Metabolic Acidosis and Subclinical Metabolic Acidosis in CKD

Abstract

Metabolic acidosis is not uncommon in CKD and is linked with bone demineralization, muscle catabolism, and higher risks of CKD progression and mortality. Clinical practice guidelines recommend maintaining serum total CO₂ at ≥22 mEq/L to help prevent these complications. Although a definitive trial testing whether correcting metabolic acidosis improves clinical outcomes has not been conducted, results from small, single-center studies support this notion. Furthermore, biologic plausibility supports the notion that a subset of patients with CKD have acid-mediated organ injury despite having a normal serum total CO₂ and might benefit from oral alkali before overt acidosis develops. Identifying these individuals with subclinical metabolic acidosis is challenging, but recent results suggest that urinary acid excretion measurements may be helpful. The dose of alkali to provide in this setting is unknown as well. The review discusses these topics and the prevalence and risk factors of metabolic acidosis, mechanisms of acid-mediated organ injury, results from interventional studies, and potential harms of alkali therapy in CKD.

Metabolic acidosis was one of the first recognized complications of kidney failure. Landmark studies identified the importance of reduced renal ammonia production in the pathogenesis of acidosis and effects of acidosis on bone demineralization and protein catabolism.^1–6 These consequences informed the clinical practice guideline recommendation to treat metabolic acidosis with alkali in CKD.^7,8 Results from single-center studies suggest that correcting acidosis may also preserve kidney function in CKD.^9–11 Although these potential benefits require confirmation in a definitive clinical trial, there is accumulating evidence that alkali might preserve kidney function in patients with CKD with normal serum total CO₂ (tCO₂) as well.^12–14 If true, this would lead to a significant paradigm shift in how alkali is utilized in CKD. However, it is unclear which patients with normal tCO₂ are most likely to benefit from alkali or the dose to prescribe in this setting. The prevalence and risk factors of metabolic acidosis, mechanisms of acid-mediated organ injury, results from interventional studies, and potential harms of alkali therapy are reviewed.

Prevalence and Risk Factors of Overt Metabolic Acidosis in CKD

Most nondialysis-requiring patients with CKD do not have metabolic acidosis, largely because of compensatory renal ammonia production and bone buffering.^4,6 When defined as a serum tCO₂ <22 mEq/L, the prevalence of metabolic acidosis is 15%.^15,16 However, its prevalence increases with worsening kidney function. For example, the prevalence of acidosis was 7% in stage 2, 13% in stage 3, and 37% in stage 4 CKD Chronic Renal Insufficiency Cohort Study (CRIC) participants,¹⁶ and similar trends were observed in the NephroTest Cohort.¹⁵

Reduced GFR is the most important risk factor for acidosis (Table 1).¹⁶ Related to GFR is the capacity to excrete acid. For example, African American Study of Kidney Disease and Hypertension (AASK) participants had 2.5-fold higher risk of incident metabolic acidosis if ammonium excretion was <15 mEq/d.¹⁷ The quantity of acid-producing (protein, grains, cheese) and alkali-producing (fruits and vegetables) foods consumed contributes as well.^10,18 Renin-angiotensin-aldosterone system (RAAS) inhibitors lower tCO₂ by attenuating aldosterone-mediated acid secretion,¹⁹ and hyperkalemia, irrespective of RAAS inhibition, lowers tCO₂ by reducing kidney ammoniagenesis.²⁰ On the other hand, diuretics increase serum tCO₂. Hence, a number of factors affect tCO₂; however, kidney function and acid excretion capacity are key determinants of metabolic acidosis in CKD.

Table 1.

Risk factors for overt metabolic acidosis in CKD after multivariate adjustment

Risk Factor	Comments
Lower GFR	2-fold higher risk for stage 3 and 7-fold higher risk for stage 4 (compared with stage 2 CKD)
Lower ammonium excretion	2.5-fold higher risk of incident acidosis if <15 mEq/d (compared with >25 mEq/d)
Hyperkalemia	2.4-fold higher risk if ≥5.0 mEq/L
Albuminuria	2-fold higher risk if ≥30 mg/gm
Smoking	43% higher risk
Anemia	40% higher risk
Higher serum albumin	35% higher risk for each 1 gm/dl increase
Diuretic use	30% lower risk
Use of ACE-i/ARB	24% higher risk

Based on references 16 and 17. ACE-i, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker.

