Abstract
Background
While higher blood pressure is known to increase proteinuria, whether increased dietary sodium as estimated from 24-hour urinary excretion correlates with increased proteinuria in patients with chronic kidney disease (CKD) is not well studied.
Methods
We measured 24-hour urine sodium, potassium, and protein excretion in 3,680 participants in the Chronic Renal Insufficiency Cohort (CRIC) study, to determine the relationship between urinary sodium and potassium and urinary protein excretion in patients with CKD. We stratified our data based on the presence or absence of diabetes given the absence of any data on this relationship, and evidence that diabetics had greater urinary protein excretion at nearly every level of urinary sodium excretion. Multiple linear regressions were used with a stepwise inclusion of covariates such as systolic blood pressure (SBP), demographics, hemoglobin A1C, and type of antihypertensive medications to evaluate the relationship between urinary electrolyte excretion and proteinuria.
Results
Our data demonstrated that urinary sodium (+1SD above the mean), as a univariate variable, explained 12% of the variation in proteinuria (β=0.29, p<.0001) with rising urinary sodium excretion associated with increasing proteinuria. The significance of that relationship was only partially attenuated with adjustment for demographic and clinical factors and the addition of 24-hour urinary potassium to the model (β=0.13, R2=0.35, p<.0001).
Conclusions
An understanding of the relationship between these clinical factors and dietary sodium may allow a more tailored approach for dietary salt restriction in patients with CKD.
Keywords: proteinuria, sodium, potassium, blood pressure
Introduction
Data from observational epidemiological studies and randomized controlled clinical trials demonstrate an association between dietary sodium intake and increased blood pressure, with few exceptions [1–5]. Moreover, in small clinical studies in patients with diabetic and non-diabetic kidney disease and proteinuria, there is evidence that increasing dietary salt intake offsets both the antihypertensive and antiproteinuric effect of drugs that block the renin-angiotensin system [6, 7], whether used alone or in conjunction with thiazide diuretics [7]. The influence of other clinical and demographic factors which may affect the relationship of dietary salt and proteinuria has also not been well studied.
Higher dietary salt intake may affect the kidneys through both blood pressure- dependent and blood pressure-independent effects [8, 9]. Although the mechanisms of direct effect of dietary salt on the kidney is not well understood, experimental studies indicate that there may be a relationship between salt and endothelial dysfunction, primarily through an increase in oxidative stress [10], with less production of endogenous nitric oxide [11]. Structural changes of vascular beds in the brains of spontaneously hypertensive rats are also affected by dietary salt, independent of any change in systemic blood pressure [12].
Given the known relationships between increasing dietary salt and increasing blood pressure, particularly in individuals with chronic kidney disease (CKD), and the evidence from small clinical studies indicating that increasing dietary sodium intake offsets both the antihypertensive and antiproteinuric effects of drugs that block the renin-angiotensin system, it is important to understand the contribution of demographics, other medical comorbidities, and anti-hypertensive treatment, with this important dietary electrolyte on proteinuria in a large cohort of patients with CKD. Thus, we performed multiple linear regressions in this large, well characterized cohort to examine the contribution of plausible demographic and clinical variables and antihypertensive treatment that could influence the relationship of urinary sodium on 24 hour urine protein excretion. These observations may be particularly important in formulating healthcare recommendations regarding diet in specific populations of individuals with CKD.
Methods
Study population
The CRIC study included a racially and ethnically diverse group of adults 21–74 years of age, with both diabetic and non-diabetic kidney disease. Details on the assembly of the cohort are described elsewhere [13]. Inclusion in the CRIC study was based on age- specific estimated glomerular filtration rate (eGFR) levels as following: eGFR of 20–70 ml/min/1.73m2 for participants aged 21–44 years, 20–60 ml/min/1.73m2 for participants aged 45–64, and 20–50 ml/min/1.73m2 for participants aged 65–74. A total of 3,939 participants were recruited into the CRIC study, of which 3,680 had complete 24-hour urine collections. CRIC participants were recruited between June 2003 and March 2007 from 13 sites and 7 centers in the United States (Baltimore, Maryland; Philadelphia, Pennsylvania; Cleveland, Ohio; Ann Arbor/Detroit, Michigan; Chicago, Illinois; New Orleans, Louisiana; and Oakland/San Francisco, California).
