INTRODUCTION — In the absence of contraindications, metformin is considered the initial medication of choice for hyperglycemia in type 2 diabetes (table 1). Metformin therapy (in the absence of contraindications) can be initiated, concurrent with lifestyle intervention, at the time of diabetes diagnosis [1,2].
The pharmacology, efficacy, contraindications, and side effects of metformin for the treatment of diabetes will be reviewed here. A general discussion of initial treatment of type 2 diabetes and the role of metformin in the prevention of diabetes, in the treatment of polycystic ovary syndrome, and in gestational diabetes are reviewed separately.
●(See "Initial management of hyperglycemia in adults with type 2 diabetes mellitus".)
●(See "Prevention of type 2 diabetes mellitus", section on 'Metformin'.)
●(See "Metformin for treatment of the polycystic ovary syndrome".)
●(See "Gestational diabetes mellitus: Glucose management and maternal prognosis", section on 'Metformin'.)
●(Related Pathway(s): Diabetes: Initial therapy for non-pregnant adults with type 2 DM.)
MECHANISM OF ACTION — Metformin's major effect is to decrease hepatic glucose output by inhibiting gluconeogenesis [3-5]. In addition, metformin increases insulin-mediated glucose utilization in peripheral tissues (such as muscle and liver), particularly after meals, and has an antilipolytic effect that lowers serum free fatty acid concentrations, thereby reducing substrate availability for gluconeogenesis [6-8]. As a result of the improvement in glycemic control, serum insulin concentrations decline slightly [9,10]. Metformin has also been shown to decrease food intake and body weight [11,12].
Metformin suppresses gluconeogenesis by inhibiting a specific mitochondrial isoform of glycerophosphate dehydrogenase (mGPD), an enzyme responsible for converting glycerophosphate to dihydroxyacetone phosphate, thereby preventing glycerol from contributing to the gluconeogenic pathway [13,14]. In addition, inhibition of mGPD leads to accumulation of cytoplasmic NADH and a decrease in the conversion of lactate to pyruvate, limiting lactate contributions to hepatic gluconeogenesis. Excess glycerol and lactate are released into the plasma.
Metformin also activates the enzyme AMP-activated protein kinase (AMPK) in hepatocytes, which appears to be the mechanism by which metformin lowers serum lipid concentrations [15,16]. AMPK-dependent inhibitory phosphorylation of acetyl-CoA carboxylases Acc1 and Acc2 suppresses lipogenesis and lowers cellular fatty acid synthesis in liver and muscle [17,18]. Metformin works through the Peutz-Jeghers protein, LKB1, to regulate AMPK [19]. LKB1 is a tumor suppressor, and activation of AMPK through LKB1 may play a role in inhibiting cell growth [20]. (See 'Cancer incidence' below.)
SUGGESTED APPROACH TO THE USE OF METFORMIN
Patient selection — In the absence of specific contraindications, metformin is considered initial pharmacologic therapy for most patients with type 2 diabetes because of glycemic efficacy, absence of weight gain and hypoglycemia, general tolerability, long-term safety profile, and low cost (table 1) [1,2,21,22]. It can be initiated at the time of diabetes diagnosis, along with consultation for lifestyle intervention. For highly motivated patients with glycated hemoglobin (A1C) near target (<7.5 percent), a three- to six-month trial of lifestyle modification before initiating metformin is reasonable. Other options for initial therapy are discussed elsewhere. (See "Initial management of hyperglycemia in adults with type 2 diabetes mellitus", section on 'Choice of initial therapy'.)
After a successful initial response to metformin, the majority of patients have recurrence of hyperglycemia and require the addition of a second oral or an injectable agent. For patients who require additional therapy, there are a number of medication classes that are available and can be used in combination with metformin. (See "Management of persistent hyperglycemia in type 2 diabetes mellitus", section on 'Our approach'.)
Contraindications — Metformin is contraindicated in patients with factors predisposing to lactic acidosis.
These predisposing factors/contraindications are:
●Impaired kidney function (estimated glomerular filtration rate [eGFR] <30 mL/min/1.73 m2)
●Concurrent active or progressive severe liver disease
●Active alcohol abuse
●Unstable or acute heart failure at risk of hypoperfusion
●Past history of lactic acidosis during metformin therapy
●Decreased tissue perfusion or hemodynamic instability due to infection or other causes
The exact degree of kidney, cardiac, and liver function required for the safe use of metformin remain uncertain [23]. Improved clinical outcomes with metformin have been reported in observational studies of patients with diabetes and heart failure [24-26], reduced kidney function (eGFR 45 to 60 mL/min/1.73 m2) [26-28], or chronic liver disease with hepatic impairment [26]. In one systematic review of 17 observational studies comparing regimens with and without metformin, metformin use was associated with lower all-cause mortality among patients with heart failure, renal impairment, or chronic liver disease with hepatic impairment [26]. In addition, metformin use in patients with renal impairment or heart failure was associated with fewer heart failure readmissions.
