Bicarbonate drops in diabetic ketoacidosis (DKA) because the body uses it up trying to neutralize a flood of acids it cannot stop producing. When insulin is absent or severely deficient, the liver ramps up production of ketoacids, and the blood’s main chemical buffer, bicarbonate, gets consumed in the effort to keep blood pH from plummeting. The degree of bicarbonate depletion tracks closely with how severe the episode is, and understanding the chemistry behind that drop explains much of what clinicians watch for and how they treat the condition.
What Happens When Insulin Disappears
Insulin does more than regulate blood sugar. It also acts as a brake on fat breakdown. When insulin levels crash, whether from missed doses in type 1 diabetes, severe illness, or other triggers, that brake releases. Fat tissue begins dumping large quantities of free fatty acids into the bloodstream, and the liver eagerly takes them up. Under normal conditions the liver would burn those fatty acids for energy at a moderate pace or convert them to storage forms. But without insulin, and with rising levels of the counter-hormone glucagon, the liver shifts into a mode where it converts fatty acids into ketone bodies at a much faster rate than the rest of the body can use them.
Research in animal models showed that administering either anti-insulin serum or glucagon to fed rats switched the liver from a “nonketogenic” to a “ketogenic” profile within just one hour, with corresponding increases in the production of the ketoacids acetoacetate and beta-hydroxybutyrate.1The Journal of Clinical Investigation. Hormonal control of ketogenesis. Rapid activation of hepatic ketogenic capacity in fed rats by anti-insulin serum and glucagon. That speed matters clinically. DKA can develop within hours when insulin delivery is interrupted, and the liver’s rapid pivot to ketone production is the reason.
How Ketoacids Consume Bicarbonate
The blood maintains a tightly regulated pH, and bicarbonate is its most important chemical buffer. When an acid enters the bloodstream, bicarbonate reacts with it: the bicarbonate ion combines with the hydrogen ion released by the acid, producing carbon dioxide and water. The CO₂ is exhaled through the lungs, and the acid is effectively neutralized. Under normal circumstances, the kidneys regenerate bicarbonate at roughly the same rate it gets used, keeping the system in balance.
In DKA, that balance collapses. The liver is pumping out acetoacetate and beta-hydroxybutyrate far faster than the kidneys can regenerate bicarbonate. Each molecule of ketoacid that enters the blood releases a hydrogen ion, which immediately gets neutralized by a molecule of bicarbonate. As the ketoacids accumulate, bicarbonate is consumed faster and faster. When the supply runs low, the blood pH begins to drop measurably, and the clinical picture of metabolic acidosis sets in.
This is not a slow process. In a study of emergency department patients presenting with DKA, the average bicarbonate level was around 11 to 13 mmol/L, roughly half to a third of the normal range, and the average blood pH had already fallen to about 7.2, well below the normal value of 7.4.2Annals of Emergency Medicine. Comparison of Arterial and Venous Blood Gas Values in the Initial Emergency Department Evaluation of Patients With Diabetic Ketoacidosis The anion gap in those same patients averaged about 27 mmol/L, indicating a large quantity of unmeasured anions, mainly ketoacids, circulating in the blood.
The Kidneys Cannot Keep Up
Normally the kidneys are bicarbonate factories. They reabsorb almost all filtered bicarbonate in the proximal tubule and generate new bicarbonate through a process tied to acid excretion. But DKA overwhelms this system from multiple angles. The sheer volume of ketoacids being filtered exceeds the kidney’s ability to reclaim them. Some ketoacid anions spill into the urine, carrying sodium and potassium with them rather than being excreted alongside ammonium. When ketoacid salts leave the body as sodium or potassium salts instead of ammonium salts, that represents an indirect loss of bicarbonate, because the sodium or potassium that leaves with the ketoanion would otherwise have been paired with a newly generated bicarbonate ion headed back to the blood.
On top of that, severe dehydration, a hallmark of DKA caused by the osmotic diuresis from high blood glucose, reduces blood flow to the kidneys. With less blood flowing through, the kidneys’ capacity to excrete acid and regenerate bicarbonate diminishes further. The result is a vicious cycle: the sicker you get, the less help your kidneys can provide.
How the Body Tries to Fight Back
When the blood becomes acidic, the body does not sit idle. The respiratory system kicks in as a compensatory mechanism. Chemoreceptors sense the falling pH and trigger faster, deeper breathing. In mild acidosis this shows up as an increased breathing rate. In more severe DKA, it progresses to a distinctive pattern called Kussmaul breathing, characterized by deep, labored, sighing respirations.3PubMed Central. Effects of diabetic ketoacidosis in the respiratory system The goal of this exaggerated breathing is to blow off carbon dioxide as fast as possible. Since CO₂ is itself an acid when dissolved in blood, lowering CO₂ levels partially compensates for the bicarbonate deficit and slows the pH drop.
This respiratory compensation buys time but cannot fix the underlying problem. As long as ketoacid production continues unchecked, bicarbonate will keep falling. Kussmaul breathing is actually a useful clinical sign: its presence tells clinicians that the acidosis is significant enough to provoke a strong compensatory response. When it disappears during treatment, that is a good sign that the metabolic picture is improving.
