Validating the Use of a Quick-Read Glucometer to Assess the Glycemic Response to Short-Term Capture Stress in Two Species of Snake, Nerodia sipedon and Sistrurus miliarius
The vertebrate stress response involves the release of adrenal hormones which make stored energy available to respond to stressors. The ability to effectively mobilize and distribute glucose is an important component of successful stress coping mechanisms. While many studies of reptiles have examined the hormonal response to stress under diverse contexts, comparatively few studies have focused on the glycemic response. Here we analyze the glycemic response to short-term capture stress in two species of snake, the Northern Water Snake (Nerodia sipedon) and the Pygmy Rattlesnake (Sistrurus miliarius), across reproductive contexts using a FreeStyle Lite (FSL) glucometer. The FSL is a point-of-care glucometer that provides a fast, inexpensive, and efficient method to monitor glucose concentrations in human whole blood. However, despite increasing use in reptiles, no studies have thoroughly validated their use in snakes or other squamates. We used a traditional colorimetric glucose assay to validate results using the FSL on both fresh whole blood and stored plasma. Hyperglycemia was induced by short-term capture stress in both species. The hyperglycemic response did not vary by sex, reproductive condition, or energetic status (body condition index, BCI). However, baseline and stress-induced glucose was higher in male compared to non-reproductive female N. sipedon. In S. miliarius, the magnitude of the glycemic response was positively correlated with cloacal body temperature. Baseline glucose was below the limit of detection in many individuals of both species, which may limit the utility of the FSL and other handheld glucometers in studies that seek to describe baseline values. Glucose estimates were strongly correlated when comparing all methods, indicating that the FreeStyle Lite is a potentially valuable tool to effectively compare the glycemic response to experimental treatment or to track changes in the glycemic response of snakes over time. However, significant bias was observed when comparing the three methodologies, suggesting that caution should be used when comparing results from the FSL and other point-of-care glucometers to reference intervals generated using other methodologies.
Scatterplots with linear regression lines describing the relationship between glucose concentrations calculated using colorimetric assay kits and the FreeStyle Lite glucometer readings for whole blood (open circles; Y = 0.79X – 0.22) and thawed plasma (closed circles; Y = 1.13X – 13.64) taken from S. miliarius in central Florida. The solid line represents a perfect isometric relationship.
Bias plots showing the discordance in glucose concentrations calculated using three different methods (whole blood FSL, plasma FSL, and colorimetric kit) across the range of reported mean glucose ((Method X + Method Y)/2) concentrations in S. miliarius sampled in central Florida (A–C) and N. sipedon sampled in southern New Jersey (D–F). Bias estimates (B) ±95% confidence level averaged across the range of measurements are provided in the top left corner of each plot. All linear relationships (solid lines) were statistically significant, with the exception of panel D.
Scatterplots with linear regression lines describing the relationship between glucose concentrations calculated using colorimetric assay kits and the FreeStyle Lite glucometer readings for whole blood (open circles; Y = 0.751X + 8.02) and thawed plasma (closed circles; Y = 1.196X – 3.66) taken from N. sipedon in southern New Jersey. The solid line represents a perfect isometric relationship.
Dilution curves created using pre-stress and post-stress pooled Nerodia sipedon plasma samples spiked with glucose standard. Predicted glucose concentrations were calculated using the initial concentration measured using the FreeStyle Lite glucometer (FSL) plus the known quantity of added glucose in the standard. Confidence limits on regression parameters (slope and intercept) did not indicate a significant deviation from isometry.
Mean and standard error plasma glucose pre- and post-30-minute capture stress in pregnant (Preg), non-reproductive female (NRF), and male S. miliarius. Bars that share a letter were not significantly different in pairwise comparisons. Sample size is indicated at the base of each bar.
Scatterplot showing the relationship between cloacal body temperature and delta glucose within each reproductive category in free-living S. miliarius. Delta glucose was significantly correlated with cloacal body temperature (P = 0.037) but was not significantly different among reproductive categories (P = 0.531).
Mean and standard error plasma glucose pre- and post-30-minute capture stress in vitellogenic/pregnant female (Repro), non-reproductive female (NRF), and male N. sipedon. Bars that share a letter were not significantly different in pairwise comparisons. Sample size is indicated at the base of each bar.
Contributor Notes
Associate Editor: C. Bevier.