Metabolic Acidosis

Updated: Sep 13, 2022
  • Author: Christie P Thomas, MBBS, FRCP, FASN, FAHA; Chief Editor: Vecihi Batuman, MD, FASN  more...
  • Print

Practice Essentials

Metabolic acidosis is a clinical disturbance characterized by an increase in plasma acidity. [1] Metabolic acidosis should be considered a sign of an underlying disease process. Identification of this underlying condition is essential to initiate appropriate therapy. (See Etiology, DDx, Workup, and Treatment.)

Understanding the regulation of acid-base balance requires appreciation of the fundamental definitions and principles underlying this complex physiologic process. (See Background.)

Go to Pediatric Metabolic Acidosis and Metabolic Acidosis in Emergency Medicine for complete information on those topics.



Basic definitions

An acid is a substance that can donate hydrogen ions (H+). A base is a substance that can accept H+ ions. The ion exchange occurs regardless of the substance's charge.

Strong acids are those that are completely ionized in body fluids, and weak acids are those that are incompletely ionized in body fluids. Hydrochloric acid (HCl) is considered a strong acid because it is present only in a completely ionized form in the body, whereas carbonic acid (H2 CO3) is a weak acid because it is ionized incompletely, and, at equilibrium, all three reactants are present in body fluids. See the reactions below.

H2 CO3 (acid)↔H+ + HCO3- (base)

HCl↔H+ + Cl-

The law of mass action states that the velocity of a reaction is proportional to the product of the reactant concentrations. On the basis of this law, the addition of H+ or bicarbonate (HCO3-) drives the reaction shown below to the left.

H2 CO3 (acid)↔H+ + HCO3- (base)

In body fluids, the concentration of hydrogen ions ([H+]) is maintained within very narrow limits, with the normal physiologic concentration being 40 nEq/L. The concentration of HCO3- (24 mEq/L) is 600,000 times that of [H+]. The tight regulation of [H+] at this low concentration is crucial for normal cellular activities because H+ at higher concentrations can bind strongly to negatively charged proteins, including enzymes, and impair their function.

Under normal conditions, acids and, to a lesser extent, bases are being added constantly to the extracellular fluid compartment, and for the body to maintain a physiologic [H+] of 40 nEq/L, the following three processes must take place:

  • Buffering by extracellular and intracellular buffers
  • Alveolar ventilation, which controls PaCO 2
  • Renal H + excretion, which controls plasma HCO 3 -


Buffers are weak acids or bases that are able to minimize changes in pH by taking up or releasing H+. Phosphate is an example of an effective buffer, as in the following reaction:

HPO42- + (H+)↔H2 PO4-

Upon addition of an H+ to extracellular fluids, the monohydrogen phosphate binds H+ to form dihydrogen phosphate, minimizing the change in pH. Similarly, when [H+] is decreased, the reaction is shifted to the left. Thus, buffers work as a first-line of defense to blunt the changes in pH that would otherwise result from the constant daily addition of acids and bases to body fluids.

The major extracellular buffering system is HCO3-/H2 CO3; its function is illustrated by the following reactions:

H2 O + CO2 ↔ H2CO3 ↔ H+ + HCO3-

One of the major factors that makes this system very effective is the ability to control PaCO2 by changes in ventilation. As can be noted from this reaction, increased carbon dioxide (CO2) concentration drives the reaction to the right, whereas a decrease in CO2 concentration drives it to the left. Put simply, adding an acid load to the body fluids results in consumption of HCO3- by the added H+, and the formation of carbonic acid; the carbonic acid, in turn, forms water and CO2. CO2 concentration is maintained within a narrow range via the respiratory drive, which eliminates accumulating CO2. The kidneys regenerate the HCO3- consumed during this reaction.

This reaction continues to move to the left as long as CO2 is constantly eliminated or until HCO3- is significantly depleted, making less HCO3- available to bind H+. That HCO3- and PaCO2 can be managed independently (by kidneys and lungs, respectively) makes this a very effective buffering system. At equilibrium, the relationship between the 3 reactants in the reaction is expressed by the Henderson-Hasselbalch equation, which relates the concentration of dissolved CO2 (ie, H2CO3) to the partial pressure of CO2 (0.03 × PaCO2) in the following way:

pH = 6.10 + log ([HCO3-]/0.03 × PaCO2)

Alternatively, [H+] = 24 × PaCO2/[HCO3-]

Note that changes in pH or [H+] are a result of relative changes in the ratio of PaCO2 to [HCO3-] rather than to absolute change in either one. In other words, if both PaCO2 and [HCO3-] change in the same direction, the ratio stays the same and the pH or [H+] remains relatively stable. To diminish the alteration in pH that occurs when either HCO3- or PaCO2 changes, the body, within certain limits, changes the other variable in the same direction.

In chronic metabolic acidosis, intracellular buffers (eg, hemoglobin, bone) may be more important than HCO3- when the extracellular HCO3- level is low.

Renal acid handling

Acids are added daily to the body fluids. These include volatile (eg, carbonic) and nonvolatile (eg, sulfuric, phosphoric) acids. The metabolism of dietary carbohydrates and fat produces approximately 15,000 mmol of CO2 per day, which is excreted by the lungs. Failure to do so results in respiratory acidosis.

The metabolism of proteins (ie, sulfur-containing amino acids) and dietary phosphate results in the formation of nonvolatile acids, H2 SO4 and H3 PO4. These acids first are buffered by the HCO3-/H2 CO3 system as follows:

H2SO4 + 2NaHCO3 ↔ Na2SO4 + 2H2CO3 ↔ 2H2O + CO2

The net result is buffering of a strong acid (H2 SO4) by 2 molecules of HCO3- and production of a weak acid (H2 CO3), which minimizes the change in pH. The lungs excrete the CO2 produced, and the kidneys, to prevent progressive HCO3- loss and metabolic acidosis, replace the consumed HCO3- (principally by H+ secretion in the collecting duct). Some amino acids (ie, glutamate, aspartate) result in the formation of citrate and lactate, which, in turn, will be converted to HCO3-. The net result, in a typical American diet, is an acid load in the range of 50-100 mEq of H+ per day.

