Skip to main content
NCBI home page
Nephrology Dialysis Transplantation logoLink to Nephrology Dialysis Transplantation
. 2017 Feb 4;33(2):274–283. doi: 10.1093/ndt/gfw426

Osmotic indices and kidney concentrating activity: population-based data on correlates and prognostic power

Massimo Cirillo 1,2,✉, Giancarlo Bilancio 1,2, Cinzia Lombardi 3, Pierpaolo Cavallo 4, Oscar Terradura Vagnarelli 5, Alberto Zanchetti 6, Martino Laurenzi 5
PMCID: PMC5837546  PMID: 28339633

Abstract

Background

Research data are limited on indices of osmotic equilibrium and of kidney concentrating activity (KCA). This study investigated correlates and prognostic power of these indices in a sample of the general population.

Methods

Urine osmolality (U-osm), plasma osmolality (P-osm), plasma creatinine and other variables were measured by the Gubbio Study for the 1988–92 exam (baseline). Plasma creatinine and other variables were re-measured in the 2001–07 exam (follow-up). KCA was assessed as the U-osm/P-osm ratio and kidney function as estimated glomerular filtration rate (eGFR).

Results

Baseline data were complete in 4220 adults, of whom 852 died before follow-up and 2795 participated in the follow-up. At baseline, the following independent cross-sectional associations were identified: female sex and higher urine flow with lower values of U-osm, P-osm and U-osm/P-osm ratio (P < 0.01); obesity with higher values of U-osm, P-osm and U-osm/P-osm ratio (P < 0.01); older age and lower eGFR with lower U-osm, lower U-osm/P-osm ratio and higher P-osm (P < 0.05); hypertension and smoking with lower U-osm and lower U-osm/P-osm ratio (P < 0.05) but not with P-osm. From baseline to follow-up, the annualized rate was 1.26% for mortality and −0.74 ± 0.76 mL/min × 1.73 m2 for eGFR change. Mortality was independently predicted by baseline U-osm and baseline U-osm/P-osm ratio (hazard ratio for one higher standard deviation was ≤0.91, 95% confidence interval was ≤0.97, P < 0.01), but not by baseline P-osm. The eGFR change was not independently predicted by baseline values of U-osm, P-osm and U-osm/P-osm ratio (P ≥ 0.4).

Conclusions

Sex, age, obesity, eGFR, urine flow, hypertension and smoking independently associated with U-osm and KCA. U-osm and KCA independently predicted mortality, but not kidney function change over time.

Keywords: kidney function, mortality, osmolality, urine concentration

INTRODUCTION

The water–osmole balance depends on several factors, which include intrinsic renal factors, like the urine concentration activity of the distal nephron, and non-renal factors, like the intake or the metabolic generation of water and osmolytes, the non-renal losses of water and osmolytes, and the secretion of antidiuretic hormone [1]. A tenet of medicine is that the urine osmolality (U-osm) and the kidney concentrating activity (KCA) are low in persons with reduced kidney function [2]. Actually, research data on this are limited and inconsistent [3, 4]. An association between low kidney function, assessed as creatinine clearance, and low 24-h U-osm was reported by a meta-analysis of 10 studies including 248 healthy volunteers, 117 women of a randomized controlled trial, 378 hypertensives of a dietary controlled trial, 65 diabetics and 181 patients with chronic kidney disease [3]. A later case–control study reported that kidney function assessed as estimated glomerular filtration rate (eGFR) was not associated with U-osm and KCA in analyses including 10 healthy volunteers and 41 chronic kidney disease patients with eGFR ranging from 142 to 12 mL/min × 1.73 m2 [4]. As for other determinants of KCA, previous reports indicated that U-osm is affected by gender (higher in men) [3, 5–8], age (higher in youngsters) [3, 7, 9–12] and time of collection (higher during night-time) [13]. Finally, small clinical studies indicated an association of kidney disease or hypertension with altered circadian rhythms of single urine osmolytes or of overall U-osm [14–22]. While the prognostic power for mortality and kidney function decline over time has been extensively investigated for kidney function markers like eGFR and albuminuria/proteinuria [23, 24], data are missing for the prognostic power of KCA and osmotic indices. Therefore, the present study was designed to investigate KCA and osmotic indices in a sample of the general population with the focus on correlates and prognostic power of these variables for mortality and kidney function change over time.

MATERIALS AND METHODS

The Gubbio study is a population-based investigation ongoing since 1982 in the city of Gubbio, Italy [25, 26]. The study adheres to the Declaration of Helsinki and its activities included institutional committee approval and an informed consent. Previous papers have reported study design, response rates and characteristics of the Gubbio cohort [25, 26]. The main exams of the study were performed in 1983–85, 1988–92 and 2001–07. Local sections of national registries provided data about mortality and renal failure from 1983 to 2007. The present article analyses data included from the 1988–92 exam to the 2001–07 exam (from here on referred to as ‘baseline’ and ‘follow-up’, respectively). The baseline exam included: a timed urine collection under fed condition from the first void after the evening meal to the first void at morning wake-up; an early morning venous blood sample under fasting conditions; and a brief medical visit with the administration of standardized questionnaires [26–28]. The follow-up exam included an early morning venous blood sample under fasting conditions and a brief medical visit with the administration of standardized questionnaires, but it did not include the collection of urine samples. The Supplementary data of the article reports the results of ancillary analyses for data collected from the 1983–85 exam to the baseline exam. These ancillary analyses were carried out to assess the consistency of results across different exams with urine/blood samples collected at different times of the day. The 1983–85 exam included the collection of urine/blood samples in the morning or in the afternoon together with a brief medical visit and the administration of standardized questionnaires [25, 26, 29].

