Endourology and Stones

Does Hypertension Impact 24-Hour Urine

Parameters in Patients With
Nephrolithiasis?
Christopher Hartman, Justin I. Friedlander, Daniel M. Moreira, David A. Leavitt,
David M. Hoenig, Arthur D. Smith, and Zeph Okeke
OBJECTIVE To examine the differences in 24-hour urine parameters and stone composition between patients
with and without systemic hypertension (HTN) in a large cohort of stone formers.
MATERIALS AND We performed a retrospective review over a 10-year period of patients with stone, who had
METHODS completed a 24-hour urinalysis (Litholink) and for whom demographic information was available,
including the presence of HTN. Univariate and multivariate analyses were performed, comparing
the 24-hour urinalysis profiles of patients with HTN with that of normotensive patients.
RESULTS Of the 1115 patients eligible for inclusion, 442 patients (40%) had HTN and 673 (60%) did not.
Patients with HTN were significantly older, had a higher body mass index, and had a greater
number of comorbid conditions than normotensive patients. Univariate analysis revealed
significantly lower urine pH, calcium, supersaturation (SS) of calcium oxalate (CaOx) and SS
calcium phosphate (all P <.05) in patients with HTN. Multivariate analysis showed significantly
lower calcium, citrate, and SS CaOx in patients with HTN (all P <.05).
CONCLUSION Our results demonstrate lower levels of calcium and SS CaOx on univariate and multivariate
analysis, as well as lower levels of citrate on multivariate analysis in patients with HTN. These
results suggest that lower levels of citrate may contribute to stone formation to a greater degree in
patients with HTN than abnormalities in calcium metabolism. UROLOGY 85: 539e543, 2015.

2015 Elsevier Inc.

I
t is estimated that approximately 29%-31% (58-65 found to be greater than that of patients without
million people) of the US adult population have nephrolithiasis.7,12,13
systemic hypertension (HTN), whereas the most Studies investigating lithogenic risk have identified
recent estimates for kidney stone prevalence show the rate increased urinary calcium excretion as a modifiable factor
has grown to 8.8% or 1 in 11 people.1-3 Among the general to prevent stone recurrence in hypertensive stone-
public, HTN is more commonly linked with conditions forming patients.4,10 There is also evidence that these
such as cardiovascular disease, end-stage renal disease, and patients have abnormalities in urinary citrate; however,
cerebrovascular disease; yet, there is a well-established as- the evidence for this is mixed despite hypocitraturia being
sociation between HTN and nephrolithiasis, as numerous an easily modifiable factor, either via dietary changes or
studies have identified HTN as an independent predictor with medication. To date, only a handful of studies have
of kidney stone disease.4-8 Some authors have suggested compared urinary metabolic profiles of patients with and
that abnormalities in renal calcium metabolism exist without HTN, most of which have used a limited sample
among patients with HTN, leading to increased urinary size, possibly understating a more accurate profile.
calcium excretion; however, consensus regarding specific Therefore, the primary objective of our study was to
physiological mechanisms linking the 2 conditions remains examine differences in 24-hour urine parameters between
to be fully determined.9-11 Additionally, among patients patients with and without HTN in a large contemporary
with urinary stone disease, the incidence of HTN has been cohort of kidney stone-forming patients. The secondary
objective was to analyze the differences in stone compo-
Financial Disclosure: The authors declare that they have no relevant financial interests.
sition between patients with and without HTN.
From the Department of Urology, Hofstra North Shore-LIJ School of Medicine, The
Arthur Smith Institute for Urology, New Hyde Park, NY; and the Department of
Urology, Fox Chase/Einstein Urologic Institute, Philadelphia, PA
Address correspondence to: Christopher Hartman, M.D., Department of Urology,
MATERIALS AND METHODS
Hofstra North Shore-LIJ School of Medicine, The Arthur Smith Institute for Urology,
450 Lakeville Road, Suite M41, New Hyde Park, NY 11042. E-mail:
Study Sample
CHartmanMD@gmail.com After obtaining institutional review board approval, data from
Submitted: September 3, 2014, accepted (with revisions): December 5, 2014 adult patients undergoing 24-hour urine analysis at The Arthur

