Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Seasonal Changes in the Combined Glucose‐Insulin Tolerance Test in Normal Aged Horses

Seasonal Changes in the Combined Glucose‐Insulin Tolerance Test in Normal Aged Horses Abbreviations ANOVA analysis of variance AUCg area under the glucose curve BCS body condition score CGIT combined glucose‐insulin tolerance test CV coefficient of variation DST dexamethasone suppression test EMS equine metabolic syndrome G : I glucose to insulin ratio GC glucose clearance MSres residual mean square PPID pituitary pars intermedia dysfunction Equine metabolic syndrome (EMS) is a condition of altered glucose homeostasis and fat metabolism in ponies and horses characterized by obesity, regional adiposity, and laminitis. The underlying pathophysiology of EMS entails relative resistance or insensitivity of the tissues to insulin which results in a variety of physiologic changes at the tissue level. Equine laminar tissue may be sensitive to these physiologic alterations, and laminitis is a common and potentially life‐threatening sequela of EMS. Diagnosis of EMS can be difficult. Baseline glucose and insulin concentrations do not specifically measure insulin sensitivity and often are affected by factors such as stress, pain, and recent feed intake. Provocative tolerance tests measure the response of the glucose‐insulin axis to the administration of exogenous glucose or insulin. Glucose clamp techniques or intravenous glucose tolerance test with frequent sampling give the best indication of whole body insulin sensitivity, but are difficult to perform outside a research setting. The combined glucose‐insulin tolerance test (CGIT), developed by Eiler et al, combines the insulin and glucose tolerance tests to provide information on the response of the tissues to both substances. Although the CGIT does not directly measure tissue insulin sensitivity, it does provide valuable information on glucose and insulin homeostasis in normal and insulin‐resistant horses and can be used as diagnostic test for EMS. The CGIT can be performed in less than 2 hours and can be easily repeated to monitor treatment success or failure. Repeated measuring to monitor suspect individuals or to follow treatment progress in affected horses raises concerns over the repeatability of the CGIT and the possible influence of seasonal changes. Although the CGIT is repeatable over a short period of time (weeks), to our knowledge, no studies have evaluated seasonal changes. Ponies with EMS have increased risk of developing laminitis in late spring and early summer. This seasonality may be because of changes in grass nonstructural carbohydrate content, but seasonal alterations in hormones also may contribute to increased risk of laminitis in the spring. In fact, horses with pituitary pars intermedia dysfunction (PPID) can have increased insulin concentrations and insulin resistance secondary to the disease process. Recent research in PPID has shown significant seasonal changes in the hormones secreted by the pituitary gland, especially in the fall months. Indeed, recent work evaluating seasonal changes in the hypothalamic‐pituitary axis and seasonal changes in pasture‐associated laminitis has demonstrated variability in endogenous insulin concentrations and other measures of insulin resistance. , Three of these studies demonstrated subtle significant seasonal changes with increases in either insulin resistance or insulin concentrations in the summer or fall months , but no seasonal changes in glucose or insulin concentrations across seasons in healthy horses or horses with EMS were detected in a fourth study. The purpose of our study was to determine the effects of season on the CGIT. Our hypothesis was that the CGIT would show seasonal changes with abnormal glucose and insulin homeostasis in the summer and fall. Materials and Methods Horses Nine horses belonging to the Auburn University Large Animal Hospital herd were used in this study. The horses had neither evidence of regional adiposity, obesity, or laminitis during the study nor a history of laminitis. Four mares, 3 geldings, and 2 stallions with an average age of 14.8 ± 4.5 years (mean ± SD; range: 11–26 years) were used. Body condition score and body weight were assessed at the beginning and end of the study. Mean neck circumference was calculated for each horse during the study as previously described. Breeds included American Paint horse (n = 2), Thoroughbred (2), Warmblood (2), American Quarter horse (1), Tennessee Walking horse (1), and draft cross (1). All horses were assessed to be healthy by physical examination and results of CBC, serum biochemistry, and a CGIT at the beginning of the study. The horses did not have clinical signs consistent with PPID or EMS at any time during the study and had normal dexamethasone suppression test (DST) results (19‐hour cortisol concentration ≤30 nmol/L) at the start (February 2008) and end of the study (November 2008). The study protocol was approved by the Institutional Animal Care and Use Committee at Auburn University. Experimental Design The horses were housed in paddocks with no or minimal grass and were fed a constant diet of free choice coastal Bermuda hay. Some horses received 1–2 pounds of a pelleted grain once or twice a day depending on body condition. The diet and activity level did not change for any of the horses throughout the duration of the study. For the DST, 40 μg/kg dexamethasone was administered IM. Blood for measurement of serum cortisol concentration was collected before and 19 hours after dexamethasone administration. The blood was allowed to fully clot at room temperature and then centrifuged for 10 minutes; serum was removed and stored at −80°C until analysis. For the CGIT, the horses were housed in stalls with free choice hay and water, but no grain for at least 5 