RESUMEN
BACKGROUND: Serum fructosamine (sFA) is used to assess glycaemic control in dogs with diabetes mellitus (DM). Nevertheless, its interpretation is hindered by several limitations. METHODS: This retrospective study evaluates the long-term diagnostic performance of sFA for monitoring clinical control of DM. sFA, bodyweight, appetite, presence of polyuria/polydipsia and clinical scores (CS; well-controlled DM, CS-0; uncontrolled DM, CS-1) were recorded. RESULTS: The study included 75 dogs (321 visits; median 3 visits/dog; range 1-19). Mean sFA was higher (p < 0.001) on visits with CS-1 (584 µmol/L; 95% confidence interval [95% CI] 561-608) than on visits with CS-0 (506 µmol/L; 95% CI 484-528). Increases in sFA increased the odds ratio for CS-1 (1.37; 95% CI 1.24-1.52, p < 0.001). sFA was moderately predictive of the CS (area under receiver operating characteristic curve = 0.75; 95% CI 0.70-0.80; p < 0.0001), with a 486 µmol/L cutoff yielding 80% sensitivity and 59% specificity. Mean sFA was lower (p = 0.005) when hypoglycaemic episodes were suspected (496 µmol/L; 95% CI 450-541) than in their absence (572 µmol/L; 95% CI 548-596). sFA is moderately accurate for classifying CS in diabetic dogs. Decreasing sFA over follow-ups indicates improved CS but might suggest occurrence of hypoglycaemic episodes. LIMITATIONS: Retrospective design, variable treatments and comorbidities are limitations of this study. CONCLUSION: sFA has a moderate clinical utility in the long-term monitoring of diabetic dogs, but may serve as a first-line, accessible diagnostic tool. Discordant CS and sFA evaluation, or decreased sFA, warrants additional monitoring (i.e., continuous glucose monitoring).