ª 2015 Elsevier Inc. http://dx.doi.org/10.1016/j.urology.2014.12.013 539

All Rights Reserved 0090-4295/15
Smith Institute for Urology in New Hyde Park, NY, between Table 1. Baseline patient characteristics by systemic
March 2002 and February 2012 were combined into the study hypertension status
database, as we did in previous studies.14 The cohort consisted of
No HTN HTN
new patients at a busy stone clinic in Long Island, NY, above
the stone belt, under the care of one of the 2 endourologists Characteristic N (%) N (%) P Value*
(A.D.S. and Z.O.). Patients were excluded if they were Age, mean (SD), y 53.7 (15.0) 64.5 (12.2) <.001
aged <18 years, did not provide a medical history, or if they did BMI, mean (SD), 28.2 (6.3) 31.3 (7.5) <.001
not complete a satisfactory 24-hour urinalysis as defined by 24- kg/m2
hour creatinine excretion of
800 mg for men and
600 mg for Gender .115
women. Diagnosis of HTN was based on self-reported medical Male 367 (54.5) 263 (59.5)
history and/or current use of antihypertensive medication. Female 306 (45.5) 179 (40.5)
Diabetes mellitus 59 (8.8) 122 (27.6) <.001
Hyperlipidemia 99 (14.7) 213 (48.2) <.001
Twenty-four-hour Urine Collection and Analysis Statin 84 (12.5) 192 (43.4) <.001
Patients presenting at their initial visit for nephrolithiasis were Thiazide 5 (0.7) 53 (12.0) <.001
Potassium citrate 44 (6.5) 35 (7.9) .447
counseled on how to appropriately collect a 24-hour urinalysis
Allopurinol 17 (2.5) 28 (6.3) .003
specimen. In the case of patients who provided two 24-hour Urinary diversion 6 (0.9) 3 (0.7) 1.000
urinalysis collections, only the first specimen was included in
the analysis, corresponding to the pretreatment urine collection. BMI, body-mass index; HTN, systemic hypertension; SD, standard
deviation.
Specimens were then analyzed by the Litholink corporation for * The chi-square test for categorical variables and the Student
a number of characteristics, including urinary pH, calcium, t test for continuous variables.
citrate, uric acid (UA), sodium, potassium, sulfate, oxalate,
chloride, magnesium, phosphate, creatinine, supersaturation
(SS) UA, SS calcium oxalate (CaOx), SS calcium phosphate RESULTS
(CaP), and volume. Of the 1115 patients available for inclusion in this study,
442 (40%) reported a history of HTN or were taking an
Stone Composition Analysis antihypertensive medication, and 673 (60%) did not
Stone composition analysis was performed using infrared spec- have HTN at the time of urinalysis. Table 1 presents the
troscopy at Mayo Medical Laboratories (Rochester, MN) on a baseline characteristics of patients in this study. There
subgroup of patients for which these data were available. was no statistically significant difference in gender among
Representative specimens from all identifiable layers of the the cohort. Hypertensive patients were older than those
calculus were taken. Each specimen was weighed and then without HTN (mean age  standard deviation [SD], 64.5
crushed into a fine powder for analysis. The infrared spectrum of  12.2 vs 53.7  15.0 years; P <.001), had a significantly