Variables

Samples frozen at −80°C were used for measurements of U-osm, plasma osmolality (P-osm) and plasma creatinine. For osmolality measurements, urine/plasma samples were quickly thawed at 37°C [30] and soon after read by an automated osmometer (2020, Advanced Instruments Inc., Norwood, MA, USA). In agreement with previous data [30], subgroup data in 149 examinees indicated that the freezing/thawing procedure had minor effects on osmolality measurements (correlation coefficient between frozen/thawed samples and fresh samples = 0.963 for U-osm and 0.873 for P-osm; % difference between frozen/thawed samples and fresh samples = −5.3% for U-osm and −1.9% for P-osm). For creatinine measurements, previous data showed that creatinine is stable during long-term storage and after thawing/refreezing [31]. Plasma creatinine was measured by automated biochemistry (Express Plus, Bayer Diagnostic, Tarrytown, NY, USA) using a kinetic alkaline picrate assay with IDMS-traceable standardization for eGFR calculation [31]. Plasma total cholesterol, plasma uric acid and other lab variables were measured in fresh samples using automated biochemistry and quality controls [25–29]. Blood pressure and anthropometry were measured at each exam by standardized methods as reported [25–29].

U-osm and the U-osm/P-osm ratio were taken as indices of KCA [1–3, 32, 33]. Kidney function was indexed as eGFR [34], calculated by the equation of the Chronic Kidney Disease Epidemiology Collaboration research group [35, 36]. Hypertension was defined as systolic pressure ≥140 mmHg or diastolic pressure ≥ 90 mmHg or anti-hypertensive drug treatment. Hypercholesterolaemia was defined as plasma cholesterol ≥6.20 mmol/L or regular drug treatment for high plasma cholesterol. Diabetes was defined as reported diagnosis of diabetes or regular anti-diabetic treatment with diet, drugs or insulin. Hyperuricaemia was defined as plasma uric acid ≥420 µmol/L. Obesity was defined as body mass index ≥30 kg/m2.

Statistics

Analyses on the presence and the independence of cross-sectional associations were performed by analysis of variance (ANOVA) and multi-variable linear regression with U-osm, P-osm and U-osm/P-osm ratio alternatively used as dependent variable. Analyses about the prognostic power targeted mortality and change in kidney function over time. Single- and multi-variable Cox regression were used to investigate the presence and the independence of the association with mortality of baseline values of U-osm, P-osm and U-osm/P-osm ratio alternatively used as independent variables. Kidney function change over time was calculated as annualized eGFR change from baseline exam to follow-up exam, that is, as follow-up eGFR minus baseline eGFR divided by the interval duration between the baseline exam and the follow-up exam. Single- and multi-variable linear regression were used to investigate the presence and the independence of associations with annualized eGFR change of baseline values of U-osm, P-osm and U-osm/P-osm ratio alternatively used as independent variables. Multi-variable analyses about the prognostic power were re-run in separate subgroups to investigate the consistency of results and the existence of possible interactions. Statistical procedures were performed by IBM-SPSS Statistics 19. Data include 95% confidence interval (CI).

RESULTS

Descriptive statistics

The target cohort consisted of 4679 adult examinees at baseline (aged 18–97 years). The study cohort consisted of 4220 examinees with complete baseline data for variables included in the analysis (90.2% of the target cohort). Of these baseline examinees, 852 died before the end of the follow-up exam (person*years product = 67 290; annual mortality rate = 1.26%). Of the 3368 survivors, 2795 participated in the follow-up exam (response rate = 83.0%) and all of them had complete follow-up data. Table 1 reports the descriptive statistics at both the exams.

Table 1.

Descriptive statistics at baseline and follow-up: prevalence or mean ± SD

Baseline Follow-up
Date of exam 1988–92 2001–07
Number of persons 4220 2795
% men (n) 45.2% (1906) 43.7% (1221)
Age, years 50.2 ± 17.6 59.7 ± 13.4
 18–44 years, % (n) 38.4% (1620) 16.5% (460)
 45–64 years, % (n) 36.8% (1552) 46.3% (1293)
 ≥ 65 years, % (n) 24.8% (1048) 37.3% (1042)
Obesity, % (n) 20.2% (852) 21.6% (604)
Plasma creatinine, mg/dL 0.91 ± 0.14 0.91 ± 0.14
eGFR, mL/min × 1.73 m2 86.3±17.0 79.9±15.2
 ≥90, % (n) 43.4% (1831) 28.2% (788)
 89–60, % (n) 50.8% (2145) 62.6% (1751)
 <60, % (n) 5.8% (244) 9.2% (256)
Hypertension, % (n) 34.6% (1461) 52.1% (1457)
Hypercholesterolaemia, % (n) 30.7% (1295) 28.3% (791)
Smoking, % (n) 29.5% (1244) 21.8% (608)
Diabetes, % (n) 6.4% (272) 8.6% (241)
Hyperuricaemia, % (n) 6.0% (252) 8.5% (238)
On anti-hypertensive drug treatment, % (n) 17.4% (736) 39.8% (1113)
On diuretic treatment, % (n) 12.9% (544) 20.5% (574)
Urine flow,a mL/min 0.92 ± 0.49 n.m.
U-osm,a mosm/kg 739 ± 155 n.m.
P-osm,b mosm/kg 283.4 ± 6.1 n.m.
U-osm/P-osm ratio 2.61 ± 0.55 n.m.
Open in a new tab

n.m., not measured.

a

Timed urine collection from first void after evening meal to first morning wake-up void.

b

Early morning blood sample collected under fasting conditions.

