What are the benefits and risks of using metformin in the treatment of chronic kidney disease?

Key messages

- This study did not find enough evidence to show that metformin had any effect on kidney function. Our findings may have been limited by the fact that most of the studies we looked at were conducted in two specific patient populations (those with cystic kidney disease or diabetes) and often did not fully report effects on kidney function.

- Further evidence in the form of large, well-designed randomised trials is needed to more robustly assess whether metformin can be a long-term protective treatment in those with chronic kidney disease.

Why use metformin to prevent the progression of chronic kidney disease?

Chronic kidney disease reduces the kidneys' ability to remove waste products from the body. Some evidence suggests that metformin, a medication traditionally used for diabetes, may help slow the progression of kidney disease and maintain kidney function.

What did we want to find out?

We wanted to find out if metformin improved or maintained kidney function, reduced the chance of kidney failure, or reduced the chance of people dying. We also wanted to find out if metformin was associated with any unwanted side effects in people with kidney disease.

What did we do?

We searched for studies that investigated metformin use compared with placebo or other medications in people with chronic kidney disease with or without diabetes and measured kidney function. We compared and summarised the studies' results and rated our confidence in the evidence-based on factors such as study methods, clarity, transparency, and the number of included patients.

What did we find?

We found eleven studies that included 8449 participants. Four studies included people with cystic kidney diseases, and seven studies included people with diabetes. After reviewing the evidence, it remained uncertain whether metformin improves or maintains kidney function or whether it reduces the long-term chance of developing kidney failure or death. Metformin usage increased the likelihood of known side effects such as gastrointestinal disturbances.

What are the limitations of the evidence?

For most of the kidney-related and long-term outcomes we assessed, the results did not provide any degree of certainty with regard to the potential effects of metformin. We found that the quality of the studies was highly variable. Some were reported only as brief abstract summaries providing few details on the study methods and results, while others were deemed to be of low quality due to poor study design, small number of patients, and bias due to discrepancies between pre-defined methods and what was eventually published. Only one out of 11 studies were deemed to be of a high research quality.

How up to date is this evidence?

The evidence is up to date to 19 July 2023.

Authors' conclusions: 

This review highlights the lack of RCTs reporting on the effects of metformin on kidney function, particularly in patients with CKD. Future research in this field requires adequately powered RCTs comparing metformin to placebo or standard care in those with CKD. Seven ongoing studies were identified in this review, and future updates, including their findings, may further inform the results of this review.

Read the full abstract...

Metformin has been used in the management of diabetes for decades. It is an effective, low-cost intervention with a well-established safety profile. Emerging evidence suggests that metformin targets a number of pathways that lead to chronic kidney damage, and long-term use may, therefore, slow the rate of kidney function decline and chronic kidney disease (CKD) progression.


To evaluate the effect of metformin therapy on kidney function decline in patients with CKD with or without diabetes mellitus and assess the safety and dose tolerability in this population.

Search strategy: 

We searched the Cochrane Kidney and Transplant Register of Studies up to 19 July 2023 with assistance from an Information Specialist using search terms relevant to this review. Studies in the Register are identified through searches of CENTRAL, MEDLINE, and EMBASE, conference proceedings, the International Clinical Trials Registry Platform (ICTRP) Search Portal and ClinicalTrials.gov.

Selection criteria: 

We included randomised controlled trials (RCTs) that reported kidney-related outcomes with a minimum duration of 12 months delivery of the metformin intervention and whose eligibility criteria included adult participants with either i) a diagnosis of CKD of any aetiology and/or ii) those with a diagnosis of diabetes mellitus. Comparisons included placebo, no intervention, non-pharmacological interventions, other antidiabetic medications or any other active control.

Studies that included patients on any modality of kidney replacement therapy were excluded.

Data collection and analysis: 

Two authors independently carried out data extraction using a standard data extraction form. The methodological quality of the included studies was assessed using the Cochrane risk of bias tool. Summary estimates of effect were obtained using a random-effects model, and results were expressed as risk ratios (RR) and their 95% confidence intervals (CI) for dichotomous outcomes and mean difference (MD) and 95% CI for continuous outcomes. Confidence in the evidence was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach.

Main results: 

This review included 11 studies reporting on 8449 randomised participants. Studies were conducted in patient populations with Autosomal Dominant Polycystic Kidney Disease (ADPKD) (four studies) or diabetes mellitus (seven studies). Six studies compared metformin with no active control, four studies compared metformin with active controls (rosiglitazone, glyburide, pioglitazone, or glipizide), and one study included treatment arms that randomised to either metformin, diet and lifestyle modifications, or other antidiabetic therapies.

The risk of bias in included studies varied; two studies were abstract-only publications and were judged to have a high risk of bias in most domains. Other included publications were judged to have a low risk of bias in most domains. Across comparisons, GRADE evaluations for most outcomes were judged as low or very low certainty, except for those relating to side effects, tolerance, and withdrawals, which were judged as moderate certainty.

The evidence suggests that compared to placebo, metformin may result in i) a slightly smaller decline in kidney function (3 studies, 505 participants: MD 1.92 mL/min, 95% CI 0.33 to 3.51; I2 = 0%; low certainty), ii) very uncertain effects on the incidence of kidney failure (1 study, 753 participants: RR 1.20, 95% CI 0.17 to 8.49), iii) little or no effect on death (3 studies, 865 participants: RR 1.00, 95% CI 0.76 to 1.32; I2 = 0%; moderate certainty), iv) little or no effect on the incidence of serious adverse events (3 studies, 576 participants: RR 1.15, 95% CI 0.76 to 1.72; I2 = 0%; moderate certainty), and v) likely higher incidence of intolerance leading to study withdrawal than placebo (4 studies, 646 participants: RR 2.19, 95% CI 1.46 to 3.27; I2 = 0%; moderate certainty). The certainty of the evidence for proteinuria was very uncertain.

Compared to other active controls (rosiglitazone, glyburide, pioglitazone, or glipizide), metformin i) demonstrated very uncertain effects on kidney function decline, ii) may result in little or no difference in death (3 studies, 5608 participants: RR 0.95 95% CI 0.63 to 1.43; I2 = 0%; low certainty), iii) probably results in little or no difference in intolerance leading to study withdrawal (3 studies, 5593 participants: RR 0.92, 95% CI, 0.79 to 1.08; I2 = 0%; moderate certainty), iv) probably results in little or no difference in the incidence of serious adverse events (2 studies, 5545 participants: RR 1.16, 95% CI 0.79 to 1.71; I2 = 0%; moderate certainty), and v) may increase the urinary albumin-creatinine ratio (2 studies, 3836 participants: MD 14.61, 95% CI 8.17 to 21.05; I2 = 0%; low certainty). No studies reported the incidence of kidney failure.