Residency · Residency · Nephrology

Acid-Base Physiology and Arterial Blood Gas Interpretation

Introduction

Normal arterial pH is maintained within a remarkably narrow range of 7.35 to 7.45, corresponding to a hydrogen ion concentration of 35 to 45 nanomoles per liter. The body generates an enormous acid load daily: approximately 15,000 mmol of volatile acid in the form of carbon dioxide through oxidative metabolism, and approximately 1 mmol per kilogram per day (roughly 70 mEq) of non-volatile, or fixed, acid from the metabolism of sulfur-containing amino acids in dietary protein and from phospholipid metabolism. Three lines of defense maintain pH homeostasis: chemical buffering, which acts instantaneously; respiratory compensation, which adjusts PaCO2 over minutes to hours; and renal compensation, which modulates bicarbonate regeneration and acid excretion over hours to days.

Buffer Systems

Bicarbonate-CO2 Buffer System

The bicarbonate-carbon dioxide buffer system is the most important extracellular buffer, described by the equilibrium reaction CO2 + H2O equilibrating with H2CO3, which dissociates into H+ and HCO3-. The Henderson-Hasselbalch equation expresses this relationship quantitatively: pH equals 6.1 plus the logarithm of the ratio of bicarbonate concentration to 0.03 times PaCO2. The unique physiologic importance of this system lies in its open nature -- CO2 is volatile and can be excreted by the lungs, meaning the denominator of the Henderson-Hasselbalch equation is under respiratory control, while the numerator (bicarbonate) is under renal control. Normal values are a bicarbonate concentration of approximately 24 mEq/L and a PaCO2 of approximately 40 mmHg.

Non-Bicarbonate Buffers

Hemoglobin is the primary intracellular buffer in red blood cells, and deoxyhemoglobin is a more effective buffer than oxyhemoglobin, which has important implications for oxygen delivery in acidotic states. The phosphate buffer system, with a pKa of 6.8, is an important intracellular buffer and plays a particularly significant role in urinary buffering of excreted hydrogen ions. Plasma and intracellular proteins, especially albumin, contribute substantially to buffering capacity. Bone mineral, including calcium carbonate and calcium phosphate, provides a reservoir for buffering in chronic metabolic acidosis, but this mobilization of bone buffer contributes to the metabolic bone disease seen in chronic kidney disease.

Renal Acid-Base Handling

Bicarbonate Reclamation (Proximal Tubule)

Approximately 4320 mEq of bicarbonate is filtered daily across the glomerulus, and roughly 80 percent of this enormous load is reclaimed in the proximal convoluted tubule. The mechanism involves secretion of hydrogen ions into the tubular lumen via the apical NHE3 sodium-hydrogen exchanger. These secreted hydrogen ions combine with filtered bicarbonate to form carbonic acid, which is rapidly dehydrated to CO2 and water by carbonic anhydrase IV, an enzyme anchored to the luminal brush border membrane. The CO2 diffuses freely into the tubular cell, where intracellular carbonic anhydrase II catalyzes its rehydration to carbonic acid, which dissociates into hydrogen ions (recycled back to the lumen via NHE3) and bicarbonate ions. The newly generated bicarbonate exits the cell across the basolateral membrane via the Na-HCO3 cotransporter NBCe1 (encoded by SLC4A4). The remaining 15 to 20 percent of filtered bicarbonate is reclaimed in the thick ascending limb and collecting duct.

Net Acid Excretion (Collecting Duct)

Net acid excretion, the quantity that matches daily fixed acid production, equals the sum of titratable acid and ammonium excretion minus any urinary bicarbonate loss. Titratable acid consists primarily of the phosphate buffer system, where filtered dibasic phosphate (HPO4 2-) accepts secreted hydrogen ions to form monobasic phosphate (H2PO4-), accounting for approximately 10 to 30 mEq per day of acid excretion.

