Safety and efficacy of sodium bicarbonate for treating metabolic acidosis in chronic kidney disease: A systematic review and meta-analysis

Abstract

BACKGROUND

Kidney dysfunction and reduced filtration capacity due to chronic kidney disease (CKD) lead to a shift in the body's acid-base balance, ultimately causing metabolic acidosis (MA). Sodium bicarbonate has been used as a supplement to alleviate the symptoms and reverse the acidosis, and it may even slow the progression of CKD. However, its safety profile and overall effectiveness are uncertain.

AIM

To conduct a systematic review and meta-analysis of clinical trials assessing sodium bicarbonate's safety and efficacy for treating CKD-induced MA.

METHODS

Medline, Scopus, EMBASE, and Cochrane Central were systematically searched from inception until May 2024 to select all relevant randomized control trials (RCTs) and non-RCT (NRCTs) evaluating the effectiveness of sodium bicarbonate in correcting MA in end-stage renal disease patients. In addition, ClinicalTrials.gov, Medrxiv.org, and Google Scholar were searched for other literature. A random-effects meta-analysis was performed to derive mean differences (MD) and risk ratios (RR) with their 95%CI for continuous and dichotomous outcomes respectively.

RESULTS

Following a systematic search of the databases, 20 RCTs and 2 and NRCTs comprising 2932 patients were included in our study. The results revealed that sodium bicarbonate significantly increased serum bicarbonate in CKD patients (MD: 2.59, 95%CI: 0.95-4.22; P = 0.02; I² = 95%). However, there was a non-significant increase in estimated glomerular filtration rate (eGFR) in patients on sodium bicarbonate therapy (MD: 0.93, 95%CI: -1.88-3.75; P = 0.52; I² = 93%). Upon assessment of the safety profile of sodium bicarbonate, no significant association was found in the outcomes of death/prolonged hospitalization (RR: 1.05, 95%CI: 0.84-1.32; P = 0.66; I² = 0%), or gastrointestinal disorders (RR: 1.64, 95%CI: 0.35-7.66; P = 0.53; I² = 76%), or worsening edema (RR: 1.26, 95%CI: 0.94-1.68; P = 0.12; I² = 37%) when compared to control.

CONCLUSION

Sodium bicarbonate therapy may halt worsening kidney function by correcting serum bicarbonate levels and treating MA. Although sodium bicarbonate does not significantly improve the eGFR, it may potentially prevent CKD progression while maintaining an overall favorable safety profile.

Key Words: Chronic kidney disease; Metabolic acidosis; Safety and efficacy; Sodium bicarbonate

Core Tip: Sodium bicarbonate is a systemic alkalizer that increases serum bicarbonate, buffers hydrogen ions, and raises blood pH, thereby reversing the clinical manifestations of acidosis. It is often prescribed for the treatment of metabolic acidosis (MA) seen in severe renal disease. Studies have been conducted to assess the safety and efficacy of sodium bicarbonate as a pH equalizer. This meta-analysis aims to evaluate the safety profile and clinical efficacy of sodium bicarbonate in treating MA in patients with chronic kidney disease.

INTRODUCTION

Chronic kidney disease (CKD) refers to a progressive decline in kidney function over time. According to the Kidney Disease Improving Global Outcomes guidelines, CKD is characterized by reduced kidney function, as indicated by an estimated glomerular filtration rate (eGFR) of less than 60 mL/min per 1.73 m²[1]. Although the etiology of CKD varies globally, diabetes and hypertension are known to be the leading causes of CKD worldwide[2].

By eliminating hydrogen ions in the urine and reabsorbing bicarbonate from the urine, the kidneys typically play a significant role in maintaining the pH balance of extracellular fluid[3]. If compromised, it may lead to a variety of unfavorable consequences, including hyperlipidemia, anemia, metabolic bone disease, and cardiovascular disease[4]. Metabolic acidosis (MA), defined as an increase in hydrogen ion concentration in the systemic circulation resulting in a serum HCO₃ less than 24 mEq/L, is the most frequent complication among patients with CKD[5].