To maintain normal pH, the kidneys must perform two physiologic functions. The first is to reabsorb all the filtered HCO3- (any loss of HCO3- is equal to the addition of an equimolar amount of H+), a function principally of the proximal tubule. The second is to excrete the daily H+ load (loss of H+ is equal to addition of an equimolar amount of HCO3-), a function of the collecting duct.

Bicarbonate reabsorption

With a serum HCO3- concentration of 24 mEq/L, the daily glomerular ultrafiltrate of 180 L, in a healthy subject, contains 4300 mEq of HCO3-, all of which has to be reabsorbed. Approximately 90% of the filtered HCO3- is reabsorbed in the proximal tubule, and the remainder is reabsorbed in the thick ascending limb and the medullary collecting duct.

The 3Na+ -2K+/ATPase (sodium-potassium/adenosine triphosphatase) provides the energy for this process, which maintains a low intracellular Na+ concentration and a relative negative intracellular potential. The low Na+ concentration indirectly provides energy for the apical Na+/H+ exchanger, NHE3 (gene symbol SLC9A3), which transports H+ into the tubular lumen. H+ in the tubular lumen combines with filtered HCO3- in the following reaction:

HCO3- + H+ ↔ H2CO3 ↔H2O + CO2

Carbonic anhydrase (CA IV isoform) present in the brush border of the first 2 segments of the proximal tubule accelerates the dissociation of H2 CO3 into H2O + CO2, which shifts the reaction shown above to the right and keeps the luminal concentration of H+ low. CO2 diffuses into the proximal tubular cell perhaps via the aquaporin-1 water channel, where carbonic anhydrase (CA II isoform) combines CO2 and water to form HCO3- and H+. The HCO3- formed intracellularly returns to the pericellular space and then to the circulation via the basolateral Na+/3HCO3- cotransporter, NBCe1-A (gene symbol SLC4A4).

In essence, the filtered HCO3- is converted to CO2 in the lumen, which diffuses into the proximal tubular cell and is then converted back to HCO3- to be returned to the systemic circulation, thus reclaiming the filtered HCO3-.

Acid excretion

Excretion of the daily acid load (50-100 mEq of H+) occurs principally through H+ secretion by the apical H+/ATPase in α-intercalated cells of the collecting duct.

HCO3- formed intracellularly is returned to the systemic circulation via the basolateral Cl-/HCO3- exchanger, AE1 (gene symbol SLC4A1), and H+ enters the tubular lumen via 1 of 2 apical proton pumps, H+/ATPase or H+ -K+/ATPase. The secretion of H+ in these segments is influenced by Na+ reabsorption in the adjacent principal cells of the collecting duct. The reabsorbed Na+ creates a relative lumen negativity, which decreases the amount of secreted H+ that back-diffuses from the lumen.

Hydrogen ions secreted by the kidneys can be excreted as free ions but, at the lowest achievable urine pH of 5.0 (equal to free H+ concentration of 10 µEq/L), would require excretion of 5000-10,000 L of urine a day. Urine pH cannot be lowered much below 5.0 because the gradient against which H+/ATPase has to pump protons (intracellular pH 7.5 to luminal pH 5) becomes too steep. A maximally acidified urine, even with a volume of 3 L, would thus contain a mere 30 µEq of free H+. Instead, more than 99.9% of the H+ load is excreted buffered by the weak bases NH3 or phosphate.

Titratable acidity

The amount of secreted H+ that is buffered by filtered weak acids is called titratable acidity. Phosphate as HPO42- is the main buffer in this system, but other urine buffers include uric acid and creatinine.

H2PO4 ↔ H+ + HPO42-

The amount of phosphate filtered is limited and relatively fixed, and only a fraction of the secreted H+ can be buffered by HPO42-.


A more important urine-buffering system for secreted H+ than phosphate, ammonia (NH3) buffering occurs via the following reaction:

NH3 + H+ ↔ NH4+

Ammonia is produced in the proximal tubule from the amino acid glutamine, and this reaction is enhanced by an acid load and by hypokalemia. Ammonia is converted to ammonium (NH4+) by intracellular H+ and is secreted into the proximal tubular lumen by the apical Na+/H+ (NH4+) antiporter.

The apical Na+/K+ (NH4+)/2Cl- cotransporter in the thick ascending limb of the loop of Henle then transports NH4+ into the medullary interstitium, where it dissociates back into NH3 and H+. The NH3 enters the collecting duct epithelial cells via the basolateral ammonia transporters, RhBG and RhCG, and then is transported into the lumen of the collecting duct via apical RhCG, where it is available to buffer H+ ions and becomes NH4+. NH4+ is trapped in the lumen and excreted as the Cl salt, and every H+ ion buffered is an HCO3- gained to the systemic circulation.

The increased secretion of H+ in the collecting duct shifts the equation to the right and decreases the NH3 concentration, facilitating continued diffusion of NH3 from the interstitium down its concentration gradient into the collecting duct lumen, allowing more H+ to be buffered. The kidneys can adjust the amount of NH3 synthesized to meet demand, making this a powerful system to buffer secreted H+ in the urine. [2]

Acidosis and alkalosis

In healthy people, blood pH is maintained at 7.39-7.41, and because pH is the negative logarithm of [H+] (pH = - log10 [H+]), an increase in pH indicates a decrease in [H+] and vice versa. An increase in [H+] and a fall in pH are termed acidemia, and a decrease in [H+] and an increase in pH are termed alkalemia. The underlying disorders that lead to acidemia and alkalemia are acidosis and alkalosis, respectively. Metabolic acidosis is a primary decrease in serum HCO3 - concentration and, in its pure form, manifests as acidemia (pH < 7.40).

Rarely, metabolic acidosis can be part of a mixed or complex acid-base disturbance in which two or more separate metabolic or respiratory derangements occur together. In these instances, pH may not be reduced or the HCO3- concentration may not be low.

As a compensatory mechanism, metabolic acidosis leads to alveolar hyperventilation with a fall in PaCO2. Normally, PaCO2 falls by 1-1.3 mm Hg for every 1-mEq/L fall in serum HCO3- concentration, a compensatory response that can occur fairly quickly. If the change in PaCO2 is not within this range, then a mixed acid-base disturbance is present. For example, if the decrease in PaCO2 is less than the expected change, then a primary respiratory acidosis also is present.

