Metformin Associated Acidosis

Dr. Chris Parker

The Case

You are on call during your toxicology rotation and the following case is called into the poison center. 


ABG 7.23/29/101

Na 131

K 4.8

Cl 92

HCO3 10

BUN 26

Cr 1.78

Gluc 196

AG 29 (previously 20)

Lactate 15.7

WBC 23.2

Hct 22.9

Hgb 9 (previously 16)

Plt 59 (previously 150)

Alb 2.5

AST 42

ALT 102

ALP 45

Bili 7.9

INR 2.3

PT 26.9

PTT 56.6

UDS +opiates

EtOH neg

APAP neg

ASA 8, down trended to 7

Betahydroxyurea 0.8

Serum Osm 301 (calculated Osms 286 with a gap of 15)


The poison center was consulted with concern for possible toxic alcohol ingestion. After discussing the case with your attending, you recommended starting fomepizole while the toxic alcohol levels are pending and reiterate the importance of aggressive supportive care. Subsequent labs demonstrated that his methanol and ethylene glycol were negative. The question was raised as to whether this could be a case of metformin-associated lactic acidosis. 


The biguanide metformin is a first-line oral antihyperglycemic agent used to treat type 2 diabetes mellitus. 

Metformin decreases hepatic gluconeogenesis from lactate and improves peripheral uptake of glucose. It has been postulated that under certain conditions metformin can cause mitochondrial inhibition, resulting in the release of free protons, which leads to increased lactate concentrations (1). 

Another biguanide, phenformin, was removed from the US market in 1977 due to association with a life-threatening metabolic acidosis with elevated lactic acid (1). However, this adverse event is 20x less common in metformin.

This process has been termed metformin-associated lactic acidosis (MALA), also referred to as metabolic acidosis with an elevated lactate concentration.

The biochemical and pathophysiologic process involving this condition is complex, but metformin is typically being taken at doses that exceed recommendations when lactic acidosis occurs. Risk factors for MALA include renal dysfunction, cardiorespiratory insufficiency, septicemia, liver disease, a history of MALA, advanced age, alcohol abuse, and radiologic contrast exposure (1). A retrospective study by Dell’Aglio et al. found no deaths occurring in patients who documented the lowest pH above 6.9, peak serum lactate concentration less than 25 mmol/L, or peak metformin level less than 50 mcg/ml (2). 

Common symptoms include nausea, vomiting, abdominal pain, malaise, myalgias, and dizziness. Physical exam findings are non-specific, but severe presentations may have blindness, confusion, respiratory insufficiency, hypothermia, or hypotension (1). Labs should be used to rule out alternative causes. Serum metformin concentration can be ordered, but it is not required if there is a high clinical suspicion.  

Management includes aggressive supportive care. A definitive airway should be obtained and pressors initiated, if indicated. Sodium bicarbonate drips are controversial in severe metabolic acidosis, but initiation has been proposed when the serum bicarbonate concentration is less than 5 mEq/L (1). Unfortunately, metformin is only moderately dialyzable due to a large volume of distribution. There are case reports of extracorporeal membrane oxygenation (ECMO) bridging patients to recovery after metformin overdose (3). ECMO should be considered if serum lactate level is greater than 20 mmol/L, pH is less than or equal to 7.00, or with failure of standard therapy (4).


The question regarding MALA is whether metformin or other risk factors for lactic acidosis caused the presenting adverse event. As demonstrated in the present case, the patient has a gastrointestinal illness with associated hemorrhage, dehydration, and acute kidney injury, which puts him at risk for elevated metformin levels. He also has a history of alcohol abuse, septicemia, liver disease, and cardiorespiratory failure. Each of these conditions increases his risk for MALA. Unfortunately, all of them could also independently cause a lactic acidosis. 

Other questions regarding MALA have been raised. The incidence of lactic acidosis in diabetics is reported at 9.7 cases per 100,000 patient-years regardless of metformin use, suggesting no causal relationship (5). Furthermore, prescribed metformin is typically supratherapeutic in reported cases of MALA, but pharmacokinetic modeling of these cases suggests pre-admission plasma concentrations did not exceed the therapeutic range (6). This raises the question of whether metformin is the primary cause of the lactic acidosis in most published cases or just playing a contributory role. 

The literature supports that most published cases of MALA have risk factors that on their own could have led to the lactic acidosis (6,7). It has been argued that true metformin-induced lactic acidosis requires no other confounding conditions that could contribute to the presenting lactic acidosis. Kuan et al. proposed four key criteria for assessment of causality, including: 

1. Use of metformin prior to onset

2. Whether an alternative etiology may have caused the presentation and to what degree

3. Measured lactic acidosis

4. Removal of metformin improves the patient’s condition (6)

For most cases, including our present patient, these criteria are not very helpful. For example, we are uncertain if he has been compliant with his metformin, and his risk factors for MALA are highly suspect for causing his lactic acidosis. Finally, though he did improve with discontinuation of metformin, as we will discuss, he also had aggressive supportive therapy correcting other possible etiologies. 

Regarding our patient, it is safe to conclude that he was very sick on presentation with multiple system failure. It is possible that his metformin contributed to his elevated lactate, but it is unlikely that this was the primary etiology due to his confounding conditions. It would have been helpful to know his metformin level at presentation, but this was not ordered and would not have changed his management. This case demonstrates how difficult it can be to make the diagnosis of MALA and highlights the controversy surrounding this condition. What is most important is that we avoid premature closure and keep a broad differential. These patients require a thorough workup and aggressive supportive care, regardless of whether the diagnosis of MALA is established.


Take Home Points