One of the main pathophysiological factors leading to the development of MA in CKD patients is a reduced ability to reabsorb filtered bicarbonate in conjunction with a compromised capacity to generate sufficient bicarbonate[6]. Acid retention because of this makes it difficult to balance the net endogenous acid load. It causes muscle atrophy, bone disease development or aggravation, hypoalbuminemia, elevated inflammation, CKD progression, protein malnutrition, changes in insulin, leptin, and growth hormone, and a higher death rate[7,8].

Several therapeutic options are available to counter MA, including alkali therapy, dietary changes, and hydration[9,10]. Of them, bicarbonate supplementation is the most important. It lowers the systemic acid load and neutralizes excess hydrogen ions to restore acid-base equilibrium. This can slow the progression of CKD and postpone the start of dialysis[11]. In contrast to other therapies like dietary adjustments, which may have a later onset of action, it helps safeguard bone health by avoiding acid-induced bone demineralization. This can be substantiated by a study where it was reported that correcting MA with oral sodium bicarbonate in patients with CKD stage III-V notably diminishes the progression from stage V CKD to dialysis and enhances patient survival, regardless of their initial renal function status[12]. While some articles highlight the benefits of sodium bicarbonate therapy, others do not show improvement in patient outcomes[13]. This discrepancy may be due to the adverse effects of sodium bicarbonate infusion for the treatment of MA, such as fluid overload, hypokalemia, hypocalcemia, alkalemia, and hypernatremia[14]. However, current evidence points to the potential benefits of sodium bicarbonate supplementation in CKD[9].

A meta-analysis was carried out to assess the advantages and safety of employing sodium bicarbonate supplementation to treat MA in patients with CKD. Compared to individual studies alone, this method offers a more thorough and trustworthy conclusion, which helps to improve comprehension of the overall effect of this therapy option. In addition to highlighting research gaps, this in-depth review supports evidence-based decision-making. It helps build well-informed healthcare policies, which improve patient care in managing CKD.

MATERIALS AND METHODS

This meta-analysis was conducted in accordance with the PRISMA guidelines and conforms to the framework set by the Cochrane Collaboration.

Data sources and search strategy

Medical literature databases, including Medline, Scopus, EMBASE, and Cochrane Central, were searched from inception until May 2024 using a search strategy based on PICO (patient, intervention, control, and outcomes). Search terms and MeSH terms used were “sodium bicarbonate”; “metabolic acidosis”; “chronic kidney disease”; and CKD. Boolean operators (“AND”, and “OR”) were used to form a search string. No limitations or filters were applied. Gray literature was identified via ClinicalTrials.gov, Medrxiv.org, and Google Scholar. A further expanded search was done into the references of the included studies to find any missing literature. A detailed description of the search strategy has been summarized in Supplementary Table 1.

Study selection

Two authors (Saad Khalid Hassan and Fizzah Batool) independently screened the results based on the title and abstract after duplicate removal was conducted on the EndNote Reference library (Version X7.5; Clarivate Analytics, Philadelphia, Pennsylvania). Any discrepancies were resolved through a discussion with a third author (Abdul Hannan Siddiqui). The inclusion criteria mandated to include: (1) Studies with adults over 18 years of age; (2) Patients having CKD with an eGFR of <15 to 60 mL/min per 1.73 m²; (3) Articles comparing sodium bicarbonate to a control; (4) Relevant primary or secondary outcomes; (5) Studies published in the English language; (6) Per oral intervention either as a powder or in capsule form; and (7) Original articles comprising of randomized controlled trials, non-randomized controlled trials or cohort studies. Meanwhile, letters, reviews, case reports, non-comparative studies, and studies with inadequate outcomes data were excluded.

Study outcomes

Our primary outcomes of interest were: (1) Changes in eGFR (mL/min); (2) Incidence of death/prolonged hospitalization, and (3) Changes in serum HCO₃ (mmol/L). The secondary outcomes of interest included: (1) Changes in edema; and (2) Gastrointestinal (GI) adverse effects.