At baseline, urinary albumin/creatinine ratio was measured only in the subgroup aged 45–64 years [mean ± standard deviation (SD) = 14.8 ± 65.6 mg/g] [37]. From the baseline exam to the follow-up exam, mean ± SD was 14.1 ± 2.2 years for duration of follow-up, −9.9 ± 10.1 mL/min × 1.73 m2 for absolute eGFR change and −0.74 ± 0.75 mL/min × 1.73 m2 for annualized eGFR change. The eGFR change was a decrease in 82.5% of the examinees. Descriptive statistics for ancillary analyses are presented in Supplementary data, Table S1.

Baseline cross-sectional associations

Figures 1–3 summarize the results of the non-adjusted ANOVA of U-osm, P-osm and U-osm/P-osm ratio by strata of age, eGFR and urine flow in men and women separately. Older age associated with lower U-osm, higher P-osm and lower U-osm/P-osm ratio in both sexes (Figure 1). Lower GFR associated with lower U-osm, higher P-osm and lower U-osm/P-osm ratio in both sexes (Figure 2). Higher urine flow associated with lower U-osm and lower U-osm/P-osm ratio in both sexes and with higher P-osm in women only (Figure 3). The general trend across strata of age, eGFR and urine flow was that, compared with men, women had lower values of U-osm, P-osm and U-osm/P-osm ratio with the exceptions of women aged ≥65 years (higher P-osm than men of the same age, Figure 1), women with eGFR ≥90 mL/min × 1.73 m2 (slightly higher U-osm/P-osm ratio than men of the same eGFR stratum, Figure 2) and women with urine flow in the lowest 25% of the distribution (slightly higher U-osm/P-osm ratio than men of the same urine flow stratum, Figure 3).

FIGURE 1.

FIGURE 1

Open in a new tab

Mean ± SE at baseline of U-osm, P-osm and U-osm/P-osm ratio by sex and stratum of age: 18–44 years, 45–64 years and ≥65 years. Men = closed circles and solid line; women = open circles and dotted line. SE is not shown when it did not exceed the size of the circles. P-values by non-adjusted ANOVA for the effect of age. Number of men and women, respectively, per age stratum: 18–44 years = 780 and 840; 45–64 years = 688 and 864; ≥65 years = 438 and 610.

FIGURE 3.

FIGURE 3

Open in a new tab

Mean ± SE at baseline of U-osm, P-osm and U-osm/P-osm ratio by sex and strata of urine flow: <0.574 (lowest quartile), 0.574–1.128 (intermediate quartiles) and ≥1.129 mL/min (highest quartile). Men = closed circles and solid line; women = open circles and dotted line. SE is not shown when it did not exceed the size of the circles. P-values by non-adjusted ANOVA for the effect of urine flow stratum. Number of men and women, respectively, per stratum of urine flow: lowest quartile (<0.574 mL/min) = 388 and 667; intermediate quartiles (0.574–1.128 mL/min) = 1055 and 1055; highest quartile (≥1.129 mL/min) = 463 and 592.

FIGURE 2.

FIGURE 2

Open in a new tab

Mean ± SE at baseline of U-osm, P-osm and U-osm/P-osm ratio by sex and stratum of eGFR: <60, 60–89, and ≥90 mL/min × 1.73 m2. Men = closed circles and solid line; women = open circles and dotted line. SE is not shown when it did not exceed the size of the circles. P-values by non-adjusted ANOVA for the effect of eGFR stratum. Number of men and women, respectively, per eGFR stratum: <60 mL/min × 1.73 m2 = 58 and 186; 60–89 mL/min × 1.73 m2 = 780 and 1365; ≥ 90 mL/min × 1.73 m2 = 1068 and 763.

Figure 4 summarizes the results for both sexes combined together of the non-adjusted ANOVA of U-osm, P-osm and U-osm/P-osm ratio between hypertensives and non-hypertensives, between examinees with hypercholesterolaemia and examinees without hypercholesterolaemia, between smokers and non-smokers, between diabetics and non-diabetics, between obese examinees and non-obese examinees and between examinees with hyperuricaemia and examinees without hyperuricaemia. Hypertension, hypercholesterolaemia, diabetes, obesity and hyperuricaemia were associated with lower U-osm, higher P-osm and lower U-osm/P-osm ratio. Vice versa, smoking was associated with higher U-osm, lower P-osm and higher U-osm/P-osm ratio. Findings were similar in analyses in men and women separately (data not shown).

FIGURE 4.

FIGURE 4

Open in a new tab

Mean ± SE at baseline of U-osm, P-osm and U-osm/P-osm ratio by presence/absence of hypertension, hypercholesterolaemia, smoking, diabetes, obesity and hyperuricaemia (presence = dark grey bar; absence = light grey bar). P-values by non-adjusted ANOVA for comparison of the presence of a given disorder to its absence. Number of persons per group is shown in Table 1 or can be derived from data shown in Table 1.

Table 2 summarizes the results of multi-variable regression analyses. Female sex independently associated with lower values of U-osm, P-osm and U-osm/P-osm ratio. Older age independently associated with lower U-osm, higher P-osm and lower U-osm/P-osm ratio. Obesity independently associated with higher values of U-osm, P-osm and U-osm/P-osm ratio. Lower eGFR independently associated with lower U-osm, higher P-osm and lower U-osm/P-osm ratio. Hypertension and smoking independently associated with lower values of U-osm and U-osm/P-osm ratio, but not with different P-osm. Higher urine flow independently associated with lower values of U-osm, P-osm and U-osm/P-osm ratio. Diabetes, hypercholesterolaemia and hyperuricaemia did not independently associate with any of the three dependent variables. Findings were similar after exclusion of the examinees reporting treatment with anti-hypertensive drugs or with diuretics (data not shown). Multi-variable regression gave similar results in ancillary cross-sectional analyses of data from a previous exam with urine/blood samples collected in different times of the day (Supplementary data, Table S2).