On the basis of these and other studies, the US Food and Drug Administration (FDA) revised its labeling of metformin, which previously had identified metformin as contraindicated in individuals with serum creatinine levels ≥1.4 mg/dL (124 micromol/L) and ≥1.5 mg/dL (133 micromol/L), respectively [29]. In the updated labeling, use of metformin is contraindicated in patients with an eGFR <30 mL/min/1.73 m2, and initiation of metformin is not recommended in patients with an eGFR between 30 and 45 mL/min/1.73 m2. For patients taking metformin whose eGFR falls below 45 mL/min/1.73 m2, the benefits and risk of continuing treatment should be assessed, whereas metformin should be discontinued if the eGFR falls below 30 mL/min/1.73 m2.
The following is our approach to the administration of metformin:
●For patients with an eGFR <30 mL/min/1.73 m2, we do not prescribe metformin.
●For patients with an eGFR ≥45 mL/min/1.73 m2, we prescribe full dose.
●For patients with an eGFR of 30 to 44 mL/min/1.73 m2 and in the absence of active kidney disease (eg, a glomerulonephritis), some UpToDate authors and editors would not initiate metformin, whereas others would reduce the metformin dose by half (no more than 1000 mg per day) and increase the frequency of kidney function monitoring, although there are little or no data to support the glycemic efficacy [30,31] and safety of the latter approach. Lower doses of metformin may not produce the desired lowering of glycemia and may not be safer.
●For patients taking metformin whose eGFR falls below 45 mL/min/1.73 m2, we reduce the metformin dose by half (no more than 1000 mg per day) with patient education to stop metformin for dehydration, nausea, or vomiting and more frequent testing of eGFR, although there are few data to support the efficacy and safety of this approach.
●For patients taking metformin whose eGFR falls below 30 mL/min/1.73 m2, we discontinue metformin.
●We generally avoid metformin in patents with known or unknown conditions that cause acidosis.
●We do not view stable compensated heart failure as a contraindication to metformin use.
●We advise patients with an eGFR between 30 and 60 mL/min/1.73 m2 or stable heart failure to stop taking metformin if they develop acute hypoxemia or any condition associated with acute hypoxemia, dehydration, or sepsis (eg, influenza, urinary tract infection) until the condition has resolved.
●We do not view fatty liver disease generally as a contraindication unless there are major manifestations, such as reduced synthetic function or cirrhosis.
●We prefer to hold metformin in patients who are about to receive intravenous iodinated contrast material (with potential for contrast-induced renal failure) if they are at increased risk for lactic acidosis independent of metformin. Such patients include those with vascular instability, hypotension, and potential hypoperfusion. Metformin should not be restarted until eGFR can be retested (usually two to five days after the procedure) and confirmed to be >30 mL/min/1.73 m2.
In addition, some UpToDate authors and editors discontinue metformin prior to any radiologic procedures with intravenous or intra-arterial contrast in patients with eGFR <45 mL/min/1.73 m2, since these patients are at higher risk for metformin-induced lactic acidosis if acute kidney injury were to develop. Metformin may be restarted several days after the procedure if the patient is back to a usual diet and activity level. In patients at high risk for acute kidney injury, renal function should be reassessed prior to resuming metformin.
The relationship among metformin use, intravenous contrast administration, and the occurrence of lactic acidosis is not well studied. The rationale for stopping metformin prior to intravenous iodinated contrast is to avoid the potential for high plasma metformin concentrations (and lactic acidosis) if the patient develops contrast-induced acute renal failure. In general, mortality in reported cases of metformin-induced lactic acidosis may be as high as 50 percent [32].
In a systematic review of studies and evidence-based guidelines on the use of intravenous contrast in patients taking metformin, the only available data were from case reports and case series [33]. The majority of cases of metformin-related lactic acidosis occurred in patients with abnormal renal function who received intravenous contrast medium. The risk of metformin-induced lactic acidosis in patients with normal renal function who receive intravenous contrast is unknown but appears to be low.