Data extraction

The study design, study population, number of patients in each group, general patient characteristics (age and gender), sample size, comorbidities, follow-up period, primary and secondary endpoints were all extracted and verified onto an Excel sheet by two independent authors (Tatheer Zehra and Syed Shabbeer Rizvi). A third reviewer (Abdul Hannan Siddiqui) resolved discrepancies through consensus and discussion.

Quality assessment

Using the Revised Cochrane risk-of-bias tool for randomized controlled trials (ROB 2), two authors (Saad Usman and Shayan Khan) independently evaluated the quality of the clinical trials. Studies were examined for missing data, selective reporting of results, randomization of participants to exposure, and creation of allocation sequences. The Risk of Bias in Non-randomized Studies of Interventions tool was used to evaluate the quality of non-randomized clinical trials. A third author (Abdul Hannan Siddiqui) was sought to resolve any disputes.

Statistical analysis

All relevant meta-analyses of this study were conducted using RevMan (version 5.4.1; Copenhagen: The Nordic Cochrane Center, The Cochrane Collaboration, 2020), while the meta-regression analyses were carried out using Comprehensive Meta-Analysis (Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Comprehensive Meta-Analysis Version 3. Englewood, NJ: Biostat; 2013). The risk ratios (RR) and their 95%CI were pooled by using a random-effects model for both the meta-analyses and meta-regression. For every outcome, a P value < 0.05 was taken to be statistically significant. Using the Higgins I² test, heterogeneity was evaluated. The I² values of 25%-50% were considered mild, 50%-75% moderate, and > 75% significant heterogeneity. Sensitivity analysis was used to examine differences in the significance of the outcomes in studies with a > 50% degree of heterogeneity. The meta-regression results were presented as coefficients (Coeff) and P values.

RESULTS

Study selection and study characteristics

A total of 1397 studies were collected initially from all the databases. From them, 22 studies were included in our final analysis. They comprised 20 randomized control trials (RCTs)[15-34] and 2 clinical trials[35,36]. The sample size included 2932 patients (intervention: n = 1545; control: n = 1387), comprising 52.7% males. The overall mean age of patients was found to be 58 years. The average follow-up time was around 14.84 months. Most of these trials were conducted in the United States (Table 1 and Table 2). The PRISMA flow chart presents the detailed study selection process (Figure 1).

Figure 1 Preferred reporting items for systematic review and meta-analyses flowchart.

Table 1 Study characteristics.