each specimen was recorded, and the resulting spectrum was greater BMI (mean  SD, 31.3  7.5 vs 28.2  6.3 kg/
compared against reference spectra of all known calculus com-
m2; P <.001), and were significantly more likely to have
ponents, including CaOx monohydrate, CaOx dihydrate, cal-
cium carbonate, UA, apatite, brushite, struvite, and ammonium
the comorbid conditions of DM (mean  SD, 27.6% vs
urate. 8.8%; P <.001) and hyperlipidemia (mean  SD, 48.2%
vs 14.7%; P <.001).
Univariate analysis demonstrated that patients with
Statistical Analysis HTN had a significantly higher 24-hour urine sodium
Comparisons of baseline patient characteristics between patients (mean  SD, 178.7  76.6 vs 169.4  74.5 mmol;
with HTN and those without HTN were performed using the P <.05) and potassium (mean  SD, 65.7  30.6 vs 61.4
Student t test for continuous variables and the chi-squared test
 27.3 mEq; P ¼ .02) than patients without HTN. Hy-
for categorical variables. Univariate comparisons of urinary el-
pertensive stone formers also had a significantly lower pH
ements between patients with and without HTN were per-
formed with the Student t test. Multivariate linear regression (mean  SD, 5.91  0.58 vs 6.02  0.56; P ¼ .002),
was used to analyze the association of urinary composition and calcium (mean  SD, 178.2  120 vs 203.4  113.1 mg;
HTN status, adjusting for patient characteristics including age, P <.001), SS CaOx (mean  SD, 6.05  3.64 vs 7.03 
gender, body mass index (BMI), diabetes mellitus (DM), thia- 3.57; P <.001), and SS CaP (mean  SD, 0.82  0.85 vs
zide use, potassium citrate use, allopurinol use, the presence of a 1.11  0.91; P <.001) than stone formers without HTN.
urinary diversion, 24-hour urine volume, and creatinine. Dif- No differences were observed on univariate analysis be-
ferences in stone composition between patients with and tween patients with HTN and without HTN for 24-hour
without HTN were performed using the Student t test for each urinary citrate, UA, sulfate, oxalate, magnesium, phos-
stone component (CaOx monohydrate, CaOx dihydrate, cal- phate, creatinine, SS UA, or volume (Table 2).
cium carbonate, UA, apatite, brushite, struvite, and ammonium
Multivariate analysis adjusting for patient characteris-
urate). We also performed an exploratory analysis to evaluate
tics including gender, age, BMI, DM, potassium citrate
the association of HTN with stone composition, adjusting for
DM and BMI using logistic regression. All analyses were per- use, thiazide use, allopurinol use, the presence of a urinary
formed using R, version 2.15.2 (R Foundation for Statistical diversion, 24-hour urine volume, and creatinine demon-
Computing, Vienna, Austria), statistical analysis software and strated that compared with patients without HTN, hy-
were 2 tailed. A P value of <.05 was considered significant, and pertensive stone formers had significantly lower urinary
confidence intervals were calculated at 95%. calcium (mean difference ¼ 20.13 mg; P ¼ .007),