Effects of Metabolic Acidosis on Bone and Muscle in CKD

The contribution of metabolic acidosis to bone demineralization and protein catabolism were cited by the National Kidney Foundation Kidney Disease Outcomes Quality Initiative guidelines as reasons to maintain serum tCO₂ ≥22 mEq/L using oral alkali as necessary in CKD (Figure 1A),^7,8 a recommendation subsequently supported by Kidney Disease: Improving Global Outcomes guidelines.²¹ Landmark studies demonstrated that renal insufficiency stimulates skeletal muscle proteolysis through acidification-dependent ubiquitination and that bone releases calcium carbonate to buffer acid leading to hypercalciuria and loss of bone mineral.^2–6 In vivo studies also showed that extracellular acidification increases activity of osteoclasts and inhibits activity of osteoblasts.²²

Figure 1.

Urinary acid excretion may identify patients with CKD and normal tCO₂ who have subclinical metabolic acidosis. (A) The current metabolic acidosis treatment paradigm is to withhold alkali until serum tCO₂ falls below 22 mEq/L, in which case only 15% of patients with CKD receive alkali therapy. (B) An alternative paradigm assumes that patients with CKD and tCO₂ below a threshold value (represented as X), have either overt acidosis or subclinical metabolic acidosis and are potential candidates for alkali therapy. (C) Another possible approach would be to treat patients with CKD with tCO₂<22 mEq/L and those with subclinical metabolic acidosis defined by a normal tCO₂ and urinary acid excretion below a threshold value (represented as Y).

Although it is biologically plausible that treatment of metabolic acidosis improves musculoskeletal health in CKD, evidence from interventional studies is limited to a few studies with relatively few participants. With respect to muscle, sodium bicarbonate treatment improved sit-stand time in a single-arm study of 20 patients with CKD with serum tCO₂ 20–24 mEq/L; however, there was no effect on hand grip strength.²³ Correction of acidosis with sodium bicarbonate (n=67) improved mid-arm muscle circumference, suggesting increased muscle mass, compared with usual care (n=67) over 2 years.⁹ In terms of bone health, treatment of metabolic acidosis (n=20) attenuated parathyroid hormone elevations over 3 months as compared with untreated patients with metabolic acidosis (n=20).²⁴ Hence, evidence that treatment of metabolic acidosis improves musculoskeletal health in CKD is sparse. In kidney transplant recipients, treatment with potassium citrate (n=19) normalized low tCO₂ and improved bone histology and turnover markers compared with potassium chloride over 1 year (n=11), and in ESRD, correction of acidosis ameliorated protein catabolism and improved bone histology.^25–29 Nevertheless, results from these non-CKD populations should be cautiously applied to those with CKD, and better evidence that treatment of metabolic acidosis improves musculoskeletal health is needed.

Effects of Nonvolatile Acids on the Kidney

Maintenance of normal systemic tCO₂ and pH is a principal kidney function. However, kidney compensatory mechanisms to maintain normal acid-base homeostasis in response to nonvolatile acids may cause further kidney injury in CKD. In the healthy state, kidneys respond to these acids by enhancing urinary acid excretion, mainly as ammonium.³⁰ With reduced kidney function, total urinary ammonium excretion declines, but ammonia production per nephron markedly increases to accommodate nephron loss and ongoing acid exposure.^1,17,31,32 This compensatory response to facilitate acid excretion leads to high local intrarenal concentrations of ammonia. This detrimentally activates the alternative pathway of complement through nucleophilic disruption of the internal thioester bond of the complement protein C3 by lone electron pairs on NH₃ leading to a cascade of events that ultimately promotes tubulointerstitial fibrosis.³³ This mechanism of kidney injury was observed in animals with subtotal nephrectomy and animals with normal renal mass and hypokalemia, and alkali treatment attenuated kidney injury in both models by suppressing kidney ammoniagenesis.^34,35 In humans, treatment of metabolic acidosis reduced urine, but not serum, levels of complement activation proteins, suggesting that intrarenal complement activation was ameliorated by lowering kidney ammoniagenesis.³⁶ However, a direct link between intrarenal ammonia and kidney fibrosis in humans has yet to be established.