Table 2.

Multi-variable regression models for analyses on cross-sectional associations at baseline of sex, age, eGFR and other independent variables with U-osm, P-osm and U-osm/P-osm ratio (dependent variable)

Independent variable Dependent variable
U-osm
 P-osm
 U-osm/P-osm ratio
beta P beta P beta P
Sex (men/women = 1/2) −0.102 <0.001 −0.177 <0.001 −0.082 <0.001
Age −0.050 0.017 0.129 <0.001 −0.063 0.002
Obesity (yes/no = 1/0) 0.096 <0.001 0.043 0.005 0.091 <0.001
eGFR 0.144 <0.001 −0.151 <0.001 0.158 <0.001
Hypertension (yes/no = 1/0) −0.042 0.003 0.008 0.642 −0.042 0.003
Smoking (yes/no = 1/0) −0.026 0.038 −0.016 0.288 −0.024 0.046
Urine flow −0.564 <0.001 −0.047 0.003 −0.556 <0.001
R2 0.408 0.089 0.413
Open in a new tab

Results are shown as partial correlation coefficient (beta), P and R squared (R2).

Presence/absence at baseline of diabetes, hypercholesterolaemia and hyperuricaemia were included in the models, but did not significantly associate with dependent variables in any analysis (absolute values of beta ≤0.020).

Prognostic power for mortality

Table 3 summarizes the results of Cox regression for the analyses on the association with mortality of three different independent variables: baseline U-osm, baseline P-osm and baseline U-osm/P-osm ratio. For each one of these independent variables, four different models were analysed as follows: model 1 without control for any covariate (non-adjusted model) and models 2–4 with control for covariates as listed in Table 3 (multi-variable models). U-osm and U-osm/P-osm ratio inversely associated with mortality in all models (i.e. higher values associated with lower mortality and vice versa). Survival curves by tertile of U-osm/P-osm ratio showed that the association of baseline U-osm/P-osm ratio with mortality was constant over time both with and without controlling for other variables (Figure 5). Survival curves were similar for baseline U-osm (data not shown). P-osm associated with mortality in the non-adjusted model only. Findings were similar when the analyses were repeated with plasma sodium in the place of P-osm to get rid of the confounding of plasma urea and plasma glucose on P-osm (data not shown). Findings about the prognostic power for mortality were similar after excluding the examinees reporting treatment with anti-hypertensive drugs or diuretics, and in ancillary analyses of data from another exam with urine/blood samples collected at different times of the day (Supplementary data, Table S3).

Table 3.

Cox regression models for analyses about the associations of baseline values of U-osm, P-osm and U-osm/P-osm ratio (independent variables) with mortality

Model Covariates in the model Independent variable
U-osm
 P-osm
 U-osm/P-osm ratio
HR (95% CI) P HR (95% CI) P HR (95% CI) P
1 None (non-adjusted) 0.65 (0.61–0.69) <0.001 1.35 (1.27–1.44) <0.001 0.63 (0.60–0.67) <0.001
2 Sex and baseline age 0.91 (0.85–0.98) 0.009 1.04 (0.97–1.12) 0.298 0.91 (0.84–0.97) 0.006
3 Sex, baseline age and baseline eGFR 0.92 (0.86–0.99) 0.019 1.03 (0.96–1.11) 0.458 0.91 (0.85–0.98) 0.014
4 Sex, baseline age, baseline eGFR and other variablesa 0.91 (0.84–0.97) 0.007 1.03 (0.95–1.10) 0.507 0.90 (0.84–0.97) 0.005
Open in a new tab

Results are shown as mortality HR for one SD higher independent variable and 95% CI in four different models: model 1 without control (non-adjusted) and models 2–4 with control for listed covariates.

Baseline SD of U-osm, P-osm and U-osm/P-osm ratio are shown in Table 1.

a

Other covariates included: presence/absence at baseline of hypertension, hypercholesterolaemia, smoking, diabetes, obesity and hyperuricaemia.

FIGURE 5.

FIGURE 5

Open in a new tab

Survival curves by tertile of baseline U-osm/P-osm ratio. Cox regression for men and women combined in non-adjusted model (left panel) and in model adjusted for sex, baseline age, baseline eGFR and presence/absence at baseline of hypertension, hypercholesterolaemia, smoking, diabetes, obesity and hyperuricaemia (right panel). Tertiles 1, 2 and 3 were defined using the following thresholds of baseline U-osm/P-osm ratio: <2.42, 2.42–2.85 and >2.85. Number of persons per tertile = 1407, 1406 and 1407; number of death per tertile = 446, 270 and 136, for tertiles 1, 2 and 3, respectively.