The American College of Radiology suggests there is no need to discontinue metformin prior to or following the intravenous administration of iodinated contrast media in patients with no evidence of acute kidney injury and with eGFR ≥30 mL/min/1.73 m2 [34]. This recommendation was presumably predicated on the increasing uncertainty regarding the role of contrast-dye procedures in acute kidney injury [35] and the overall rarity of lactic acidosis in metformin-treated patients (see 'Lactic acidosis' below). Until more data are available, however, and taking into account the morbidity and mortality of metformin-associated lactic acidosis, we prefer to hold metformin in patients at increased risk for lactic acidosis (eg, vascular instability, hypotension, potential hypoperfusion) independent of metformin, as described in the bullet above. (See "Contrast-associated and contrast-induced acute kidney injury: Clinical features, diagnosis, and management".)
Pretreatment evaluation — Prior to starting metformin, we review serum creatinine (with eGFR), liver function tests, and A1C. The general evaluation for patients with newly diagnosed diabetes is reviewed separately. (See "Clinical presentation, diagnosis, and initial evaluation of diabetes mellitus in adults", section on 'Comprehensive history'.)
Dosing — For initial dosing of metformin, we prefer immediate-release rather than extended-release preparations. Metformin is absorbed rapidly from the small intestine, with peak plasma concentrations attained in two hours. It is not bound to plasma proteins, is not metabolized, and is rapidly excreted in the urine [36,37].
●Immediate-release metformin – Immediate-release metformin is available as 500, 850, or 1000 mg tablets and should be taken with meals (to reduce gastrointestinal side-effects) twice daily.
•For patients with eGFR ≥45 mL/min/1.73 m2, we begin with 500 mg once daily with the evening meal and, if tolerated, add a second 500 mg dose with breakfast. The dose can be increased slowly (one tablet every one to two weeks) until reaching the usual effective dose (1500 to 2000 mg/day). In one seminal study, the maximum recommended dose of 2550 mg/day (850 mg three times daily) provided only marginally better glycemic control than 1500 or 2000 mg and was associated with a higher incidence of gastrointestinal side effects [31]; however, based on individual glycemic response and side effects, the maximum recommended dose may be beneficial.
•For patients with eGFR of 30 to 44 mL/min/1.73 m2, we begin with 500 mg once daily with the evening meal and, if tolerated, add a second 500 mg dose with breakfast. We do not increase the dose further (ie, no more than 1000 mg per day), although there are little or no data to support the glycemic efficacy and safety of this approach. We educate patients regarding the risk of taking metformin if acute kidney injury develops and the importance of discontinuation in situations that may increase this risk. We typically increase the frequency of kidney function monitoring. (See 'Monitoring' below.)
•For patients with eGFR <30 mL/min/1.73 m2, metformin should not be prescribed. (See 'Contraindications' above.)
For patients who have gastrointestinal intolerance of metformin, slower titration, ensuring that the patient is taking the medication with food, or switching to an extended-release formulation may improve tolerability.
●Extended-release metformin – Extended-release metformin, typically taken once daily with the evening meal, is available in 500 and 750 mg tablets, as well as in an oral suspension (500 mg/5 mL). Patients may be switched from metformin to the once-daily, extended-release preparation at the same total daily dose, up to 2000 mg once daily. Since metformin should be taken with meals, the ability to administer the entire dose once daily may promote adherence [38].
Randomized comparisons of immediate-release and extended-release metformin have demonstrated that the preparations are equally effective, with extended-release metformin having similar or slightly improved gastrointestinal tolerability [39,40].
Monitoring — For patients taking metformin, we measure:
●A1C every three to six months.
●Serum creatinine annually.
●Vitamin B12 annually, particularly in individuals at risk for developing B12 deficiency due to decreased intake (eg, vegan diet) or absorption (eg, bariatric surgery). Metformin lowers serum vitamin B12 concentrations in 5 to 10 percent of patients who have been taking metformin for five years. (See 'Vitamin B12 deficiency' below.)
General monitoring for glycemic control and diabetes-related complications is reviewed separately (table 2). (See "Overview of general medical care in nonpregnant adults with diabetes mellitus", section on 'Blood glucose monitoring and target A1C' and "Overview of general medical care in nonpregnant adults with diabetes mellitus", section on 'Diabetes-related complications'.)