The only definitive way to diagnose metabolic acidosis is by simultaneous measurement of serum electrolytes and arterial blood gases (ABGs) that shows both pH and PaCO2 to be low; calculated HCO3- also is low. (See Metabolic Alkalosis for a discussion of the difference between measured and calculated HCO3- concentrations.)

A normal serum HCO3- level does not rule out the presence of metabolic acidosis, because a drop in HCO3- from a high baseline (ie, preexisting metabolic alkalosis) can result in a serum HCO3- level that is within the reference range, concealing the metabolic acidosis.

In general, patients with kidney failure tend to have a serum HCO3- level greater than 12 mEq/L, and buffering by the skeleton prevents further decline in serum HCO3-. Note that patients with hypobicarbonatemia from kidney failure cannot compensate for additional HCO3- loss from an extrarenal source (eg, diarrhea), and severe metabolic acidosis can develop rapidly.

In persons with chronic uremic acidosis, bone salts contribute to buffering, and the serum HCO3- level usually remains greater than 12 mEq/L. This bone buffering can lead to significant loss of bone calcium, with resulting osteopenia and osteomalacia.

Anion gap

Plasma, like any other body fluid compartment, is neutral; total anions match total cations. The major plasma cation is Na+, and major plasma anions are Cl- and HCO3-. Extracellular anions present in lower concentrations include phosphate, sulfate, and some organic anions, while other cations present include K+, Mg2+, and Ca2+. The anion gap (AG) is the difference between the concentration of the major measured cation Na+ and the major measured anions Cl- and HCO3-.

An increase in the AG can result from either a decrease in unmeasured cations (eg, hypokalemia, hypocalcemia, hypomagnesemia) or an increase in unmeasured anions (eg, hyperphosphatemia, high albumin levels). In certain forms of metabolic acidosis, other anions accumulate; by recognizing the increasing AG, the clinician can formulate a differential diagnosis for the cause of that acidosis. [3]

The reaction below indicates that the addition of an acid (HA, where H+ is combined with an unmeasured anion A-) results in the consumption of HCO3- with an addition of anions that will account for the increase in the AG. Metabolic acidosis is classified on the basis of AG into normal- (also called non-AG or hyperchloremic metabolic acidosis [4] ) and high-AG metabolic acidosis.

HA + NaHCO3 ↔ NaA + H2CO3 ↔ CO2 + H2O

Urinary AG

Calculating the urine AG is helpful in evaluating some cases of non-AG metabolic acidosis. The major measured urinary cations are Na+ and K+, and the major measured urinary anion is Cl-.

Urine AG = ([Na+] + [K+]) - [Cl-]

The major unmeasured urinary anions and cations are HCO3- and NH4+, respectively. HCO3- excretion in healthy subjects is usually negligible, and average daily excretion of NH4+ is approximately 40 mEq/L, which results in a positive or near-zero gap. In the face of metabolic acidosis, the kidneys increase the amount of NH3 synthesized to buffer the excess H+ and NH4 Cl excretion increases. The increased unmeasured NH4+ thus increases the measured anion Cl- in the urine, and the net effect is a negative AG, representing a normal response to systemic acidification. Thus, the finding of a positive urine AG in the face of non-AG metabolic acidosis points toward a renal acidification defect (eg, renal tubular acidosis [RTA]).

Caveats to urinary anion gap testing

The presence of ketonuria makes this test unreliable because the negatively charged ketones are unmeasured and urine AG will be positive or zero despite the fact that renal acidification and NH4+ levels are increased. Moreover, severe volume depletion from extrarenal NaHCO3 loss causes avid proximal Na+ reabsorption, with little Na+ reaching the lumen of the collecting duct to be reabsorbed in exchange for H+. Limiting H+ excretion reduces NH4+ excretion and may make the urine AG become positive.

Potassium and renal acid secretion

Renal acid secretion is influenced by serum K+ and may result from the transcellular shift of K+ when intracellular K+ is exchanged for extracellular H+ or vice versa. In hypokalemia, an intracellular acidosis can develop; in hyperkalemia, an intracellular alkalosis can develop. HCO3- reabsorption is increased secondary to relative intracellular acidosis. The increase in intracellular H+ concentration promotes the activity of the apical Na+/H+ exchanger.

Renal production of NH3 is increased in hypokalemia, resulting in an increase in renal acid excretion. The increase in NH3 production by the kidneys may be significant enough to precipitate hepatic encephalopathy in patients who have advanced liver disease. Correcting the hypokalemia can reverse this process.

Patients with hypokalemia may have relatively alkaline urine because hypokalemia increases renal ammoniagenesis. Excess NH3 then binds more H+ in the lumen of the distal nephron and urine pH increases, which may suggest RTA as an etiology for non-AG acidosis. However, these conditions can be distinguished by measuring urine AG, which will be negative in patients who have normal NH4+ excretion and positive in patients with RTA. The most common cause for hypokalemia and metabolic acidosis is GI loss (eg, diarrhea, laxative use). Other less common etiologies include renal loss of potassium secondary to RTA or salt-wasting nephropathy. The urine pH, the urine AG, and the urinary K+ concentration can distinguish these conditions.

Hyperkalemia has an effect on acid-base regulation opposite to that observed in hypokalemia. Hyperkalemia impairs NH4+ excretion through reduction of NH3 synthesis in the proximal tubule and reduction of NH4+ reabsorption in the thick ascending limb, resulting in reduced medullary interstitial NH3 concentration. This leads to a decrease in net renal acid secretion and is a classic feature of primary or secondary hypoaldosteronism. Consistent with the central role of hyperkalemia in the generation of the acidosis, lowering the serum K+ concentration can correct the associated metabolic acidosis.