In multi-variable Cox regression with control for sex, age, eGFR and other variables, one SD higher baseline U-osm/P-osm ratio consistently associated with hazard ratio (HR) of mortality <1 in the following subgroups: men and women (HR = 0.89 and 0.92), examinees both below and above the median age (HR = 0.87 and 0.91), obese and non-obese examinees (HR = 0.83 and 0.93), examinees with eGFR <90 mL/min × 1.73 m2 and examinees with eGFR ≥90 mL/min × 1.73 m2 (HR = 0.90 and 0.91), hypertensive and non-hypertensive examinees (HR = 0.86 and 0.96), examinees with hypercholesterolaemia and examinees without hypercholesterolaemia (HR = 0.91 and 0.90), smokers and non-smokers (HR = 0.78 and 0.95), diabetics and non-diabetics (HR = 0.77 and 0.93), examinees with hyperuricaemia and examinees without hyperuricaemia (HR = 0.76 and 0.92), examinees with urine flow below the median and examinees with urine flow above the median (HR = 0.88 and 0.89) and examinees who at baseline were with age <65 years, with eGFR ≥60 mL/min × 1.73 m2, non-diabetic and non-hypertensive (HR = 0.92). Findings were also consistent when the Cox model for examinees with baseline age 45–64 years was controlled for the urinary/albumin to creatinine ratio (HR = 0.79). Findings were similar for baseline U-osm (data not shown).

Prognostic power for the change of kidney function over time

Table 4 summarizes the results of linear regression analyses of the association with annualized eGFR change of three different independent variables: baseline U-osm, baseline P-osm and baseline U-osm/P-osm ratio. For each one of these independent variables, four different models were analysed as follows: model 1 without control for any covariate (non-adjusted model) and models 2–4 with control for covariates as listed in Table 4 (multi-variable models). In non-adjusted models, baseline values of U-osm and U-osm/P-osm ratio directly associated with annualized eGFR change (i.e. higher values associated with less negative eGFR change). These associations were not significant when controlling for sex, baseline age, baseline eGFR and other variables. Baseline differences in P-osm associated with annualized eGFR change only in model 2. Findings were similar when analyses were repeated with plasma sodium in the place of P-osm to eliminate the confounding effect of plasma urea and plasma glucose on P-osm (data not shown). Findings were similar also when excluding the examinees reporting treatment with anti-hypertensive drugs or diuretics (data not shown). Figure 6 shows the annualized eGFR change by tertile of baseline U-osm/P-osm ratio in non-adjusted ANOVA and in ANOVA with control for other variables. Findings were similar for baseline U-osm and in the analyses for the subgroups listed above (data not shown).

Table 4.

Linear regression for analyses about the associations of baseline values of U-osm, P-osm and U-osm/P-osm ratio (independent variables) with annualized eGFR change from baseline exam to follow-up exam (dependent variable)

Model Covariates in the model Independent variable
U-osm
 P-osm
 U-osm/P-osm ratio
beta P beta P beta P
1 None (non-adjusted model) 0.059 0.002 −0.002 0.927 −0.002 0.002
2 Sex and baseline age −0.005 0.791 0.058 0.002 0.058 0.554
3 Sex, baseline age and baseline eGFR 0.011 0.487 0.003 0.832 0.003 0.493
4 Sex, baseline age, baseline eGFR and other variablesa 0.012 0.456 0.004 0.795 0.004 0.465
Open in a new tab

Results are shown as partial correlation coefficient (beta) in four different models: model 1 without control for co-variates (non-adjusted) and models 2–4 with control for listed covariates.

a

Other covariates included: presence/absence at baseline of hypertension, hypercholesterolaemia, smoking, diabetes, obesity and hyperuricaemia.

FIGURE 6.

FIGURE 6

Open in a new tab

Mean ± SE of annualized eGFR change from baseline to follow-up plotted over mean U-osm/P-osm ratio of tertiles of baseline urine/plasma osmolality ratio in non-adjusted ANOVA (left panel) and in ANOVA adjusted for sex, baseline age, baseline eGFR and presence/absence at baseline of hypertension, hypercholesterolaemia, smoking, diabetes, obesity and hyperuricaemia (right panel). P-values for comparison among tertiles. Tertiles were defined using the same thresholds of baseline U-osm/P-osm ratio used in analyses of Figure 5. Number of persons per tertile = 783, 953 and 1059.

DISCUSSION

Cross-sectional results of the study were that sex, age, obesity, kidney function, urine flow, hypertension and smoking independently associated with osmotic indices or KCA. Longitudinal results of the study were that KCA independently predicted mortality, but did not predict kidney function change over time.

Compared to previous papers, these results confirmed and extended the reports of cross-sectional associations of sex, age and kidney function with KCA [2–11]. Other results of the study were novel and could not be compared to the results of previous papers.

With regard to the interpretation of cross-sectional results, female sex and higher urine flow independently associated with lower U-osm, lower P-osm and lower U-osm/P-osm ratio. The association of female sex with lower KCA and lower P-osm could be explained by effects of female hormones on extracellular volume (expansion) and antidiuretic hormone secretion (reduction) [5, 6], hence by a condition of relative hypotonic hypervolaemia. Higher urine flow reflected higher hydration level because it associated with lower P-osm after the completion of the urine collection. Thus, as for female sex, the association of higher urine flow with lower KCA was likely explained by a relative hypotonic hypervolaemia [38]. These interpretations were in accordance with the previous evidence of lower indices of antidiuretic hormone secretion in women and in persons with higher urine flow [39, 40]. Obesity independently associated with higher U-osm, higher P-osm and higher U-osm/P-osm ratio. An obesity-related excess in the intake or in the metabolic generation of osmolytes could be the underlying mechanism. An overload of osmolytes could theoretically up-regulate P-osm and, secondarily, KCA. Older age and lower eGFR independently associated with lower values of U-osm and U-osm/P-osm ratio, but with higher P-osm. Renal alterations at the tubulo-interstitial level have been reported in the elderly and in patients with low kidney function [12, 41, 42]. Such alterations could reduce the urine concentration activity of the distal nephron and could thus explain the association of older age and lower eGFR with low KCA. The possibility of a low secretion of antidiuretic hormone was unlikely given the report of normal-to-high indices of antidiuretic hormone secretion in older ages and in individuals with low kidney function [40]. The possibility of low renal sensitivity to antidiuretic hormone could not be excluded. The association with higher P-osm likely reflected different mechanisms for older age and for lower eGFR. An ageing-related reduction in thirst could be the major determinant of the association between older age and higher P-osm, although it could not be excluded that a concomitant reduction in KCA and a relative renal water leak favoured the trend towards higher P-osm [12]. In contrast, high plasma urea should play a major role in the association of lower eGFR with higher P-osm given that urea concentration is a determinant of P-osm and increases when kidney function decreases. Hypertension and smoking independently associated with lower U-osm and lower U-osm/P-osm ratio but not with altered P-osm. As proposed for older age and lower eGFR, the presence of renal alterations at the tubulo-interstitial level was a likely underlying mechanism also for hypertension and smoking because such alterations can be induced by hypertension [43–45] and smoking [46]. The natriuretic effects of anti-hypertensive drugs could play a role in treated hypertensives because sodium is a major urine osmolyte [32, 33]. The possibility of low KCA due to low secretion of antidiuretic hormone appeared unlikely because hypertension and smoking associated with higher indices of antidiuretic hormone secretion in previous reports [40, 47].