Upregulation of systemic and kidney endothelin-1 (ET-1) levels is another compensatory response to nonvolatile acids that helps maintain normal tCO₂ and pH but promotes kidney injury. ET-1 facilitates acid excretion by stimulating proximal and distal Na⁺/H⁺ exchange, reducing distal bicarbonate secretion through nitric oxide, and stimulating adrenal aldosterone release to facilitate H⁺-ATPase activity.^37,38 However, ET-1 promotes kidney injury, proteinuria, inflammation, and kidney fibrosis,³⁹ and endothelin A receptor antagonism prevented GFR decline in subtotal nephrectomy animals on an acid-producing diet.⁴⁰

In addition to ET-1–mediated aldosterone upregulation, systemic and kidney angiotensin II levels are increased in subtotal nephrectomy animals with normal tCO₂ but interstitial acid accumulation, determined by microdialysis of the kidney cortex. Eliminating kidney interstitial acid accumulation with sodium bicarbonate reduced angiotensin II levels and preserved GFR better than angiotensin II receptor blockade, suggesting that amelioration of acid accumulation exerts renoprotective effects above and beyond angiotensin II inhibition.^41,42 Thus, compensatory upregulation of kidney ammonia production and ET-1 and RAAS activity to preserve systemic tCO₂ and pH are hypothesized to be detrimental in CKD. Interstitial acid accumulation may also directly contribute to kidney injury, along with intrarenal inflammation, insulin resistance, and oxidative stress.^43–47

Subclinical Metabolic Acidosis in CKD

These compensatory responses to nonvolatile acids suggest that acid-mediated kidney injury might be occurring even in the setting of normal tCO₂ in CKD. This state of nonvolatile acid–mediated organ injury despite having a normal tCO₂ has been referred to as eubicarbonatemic metabolic acidosis,⁴⁸ preclinical metabolic acidosis,⁴⁹ or subclinical metabolic acidosis.^50,51 In persons with preserved kidney function, subclinical metabolic acidosis might be observed in conditions such as chronic diarrhea or high dietary protein intake.⁴⁸ In these cases, enhanced urinary ammonium excretion might identify individuals with preserved kidney function who have subclinical metabolic acidosis despite having normal serum tCO₂.⁴⁸ A low-normal tCO₂ might be a sign of subclinical metabolic acidosis as well (Figure 1B). In AASK, United States veterans with CKD, and patients with CKD at Cleveland Clinic Foundation, those with low-normal tCO₂ had higher risk of mortality or GFR decline than those with high-normal tCO₂.^52–54 Conceivably, those with low-normal tCO₂ have higher risk of these outcomes because they require a greater compensatory response to maintain normal tCO₂ than those with high-normal tCO₂ and therefore have more acid-mediated organ injury.

Alternatively, urinary acid excretion measurements might aid in the diagnosis of subclinical metabolic acidosis in CKD. Higher urinary acid excretion in CKD might signal a high-acid diet. On the other hand, lower urinary acid excretion in CKD might indicate impaired tubular capacity to produce and excrete acid with tissue acid accumulation. Hence, a u-shaped relationship between urinary acid excretion and adverse outcomes might be expected. However, in the CRIC study, the NephroTest Cohort, and AASK, lower, but not higher, urinary acid excretion was associated with higher risk of ESRD or death.^17,32,55 In AASK, those with tCO₂≥22 mEq/L and lower ammonium excretion had (1) higher risk of ESRD or death than those with tCO₂≥22 mEq/L and higher ammonium excretion and (2) 2.5-fold higher risk of incident acidosis at 1 year, consistent with low urinary ammonium excretion signaling impaired acid excretion with acid retention rather than a low-acid or high-alkali diet or metabolic alkalosis.¹⁷ Thus, quantification of urinary acid excretion, particularly ammonium, may be a useful guide to identify those with subclinical metabolic acidosis but at high risk of overt metabolic acidosis and poor outcomes (Figure 1C).

Other urine measurements may identify patients with CKD with subclinical metabolic acidosis. Higher net endogenous acid production calculated from urine urea nitrogen and potassium measurements was associated with GFR decline in AASK.⁵⁶ Urine sulfate and citrate levels have also been proposed.⁴⁸ Urine sulfate is largely an indicator of animal protein intake⁴⁸; hence, higher urine sulfate levels might be expected to associate with worse outcomes in CKD. However, higher urine sulfate was associated with lower risk of mortality in kidney transplant recipients and lower risk of renal events in diabetic nephropathy.^57,58 Hence, its utility as an index of dietary acid–related subclinical metabolic acidosis in CKD is uncertain. Quantification of urinary citrate levels might also be informative, because citrate excretion decreases in response to metabolic acidosis⁵⁹; however, a relationship between urinary citrate and clinical outcomes in CKD has not been reported. These are possible ways of defining subclinical metabolic acidosis in CKD; however, the best way to identify these individuals is not yet certain.