hours before the start of the experiment. This grain fast is similar to the fasting duration recommended in the 2010 ACVIM Consensus statement on EMS. An IV jugular catheter for blood sampling was placed at least 2 hours before the start of the test. Horses were tested on the same day in a staggered fashion starting at approximately 9 AM and finishing at approximately 3 PM. The CGIT consisted of rapid IV administration of 150 mg/kg of 50% dextrose and 0.1 U/kg of regular insulin in the noncatheterized jugular vein. Blood for glucose and insulin analysis was collected at all time points from the catheter after removal of at least 5 mL of waste blood. Blood glucose samples were taken before (T = 0) and 1, 5, 15, 30, 45, 60, 75, 90, and 150 minutes after administration of the dextrose and insulin, and were placed in heparinized tubes on ice. Samples were transported to the laboratory within 30 minutes and were analyzed by spectrophotometric analysis. Serum samples for measurement of insulin concentration were taken before (T = 0) and 5 and 75 minutes after administration of dextrose and insulin. The samples were placed in tubes with no anticoagulant or additive and allowed to fully clot at room temperature. Samples then were transported to the laboratory, centrifuged for 10 minutes and serum was removed and stored at −80°C until analysis. Serum cortisol and insulin concentrations were measured in duplicate using batched frozen samples by commercially available radioimmunoassays that have been validated in the horse. The CGIT tests were performed in 2008 in the months of February (control), May, June, August, September, and November. Daylight length during each testing period was 11.25, 13.5, 14.5, 13.5, 12, and 10.75 hours, respectively. Analysis of CGIT Results Area under the glucose curve (AUCg) was calculated using the trapezoidal method with commercial software. Positive phase duration was defined as the time from the start of the CGIT to the time the glucose concentration returned to baseline. Time to nadir was defined as the time from the start of the CGIT to the lowest measured glucose concentration. Positive phase glucose clearance was calculated by dividing the difference between the highest measured (T = 1 or T = 5 minutes) and baseline glucose concentrations by the difference in time from the highest measured glucose concentration to the end of the positive phase. Negative phase glucose clearance was calculated by dividing the difference between the baseline glucose concentration and the glucose nadir by the difference in time from the end of the positive phase and the lowest glucose concentration. The baseline glucose : insulin ratio (G : I) was calculated by dividing the baseline glucose concentration by the baseline insulin concentration. See Figure A for a graphical representation of these calculations. Horses were defined as having an abnormal CGIT if positive phase duration was >45 minutes or if the insulin concentration did not return to normal baseline concentration (<20 μU/mL) by T = 75 minutes. Baseline insulin concentration was not used to define an abnormal CGIT because concentrations were measured before the start of the test and horses were fed hay before and during the test. Results of the combined glucose‐insulin tolerance test ( CGIT ) in 9 horses. (A) Mean plasma glucose concentrations for each time point of the CGIT for each month tested. a‐ glucose concentration at the beginning of the CGIT ; b‐ highest measured glucose concentration; c‐ glucose concentration returned to baseline; d‐ glucose nadir; e‐ time of baseline glucose; f‐ time at highest measured glucose; g‐ time glucose returned to baseline; h‐time of glucose nadir. Positive phase duration = g‐e; time to nadir = h‐e; positive phase glucose clearance = (b‐c)/(g‐e); negative phase glucose clearance = (d‐c)/(h‐g). (B) Box plots of plasma glucose concentrations across time for all months combined. Values denoted by different letters are significantly different ( P < .05). (C) Box plots of the area under the glucose curve ( AUC g) for each month tested. The AUC g was calculated using the trapezoidal method with commercial software. *Indicates that February was significantly different from A ugust and N ovember #Indicates that J une was significantly different from N ovember ( P < .05). The box indicates the interquartile range (25–75%), the line in the box denotes the median value, and + signifies the mean. The upper whisker represents the maximum and the lower whisker the minimum. Statistical Analysis Data are reported as mean ± SD. Data were analyzed for normality using the D'Agostino and Pearson omnibus normality test. Nonnormal data were transformed and retested for normality. Natural log transformation normalized the data of baseline insulin, T = 75 minute insulin, and G : I. The highest measured glucose concentration remained nonparametric despite transformation and time to nadir was noncontinuous data, hence non parametric statistical analysis (Friedman's test) was performed on these data sets. Mean glucose concentrations during the CGIT were compared across months (treatment) and time (postinsulin and glucose administration) using repeated measures 2‐way analysis of variance (ANOVA). When significant effects of treatment (month), time (postinsulin and glucose administration) or the interaction (month × time) were identified at the P < .05 level, posthoc pair‐wise comparisons were made using Bonferroni analysis. Mean AUCg, baseline glucose concentration, baseline insulin concentration, baseline G : I, glucose nadir, T = 75 minute insulin concentration, positive phase duration, positive phase glucose clearance, and negative phase glucose clearance were compared across month using repeated measures 1‐way ANOVA. When significant effects of treatment (month) were identified at the P < .05 level, post