Procedures
At each CRIC study visit, several core measurements were ascertained and have been described elsewhere [13]. Most pertinent to this study, three brachial blood pressure measures were obtained in the sitting position after at least five minutes of quiet rest by trained and certified staff according to standard protocol. A Tyco aneroid sphygmomanometer was used with cuff size based on the participant’s arm circumference. Participants were advised to refrain from coffee, tea, or alcohol intake, cigarette smoking and vigorous exercise for at least 30 minutes before their examination. All blood pressure observers were certified for the blood pressure measurement protocol.
A 24-hour urine collection for protein and creatinine was obtained from each participant on the day of the blood pressure measures. Protocol-specified laboratory measurements were obtained on participants at each annual visit. Laboratory testing was performed at the central laboratory of the University of Pennsylvania. The eGFR was determined according to the abbreviated MDRD formula [14], using creatinine values calibrated to the Cleveland Clinic laboratory. Ethnicity was self-described by each participant.
Statistical Analyses
Continuous variables are presented as mean +/− standard deviation or (95% confidence interval). Categorical variables are expressed as N (%). Certain variables (urinary sodium and clinical data) were pre-specified in the basic model. In the analyses, we initially fit a joint model including both diabetics and non-diabetics. The log-transformed 24-hour urine protein was designated as the outcome variable in this cross-sectional analysis and measures including systolic and diastolic blood pressure, demographics, glycohemoglobin, type of antihypertensive medication(s), and urinary potassium were considered as key predictors, because of their possible relationship to proteinuria. Multiple linear regressions were performed sequentially. First, urinary sodium and clinical site were included in the model. Next, we included blood pressure, followed by several demographic and clinical variables known to be associated with proteinuria were designated as key confounders, including age, ethnicity, gender, estimated GFR, smoking status, and waist circumference. Subsequently, glucose levels, glycohemoglobin, antihypertensive medications (diuretics, renin-angiotensin system blockers, and calcium channel blockers), 24-hour urine potassium, and urinary sodium: potassium ratio was entered into the regression model. We included the urinary sodium: potassium ratio as a variable, as prior studies have indicated that higher urinary potassium especially patients with higher urinary sodium may be associated with lower blood pressure and target organ injury [15]. All analyses were subsequently stratified based on diabetic status, given the sufficient numbers of participants to examine whether there may be differences in the relationship between urine electrolytes and proteinuria in diabetic versus non-diabetic kidney disease. Figure 1 was obtained from a loess (locally weighted polynomial regression) fit. Confidence intervals were computed locally assuming linear relationships, independent errors, and normality. R-squared is the proportion of variability that is accounted for by the model. All analyses were executed in SAS 9.2 (SAS Institute, Cary, NC).
Figure 1.
This figure depicts the unadjusted data of the 24-hour urine protein excretion in grams, in relation to the 24 hour urine excretion of sodium in millimoles for both diabetes and non-diabetics. Confidence intervals are represented by dotted lines. Vertical lines at bottom of figure indicate the number of patients.
Results
Demographic characteristics (Table 1)
Table 1.
Clinical and Demographic Characteristics
| Non-Diabetics | Diabetics | ||||||
|---|---|---|---|---|---|---|---|
| All n=3673 |
UPr <=0.15 n=1094 |
UPr >0.15 n=813 |
p | UPr <=0.15 n=606 |
UPr >0.15 n=1160 |
p | |
| Demographics / Descriptive | . | ||||||
| Age (years) | 58.4 (10.9) | 59.6 (10.4) | 54.1 (12.8) | <.001 | 61.7 (9.1) | 58.5 (9.8) | <.001 |
| Male | 2020 (55%) | 511 (46.6%) | 516 (63.3%) | <.001 | 267 (44%) | 726 (62.5%) | <.001 |
| Female | 1660 (45%) | 585 (53.4%) | 299 (36.7%) | . | 340 (56%) | 436 (37.5%) | . |
