Patient-Oriented, Translational Research: Research Article

Nephrology
American Journal of

Am J Nephrol 2019;49:438–448 Received: January 31, 2019

Accepted: March 14, 2019
DOI: 10.1159/000500042 Published online: April 17, 2019

Fruit and Vegetable Treatment of Chronic

Kidney Disease-Related Metabolic Acidosis
Reduces Cardiovascular Risk Better than
Sodium Bicarbonate
Nimrit Goraya a, b Yolanda Munoz-Maldonado c Jan Simoni d
     

Donald E. Wesson e, f  

a Departments of Internal Medicine, Texas A&M College of Medicine, Temple, TX, USA; b Departments of Internal
   

Medicine, Baylor Scott and White Health, Temple, TX, USA; c Statistical Savvy Consulting, Georgetown, TX, USA;
 

d Department of Surgery, Texas Tech University Health Sciences Center, Lubbock, TX, USA; e Department of Internal
   

Medicine, Texas A&M Health Sciences Center College of Medicine, Dallas, TX, USA; f Baylor Scott and White Health
 

and Wellness Center, Dallas, TX, USA

Keywords rate (eGFR) course as the primary analysis and on indicators

Bicarbonate · Dietary alkali · Metabolic acidosis · Health
outcomes · Acid · Alkali · Cardiovascular disease risk ·
Chronic kidney disease · Chronic metabolic acidosis · Diet ·
Myocardial infarction · Stroke
Abstract
Background: Current guidelines recommend treatment of
metabolic acidosis in chronic kidney disease (CKD) with so-
dium-based alkali. We tested the hypothesis that treatment
with base-producing fruits and vegetables (F + V) better im-
proves cardiovascular disease (CVD) risk indicators than oral
sodium bicarbonate (NaHCO3). Methods: We randomized
108 macroalbuminuric, matched, nondiabetic CKD patients
with metabolic acidosis to F + V (n = 36) in amounts to reduce
dietary acid by half, oral NaHCO3 (HCO3, n = 36) 0.3 mEq/kg
bw/day, or to Usual Care (UC, n = 36) to assess the 5-year ef-
fect of these interventions on estimated glomerular filtration
of CVD risk as the secondary analysis. Results: Five-year plas-
ma total CO2 was higher in HCO3 and F + V than UC but was
not different between HCO3 and F + V (difference p value <
0.01). Five-year net eGFR decrease was less in HCO3 (mean
–12.3, 95% CI –12.9 to –11.7 mL/min/1.73 m2) and F + V
(–10.0, 95% CI –10.6 to –9.4 mL/min/1.73 m2) than UC (–18.8,
95% CI –19.5 to –18.2 mL/min/1.73 m2; p value < 0.01) but
was not different between HCO3 and F + V. Five-year systolic
blood pressure was lower in F + V than UC and HCO3 (p value
<0.01). Despite similar baseline values, F + V had lower low-
density lipoprotein, Lp(a), and higher serum vitamin K1 (low
serum K1 is associated with coronary artery calcification)
than HCO3 and UC at 5 years. Conclusion: Metabolic acidosis
improvement and eGFR preservation were comparable in
CKD patients treated with F + V or oral NaHCO3 but F + V bet-
ter improved CVD risk indicators, making it a potentially bet-
ter treatment option for reducing CVD risk.
© 2019 S. Karger AG, Basel
130.237.165.40 - 4/21/2019 4:10:32 AM

© 2019 S. Karger AG, Basel Donald E. Wesson, MD, MBA

Stockholm University Library

Department of Internal Medicine, Texas A&M College of Medicine

Baylor Scott and White Health and Wellness Center
E-Mail karger@karger.com
4500 Spring Avenue, Dallas, TX 75210 (USA)
www.karger.com/ajn
Downloaded by:

E-Mail Donald.wesson @ BSWHealth.org

 Introduction
Individuals with chronic kidney disease (CKD) suffer
premature death and do so more commonly than devel-
oping end-stage kidney disease [1]. This enhanced mor-
tality is due predominantly to increased risk for, and ear-
lier onset of, cardiovascular disease (CVD) [1, 2], includ-
ing myocardial infarction [3] and cerebrovascular
accident or stroke [4]. Metabolic acidosis further increas-
es CVD and overall mortality risk in CKD [5]. Metabolic
acidosis prevalence increases as estimated glomerular fil-
tration rate (eGFR) declines [6] and the acid (H+)-pro-
ducing diets typical of developed societies [7] contribute
to metabolic acidosis in CKD patients with reduced GFR
[8]. Although current guidelines recommend treatment
of metabolic acidosis in CKD with sodium (Na+)-based
alkali [9], base-producing fruits and vegetables (F + V)
that are generally low in diets of US adults overall [10] and
are even lower in those with CKD [11], comparably im-
prove CKD-related metabolic acidosis [12, 13]. A diet
high in F + V is associated with reduced CVD mortality
in the general population [14] and in patients with CKD
[15, 16]. Such a diet is also associated with reduced low-
density lipoprotein cholesterol (LDL-c) [17], and de-
creased LDL-c predicts a reduction in major CVD events
in CKD [18]. Furthermore, high F + V diets are associ-
ated with reduced risk of coronary artery calcification
(CAC) [19], which independently predicts CVD risk over
and above more traditional CVD risk factors in the gen-
eral population [20], is more common and more severe
in CKD [21], and has stronger association with risk of
CVD and mortality in CKD patients [22]. Vitamin K-de-
pendent matrix Gla protein contributes to the physiolog-
ic prevention of CAC [23], and low vitamin K status com-
mon in CKD patients [24] might be corrected by in-
creased F + V intake. We tested the hypothesis that
treating metabolic acidosis in CKD with currently recom-
mended sodium bicarbonate (NaHCO3) or F + V helps
preserve eGFR but that F + V better improves CVD risk
indicators and yields fewer adverse CVD events.
Methods
We designed this randomized matched interventional study
to compare the effect of 60 months of dietary acid reduction
with added daily oral NaHCO3 (HCO3) or added F + V on the
primary analysis of Cystatin C-calculated eGFR and the second-
ary analyses of effect on urine kidney injury parameters and on
indicators of CVD risk. We published a 3-year analysis of the
effect on kidney parameters of these same patients [13]. Patients
had CKD stage 3 (eGFR 30–59 mL/min/1.73 m2) macroalbu-
Metabolic Acidosis Treatment and
Cardiovascular Risk
minuric (urine albumin-to-creatinine ratio > 200 mg/g creati-
nine) hypertension-associated, but non-diabetic, nephropathy,
and metabolic acidosis characterized by plasma total CO2
(PTCO2) > 22 mM but < 24 mM. We chose these PTCO2 criteria
to study patients with a level of metabolic acidosis not warrant-
ing alkali treatment by extant guidelines [25], allowing us ethi-
cally to randomize some to no alkali treatment (usual care
[UC]). Primary outcome was effect of these interventions on net
eGFR change. Secondary outcomes were urine excretion of kid-
ney injury parameters, body mass index (BMI), systolic blood
pressure (SBP), plasma levels of LDL-c and high-density lipo-
protein cholesterol (HDL-c). Post hoc we included plasma Lp(a)
and vitamin K1 as well as course of the use of the following
top  7  medications taken by the groups: enalapril, diltiazem,
clonidine, hydrochlorthiazide, furosemide, atorvastatin, and at-
enolol.
Earlier studies of a separate group of patients [26] suggested
the need for 33 patients in each of the 3 groups to achieve 80%
power and significance level of 0.05 to detect eGFR benefit of the
interventions among the groups (assuming NaHCO3 and F + V
were equally effective in preserving eGFR) compared with UC.
We identified individuals from our clinic system as described
[27] who met these inclusion criteria: (1) nonmalignant hyper-
tension; (2) eGFR by Modification of Diet in Renal Disease for-
mula [28] = 30–59 mL/min/1.73 m2; (3) PTCO2 < 25 mM (the
lower limit of normal in our clinical laboratory) and >22 mM but
< 24 mM upon follow-up measurement using ultrafluoremetry
[29]; (4) macroalbuminuria because of the higher risk for subse-
quent GFR decline [30] than normoalbuminuria or microalbu-
minuria, so as to include participants at increased risk for ne-
phropathy progression and thereby enhance our ability to detect
an intervention effect on nephropathy progression; (5) able to
tolerate angiotensin-converting enzyme inhibition (ACEI); (6)
nonsmoking defined as not having smoked tobacco for ≥1 year
prior to study because smoking is associated with exacerbation of
hypertension-associated nephropathy [30–32]; (7) no diabetes or
CVD on their problem lists and no clinical evidence of CVD; (8)
≥2 primary care physician visits in the preceding year, showing
compliance with clinic visits; and (9) age ≥18 years and able to
give consent. Exclusion criteria were (1) primary kidney disease
or findings consistent thereof such as ≥3 red blood cells per high-
powered field of urine or urine cellular casts; (2) history of dia-
betes or fasting blood glucose ≥110 mg/dL; (3) current pregnan-
cy, history of malignances, chronic infections, or clinical evi-
dence of CVD; (4) peripheral edema or diagnoses associated with
edema such as heart/liver failure or nephrotic syndrome; (5)
baseline plasma potassium concentration ([K+]) >4.6 mEq/L be-
cause observation of our cadre of such patients collected over
many years and published studies [12, 13] show that those with
CKD stage 4 and with [K+] < 4.6 mEq/L were at very low risk of
subsequently developing [K+] > 5.0 mEq/L; and (6) patients tak-
ing or who could not stop taking drugs other than ACEIs that
limit urine K+ excretion. Preliminary studies showed no patient
fitting the eGFR criteria with plasma [K+] ≤4.6 mEq/L given F +
V developed plasma [K+] > 5.0 mEq/L after 4 weekly measure-
ments of plasma [K+] despite ACEI and increased F + V intake
which increases dietary K+.
We identified 961 participants between 1998 and 2005 meet-
ing study criteria and matched 108 for age, sex, ethnicity (black
vs. white or Hispanic), eGFR, and urine albumin excretion (Ualb)
130.237.165.40 - 4/21/2019 4:10:32 AM
Am J Nephrol 2019;49:438–448 439
Stockholm University Library
DOI: 10.1159/000500042
Downloaded by:
 ceipt of F + V, we randomized patients so that the study was con-
Protocol trolled but not blinded. We measured dietary H+ as potential re-
nal acid load, pre- and post-interventions, using 3-day diaries of
961 CKD 3
patients screened food intake and a published formula [34]. We assessed compli-
ance to prescribed NaHCO3 and F + V through reductions in
Triplicate match for age, sex, ethnicity, urine net acid excretion and did not monitor participant
eGFR, urine albumin ­NaHCO3 or F + V intake. We had serum and blood data for up
to 4 years for all participants but no year-5 data for 3 of UC, 3 of
108 CKD 3 HCO3, and 3 of F + V patients. There were 3 deaths in UC, 1 in
patients
HCO3, and none in F + V.
Randomized within triplicates
We made the diagnosis of hypertension-associated nephropa-
thy as the exclusive nephropathy cause clinically by excluding pa-
0.3 mEq/kg.bw tients with systemic diseases associated with nephropathy, ne-
Usual care F+V phrotic-range proteinuria, and urine abnormalities except albu-
NaHCO3
minuria. None had a kidney biopsy. We excluded secondary causes
of hypertension such as renal artery stenosis and hyperaldosteron-
Urine + serum Urine + serum Urine + serum ism clinically. We did not perform kidney Doppler studies or plas-
parameters parameters parameters ma aldosterone-to-renin ratios. All had SBP reduced toward
(n = 36) (n = 36) (n = 36) <130 mm Hg using a described protocol [31] with regimens in-
cluding ACEIs as per extant guidelines for participants with hyper-
5 years 5 years
tension and albuminuria [35]. We prescribed enalapril for ACEI
Urine + serum Urine + serum Urine + serum because it was the only indigent formulary ACE inhibitor when
parameters parameters parameters these studies began. We instituted these kidney-protective strate-
(n = 33) (n = 33) (n = 33) gies with the goal of preventing transition of these CKD stage
3 (eGFR = 30–59 mL/min/1.73 m2) patients to stage 4 (eGFR =
15–29 mL/min/1.73 m2) with its associated increase in CVD risk
Fig. 1. Protocol outline. CKD, chronic kidney disease; n, number [2]. We instituted CVD risk reduction as per extant guidelines [36]
of participants; UC, usual care; F + V, fruits + vegetables; eGFR, and prescribed atorvastatin for statin indications because it was the
estimated glomerular filtration rate. only indigent formulary statin when these studies began. We mea-
sured plasma and urine parameters, SBP, eGFR, and BMI before
and after intervention.
Our local IRB approved the protocol in accordance with The
who agreed to randomization as outlined in Figure 1. When we Code of Ethics of the World Medical Association (Declaration of
identified 3 so-matched individuals, we prospectively random- Helsinki) for experiments involving humans. We did not register
ized them to UC, HCO3, or F + V and began their 5-year follow- the study on the clinical trials website because we recruited pa-
up. These 108 recruited patients were similar in recruiting crite- tients prior to development of that website.
ria to those not recruited. The 36 F + V patients received We measured plasma and urine creatinine and urine albumin
an amount of fresh F + V free of charge to reduce their dietary using the Sigma Diagnostics Creatinine Kit (Procedure No. 555,
acid by half as done previously [12, 13, 33], focusing primarily on Sigma Diagnostics) [31]. Plasma Cystatin-C was measured with
F + V that are particularly base producing [34]. We provided fruit a particle-enhanced immunonephelometric assay (N Latex Cys-
predominantly as apples, apricots, oranges, peaches, pears, rai- tatin C; Dade Behring, Somerville, NJ, USA) with a nephelometer
sins, and strawberries. We provided vegetables predominantly as (BNII; Dade Bering) [37]. We measured urine and PTCO2 with
carrots, cauliflower, eggplant, lettuce, potatoes, spinach, toma- ultrafluoremetry [29]. We used the Abcam (Cambridge, MA,
toes, and zucchini. We gave no specific dietary instructions and USA) cholesterol assay kit (ab65390) to measure LDL and HDL.
allowed patients to integrate the provided F + V into their ad lib
diets as they wished. The necessary amount of F + V was pre- Statistical Analysis
scribed by a dietician and distributed from a community center We described characteristics of the sample using frequencies
food bank and averaged 2–4 cups (1 cup is ~200 g, allowing for with percentages for categorical variables and means with SDs for
volume differences among foods) daily, depending on the base- all. To account for the dependency induced by the matching
producing capacity of the administered F + V. To assure that each scheme utilized in the sample design [38], we used mixed effects
F + V patient received sufficient F + V and did not spread the models to assess if there was a treatment effect, adjusting for a
prescribed amount throughout their families, we gave the amount random participant effect and using a variance component struc-
prescribed for a study patient to each household member. We did ture [39]. Estimates of contrasts between the group difference
not measure F + V intake beyond 3-day diaries. The 36 HCO3 from 5-year to baseline were provided along with 95% CIs using
patients received oral NaHCO3 0.3 mEq/Kg/day (average patient Tukey’s adjustment. We set statistical significance at 0.05 for a
dose was 25.2 mEq/day), an amount estimated to reduce dietary family wise error rate. Because we performed 46 comparisons, we
acid about half, as scored tablets containing 10 mEq NaHCO3 used the Holms-Bonferroni method to adjust the p values and
with sucrose in 3 divided daily doses. We dosed each patient to control the familywise error rate. Software used was R version
the nearest one-half tablet by lean body weight in kg (e.g., a 70 kg 3.5.0 [40] and the packages multgee [41] and nlme [42] and
patient received 2.0 tablets/day). Because we could not blind re- PASS16 [43].
130.237.165.40 - 4/21/2019 4:10:32 AM