hoc pair‐wise comparisons were made using Tukey's analysis. The highest measured glucose concentration and time to nadir were compared across months using the Friedman test, and when significant effects of treatment (month) were identified at the P < .05 level, post hoc pair‐wise comparisons were made using Dunn's multiple comparison tests. To evaluate the repeatability of the CGIT for an individual horse across months, repeatability coefficients and coefficients of variation (CVs) were calculated for the parametric variables AUCg, baseline glucose, baseline insulin, G : I, positive phase duration, glucose clearance, and glucose concentration at each time point. Repeatability coefficients were calculated as √2 × 1.96s w where s w is the estimate of the within‐subjects standard deviation calculated from the ANOVA table as the square root of the residual mean square (MSres). The CVs for the glucose concentration at each time point or the CVs for each parameter (eg, AUCg, G : I) were calculated for the 6 measurements (each month) for each horse. Correlations between parameters were performed using Pearson correlation coefficients. Statistical analysis was performed using a commercial software package. Results Body condition score (BCS) out of 9 and body weight were assessed in February and November. Individual differences in BCS were ≤0.5 between the initial (6.3 ± 0.5) and final (5.9 ± 0.6) assessment and were not significantly different ( P = .21). The average body weight was 587 ± 77 kg at the beginning of the study and 575 ± 82 kg at the end of the study and also was not significantly different ( P = .93). Mean neck circumference of all horses in the study was 95 ± 6.2 cm. All horses had normal CGIT results (positive phase duration <45 minutes and T = 75 minutes insulin concentrations ≤20 μU/mL) and normal DST results (19‐hour cortisol concentrations <30 nmol/L) at the beginning of the study (February). At the end of the study (November), all horses had normal positive phase durations and DST results but 2 horses had mildly increased T = 75 minutes insulin concentrations (24.2 and 26.8 μU/mL). The repeatability coefficients and CV for all parameters of the glucose curve are presented in Table . The repeatability coefficient describes the levels that 95% of the measurements should fall within in order for the variation to be caused by an individual variation rather than external factors, such as season. The CVs for negative phase GC, baseline insulin, and G : I ratio were quite high, indicating that these values were not repeatable. Assessment of repeatability of the CGIT over months Repeatability Coefficient Coefficient of Variation (CV)% Glucose concentration at each time point n/a Range: 2.9–15.2, Mean: 9.0 ± 4.5 Baseline glucose concentration 22.0 mg/dL 8.0 ± 4.0 AUCg 3.6 × 10 3 mg/dL/min 12.4 ± 3.4 Glucose nadir 30.8 mg/dL 23.7 ± 10.8 Positive phase duration 19.4 min 21 ± 10.4 Positive phase glucose clearance 4.2 mg/dL/min 23.7 ± 8.0 Negative phase glucose clearance 1.0 mg/dL/min 33.3 ± 10.1 Baseline insulin concentration 37.2 μU/mL 73 ± 38.8 Baseline glucose : insulin ratio 25.0 62 ± 22.6 No observable adverse effects of either test were detected in mares or geldings, but during CGIT testing, 1 stallion consistently showed mild signs of hypoglycemia (eg, tremors, muscle fasciculations, sweating) 60–90 minutes postglucose and insulin administration. During this time, the stallion was monitored closely for worsening of signs and glucose was available for administration. The episodes were <5 minutes in duration. Blood glucose concentrations in this stallion subsequently were found to be 21–33 mg/dL. The 2nd stallion also had markedly low blood glucose nadirs (23–36 mg/dL at 60–90 minutes) but did not show any clinical signs. Overall, results of the CGIT were consistent among months, although a few significant differences were identified. Mean plasma glucose concentrations during the CGIT were significantly different across months ( P < .001; Fig A) and time ( P < .001; Fig B). The interaction between month and time was not significant, hence the AUCg was used to compare months (Fig C). February had a significantly higher AUCg than August ( P < .05) and November ( P < .001), and June had a significantly higher AUCg than November ( P < .05). Within the CGIT test, glucose concentrations at all time points were significantly different ( P < .01) from each other with the following exceptions: T = 0, T = 30, and T = 150 minutes were not different from each other and T = 45, T = 60, T = 75, and T = 90 minutes were not significantly different from each other (Fig B). Parameters of the glucose and insulin responses during the CGIT are summarized in Table . No significant differences were detected across months in baseline glucose concentration ( P = .054), the highest measured glucose concentration ( P = .69), positive phase duration ( P = .081; Fig ), time to glucose nadir ( P = .36), glucose clearance during the positive phase ( P = .35), glucose clearance during the negative phase ( P = .13), baseline insulin concentration ( P = .22, Fig ), T = 75 minutes insulin concentration ( P = .96, Fig ), or G : I ( P = .26). Nadir glucose concentration was significantly different across months ( P = .009), with February being significantly higher than November ( P < .05). There was no correlation between body weight and the AUCg (R = .212, P = .31) or the glucose nadir (R = .144, P = .49). Box plots of the positive phase duration of glucose during the combined glucose‐insulin tolerance test in 9 horses. Positive phase duration was defined as the time from the start of the test to the time the glucose concentration returned to baseline. No significant differences were found across months. See Figure for additional explanation. Mean + SD insulin concentrations at baseline (T = 0) and T = 75 minutes