| Race/Ethnicity: Non-Hispanic White | 1590 (43%) | 615 (56.1%) | 345 (42.3%) | <.001 | 269 (44.3%) | 361 (31.1%) | <.001 |
| Non-Hispanic Black | 1530 (42%) | 387 (35.3%) | 358 (43.9%) | . | 256 (42.2%) | 529 (45.5%) | . |
| Hispanic | 415 (11%) | 55 (5%) | 78 (9.6%) | . | 60 (9.9%) | 222 (19.1%) | . |
| Other | 145 (4%) | 39 (3.6%) | 34 (4.2%) | . | 22 (3.6%) | 50 (4.3%) | . |
| Current smoking | 479 (13%) | 125 (11.4%) | 146 (17.9%) | <.001 | 51 (8.4%) | 157 (13.5%) | 0.002 |
| Body Mass Index (kg/m2) | 32.1 (7.9) | 30.0 (6.8) | 30.8 (7.5) | 0.011 | 34.8 (8.7) | 33.7 (7.8) | 0.007 |
| Waist (cm) | 106.0 (17.5) | 100.7 (15.6) | 102.7 (17.4) | 0.010 | 111.1 (17.4) | 110.6 (17.6) | 0.531 |
| Medications | . | . | |||||
| ACE inhibitor / ARB use | 2517 (69%) | 584 (53.5%) | 532 (66%) | <.001 | 502 (83.1%) | 899 (77.9%) | 0.010 |
| Diuretics | 2179 (60%) | 556 (51%) | 369 (45.8%) | 0.026 | 435 (72%) | 819 (71%) | 0.644 |
| Calcium Channel blockers | 1488 (41%) | 321 (29.4%) | 344 (42.7%) | <.001 | 233 (38.6%) | 590 (51.1%) | <.001 |
| Clinical and Lab measures | . | . | |||||
| Total Cholesterol | 183.0 (45.0) | 188.2 (39.2) | 188.5 (44.7) | 0.858 | 170.7 (40.5) | 180.6 (50.7) | <.001 |
| Non-HDL (mg/dl) | 135.4 (43.1) | 136.3 (37.6) | 141.4 (42.8) | 0.007 | 124.3 (39.2) | 136.2 (48.7) | <.001 |
| Systolic BP (mm Hg) | 128.0 (21.9) | 120.3 (18.5) | 127.6 (21.8) | <.001 | 123.0 (18.5) | 138.3 (22.5) | <.001 |
| Diastolic BP (mm Hg) | 71.3 (12.7) | 70.5 (11.4) | 76.3 (13.2) | <.001 | 65.0 (11.1) | 71.8 (12.8) | <.001 |
| eGFR (ml/min/1.73 m2) | 42.8 (13.4) | 48.2 (12.8) | 40.1 (13.6) | <.001 | 44.7 (12.7) | 38.8 (12.3) | <.001 |
| eGFR (CRIC) (ml/min/1.73m2) | 44.9 (16.7) | 53.4 (17.3) | 41.7 (16.0) | <.001 | 46.5 (15.9) | 38.4 (13.2) | <.001 |
| Glucose (mg/dL) | 114.9 (51.2) | 91.5 (11.3) | 91.6 (11.9) | 0.966 | 129.7 (50.4) | 145.5 (69.3) | <.001 |
Abbreviations: UPr = 24 hour urine protein (grams), ACE = angiotensin converting enzyme, ARB = angiotensin receptor blocker, HDL = high-density lipoprotein, BP = blood pressure, eGFR = estimated glomerular filtration rate
The entire CRIC cohort of 3,939 participants included 3,673 who had complete 24-hour urine collections (non-diabetic, n=1,907, and diabetic, n=1,766). Table 1 exhibits important demographic and clinical aspects of the CRIC participants. Note that in the analytic cohort, ethnicity percentage was similar between minority populations and non-Hispanic whites, and there were slightly more men. A large proportion of patients were on renin- angiotensin system blocking drugs (69%), and only 13% were current smokers.
The mean blood pressure of 128/71 mmHg indicated good control, but the mean waist circumference and body mass index were high (106 cm and 32kg/m2, respectively). Mean estimated GFR was 42.8 ml/min/1.73m2,and mean 24-hour urine protein was 790 mg.
Patients with more than 150 mg/day of proteinuria were more likely to be current smokers and receiving renin-angiotensin blockers and calcium channel blockers. In addition, they had higher non-HDL cholesterol and higher systolic and diastolic blood pressure, and if diabetic, a higher glycosylated hemoglobin. They also had a lower estimated GFR.
Table 2 depicts the mean and median urinary sodium, urinary potassium, and urinary sodium to potassium ratio in the CRIC participants. This analysis was considered as there is some evidence that increased dietary potassium may offset some of the adverse effects of dietary sodium [15]. The table also divides the participants into non-diabetics and diabetics and splits them further into two groups based on if their 24-hour urinary protein excretion is less than or equal to 150 mg per day or above this level. Both non-diabetics and diabetics had a similar prevalence of increased urinary sodium excretion in the groups with more than 150 mg of protein per day. Urinary potassium excretion did not differ in the non-diabetics across levels of proteinuria. However, in diabetics, the more proteinuric participants excreted slightly more potassium in their urine compared to the group with less proteinuria (p=0.06). Likewise, the sodium to potassium ratio in the urine was statistically greater in non-diabetics with more than 150 mg/day of proteinuria (p<.001), whereas in diabetics, there was no difference in the sodium to potassium ratio based on the amount of proteinuria.