CLINICAL OUTCOMES
Glycemic efficacy — Metformin is considered initial pharmacologic therapy for most patients with type 2 diabetes because of glycemic efficacy, absence of weight gain and hypoglycemia, general tolerability, and favorable cost (table 1) [41]. In a now classic study, the United States Multicenter Metformin Study Group, for example, randomly assigned patients with type 2 diabetes and obesity who were inadequately controlled on diet alone to either metformin or placebo [42]. After 29 weeks, the mean A1C concentration was 7.1 percent in the metformin group as compared with 8.6 percent in the placebo group; these findings have been borne out in multiple subsequent trials.
The glycemic efficacy of metformin is reviewed in more detail elsewhere. (See "Initial management of hyperglycemia in adults with type 2 diabetes mellitus", section on 'Glycemic efficacy'.)
Weight loss — In individuals who are obese, metformin promotes modest weight reduction or at least weight stabilization (figure 1) [9,43]. This is in contrast to the weight gain often associated with insulin or sulfonylurea treatment [9,43]. In one large study, for example, patients treated with glyburide gained an average of 1.6 kg, whereas those receiving metformin lost 2.9 kg [44]. In a trial comparing metformin with a long-acting glucagon-like peptide-1 (GLP-1) receptor agonist, dulaglutide, weight loss at 52 weeks was similar in the two groups (-2.29 and -2.22 kg for dulaglutide 1.5 mg and metformin, respectively) [45].
Cardiovascular effects — Metformin does not have adverse cardiovascular effects, and it appears to decrease cardiovascular events in certain populations [41,46-51]. As examples:
●In the United Kingdom Prospective Diabetes Study (UKPDS), patients with type 2 diabetes and obesity who were assigned initially to receive metformin rather than sulfonylurea or insulin therapy had a decreased risk of the aggregate diabetes-related endpoint (endpoints included both macrovascular and microvascular complications) and all-cause mortality [48]. During the postinterventional observation period of the UKPDS, reductions in the risk of macrovascular complications were maintained in the metformin group [52]. (See "Glycemic control and vascular complications in type 2 diabetes mellitus", section on 'Intensive glycemic control'.)
●In another trial, 390 patients treated with insulin were randomly assigned to metformin versus placebo [46]. After four years, mean A1C (7.5 versus 7.9 percent) and body weight (85 versus 90 kg) were significantly lower in the metformin group. In addition, there was a decrease in the risk of the secondary macrovascular endpoint, which was a composite of 13 vascular events including myocardial infarction (MI), heart failure, stroke, amputation, and sudden death (event rates 15 versus 18 percent, adjusted hazard ratio [HR] 0.6, 95% CI 0.4-0.9).
In a subsequent meta-analysis of 170 trials and 25 observational studies evaluating the effects of oral or injectable diabetes medications as monotherapy and in combination with other oral agents or insulin on cardiovascular mortality, intermediate outcomes (A1C, body weight, lipid profiles), and adverse events, metformin was associated with lower long-term cardiovascular mortality compared with sulfonylurea monotherapy (based upon findings from two randomized trials and three observational studies) [41].
In one of the trials included in the meta-analysis, 304 Chinese patients with established coronary heart disease and type 2 diabetes were randomly assigned to metformin versus glipizide [47]. Lifestyle intervention and other treatment for coronary heart disease were similar in both groups. After three years, the mean achieved A1C level was similar (7 and 7.1 percent) in the two groups. However, body weight, waist circumference, and body mass index (BMI) were significantly lower in the metformin group. A similar proportion of patients in each group received insulin (30 and 25 patients in metformin and glipizide groups, respectively). After a median follow-up of five years, there were fewer cardiovascular events (composite of nonfatal MI [5 versus 6], stroke [10 versus 15], arterial revascularization [21 versus 25], or death from cardiovascular or any cause [7 versus 14]) in the metformin group (total events 43 versus 60; HR 0.54, 95% CI 0.3-0.9). The main limitation of this trial was the small number of events. However, the results support the use of metformin, particularly in patients with coronary heart disease.
Metformin has lipid-lowering activity, resulting in a decrease in serum triglyceride and free fatty acid concentrations, a small decrease in serum low-density lipoprotein (LDL) cholesterol concentrations, and a very modest increase in serum high-density lipoprotein (HDL) cholesterol concentrations [10,36,53].