Ref.	Study design	Study sample	Intervention/control name		Sample size		Follow-up duration	Dosage of sodium bicarbonate
			Intervention	Control	Intervention	Control	Months
Mathur et al[28]	RCT	Mild to moderate CKD with serum creatinine < 5 mg/dL	Sodium bicarbonate	Placebo	20	20	3	1.2 mEq/kg
Ashurst et al[27]	RCT	Creatinine clearance 15 to 30 ml/min per 1.73 m2 and serum bicarbonate 16 to 20 mmol/L	Sodium bicarbonate	Standard care	67	67	24	600 mg
Goraya et al[26]	RCT	eGFR, 15-29 ml/min per 1.73 m2 and PTCO2 level < 22 mm	Sodium bicarbonate	Fruits and vegetables	35	36	12	1.0 mEq/kg
Abramowitz et al[35]	Clinical trial	eGFR 15-45 ml/min per 1.73 m2 and serum bicarbonate 20-24 mEq/L	Sodium bicarbonate	Placebo	20	20	3	1.0 mEq/kg
Goraya et al[25]	RCT	eGFR 30-59 ml/min and metabolic acidosis characterized by plasma TCO2 > 22 but < 24 mmol/L	Sodium bicarbonate	Standard care	36	36	36	0.3 mEq/kg
Jeong et al[36]	Clinical trial	eGFR < 15-30 mL/min per 1.73 m2 and CO2 less than 22 mEq/L	Sodium bicarbonate	Standard care	40	40	12	1000 mg
Bellasi et al[29]	RCT	CKD-3b-4 with DM2 and serum bicarbonate levels below 24 mEql/L	Sodium bicarbonate	Standard care	71	74	12	0.5 mmol/kg
Yan et al[30]	RCT	eGFR of 15-60 ml/min/1.73 m2, serum HCO3- level of 16-20 mmol/L	Sodium bicarbonate	Placebo	42	42	4.25	N/A
Kendrick et al[31]	RCT	eGFR 15-44 ml/min per 1.73 m2 with low serum bicarbonate levels (16-21 mEq/L)	Sodium bicarbonate	Control group	10	10	6	650 mg
Di Iorio et al[22]	RCT	CKD stage 3-5, serum bicarbonate > 18 and < 24 mmol/L	Sodium bicarbonate	Standard care	376	364	36	N/A
Aigner et al[24]	RCT	CKD stage 3-4 with serum HCO3- ≤ 21 mmol/L	Sodium bicarbonate	Rescue group	18	17	1	840 mg
Goraya et al[23]	RCT	eGFR 30-59 mL/min/1.73 m2 and plasma total CO2 > 22 mM but < 24 mM	Sodium bicarbonate	Standard care	36	36	60	0.3 mEq/kg
Raphael et al[17]	RCT	eGFR 20-44 or 45-59 ml/min per 1.73 m2 and serum bicarbonate 20-28 meq/L	Sodium bicarbonate	Placebo	HD: 90; LD: 90	52	2.25	0.8 mEq/kg
Kittiskulnam et al[28]	RCT	eGFR 15-59 mL/min/1.73 m2, serum bicarbonate level < 22 mEq/L	Sodium bicarbonate	LD sodium bicarbonate	21	21	4	300 mg
Melamed et al[19]	RCT	eGFR of 15-59 ml/min/1.73m2, and serum bicarbonate levels 20 to 26 mEq/L	Sodium bicarbonate	Placebo	74	75	24	0.4 mEq/kg
Dubey et al[33]	RCT	CKD stages 3 and 4 and serum bicarbonate levels < 22 mEq/dL	Sodium bicarbonate	Standard care	94	94	6	0.5 mEq/kg
Witham et al[20]	RCT	CKD stages 3 and 4, and serum bicarbonate levels < 22 mEq/dL	Sodium bicarbonate	Placebo	152	148	24	500 mg
Alva et al[21]	RCT	eGFR of 15-30 mL/min/1.73 m2 and serum bicarbonate 10-20 mm/L	Sodium bicarbonate	Standard care	33	34	9	600 mg
Gaggl et al[34]	RCT	eGFR between 60 and 15 mL/min per 1.73 m2 and a serum HCO3- of ≤ 21 mmol/L	Sodium bicarbonate	Rescue group	24	23	2	N/A
Bovée et al[16]	RCT	eGFR 15-30 ml/min/1.73m2 and plasma bicarbonate levels between 15.0 and 24.0 mmol/L	Sodium bicarbonate	Sodium chloride	15	15	1	1000 mg
Mohebbi et al[32]	RCT	eGFR between 15 mL/min per 1.73 m² and 89 mL/min per 1.73 m² and serum bicarbonate of 22 mmol/L or less	Sodium bicarbonate	Placebo	119	121	24	500 mg
Sorohan et al[15]	RCT	eGFR between 45 and 15 mL/min/1.73 m2 and serum bicarbonate between 10 and 22 mmol/L	Sodium bicarbonate	Sodium citrate	62	62	12	600 mg

RCT: Randomized controlled trial; eGFR: Estimated glomerular filtration rate; CKD: Chronic kidney disease; HD: High dose; LD: Low dose; N/A: Not applicable.

Table 2 Baseline patient characteristics, n (%)/mean ± SD.