Ammonium excretion is quantitatively the most important component of net acid excretion, normally accounting for 40 to 60 mEq per day but capable of increasing 5- to 10-fold during chronic acidosis. Ammonia and ammonium are produced in the proximal tubule from the metabolism of glutamine by glutaminase, which generates glutamate and ammonium; further metabolism produces alpha-ketoglutarate along with two molecules of ammonium and two molecules of bicarbonate, which are returned to the systemic circulation as new bicarbonate. Chronic acidosis upregulates both glutaminase and phosphoenolpyruvate carboxykinase (PEPCK), enhancing this pathway. Ammonium secreted into the proximal tubular lumen is reabsorbed in the thick ascending limb by substituting for potassium on the NKCC2 transporter, then recycled through the medullary interstitium. In the collecting duct, ammonia diffuses from the medullary interstitium into the lumen, where it is trapped as ammonium by the hydrogen ions secreted by the apical H-ATPase and H-K-ATPase of type A intercalated cells through the process of non-ionic diffusion trapping.

<image>Detailed diagram of renal acid-base handling showing three panels: (1) Bicarbonate reclamation in the proximal tubule with NHE3 on the apical membrane, carbonic anhydrase IV in the lumen, carbonic anhydrase II intracellularly, and NBCe1 on the basolateral membrane. (2) Ammonium production and handling showing glutamine metabolism in the proximal tubule cell producing NH4+ and HCO3-, NH4+ secretion into the lumen, reabsorption via NKCC2 in the TAL, medullary recycling, and trapping in the collecting duct. (3) Type A intercalated cell in the collecting duct showing apical H-ATPase and H-K-ATPase, intracellular carbonic anhydrase II, and basolateral AE1 (Cl-/HCO3- exchanger). Label all transporters and enzymes.</image>

Systematic ABG Interpretation

Step-by-Step Approach

A systematic approach to arterial blood gas interpretation ensures that no disorder is missed, including complex mixed acid-base disturbances. The first step is to assess the pH to determine whether acidemia (pH less than 7.35) or alkalemia (pH greater than 7.45) is present. The second step is to identify the primary disorder by comparing the direction of change in PaCO2 and bicarbonate with the pH: metabolic acidosis presents with low pH and low bicarbonate, metabolic alkalosis with high pH and high bicarbonate, respiratory acidosis with low pH and high PaCO2, and respiratory alkalosis with high pH and low PaCO2. The third step is to calculate the expected compensation using the appropriate formula and determine whether the observed compensation is appropriate. The fourth step is to calculate the anion gap, and the fifth step, if the anion gap is elevated, is to calculate the delta-delta ratio. The sixth step is to assess for additional disorders that may be revealed by inappropriate compensation or an abnormal delta-delta ratio.

Compensation Rules

The compensation rules are essential clinical tools. For metabolic acidosis, Winter's formula predicts the expected PaCO2: 1.5 times the bicarbonate concentration plus 8, with a margin of plus or minus 2. For metabolic alkalosis, PaCO2 rises approximately 0.7 mmHg for each 1 mEq/L rise in bicarbonate, or equivalently, PaCO2 equals 0.7 times the bicarbonate plus 21. For acute respiratory acidosis, bicarbonate rises by 1 mEq/L per 10 mmHg rise in PaCO2 through chemical buffering alone. For chronic respiratory acidosis, bicarbonate rises by 3.5 mEq/L per 10 mmHg rise in PaCO2 as renal compensation develops over 3 to 5 days. For acute respiratory alkalosis, bicarbonate falls by 2 mEq/L per 10 mmHg fall in PaCO2. For chronic respiratory alkalosis, bicarbonate falls by 5 mEq/L per 10 mmHg fall in PaCO2.

Anion Gap

The anion gap, calculated as sodium minus the sum of chloride and bicarbonate, represents unmeasured anions in the plasma. The normal anion gap is approximately 12 mEq/L, though this varies by laboratory methodology. Because albumin is a major unmeasured anion, the anion gap must be corrected for hypoalbuminemia: the corrected AG equals the observed AG plus 2.5 times the difference between 4.0 g/dL and the measured albumin level. Each 1 g/dL decrease in albumin reduces the expected anion gap by approximately 2.5 mEq/L. An anion gap exceeding 20 mEq/L is virtually always pathologic.