540 UROLOGY 85 (3), 2015

Table 2. Univariate analysis of 24-hour urine composition Table 3. Multivariate adjusted differences in 24-hour urine
by HTN status composition comparing patients with hypertension with
those without hypertension
P
Parameter No HTN HTN Value* P
pH 6.02 (0.56) 5.91 (0.58) .002 Parameter Difference* 95% CI Valuey
Calcium (mg) 203.4 (113.1) 178.2 (120.0) <.001 pH 0.013 0.082 to 0.056 .709
Citrate (mg) 536.4 (331.4) 511.9 (323.5) .215 Calcium (mg) 20.13 34.69 to 5.57 .007
Uric acid (g) 0.69 (0.24) 0.68 (0.26) .406 Citrate (mg) 74.04 115.32 to 32.76 <.001
Sodium (mmol) 169.4 (74.5) 178.7 (76.6) .046 Uric acid (g) 0.014 0.038 to 0.010 .232
Potassium 61.4 (27.3) 65.7 (30.6) .016 Sodium (mmol) 2.08 5.84 to 10.00 .605
(mEq) Potassium (mEq) 1.45 4.65 to 1.75 .372
Sulfate (mmol) 43.7 (19.3) 44.5 (19.8) .524 Sulfate (mmol) 0.65 2.47 to 1.17 .478
Oxalate (mg) 38.8 (16.6) 40.1 (15.7) .199 Oxalate (mg) 0.42 2.30 to 1.46 .659
Magnesium (mg) 98.7 (43.2) 95.9 (44.5) .290 Magnesium (mg) 3.45 8.64 to 1.74 .195
Phosphate (g) 0.97 (0.39) 0.94 (0.39) .239 Phosphate (g) 0.043 0.078 to 0.008 .001
Creatinine (mg) 1593 (536.3) 1578 (535.6) .653 Creatinine (mg) 24.2 73.6 to 25.2 .341
SS UA 1.13 (0.97) 1.19 (0.94) .293 SS UA 0.022 0.130 to 0.086 .689
SS CaOx 7.03 (3.57) 6.05 (3.64) <.001 SS CaOx 0.648 1.060 to 0.236 .002
SS CaP 1.11 (0.91) 0.82 (0.85) <.001 SS CaP 0.083 0.197 to 0.031 .151
Volume (L) 1.90 (0.87) 1.96 (0.81) .203 Volume (L) 0.025 0.141 to 0.091 .653
SS CaOx, supersaturation calcium oxalate; SS CaP, supersatura- CI, confidence interval; other abbreviations as in Table 2.
tion calcium phosphate; SS UA, supersaturation uric acid; other * Referent is to patients without hypertension.
abbreviation as in Table 1. y
Multinomial linear regression.
* Student t test.
Table 4. Average stone composition by hypertension
citrate (mean difference ¼ 74.04 mg; P <.001), phos- status
phate (mean difference ¼ 0.043 g; P ¼ .001), and SS
No HTN HTN
CaOx (mean difference ¼ 0.648; P ¼ .002). There was P
no difference on multivariate analysis between patients Stone Composition Mean (%) Mean (%) Value*
with HTN and those without HTN for urinary pH, UA, Calcium oxalate 47.2 45.3 .618
sodium, potassium, sulfate, oxalate, magnesium, creati- monohydrate
Calcium oxalate dihydrate 11.2 5.0 <.001
nine, SS UA, SS CaP, or volume (Table 3).
Calcium carbonate 0.8 0.6 .356
In a separate analysis of 436 patients for which stone Uric acid 13.2 28.8 <.001
composition was available, patients without HTN were Apatite 21.9 14.1 .002
found to have a significantly greater proportion of CaOx Brushite 3.0 1.3 .160
dihydrate (11.2% vs 5.0%; P <.001) and apatite (21.9% Struvite 2.5 4.2 .225
Ammonium urate 0.2 0.7 .265
vs 14.1%; P ¼ .002) in their stones than patients with
HTN. Conversely, patients with HTN had a greater Abbreviation as in Table 1.
* Student t test.
proportion of UA (28.8% vs 13.2%; P <.001) in their
stones than patients without HTN. There was no differ-
ence in the proportion of other stone components be- this would include the treatment of metabolic abnor-
tween hypertensive and normotensive patients (Table 4). malities detected in the serum and 24-hour urine analysis.
Given that these results may be confounded by the fact To identify the potentially treatable urinary metabolic
that HTN is associated with DM and BMI, and both are abnormalities associated with HTN, we sought to analyze
known to be associated with UA stones, we performed the differences in 24-hour urine parameters between pa-
exploratory analyses of stone composition by HTN status, tients with and without HTN in a cohort of kidney stone-
adjusting for DM and BMI. In these analyses, the results forming patients. In addition, we sought to evaluate the
were virtually unchanged (Supplementary Tables 1, 2). differences in stone composition between patients with
and without HTN.
Previous studies have investigated differences in urine
COMMENT composition between patients with HTN compared with
HTN and nephrolithiasis are common medical condi- normotensive patients. For example, Eisner et al found
tions in the United States, both of which are growing in that patients with HTN having stones excreted signifi-
prevalence. Studies have identified an increased risk of cantly more calcium in their urine than normotensive
HTN in patients with kidney stones, whereas others have patients with a history of nephrolithiasis (25.6 mg more
shown an increased risk of stones in patients with per day). Additionally, they found that the relative risk
HTN.1,2,5,10 Although the specific physiological mecha- (RR) of having HTN significantly increased as calcium
nisms linking these 2 conditions have not been described, excretion increased (RR ¼ 1.29; P ¼ .03). This rela-
it is important for urologists to use readily available means tionship did not exist for citrate excretion (RR ¼ 0.94;
to lower the risk of kidney stone formation specifically P ¼ .56).10 Conversely, Taylor et al examined data from
tailored for patients with and without HTN. At present, the Nurses’ Health Study and Health Professionals