Table 2.
Urine Laboratory Data
| Non-Diabetics | Diabetics | |||||||
|---|---|---|---|---|---|---|---|---|
| All n=3673 |
UPr <=0.15 n=1094 |
UPr >0.15 n=813 |
p | UPr <=0.15 n=606 |
UPr >0.15 n=1160 |
p | ||
| Urinary Lab Values |
||||||||
| Urinary Sodium excretion (mmol/24 hr) | Mean (SD) | 162.0 (77.6) | 146.2 (73.3) | 171.2 (80.1) | <.001 | 158.8 (71.6) | 172.0 (80.4) | <.001 |
| Median (IQR) | 150.9 (108.5 – 201.6) | 135.3 (94.5 – 182.5) | 161.1 (117.3 – 214.2) | <.001 | 148.6 (109.7 – 198.8) | 162.0 (115.7 – 213.4) | 0.002 | |
| Urinary Potassium excretion (mmol/24 hr) | Mean (SD) | 55.3 (26.3) | 55.1 (26.3) | 54.1 (24.9) | 0.375 | 54.1 (24.3) | 56.9 (28.2) | 0.034 |
| Median (IQR) | 51.3 (37.5 – 68.6) | 52.5 (35.6 – 69.2) | 50.1 (36.7 – 67.1) | 0.312 | 49.6 (38.3 – 67.9) | 51.9 (39.0 – 69.1) | 0.061 | |
| Urinary Sodium/Potassium Ratio | Mean (SD) | 3.3 (1.6) | 3.0 (1.7) | 3.6 (1.7) | <.001 | 3.2 (1.4) | 3.3 (1.5) | 0.198 |
| Median (IQR) | 3.0 (2.2 – 4.0) | 2.7 (1.9 – 3.8) | 3.2 (2.3 – 4.4) | <.001 | 3.0 (2.3 – 3.9) | 3.1 (2.3 – 4.0) | 0.344 | |
Abbreviations: UPr = 24-hour urine protein, IQR= Interquartile range
Figure 1 depicts the unadjusted data of urine protein excretion in grams per 24 hours compared to urinary sodium excretion (millimoles per 24 hours) for both diabetics and non-diabetics. Note the similar upward trend between the unadjusted urinary sodium excretion and 24hour urine protein excretion in both diabetics and non-diabetics. Overall, diabetics had greater 24-hour urinary protein excretion at almost every level of urinary sodium excretion except at the highest levels of urinary sodium (greater than 350 mmol/24 hours).
Regression Model
Table 3 shows that our basic model including urinary sodium and clinical site. These two covariates in the joint model for participants with a urine sodium excretion value more than one SD above the norm yielded similar β coefficients and R2 for both diabetics and non-diabetics, indicating a consistent positive association between urinary sodium and proteinuria across the populations studied. Yet, as a univariate variable, urinary sodium explained only 12% of the variation in proteinuria. Progressive adjustment for demographic and clinical variables did not remove the cross-sectional relationship between urinary sodium and proteinuria. This relationship persisted for both diabetics and non-diabetics. The full models are displayed in the supplement. We describe the influence of demographic and clinical variables in the next few paragraphs.
Table 3.
Multivariable regression model to examine the relationship between 24-hour urine sodium and the log transformed 24-hour urine protein
| All n=3673 |
Non-Diabetics n=1904 |
Diabetics n=1766 |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| MODEL | ROWLABEL | Estimate | p | R-SQ | Estimate | p | R-SQ | Estimate | p | R-SQ |
| UNa, CRIC Site only | Urinary Sodium (/SD) | 0.287 | <.0001 | 11.6% | 0.292 | <.0001 | 12.2% | 0.234 | <.0001 | 8.7% |
| + SBP | Urinary Sodium (/SD) | 0.272 | <.0001 | 21.1% | 0.282 | <.0001 | 15.0% | 0.238 | <.0001 | 22.9% |
| demographics, SBP | Urinary Sodium (/SD) | 0.146 | <.0001 | 31.4% | 0.181 | <.0001 | 25.1% | 0.113 | 0.0014 | 35.5% |
| demographics, SBP, glucose | Urinary Sodium (/SD) | 0.151 | <.0001 | 34.6% | 0.192 | <.0001 | 25.3% | 0.119 | 0.0009 | 36.5% |
| demographics, SBP, glucose, U K | Urinary Sodium (/SD) | 0.132 | <.0001 | 34.6% | 0.192 | <.0001 | 25.3% | 0.075 | 0.0648 | 36.7% |
| demographics, SBP, glucose, meds | Urinary Sodium (/SD) | 0.145 | <.0001 | 35.1% | 0.181 | <.0001 | 26.6% | 0.115 | 0.0013 | 37.2% |
| No UNa, demographics, SBP, glucose, meds | Age (/10 years) | −0.434 | <.0001 | 34.3% | −0.369 | <.0001 | 25.1% | −0.513 | <.0001 | 36.8% |
Abbreviations: SBP = systolic blood pressure, UK = urinary potassium, UNa = urinary sodium
In the joint model, with the addition of SBP to the model, one can see a slight decrease in the β coefficient relating urinary sodium to proteinuria. Similar changes occurred in both diabetic and non-diabetic populations. There was a substantial increase in the R2 with the inclusion of SBP for both diabetic and non-diabetic populations indicating that SBP explains some of the variability.