Delta-Delta Ratio (Delta Gap)

The delta-delta ratio is calculated by dividing the change in anion gap (measured AG minus normal AG of 12) by the change in bicarbonate (24 minus measured bicarbonate). A ratio between 1 and 2 indicates a pure anion gap metabolic acidosis. A ratio less than 1 indicates a concurrent non-anion gap metabolic acidosis superimposed on the anion gap acidosis, meaning bicarbonate has fallen more than expected for the degree of anion gap elevation. A ratio greater than 2 indicates a concurrent metabolic alkalosis, meaning bicarbonate is higher than expected for the degree of anion gap elevation.

<image>Flowchart for systematic arterial blood gas interpretation. Begin with pH assessment (acidemia vs alkalemia vs normal). Branch to identify primary disorder based on PaCO2 and HCO3- values. For each primary disorder, show the compensation formula with a decision box: "Is compensation appropriate?" If not, identify the additional mixed disorder. Include an anion gap calculation step for metabolic acidosis with the albumin-corrected AG formula, followed by delta-delta ratio assessment with interpretation ranges (<1, 1-2, >2). Use color coding: red for acidosis pathways, blue for alkalosis pathways.</image>

Stewart (Physicochemical) Approach

Independent Variables

The Stewart approach to acid-base analysis identifies three independent variables that determine pH: the strong ion difference (SID), the total concentration of weak acids (Atot), and PaCO2. The apparent SID is calculated as the sum of strong cations (sodium, potassium, calcium, magnesium) minus the sum of strong anions (chloride, lactate), yielding a normal value of approximately 38 to 42 mEq/L. A decrease in SID produces acidosis, while an increase in SID produces alkalosis. The total weak acid concentration, Atot, is determined primarily by albumin and phosphate concentrations. Hypoalbuminemia, common in ICU patients, produces an alkalotic effect, which is often unrecognized in traditional acid-base analysis.

Strong Ion Gap (SIG)

The strong ion gap, calculated as the difference between the apparent SID and the effective SID (which accounts for the buffering contributions of bicarbonate, albumin, and phosphate), is analogous to the traditional anion gap. A SIG exceeding 2 indicates the presence of unmeasured strong anions. The SIG may be more precise than the traditional anion gap in critically ill patients with hypoalbuminemia, where the traditional AG can appear falsely normal despite the presence of pathologic unmeasured anions.

Clinical Utility

The Stewart approach provides a mechanistic explanation for hyperchloremic acidosis following normal saline infusion: saline has an SID of zero (154 mEq/L each of sodium and chloride), and infusing large volumes reduces the plasma SID, producing acidosis. Balanced crystalloids such as lactated Ringer's and Plasmalyte have an SID of approximately 28 to 29, closer to plasma SID, and therefore cause less acidosis. The SMART trial demonstrated that balanced crystalloids, compared with normal saline, reduced the composite outcome of death, new need for dialysis, and persistent renal dysfunction in critically ill adults.

<image>Table-format visual summary comparing the traditional Henderson-Hasselbalch approach to acid-base analysis with the Stewart physicochemical approach. Show the three independent variables of the Stewart approach (SID, Atot, PaCO2) and how changes in each variable produce acidosis or alkalosis. Include clinical examples for each: decreased SID from hyperchloremia or unmeasured anions, decreased Atot from hypoalbuminemia, and increased PaCO2 from hypoventilation. Show how normal saline (SID = 0) reduces plasma SID and causes acidosis versus balanced crystalloids (SID ~28-29) which preserve plasma SID.</image>

Respiratory Acid-Base Disorders

Respiratory Acidosis (Hypoventilation)

Respiratory acidosis results from alveolar hypoventilation, causing CO2 retention and an increase in PaCO2. Acute causes include CNS depression from opioids, sedatives, or stroke; neuromuscular disease such as myasthenia gravis, Guillain-Barre syndrome, or amyotrophic lateral sclerosis; airway obstruction; and severe pneumonia or ARDS. Chronic causes include COPD, obesity hypoventilation syndrome, chronic neuromuscular disease, and chest wall deformities. Renal compensation differs between acute and chronic states: in acute respiratory acidosis, buffering is limited to chemical mechanisms, providing only modest bicarbonate increases, while in chronic respiratory acidosis, the kidneys increase ammonium excretion and bicarbonate generation over 3 to 5 days, providing substantially more effective compensation.