during the combined glucose‐insulin tolerance test in 9 horses. No significant differences were found across months. Mean ± SD parameters for the combined glucose‐insulin tolerance test compared across months Feb May Jun Aug Sep Nov AUCg (×10 3 mg/dL/min) 13.2 ± 2.2 a 11.7 ± 1.6 12.5 ± 2.3 ab 11.2 ± 1.6 bc 11.6 ± 2.3 10.4 ± 2.4 c Baseline glucose concentration (mg/dL) 100 ± 7 93 ± 7 98 ± 9 91 ± 4 90 ± 4 94 ± 14 Highest measured glucose concentration (mg/dL) 250 ± 23 245 ± 32 233 ± 35 237 ± 18 229 ± 24 236 ± 22 Glucose nadir (mg/dL) 55 ± 18 a 42 ± 12 52 ± 19 41 ± 14 47 ± 19 37 ± 16 b Positive phase duration (min) 30 ± 8 27 ± 7 27 ± 10 27 ± 7 33 ± 16 23 ± 5 Time to nadir (min) 77 ± 16 70 ± 13 75 ± 13 77 ± 30 83 ± 29 67 ± 17 Positive phase glucose clearance (mg/dL/min) 5.7 ± 2.1 6.0 ± 1.3 5.6 ± 1.9 5.8 ± 1.6 5.7 ± 3.6 6.8 ± 2.6 Negative phase glucose clearance (mg/dL/min) 1.1 ± 0.6 1.2 ± 0.4 1.0 ± 0.3 1.2 ± 0.5 1.0 ± 0.6 1.4 ± 0.5 Baseline insulin concentration (μU/mL) 11.4 ± 6.5 7.2 ± 3.9 28.8 ± 31.5 11.0 ± 6.4 10.4 ± 10.1 13.8 ± 15.4 75 minutes insulin concentration (μU/mL) 11.0 ± 5.3 10.8 ± 9.2 10.5 ± 6.7 10.6 ± 3.0 9.7 ± 4.3 11.9 ± 7.9 Baseline glucose : insulin ratio 11.0 ± 5.0 16.4 ± 8.1 11.5 ± 12.3 10.7 ± 5.1 15.6 ± 12.8 12.1 ± 7.8 Within rows, values with different letter superscripts are significantly different ( P < .05). On the basis of our definition of an abnormal CGIT (positive phase duration >45 minutes or T = 75 minutes insulin concentration >20 μU/mL), mean CGIT results for all months were normal, but individual horses had abnormal CGIT results. In September, 2/9 horses (1 gelding and 1 mare) had glucose concentrations marginally above baseline (4 mg/dL and 11 mg/dL above baseline) at 45 minutes, resulting in positive phase durations of 49.5 and 60.5 minutes, respectively. Abnormal T = 75 minutes insulin concentrations were seen in 4/36 CGIT tests. These included 1 mare in May (34.3 μU/mL), another mare in June (21.0 μU/mL), and 2 geldings in November (24.2 and 26.8 μU/mL). Discussion The CGIT is a clinically applicable and economical method of assessing glucose and insulin homeostasis in horses. Measures of glucose and insulin responsiveness in dynamic tests such as the CGIT may have less reliable results in insulin‐resistant animals because of changes in the relative contribution of hepatic glucose production, hence understanding extraneous factors that could affect the results of the CGIT in normal animals is essential to interpreting correctly the results in abnormal animals. Given the seasonal changes reported in baseline insulin concentration in the southeastern United States, it was important to assess whether or not season affected the results of the CGIT. No significant differences were detected among months in parameters that are used to define the normal response to the CGIT such as positive phase duration, glucose clearance, and 75 minute insulin concentration. In addition, none of the means for any of the parameters of the CGIT were in the range for an insulin‐resistant horse for any month. There were, however, significant differences among months in the AUCg and glucose nadir. Frank et al demonstrated that normal horses have lower glucose nadirs and increased glucose clearance when compared with insulin‐resistant horses, which results in a lower AUCg (mean of 9.2 × 10 3 mg/dL/min in lean horses versus a mean of 15.5 × 10 3 mg/dL/min in obese/insulin‐resistant horses). Thus, the decreases in AUCg and glucose nadir seen in the fall months in this study may indicate a slight increase in insulin‐mediated glucose clearance in that season rather than the decrease that was expected. Recent literature has demonstrated a seasonal change in the activity of the hypothalic‐pituitary axis in normal and PPID‐affected horses. Some horses with PPID have been diagnosed with insulin resistance that is thought to be secondary to hormonal changes associated with PPID. Theoretically, the increase in pituitary activity in the fall would result in a decrease in insulin sensitivity in these months. This change in insulin sensitivity was not seen in the horses in our study. The horses in our study were tested for PPID by use of the DST in February and November. Although this test is not 100% accurate, the negative results seen in November (late fall), close to the time when false positive results may occur, suggest that the horses in our study were not affected by PPID. In addition, seasonal changes in insulin and glucose homeostasis are likely to be much less pronounced in normal horses than in horses with PPID because the changes in the hypothalamic‐pituitary axis are less extreme in these normal horses. Extraneous factors, as compared with endogenous physiological reasons, also need to be considered for changes in glucose and insulin homeostasis. For example, weight changes could affect glucose and insulin dynamics independent of season. However, no significant changes in BCS or body weight were detected over the study period, and no significant correlations between body weight and AUCg or glucose nadir were identified. With CVs of <15%, especially over months, the CGIT was shown to be repeatable for the parameters AUCg, baseline glucose, and glucose concentrations at each time point. The variation was much higher and the repeatability much lower for positive phase GC, positive phase duration, and negative phase GC. Baseline insulin and G : I were not repeatable over months. The variability in G : I is likely caused by the variability in baseline insulin (discussed below), because baseline glucose was very repeatable. Extraneous factors likely influenced these results because the measurements were taken before the start of the CGIT. Another important factor is the diet of the study population. In the previous studies that evaluated metabolic changes across time, the horses were