With the subsequent stepwise inclusion of demographic variables for the whole cohort, a decline in the β coefficient to 0.15 was observed. These changes were similar in both non-diabetics (0.28 to 0.18) and diabetics, (0.24 to 0.11),. Within the demographic category, female gender was associated with a negative slope, whereas non-Hispanic blacks and Hispanics had a significant positive slope. It is possible that these slope differences may associate with differential risk for progression of renal disease.
With the subsequent inclusion of glucose measures in the joint model overall, the β coefficient increases slightly. Similar modest increases in the β coefficient with glucose measures occur in both non-diabetics and diabetics. In this section of the joint model, glycosylated hemoglobin was a more important predictive value for the β coefficient than the single glucose measurement. Thus, glucose measures did not add substantially to explain the relationship of urinary sodium to proteinuria.
Next, we included antihypertensive medications in the model. In the joint model, one can see little change in the β coefficient relating urinary sodium to proteinuria. This was true for both non-diabetics and diabetics. There was also a difference in the β coefficient for the whole cohort, as well as the diabetics, when one compared calcium channel blockers against diuretics and angiotensin converting enzyme (ACE) inhibitors, suggesting differential glomerular hemodynamic effects of these medications. There were similar changes in the joint model for both diabetics and non-diabetics.
In the next model, we evaluated the influence of 24-hour urine potassium excretion. This addition had little impact on the association of urinary sodium and proteinuria. The β coefficient and R2 values are negligibly changed, indicating little influence on the relationship of urinary potassium on the relationship of urinary sodium to proteinuria. We saw minimal influence of the sodium: potassium ratio as well (data not shown).
In the final section of Table 3, we removed urinary sodium from the model to examine the independent relationship of each demographic and clinical variable on proteinuria. Of note, one can see the powerful relationship of SBP and demographic factors on proteinuria independent of urinary sodium.
Discussion
Secondary analyses of clinical trials and observational studies in patients with CKD have demonstrated the importance of proteinuria as a predictor of adverse renal outcomes [16–18]. Moreover, epidemiological studies have demonstrated the value of proteinuria as a biomarker of increased risk for cardiovascular events [19]. Clinical studies have demonstrated that increasing dietary sodium intake is associated with increasing blood pressure and increasing proteinuria in patients with CKD [6, 7, 20–22]. The goal of our study was to evaluate the cross-sectional relationship between a 24-hour urinary measurement of sodium and potassium and proteinuria in almost 3700 participants in the CRIC study. We stratified our analysis based on the presence or absence of diabetes as there were no prior studies examining this relationship, and there could be potential differences of clinical relevance. In addition, as illustrated in Figure 1, diabetic patients exhibited a greater 24 hour urine protein excretion at almost every level of urinary sodium. In this analysis, we demonstrate that there is a correlation between urinary sodium excretion and 24-hour urinary protein. However, our results indicate that urinary sodium (+1 standard deviation above the mean) as a single variable only explains 12% of the variation of proteinuria. Moreover, with the stepwise inclusion of other demographic variables, the β coefficient declines indicating the apparent independent contribution of 24-hour urinary is reduced after considering a variety of sodium demographic and clinical factors. Similarly, with the removal of urinary sodium from the model, one can see the powerful influence of SBP and demographic factors on 24-hour urine protein measures.
