6):E607-E614. 43. Field M, Stanton B, Giebisch G. Influence of ADH on renal potassium handling: a micropuncture and microperfusion study. Kidney Int. 1984;25(3):502-511. 44. Amorim J, Malnic G. V(1) receptors in luminal action of vasopressin on distal K(1) secretion. Am J Physiol Renal Physiol. 2000;278(5):F809-F816.

Am J Kidney Dis. 2015;65(5):719-727

45. Brimble M, Dyball R. Characterization of the responses of oxytocin- and vasopressin-secreting neurones in the supraoptic nucleus to osmotic stimulation. J Physiol. 1977;271(1):253-271. 46. Bourque C. Osmoregulation of vasopressin neurons: a synergy of intrinsic and synaptic processes. Prog Brain Res. 1998;119:59-76. 47. Robertson G, Ganguly A. Osmoregulation and baroregulation of plasma vasopressin in essential hypertension. J Cardiovasc Pharmacol. 1986;8(suppl 7):S87-S91.

727

Modifiable factors associated with copeptin concentration: a general population cohort.

Vasopressin plays an important role in maintaining volume homeostasis. However, recent studies suggest that vasopressin also may play a detrimental ro...
347KB Sizes 0 Downloads 3 Views