Ref.	Age (year)		Male		eGFR (mL/min)		Serum HCO3- (mmol/L)		Serum K+ (mmol/L)		Serum Na+ (mmol/L)
	Intervention	Control	Intervention	Control	Intervention	Control	Intervention	Control	Intervention	Control	Intervention	Control
Mathur et al[28]	37.5 ± 17	43.5 ± 10.5	12 (60)	13 (65)	-	-	19.49 ± 5.51	19.35 ± 3.74	-	-	-	-
Ashurst et al[27]	54.78 ± 2.56	54.77 ± 2.34	35 (52)	34 (51)	-	-	19.8 ± 2.2	19.9 ± 1.5	-	-	-	-
Goraya et al[26]	54.2 ± 5.3	53.9 ± 6.9	18 (51)	20 (56)	-	-	-	-	4.1 ± 0.2	4.1 ± 0.2	139.9 ± 1.2	139.8 ± 1.3
Abramowitz et al[35]	63.1 ± 10.8	-	8 (40)	-	32.9 ± 8.9	-	23.0 ± 2.4	-	4.6 ± 0.4	-	-	-
Goraya et al[25]	53.6 ± 5.3	53.9 ± 4.8	16 (44)	16 (44)	-	-	-	-	4.24 ± 0.12	4.29 ± 0.14	-	-
Jeong et al[36]	53.3 ± 13.5	55.8 ± 12.7	21 (52.5)	26 (57.5)	16.7 ± 6.1	17.7 ± 6.4			5.25 ± 0.71	5.44 ± 0.63	-	-
Bellasi et al[29]	64.9 ± 11.8	66.0 ± 12.9	47 (66)	36 (48)	-	-	21.2 ± 1.9	21.6 ± 2.0	-	-	-	-
Yan et al[30]	53.10 ± 6.11	53.10 ± 8.66	25 (59.5)	24 (57.1)	19.09 ± 2.26	18.43 ± 1.89	16.49 ± 1.22	16.14 ± 1.40	-	-	-	-
Kendrick et al[31]	59 ± 12	-	10 (50)	-	25 ± 8	24 ± 8	19.3 ± 2.9	19.7 ± 2.3	-	-	-	-
Di Iorio et al[22]	67.6 ± 15.1	68.1 ± 14.7	234 (62.2)	224 (61.5)	33.4 ± 12.4	36.9 ± 10.8	21.7 ± 2.6	21.4 ± 2.1	4.9 ± 0.5	4.9 ± 0.6	139 ± 2.9	139 ± 2.8
Aigner et al[24]	59.94 ± 12.46	53.12 ± 17.74	12 (66.7)	11 (52.9)	23.64 ± 5.87	27.45 ± 10.05	18.78 ± 1.59	19.28 ± 1.55	4.83 ± 0.47	4.94 ± 0.61	140 ± 3.03	139 ± 2.6
Goraya et al[23]	53.6 ± 5.3	53.9 ± 4.8	16 (44.4)	16 (44.4)	39.6 ± 6.6	39.5 ± 6.9	-	-	-	-	-	-
Raphael et al[17]	HD: 67 ± 11 LD: 66 ± 14	66 ± 11	HD: 61 (68) LD: 69 (60)	39 (75)	HD: 36 ± 10 LD: 37 ± 10	35 ± 9	HD: 24 ± 2 LD: 24 ± 2	24 ± 2	4.4 ± 0.4	4.5 ± 0.5	-	-
Kittiskulnam et al[28]	61.2 ± 9.8	61.2 ± 9.8	12 (57.1)	12 (57.1)	32.4 ± 14.1	-	21.0 ± 2.1	-	-	-	-	-
Melamed et al[19]	60.3 ± 14.1	61.6 ± 10.9	36 (49)	33 (44)	36.4 ± 11.4	36.2 ± 11.1	24.0 ± 2.2	24.1 ± 2.6	4.5 ± 0.5	4.6 ± 0.5	-	-
Dubey et al[33]	50.12 ± 11.6	50.30 ± 11.4	68 (72.3)	66 (70.2)	29.2 (27-31.3)	31.5 (29.3-33.8)	18.1 (17.7-18.6)	18.1 (17.6-18.6)	-	-	-	-
Witham et al[20]	73.9 ± 7.6	74.0 ± 6.6	110 (72.4)	104 (70.3)	19.7 ± 6.5	18.2 ± 6.4	20.6 ± 2.6	20.1 ± 2.5	4.9 ± 0.5	4.9 ± 0.5	-	-
Alva et al[21]	-	-	23 (69.7)	25 (73.53)	22.39 ± 4.08	21.21 ± 4.37	16.62 ± 3.05	16.84 ± 2.17	-	-	-	-
Gaggl et al[34]	58 ± 12.8	56 ± 16.5	15 (62)	14 (61)	25.33 ± 7.5	27 ± 9.7	18.8 ± 1.8	19.2 ± 1.5	-	-	-	-
Bovée et al[16]	61 ± 17	61 ± 14	11 (73)	14 (93)	21 ± 6	22 ± 3	20.8 ± 3.9	21.8 ± 2.9	5.1 ± 0.7	5.0 ± 0.6	-	-
Mohebbi et al[32]	55.7 ± 13.2	55.3 ± 13.8	82 (69)	85 (70)	48.2 ± 16.3	47.7 ± 15.8	21.3 ± 2.6	21.0 ± 2.7	-	-	-	-
Sorohan et al[15]	57.37 ± 10.25	57.95 ± 10.22	35 (56.5)	35 (56.5)	23.68 ± 9.05	24.93 ± 6.70	17.16 ± 1.62	17.53 ± 2.05	5.00 ± 0.34	4.88 ± 0.44	138.71 ± 1.41	138.44 ± 2.61