Alkali May Preserve GFR in Patients with CKD and Subclinical Metabolic Acidosis

Results from alkali interventional studies in hypertensive patients with CKD with normal tCO₂ support the notion that subclinical metabolic acidosis causes kidney injury. In stage 2 hypertensive patients with CKD with mean tCO₂ approximately 26 mEq/L, those treated with sodium bicarbonate (n=40) had higher GFR than those in sodium chloride (n=40) and placebo arms (n=40) after 5 years, despite having similar baseline GFR.¹⁴ Similarly, treatment with either sodium bicarbonate or fruits and vegetables better preserved GFR than usual care in stage 3 hypertensive patients with CKD with tCO₂ 22–24 mEq/L over 3 years.¹² Finally, 30 days of fruits and vegetables or sodium bicarbonate lowered albuminuria and other kidney injury markers in hypertensive stage 2, but not stage 1, patients with CKD with tCO₂>24.5 mEq/L.¹³ The effects of oral alkali on bone in patients with CKD with normal tCO₂ have not been evaluated. However, in persons without CKD, oral alkali reduces urinary calcium excretion and plasma C-terminal crosslinking telopeptide of type 1 collagen (CTX) levels and increases bone mineral density.^60–62

Despite the promising results and biologic plausibility, the use of pharmacologic alkali in patients with CKD with normal tCO₂ is not recommended unless results from a well designed, multicenter trial are supportive. Results from several ongoing studies evaluating pleiotropic effects of sodium bicarbonate in patients with CKD with normal tCO₂ will provide additional insight.^63–65 One of the main challenges moving forward is that the dose of alkali to prescribe in patients with normal tCO₂ is uncertain. In metabolic acidosis, alkali is typically titrated until the serum tCO₂ is ≥22 mEq/L. In those with normal tCO₂, weight-based sodium bicarbonate doses of 0.3–0.5 mEq/kg per day have been administered.^12,14,63,64 Nevertheless, sodium bicarbonate seems to be safe in this dose range and raises the possibility that higher doses might be more efficacious and safe in CKD. In a 6-week dose-escalation study of 20 patients with stage 3b/4 CKD with a tCO₂ of 20–24 mEq/L, sodium bicarbonate doses up to 1 mEq/kg per day were well tolerated and improved quadricep strength.²³ Whether a dose in this range is safe and tolerable over a longer period is uncertain. This is the overarching goal of the CKD Pilot Studies Consortium Bicarbonate Administration to Stabilize eGFR Pilot Study. Briefly, 192 patients with CKD with a tCO₂ of 20–28 mEq/L will receive one of two doses of sodium bicarbonate (0.5 or 0.8 mEq/kg daily) or placebo over 28 weeks. The main objectives are to determine the safety, tolerability, and compliance of these doses as well as the effect on urine ammonium excretion as a pharmacodynamics assessment.⁶⁵ Another challenge moving forward is that it is not clear how best to identify patients with normal tCO₂ who would be most likely to benefit during a clinical trial of reasonable duration. Urinary acid excretion measurements may be helpful as alluded to, but none of the published or ongoing studies have used these to select participants. This may be important to consider in future interventional studies.

Alkali May Preserve Kidney Function in Patients with CKD and Overt Metabolic Acidosis

In an open-label study of 134 participants with CKD stage 4/5 and metabolic acidosis, de Brito-Ashurst et al.⁹ found that treatment of acidosis by titrating sodium bicarbonate toward a goal tCO₂ of ≥23 mEq/L slowed GFR decline and reduced the risk of ESRD. Phisitkul et al.¹¹ reported results from a nonrandomized, open-label comparison of hypertensive patients with CKD with metabolic acidosis who tolerated sodium citrate (n=30) to similar individuals who were intolerant of sodium citrate and sodium bicarbonate (n=29), and those who tolerated sodium citrate had less GFR decline. Nutritional therapy with fruits and vegetables (n=36) to lower the potential renal acid load by 50% led to similar reductions in kidney injury markers at 1 year as 1.0 mEq/kg daily of sodium bicarbonate (n=35) in stage 4 hypertensive patients with CKD with metabolic acidosis and serum potassium <4.6 mEq/L.¹⁰ Although the lower alkali dose with fruits and vegetables raised tCO₂ less than sodium bicarbonate, it lowered BP more.¹⁰ The accompanying dietary fiber with fruits and vegetables may also explain similar changes in these markers despite the lower alkali dose.^66,67 Importantly, serum potassium did not increase substantially in the fruits and vegetables arm.¹⁰

The results from these single-center studies support the possibility that nutritional and pharmacologic treatment of overt metabolic acidosis preserves renal function in CKD. Although clinical practice guidelines have recommended treatment of metabolic acidosis in CKD for nearly two decades now, a large-scale, well designed clinical trial testing pleiotropic effects of correcting metabolic acidosis on muscle, bone, and kidney health in CKD remains necessary.