Causes and diagnostic considerations

Metabolic acidosis is typically classified according to whether the anion gap (AG) is normal (ie, non-AG) or high. Non-AG metabolic acidosis is also characterized by hyperchloremia and is sometimes referred to as hyperchloremic acidosis. Calculation of the AG is thus helpful in the differential diagnosis of metabolic acidosis. [3, 5]

Normal anion gap metabolic acidosis 

Hyperchloremic or non-AG metabolic acidosis occurs principally when HCO3- is lost from either the GI tract or the kidneys or because of a renal acidification defect. Some of the mechanisms that result in a non-AG metabolic acidosis are the following:

  • Addition of HCl to body fluids: H + buffers HCO 3 - and the added Cl - results in a normal AG.
  • Loss of HCO 3 - from the kidneys or the GI tract: The kidneys reabsorb sodium chloride to maintain volume.
  • Rapid volume expansion with normal saline: This results in an increase in the chloride load that exceeds the renal capacity to generate equal amounts of HCO 3 -.

Causes of non-AG metabolic acidosis can be remembered with the mnemonic ACCRUED:

  • Acid load
  • Chronic kidney disease (CKD)
  • Carbonic anhydrase inhibitors
  • Renal tubular acidosis (RTA)
  • Ureteroenterostomy
  • Expansion/extra-alimentation
  • Diarrhea

Conditions that may cause a non-AG metabolic acidosis are as follows:

  • GI loss of HCO 3 - - Diarrhea
  • Enterocutaneous fistula (eg, pancreatic) - Enteric diversion of urine (eg, ileal loop bladder), pancreas transplantation with bladder drainage
  • Renal loss of HCO 3 - - Proximal RTA (type 2), carbonic anhydrase inhibitor therapy (including topiramate [6] )
  • Failure of renal H + secretion - Distal RTA (type 1), hyperkalemic RTA (type 4), kidney failure
  • Acid infusion - Ammonium chloride, hyperalimentation
  • Other - Rapid volume expansion with normal saline

Causes of non-AG metabolic acidosis are discussed in more detail below.

High anion gap metabolic acidosis

High AG metabolic acidosis warrants consideration of the following:

  • Lactic acidosis – L-lactate, D-lactate
  • Ketoacidosis - Beta-hydroxybutyrate, acetoacetate
  • CKD - High-AG chronic metabolic acidosis is seen in later stages of CKD, as a result of loss of nephron mass and impaired glomerular elimination of organic acid residues [7]
  • Ingestions - Salicylate, methanol, formaldehyde (formate), ethylene glycol (glycolate, oxalate), paraldehyde (organic anions), phenformin/metformin [8]
  • Infusions - Propylene glycol (D-lactate, L-lactate)
  • Pyroglutamic acid (5-oxoprolinemia) - Typically seen in malnourished, chronically ill women with a history of long-term acetaminophen ingestion [9]
  • Massive rhabdomyolysis (release of H + and organic anions from damaged muscle)

Several mnemonics are used to help recall of the differential diagnosis of high anion gap acidosis. Three are as follows:

  • MUDPILES: Methanol; Uremia; Diabetic ketoacidosis (DKA); Paraldehyde, phenformin; Iron, isoniazid; Lactic (ie, carbon monoxide [CO], cyanide); Ethylene glycol; Salicylates

  • DR. MAPLES: DKA; Renal; Methanol; Alcoholic ketoacidosis; Paraldehyde, phenformin; Lactic (ie, CO, HCN); Ethylene glycol; Salicylates

  • SLUMPED: Salicylate, Lactate, Uremia, Methanol, Paraldehyde, Ethylene glycol, Diabetes)

A more current mnemonic is GOLD MARK, which incorporates newly recognized forms of metabolic acidosis and eliminates P for paraldehyde as that is now rarely seen. [10]

  • GOLD MARK: Glycols (ethylene and propylene), Oxoproline, Lactate, D-lactate, Methanol, Aspirin, Renal failure, Ketoacidosis

Plasma osmolality and the osmolar gap can be helpful in determining the cause of high AG acidosis. Plasma osmolality can be calculated using the following equation:

Posm = [2 × Na+]+[glucose in mg/dL]/18+[BUN in mg/dL]/2.8

Posm can also be measured in the laboratory, and because other solutes normally contribute minimally to serum osmolality, the difference between the measured and the calculated value (osmolar gap) is no more than 10-15 mOsm/kg. In certain situations, unmeasured osmotically active solutes in the plasma can raise the osmolar gap (eg, mannitol, radiocontrast agents).

The osmolar gap can also be a clue to the nature of the anion in high-AG acidosis because some osmotically active toxins also cause a high-AG acidosis. Methanol, ethylene glycol, and acetone are classic poisons that increase the osmolar gap and AG; measuring the osmolar gap can help narrow the differential diagnosis of high-AG acidosis.

Causes of AG metabolic acidosis are discussed in more detail below.

Specific causes of hyperchloremic or non-AG metabolic acidosis

Loss of HCO3- via the GI tract

The secretions of the GI tract, with the exception of the stomach, are relatively alkaline, with high concentrations of base (50-70 mEq/L). Significant loss of lower GI secretions results in metabolic acidosis, especially when the kidneys are unable to adapt to the loss by increasing net renal acid excretion.

Such losses can occur in diarrheal states, fistula with drainage from the pancreas or the lower GI tract, and sometimes vomiting if it occurs as a result of intestinal obstruction. When pancreatic transplantation is performed, the pancreatic duct is sometimes diverted into the recipient bladder, from where exocrine pancreatic secretions are lost in the final urine. Significant loss also occurs in patients who abuse laxatives, which should be suspected when the etiology for non-AG metabolic acidosis is not clear.

Urine pH will be less than 5.3, with a negative urine AG reflecting normal urine acidification and increased NH4+ excretion. However, if distal Na+ delivery is limited because of volume depletion, theurine pH cannot be lowered maximally.

Replacing the lost HCO3- on a daily basis can treat this form of metabolic acidosis.

Distal RTA (type 1) (see the Table below)

The defect in this type of RTA is a decrease in net H+ secreted by the A-type intercalated cells of the collecting duct. As mentioned previously, H+ is secreted by the apical H+–ATPase and, to a lesser extent, by the apical K+/H+–ATPase. The K+/H+–ATPase seems to be more important in K+ regulation than in H+ secretion. The secreted H+ is then excreted as free ions (reflected by urine pH value) or titrated by urinary buffers, phosphate, and NH3. A decrease in the amount of H+ secreted results in a reduction in its urinary concentration (ie, increase in urine pH) and a reduction in total H+ buffered by urinary phosphate or NH3.