allowed to graze free choice. Fluctuations in grass non structural carbohydrate concentrations through different seasons could account for changes in glucose and insulin homeostasis in those studies. The horses in our study were given free choice Bermuda grass hay with very limited grazing, similar to the protocol followed by Eiler et al in the initial description of the CGIT. The hay was purchased in large quantities from the same vendor to reduce variability, but carbohydrate analysis of the hay was not performed, hence it is unknown if starch content in the hay affected the results of the CGIT. However, Eiler et al reported that free‐choice water and grass hay did not affect glycemia during the CGIT. Baseline insulin concentrations were ≤20 μU/mL in 30/36 (83%) measurements and ≤40 μU/mL in 32/36 (89%) measurements in our study. No significant differences were detected across months in mean baseline insulin concentrations. Increases in baseline insulin concentrations in June occurred in 3 horses that had baseline insulin concentrations of 56.2, 56.5, and 91.5 μU/mL. All 3 horses had normal insulin concentrations by 75 minutes (≤20 μU/mL) and had normal glucose curves during the June CGIT. The baseline insulin concentrations in these horses may have been affected by hay feeding as recently reported. The findings of Chumbler et al were not reported before the performance of our study. Thus, hay was given free‐choice to all animals before and during the test as previously recommended. The glucose and insulin parameters found in this study are similar to those reported by Eiler et al, but the baseline G : I and AUCg were higher overall in our study (Table ) than those previously reported by Frank et al. Our study demonstrated a baseline G : I of 10.7–16.9 over the months compared with a mean of 6.4 (range: 3.2–11.6) at a single time point reported for normal horses by Frank et al. Our study also had higher AUCg concentrations over the months (10.4–13.2 × 10 3 mg/dL/min) than those reported by Frank et al at a single time point (7.7–10.1 × 10 3 mg/dL/min). The use of lithium heparin tubes in our study versus sodium fluoride tubes used by Frank et al likely explains most of the difference between studies, because sodium fluoride decreases blood glucose concentrations by approximately 7–12%. Lithium heparin tubes were chosen in this study because they are readily available in most veterinary hospitals and provide accurate glucose measurements if analyzed within 30 minutes. Eiler et al reported that insulin concentrations should return to baseline by 75 minutes during the CGIT. In our study, 32/36 insulin concentrations at 75 minutes (89%) were ≤20 μU/mL. Thus, most horses in most months had insulin concentrations return to baseline by T = 75 minutes. The 2 stallions used in the study had an apparent increase in insulin‐mediated glucose clearance compared to the geldings and mares. In fact, 1 stallion showed clinical signs of hypoglycemia (eg, muscle fasciculations, sweating, tremors) during the nadir portion of the CGIT when blood glucose concentrations were very low (21–33 mg/dL). The other stallion also had very low glucose concentrations during the nadir portion of the CGIT (23–36 mg/dL), but did not show clinical signs. Both horses had BCS of 6/9, and they had similar amounts of turnout and diet as the remaining horses in the study. Thus, a possible sex difference may exist in insulin sensitivity, and additional care may need to be taken when performing the CGIT on stallions. In humans, lower serum testosterone concentrations are associated with insulin resistance in men. Studies have shown a significant difference in insulin sensitivity (although within reference ranges) in mares depending on the stage of estrous cycle or pregnancy. Additional research is needed to fully evaluate the influences of sex hormones on the results of the GCIT in horses. In conclusion, our research demonstrated no seasonal change in the majority of parameters used to assess glucose and insulin dynamics in the CGIT using similarly aged horses maintained on a static diet, and none of the parameters were within the ranges for insulin‐resistant horses. The significant differences that were identified in the AUCg and glucose nadir were suggestive of a slight increase in insulin sensitivity in the summer and fall months. Taken as a whole, season does not cause clinically relevant changes in the results of the CGIT in normal horses in the southeastern United States. Acknowledgment We acknowledge Auburn University technicians and students Brittany Chaddick, Amy Bley, Michelle Brown, Ann Busch, Charles Smith, Heather Edwards, and Kathleen O'Donnell for technical assistance. Footnotes Schreiber CM, Stewart AJ, Behrend EN, et al. Seasonal variation in diagnostic tests for pituitary pars intermedia dysfunction in normal aged geldings. J Am Vet Med Assoc. In press Dexamethasone (2 mg/mL), Phoenix Pharmaceutical Inc, St Joseph, MO Dextrose 50%, Abbott Laboratories, Abbott Park, IL Humulin R (100 U/mL), Eli Lilly and Company, Indianapolis, IN Boehringer Mannheim/Hitachi 911 system, Boehringer Mannheim Corp, Indianapolis, IN Coat‐A‐Count Cortisol In‐vitro Diagnostic Test Kit, Siemens Medical Solutions Diagnostics, Los Angeles, CA Coat‐A‐Count Insulin In‐vitro Diagnostic Test Kit, Siemens Medical Solutions Diagnostics GraphPad Prism version 5.00 for Windows, GraphPad Software, San Diego, CA Chumbler NS, Toth F, Elliott SB, Frank N. Effects of sampling time and hay feeding on blood glucose, insulin and adrenocorticotropin hormone (ACTH) concentrations in horses. 27th Annual Forum of the ACVIM 2009 (abstract) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Veterinary Internal Medicine Wiley

Seasonal Changes in the Combined Glucose‐Insulin Tolerance Test in Normal Aged Horses

Download PDF
7 pages

Loading next page...
