Back

Increased Insulin Resistance and Hyperglycemia in Long COVID disease: A Systematic Review and Meta-Analysis

Almulla, A. F.; Zhang, Y.; Tunvirachaisakul, C.; Carvalho, A. F.; Maes, M.

2025-11-02 infectious diseases
10.1101/2025.11.01.25339290 medRxiv
Show abstract

Backgroundaccumulating evidence suggests that Long COVID (LC) is mediated by chronic immune activation, oxidative stress, and metabolic dysregulation. These processes may impair glucose homeostasis and promote insulin resistance (IR). However, no prior meta-analysis has systematically and quantitatively evaluated IR indices and related biomarkers in LC compared with normal controls. ObjectivesTo systematically review and meta-analyze composite and solitary indices of IR, {beta}-cell function, and adipokine levels in individuals with LC compared with normal controls. MethodsPubMed, SCOPUS, and Google Scholar databases were searched for relevant studies from inception to August 2025. Sixty-three eligible studies were included, comprising 12,409 participants--5,891 LC patients and 6,518 normal controls. ResultsLC disease is characterized by elevated global IR (standardized mean difference, SMD = 0.395; 95% confidence intervals, CI: 0.226;0.563), Fasting insulin + C-peptide + FBG composite score (SMD = 0.605; 95% CI: 0.306;0.904) and acute + chronic glycemia composite scores (SMD = 0.424; 95% CI: 0.258;0.590). Furthermore, significant increases in HOMA-IR (SMD = 0.621; 95% CI: 0.379;0.863), Insulin (SMD = 0.488; 95% CI: 0.202;0.774), HbA1c (SMD = 0.308; 95% CI: 0.053;0.563), and fasting and random blood glucose (SMD = 0.831; 95% CI: 0.271;1.391, SMD = 0.396; 95% CI: 0.188;0.605) alongside reduced HOMA-%B and HOMA-%S were observed in LC patients versus normal controls. No publication bias observed in the results. ConclusionThe current study suggests that LC disease is characterized by persistent insulin resistance, hyperglycemia, and {beta}-cell dysfunction, suggesting sustained metabolic disturbances beyond the acute phase

Matching journals

The top 14 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.