Grading Severity by Bicarbonate and pH
Clinicians use the bicarbonate level alongside pH and the anion gap to classify how bad a DKA episode is. The American Diabetes Association classification, widely used in clinical practice, breaks it into three tiers:4PubMed Central. Comparisons of biochemical parameters and diabetic ketoacidosis severity in adult patients with type 1 and type 2 diabetes
- Mild DKA: pH above 7.25, bicarbonate at or above 15 mmol/L, anion gap above 10.
- Moderate DKA: pH between 7.0 and 7.24, bicarbonate between 10 and 15 mmol/L, anion gap above 12.
- Severe DKA: pH below 7.0, bicarbonate below 10 mmol/L, anion gap above 12.
Bicarbonate below 10 signals that the buffering system is nearly exhausted. At that point the body is running out of chemical tools to prevent the pH from crashing into dangerous territory, and the risk of organ dysfunction rises sharply. This grading system is not just academic bookkeeping; it directly influences treatment decisions, including how aggressively fluids are given, whether intensive care monitoring is needed, and in very rare cases whether exogenous bicarbonate should be considered.
When Lactic Acid Piles On
Ketoacids are not the only acids that may be elevated during DKA. Lactic acid frequently rises as well, and when it does, the bicarbonate takes a double hit. Elevated lactate in DKA appears to result partly from the severe dehydration and poor tissue perfusion that accompany the condition. When tissues do not get enough oxygen-rich blood, they shift to anaerobic metabolism, producing lactic acid as a byproduct.
A study examining lactate levels in DKA patients found that higher lactate was associated with lower pH, lower blood pressure, and higher illness severity scores, all consistent with the idea that poor perfusion drives much of the lactate rise.5PubMed Central. Lactic Acidosis in Diabetic Ketoacidosis: A Marker of Severity or Alternate Substrate for Metabolism But the same researchers also found a positive correlation between lactate and glucose levels, suggesting that the metabolic derangement itself, not just poor perfusion, contributes. In practical terms, when lactate is elevated alongside ketoacids, bicarbonate is being consumed even faster than the ketoacid load alone would predict, and the clinical picture tends to be more severe.
A Comparison With Alcoholic Ketoacidosis
DKA is not the only condition that produces dangerous levels of ketoacids. Alcoholic ketoacidosis (AKA) can look similar on basic lab work, but the underlying chemistry differs in ways that affect how bicarbonate behaves. In a study comparing the two conditions, patients presenting with DKA had dramatically higher blood glucose (about 32 mmol/L versus 6.6 mmol/L in AKA) despite similar concentrations of the ketoacid beta-hydroxybutyrate.6Journal of Critical Care. Differences in metabolic and hormonal milieu in diabetic- and alcohol-induced ketoacidosis The ratio of beta-hydroxybutyrate to acetoacetate was also different: about 3 to 1 in DKA versus 7 to 1 in AKA. That ratio matters diagnostically because some older lab tests only detect acetoacetate, meaning AKA can appear deceptively mild on a standard urine ketone dip even when the total ketoacid burden is high.
From the bicarbonate perspective, both conditions consume it through the same buffering reaction. But the much higher glucose in DKA adds a layer of osmotic diuresis and dehydration that is less pronounced in AKA, further taxing the kidneys’ ability to regenerate bicarbonate. AKA also tends to resolve faster with glucose and fluid administration alone, partly because the underlying insulin deficiency in DKA is the harder problem to correct.
SGLT2 Inhibitors and Euglycemic DKA
A newer twist on the DKA story involves a class of diabetes medications called SGLT2 inhibitors, which work by causing the kidneys to excrete glucose in the urine. These drugs can occasionally trigger a form of DKA where blood sugar stays near normal, a condition called euglycemic DKA. The low bicarbonate in this variant has an additional mechanism layered on top of the classic one.
SGLT2 inhibitors alter how the kidney handles energy. By blocking glucose reabsorption in the proximal tubule, the drug reduces the kidney’s demand for ATP. In response, the kidney dials down several energy-producing pathways, including ones responsible for generating new bicarbonate and for metabolizing ketoacids. The practical result is that the kidney excretes ketoacid anions as sodium and potassium salts rather than as ammonium salts, which amounts to an indirect loss of bicarbonate from the body.7PubMed Central. Euglycemic Ketoacidosis as a Complication of SGLT2 Inhibitor Therapy On top of that, the kidney’s ability to produce new bicarbonate through ammoniagenesis is impaired, and chloride levels tend to rise as bicarbonate falls, creating a persistent hyperchloremic metabolic acidosis that compounds the problem.8Annals of the American Thoracic Society. Sodium-Glucose Cotransporter 2 Inhibitor–induced Ketoacidosis and Non-Ketotic Acidosis: When Is There a Role for Bicarbonate?