Quality assessment

Upon our assessment of the studies included, most RCTs were classified as having an overall "low" risk of bias. Studies by Witham et al[20], Alva et al[21] and Aigner et al[24] did show some concerns, while studies done by Sorohan et al[15], Yan et al[30], and Kendrick et al[31] were classified as having an overall "high" risk of bias (Supplementary Figure 1). The non-randomized studies of the effects of interventions were classified as having an overall "low to moderate" risk of bias owing to their robust methodology (Supplementary Figure 2).

Primary outcomes

Out of the 22 articles, nine studies reported data on change in eGFR, nine trials reported data on change in serum bicarbonate, and five trials reported the risk of death/prolonged hospitalization. Compared with the comparison group, the change in eGFR from baseline to trial completion was statistically insignificant in the intervention group (MD: 0.93, 95%CI: -1.88-3.75; P = 0.52; I² = 93%) (Figure 2). Regarding the change in serum bicarbonate concentration, the analysis showed a statistically significant difference between the two groups (MD: 2.59, 95%CI: 0.95-4.22; P = 0.002; I² = 95%) (Figure 3). The random effects analysis showed a statistically insignificant risk for death/prolonged hospitalization with the intervention compared to the comparison group (RR: 1.05, 95%CI: 0.84-1.32; P = 0.66; I² = 0%) (Figure 4).

Figure 2 Forest plot of changes in estimated glomerular filtration rate. IV: Inverse variance.

Figure 3 Forest plot of change in serum bicarbonate. IV: Inverse variance.

Figure 4 Forest plot of death/prolonged hospitalization outcome. M-H: Mantel-Haenszel.

Secondary outcomes

Compared to the comparison group, treatment with sodium bicarbonate did not significantly increase the risk of GI disorders, as reported by three trials. (RR: 1.64, 95%CI: 0.35-7.66; P = 0.53; I² = 76%) (Figure 5). Similarly, outcomes data from three trials did not show any significant difference in the risk of worsening edema (RR: 1.26, 95%CI: 0.94-1.68; P = 0.12; I² = 37%) between the two groups (Figure 6).

Figure 5 Forest plot of gastrointestinal disorders outcome. M-H: Mantel-Haenszel.

Figure 6 Forest plot of worsening edema outcome. M-H: Mantel-Haenszel.

Sensitivity analysis

A sensitivity analysis was performed for all outcomes except for death/prolonged hospitalization and worsening edema due to the low heterogeneity observed in data. For the change in eGFR outcome, removing Witham et al[20] and Goraya et al[23] significantly reduced heterogeneity (I² = 93% to I² = 34%, P = 0.48 to P = 0.21). However, the results were statistically nonsignificant (Supplementary Figure 3). Change in serum bicarbonate saw a great reduction in heterogeneity when Witham et al[20] and Dubey et al[33] were removed (I² = 95% to I² = 38%, P = 0.03 to P = 0.01) (Supplementary Figure 4), while removing Mohebbi et al[32] in GI disorders showed a smaller decrease in heterogeneity (I² = 76% to I² = 72%, P = 0.52 to P = 0.26) (Supplementary Figure 5). While the overall status of change in eGFR and GI disorders is still insignificant, change in serum bicarbonate remains significant even after the decrease in heterogeneity.