440 Am J Nephrol 2019;49:438–448 Goraya/Munoz-Maldonado/Simoni/

Stockholm University Library

DOI: 10.1159/000500042 Wesson

Downloaded by:
Table 1. Baseline characteristics. All variables were approximately normal. A mixed effects models was used to adjust for the depen-
dency between the participants induced by the matching design. The adjusted p value is the p value after adjusting all p values reported
in the tables using the Holm-Bonferroni method for multiple comparisons

Variable UC HCO3 F+V p value Adjusted

(n = 36) (n = 36) (n = 36) p value

Matching variables
Age, years 53.9 (4.8) 53.6 (5.3) 53.5 (5.2)
Gender
Female 20 (55.6) 20 (55.6) 20 (55.6)
Male 16 (44.4) 16 (44.4) 16 (44.4)
Ethnicity
White 5 (13.9) 8 (22.2) 6 (16.7)
African American 19 (52.8) 19 (52.8) 19 (52.8)
Hispanic 12 (33.3) 9 (25.0) 11 (30.5)
eGFR 39.5 (6.9) 39.6 (6.6) 39.4 (6.4)
Ualb 315 (73.1) 317 (71.6) 318 (71.2)
Outcome variables
PTCO2 22.9 (0.6) 23.1 (0.6) 22.9 (0.6) 0.619 1.000
BMI 28.2 (2.1) 28.3 (2.1) 28.8 (2.1) 0.434 1.000
SBP, mm Hg 158.6 (10.7) 165.8 (10.7) 163.3 (10.7) 0.017 0.374
LDL, mg/dL 151.9 (19.0) 157.2 (19.0) 161.8 (19.0) 0.102 1.000
HDL, mg/dL 50.6 (5.9) 51.5 (5.9) 50.9 (5.9) 0.820 1.000
Lp(a), mg/dL 81.8 (13.0) 81.9 (13.0) 84.3 (13.0) 0.438 1.000
Vitamin K1, mM 0.44 (0.20) 0.45 (0.24) 0.43 (0.18) 0.845 1.000
UNaV, mmol/8 h 74.9 (6.1) 74.3 (6.1) 73.9 (6.1) 0.779 1.000
UKV, mmol/8 h 33.5 (5.6) 33.1 (5.6) 34.0 (5.6) 0.791 1.000
Fruit intake, cups/day 0.98 (0.24) 0.95 (0.24) 0.99 (0.24) 0.787 1.000

The first variables are the variables the participants were matched by. No p value is provided since, by design, the groups are similar.
Numerical variables are reported as mean (SD) or n (%).
UC = group treated according to extant guidelines but without dietary acid reduction therapy; HCO3 = group given dietary acid re-
duction therapy with oral NaHCO3 to treat metabolic acidosis; F + V = group given dietary acid reduction therapy as base-producing
fruits and vegetables to treat metabolic acidosis.
Linear Mixed models were used to adjust for the clustering due to matching and to calculate the SDs.
SBP, systolic blood pressure; eGFR, estimated glomerular filtration rate; PTCO2, plasma total CO2 (nearly identical to plasma
[HCO3]); BMI, body mass index; LDL, low-density lipoprotein cholesterol; Lp(a), lipoprotein (a); HDL, high-density lipoprotein cho-
lesterol; UC, usual care; Ualb, urine albumin.

Results 5  years, whether done with oral NaHCO3 (HCO3) or

Table 1 shows no difference in the listed demographic
characteristics among the 3 groups, as well as in eGFR and
Ualb (these were matching variables so we performed no
testing). Baseline PTCO2, BMI, LDL-c, HDL-c, Lp(a), vi-
tamin K, 8-h urine excretion of sodium (UNaV), potas-
sium (UKV), and cups/day of F + V eaten were not dif-
ferent among groups. Although baseline SBP was higher
in HCO3 than the other treatment groups, the difference
was not significant after adjusting the p value for the mul-
tiple tests performed.
Figure 2 shows the course of PTCO2 in the 3 groups.
Dietary acid reduction yielded a net PTCO2 increase at
F  +  V, consistent with improved metabolic acidosis, as
shown in Table 2. By contrast, UC who received no di-
etary H+ reduction had a net PTCO2 decrease, consistent
with worsened metabolic acidosis. Table 2 shows a great-
er BMI decrease from baseline in F + V than both HCO3
and UC at 5 years, but the BMI change in HCO3 was not
different from UC. Similarly, Table 2 shows a greater net
LDL-c decrease at 5 years in F + V than both HCO3 and
UC, but the LDL-c net decrease in HCO3 was not differ-
ent from UC. There were no differences in HDL-c chang-
es among groups. Table 2 shows that compared to base-
line, at 5 years that there was a net increase in Lp(a) in UC,
no significant change in HCO3, and a net decrease in
130.237.165.40 - 4/21/2019 4:10:32 AM

Metabolic Acidosis Treatment and Am J Nephrol 2019;49:438–448 441

Stockholm University Library

Cardiovascular Risk DOI: 10.1159/000500042

Downloaded by:

25
24
PTCO2, mM
23
22
Fig. 2. Course of PTCO2 during 5-year fol-
low-up of patients undergoing UC (n = 36),
0.3 mEq/kg bw NaHCO3 (HCO3, n = 36),
or receiving base-producing F + V. PTCO2,
plasma total CO2, UC, usual care; F + V,
fruits + vegetables.
F + V. The net change was significantly different from UC
in both HCO3 and F + V and was greater in F + V than
HCO3. Table 2 shows that at 5 years, there was a greater
net increase in vitamin K1 compared to baseline in F + V
but not in HCO3 or UC, with the change in UC and HCO3
being not different from zero.
Table 2 shows a net decrease in SBP at 5 years for all
3 groups, but the net decrease was greater in F + V than
HCO3 and UC. The net SBP decrease for HCO3 was not
statistically different from UC. Furthermore, a greater
percentage of F + V (mean 89%, 95% CI 73–97) achieved
the SBP goal of ≤130 mm Hg than HCO3 (17%, 95% CI
7–34) and UC (27%, 95% CI 13–46). Figure 3 shows the
course of eGFR in the 3 groups. Although Table 2 shows
a net eGFR decrease at 5 years for all 3 groups, both F +
V and HCO3 had a smaller net eGFR decrease than UC
at 5 years. Furthermore, a greater percentage of F + V
(40%, 95% CI 42–76%) and HCO3 (34%, 95% CI 46.5–
80.3%) achieved the goal of maintaining eGFR ≥30 mL/
min/1.73 m2 than UC (6%, 95% CI 1–20%). To assess if
BMI played a role in the eGFR decrease, we ran a model
with both BMI difference at 5 years and treatment group
including interaction between BMI and treatment
group. Neither the interaction term nor the BMI differ-
ence was statistically significant (p values = 0.42 and
0.84, respectively), indicating that the BMI difference
442 Am J Nephrol 2019;49:438–448
DOI: 10.1159/000500042
UC HCO3 F+V
0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5
Time, years
was not a significant factor in the net eGFR decrease at
5 years.
Five-year UNaV was lower in F + V than UC but that
for HCO3 was higher than UC as in Table 2. There was no
net UNaV change in UC, but F + V had a net decrease
while HCO3 had a net UNaV increase. The change in
UNaV for both F + V and HCO3 was significantly differ-
ent from the change for UC. On the other hand, Table 2
shows that 5-year UKV was higher in F + V than both
HCO3 and UC, but UKV in HCO3 and UC were not dif-
ferent from each other. There was no net UKV change in
HCO3 and UC, F + V had a net UKV increase, and the net
UKV change for F + V was greater than that for both
HCO3 and UC. Table 2 also shows that patient-recorded
5-year F + V consumption was higher in the F + V group
than both HCO3 and UC, but F + V consumption in
HCO3 and UC were not different from each other. There
was no net F + V consumption change in HCO3 and UC,
and the net F + V consumption change for F + V was
greater than that for both HCO3 and UC.
Table 3 shows the median change in medication dosage
at 5-year follow-up and the distribution of number of pa-
tients within each of the 3 groups who at 5 years had a de-
crease, no change, or an increase in the dose of the 7 top
prescribed medications for patients. The table shows that
there were significant differences among the 3 groups in
130.237.165.40 - 4/21/2019 4:10:32 AM
Goraya/Munoz-Maldonado/Simoni/
Stockholm University Library
Wesson
Downloaded by:
Table 2. Five-year level and net change (5-years minus baseline) estimates. The 5-year level p values correspond to comparisons of the
5-years average levels by intervention, while the adjusted p values are calculated using the Holm-Bonferroni method for multiple com-
parison. The contrast p values correspond to comparing the pairwise difference in change (5-years minus baseline) between interven-
tions using Tukey’s adjustment. Mixed-effects models were used to account for dependence induced by the matching design. Tukey
adjustment was used for the contrast p values of HCO3 and F + V groups vs. UC. The contrast adjusted p values are the Tukey’s p values
adjusted for the multiple comparison

Variable 5-year 5-year 5-year Net change 95% CI Contrast

estimate p value adjusted Estimate adjusted
p value p value

PTCO2 <0.001 0.005

UC (reference) 21.9 (0.4) –1.03 –1.25 to –0.81
HCO3 23.9 (0.4) 0.87 0.66 to 1.08 0.005
F+V 23.8 (0.4) 0.89 0.68 to 1.10 0.005
BMI <0.001 0.005
UC (reference) 27.8 (1.8) –0.5 –0.9 to –0.1
HCO3 28.4 (1.8) 0.1 –0.3 to 0.5 1.000
F+V 26.6 (1.8) –2.2 –2.6 to –1.8 0.005
LDL, mg/dL <0.001 0.005
UC (reference) 137.4 (10.9) –14.3 –20.6 to –9.2
HCO3 134.9 (10.9) –21.5 –27.0 to –16.0 1.000
F+V 115.8 (10.9) –45.5 –51.0 to –40.0 0.005
HDL, mg/dL 0.028 0.560
UC (reference) 52.0 (4.4) 1.6 0.3 to 2.9
HCO3 53.5 (4.4) 2.0 0.7 to 3.2 1.000
F+V 54.9 (4.4) 3.8 2.5 to 5.1 0.779
Lp(a), mg/dL <0.001 0.005
UC (reference) 88.2 (11.7) 5.4 3.5 to 7.3
HCO3 81.7 (11.7) –0.2 –2.1 to 1.6 0.005
F+V 73.7 (11.7) –10.9 –12.7 to –9.0 0.005
Vitamin K1, mM <0.001 0.005
UC (reference) 0.44 (0.2) –0.003 –0.01 to 0.03
HCO3 0.46 (0.2) 0.006 –0.03 to 0.04 1.000
F+V 0.72 (0.1) 0.29 0.25 to 0.323 0.005
SBP, mm Hg <0.001 0.005
UC (reference) 134.4 (4.7) –24.2 –28.3 to –20.0
HCO3 135.5 (4.7) –29.5 –33.6 to –24.4 1.000
F+V 125.4 (4.7) –37.5 –41.5 to –33.4 0.005
eGFR <0.001 0.005
UC (reference) 20.7 (5.8) –18.8 –19.5 to –18.2
HCO3 27.3 (5.7) –12.3 –12.9 to –11.7 0.005
F+V 29.4 (5.8) –10.0 –10.6 to –9.4 0.005
UNaV, mmol/8 h <0.001 0.005
UC (reference) 74.6 (6.0) –0.2 –1.2 to 0.8
HCO3 84.1 (6.0) 7.6 6.6 to 8.5 0.005
F+V 66.0 (6.0) –7.7 –8.6 to –6.7 0.005
UKV, mmol/8 h <0.001 0.005
UC (reference) 33.9 (5.1) –0.2 –1.0 to 0.5
HCO3 33.9 (5.0) 1.0 0.2 to 1.7 1.000
F+V 41.8 (5.0) 7.8 7.0 to 8.5 0.005
F + V intake, cups/day <0.001 0.005
UC (reference) 1.0 (0.3) 0.0 –1.0 to 0.09
HCO3 1.0 (0.3) 0.0 –0.07 to 0.1 1.000
F+V 3.1 (0.3) 2.1 2.0 to 2.2 0.005