Previous small studies have demonstrated that increasing urinary sodium excretion, as a measure of dietary sodium ingestion, is associated with an attenuation of the antihypertensive and antiproteinuric effects of drugs that block the renin-angiotensin system such as ACE inhibitors or angiotensin receptor blockers [6, 7]. This is evident whether these medicines are used alone or in conjunction with thiazide diuretics [7]. However, none of these studies adjusted for clinical and demographic factors as we have, nor do they have such a large number of patients with both diabetic and non-diabetic kidney disease. Our cross-sectional study does not provide specific information on whether or not a high level of urinary sodium excretion may attenuate the effects of renin-angiotensin system blockade on proteinuria. What it does provide are a number of interesting associations between urinary sodium and protein excretion in patients with CKD, taking into account several clinical and demographic variables. When adjusting for clinical and demographic factors, there was a reduced contribution of urinary sodium on 24-hour urine protein excretion. We suspect that this may in part be related to the well- controlled blood pressure in our participants (mean 128/71 mmHg). If greater dietary salt intake increases proteinuria through blood pressure-dependent effects, conceivably it would be quenched under circumstances of more effective control of blood pressure. However, one could also view that the persistency of the relationship between urinary sodium excretion and proteinuria despite adjustment for confounders suggests that there are blood pressure-independent effects of salt on glomerular permeability to proteins or renal tubular epithelial cell uptake mechanisms. Some investigators have suggested that increasing dietary salt may induce endothelial dysfunction through oxidative stress and diminish the production of endogenous nitrous oxide [23]. Likewise, other investigators have demonstrated the effect of increasing dietary salt on vascular beds in the brain in hypertensive rats independent of changes in blood pressure [9]. Thus, there may be vascular effects of dietary salt which occur despite well-treated blood pressure.
Amongst the demographic variables, it is notable that female gender was associated with a lesser risk of proteinuria, whereas non-Hispanic Blacks and Hispanics had significantly more proteinuria. These factors were important in diminishing the association of urinary sodium excretion and urinary protein excretion. Not surprisingly, SBP, which was associated with more proteinuria, diminished the contribution of urinary sodium excretion on 24-hour urine protein when added to the model. Hemoglobin A1c was correlated with increased proteinuria, but did little to the overall model with stepwise inclusion. There were also differences in the association of urinary sodium excretion with proteinuria based on the different antihypertensive medications. Both diabetics and non-diabetics had greater β coefficients with calcium channel blockers than they did with diuretics or ACE inhibitors. This may be related to the effects of calcium channel blockers to preferentially dilate the afferent glomerular arteriole. Overall, however, the use of antihypertensive medications did not alter the association of urinary sodium excretion and proteinuria. Similarly, measurement of 24-hour urine potassium excretion did not appear to alter the contribution of urinary sodium on urinary protein excretion. It has been suggested that increasing dietary potassium ingestion can offset the antihypertensive effects of increased dietary sodium [15]. It is possible that with higher levels of blood pressure than what was observed with our participants (128/71 mmHg), one could see a potential vasodepressor response to increased potassium ingestion. It is also possible that with higher blood pressures, increasing dietary potassium could be associated with reduced urinary protein excretion. Our patients received no specific dietary counseling during this cohort study except that during the course of their routine clinical care.
Our results have some limitations. It is important to consider that the CRIC population that we studied is not necessarily representative of the US chronic kidney disease population. We describe only a cross-sectional relationship between urinary sodium and urinary protein excretion after adjustment for a variety of clinical and demographic factors. Urinary electrolyte excretion may not be an accurate representation of dietary electrolyte consumption. In addition, only a single 24-hour urine collection was available. Others have described that 3 overnight urine collections may be a more accurate way to calculate dietary ingestion of sodium [24]. However, the advantage of our study is the large number of cross-sectional measures in both diabetics and non-diabetics who have well-controlled blood pressure, many of whom were receiving renin angiotensin system blocking therapy. Thus, these observations may provide perspective for the relative association of dietary salt on proteinuria in patients with CKD whose blood pressure is well-controlled on a renin-angiotensin blocking regimen. As others have suggested [25, 26] dietary sodium modification may be an important consideration in many, if not most, patients with CKD.
Our results suggest that in the aggregate, there is relationship between 24-hour urinary sodium excretion and proteinuria. We clearly demonstrate that a number of clinical and demographic factors reduce the contribution of dietary sodium on 24-hour excretion. A more complete understanding on the longitudinal relationship between dietary electrolytes as reflected in urinary electrolyte excretion is needed. This may allow a more rational approach to diet modification as a means of facilitating proteinuria reduction in patients with CKD.