Type 1 RTA should be suspected in any patient with non-AG metabolic acidosis and a urine pH greater than 5.0. Patients have a reduction in serum HCO3- to various degrees, in some cases to less than 10 mEq/L. They are able to reabsorb HCO3- normally, and their fractional excretion (FE) of HCO3- is less than 3%. The disorder has been classified into 4 types—secretory, rate dependent, gradient, and voltage dependent—based on the nature of the defect.

Several different mechanisms are implicated in the development of distal RTA. These include a defect in 1 of the 2 proton pumps, H+–ATPase or K+/H+–ATPase, that can be acquired or congenital. This may lead to loss of function (ie, secretory defect) or a reduction in the rate of H+ secretion (ie, rate-dependent defect).

Another mechanism is a defect in the basolateral Cl-/HCO3- exchanger, AE1, or the intracellular carbonic anhydrase that can be acquired or congenital. This also causes a secretory defect.

Back-diffusion of the H+ from the lumen via the paracellular or transcellular space is another mechanism; this occurs if the integrity of the tight junctions is lost or permeability of the apical membrane is increased (ie, permeability or gradient defect). With a urine pH of 5.0 and an interstitial fluid pH of 7.4, the concentration gradient facilitating back-diffusion of free H+, under conditions of increased permeability of the collecting duct epithelia, is approximately 250-fold.

A defect in Na+ reabsorption in the collecting duct would decrease the electrical gradient favoring the secretion of H+ into the tubular lumen (ie, voltage-dependent defect). This can occur, for instance, in severe volume depletion with decreased luminal Na+ delivery to this site.

The serum potassium level typically is low in patients with distal RTA because defects in H+ secretion or back-diffusion of H+ tend to increase urinary K+ wasting. Potassium wasting occurs from one or more of the following factors:

  • Decreased net H+ secretion results in more Na+ reabsorption in exchange for K+ secretion.

  • The drop in serum HCO3- and, therefore, filtered HCO3-, reduces the amount of Na+ reabsorbed by the Na+/H+ exchanger in the proximal tubule, leading to mild volume depletion. The associated activation of the renin-angiotensin-aldosterone system increases K+ secretion in the collecting duct.

  • A possible defect in K+/H+–ATPase results in decreased H+ secretion and decreased K+ reabsorption.

The serum K+ level can be high if the distal RTA is secondary to decreased luminal Na+ in the distal nephron. Na+ reabsorption in the principal cells of the collecting duct serves as the driving force for K+ secretion. In this case, the patient has hyperkalemia and acidosis; the disorder is also called voltage-dependent or hyperkalemic type 1 acidosis.

Urine AG is positive and urine pH is high secondary to the renal acid secretion defect. Urine pH also can be high in patients with type 2 RTA if their serum HCO3- level is higher than the renal threshold for reabsorption, typically when a patient with type 2 RTA is on HCO3- replacement therapy. Administration of an HCO3- load leads to a marked increase in urine pH in those who have type 2 RTA, while those with type 1 RTA have a constant urine pH unless their acidosis is overcorrected.

Patients with type 1 RTA may develop nephrocalcinosis and nephrolithiasis. This is thought to occur for the following reasons:

  • Patients have a constant release of calcium phosphate from bones to buffer the extracellular H +.
  • Patients have decreased reabsorption of calcium and phosphate, leading to hypercalciuria and hyperphosphaturia.
  • Patients have relatively alkaline urine, which promotes calcium phosphate precipitation.
  • Metabolic acidosis and hypokalemia lead to hypocitraturia, a risk factor for stones. Citrate in the urine complexes calcium and inhibits stone formation.

The causes of distal RTA are shown as follows. Type 1 RTA occurs sporadically, although genetic forms have been reported.

  • Primary - Genetic or sporadic
  • Drug-related - Amphotericin B, lithium, analgesics, ifosfamide, topiramate, toluene
  • Autoimmune disease - Systemic lupus erythematosus, chronic active hepatitis, Sjögren syndrome, rheumatoid arthritis, primary biliary cirrhosis
  • Related to other systemic disease - Sickle cell disease, hyperparathyroidism, light chain disease, cryoglobulinemia, Wilson disease, Fabry disease
  • Tubulointerstitial disease - Obstructive uropathy, transplant rejection, medullary cystic kidney disease, hypercalciuria

The genetic forms of type 1 RTA are the following:

  • Autosomal dominant: Heterozygous mutations in the basolateral Cl-/HCO3- exchanger, AE1 (gene symbol SLC4A1), cause a dominant form of distal RTA with nephrocalcinosis and osteomalacia. Some patients with this disorder can be relatively asymptomatic and present in later years, while others present with severe disease in childhood. The disorder is allelic with one form of hereditary spherocytosis, but each disease is caused by distinct mutations in the same gene.

  • Autosomal recessive: This form of the disease may occur with or without sensorineural deafness. The type that occurs with deafness involves homozygous mutations in the B subunit of H+ –ATPase (gene symbol ATP6B1) in the A-type intercalated cells. The type that occurs without deafness involves homozygous mutations in the accessory N1 subunit of H+ –ATPase (gene symbol ATP6N1B). Homozygous or compound heterozygous mutations in AE1 also cause a recessive form of distal RTA that manifests in childhood with growth retardation and nephrocalcinosis that may lead to renal insufficiency. Heterozygous carriers have autosomal dominant ovalocytosis but normal renal acidification.

Proximal (type 2) RTA

The hallmark of type 2 RTA is impairment in proximal tubular HCO3- reabsorption. In the euvolemic state and in the absence of elevated levels of serum HCO3-, all filtered HCO3- is reabsorbed, 90% of which is in the proximal tubule. Normally, HCO3- excretion occurs only when serum HCO3- exceeds 24-28 mEq/L. Patients with type 2 RTA, however, have a lower threshold for excretion of HCO3-, leading to a loss of filtered HCO3- until the serum HCO3- concentration reaches the lower threshold. At this point, bicarbonaturia ceases and the urine appears appropriately acidified. Serum HCO3- typically does not fall below 15 mEq/L because of the ability of the collecting duct to reabsorb some HCO3-.