F + V, fruits and vegetables; UC, usual care; PTCO2, plasma total CO2; BMI, body mass index; LDL, low-density lipoprotein; HDL,
high-density lipoprotein; SBP, systolic blood pressure; eGFR, estimated glomerular filtration rate; UNaV, urine excretion of sodium;
UKV, urine excretion of potassium.
130.237.165.40 - 4/21/2019 4:10:32 AM

Metabolic Acidosis Treatment and Am J Nephrol 2019;49:438–448 443

Stockholm University Library

Cardiovascular Risk DOI: 10.1159/000500042

Downloaded by:

60
55
50
45
eGFR, mL/min/1.73 m2
40
35
30
25
20
Fig. 3. Course of eGFR during 5-year fol- 15
low-up of patients undergoing UC (n = 36),
0.3 mEq/kg bw NaHCO3 (HCO3, n = 36), 10
or receiving base-producing F + V. eGFR,
estimated glomerular filtration rate, UC,
usual care; F + V, fruits + vegetables.
the percentage of patients who had changes in doses of
clonidine and atorvastatin, but the adjusted p value for
enalapril was not significant. For each of these drugs,
F + V had a greater number of patients who had had a de-
crease in dosage than HCO3 and UC. Descriptively, more
F + V patients had a decrease in dosage of diltiazem and
hydrochlorthiazide. There was no difference in the num-
ber of distribution of patients who had changes in drug
doses among the 3 groups for furosemide and atenolol.
Discussion
The current study showed that dietary acid reduction
with either NaHCO3 or F + V yielded comparable im-
provement of metabolic acidosis in patients with CKD,
but F + V yielded better improvement in CVD risk indi-
cators including SBP, LDL, Lp(a), BMI, and serum vita-
min K1 as secondary outcomes of these studies in which
the primary outcome was change in eGFR. These data
support the need for subsequent studies with larger sam-
ple size and longer follow-up to test if the better improve-
ment in CVD risk indicators with F + V translates into
fewer adverse CVD outcomes than treatment with oral
NaHCO3. If such follow-up studies show that F + V com-
pared to NaHCO3 treatment of metabolic acidosis indeed
444 Am J Nephrol 2019;49:438–448
DOI: 10.1159/000500042
UC HCO3 F+V
0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5
Time, years
yields fewer CVD outcomes, F + V might be preferred
initial therapy for CKD-related metabolic acidosis, de-
spite it being more challenging than NaHCO3 to imple-
ment, given its potential to concomitantly reduce the sub-
stantial CVD risk in patients with CKD [2–4].
We previously reported in interim analysis for these
same patients that both dietary H+ reduction interven-
tions better preserved eGFR than UC at 3 years [13]. The
current 5-year eGFR results for these same patients are
qualitatively the same, supporting that these interven-
tions provide long-term eGFR protection.
Management approaches that have successfully re-
duced CVD risk in non-CKD patients might require
modification in CKD to achieve similar success. Some re-
ports suggest that traditional CVD risk factors such as
increased serum lipids, including LDL-c, are less predic-
tive of CVD risk in CKD compared to non-CKD patients
[44, 45]. Recent reports, however, support that increased
LDL-c is indeed predictive of major CVD events in pa-
tients with CKD and that its reduction is associated with
a reduced risk of CVD events [18]. Also, statins might be
less effective in lowering LDL-c as eGFR declines [46]
and/or might promote vascular calcification through in-
hibition of vitamin K synthesis [47], suggesting that this
standard CVD risk reduction intervention might be less
effective, or possibly even detrimental, in patient with
130.237.165.40 - 4/21/2019 4:10:32 AM
Goraya/Munoz-Maldonado/Simoni/
Stockholm University Library
Wesson
Downloaded by:
Table 3. Number of participants (percentage) that had increase, no change, or decrease in dosage for each of the prescribed medication
at the end of the 5-year study. Because each medication has different dosage presentation, the median (range) of the change in dose is
also reported. The p value corresponds to a multinomial ordinal model testing if the treatment is a predictor of change in dosage. The
adjusted p value was calculated using the Holm-Bonferroni method for multiple comparisons

Variable UC HCO3 F+V p value Adjusted

(n = 36) (n = 36) (n = 36) p value

Enalapril 0.026 0.546

Median (range), mg/day 0 (–10 to 10) 0 (–10 to 5) –5 (–10 to 0)
Decrease 14 (38.9) 17 (47.2) 25 (69.4)
No change 16 (44.4) 15 (41.7) 11 (30.4)
Increase 6 (16.6) 4 (11.2) 0 (0.0)
Diltiazem*
Median (range), mg/day 0 (–60 to 90) 0 (–120 to 60) –135 (–240 to –90)
Decrease 4 (11.1) 7 (19.4) 36 (100.0)
No change 24 (66.7) 24 (66.7) 0 (0.0)
Increase 8 (22.2) 5 (13.9) 0 (0.0)
Clonidine <0.001 0.005
Median (range), mg/day 0 (–0.3 to 0.1) 0 (–0.3 to 0.3) 0.6 (–0.9 to 0)
Decrease 14 (38.9) 9 (25.0) 34 (94.4)
No change 21 (58.3) 25 (69.4) 2 (5.6)
Increase 1 (2.8) 2 (5.6) 0 (0.0)
Atenodol 0.109 1.000
Median (range), mg/day 0 (–50 to 0) 0 (–50 to 0) 0 (–100 to 0)
Decrease 5 (13.9) 2 (5.6) 8 (22.2)
No change 31 (86.1) 34 (94.4) 28 (77.8)
Increase 0 (0.0) 0 (0.0) 0 (0.0)
HCTZ◆
Median (range), mg/day 0 (–37.5 to 0)
Decrease 0 (0.0) 0 (0.0) 7 (19.4)
No change 36 (100.0) 36 (100.0) 29 (80.6)
Increase 0 (0.0) 0 (0.0) 0 (0.0)
Furosemide 0.099 1.000
Median (range), mg/day 0 (0 to 20) 0 (–20 to 0) 0 (–40 to 0)
Decrease 1 (2.8) 2 (5.6) 6 (16.7)
No change 33 (91.6) 34 (94.4) 30 (83.3)
Increase 2 (5.6) 0 (0.0) 0 (0.0)
Atorvastatin <0.001 0.005
Median (range), mg/day 0 (–40 to 0) 0 (–40 to 0) –40 (–80 to 0)
Decrease 14 (38.9) 12 (33.3) 34 (94.4)
No change 22 (61.1) 24 (66.7) 2 (5.6)
Increase 0 (0.0) 0 (0.0) 0 (0.0)

* No test was possible since 100% of the F + V group had a decrease making the matrix used in the calculation nonpositive definite.
◆ No test was possible since in the UC and the HCO group 100% of the participants had no change in dose.
3
UC, usual care; F + V, fruits and vegetables.