With regard to the interpretation of longitudinal results, U-osm and U-osm/P-osm ratio independently predicted the mortality risk in the Gubbio cohort. Previous reports showed that markers of kidney dysfunction such as eGFR and albuminuria independently predicted the mortality risk in the Gubbio cohort similar to other samples of the general population [37, 48]. Thus, given that KCA is regarded as an index of distal nephron functions, the present results suggested that a dysfunction of the distal nephron could be an additional predictor of the risk of mortality. The novelty of the finding and its observational characteristics limited the inference about underlying mechanisms, as for other renal indices predicting the risk of mortality [48]. Low KCA could be either a mere marker or a true determinant of mortality risk. Theoretically, low KCA could reflect a kidney damage at the tubulo-interstitial level and/or at the non-glomerular vascular level, which can be adequately assessed neither by renal indices like eGFR or albuminuria nor by the presence and by the severity of other risk factors. Alternatively, low KCA could affect the homoeostasis of substances that are handled by the distal nephron and that play a role in the mortality risk. A further possible explanation could be that low KCA implied a compensatory increase in the secretion of antidiuretic hormone which, in turn, associated with unfavourable effects on mortality [49]. The lack of an independent association of P-osm with mortality did not support mechanisms involving a renal water leak and a secondary trend towards a relative hypertonic dehydration per se. Finally, the lack of an independent longitudinal association of KCA with eGFR change over time supported the view that the cross-sectional association between baseline eGFR and baseline KCA reflected an influence of lower eGFR on KCA rather than the opposite.

The practical implications of the results concerned the clinical evaluation of P-osm and kidney disease. The present results indicated that demographics, obesity and kidney function account for part of the variability in P-osm. If other studies will confirm the association between KCA and mortality, a KCA assessment should be included in the diagnostic work-up required for the definition of the prognosis of kidney disease.

The main limitations of this study were the observational design, the lack of data for 24-h urine, the measurement of antidiuretic hormone and various ethnic groups. The observational design limited the assessment of cause–effect relationships, which can hardly be investigated in humans for factors like gender and age. The study design also limited the possibility to define the temporal sequence leading to lower KCA and its relationships with eGFR reduction. The lack of 24-h urine was compensated for by the ancillary analyses, which showed similar results using data from another exam with urine collections at different times of the day. The lack of data for antidiuretic hormone could affect the interpretation of results but not their significance. The lack of information for various ethnic groups reflected the structure of the population sample under study.

Briefly, this observational population-based study reported on KCA indices and P-osm. KCA and P-osm were concurrently lower in females and in individuals with higher urine flow, whereas they were concurrently higher in obese examinees. Thus, the associations with KCA of sex, urine flow and obesity appeared as KCA adaptations secondary to differences in P-osm, that is, to differences in the hydration level or in the osmolyte load. Vice versa, the associations with KCA of older age, lower kidney function, hypertension and smoking appeared independent of P-osm and reasonably reflected intrinsic renal alterations due to or related to these traits. KCA independently predicted the mortality risk, suggesting that its assessment could be useful for the prognosis of kidney disease.

SUPPLEMENTARY DATA

Supplementary data are available online at http://ndt.oxfordjournals.org.

Supplementary Material

Supplementary Data
Click here for additional data file. (27.3KB, docx)

ACKNOWLEDGEMENTS

The Gubbio Study was made possible thanks to the enthusiasm of the people of Gubbio and to the support of its municipal and health authorities and community leaders. Economic support was given in the past by Merck, Sharp & Dohme – Italy, US National Heart, Lung, and Blood Institute (Grant R01HL-40397-02), and Ministero Italiano di Università e Ricerca (Grant # 068034). The study was realized in collaboration with Gubbio Civil Hospital, ‘Federico II’ University of Naples, Italy, University of Milan, Italy, Northwestern University of Chicago, Illinois, Istituto Superiore di Sanità, Italy, University of Salerno, Italy. None of the sponsors had any role in study design; collection, analysis and interpretation of data; writing the report; and the decision to submit the report for publication.

CONFLICT OF INTEREST STATEMENT

The results presented in this article have not been published previously in whole or part, except in abstract format. There is no conflict of interest to disclose.