Acknowledgements
This report work was supported by the following institutions: University of Maryland GCRC (M01 RR-16500), University of Pennsylvania CTRC CTSA (UL1 RR-024134), The Johns Hopkins University (UL1 RR-025005), Case Western Reserve University Clinical and Translational Science Collaborative (University Hospitals of Cleveland, Cleveland Clinical Foundation, and MetroHealth UL1 RR-024989), University of Michigan (GCRC grant M01 RR-000042, CTSA grant UL1 RR-024986), University of Illinois at Chicago CTSA (UL1RR029879), Tulane/LSU/Charity Hospital General Clinical Research Center (RR-05096), Kaiser NIH/NCRR UCSF-CTSI (UL1 RR-024131).
We thank Tia A. Paul, University of Maryland School of Medicine, Baltimore, MD, for expert secretarial support.
Footnotes
Conflicts of Interest: None
Reference List
- 1.Bibbins-Domingo K, Chertow GM, Coxson PG, Moran A, Lightwood JM, Pletcher MJ, et al. Projected effect of dietary salt reductions on future cardiovascular disease. N Engl J Med. 2010;362(7):590–599. doi: 10.1056/NEJMoa0907355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Graudal NA, Galloe AM, Garred P. Effects of sodium restriction on blood pressure, renin, aldosterone, catecholamines, cholesterols, and triglyceride: a meta-analysis. JAMA. 1998;279(17):1383–1391. doi: 10.1001/jama.279.17.1383. [DOI] [PubMed] [Google Scholar]
- 3.Midgley JP, Matthew AG, Greenwood CM, Logan AG. Effect of reduced dietary sodium on blood pressure: a meta-analysis of randomized controlled trials. JAMA. 1996;275(20):1590–1597. doi: 10.1001/jama.1996.03530440070039. [DOI] [PubMed] [Google Scholar]
- 4.Smith-Spangler CM, Juusola JL, Enns EA, Owens DK, Garber AM. Population strategies to decrease sodium intake and the burden of cardiovascular disease: a cost-effectiveness analysis. Ann Intern Med. 2010;152(8):481–483. doi: 10.7326/0003-4819-152-8-201004200-00212. [DOI] [PubMed] [Google Scholar]
- 5.Stolarz-Skrzypek K, Kuznetsova T, Thijs L, Tikhonoff V, Seidlerova J, Richart T, et al. Fatal and nonfatal outcomes, incidence of hypertension, and blood pressure changes in relation to urinary sodium excretion. JAMA. 2011;305(17):1777–1785. doi: 10.1001/jama.2011.574. [DOI] [PubMed] [Google Scholar]
- 6.Heeg JE, de Jong PE, van der Hem GK, de ZD. Efficacy and variability of the antiproteinuric effect of ACE inhibition by lisinopril. Kidney Int. 1989;36(2):272–279. doi: 10.1038/ki.1989.190. [DOI] [PubMed] [Google Scholar]
- 7.Vogt L, Waanders F, Boomsma F, de ZD, Navis G. Effects of dietary sodium and hydrochlorothiazide on the antiproteinuric efficacy of losartan. J Am Soc Nephrol. 2008;19(5):999–1007. doi: 10.1681/ASN.2007060693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Messerli FH, Schmieder RE, Weir MR. Salt. A perpetrator of hypertensive target organ disease? Arch Intern Med. 1997;157(21):2449–2452. doi: 10.1001/archinte.157.21.2449. [DOI] [PubMed] [Google Scholar]
- 9.Tobian L, Hanlon S. High sodium chloride diets injure arteries and raise mortality without changing blood pressure. Hypertension. 1990;15(6 Pt 2):900–903. doi: 10.1161/01.hyp.15.6.900. [DOI] [PubMed] [Google Scholar]
- 10.Weir MR, Dengel DR, Behrens MT, Goldberg AP. Salt-induced increases in systolic blood pressure affect renal hemodynamics and proteinuria. Hypertension. 1995;25(6):1339–1344. doi: 10.1161/01.hyp.25.6.1339. [DOI] [PubMed] [Google Scholar]
- 11.Jones-Burton C, Mishra SI, Fink JC, Brown J, Gossa W, Bakris GL, et al. An in-depth review of the evidence linking dietary salt intake and progression of chronic kidney disease. Am J Nephrol. 2006;26(3):268–275. doi: 10.1159/000093833. [DOI] [PubMed] [Google Scholar]
- 12.Perry IJ, Beevers DG. Salt intake and stroke: a possible direct effect. J Hum Hypertens. 1992;6(1):23–25. [PubMed] [Google Scholar]