CKD. Furthermore, CAC is more common and more se-
vere in CKD [21] and has a stronger association with risk
of CVD and mortality in CKD patients [22]. Moreover,
F + V reduced LDL-c [17] and diets high in F + V are as-
sociated with reduced CAC [19, 20], additionally sup-
porting F + V as an effective adjunct to standard CVD risk
reduction strategies to reduce the high CVD risk in CKD
patients [2] which metabolic acidosis increases further
[5]. Large-scale studies with longer follow-up will better
test this hypothesis.
Although epidemiologic studies suggest that metabol-
ic acidosis further increases the already increased CVD
risk in CKD [5], the present studies support that correc-
tion of metabolic acidosis with F + V more comprehen-
sively improved the examined CVD risk factors than cor-
rection with NaHCO3. Among the examined CVD risk
130.237.165.40 - 4/21/2019 4:10:32 AM

Metabolic Acidosis Treatment and Am J Nephrol 2019;49:438–448 445

Stockholm University Library

Cardiovascular Risk DOI: 10.1159/000500042

Downloaded by:
parameters, the only one that was significantly better in
HCO3 than UC was lower 5-year Lp(a) in HCO3 than UC.
This lower 5-year Lp(a) in HCO3 was due to a net Lp(a)
increase in UC associated with no net change in Lp(a) in
HCO3 patients. Because a decrease in GFR has been as-
sociated with an increase in Lp(a) [48], lower eGFR, and
the greater net decreased eGFR in UC compared to HCO3
patients might have contributed to the increase in Lp(a)
over 5 years and so the higher Lp(a) in UC than HCO3 at
5 years. Relatedly, better-preserved eGFR in HCO3 might
have contributed to maintaining Lp(a) in HCO3 in the
face of a significantly greater eGFR decline in UC. In ad-
dition, F + V compared to HCO3 patients had comparable
PTCO2; yet, F + V had greater improvement in the exam-
ined CVD risk parameters as stated. Further studies will
clarify the effect of metabolic acidosis, independent of
GFR, on Lp(a) and on mortality in CKD.
Because CVD-related mortality is inversely related to
GFR [49], better-preserved eGFR with dietary H+ reduc-
tion might lead to fewer adverse CVD events in the inter-
vention groups than UC with longer follow-up. Because
both Na+-based alkali therapy [13, 26, 50] and F + V [13,
51] preserved eGFR in CKD patients with metabolic aci-
dosis, these 2 dietary acid reduction interventions poten-
tially reduce CVD risk by slowing eGFR decline rate. De-
spite comparable eGFR preservation, F + V improved the
examined CVD risk parameters better than HCO3. Lower
eGFR is associated with faster CAC progression [52] that
possibly contributes to higher CVD risk associated with
reduced eGFR [49]. The F + V benefit on CAC [21, 22],
possibly mediated by increased vitamin K1 levels, and
other described benefits of F + V on the examined risk
parameters might mediate fewer subsequent adverse
CVD events in F + V.
Predictive models for CVD mortality risk have begun
incorporating eGFR and albuminuria. One such model
deems CKD patients fitting the recruitment criteria for the
present studies of eGFR (30–59 mL/min/1.73 m2) and
macroalbuminuria to be at “high” risk with a 10-year CVD
mortality of 5–10% [53]. The present 5-year follow-up
yielded 3 deaths in UC, one in HCO3, and none in F + V in
the groups of 36 patients each. Because about 13% of pa-
tients with the recruited eGFR range have metabolic acido-
sis [54], only a minority of such patients might have their
CVD-related mortality reduced by correction of their met-
abolic acidosis alone. This favors the F + V over the NaH-
CO3 intervention that appears to contribute CVD risk re-
duction beyond that due to metabolic acidosis alone.
Limitations of the study include the CVD risk indica-
tors described and CVD risk indicators being secondary
446 Am J Nephrol 2019;49:438–448
DOI: 10.1159/000500042
and not primary. Also, all participants had nondiabetic,
hypertension-associated CKD so follow-up studies must
include individuals with other CKD causes, importantly
diabetes, to see if the outcomes reported would be similar.
In addition, most participants were either black or His-
panic and so follow-up studies should include a more di-
verse panel of participants.
In summary, these studies show that dietary H+ reduc-
tion treatment of CKD-related metabolic acidosis with
F + V and NaHCO3 yielded comparable improvement of
metabolic acidosis and comparable eGFR preservation.
Nevertheless, F + V yielded better improvement of CVD
risk parameters compared to HCO3 and UC patients.
These data support the need for follow-up studies to ex-
amine this potential adjunct to standard CVD risk reduc-
tion strategies in CKD participants with metabolic acido-
sis.
Acknowledgments
We are thankful to the nursing and clerical staff of the Internal
Medicine Clinic of the Department of Internal Medicine at Texas
Tech University Health Sciences Center for their assistance, to the
Inside Out Community Outreach Program of Lubbock, Texas, and
to the Academic Operations division of Baylor Scott and White
Health for making these studies possible.
Disclosure Statement
Dr. Donald E. Wesson has part of his salary paid through his
employing institution for consultant work done for Tricida, Inc.
None of the remaining authors has conflicts to disclose.
Funding Source
This work was supported by funds from the Larry and Jane
Woirhaye Memorial Endowment in Renal Research the Texas
Tech University Health Sciences Center, by the Statistics Depart-
ment of Scott and White Healthcare, and by the Academic Opera-
tions Division at Scott and White Healthcare.
References   1 Keith DS, Nichols GA, Gullion CM, Brown
JB, Smith DH. Longitudinal follow-up and
outcomes among a population with chronic
kidney disease in a large managed care orga-
nization. Arch Intern Med. 2004 Mar;164(6):
659–63.
  2 US Renal Data System. USRDS 2016 Annual
Data Report. Bethesda (MD): The National In-
stitutes of Health, National Institute of Diabe-
tes and Digestive and Kidney Diseases; 2017.
130.237.165.40 - 4/21/2019 4:10:32 AM
Goraya/Munoz-Maldonado/Simoni/
Stockholm University Library
