The effect of sodium bicarbonate administration on renal function: A systematic review and meta-analysis of clinical trials

Background Metabolic acidosis is a prevalent condition in patients with chronic kidney disease (CKD). Although sodium bicarbonate is extensively used for management of metabolic acidosis, its e�cacy has not been summarized in previous review studies. Objective To conduct a systematic review and meta-analysis to estimate the overall effects of sodium bicarbonate on indices of renal function in patients with CKD.


Introduction
Chronic kidney disease (CKD) is a progressive condition that gradually induces serious health complications, including hyperlipidemia, cardiovascular and bone disease, and electrolyte disturbance (1).For industrialized and developing countries, CKD is a dangerous condition due to the wide-scale consumption of animal and fruit-based acids; where such acid-inducing diets can cause metabolic acidosis and exacerbate CKD (2).These acid loads provoke endothelin and aldosterone production and induce brosis, progressive decline in kidney function and adverse cardiovascular effects (3,4).However, these effects may be reversed by alkali supplementation; indeed, the National Kidney Foundation advocates alkali therapy in acidotic patients.Sodium bicarbonate is a major component of alkali therapy (5), where previous studies have shown that there is an inverse association between sodium bicarbonate administration and acidosis (6, 7).Also, sodium bicarbonate therapy has a signi cant increasing effect on serum bicarbonate level (8), thereby conferring a bene cial effect on the bicarbonate buffer system (9).Nevertheless, the e cacy of sodium bicarbonate therapy for CKD patients without signi cant acidosis in order to delay CKD progression remains unclear.Several studies assessed the effect of sodium bicarbonate therapy on indices of renal function (10); however, no systematic review and meta-analysis has evaluated published data regarding the e cacy of sodium bicarbonate in slowing down CKD progression.Also, the overall effect of sodium bicarbonate administration on indices of renal function is unclear.Therefore, the aim of the present study was to summarize the effect of sodium bicarbonate intake on kidney function in patients with CKD.

Results
As illustrated in Figure 1, we identi ed 1534 records (613 from the Medline, 521 from the Web of Science, and 400 from the Scopus) through literature searches.After removing duplicate reports, 1217 studies were screened against inclusion and exclusion criteria.Subsequently, 1167 reports were excluded because they were animal studies (n=139), editorial letter (n=8), review/systematic review studies (n=23), or not relevant (n=997).Fifty remaining studies were scrutinized in detail, and 37 articles were excluded because of: 1) using intravenous sodium bicarbonate, 2) co-supplementation of sodium bicarbonate with vitamin C, 3) using vegetables and fruits as an intervention in control group, 4) not reporting relevant data, or 5) enrolling subjects with acute kidney injury.Finally, 13 studies met all inclusion and exclusion criteria were included in this systematic review and meta-analysis (3,4,(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21) Characteristics of the included studies are presented in Table 1.Participants were from Hong Kong, Israel, Thailand, Chinese, Brazil, Italy, South Korea, England, Thailand, and the USA.All studies enrolled both genders.Participants in all studies were pre-dialysis patients, except for two studies that recruited peritoneal dialysis patients (1,7).The mean age ranged from 54 to 72 years old.The design of three studies was pre-post (3,5,6) and the remaining studies were parallel randomized clinical trials.Five studies administered an increasing dose of sodium bicarbonate (3,4,8,12,13) and the remaining trials used a xed dose.------------ The results of quality scores allocation showed that all studies were high-quality, except for two studies that received low scores (5,6).