Type 2 RTA can be found as a solitary proximal tubular defect, in which reabsorption of HCO3- is the only abnormality (rare) such as with homozygous mutations in SLC4A4. More commonly, it is part of a more generalized defect of the proximal tubule characterized by glucosuria, aminoaciduria, and phosphaturia, also called Fanconi syndrome.

Dent disease or X-linked hypercalciuric nephrolithiasis is one example of a generalized proximal tubular disorder characterized by an acidification defect, hypophosphatemia, and hypercalciuria and arises from mutations in the renal chloride channel gene (CLCN5). Homozygous mutations in SCL34A1 also cause a genetic form of Fanconi syndrome.

The proximal tubule is the site where bulk reabsorption of ultrafiltrate occurs, driven by the basolateral Na+/K+ –ATPase. Any disorder that leads to decreased ATP production or a disorder involving Na+ -K+ –ATPase can result in Fanconi syndrome. In principle, loss of function of the apical Na+/H+ antiporter or the basolateral Na+/3HCO3- cotransporter or the intracellular carbonic anhydrase results in selective reduction in HCO3- reabsorption.

Patients with type 2 RTA typically have hypokalemia and increased urinary K+ wasting. This is thought, in part, to be due to an increased rate of urine flow to the distal nephron caused by the reduced proximal HCO3- reabsorption and, in part, to be due to activation of the renin-angiotensin-aldosterone axis with increased collecting duct Na+ reabsorption from the mild hypovolemia induced by bicarbonaturia. Administration of alkali in those patients leads to more HCO3- wasting and can worsen hypokalemia unless K+ is replaced simultaneously.

The diagnosis of type 2 RTA should be suspected in patients who have a normal-AG metabolic acidosis with a serum HCO3- level usually greater than 15 mEq/L and acidic urine (pH < 5.0). Those patients have an FEHCO3- less than 3% when their serum HCO3- is low. However, raising serum HCO3- above their lower threshold and closer to normal levels results in significant HCO3- wasting and an FEHCO3 exceeding 15%.

FEHCO3- = (urine [HCO3-] X plasma [creatinine] / plasma [HCO3-]) X urine [creatinine] X 100

Some patients with type 2 RTA tend to have osteomalacia, a condition that can be observed in any chronic acidemic state, although it is more common in persons with type 2 RTA. The traditional explanation is that the proximal tubular conversion of 25(OH)-cholecalciferol to the active 1,25(OH)2-cholecalciferol is impaired. Patients with more generalized defects in proximal tubular function (as in Fanconi syndrome) may have phosphaturia and hypophosphatemia, which also predispose to osteomalacia.

The following are causes of proximal RTA:

  • Primary - Genetic or sporadic
  • Inherited systemic disease - Wilson disease, glycogen storage disease, tyrosinemia, Lowe syndrome, cystinosis, fructose intolerance
  • Related to other systemic disease - Multiple myeloma, amyloidosis, hyperparathyroidism, Sjögren syndrome
  • Drug- and toxin-related - Carbonic anhydrase inhibitors, ifosfamide, gentamicin, valproic acid, lead, mercury, streptozotocin

Isolated proximal RTA occurs sporadically, although an inherited form has recently been described. Homozygous mutations in the apical Na+/3HCO3- cotransporter have been found in 2 kindred with proximal RTA, band keratopathy, glaucoma, and cataracts. A form of autosomal recessive osteopetrosis with mental retardation is associated with a mixed RTA with features of both proximal and distal disease (called type 3). The mixed defect is related to the deficiency of carbonic anhydrase (CA II isoform) normally found in the cytosol of the proximal tubular cells and the intercalated cells of the collecting duct. The most common cause of acquired proximal RTA in adults follows the use of carbonic anhydrase inhibitors.

Type 4 RTA

This is the most common form of RTA in adults and results from aldosterone deficiency or resistance. The collecting duct is a major site of aldosterone action; there it stimulates Na+ reabsorption and K+ secretion in the principal cells and stimulates H+ secretion in the A-type intercalated cells. Hypoaldosteronism, therefore, is associated with decreased collecting duct Na+ reabsorption, hyperkalemia, and metabolic acidosis.

Hyperkalemia also reduces proximal tubular NH4+ production and decreases NH4+ absorption by the thick ascending limb, leading to a reduction in medullary interstitial NH3 concentration. This diminishes the ability of the kidneys to excrete an acid load and worsens the acidosis.

Because the function of H+ –ATPase is normal, the urine is appropriately acidic in this form of RTA. Correction of hyperkalemia leads to correction of metabolic acidosis in many patients, pointing to the central role of hyperkalemia in the pathogenesis of this acidosis.

Almost all patients with type 4 RTA manifest varying degrees of hyperkalemia, which commonly is asymptomatic. The etiology of hyperkalemia is multifactorial and related to the presence of hypoaldosteronism in conjunction with a degree of renal insufficiency. The acidosis and hyperkalemia, however, are out of proportion to the degree of renal failure.

The following findings are typical of type 4 RTA:

  • Mild-to-moderate chronic kidney disease (stages 2-3) in most patients, with a creatinine clearance of 30-60 mL/min
  • Hyperkalemia
  • Hypoaldosteronism
  • Diabetes mellitus (in approximately 50% of patients)

Type 4 RTA should be suspected in any patient with a mild non-AG metabolic acidosis and hyperkalemia. The serum HCO3- level is usually greater than 15 mEq/L, and the urine pH is less than 5.0 because these patients have a normal ability to secrete H+. The primary problem is hyperkalemia from aldosterone deficiency or end organ (collecting duct) resistance to the action of aldosterone. This can be diagnosed by measuring the transtubular potassium gradient (TTKG).

TTKG = urine K+ X serum osmolality/serum K+ X urine osmolality

A TTKG greater than 8 indicates that aldosterone is present and the collecting duct is responsive to it. A TTKG less than 5 in the presence of hyperkalemia indicates aldosterone deficiency or resistance. For the test to be interpretable, the urine Na+ level should be greater than 10 mEq/L and the urine osmolality should be greater than or equal to serum osmolality.