1
 
/lp/wiley/seasonal-changes-in-the-combined-glucose-insulin-tolerance-test-in-x9Y0l0CmjV

References (34)

Publisher
Wiley
Copyright
© 2012 American College of Veterinary Internal Medicine
eISSN
1939-1676
DOI
10.1111/j.1939-1676.2012.00939.x
pmid
22594619
Publisher site
See Article on Publisher Site

Abstract

Abbreviations ANOVA analysis of variance AUCg area under the glucose curve BCS body condition score CGIT combined glucose‐insulin tolerance test CV coefficient of variation DST dexamethasone suppression test EMS equine metabolic syndrome G : I glucose to insulin ratio GC glucose clearance MSres residual mean square PPID pituitary pars intermedia dysfunction Equine metabolic syndrome (EMS) is a condition of altered glucose homeostasis and fat metabolism in ponies and horses characterized by obesity, regional adiposity, and laminitis. The underlying pathophysiology of EMS entails relative resistance or insensitivity of the tissues to insulin which results in a variety of physiologic changes at the tissue level. Equine laminar tissue may be sensitive to these physiologic alterations, and laminitis is a common and potentially life‐threatening sequela of EMS. Diagnosis of EMS can be difficult. Baseline glucose and insulin concentrations do not specifically measure insulin sensitivity and often are affected by factors such as stress, pain, and recent feed intake. Provocative tolerance tests measure the response of the glucose‐insulin axis to the administration of exogenous glucose or insulin. Glucose clamp techniques or intravenous glucose tolerance test with frequent sampling give the best indication of whole body insulin sensitivity, but are difficult to perform outside a research setting. The combined glucose‐insulin tolerance test (CGIT), developed by Eiler et al, combines the insulin and glucose tolerance tests to provide information on the response of the tissues to both substances. Although the CGIT does not directly measure tissue insulin sensitivity, it does provide valuable information on glucose and insulin homeostasis in normal and insulin‐resistant horses and can be used as diagnostic test for EMS. The CGIT can be performed in less than 2 hours and can be easily repeated to monitor treatment success or failure. Repeated measuring to monitor suspect individuals or to follow treatment progress in affected horses raises concerns over the repeatability of the CGIT and the possible influence of seasonal changes. Although the CGIT is repeatable over a short period of time (weeks), to our knowledge, no studies have evaluated seasonal changes. Ponies with EMS have increased risk of developing laminitis in late spring and early summer. This seasonality may be because of changes in grass nonstructural carbohydrate content, but seasonal alterations in hormones also may contribute to increased risk of laminitis in the spring. In fact, horses with pituitary pars intermedia dysfunction (PPID) can have increased insulin concentrations and insulin resistance secondary to the disease process. Recent research in PPID has shown significant seasonal changes in the hormones secreted by the pituitary gland, especially in the fall months. Indeed, recent work evaluating seasonal changes in the hypothalamic‐pituitary axis and seasonal changes in pasture‐associated laminitis has demonstrated variability in endogenous insulin concentrations and other measures of insulin resistance. , Three of these studies demonstrated subtle significant seasonal changes with increases in either insulin resistance or insulin concentrations in the summer or fall months , but no seasonal changes in glucose or insulin concentrations across seasons in healthy horses or horses with EMS were detected in a fourth study. The purpose of our study was to determine the effects of season on the CGIT. Our hypothesis was that the CGIT would show seasonal changes with abnormal glucose and insulin homeostasis in the summer and fall. Materials and Methods Horses Nine horses belonging to the Auburn University Large Animal Hospital herd were used in this study. The horses had neither evidence of regional adiposity, obesity, or laminitis during the study nor a history of laminitis. Four mares, 3 geldings, and 2 stallions with an average age of 14.8 ± 4.5 years (mean ± SD; range: 11–26 years) were used. Body condition score and body weight were assessed at the beginning and end of the study. Mean neck circumference was calculated for each horse during the study as previously described. Breeds included American Paint horse (n = 2), Thoroughbred (2), Warmblood (2), American Quarter horse (1), Tennessee Walking horse (1), and draft cross (1). All horses were assessed to be healthy by physical examination and results of CBC, serum biochemistry, and a CGIT at the beginning of the study. The horses did not have clinical signs consistent with PPID or EMS at any time during the study and had normal dexamethasone suppression test (DST) results (19‐hour cortisol concentration ≤30 nmol/L) at the start (February 2008) and end of the study (November 2008). The study protocol was approved by the Institutional Animal Care and Use Committee at Auburn University. Experimental Design The horses were housed in paddocks with no or minimal grass and were fed a constant diet of free choice coastal Bermuda hay. Some horses received 1–2 pounds of a pelleted grain once or twice a day depending on body condition. The diet and activity level did not change for any of the horses throughout the duration of the study. For the DST, 40 μg/kg dexamethasone was administered IM. Blood for measurement of serum cortisol concentration was collected before and 19 hours after dexamethasone administration. The blood was allowed to fully clot at room temperature and then centrifuged for 10 minutes; serum was removed and stored at −80°C until analysis. For the CGIT, the horses were housed in stalls with free choice hay and water, but no grain for at least 5 hours before the start of the experiment. This grain fast is similar to the fasting duration recommended in the 2010 ACVIM Consensus statement on EMS. An IV jugular catheter for blood sampling was placed