Use of FreeStyle Libre for continuous glucose monitoring in adult horses

Abstract

Objectives

To evaluate the feasibility of using the FreeStyle Libre (a continuous glucose monitoring system [CGMS]) for instantaneous continuous monitoring of interstitial glucose in adult horses and examine the applicability and accuracy of this system in horses submitted to combined glucose–insulin test (CGIT).

Design

Laboratory measurements and continuous glucose monitoring system (CGMS) readings were analyzed using a 2 × 2 factorial statistical model with repeated measures over time. This analysis assessed the effects of the test (factor 1), group (factor 2), and their interactions (test × group, test × time, and group × time). Pearson's correlation analysis was applied to blood glucose values. Mean comparisons were conducted using the t-test, and agreement between techniques was assessed via the Bland–Altman method, with a 95% confidence interval.

Setting

Field study on private horse farms in association with a veterinary school.

Animals

Ten healthy stallions were assigned to one of two groups based on their body condition scores (BCS). Group 1 (G1, n = 5) consisted of nonobese horses with a BCS of 5 or 6, while Group 2 (G2, n = 5) consisted of obese horses with a BCS of 7 or higher.

Interventions

A CGMS sensor was attached to the dorsolateral aspect of the proximal one third of each horse's neck. Laboratory blood glucose measurements and CGMS interstitial glucose readings were compared at different time points for up to 7 days after sensor fixation. Obese horses were also submitted to CGIT on Day 4.

Measurements and Main Results

A comparative analysis of glucose measurements obtained in G1 and G2 horses using the CGMS and enzymatic methods revealed significant group × time interactions (P < 0.001) and time effects (P < 0.001). No interactions were detected between group (P = 0.45), test (P = 0.62), group and test (P = 0.28), or time and test (P = 0.92). In G1 and G2, tests were significantly correlated (r = 0.84 and P = 0.00) at all time points (T0–T5). Agreement between the glucose values obtained using different methods was excellent despite a small time delay in CGMS detection of rapid changes in blood glucose.

Conclusions

It was concluded that the CGMS can be used for indirect assessment of glycemic status (ie, based on interstitial glucose measurements) in nonobese and obese adult horses submitted to CGIT.