Wesson
Downloaded by:
  3 Shroff GR, Frederick PD, Herzog CA. Renal
failure and acute myocardial infarction: clini-
cal characteristics in patients with advanced
chronic kidney disease, on dialysis, and with-
out chronic kidney disease. A collaborative
project of the United States Renal Data Sys-
tem/National Institutes of Health and the Na-
tional Registry of Myocardial Infarction. Am
Heart J. 2012 Mar;163(3):399–406.
  4 Masson P, Webster AC, Hong M, Turner R,
Lindley RI, Craig JC. Chronic kidney disease
and the risk of stroke: a systematic review and
meta-analysis. Nephrol Dial Transplant. 2015
Jul;30(7):1162–9.
  5 Kovesdy CP, Anderson JE, Kalantar-Zadeh K.
Association of serum bicarbonate levels with
mortality in patients with non-dialysis-de-
pendent CKD. Nephrol Dial Transplant. 2009
Apr;24(4):1232–7.
  6 Clase CM, Kiberd BA, Garg AX. Relationship
between glomerular filtration rate and the
prevalence of metabolic abnormalities: results
from the Third National Health and Nutri-
tion Examination Survey (NHANES III).
Nephron Clin Pract. 2007;105(4):c178–84.
  7 Remer T. Influence of nutrition on acid-base
balance—metabolic aspects. Eur J Nutr. 2001
Oct;40(5):214–20.
  8 Adeva MM, Souto G. Diet-induced metabolic
acidosis. Clin Nutr. 2011 Aug;30(4):416–21.
 9 Guidelines KD. Chapter 3: Management of
progression and complications of CKD. Kid-
ney Int Suppl (2011). 2013 Jan;3(1):73–90.
10 Moore LV, Thompson FE. Adults Meeting
Fruit and Vegetable Intake Recommenda-
tions - United States, 2013. MMWR Morb
Mortal Wkly Rep. 2015 Jul;64:709–13.
11 Fernandes AS, Ramos CK, Nerbass FB, Cup-
pari L. Diet quality of chronic kidney disease
patients and the impact of nutritional coun-
seling. J Ren Nutr. 2018 Nov;28(6):403–10.
12 Goraya N, Simoni J, Jo CH, Wesson DE. A
comparison of treating metabolic acidosis in
CKD stage 4 hypertensive kidney disease with
fruits and vegetables or sodium bicarbonate.
Clin J Am Soc Nephrol. 2013 Mar; 8(3): 371–
81.
13 Goraya N, Simoni J, Jo CH, Wesson DE. Treat-
ment of metabolic acidosis in patients with
stage 3 chronic kidney disease with fruits and
vegetables or oral bicarbonate reduces urine
angiotensinogen and preserves glomerular fil-
tration rate. Kidney Int. 2014 Nov;86:1031–8.
14 Wang X, Ouyang Y, Liu J, Zhu M, Zhao G,
Bao W, et al. Fruit and vegetable consumption
and mortality from all causes, cardiovascular
disease, and cancer: systematic review and
dose-response meta-analysis of prospective
cohort studies. BMJ. 2014 Jul;349:g4490.
15 Gutiérrez OM, Muntner P, Rizk DV, McClel-
lan WM, Warnock DG, Newby PK, et al. Di-
etary patterns and risk of death and progres-
sion to ESRD in individuals with CKD: a co-
hort study. Am J Kidney Dis. 2014 Aug;64(2):
204–13.
16 Smyth A, Griffin M, Yusuf S, Mann JF, Red-
dan D, Canavan M, et al. Diet and major renal
Metabolic Acidosis Treatment and
Cardiovascular Risk
outcomes: A prospective cohort study. The
NIH-AARP Diet and health study. J Ren
Nutr. 2016 Sep;26(5):288–98.
17 Djoussé L, Arnett DK, Coon H, Province MA,
Moore LL, Ellison RC. Fruit and vegetable
consumption and LDL cholesterol: the Na-
tional Heart, Lung, and Blood Institute Fam-
ily Heart Study. Am J Clin Nutr. 2004 Feb;
79(2):213–7.
18 Storey BC, Staplin N, Haynes R, Reith C, Em-
berson J, Herrington WG, et al.; SHARP Col-
laborative Group. Lowering LDL cholesterol
reduces cardiovascular risk independently of
presence of inflammation. Kidney Int. 2018
Apr;93(4):1000–7.
19 Miedema MD, Petrone A, Shikany JM,
Greenland P, Lewis CE, Pletcher MJ, et al. As-
sociation of fruit and vegetable consumption
during early adulthood with the prevalence of
coronary artery calcium after 20 years of fol-
low-up: The Coronary Artery Risk Develop-
ment in Young Adults (CARDIA) Study. Cir-
culation. 2015 Nov;132(21):1990–8.
20 Greenland P, Bonow RO, Brundage BH, Bu-
doff MJ, Eisenberg MJ, Grundy SM, et al.;
Society of Cardiovascular Computed To-
mography. ACCF/AHA 2007 clinical expert
consensus document on coronary artery cal-
cium scoring by computed tomography in
global cardiovascular risk assessment and in
evaluation of patients with chest pain: a re-
port of the American College of Cardiology
Foundation Clinical Expert Consensus Task
Force (ACCF/AHA Writing Committee to
Update the 2000 Expert Consensus Docu-
ment on Electron Beam Computed Tomog-
raphy). Circulation. 2007 Jan; 115(3): 402–
26.
21 Budoff MJ, Rader DJ, Reilly MP, Mohler ER
3rd, Lash J, Yang W, et al.; CRIC Study In-
vestigators. Relationship of estimated GFR
and coronary artery calcification in the
CRIC (Chronic Renal Insufficiency Cohort)
Study. Am J Kidney Dis. 2011 Oct; 58(4):
519–26.
22 Chen J, Budoff MJ, Reilly MP, Yang W, Rosas
SE, Rahman M, et al.; CRIC Investigators.
Coronary artery calcification and risk of car-
diovascular disease and death among patients
with chronic kidney disease. JAMA Cardiol.
2017 Jun;2(6):635–43.
23 Proudfoot D, Shanahan CM. Molecular
mechanisms mediating vascular calcification:
role of matrix Gla protein. Nephrology (Carl-
ton). 2006 Oct;11(5):455–61.
24 Cranenburg EC, Schurgers LJ, Uiterwijk HH,
Beulens JW, Dalmeijer GW, Westerhuis R, et
al. Vitamin K intake and status are low in he-
modialysis patients. Kidney Int. 2012 Sep;
82(5):605–10.
25 National Kidney Foundation. K/DOQI Clini-
cal Practice Guidelines for Nutrition in
Chronic Renal Failure. Am J Kidney Dis. 2000
Jun;35(6 Suppl 2):S1–140.
26 Phisitkul S, Khanna A, Simoni J, Broglio K,
Sheather S, Rajab MH, et al. Amelioration of
metabolic acidosis in patients with low GFR
Am J Nephrol 2019;49:438–448
DOI: 10.1159/000500042
reduced kidney endothelin production and
kidney injury, and better preserved GFR. Kid-
ney Int. 2010 Apr;77(7):617–23.
27 Mahajan A, Simoni J, Sheather SJ, Broglio KR,
Rajab MH, Wesson DE. Daily oral sodium bi-
carbonate preserves glomerular filtration rate
by slowing its decline in early hypertensive
nephropathy. Kidney Int. 2010 Aug; 78(3):
303–9.
28 Klahr S, Levey AS, Beck GJ, Caggiula AW,
Hunsicker L, Kusek JW, et al.; Modification of
Diet in Renal Disease Study Group. The ef-