Discussion
The results of the present systematic review and meta-analysis indicated that sodium bicarbonate could signi cantly increase GFR and creatinine clearance and decrease BUN and serum creatinine.These changes can decrease the risk of ESRD, consequently reducing the mortality of patients (17).
Our ndings were in line with some related review studies; for instance, a previous systematic review reported that alkali therapy is associated with an improvement in kidney function, which may cause a long-term bene t in slowing the progression of CKD (22).Another review study suggested that oral alkali supplementation or a reduction in dietary acid intake may slow the rate of kidney function decline and potentially reduce the risk of ESKD in patients with CKD and metabolic acidosis (23).
The results of sensitivity analysis revealed that excluding the Rizetto et al. study (14) attenuated bene cial effect of sodium bicarbonate on serum creatinine.This study showed a signi cant reduction in serum creatinine after sodium bicarbonate administration, however, the reported con dence interval for changes in serum creatinine was very short in this study compared with other included studies.Therefore, the weight of this study was high in our analysis and omitting it resulted in a notable change in overall effect size.For GRF, we found that excluding the Jeong et al study (12) attenuated observed effect of sodium bicarbonate.All included studies enrolled subjects at a certain stage of CKD, except for the Jeong et al study, which included participants with CKD4 and CKD5.Therefore, the design of this study was different from other included publications.
There are several putative mechanisms for the effect of acidosis on the CKD progression including ammonia-induced activation of the alternative complement system, overproduction of endothelin-1, angiotensin II, aldosterone and proin ammatory cytokines, and induction of tubulointerstitial brosis (24,25).It was observed that serum bicarbonate levels below the normal range increased QRS duration and arterial pressure (23).Intravenous sodium bicarbonate treatment increases plasma bicarbonate, adjusts excess hydrogen ion concentrations, raises blood pH, and reverses the clinical manifestations of acidosis (26).Alkaline oral treatment can reduce acidosis and thus reduce the progression of CKD and its complications, including protein catabolism, chronic in ammation, insulin resistance, and growth hormone (27).
Although sodium bicarbonate is a very effective drug in patients with CKD, it interacts with some of the most ubiquitous medications, such as captopril, iron compounds, multivitamins, chloroquine, corticosteroids, mesalamine, rosavastatin, sotalol, and dabigatran (28).These interactions are caused by various mechanisms, such as reduction/increases in absorption, bioavailability, serum concentration, and the effect of the medications (29).Therefore, drug-drug interaction should be considered in these cases.In the present review, the medications received by patients were not reported and drug-drug interaction was not considered, and thus, we could not include such information in our study.Future studies should consider possible interactions and contraindications between sodium bicarbonate and other prescribed medications in patients with CKD (29).
There is a concern regarding sodium content of the sodium bicarbonate (27% of sodium bicarbonate is sodium) and its potential deleterious effect of blood pressure.A clinical trial showed that acute loading of high dose (300 mg/kg) sodium bicarbonate could increase diastolic blood pressure among athletes (30).Although high dose sodium bicarbonate may be harmful, low/moderate dose sodium bicarbonate elicited no signi cant effect on blood pressure in hypertensive men.Additionally, compared with NaCl, sodium bicarbonate has been shown capable of decreasing blood pressure (31).Therefore, it seems that using sodium bicarbonate in high-doses may not be safe or advisable for patients with CKD.
This meta-analysis has some limitations and strengths.One of the limitations was using different doses of sodium bicarbonate (some studies used a xed dose and other studies used an increasing dose).Therefore, we could not draw a rm conclusion regarding intervention dose.
Another limitation was that no intervention was performed on hemodialysis patients and there was no information regarding these patients.However, a strength of this meta-analysis was that between-study heterogeneity was low among the studies; thereby allowing us to assert veracity in the reliability of our results.Also, most of the included studies were long-term interventions, and therefore, we were able to report the long-term effect of sodium bicarbonate effect on renal function.
In conclusion, this systematic and meta-analysis showed that sodium bicarbonate supplementation could elicit signi cant increases GFR and creatinine clearance and decrease BUN and creatinine.Therefore, sodium bicarbonate therapy may regarded as e cacious in ameliorating the progression of CKD.

Search strategy
This systematic and meta-analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines.A systematic literature search was carried out through the Medline (www.ncbi.nlm.nih.gov/pubmed),Web of Science and Scopus (www.scopus.com)databases until July 2020.The following terms were used in the electronic search to identify studies, with no restrictions: (''sodium bicarbonate'' OR "NaHCO 3 ") AND (''GFR'' OR ''glomerular ltration rate'' OR ''BUN'' OR "blood urea nitrogen" OR "urea" OR ''creatinine'' OR ''creatinine clearance'').In addition, the reference lists of all eligible articles were reviewed to nd more relevant studies lost in the initial search.

Eligibility criteria
We included relevant articles if they: 1) examined the effects of oral supplementation sodium bicarbonate on GFR, creatinine, creatinine clearance, serum urea, or BUN; 2) provided su cient information regarding baseline and endpoint serum levels of GFR, creatinine, creatinine clearance, or BUN; and 3) enrolled patients with CKD.No time and language limitation were applied.We excluded articles if they: 1) reported duplicate data, 2) were animal-based, reviews, letters, editorial articles, or case reports, 3) co-supplemented sodium bicarbonate with other agents, 4) used other medications in the control group.In instances of multiple publications that reported the same or overlapping data, the most recent, with the largest sample size, were included.
Quality quality selected articles was evaluated by two authors (M.T and M.H.R) using the Cochrane risk of bias tool (32).It was based on seven criteria including sequence generation, allocation sequence concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, and other potential threats to validity.Therefore, articles were scored as high risk, low risk, or unclear.

Data extraction
Two independent researchers (M.T and M.H.R) extracted relevant data.A third author (F.M) discussed and resolved any controversy.We extracted the following data: rst author name and publication year, country, sex-strati ed sample size, comorbidities among participants, mean age, design, duration, dose of intervention and intervention in comparison group.Also, mean, standard deviation or standard error of the following outcomes were extracted when available in the groups: GFR (ml/min/1.73m 2 ), BUN or serum urea (mg/dL), creatinine (mg/dL), and creatinine clearance (mL/min).

Statistical analysis
The current meta-analysis was conducted using STATA software (version 11.0; Stata Corporation).To calculate effect size, the reported net changes for main outcomes including GFR, BUN or serum urea, creatinine, and creatinine clearance in each study, respectively, were used.When net change was not reported, we used mean and SD at baseline and endpoint in sodium bicarbonate and control groups.Moreover, endpoint mean and standard error (SE) or standard deviation (SD) were used when baseline means were not reported.The overall effect size was calculated using a random-effects model and reported as weighted mean difference (WMD).We converted all reported data to the acceptable units including ml/min/1.73m 2 for GFR, mg/dL for BUN, serum urea and creatinine, and mL/min for creatinine clearance.Between-study heterogeneity was evaluated using the I-square (I 2 ) statistic.Subgroup analysis was performed when between-study heterogeneity was high.To evaluate the possible in uence of each single study on the pooled effect size, the stability of the results was checked through sensitivity analyses.Finally, Egger's regression asymmetry test and Begg's rank-correlation methods were conducted to assess the presence of publication bias.Statistical signi cance was accepted, a priori¸ at <0.05.

Declarations
Figures

Figure 1 Flow
Figure 1