Cancer incidence — Observational data suggest that use of metformin decreases cancer incidence [54-56]. In meta-analyses of predominantly case-control and cohort studies in patients with type 2 diabetes, use of metformin compared with nonuse or with use of other diabetes treatment was associated with a reduced risk of all cancers (relative risk [RR] 0.61, 95% CI 0.54-0.70) [57,58], colorectal cancer (RR 0.64, 95% CI 0.54-0.76) [57-59], and lower cancer mortality (RR 0.66, 95% CI 0.49-0.88) [57,60,61]. Among the meta-analyses, the summary effect estimates were similar. With the exception of colorectal cancer, there was significant heterogeneity among the individual studies. Allocation and time-lag bias (healthier and younger individuals with earlier-stage diabetes likelier to be using metformin) complicates interpretation of these findings.
In contrast to the observational data, a meta-analysis of randomized trials comparing metformin with a comparator (thiazolidinedione [TZD], sulfonylurea, dipeptidyl peptidase-4 [DPP-4] inhibitor, or placebo) did not show a reduction in cancer incidence [62]. The majority of the trials were not designed to explore cancer outcomes, which were not uniformly collected or adjudicated; therefore, malignancies were noted as serious adverse events. In addition, average follow-up for cancer outcomes was only four years. A longer interval may be required to adequately assess cancer outcomes. Thus, prospective clinical trial data are required to confirm or refute this protective effect.
A possible mechanism by which metformin may decrease cancer incidence is regulation of AMP-activated protein kinase (AMPK) through LKB1 [19]. LKB1 is a tumor suppressor, and activation of AMPK through LKB1 may play a role in inhibiting cell growth. Studies in Caenorhabditis elegans have suggested that inactivation of mTORC1 with subsequent inhibition of growth through induction of ACAD10 may explain the anticancer effects of metformin [63]. (See 'Mechanism of action' above.)
ADVERSE EFFECTS
Gastrointestinal — The most common side effects of metformin are gastrointestinal, including a metallic taste in the mouth, mild anorexia, nausea, abdominal discomfort, and soft bowel movements or diarrhea [36]. These symptoms are usually mild, transient, and reversible after dose reduction or discontinuation of the drug. They are minimized by taking the medication with food. In clinical trials, only 5 percent of study subjects discontinue metformin because of the gastrointestinal side effects.
Although metformin is generally well tolerated, patients taking metformin may develop gastrointestinal side effects even after many years of use. If gastrointestinal symptoms develop while taking metformin, we institute a metformin holiday, which may lead to resolution of symptoms. After a period of non-use, metformin may be successfully resumed at the same or a lower dose with a slow titration of the immediate-release or extended-release formulation.
Vitamin B12 deficiency — Metformin reduces intestinal absorption of vitamin B12 in up to 30 percent of patients and lowers serum vitamin B12 concentrations in 5 to 10 percent, but it only rarely causes megaloblastic anemia (possibly due to folic acid supplementation of the United States food supply) [64,65]. In some patients, vitamin B12 deficiency may present as peripheral neuropathy [66]. The dose and duration of use of metformin correlates with the risk of vitamin B12 deficiency [65,67]. In one study, the reduction in serum vitamin B12 appeared to be due to poor absorption of B12 in the ileum and was corrected by administration of oral calcium carbonate (1.2 g daily) [68]. In another study, supplementation with a daily multivitamin was associated with a lower prevalence of vitamin B12 deficiency [69].
Owing to data that suggest vitamin B12 deficiency is often asymptomatic and anemia is not a sensitive indicator, as well as a prevalence of vitamin B12 deficiency (or borderline low levels) in metformin-treated patients that may approach 20 percent over five years [65], routine B12 monitoring or administration may be considered in patients with poor dietary intake or absorption. (See 'Monitoring' above.)
Lactic acidosis
●Incidence – The incidence of lactic acidosis in metformin users appears to be very low [70,71]. Despite its rarity, lactic acidosis related to metformin remains a concern because of the high case-fatality rate.
In a systematic review of 347 randomized trials and prospective cohort studies representing 70,490 patient-years of metformin use and 55,451 patient-years in the comparator group, there were no cases of lactic acidosis [70]. Almost one-half of the studies allowed inclusion of patients with a serum creatinine above 1.5 mg/dL (133 micromol/L), and almost all allowed inclusion of patients with at least one standard contraindication to metformin therapy. However, the number of patients who actually had these contraindications was not presented, and therefore, the safety of metformin in the presence of standard contraindications could not be assessed.
In a subsequent retrospective cohort study evaluating the risk of hospitalization for acidosis in metformin users across the full spectrum of renal function, risk of acidosis was similar in metformin users and nonusers when estimated glomerular filtration rate (eGFR) was ≥30 mL/min/1.73 m2 [71]. Lower eGFR was associated with a higher risk of acidosis in both metformin users and nonusers; however, there was an increased risk of acidosis in metformin users when eGFR was <30 mL/min/1.73 m2. (See 'Contraindications' above.)