The hyperkalemia suppresses renal ammoniagenesis, leading to a lack of urinary buffers to excrete the total H+ load. The urine AG will be positive. Note that patients with hyperkalemic type 1 RTA have a urine pH greater than 5.5 and a low urine Na+.

The following are causes of type 4 RTA:

  • Hypoaldosteronism (low renin) - Hyporeninemic hypoaldosteronism (diabetes mellitus/mild renal impairment, chronic interstitial nephritis, nonsteroidal anti-inflammatory drugs, beta-blockers)

  • Hypoaldosteronism (high renin) - Primary adrenal defect (isolated: congenital hypoaldosteronism; generalized: Addison disease, adrenalectomy, AIDS), inhibition of aldosterone secretion (heparin, ACE inhibitors, AT1 receptor blockers)

  • Aldosterone resistance (drugs) - Diuretics (amiloride, triamterene, spironolactone), calcineurin inhibitors (cyclosporine, tacrolimus), antibiotics (trimethoprim, pentamidine)

  • Aldosterone resistance (genetic) - Pseudohypoaldosteronism (PHA) types I and II

Although type 4 RTA occurs sporadically, familial forms have been reported. The genetic forms are called PHA; PHA type 1 is characterized by hypotension with hyperkalemia and acidosis and includes an autosomal recessive and autosomal dominant form. PHA type 2 is characterized by hypertension with hyperkalemia and acidosis and is also known as Gordon syndrome and familial hyperkalemic hypertension. Note the following:

  • Autosomal recessive PHA type 1: Homozygous mutations in the alpha, beta, or gamma subunits (gene symbols SCNN1A, SCNN1B, and SCNN1G) of the collecting duct epithelial sodium channel cause a syndrome that manifests in infancy with severe salt wasting, hypotension, hyperkalemia, and acidosis. A pulmonary syndrome characterized by recurrent respiratory infections, chronic cough, and increased respiratory secretions has also been noted in some individuals.

  • Autosomal dominant PHA type 1: Heterozygous mutations in the mineralocorticoid receptor lead to a milder phenotype that is restricted to the kidneys. Unlike the autosomal recessive form, the clinical symptoms improve with age.

  • Gordon syndrome (PHA type 2): This disorder is characterized by hypertension and hyperkalemia with variable degrees of metabolic acidosis. There are at least 5 genetic loci associated with this disease. Heterozygous mutations in 1 of 2 kinases, WNK1 or WNK4, or in the CUL3 gene cause this syndrome. Heterozygous or homozygous mutations in the KLHL3 genecause an autosomal dominant or recessive form of this syndrome. A fifth locus on band 1q has been described, but the genetic defect at this locus has not yet been identified.

Table. Comparison of Types 1, 2, and 4 RTA (Open Table in a new window)


Proximal (Type 2)

Distal (Type 1)

Type 4

Primary defect

Proximal HCO3 - reabsorption

Diminished distal H+ secretion

Diminished ammoniagenesis

Urine pH

< 5.5 when serum HCO3 - is low


< 5.5

Serum HCO3 -

>15 mEq/L

Can be < 10 mEq/L

>15 mEq/L

Fractional excretion of HCO3 - (FEHCO3)

>15-20% during HCO3 - load

< 5% (can be as high as 10% in children)

< 5%

Serum K+

Normal or mild decrease

Mild-to-severe decrease*


Associated features

Fanconi syndrome


Diabetes mellitus, renal insufficiency

Alkali therapy

High doses

Low doses

Low doses


Osteomalacia or rickets

Nephrocalcinosis, nephrolithiasis


*K+ may be high if RTA is due to volume depletion.

Early renal failure

Metabolic acidosis is usual in patients with renal failure, and, in early to moderate stages of chronic kidney disease (glomerular filtration rate of 20-50 mL/min), it is associated with a normal AG (hyperchloremic). In more advanced renal failure, the acidosis is associated with a high AG.

In hyperchloremic acidosis, reduced ammoniagenesis (secondary to loss of functioning renal mass) is the primary defect, leading to an inability of the kidneys to excrete the normal daily acid load. In addition, NH3 reabsorption and recycling may be impaired, leading to reduced medullary interstitial NH3 concentration.

In general, patients tend to have a serum HCO3- level greater than 12 mEq/L, and buffering by the skeleton prevents further decline in serum HCO3-.

Note that patients with hypobicarbonatemia from renal failure cannot compensate for additional HCO3- loss from an extrarenal source (eg, diarrhea) and severe metabolic acidosis can develop rapidly.

Urinary diversion

Hyperchloremic metabolic acidosis can develop in patients who undergo a urinary diversion procedure, such as a sigmoid bladder or an ileal conduit.

This occurs through 1 of the following 2 mechanisms:

The first is the intestinal mucosa has an apical Cl-/HCO3- exchanger. When urine is diverted to a loop of bowel (as in patients with obstructive uropathy), the chloride in the urine is exchanged for HCO3-. Significant loss of HCO3- can occur, with a concurrent increase in serum Cl- concentration.

The second is intestinal mucosa reabsorbs urinary NH4+, and the latter is metabolized in the liver to NH3 and H+. This is particularly likely to occur if urine contact time with the intestinal mucosa is prolonged, as when a long loop of bowel is used or when the stoma is obstructed and when sigmoid rather than ileal loop is used. Presumably, the creation of a continent bladder also increases HCO3- loss. This disorder is not observed very frequently anymore because short-loop incontinent ureteroileostomies are used now.

Infusion of acids

The addition of an acid that contains Cl- as an ion (eg, NH4 Cl) can result in a normal-AG acidosis because the drop in HCO3- is accompanied by an increase in Cl-.

The use of arginine or lysine hydrochloride as amino acids during hyperalimentation can have the same result.

Specific causes of high-AG metabolic acidosis

Lactic acidosis

Briefly, L-lactate is a product of pyruvic acid metabolism in a reaction catalyzed by lactate dehydrogenase that also involves the conversion of nicotinamide adenine dinucleotide (NADH) to the oxidized form of nicotinamide adenine dinucleotide (NAD+). This is an equilibrium reaction that is bidirectional, and the amount of lactate produced is related to the reactant concentration in the cytosol (pyruvate, NADH/NAD+).