REFERENCES

  • 1. Sands JM, Layton HE, Fenton RA.. Urine concentration and dilution In: Tall MWChertow GMMarsden PA et al. (eds). Brenner & Rector’s The Kidney, vol. 1 Philadelphia, PA: Elsevier Saunders, 2012, 326–352 [Google Scholar]
  • 2. Harrington JT, Cohen JJ.. Clinical disorders of urine concentration and dilution. Arch Intern Med 1973; 131: 810–825 [PubMed] [Google Scholar]
  • 3. Perucca J, Bouby N, Valeix P. et al. Sex difference in urine concentration across differing ages, sodium intake, and level of kidney disease. Am J Physiol Regul Integr Comp Physiol 2007; 292: R700–R705 [DOI] [PubMed] [Google Scholar]
  • 4. Pedersen EB, Thomsen IM, Lauridsen TG.. Abnormal function of the vasopressin-cyclic-AMP-aquaporin2 axis during urine concentrating and diluting in patients with reduced renal function. A case control study. BMC Nephrol 2010; 11: 26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Stachenfeld NS, Keefe DL.. Estrogen effects on osmotic regulation of AVP and fluid balance. Am J Physiol Endocrinol Metab 2002; 283: E711–E721 [DOI] [PubMed] [Google Scholar]
  • 6. Wenner MM, Stachenfeld NS.. Blood pressure and water regulation: understanding sex hormone effects within and between men and women. J Physiol 2012; 590: 5949–5961 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Hyistendhal GM, Frokiaer J, Nielsen S. et al. Gender differences in nighttime plasma arginine vasopressin and delayed compensatory urine output in the elderly population after desmopressin. J Urol 2007; 178: 2671–2676 [DOI] [PubMed] [Google Scholar]
  • 8. Perinpam M, Ware EB, Smith JA. et al. Key influence of sex on urine volume and osmolality. Biol Sex Differ 2016; 7: 12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Phillips PA, Johnston CI, Gray L.. Disturbed fluid and electrolyte homoeostasis following dehydration in elderly people. Age Ageing 1993; 22: S26–S33 [DOI] [PubMed] [Google Scholar]
  • 10. Sands JM. Urinary concentration and dilution in the aging kidney. Semin Nephrol 2009; 29: 579–586 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Sands JM. Urine concentrating and diluting ability during aging. J Gerontol A Biol Sci Med Sci 2012; 67: 1352–1357 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Cowen LE, Hodak SP, Verbalis JG.. Age-associated abnormalities of water homeostasis. Endocrinol Metab Clin North Am 2013; 42: 349–370 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Perrier E, Demazieres A, Girard N. et al. Circadian variation and responsiveness of hydration biomarkers to changes in daily water intake. Eur J Appl Physiol 2013; 113: 2143–2151 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Hillier P, Knapp MS, Cove-Smith R.. Circadian variations in urine excretion in chronic renal failure. Q J Med 1980; 49: 461–478 [PubMed] [Google Scholar]
  • 15. Fujii T, Uzu T, Nishimura M. et al. Circadian rhythm of natriuresis is disturbed in nondipper type of essential hypertension. Am J Kidney Dis 1999; 33: 29–35 [DOI] [PubMed] [Google Scholar]
  • 16. Fukuda M, Munemura M, Usami T. et al. Nocturnal blood pressure is elevated with natriuresis and proteinuria as renal function deteriorates in nephropathy. Kidney Int 2004; 65: 621–625 [DOI] [PubMed] [Google Scholar]
  • 17. Sachdeva A, Weder AB.. Nocturnal sodium excretion, blood pressure dipping, and sodium sensitivity. Hypertension 2006; 48: 527–533 [DOI] [PubMed] [Google Scholar]
  • 18. Fukuda M, Motokawa M, Miyagi S. et al. Polynocturia in chronic kidney disease is related to natriuresis rather than to water diuresis. Nephrol Dial Transplant 2006; 21: 2172–2177 [DOI] [PubMed] [Google Scholar]
  • 19. Burnier M, Coltamai L, Maillard M. et al. Renal sodium handling and nighttime blood pressure. Semin Nephrol 2007; 27: 565–571 [DOI] [PubMed] [Google Scholar]
  • 20. Fukuda M, Yamanaka T, Mizuno M. et al. Angiotensin II type 1 receptor blocker, olmesartan, restores nocturnal blood pressure decline by enhancing daytime natriuresis. J Hypertens 2008; 26: 583–588 [DOI] [PubMed] [Google Scholar]
  • 21. Miura T, Fukuda M, Naito T. et al. Circadian rhythm of urinary potassium excretion in patients with CKD. Clin Nephrol 2012; 78: 169–173 [DOI] [PubMed] [Google Scholar]
  • 22. Fezeu L, Bankir L, Hansel B. et al. Differential circadian pattern of water and Na excretion rates in the metabolic syndrome. Chronobiol Int 2014; 31: 861–867 [DOI] [PubMed] [Google Scholar]
  • 23. Mahmoodi BK, Matsushita K, Woodward M. et al. Associations of kidney disease measures with mortality and end-stage renal disease in individuals with and without hypertension: a meta-analysis. Lancet 2012; 380: 1649–1661 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Nitsch D, Grams M, Sang Y. et al. Associations of estimated glomerular filtration rate and albuminuria with mortality and renal failure by sex: a meta-analysis. BMJ 2013; 346: f324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Laurenzi M, Cirillo M, Angeletti M. et al. Gubbio population study: baseline findings. Nutr Metab Cardiovasc Dis 1991; 1: S1–S8 [Google Scholar]