Meta-regression

We assessed age as a probable covariate having an impact on effect sizes for all outcomes and found that there was no statistically significant association between them except for change in eGFR (Coeff: -0.1337, P = 0.0417) (Supplementary Figure 6). The estimates of the effect of death/prolonged hospitalization were also assessed (Coeff: -0.0093, P = 0.8549) (Supplementary Figure 7), along with serum bicarbonate (Coeff: -0.0552, P = 0.2761) (Supplementary Figure 8), (Supplementary Table 2-Supplementary Table 4).

When evaluating the male gender as a possible covariate impacting the estimates of effect, it was statistically insignificantly associated with death/prolonged hospitalization (Coeff: -0.0308, P = 0.4425) (Supplementary Figure 9). However, it’s association with change in eGFR (Coeff: -0.1006, P = 0.0121) (Supplementary Figure 10) and serum HCO₃ (Coeff: -0.1501, P = 0.0063) (Supplementary Figure 11) was statistically significant (Supplementary Table 2-Supplementary Table 4).

When baseline serum bicarbonate was assessed as a covariate impacting the estimates of effect, it was statistically insignificantly associated with a change in eGFR (Coeff: 0.0379, P = 0.2837) (Supplementary Figure 12), death/prolonged hospitalization (Coeff: -0.0724, P = 0.5733) (Supplementary Figure 13) and serum HCO₃ (Coeff: -0.2468, P = 0.3313) (Supplementary Figure 14), (Supplementary Table 2-Supplementary Table 4).

DISCUSSION

This systematic review and meta-analysis studying the effects of sodium bicarbonate therapy on MA in CKD included 20 RCTs and 2 clinical trials (n = 2932 patients). Primary outcome analysis unveils that sodium bicarbonate was the reason for a substantial elevation in serum bicarbonate levels among CKD patients. Conversely, no significant correlation was observed between the administration of sodium bicarbonate and the occurrences of death or prolonged hospitalization in the study. Similarly, while assessing our secondary outcomes, treatment with sodium bicarbonate, compared to a comparison group, did not lead to a notable rise in the likelihood of GI disorders. Likewise, no significant change was noted in the risk of worsening edema between the two groups.

A recently conducted meta-analysis[37] shows that sodium bicarbonate has a decreasing effect on the decline of eGFR. The sodium bicarbonate group had fewer hospitalizations. Additionally, a second meta-analysis published in 2020[38] included fifteen trials, compared with placebo or no study medication, and yielded similar results. Another study[39] also showed findings consistent with our results. However, edema was found to be a significant side effect of sodium bicarbonate supplementation.

Meanwhile, our analysis showed no significant progression in the worsening of edema. Our observation is based on the latest and larger clinical trials incorporated, which did not identify edema as a significant adverse effect. This inclusion accounts for the observed distinction.

The current meta-analysis stands crucial in support of the 2024 Kidney Disease Improving Global Outcomes guidelines[1], which suggests prescribing bicarbonate supplementation to inhibit the progression of MA to levels of serum bicarbonate at or below 18 mmol/L and is also consistent with the results of previous meta-analysis. However, this study includes recent and more trials with 20 RCTs and addresses various outcomes. We reported the implications of sodium bicarbonate therapy on GI disorders, a relatively less explored domain.

The therapeutic use of sodium bicarbonate in patients with CKD is attributed to its several potential mechanisms of action in mitigating MA. In CKD, the functioning nephrons undergo hypertrophy to excessively generate NH₃ to neutralize the excess H⁺ ions through the formation of NH₄⁺[40]. This surplus production of NH₃/NH₄⁺ activates the complement system and leads to tubulointerstitial inflammation and fibrosis, which advances the decline in GFR[41]. However, maintaining normal serum HCO₃^- levels (24-26 mmol/L) through dietary alkali and sodium bicarbonate supplementation helps lower interstitial ammonium levels, thereby minimizing CKD progression by mitigating MA[41].