Daily lactate production in a healthy person is substantial (approximately 20 mEq/kg/d), and this is usually metabolized to pyruvate in the liver, the kidneys, and, to a lesser degree, in the heart. Thus, production and use of lactate (ie, Cori cycle) is constant, keeping plasma lactate low.

The major metabolic pathway for pyruvate is to acetyl coenzyme A, which then enters the citric acid cycle. In the presence of mitochondrial dysfunction, pyruvate accumulates in the cytosol and more lactate is produced.

Lactic acid accumulates in blood whenever production is increased or use is decreased. A value greater than 4-5 mEq/L is considered diagnostic of lactic acidosis.

Type A lactic acidosis occurs in hypoxic states, while type B occurs without associated tissue hypoxia.

D-lactic acidosis is a form of lactic acidosis that occurs from overproduction of D-lactate by intestinal bacteria. It is observed in association with intestinal bacterial overgrowth syndromes. D-lactate is not measured routinely when lactate levels are ordered and must be requested specifically when such cases are suspected.

Metformin-associated lactic acidosis has been reported. [11, 12]     


Free fatty acids released from adipose tissue have 2 principal fates. In the major pathway, triglycerides are synthesized in the cytosol of the liver. In the less common pathway, fatty acids enter mitochondria and are metabolized to ketoacids (acetoacetic acid and beta-hydroxybutyric acid) by the beta-oxidation pathway. Ketoacidosis occurs when delivery of free fatty acids to the liver or preferential conversion of fatty acids to ketoacids is increased.

This pathway is favored when insulin is absent (as in the fasting state), in certain forms of diabetes, and when glucagon action is enhanced.

Alcoholic ketoacidosis occurs when excess alcohol intake is accompanied by poor nutrition. Alcohol inhibits gluconeogenesis, and the fasting state leads to low insulin and high glucagon levels. These patients tend to have a mild degree of lactic acidosis. This diagnosis should be suspected in alcoholic patients who have an unexplained AG acidosis, and detection of beta-hydroxybutyric acid in the serum in the absence of hyperglycemia is highly suggestive. Patients may have more than one metabolic disturbance (eg, mild lactic acidosis, metabolic alkalosis secondary to vomiting).

Starvation ketoacidosis can occur after prolonged fasting and may be exacerbated by exercise.

DKA is usually precipitated in patients with type 1 diabetes by stressful conditions (eg, infection, surgery, emotional trauma), but it can also occur in patients with type 2 diabetes. Hyperglycemia, metabolic acidosis, and elevated beta-hydroxybutyrate confirm the diagnosis. The metabolic acidosis in DKA is commonly a high-AG acidosis secondary to the presence of ketones in the blood. However, after initiation of treatment with insulin, ketone production ceases, the liver uses ketones, and the acidosis becomes a non-AG type that resolves in a few days (ie, time necessary for kidneys to regenerate HCO3-, which was consumed during the acidosis).

Advanced renal failure

Patients with advanced chronic kidney disease (glomerular filtration rate of less than 20 mL/min) present with a high-AG acidosis. The acidosis occurs from reduced ammoniagenesis leading to a decrease in the amount of H+ buffered in the urine. The increase in AG is thought to occur because of the accumulation of sulfates, urates and phosphates from a reduction in glomerular filtration and from diminished tubular function.

In persons with chronic uremic acidosis, bone salts contribute to buffering, and the serum HCO3- level usually remains greater than 12 mEq/L. This bone buffering can lead to significant loss of bone calcium with resulting osteopenia and osteomalacia.

Salicylate overdose

Deliberate or accidental ingestion of salicylates can produce a high-AG acidosis, although respiratory alkalosis is usually the more pronounced acid-base disorder.

The increase in AG is only partly from the unmeasured salicylate anion. Increased ketoacid and lactic acid levels have been reported in persons with salicylate overdose and are thought to account for the remainder of the AG.

Salicylic acid ionizes to salicylate and H+ ion with increasing pH; at a pH of 7.4, only 0.004% of salicylic acid is nonionized, as follows:

Salicylic acid (HS)↔salicylate (S) + H+ (H+)

HS is lipid soluble and can diffuse into the CNS; with a fall in pH, more HS is formed. The metabolic acidosis thus increases salicylate entry to the CNS, leading to respiratory alkalosis and CNS toxicity.

Methanol poisoning

Methanol ingestion is associated with the development of a high-AG metabolic acidosis. Methanol is metabolized by alcohol dehydrogenase to formaldehyde and then to formic acid.

Formaldehyde is responsible for the optic nerve and CNS toxicity, while the increase in AG is from formic acid and from lactic acid and ketoacid accumulation.

Clinical manifestations include optic nerve injury that can be appreciated by funduscopic examination as retinal edema, CNS depression, and unexplained metabolic acidosis with high anion and osmolar gaps.

Ethylene glycol poisoning

Ingestion of ethylene glycol, a component of antifreeze and engine coolants, leads to a high-AG acidosis. Ethylene glycol is converted by alcohol dehydrogenase first to glycoaldehyde and then to glycolic and glyoxylic acids. Glyoxylic acid then is degraded to several compounds, including oxalic acid, which is toxic, and glycine, which is relatively innocuous.

The high AG is primarily from the accumulation of these acids, although a mild lactic acidosis also may be present.

Patients present with CNS symptoms, including slurred speech, confusion, stupor or coma, myocardial depression, and renal failure with flank pain.

Oxalate crystals are usually observed in the urine and are an important clue to the diagnosis, as is an elevated osmolar gap.



Morbidity and mortality in metabolic acidosis are primarily related to the underlying condition.

In a prospective, observational, cohort study, Maciel and Park looked at differences between survivors and nonsurvivors within a group of 107 patients suffering from metabolic acidosis on admission to an intensive care unit (ICU). [13] The authors found that although acidosis was more severe in nonsurvivors than in survivors, the proportion of acidifying variables was similar on admission between the 2 groups (with hyperchloremia being the primary cause of the acidosis).

The investigators also found that in nonsurviving patients, the degree of metabolic acidosis was similar on the day of death to the level measured when they were admitted to the ICU, but that the proportion of anions had changed. Specifically, the chloride levels in the patients had decreased, and the lactate levels had increased.