  • 26. Cirillo M, Terradura-Vagnarelli O, Mancini M. et al. Cohort profile: the Gubbio population study. Int J Epidemiol 2014; 43: 713–720 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Laurenzi M, Cirillo M, Panarelli W. et al. Baseline sodium-lithium countertransport and 6-year incidence of hypertension – The Gubbio Population Study. Circulation 1997; 95: 581–587 [DOI] [PubMed] [Google Scholar]
  • 28. Cirillo M, Ciacci C, De Santo NG.. Age, renal tubular phosphate reabsorption, and serum phosphate levels in adults. N Engl J Med 2008; 359: 864–866 [DOI] [PubMed] [Google Scholar]
  • 29. Cirillo M, Laurenzi M, Panarelli W. et al. Urinary sodium to potassium ratio and urinary stone disease. Kidney Int 1994; 46: 1133–1139 [DOI] [PubMed] [Google Scholar]
  • 30. Seifarth CC, Miertschischk J, Hahn EG. et al. Measurement of serum and plasma osmolality in healthy young humans – Influence of time and storage conditions. Clin Chem Lab Med 2004; 42: 927–932 [DOI] [PubMed] [Google Scholar]
  • 31. Myers GL, Miller WG, Coresh J. et al. Recommendations for improving serum creatinine measurement: a report from the Laboratory Working Group of the National Kidney Disease Education Program. Clin Chem 2006; 52: 5–18 [DOI] [PubMed] [Google Scholar]
  • 32. Mayo Clinic – Mayo Medical Laboratories. Interpretive Handbook, Test 9260: Osmolality, Urine, Clinical Information http://www.mayomedicallaboratories.com/interpretive-guide/?alpha=O&unit_code=9260 (12 September 2016, date last accessed)
  • 33. Scott MG, LeGrys VA, Hood JL.. Electrolytes and blood gases In: Burtis CAAshwood ERBruns DE (eds). Tietz Textbook of Clinical Chemistry and Molecular Diagnostics. St Louis, MO: Elsevier Saunders, 2012, 807–835 [Google Scholar]
  • 34. Levey AS, Coresh J, Balk E. et al. National Kidney Foundation practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Ann Intern Med 2003; 139: 137–147 [DOI] [PubMed] [Google Scholar]
  • 35. Levey AS, Stevens LA, Schmid CH. et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009; 150: 604–612 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Cirillo M, Lombardi C, Luciano MG. et al. Estimation of GFR: a comparison of new and established equations. Am J Kidney Dis 2010; 56: 802–804 [DOI] [PubMed] [Google Scholar]
  • 37. Cirillo M, Lanti MP, Menotti A. et al. Definition of kidney dysfunction as a cardiovascular risk factor: use of urinary albumin excretion and estimated glomerular filtration rate. Arch Intern Med 2008; 168: 617–624 [DOI] [PubMed] [Google Scholar]
  • 38. Anastasio P, Cirillo M, Spitali L. et al. Level of hydration and renal function in healthy humans. Kidney Int 2001; 60: 748–756 [DOI] [PubMed] [Google Scholar]
  • 39. Bhandari SS, Loke I, Davies JE. et al. Gender and renal function influence plasma levels of copeptin in healthy individuals. Clin Sci 2009; 116: 257–263 [DOI] [PubMed] [Google Scholar]
  • 40. van Gastel MD, Meijer E, Scheven LE. et al. Modifiable factors associated with copeptin concentration: a general population cohort. Am J Kidney Dis 2015; 65: 719–727 [DOI] [PubMed] [Google Scholar]
  • 41. Lamb EJ, O’Riordan SE, Delaney MP.. Kidney function in older people: pathology, assessment and management. Clin Chim Acta 2003; 334: 25–40 [DOI] [PubMed] [Google Scholar]
  • 42. Theilig F. Spread of glomerular to tubulointerstitial disease with a focus on proteinuria. Ann Anat 2010; 192: 125–132 [DOI] [PubMed] [Google Scholar]
  • 43. Fine LG, Orphanides C, Norman JT.. Progressive renal disease: the chronic hypoxia hypothesis. Kidney Int Suppl 1998; 65: S74–S78 [PubMed] [Google Scholar]
  • 44. Choi YJ, Chakraborty S, Nguyen V. et al. Peritubular capillary loss is associated with chronic tubulointerstitial injury in human kidney: altered expression of vascular endothelial growth factor. Hum Pathol 2000; 31: 1491–1497 [DOI] [PubMed] [Google Scholar]
  • 45. Meyrier A. Nephrosclerosis: update on a centenarian. Nephrol Dial Transplant 2015; 30: 1833–1841 [DOI] [PubMed] [Google Scholar]
  • 46. Odoni G, Ogata H, Viedt C. et al. Cigarette smoke condensate aggravates renal injury in the renal ablation model. Kidney Int 2002; 61: 2090–2098 [DOI] [PubMed] [Google Scholar]
  • 47. Finch CK, Andrus MR, Curry WA.. Nicotine replacement therapy-associated syndrome of inappropriate antidiuretic hormone. South Med J 2004; 97: 322–324 [DOI] [PubMed] [Google Scholar]
  • 48. Chronic Kidney Disease Prognosis Consortium, Matsushita K, van der Velde M. et al. Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet 2010; 375: 2073–2081 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49. Odermatt J, Bolliger R, Hersberger L. et al. Copeptin predicts 10-year all-cause mortality in community patients: a 10-year prospective cohort study. Clin Chem Lab Med 2016; 54: 1681–1690 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
Click here for additional data file. (27.3KB, docx)

Articles from Nephrology Dialysis Transplantation are provided here courtesy of Oxford University Press

RESOURCES