UROLOGY 85 (3), 2015 541

Follow-up Study and found that patients with lower 24- On stone composition analysis, our results show that
hour citrate excretion were significantly more likely to patients with HTN have a significantly greater proportion
have HTN than patients with higher citrate excretion of UA in their stones than normotensive patients. Given
(odds ratio [OR] ¼ 0.37 for older women, P <.001; OR ¼ that citrate has been shown in a number of studies to
0.54 for younger women, P ¼ .03; and OR ¼ 0.27 for prevent the formation of UA stones by urinary alkaliza-
men, P <.001). They found that citrate was the only tion,23,24 our results suggest that lower levels of urinary
urine constituent that was statistically different between citrate among patients with HTN could potentially be
patients with and without HTN, failing to demonstrate a leading to a greater propensity to form UA stones in these
difference in urinary calcium excretion.15 Thus, there is patients. Additionally, we demonstrated that patients
still considerable disagreement in the literature regarding with HTN were significantly more likely to have DM,
differences in urinary composition between stone formers which has been shown in multiple previous studies to be a
with and without HTN, and this is why we decided to risk factor for the formation of UA stones.25-27 Although
study this topic. In addition, we also sought to evaluate DM may be a confounder that serves as a link between
the differences in stone composition between these 2 HTN and the greater proportion of UA stones in our
groups of stone formers. cohort, a separate analysis after adjusting for DM
In our study, we found significantly lower levels of demonstrated that the results were virtually unchanged.
urinary calcium and SS CaOx in patients with HTN on Limitations of this study include the retrospective
both univariate and multivariate analysis. Additionally, design and restriction of our cohort to stone formers only.
our results demonstrate lower levels of urinary citrate Additionally, patients were seen in a single, large-volume,
among patients with HTN on multivariate analysis. Our stone clinic, and therefore, our results may not be
findings are consistent with the results of Taylor et al in generalizable to individuals in other geographic areas.
demonstrating lower levels of citrate in the urine of hy- Classification of patients as either having HTN or being
pertensive stone formers than in patients without HTN. normotensive was done by self-report on a questionnaire
Although Eisner et al found a difference in citrate that patients completed at their initial office visit for
excretion between patients with and without HTN, their nephrolithiasis, thereby potentially incorrectly labeling
results did not attain statistical significance, possibly due some patients as normotensive when they were actually
to sample size. Interestingly, we found a significant dif- undiagnosed with HTN. In addition, we were not able to
ference in urinary calcium excretion between hyperten- analyze the severity of HTN, as objective data were not
sive and normotensive stone formers on both univariate available for all patients. Finally, as some patients were
and multivariate analysis; however, greater calcium recurrent stone formers referred from other practices, we
excretion was observed in patients without HTN, which were unable to differentiate between recurrent and first-
is in direct opposition to the results found by Eisner et al time stone formers in this cohort. Nevertheless, to our
on multivariate analysis. Although their results showed a knowledge, this study is the largest to date comparing the
trend on univariate analysis toward lower calcium 24-hour urine composition of patients with HTN with
excretion in patients with HTN compared with normo- those without HTN.
tensive patients (206.7 vs 213.4 mg), this was not sta-
tistically significant (P ¼ .60). We also observed a CONCLUSION
significantly lower SS of CaOx in patients with HTN Our results demonstrate lower 24-hour urine calcium and
compared with patients without HTN on both univariate SS CaOx in patients with HTN compared with normo-
and multivariate analysis. These findings suggest that tensive patients on both univariate and multivariate an-
citrate excretion appears to have a greater impact on alyses. Additionally, multivariate analysis revealed
stone formation in patients with HTN, and calcium significantly lower 24-hour urine citrate in patients with
excretion may play a less substantial role. HTN. These results argue that lower levels of stone in-
Patients with HTN differ from normotensive patients hibitors such as citrate may play a greater role in stone
in a number of ways. For example, patients with HTN are formation in patients with HTN than deranged calcium
more likely to have a higher concentration of salt in their metabolism and should be taken into account in miti-
diets16,17 and tend to be more obese.18,19 They are, gating nephrolithiasis in these patients.
additionally, more likely to have a greater number of
comorbid conditions such as hyperlipidemia and DM20,21 References
and tend to live more sedentary lifestyles.22 A number of
1. Egan BM, Zhao Y, Axon RN. US trends in prevalence, awareness,
medications used in the treatment of HTN, such as treatment, and control of hypertension, 1988-2008. JAMA. 2010;
thiazide diuretics, may impact urine composition in these 303:2043-2050.
patients. The aforementioned dissimilarities between pa- 2. Scales CD Jr, Smith AC, Hanley JM, et al. Prevalence of kidney
tients with and without HTN may serve to explain the stones in the United States. Eur Urol. 2012;62:160-165.
3. Wright JD, Hughes JP, Ostchega Y, et al. Mean systolic and diastolic
differences seen on 24-hour urinalysis. Although we were
blood pressure in adults aged 18 and over in the United States,
able to adjust for some of these variables in our analyses, 2001-2008. Natl Health Stat Rep. 2011:1-22. 24.
studies evaluating these factors are needed to identify the 4. Borghi L, Meschi T, Guerra A, et al. Essential arterial hypertension
link between HTN and stone formation. and stone disease. Kidney Int. 1999;55:2397-2406.