- 13.Feldman HI, Appel LJ, Chertow GM, Cifelli D, Cizman B, Daugirdas J, et al. The Chronic Renal Insufficiency Cohort (CRIC) Study: Design and Methods. J Am Soc Nephrol. 2003;14(7 Suppl 2):S148–S153. doi: 10.1097/01.asn.0000070149.78399.ce. [DOI] [PubMed] [Google Scholar]
- 14.Peterson JC, Adler S, Burkart JM, Greene T, Hebert LA, Hunsicker LG, et al. Blood pressure control, proteinuria, and the progression of renal disease. The Modification of Diet in Renal Disease Study. Ann Intern Med. 1995;123(10):754–762. doi: 10.7326/0003-4819-123-10-199511150-00003. [DOI] [PubMed] [Google Scholar]
- 15.D'Elia L, Barba G, Cappuccio FP, Strazzullo P. Potassium Intake, Stroke, and Cardiovascular Disease: A Meta-Analysis of Prospective Studies. J Am Coll Cardiol. 2011;57(10):1210–1219. doi: 10.1016/j.jacc.2010.09.070. [DOI] [PubMed] [Google Scholar]
- 16.Astor BC, Matsushita K, Gansevoort RT, van d V, Woodward M, Levey AS, et al. Lower estimated glomerular filtration rate and higher albuminuria are associated with mortality and end-stage renal disease. A collaborative meta-analysis of kidney disease population cohorts. Kidney Int. 2011;79(12):1331–1340. doi: 10.1038/ki.2010.550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.de Zeeuw D, Remuzzi G, Parving HH, Keane WF, Zhang Z, Shahinfar S, et al. Proteinuria, a target for renoprotection in patients with type 2 diabetic nephropathy: Lessons from RENAAL. Kidney Int. 2004;65(6):2309–2320. doi: 10.1111/j.1523-1755.2004.00653.x. [DOI] [PubMed] [Google Scholar]
- 18.Jafar TH, Stark PC, Schmid CH, Landa M, Maschio G, Marcantoni C, et al. Proteinuria as a modifiable risk factor for the progression of non-diabetic renal disease. Kidney Int. 2001;60(3):1131–1140. doi: 10.1046/j.1523-1755.2001.0600031131.x. [DOI] [PubMed] [Google Scholar]
- 19.de Zeeuw D, Remuzzi G, Parving HH, Keane WF, Zhang Z, Shahinfar S, et al. Albuminuria, a therapeutic target for cardiovascular protection in type 2 diabetic patients with nephropathy. Circulation. 2004;110(8):921–927. doi: 10.1161/01.CIR.0000139860.33974.28. [DOI] [PubMed] [Google Scholar]
- 20.Myrvang H. Progression of renal disease: High salt intake blunts the benefit of ACE inhibitors and accelerates renal function decline. Nat Rev Nephrol. 2012;8(2):61. doi: 10.1038/nrneph.2011.198. [DOI] [PubMed] [Google Scholar]
- 21.Vegter S, Perna A, Postma MJ, Navis G, Remuzzi G, Ruggenenti P. Sodium intake, ACE inhibition, and progression to ESRD. J Am Soc Nephrol. 2012;23(1):165–173. doi: 10.1681/ASN.2011040430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Yu W, Luying S, Haiyan W, Xiaomei L. Importance and benefits of dietary sodium restriction in the management of chronic kidney disease patients: experience from a single Chinese center. Int Urol Nephrol. 2012;44(2):549–556. doi: 10.1007/s11255-011-9986-x. [DOI] [PubMed] [Google Scholar]
- 23.Nakamura T, Kataoka K, Tokutomi Y, Nako H, Toyama K, Dong YF, et al. Novel mechanism of salt-induced glomerular injury: critical role of eNOS and angiotensin II. J Hypertens. 2011;29(8):1528–1535. doi: 10.1097/HJH.0b013e328348ca95. [DOI] [PubMed] [Google Scholar]
- 24.Liu K, Dyer AR, Cooper RS, Stamler R, Stamler J. Can overnight urine replace 24-hour urine collection to asses salt intake? Hypertension. 1979;1(5):529–536. doi: 10.1161/01.hyp.1.5.529. [DOI] [PubMed] [Google Scholar]
- 25.Krikken JA, Laverman GD, Navis G. Benefits of dietary sodium restriction in the management of chronic kidney disease. Curr Opin Nephrol Hypertens. 2009;18(6):531–538. doi: 10.1097/MNH.0b013e3283312fc8. [DOI] [PubMed] [Google Scholar]
- 26.Wright JA, Cavanaugh KL. Dietary sodium in chronic kidney disease: a comprehensive approach. Semin Dial. 2010;23(4):415–421. doi: 10.1111/j.1525-139X.2010.00752.x. [DOI] [PMC free article] [PubMed] [Google Scholar]