The mechanism that triggers ketogenesis in these patients also starts at the kidney. Because the drug forces glucose out in the urine, blood sugar stays lower than it otherwise would, which means insulin secretion stays lower too. That relative insulin deficit, combined with carbohydrate depletion and rising glucagon, pushes the liver toward ketone production just as it would in classic DKA, but without the telltale high blood sugar that usually sounds the alarm.9Kidney Medicine. Why Is Bicarbonate Low in Diabetic Ketoacidosis? The danger here is that patients and sometimes clinicians do not recognize what is happening because the glucose reading looks reassuring.
Why Giving Bicarbonate Directly Rarely Helps
It might seem logical that if the problem is low bicarbonate, the solution would be to infuse bicarbonate directly. Clinicians debated this for decades, and the evidence consistently says otherwise for the majority of DKA episodes. A systematic review and meta-analysis pooling data from over 600 patients found that bicarbonate therapy provided no meaningful benefit in routine DKA management. It did not speed up the resolution of acidosis, improve pH or electrolyte outcomes, or shorten hospital stays. In fact, it was associated with potential harm, including worsened hyperglycemia and, in one analysis, prolonged hospitalization.10PubMed Central. The Role of Bicarbonate Therapy in Diabetic Ketoacidosis: A Systematic Review and Meta‐Analysis A separate review of the broader literature reached the same conclusion: the bulk of the data argue against significant benefit in mild to moderately severe acidosis and suggest possible adverse effects.11PubMed Central. Use of sodium bicarbonate and blood gas monitoring in diabetic ketoacidosis: A review
The reason is straightforward once you think about it. The low bicarbonate is a consequence, not the root cause. The root cause is uncontrolled ketoacid production driven by insulin deficiency. Giving insulin stops the flood of new acid at its source. Once insulin is on board and the liver stops overproducing ketoacids, the body’s own regeneration systems, particularly the kidneys, can rebuild bicarbonate stores. Pouring in exogenous bicarbonate without addressing the ketoacid production is like mopping the floor while the faucet is still running.
Current guidelines reserve bicarbonate therapy for extreme cases where the pH drops below about 6.9, a level at which the acidosis itself threatens heart function and survival. Even then, the evidence supporting its use in those dire circumstances is thin; it is given mostly as a physiological rescue measure while insulin and fluids work to address the underlying problem.
The Pediatric Concern
In children, there is an additional reason to avoid bicarbonate: it may increase the risk of cerebral edema, one of the most feared complications of DKA in young patients. Retrospective evidence has linked bicarbonate administration in pediatric DKA to a higher risk of cerebral edema and longer hospital stays.12PubMed Central. Bicarbonate in diabetic ketoacidosis – a systematic review The mechanism likely involves rapid shifts in blood pH and CO₂ that affect how fluid moves across the blood-brain barrier. Because of this, pediatric DKA protocols are especially cautious about bicarbonate use, and most pediatric guidelines recommend against it entirely except in extreme circumstances.
Fluid Choice and the Chloride Problem
Even the choice of intravenous fluid used to rehydrate a DKA patient affects how quickly bicarbonate recovers. Normal saline, the traditional workhorse of DKA resuscitation, contains a high concentration of chloride. When large volumes are given, the excess chloride can itself cause a form of acidosis called hyperchloremic metabolic acidosis, which further suppresses bicarbonate. In a study comparing normal saline to a balanced electrolyte solution in DKA patients, those who received the balanced fluid ended up with lower chloride levels (about 105 mmol/L versus 111 mmol/L) and higher bicarbonate after resuscitation (about 20 mmol/L versus 17 mmol/L).13PubMed. Resuscitation with balanced electrolyte solution prevents hyperchloremic metabolic acidosis in patients with diabetic ketoacidosis
This matters because clinicians often track the anion gap to determine when DKA has resolved. If normal saline is given in large volumes, the anion gap may close as ketoacids are cleared, but bicarbonate stays stubbornly low because chloride has taken its place. The patient’s lab work then looks like it has shifted from a ketoacid-driven gap acidosis to a chloride-driven non-gap acidosis, which can confuse the clinical picture and lead to unnecessarily prolonged treatment. Using balanced fluids sidesteps this problem, though the practice is still not universal in all emergency departments.
How Venous and Arterial Samples Compare
When tracking bicarbonate during DKA treatment, clinicians need reliable blood gas measurements. Arterial blood draws are the traditional gold standard but are more painful and carry slightly more risk than a simple venous draw. In the emergency department study mentioned earlier, arterial and venous bicarbonate values were highly correlated, with a correlation coefficient above 0.95, and the average venous bicarbonate was only about 1 to 2 mmol/L higher than the arterial value.2Annals of Emergency Medicine. Comparison of Arterial and Venous Blood Gas Values in the Initial Emergency Department Evaluation of Patients With Diabetic Ketoacidosis The pH difference between arterial and venous samples averaged just 0.03. This tight agreement means venous blood gases are generally sufficient for initial evaluation and ongoing monitoring in most DKA cases, sparing patients from repeated arterial sticks. If the venous bicarbonate is low enough to confirm DKA and guide treatment, an arterial sample rarely changes the clinical decision.