542 UROLOGY 85 (3), 2015

 5. Cappuccio FP, Strazzullo P, Mancini M. Kidney stones and hyper- across populations. Intersalt Cooperative Research Group. BMJ.
tension: population based study of an independent clinical associ- 1996;312:1249-1253.
ation. BMJ. 1990;300:1234-1236. 18. Aucott L, Rothnie H, McIntyre L, et al. Long-term weight loss from
6. Losito A, Nunzi EG, Covarelli C, et al. Increased acid excretion in lifestyle intervention benefits blood pressure?: a systematic review.
kidney stone formers with essential hypertension. Nephrol Dial Hypertension. 2009;54:756-762.
Transplant. 2009;24:137-141. 19. Huang Z, Willett WC, Manson JE, et al. Body weight, weight
7. Madore F, Stampfer MJ, Rimm EB, et al. Nephrolithiasis and risk of change, and risk for hypertension in women. Ann Intern Med. 1998;
hypertension. Am J Hypertens. 1998;11:46-53. 128:81-88.
8. Robertson WG, Peacock M, Baker M, et al. Studies on the preva- 20. Ames RP. Hyperlipidemia in hypertension: causes and prevention.
lence and epidemiology of urinary stone disease in men in Leeds. Br Am Heart J. 1991;122:1219-1224.
J Urol. 1983;55:595-598. 21. Lago RM, Singh PP, Nesto RW. Diabetes and hypertension. Nat
9. Cappuccio FP, Kalaitzidis R, Duneclift S, et al. Unravelling the Clin Pract Endocrinol Metab. 2007;3:667.
links between calcium excretion, salt intake, hypertension, kidney 22. Beilin LJ. Lifestyle and hypertension—an overview. Clin Exp
stones and bone metabolism. J Nephrol. 2000;13:169-177. Hypertens. 1993;1999:749-762.
10. Eisner BH, Porten SP, Bechis SK, et al. Hypertension is associated 23. Cameron MA, Sakhaee K. Uric acid nephrolithiasis. Urol Clin
with increased urinary calcium excretion in patients with neph- North Am. 2007;34:335-346.
rolithiasis. J Urol. 2010;183:576-579. 24. Pak CY, Sakhaee K, Fuller C. Successful management of uric acid
11. McCarron DA, Pingree PA, Rubin RJ, et al. Enhanced parathyroid nephrolithiasis with potassium citrate. Kidney Int. 1986;30:422-428.
function in essential hypertension: a homeostatic response to a 25. Daudon M, Lacour B, Jungers P. High prevalence of uric acid
urinary calcium leak. Hypertension. 1980;2:162-168. calculi in diabetic stone formers. Nephrol Dial Transplant. 2005;20:
12. Gillen DL, Coe FL, Worcester EM. Nephrolithiasis and increased 468-469.
blood pressure among females with high body mass index. Am J 26. Daudon M, Traxer O, Conort P, et al. Type 2 diabetes increases the
kidney Dis. 2005;46:263-269. risk for uric acid stones. J Am Soc Nephrol. 2006;17:2026-2033.
13. Madore F, Stampfer MJ, Willett WC, et al. Nephrolithiasis and risk 27. Lieske JC, de la Vega LS, Gettman MT, et al. Diabetes mellitus and
of hypertension in women. Am J kidney Dis. 1998;32:802-807. the risk of urinary tract stones: a population-based case-control
14. Moreira DM, Friedlander JI, Hartman C, et al. Using 24-hour uri- study. Am J kidney Dis. 2006;48:897-904.
nalysis to predict stone type. J Urol. 2013;190:2106-2111.
15. Taylor EN, Mount DB, Forman JP, et al. Association of prevalent
hypertension with 24-hour urinary excretion of calcium, citrate, and APPENDIX
other factors. Am J kidney Dis. 2006;47:780-789.
16. Adrogue HJ, Madias NE. Sodium and potassium in the pathogenesis SUPPLEMENTARY DATA
of hypertension. New Engl J Med. 2007;356:1966-1978. Supplementary data associated with this article can be found,
17. Elliott P, Stamler J, Nichols R, et al. Intersalt revisited: further in the online version, at http://dx.doi.org/10.1016/j.urology.
analyses of 24 hour sodium excretion and blood pressure within and 2014.12.013.

UROLOGY 85 (3), 2015 543