Is Sodium-Based Alkali Safe?

Pharmacologic management of metabolic acidosis using potassium-based alkali may cause hyperkalemia. Hence, sodium-based alkali is commonly prescribed in CKD. Given the potential of sodium-related fluid retention, fluid weight gain, worsening BP, peripheral edema, and pulmonary congestion, an important consideration is whether sodium-based alkali is safe in CKD. Several lines of evidence suggest that these fluid-related possibilities are more likely to occur when sodium is accompanied by the chloride anion rather than bicarbonate or citrate.^68,69 Although the mechanisms for this are beyond the scope of this review, hyperchloremia induces tubuloglomerular feedback, renal vasoconstriction, and, consequently, reduced GFR.^70,71

The apparent safety of sodium-based alkali is thus far supported by evidence from the previously mentioned clinical studies. In the study by de Brito-Ashurst et al.,⁹ systolic BP was similar between the sodium bicarbonate and the usual care groups at baseline and during follow-up, and although a higher percentage of sodium bicarbonate–treated participants required an increase in antihypertensive therapy (61% versus 48%; P=0.17) and loop diuretic therapy (39% versus 30%; P=0.50), the differences were not statistically significant. This suggests that BP was manageable with sodium bicarbonate. More importantly, none in the sodium bicarbonate arm discontinued treatment or were hospitalized for congestive heart failure (CHF).⁹ In the study by Phisitkul et al.,⁴⁰ there was no suggestion that systolic BP was higher in the sodium citrate group than the control group. No information about hospitalizations was mentioned and no participants died during the 2-year follow-up. These results suggest that sodium-based alkali is relatively safe in terms of fluid-related adverse effects. Uncontrolled BP (>150/90 mm Hg), CHF, or morbid obesity were exclusion criteria in the study by de Brito-Ashurst et al.⁹ BP was controlled before exposure to sodium citrate and known cardiovascular disease, peripheral edema, liver failure, or nephrotic syndrome were exclusion criteria in the study by Phisitkul et al.⁴⁰ The safety of sodium-based alkali in patients with these comorbidities needs further study.

Potential Cardiovascular Harms of Raising Serum tCO₂

Although sodium-based alkali has not raised serious safety concerns in CKD so far, raising tCO₂ may have adverse cardiovascular effects. First, tCO₂>26 mEq/L was associated with higher risk of incident CHF in two CKD cohorts.^72,73 Whether raising tCO₂ toward these levels causes CHF is uncertain. Second, studies in animal models of CKD suggest that metabolic acidosis might prevent vascular calcification.^74,75 Hence, correcting acidosis might promote or worsen vascular calcification in CKD. To date, these possibilities remain theoretic, but a well designed, long-term clinical trial is needed to evaluate the overall risk and benefits of sodium-based alkali in CKD.

Dietary Therapy of Metabolic Acidosis in CKD

Dietary strategies incorporating fruits and vegetables should be considered in the management of metabolic acidosis in CKD and are reasonable for those with normal tCO₂. In addition to providing base, the accompanying potassium, fiber, and other nutrients may be beneficial in CKD, but serum potassium concentration should be monitored closely. A dose of fruits and vegetables that reduces potential renal acid load by 50% may reduce kidney injury, including albuminuria.^10,12,13 Dietary protein reduction raises serum tCO₂.¹⁸ However, this may lead to protein malnutrition and decreased muscle mass.⁷⁶ Reduction of other acid-producing foods such as grains and cheeses can be considered.⁵¹

Summary

Metabolic acidosis is commonly treated to prevent adverse effects on bone, muscle, and kidney health. Results from interventional studies support this notion but are not definitive. There is biologic plausibility that acid-mediated organ injury occurs in a subset of patients with CKD with normal tCO₂. These patients with subclinical metabolic acidosis may also benefit from alkali. Identifying these individuals is challenging and the dose to prescribe in this setting is unclear. Despite the apparent safety, there are potential risks of raising tCO₂ using sodium-based alkali. Base therapy with fruits and vegetables is also a promising strategy to improve outcomes in CKD. Well designed, multicenter clinical trials investigating the long-term benefits and harms of alkali therapy in CKD are necessary.

Disclosures

None.