Respiratory Alkalosis (Hyperventilation)

Respiratory alkalosis is the most common acid-base disorder encountered in hospitalized patients and results from alveolar hyperventilation. Acute causes include anxiety, pain, early sepsis, pulmonary embolism, early salicylate toxicity, pregnancy, hepatic encephalopathy, and high altitude exposure. Chronic causes include hepatic cirrhosis, pregnancy, chronic hypoxemia, and central nervous system lesions. Respiratory alkalosis is unique among acid-base disorders in that chronic compensation can return the pH to near-normal values, unlike metabolic disorders where compensation never fully normalizes pH.

Mixed Acid-Base Disorders

Common Combinations

Several mixed acid-base disturbances are commonly encountered in clinical practice. The combination of anion gap metabolic acidosis with respiratory alkalosis is characteristic of salicylate toxicity and sepsis. Anion gap metabolic acidosis combined with metabolic alkalosis occurs in diabetic ketoacidosis with concurrent vomiting or uremia with nasogastric suction. Anion gap metabolic acidosis combined with non-anion gap metabolic acidosis may occur in DKA with diarrhea or toluene intoxication. The combination of respiratory acidosis with metabolic alkalosis, seen in COPD patients using diuretics, is the most common triple acid-base disorder. True triple disorders, involving all three pathologic processes simultaneously, are identified when pH is near normal but the anion gap is elevated, suggesting an anion gap metabolic acidosis with a concurrent metabolic alkalosis, and the direction of PaCO2 deviation identifies the respiratory component.

Approach to Triple Disorders

When the pH appears inappropriately normal despite an elevated anion gap, the clinician should suspect the concurrent presence of both an anion gap metabolic acidosis and a metabolic alkalosis. The respiratory component is then identified by determining whether the PaCO2 is higher than expected (concurrent respiratory acidosis) or lower than expected (concurrent respiratory alkalosis) for the degree of bicarbonate change.

Key Clinical Pearls

• Always correct the anion gap for albumin; in critically ill patients with albumin of 2.0 g/dL, the "normal" AG is only ~7 mEq/L, and an observed AG of 14 is actually significantly elevated
• Compensation is never complete for metabolic disorders (pH does not return to normal); if pH is normal with abnormal HCO3- and PaCO2, consider a mixed disorder
• Winter's formula is essential: if PaCO2 is higher than predicted, there is a concurrent respiratory acidosis; if lower, a concurrent respiratory alkalosis
• The delta-delta ratio is the most commonly tested and clinically important tool for identifying occult metabolic alkalosis superimposed on AG metabolic acidosis
• Normal saline is a "strong ion acidosis" solution (SID = 0); use balanced crystalloids (LR, Plasmalyte) when large-volume resuscitation is needed to avoid hyperchloremic metabolic acidosis

References

1. Berend K, de Vries APJ, Gans ROB. Physiological Approach to Assessment of Acid-Base Disturbances. N Engl J Med. 2014;371(15):1434-1445.
2. Emmett M, Narins RG. Clinical Use of the Anion Gap. Medicine. 1977;56(1):38-54.
3. Stewart PA. Modern quantitative acid-base chemistry. Can J Physiol Pharmacol. 1983;61(12):1444-1461.
4. Semler MW, Self WH, Wanderer JP, et al. Balanced Crystalloids versus Saline in Critically Ill Adults (SMART). N Engl J Med. 2018;378(9):829-839.
5. Kraut JA, Madias NE. Serum Anion Gap: Its Uses and Limitations in Clinical Medicine. Clin J Am Soc Nephrol. 2007;2(1):162-174.

Read this lecture as Markdown