at least 2 hours before the start of the test. Horses were tested on the same day in a staggered fashion starting at approximately 9 AM and finishing at approximately 3 PM. The CGIT consisted of rapid IV administration of 150 mg/kg of 50% dextrose and 0.1 U/kg of regular insulin in the noncatheterized jugular vein. Blood for glucose and insulin analysis was collected at all time points from the catheter after removal of at least 5 mL of waste blood. Blood glucose samples were taken before (T = 0) and 1, 5, 15, 30, 45, 60, 75, 90, and 150 minutes after administration of the dextrose and insulin, and were placed in heparinized tubes on ice. Samples were transported to the laboratory within 30 minutes and were analyzed by spectrophotometric analysis. Serum samples for measurement of insulin concentration were taken before (T = 0) and 5 and 75 minutes after administration of dextrose and insulin. The samples were placed in tubes with no anticoagulant or additive and allowed to fully clot at room temperature. Samples then were transported to the laboratory, centrifuged for 10 minutes and serum was removed and stored at −80°C until analysis. Serum cortisol and insulin concentrations were measured in duplicate using batched frozen samples by commercially available radioimmunoassays that have been validated in the horse. The CGIT tests were performed in 2008 in the months of February (control), May, June, August, September, and November. Daylight length during each testing period was 11.25, 13.5, 14.5, 13.5, 12, and 10.75 hours, respectively. Analysis of CGIT Results Area under the glucose curve (AUCg) was calculated using the trapezoidal method with commercial software. Positive phase duration was defined as the time from the start of the CGIT to the time the glucose concentration returned to baseline. Time to nadir was defined as the time from the start of the CGIT to the lowest measured glucose concentration. Positive phase glucose clearance was calculated by dividing the difference between the highest measured (T = 1 or T = 5 minutes) and baseline glucose concentrations by the difference in time from the highest measured glucose concentration to the end of the positive phase. Negative phase glucose clearance was calculated by dividing the difference between the baseline glucose concentration and the glucose nadir by the difference in time from the end of the positive phase and the lowest glucose concentration. The baseline glucose : insulin ratio (G : I) was calculated by dividing the baseline glucose concentration by the baseline insulin concentration. See Figure A for a graphical representation of these calculations. Horses were defined as having an abnormal CGIT if positive phase duration was >45 minutes or if the insulin concentration did not return to normal baseline concentration (<20 μU/mL) by T = 75 minutes. Baseline insulin concentration was not used to define an abnormal CGIT because concentrations were measured before the start of the test and horses were fed hay before and during the test. Results of the combined glucose‐insulin tolerance test ( CGIT ) in 9 horses. (A) Mean plasma glucose concentrations for each time point of the CGIT for each month tested. a‐ glucose concentration at the beginning of the CGIT ; b‐ highest measured glucose concentration; c‐ glucose concentration returned to baseline; d‐ glucose nadir; e‐ time of baseline glucose; f‐ time at highest measured glucose; g‐ time glucose returned to baseline; h‐time of glucose nadir. Positive phase duration = g‐e; time to nadir = h‐e; positive phase glucose clearance = (b‐c)/(g‐e); negative phase glucose clearance = (d‐c)/(h‐g). (B) Box plots of plasma glucose concentrations across time for all months combined. Values denoted by different letters are significantly different ( P < .05). (C) Box plots of the area under the glucose curve ( AUC g) for each month tested. The AUC g was calculated using the trapezoidal method with commercial software. *Indicates that February was significantly different from A ugust and N ovember #Indicates that J une was significantly different from N ovember ( P < .05). The box indicates the interquartile range (25–75%), the line in the box denotes the median value, and + signifies the mean. The upper whisker represents the maximum and the lower whisker the minimum. Statistical Analysis Data are reported as mean ± SD. Data were analyzed for normality using the D'Agostino and Pearson omnibus normality test. Nonnormal data were transformed and retested for normality. Natural log transformation normalized the data of baseline insulin, T = 75 minute insulin, and G : I. The highest measured glucose concentration remained nonparametric despite transformation and time to nadir was noncontinuous data, hence non parametric statistical analysis (Friedman's test) was performed on these data sets. Mean glucose concentrations during the CGIT were compared across months (treatment) and time (postinsulin and glucose administration) using repeated measures 2‐way analysis of variance (ANOVA). When significant effects of treatment (month), time (postinsulin and glucose administration) or the interaction (month × time) were identified at the P < .05 level, posthoc pair‐wise comparisons were made using Bonferroni analysis. Mean AUCg, baseline glucose concentration, baseline insulin concentration, baseline G : I, glucose nadir, T = 75 minute insulin concentration, positive phase duration, positive phase glucose clearance, and negative phase glucose clearance were compared across month using repeated measures 1‐way ANOVA. When significant effects of treatment (month) were identified at the P < .05 level, post hoc pair‐wise comparisons were made using Tukey's analysis. The highest measured glucose concentration and time to nadir were compared across months using the Friedman test, and when significant effects of treatment (month) were identified at the P < .05 level, post hoc pair‐wise comparisons were made using Dunn's multiple comparison tests. To evaluate the repeatability of the CGIT for an individual horse across months, repeatability coefficients and coefficients of variation (CVs) were