The recently published meta-analysis[37] included the studies of the age group < 65 years; meanwhile, this systematic review does not impose any such threshold for age as an inclusion or exclusion criterion. Despite the similarities in the outcomes of this meta-analysis and the aforementioned study, studies with an age limit of > 65 years as an inclusion criterion, when analyzed individually, exhibited insignificant treatment effects, with the risk of developing adverse effects on the rise. Therefore, it is suggested that more treatment options should be explored for elderly patients with CKD and MA.

To the best of our knowledge, this meta-analysis is the first study to report the GI adverse effects of sodium bicarbonate therapy in CKD patients with MA as a secondary outcome. The basis of the inclusion of GI effects as a secondary outcome stems from the fact that bicarbonate reacts with the hydrogen chloride found in the gastric lumen, leading to the formation of carbon dioxide gas, causing GI adverse effects like diarrhea, vomiting, nausea, and epigastric pain. Three of our included studies reported GI adverse effects; the study by Sorohan et al[15] analyzed these effects in sodium bicarbonate and sodium citrate therapy groups. While the comparative results were insignificant, the cumulated outcome leading to treatment withdrawal was statistically significant, with a higher relative ratio in the sodium bicarbonate group. Sodium bicarbonate therapy may be associated with edema; however, we analyzed insignificant associations with little to no requirement of initiating diuretics therapy. Further research may be warranted to explore these effects more comprehensively and to identify strategies for mitigating these adverse reactions in CKD patients.

Our study has some limitations. Firstly, we set no threshold for the maximum age of participants, which is important since it is observed that sodium bicarbonate is minimally effective on elderly patients, yet trials with patients over 65 years of age are included, which may affect the scope of our study. Second, adverse GI effects were outside the primary or secondary outcomes of many studies, raising concerns about the generalizability of the findings. Third, for some outcomes, high heterogeneity existed due to the inclusion of some studies, which, upon their removal, caused a dramatic reduction in the heterogeneity. This may be due to variations in therapeutic dosage, modalities, treatment duration, and baseline kidney function.

Additionally, most of the trials included in our meta-analysis predominantly featured male participants, with significantly fewer female patients. This gender imbalance could affect the generalizability of our findings, as the results may not fully represent female populations. Gender differences in physiology, disease progression, and treatment response could lead to varying outcomes between males and females, potentially limiting the applicability of our conclusions across both sexes. The treatment may be associated with high blood pressure; however, we did not explore this comorbidity as an outcome. The intake of vegetables and fruits as alkali therapeutic agents proves invaluable in treating MA in CKD patients. However, since this meta-analysis aimed to study the outcomes of supplementation only, we were unable to include the dietary interventions alongside the supplementation therapy.

Finally, analysis for the outcome of worsening edema failed to reach sufficient statistical power to detect significant differences between the treatment and the comparison groups. The trend showed an increased risk of worsening edema in the sodium bicarbonate group (RR: 1.26, 95%CI: 0.94-1.68), however the result was statistically non-significant. This may be attributed to the relatively small sample size (n = 471) available for this outcome. Hence, more patient data is needed to definitively assess sodium bicarbonate and whether it plays a significant role in the worsening of edema.

CONCLUSION

Based on the studies we evaluated in this meta-analysis; sodium bicarbonate therapy may halt worsening kidney function by correcting serum bicarbonate levels and treating MA. Although sodium bicarbonate does not significantly improve the eGFR, it may potentially prevent CKD progression while maintaining an overall favorable safety profile. This analysis also revealed that sodium bicarbonate did not have a statistically significant association with adverse effects like death/prolonged hospitalization, GI adverse events or worsening edema. Larger and more stringent randomized clinical trials are strongly suggested to establish the benefits and risks of sodium bicarbonate in CKD patients with MA.