Metformin-induced lactic acidosis can occur in patients with normal renal and hepatic function in the following clinical settings:
•Purposeful metformin overdose. (See "Metformin poisoning".)
•The genetic diabetes syndrome known as maternally inherited diabetes and deafness (MIDD), in which individuals are at increased risk of developing lactic acidosis with metformin therapy. (See "Classification of diabetes mellitus and genetic diabetic syndromes", section on 'Genetic defects in mitochondrial DNA'.)
•Intercurrent acute kidney injury from other causes. The presence of nausea, vomiting, or dehydration should prompt metformin discontinuation. (See "Causes of lactic acidosis".)
●Predisposing factors – More serious lactic acid accumulation occurs in patients with conditions that predispose to hypoperfusion and hypoxemia (acute or progressive renal impairment, acute or progressive heart failure, acute pulmonary decompensation, sepsis, dehydration) [36,72,73]. This finding has resulted in the development of standard contraindications to metformin, including significantly impaired renal function, heart failure, liver disease, and excessive alcohol intake. (See 'Contraindications' above.)
●Treatment – The treatment of metformin-induced lactic acidosis is discussed elsewhere. (See "Metformin poisoning".)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Diabetes mellitus in adults".)
SUMMARY AND RECOMMENDATIONS
●Initial therapy – In the absence of specific contraindications, we suggest metformin as initial therapy in most patients with type 2 diabetes (Grade 2B) (table 1). Metformin is considered initial pharmacologic therapy over other diabetes drugs because of glycemic efficacy, absence of weight gain and hypoglycemia, general tolerability, and favorable cost. It can be initiated at the time of diabetes diagnosis, along with consultation for lifestyle intervention. For highly motivated patients with glycated hemoglobin (A1C) near target (<7.5 percent), a three- to six-month trial of lifestyle modification before initiating metformin is reasonable. (See 'Clinical outcomes' above and "Initial management of hyperglycemia in adults with type 2 diabetes mellitus", section on 'Choice of initial therapy' and "Initial management of hyperglycemia in adults with type 2 diabetes mellitus", section on 'When to start'.)
●Contraindications – Metformin is contraindicated in patients with factors predisposing to lactic acidosis, including estimated glomerular filtration rate (eGFR) <30 mL/min/1.73 m2. Alternative initial treatment for patients with contraindications or intolerance to metformin is reviewed separately. (See 'Contraindications' above and "Initial management of hyperglycemia in adults with type 2 diabetes mellitus", section on 'Contraindications to or intolerance of metformin'.)
●Dosing
•eGFR ≥45 mL/min/1.73 m2 – For patients with an eGFR ≥45 mL/min/1.73 m2, we begin with 500 mg once daily with the evening meal and, if tolerated, add a second 500 mg dose with breakfast. The dose can be increased slowly (one tablet every one to two weeks) until reaching the usual effective dose (1500 to 2000 mg/day). (See 'Dosing' above.)
•eGFR of 30 to 44 mL/min/1.73 m2 – For patients with an eGFR of 30 to 44 mL/min/1.73 m2, we typically reduce the metformin dose by half (ie, no more than 1000 mg per day). We educate patients regarding the risk of taking metformin if acute kidney injury develops and the importance of discontinuation in situations that may increase this risk. We typically increase the frequency of kidney function monitoring. (See 'Dosing' above.)
●Monitoring – For patients taking metformin, we measure A1C every three to six months; serum creatinine annually; and vitamin B12 annually, particularly in patients at risk for vitamin B12 deficiency due to decreased intake (eg, vegan diet) or absorption (eg, bariatric surgery). (See 'Monitoring' above.)
●Adverse effects – The most common side effects of metformin are gastrointestinal, including a metallic taste in the mouth, mild anorexia, nausea, abdominal discomfort, and soft bowel movements or diarrhea. (See 'Adverse effects' above.)
Lactic acidosis is an extremely uncommon side effect. However, it remains a concern because of the high case-fatality rate. Most cases have occurred in patients with conditions that predispose to hypoperfusion and hypoxemia (acute or progressive renal impairment, acute or progressive heart failure, acute pulmonary decompensation, sepsis, dehydration). (See 'Lactic acidosis' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges David McCulloch, MD, who contributed to an earlier version of this topic review.