fects of dietary protein restriction and blood-
pressure control on the progression of chron-
ic renal disease. N Engl J Med. 1994 Mar;
330(13):877–84.
29 Wesson DE. Dietary HCO3 reduces distal tu-
bule acidification by increasing cellular
HCO3 secretion. Am J Physiol. 1996 Jul;271(1
Pt 2):F132–42.
30 Warmoth L, Regalado MM, Simoni J, Harrist
RB, Wesson DE. Cigarette smoking enhances
increased urine albumin excretion as a risk
factor for glomerular filtration rate decline in
primary hypertension. Am J Med Sci. 2005;
330:111–9.
31 Regalado M, Yang S, Wesson DE. Cigarette
smoking is associated with augmented pro-
gression of renal insufficiency in severe essen-
tial hypertension. Am J Kidney Dis. 2000 Apr;
35(4):687–94.
32 Roehm B, Simoni J, Prusynski J, Wesson DE.
Cigarette smoking Attenuates kidney protec-
tion by angiotensin converting enzyme inhi-
bition in nondiabetic CKD. Am J Nephrol.
2017;46(4):260–7.
33 Goraya N, Simoni J, Jo CH, Wesson DE. Di-
etary acid reduction with fruits and vegetables
or bicarbonate attenuates kidney injury in pa-
tients with a moderately reduced glomerular
filtration rate due to hypertensive nephropa-
thy. Kidney Int. 2012 Jan;81(1):86–93.
34 Remer T, Manz F. Potential renal acid load of
foods and its influence on urine pH. J Am Diet
Assoc. 1995 Jul;95(7):791–7.
35 Chobanian AV, Bakris GL, Black HR, Cush-
man WC, Green LA, Izzo JL Jr, et al.; Nation-
al High Blood Pressure Education Program
Coordinating Committee. The Seventh Re-
port of the Joint National Committee on Pre-
vention, Detection, Evaluation, and Treat-
ment of High Blood Pressure: the JNC 7 re-
port. JAMA. 2003 May;289(19):2560–72.
36 Expert Panel on Detection, Evaluation, and
Treatment of High Blood Cholesterol in
Adults. Executive Summary of The Third
Report of The National Cholesterol Educa-
tion Program (NCEP) Expert Panel on De-
tection, Evaluation, And Treatment of High
Blood Cholesterol In Adults (Adult Treat-
ment Panel III). JAMA. 2001 May; 285(19):
2486–97.
37 Erlandsen EJ, Randers E, Kristensen JH. Eval-
uation of the Dade Behring N Latex Cystatin
C assay on the Dade Behring Nephelometer II
System. Scand J Clin Lab Invest. 1999 Feb;
59(1):1–8.
130.237.165.40 - 4/21/2019 4:10:32 AM
447
Stockholm University Library
Downloaded by:
38 Breslow, NE, Day NE. Statistical methods in
Cancer Research. Volume I -The analysis of
case-control studies. 1980. Who. Internation-
al Agency for Research Cancer scientific pub-
lication No. 32.
39 Sainani K. The importance of accounting for
correlated observations. PM R. 2010 Sep;2(9):
858–61.
40 R version 3.5.0 (2018-04-23) – "Joy in Play-
ing". Copyright © 2018 The R Foundation for
Statistical Computing, Platform: x86_64-w64-
mingw32/x64 (64-bit).
41 Touloumis A. R Package multgee: A General-
ized Estimating Equations Solver for Multi-
nomial Responses. J Stat Softw. 2015; 64(8):
1–14.
42 Pinheiro J, Bates D. DebRoy S, Sarkar D, R
Core Team (2018). Nlme: Linear and Nonlin-
ear Mixed Effects Models. R Package version
3.1-137, URL: https://CRAN.R-project.org/
package=nlme.
43 Kilpatrick RD, McAllister CJ, Kovesdy CP,
Derose SF, Kopple JD, Kalantar-Zadeh K. As-
sociation between serum lipids and survival
in hemodialysis patients and impact of race. J
Am Soc Nephrol. 2007 Jan;18(1):293–303.
44 PASS 16 Power Analysis and Sample Size
Software (2018), NCSS, LLC. Kaysville, Utah,
USA, ncss.com/software/pass.
45 Shlipak MG, Fried LF, Cushman M, Manolio
TA, Peterson D, Stehman-Breen C, et al. Car-
diovascular mortality risk in chronic kidney
disease: comparison of traditional and novel
risk factors. JAMA. 2005 Apr; 293(14): 1737–
45.
46 Herrington WG, Emberson J, Mihaylova B,
Blackwell L, Reith C, Solbu MD, et al.; Choles-
terol Treatment Trialists’ (CTT) Collabora-
tion. Impact of renal function on the effects of
LDL cholesterol lowering with statin-based
regimens: a meta-analysis of individual par-
ticipant data from 28 randomised trials. Lan-
cet Diabetes Endocrinol. 2016 Oct;4(10):829–
39.
47 Chen Z, Qureshi AR, Parini P, Hurt-Camejo
E, Ripsweden J, Brismar TB, et al. Does statins
promote vascular calcification in chronic kid-
ney disease? Eur J Clin Invest. 2017 Feb;47(2):
137–48.
48 Lin J, Khetarpal SA, Terembula K, Reilly MP,
Wilson FP: Relation of atherogenic lipopro-
teins with estimated glomerular filtration rate
decline: a longitudinal study. BMC Nephrol
2015 Aug;16:130.
49 Go AS, Chertow GM, Fan D, McCulloch CE,
Hsu CY. Chronic kidney disease and the risks
of death, cardiovascular events, and hospital-
ization. N Engl J Med. 2004 Sep; 351(13):
1296–305.
50 de Brito-Ashurst I, Varagunam M, Raftery
MJ, Yaqoob MM. Bicarbonate supplementa-
tion slows progression of CKD and improves
nutritional status. J Am Soc Nephrol. 2009
Sep;20(9):2075–84.
51 Garneata L, Stancu A, Dragomir D, Stefan G,
Mircescu G. Ketoanalogue-supplemented
vegetarian very low-protein diet and CKD
progression. J Am Soc Nephrol. 2016 Jul;
27(7):2164–76.
52 Bundy JD, Chen J, Yang W, Budoff M, Go AS,
Grunwald JE, et al.; CRIC Study Investigators.
Risk factors for progression of coronary ar-
tery calcification in patients with chronic kid-
ney disease: the CRIC study. Atherosclerosis.
2018 Apr;271:53–60.
53 Perk J, De Backer G, Gohlke H, Graham I,
Reiner Z, Verschuren M, et al.; European As-
sociation for Cardiovascular Prevention &
Rehabilitation (EACPR); ESC Committee for
Practice Guidelines (CPG). European Guide-
lines on cardiovascular disease prevention in
clinical practice (version 2012). The Fifth
Joint Task Force of the European Society of
Cardiology and Other Societies on Cardiovas-
cular Disease Prevention in Clinical Practice
(constituted by representatives of nine societ-
ies and by invited experts). Eur Heart J. 2012
Jul;33(13):1635–701.
54 Raphael KL, Zhang Y, Ying J, Greene T.
Prevalence of and risk factors for reduced se-
rum bicarbonate in chronic kidney disease.
Nephrology (Carlton). 2014 Oct; 19(10):
648–54.

130.237.165.40 - 4/21/2019 4:10:32 AM

448 Am J Nephrol 2019;49:438–448 Goraya/Munoz-Maldonado/Simoni/

Stockholm University Library

DOI: 10.1159/000500042 Wesson

Downloaded by: