{"doi":"10.1093/cid/ciaf700","title":"Preventing Long COVID With Metformin","abstract":"(See the Major Article by Chaichana et al on pages e423–32.) Chaichana et al conducted a sequential trial emulation to assess whether starting metformin after severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection prevented the development of post-coronavirus 19 (COVID 19) condition or long COVID [1]. Their work provides critical validation of previously published randomized clinical trial data and is consistent with emerging data. We review this body of literature and why metformin should now be offered to outpatient adults for treating acute SARS-CoV-2 infection to prevent long COVID [2]. Repurposing a chronic diabetes medication for acute viral infection may seem counterintuitive, so we will briefly review the rationale. Early papers on biguanides were for use against viruses and malaria [3–5]. In about 1940, biguanides fell out of favor because of lactic acidosis with phenformin and buformin [3]. Metformin was less potent for lowering glucose but had fewer safety concerns and was approved for treating type 2 diabetes [3]. This led to studies of metformin's anti-inflammatory actions and cohort studies in which there was no increased risk of lactic acidosis with metformin [6–9]. In the 2000's, metformin was studied in vitro against viruses because of its host-directed, immunometabolic actions [10]. In 2020, observational, in silico computer modeling and in vitro studies of SARS-CoV-2 added to the rationale for studying metformin as acute SARS-CoV-2 treatment [10–13]. Given these multiple streams of evidence, metformin's low cost, wide availability, tolerability, and safety with no need for monitoring during short-term use, it was important to test metformin versus placebo for outpatient treatment of SARS-CoV-2 [14]. In July 2021, we added a long-term clinical outcome to the COVID-OUT randomized trial protocol and consent to assess whether treatment during acute infection prevented long-term sequelae. At that time, it was unclear which specific symptoms constituted long-term sequelae, or what cutoffs to use for each symptom's duration, severity, or frequency. Symptom-based outcomes could also not be verified in electronic health records, and we wanted an outcome that could be verified in other sources of data. Thus, to best capture the definition of long COVID in clinical practice, and to still capture the outcome if that definition evolved during the course of the trial, we decided to ask participants: “Have you been told by a medical provider that you have long COVID?” This way of ascertaining long COVID allowed us to obtain the source medical records to confirm the participant-reported diagnosis. These long COVID diagnoses were made by clinicians in the community who were blinded, not involved in the study, and who were using the resources available for diagnosing this new disease while simultaneously ruling out other issues. In the COVID-OUT trial, the metformin group had a 41% lower risk of long COVID over 10 months of follow-up [15]. We need to clarify the first line of the abstract by Chaichana et al—the authors state “A subgroup analysis of the COVID-OUT trial's long-term outcome found that starting metformin within 3 days of COVID-19 reduced PCC incidence by 63%.” This could imply that the effect was observed in only a subgroup and not in the full sample. However, the 41% lower risk was for the full sample, hazard ratio 0.59 (95% confidence interval [CI] .39–.89). The 63% lower risk they quoted, hazard ratio 0.37 (95% CI .15–.95), was among the subgroup who started study drug within 3 days of COVID-19 symptom onset. A small mechanistic randomized trial testing metformin versus placebo on viral load found metformin reduced SARS-CoV-2 by 93.2%, compared to 78.3% with placebo (P = .013); and the time to an undetectable viral load was 3.3 days for metformin and 5.6 days for placebo (P = .043) [16]. Metformin also reduced viral load relative to placebo (−0.56 log10 copies/mL) in COVID-OUT [17]. Given these","journal":"Clinical Infectious Diseases","year":2025,"id":527948,"datarank":0.0,"base_score":0.0,"endowment":0.0,"self_citation_contribution":0.0,"citation_network_contribution":0.0,"self_endowment_contribution":0.0,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":3,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":0.9531,"is_data_producer":false,"deposit_databanks":null,"is_oa":true,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":"2025-01-01","fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":1405973,"name":"David R Boulware","orcid":null,"position":1,"is_corresponding":false},{"id":1405972,"name":"Carolyn T Bramante","orcid":null,"position":0,"is_corresponding":true}],"reference_count":20,"raw_metadata":null,"created_at":"2026-07-19T02:50:44.062153Z","pmid":"41608911","pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}