Safety and efficacy of metformin for the treatment of pain: protocol for a systematic review

Abstract

Introduction

Pain remains difficult to manage due to limited efficacy and adverse effects of many existing analgesics, highlighting the need for safer non-opioid treatment strategies. Metformin is a widely prescribed oral antihyperglycaemic agent with emerging preclinical and early clinical evidence suggesting potential analgesic effects. Proposed mechanisms include activation of AMP-activated protein kinase with downstream inhibition of mammalian target of rapamycin and mitogen-activated protein kinase signalling pathways implicated in nociceptive sensitisation and chronic pain. This protocol describes a systematic review to evaluate the efficacy and safety of metformin for the treatment or prevention of pain in an adult population.

Methods and analysis

This systematic review will include randomised, double-blind, placebo or active-controlled trials evaluating orally administered metformin for the treatment or prevention of acute or chronic pain in adult human participants. Searches will be conducted in MEDLINE (Ovid MEDLINE ALL), EMBASE and the Cochrane Central Register of Controlled Trials from inception alongside ClinicalTrials.gov, the WHO International Clinical Trials Registry Platform and reference list screening. Two reviewers will independently screen studies, extract data and assess risk of bias using the Cochrane Risk of Bias 2.0 tool. Outcomes will include pain intensity, pain relief, responder outcomes, functional interference, analgesic use and adverse events. Where appropriate, random-effects meta-analyses will be performed. Certainty of evidence will be evaluated using the Grading of Recommendations Assessment, Development and Evaluation approach.

Ethics and dissemination

Ethics approval is not required as this study will synthesise published data. Findings will be disseminated through peer-reviewed publication and presentation in scientific conferences.

PROSPERO registration number

CRD420261296816.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The review methods were established in advance through protocol registration which includes predefined eligibility criteria, outcomes and planned approach to evidence synthesis.

• A comprehensive search strategy across multiple databases and trial registries will be employed.

• Risk of bias will be assessed using the Cochrane Risk of Bias 2.0 tool, and certainty of evidence will be evaluated using the Grading of Recommendations Assessment, Development and Evaluation approach.

• Inclusion will not be restricted to English-language studies to prevent language bias.

• Anticipated heterogeneity in pain conditions, metformin dosing and outcome measures may limit quantitative synthesis, thus structured narrative synthesis will be used where meta-analysis is inappropriate.

Introduction

Pain is a highly prevalent and debilitating condition affecting an estimated 20% of adults globally and is associated with substantial personal, societal and economic burden.¹ Despite extensive research and the availability of multiple pharmacological therapies, many existing analgesics, particularly opioids, offer questionable long-term benefit for chronic pain while carrying a measurable risk of addiction.² Rates of carefully diagnosed opioid addiction among patients treated for pain average up to 8% with substantially higher rates of misuse and aberrant use reported to be 15 - 26%.^{3 4} These limitations, including tolerance, dependence and addiction risk, underscore the need for safer and more effective non-opioid treatment strategies.

Metformin is a first-line oral antihyperglycaemic agent with an established safety profile and widespread global use.⁵ Metformin has also been investigated for therapeutic effects beyond glycaemic control, with evidence suggesting activity across metabolic, inflammatory and cellular signalling pathways relevant to disease modification.⁶ Beyond its metabolic effects, metformin has demonstrated analgesic properties in preclinical models of inflammatory and neuropathic pain. These effects are thought to be mediated through indirect activation of AMP-activated protein kinase. This leads to downstream inhibition of mammalian target of rapamycin (mTOR) and mitogen-activated protein kinase signalling pathways involved in nociceptive sensitisation, neuroinflammation and maladaptive plasticity associated with chronic pain states.^{7 8}

Emerging clinical trials have begun to evaluate metformin in a variety of pain conditions, including osteoarthritis, neuropathic pain, postoperative pain, fibromyalgia and other musculoskeletal disorders.8,13 Evidence that metformin may have analgesic effects for osteoarthritis has recently been reviewed;¹⁴ however, the goal of the current review is to comprehensively synthesise evidence of metformin analgesia across all painful conditions that have been studied. A rigorous synthesis of randomised controlled trial evidence across multiple pain disorders is needed to clarify the potential role of metformin as a repurposed analgesic therapy and to guide future clinical research. This review aims to address this objective.

Objectives

The objective of this systematic review is to evaluate the efficacy and safety of metformin for the treatment or prevention of acute and chronic pain. Using the PICOTS framework: Population: adult human participants (≥18 years) experiencing acute or chronic pain of any aetiology. Intervention: orally administered metformin at any dose, formulation, frequency and duration. Comparator: placebo, usual care, standard analgesic therapy or active comparators. Outcomes: primary outcomes are participant-reported pain intensity and pain relief; secondary outcomes include functional interference, analgesic consumption, adverse events, emotional function and other pain-related measures. Time: short-term (<1 week), medium-term (1 week to 3 months) and long-term (>3 months) follow-up. Study design: randomised controlled trials.

Methods and analysis

This protocol has been registered with the International Prospective Register of Systematic Reviews (PROSPERO: CRD420261296816). This protocol is reported in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) guidelines.¹⁵ The review will be conducted and reported in accordance with the PRISMA 2020 statement.¹⁶

Sources of evidence

We will conduct a comprehensive search of the Cochrane Central Register of Controlled Trials, MEDLINE (Ovid MEDLINE ALL) and EMBASE from their inception to the date the final searches are run. The search strategy will include terms relating to metformin and other biguanides, as well as pain-related conditions and outcomes, including acute and chronic pain, neuropathic pain, inflammatory pain, musculoskeletal pain and postoperative pain. Search strategies will be developed in consultation with an experienced health sciences librarian specialising in systematic review methodology. The full search strategies are reported in online supplemental appendix 1.

In addition, we will screen the reference lists of all included randomised controlled trials and relevant review articles for additional eligible studies. To identify ongoing or unpublished trials, we will search ClinicalTrials.gov and the WHO International Clinical Trials Registry Platform (WHO ICTRP).

Report selection

Types of studies

We will include randomised, double-blind, placebo-controlled or active-controlled clinical trials evaluating the efficacy and/or safety of metformin for pain-related outcomes. We will exclude non-randomised studies, observational designs, case reports, case series, conference abstracts without sufficient data, study protocols without results and preclinical or animal studies. Double-blind designs are required given the subjective nature of pain outcomes and their susceptibility to expectation and placebo effects. Non-blinded trials are excluded as they carry high risk of performance and detection bias for subjective endpoints. The blinding status of all included trials will be assessed using the Cochrane Risk of Bias 2.0¹⁷ tool and subsequently reported.

Types of participants

Eligible studies will include adult human participants (≥18 years) experiencing acute or chronic pain of any aetiology. The restriction to adults (≥18 years) reflects the established evidence base, which derives predominantly from adult populations. Although metformin holds regulatory approval for paediatric use in multiple jurisdictions, paediatric pain trials are anticipated to be few and identifying this gap is an important direction for future systematic reviews. Participants must have pain assessed using validated patient-reported outcome measures such as the visual analogue scale (VAS), Numeric Rating Scale (NRS) or other comparable tools. Trials in which metformin is administered for non-pain indications (eg, diabetes, obesity) will be included if prespecified pain-related outcomes are reported. Pain outcomes will be considered prespecified if they are explicitly designated as primary or secondary outcomes in the trial registration record, published protocol or methods section of the trial report before unblinding.

Types of interventions

Eligible interventions will include orally administered metformin (any oral formulation) at any dose, frequency and duration. Metformin may be administered for the treatment of an established pain condition or for the prevention or reduction of anticipated pain (eg, perioperative use). Trials in which metformin is used as part of combination therapy will be eligible if cointerventions are balanced across study arms, defined as the same analgesic comedication given at equivalent doses, frequencies and duration, so that observed differences in pain-related outcomes can be attributed to metformin.

Comparators

Eligible comparators will include placebo, usual care, standard pain management or other active comparator interventions, provided the study design permits attribution of observed effects to metformin.

Data collection, extraction and management

Two reviewers will independently screen eligible studies and extract data using a standardised data extraction form. Extracted data will include study design, pain condition, participant characteristics, eligibility criteria, sample size, intervention, comparator details, including metformin dose, formulation, frequency and duration, cointerventions, outcome measures, follow-up durations and adverse events.

Data extractions will be performed independently by both reviewers, with discrepancies resolved through discussion and consensus. If necessary, a third reviewer will be consulted.

Types of outcome measures

We will include studies reporting validated participant-reported measures of pain intensity or pain relief (eg, VAS or NRS). Additional outcomes may include functional measures, opioid consumption and adverse events, including serious adverse events related to metformin.

Primary outcomes

Primary outcomes will include participant reported pain intensity or pain relief, measured using validated instruments such as NRS, VAS. Where available, responder outcome, ≥30% or ≥50% pain reduction, consistent with Initiative on Methods, Measurement and Pain Assessment in Clinical Trials recommendations will be included.¹⁸ Outcomes will be assessed and reported at all available time points, categorised as: short term (up to 1 week postintervention), medium-term (greater than 1 week to 3 months postintervention) and long-term (greater than 3 months postintervention). Where multiple assessments are reported within a time category, the assessment closest to the end of the defined period will be used for pooling.

Secondary outcomes

1. Other pain-related outcomes (eg, physical function, sleep, interference with activities).

2. Emotional function and pain-related comorbidities (eg, measures of anxiety or depression).

3. Analgesic consumption (eg, concomitant rescue medication use, opioid-sparing effects).

4. Withdrawals due to lack of efficacy, adverse events or any cause.

5. Specific adverse events of interest (eg, gastrointestinal side effects, hypoglycaemia, lactic acidosis, weight change).

6. Neuropathy and sensory outcomes (eg, changes in temperature sensation, altered light touch, nerve conduction velocity or other neurophysiological measures).

Adverse event severity will be extracted according to the grading system used in each original trial. For the trials that use a standardised system, such as the Common Terminology Criteria for Adverse Events, these grades will be reported; otherwise, severity classification will be summarised as described by the trial author.

Search methods for identification of studies

Electronic searches

No language restrictions are applied. The full details of the search strategies are reported in online supplemental appendix 1.

Searching other sources

ClinicalTrials.gov and the WHO ICTRP will be searched to identify ongoing or unpublished trials. Reference lists of included studies and relevant reviews will also be screened.

Data collection and analysis

Selection of studies

Search results will be exported to the Covidence screening tool and duplicates will be identified and removed. Two reviewers (NZ and CH) will independently screen titles and abstracts, followed by full-text review of potentially eligible studies. At the full-text stage, reasons for exclusion will be recorded in Covidence and reported in the PRISMA flow diagram. Disagreements at any stage will be resolved through discussion and if consensus is not reached, a third reviewer (IG) will mediate. Study selection will be summarised using a PRISMA flow diagram.

Data extraction and management

Two reviewers will independently extract data using a standardised form, capturing information about the chronic/acute pain condition, number of participants treated, participant characteristics, type of metformin used, other study drugs used, dose and frequency and route of administration of metformin and other study drugs, study duration and follow-up, study design, outcome measures of interest and results, and adverse events. Disagreements at any stage will be resolved by discussion and, if necessary, consultation with a third reviewer.

Risk of bias assessment in included studies

Risk of bias will be assessed independently by two reviewers using the Cochrane Risk of Bias 2.0¹⁷ tool at the outcome level. Each domain and overall risk of bias will be judged as low risk, some concerns or high risk, with disagreements resolved by consensus or a third reviewer. Risk of bias will be assessed at the outcome level across the following domains:

1. Bias arising from the randomisation process.

2. Bias due to deviation from intended interventions.

3. Bias due to missing outcome data.

4. Bias in measurement of the outcome.

5. Bias in selection of the reported outcome.

Measures of treatment effect

For dichotomous outcomes such as responder outcomes, adverse events, treatment effects will be expressed as risk ratios (RRs) with 95% CIs. Where appropriate, absolute risk differences will also be calculated. For continuous outcomes (eg, pain intensity scores), mean differences (MDs) will be used when outcomes are measured on the same scale, and standardised MDs (SMDs) when different validated scales are used. All effect estimates will be reported with 95% CIs.

Dealing with missing data

Where possible, analyses will be conducted using an intention-to-treat approach, including all randomised participants who received at least one dose of study medication and provided at least one post-baseline assessment. The handling of missing data in individual trials will be documented. We will assess and report methods used to address missing outcome data due to participant withdrawal, given their potential to influence effect estimates.

Data synthesis

When at least three trials are sufficiently similar in terms of population, intervention, comparator and outcomes, random-effects meta-analysis will be performed. Dichotomous outcomes will be pooled using RRs, and continuous outcomes using MDs or SMDs, as appropriate. Statistical heterogeneity will be assessed using the I² statistic and visual inspection of forest plots. I² values of 0-40% will be considered not important, 30-60% moderate, 50-90% substantial and 75-100% considerable heterogeneity. Where substantial or considerable heterogeneity is identified (I²≥50%), we will explore potential explanations through predefined subgroup analyses before pooling and will consider whether quantitative synthesis remains appropriate. The χ² test (p<0.10) will be used as a supplementary indicator of significant heterogeneity.

If quantitative synthesis is not appropriate due to limited data or substantial clinical or methodological heterogeneity, findings will be summarised using a structured narrative synthesis, grouping studies by pain condition, metformin regimen and comparator type. Narrative synthesis will be the primary approach when clinical or methodological heterogeneity precludes scientifically justified pooling.

Anticipated challenges

Given the diversity of pain conditions and trial designs in this area, we anticipate variability in sample size, populations, interventions and outcomes. To address this, studies will be grouped according to clinically relevant characteristics, including pain type and treatment duration before synthesis. Meta-analysis will be done when studies are appropriately comparable, while structured narrative synthesis will be used when clinical or methodological differences make pooling unfeasible. Small trial sizes and imprecision will be considered when interpreting effect estimates and assessing certainty of evidence using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework.¹⁹

Subgroup and sensitivity analyses

Where sufficient data are available, exploratory subgroup analyses will be conducted to investigate potential sources of heterogeneity. Planned subgroups include pain type such as neuropathic, musculoskeletal, postoperative, pain duration (acute vs chronic), metformin dose, treatment duration, comparator type and presence of diabetes or obesity.

Sensitivity analyses will be performed by excluding studies judged to be at high risk of bias to assess the robustness of the findings.

Assessment of meta-biases

Where sufficient studies (≥10) are available for a given outcome, publication bias will be assessed using funnel plots and, where appropriate, Egger’s regression test. Selective outcome reporting within studies will be considered as part of the Risk of Bias 2.0 assessment.¹⁷

Sample size consideration

As this is a systematic review, no formal sample size calculation is applicable; the sample size will be determined by the number and size of eligible randomised controlled trials identified through the comprehensive literature search. Where meta-analysis is feasible, the precision of pooled estimates will be reflected by 95% CIs, and the certainty of evidence will be assessed using the GRADE framework.¹⁹

Quality of evidence

Evidence certainty will be evaluated using the GRADE framework¹⁹ and reported in a summary of findings table.

Progress

The protocol has been registered in the PROSPERO review registry (CRD420261296816). Database searches have been completed. Study screening is completed. Data extraction and analysis are expected to be completed by September 2026.

Patient and public involvement

Patients and the public were not involved in the design, conduct, reporting or dissemination plans of this research.

Ethics and dissemination

Formal ethical approval is not required, as this study involves secondary analysis of published data. The findings of this systematic review will be disseminated through publication in a peer-reviewed journal and presentation at scientific conferences.