calculated for the parametric variables AUCg, baseline glucose, baseline insulin, G : I, positive phase duration, glucose clearance, and glucose concentration at each time point. Repeatability coefficients were calculated as √2 × 1.96s w where s w is the estimate of the within‐subjects standard deviation calculated from the ANOVA table as the square root of the residual mean square (MSres). The CVs for the glucose concentration at each time point or the CVs for each parameter (eg, AUCg, G : I) were calculated for the 6 measurements (each month) for each horse. Correlations between parameters were performed using Pearson correlation coefficients. Statistical analysis was performed using a commercial software package. Results Body condition score (BCS) out of 9 and body weight were assessed in February and November. Individual differences in BCS were ≤0.5 between the initial (6.3 ± 0.5) and final (5.9 ± 0.6) assessment and were not significantly different ( P = .21). The average body weight was 587 ± 77 kg at the beginning of the study and 575 ± 82 kg at the end of the study and also was not significantly different ( P = .93). Mean neck circumference of all horses in the study was 95 ± 6.2 cm. All horses had normal CGIT results (positive phase duration <45 minutes and T = 75 minutes insulin concentrations ≤20 μU/mL) and normal DST results (19‐hour cortisol concentrations <30 nmol/L) at the beginning of the study (February). At the end of the study (November), all horses had normal positive phase durations and DST results but 2 horses had mildly increased T = 75 minutes insulin concentrations (24.2 and 26.8 μU/mL). The repeatability coefficients and CV for all parameters of the glucose curve are presented in Table . The repeatability coefficient describes the levels that 95% of the measurements should fall within in order for the variation to be caused by an individual variation rather than external factors, such as season. The CVs for negative phase GC, baseline insulin, and G : I ratio were quite high, indicating that these values were not repeatable. Assessment of repeatability of the CGIT over months Repeatability Coefficient Coefficient of Variation (CV)% Glucose concentration at each time point n/a Range: 2.9–15.2, Mean: 9.0 ± 4.5 Baseline glucose concentration 22.0 mg/dL 8.0 ± 4.0 AUCg 3.6 × 10 3 mg/dL/min 12.4 ± 3.4 Glucose nadir 30.8 mg/dL 23.7 ± 10.8 Positive phase duration 19.4 min 21 ± 10.4 Positive phase glucose clearance 4.2 mg/dL/min 23.7 ± 8.0 Negative phase glucose clearance 1.0 mg/dL/min 33.3 ± 10.1 Baseline insulin concentration 37.2 μU/mL 73 ± 38.8 Baseline glucose : insulin ratio 25.0 62 ± 22.6 No observable adverse effects of either test were detected in mares or geldings, but during CGIT testing, 1 stallion consistently showed mild signs of hypoglycemia (eg, tremors, muscle fasciculations, sweating) 60–90 minutes postglucose and insulin administration. During this time, the stallion was monitored closely for worsening of signs and glucose was available for administration. The episodes were <5 minutes in duration. Blood glucose concentrations in this stallion subsequently were found to be 21–33 mg/dL. The 2nd stallion also had markedly low blood glucose nadirs (23–36 mg/dL at 60–90 minutes) but did not show any clinical signs. Overall, results of the CGIT were consistent among months, although a few significant differences were identified. Mean plasma glucose concentrations during the CGIT were significantly different across months ( P < .001; Fig A) and time ( P < .001; Fig B). The interaction between month and time was not significant, hence the AUCg was used to compare months (Fig C). February had a significantly higher AUCg than August ( P < .05) and November ( P < .001), and June had a significantly higher AUCg than November ( P < .05). Within the CGIT test, glucose concentrations at all time points were significantly different ( P < .01) from each other with the following exceptions: T = 0, T = 30, and T = 150 minutes were not different from each other and T = 45, T = 60, T = 75, and T = 90 minutes were not significantly different from each other (Fig B). Parameters of the glucose and insulin responses during the CGIT are summarized in Table . No significant differences were detected across months in baseline glucose concentration ( P = .054), the highest measured glucose concentration ( P = .69), positive phase duration ( P = .081; Fig ), time to glucose nadir ( P = .36), glucose clearance during the positive phase ( P = .35), glucose clearance during the negative phase ( P = .13), baseline insulin concentration ( P = .22, Fig ), T = 75 minutes insulin concentration ( P = .96, Fig ), or G : I ( P = .26). Nadir glucose concentration was significantly different across months ( P = .009), with February being significantly higher than November ( P < .05). There was no correlation between body weight and the AUCg (R = .212, P = .31) or the glucose nadir (R = .144, P = .49). Box plots of the positive phase duration of glucose during the combined glucose‐insulin tolerance test in 9 horses. Positive phase duration was defined as the time from the start of the test to the time the glucose concentration returned to baseline. No significant differences were found across months. See Figure for additional explanation. Mean + SD insulin concentrations at baseline (T = 0) and T = 75 minutes during the combined glucose‐insulin tolerance test in 9 horses. No significant differences were found across months. Mean ± SD parameters for the combined glucose‐insulin tolerance test compared across months Feb May Jun Aug Sep Nov AUCg (×10 3 mg/dL/min) 13.2 ± 2.2 a 11.7 ± 1.6 12.5 ± 2.3 ab 11.2 ± 1.6 bc 11.6 ± 2.3 10.4 ± 2.4 c Baseline glucose concentration (mg/dL) 100 ± 7 93 ± 7 98 ± 9 91 ± 4 90 ± 4 94 ± 14 Highest measured glucose concentration (mg/dL) 250 ± 23 245 ± 32 233 ± 35 237 ± 18 229 ± 24 236 ± 22 Glucose nadir (mg/dL) 55 ± 18 a 42 ± 12 52 ± 19 41 ± 14 47 ± 19 37 ± 16 b Positive phase duration (min) 30 ± 8 27 ± 7 27 ± 10 27 ± 7 33 ± 16 23 ± 5 Time to nadir (min) 77 ± 16 70 ± 13 75 ± 13 77 ± 30 83 ± 29 67 ± 17 Positive phase glucose clearance (mg/dL/min) 5.7 ± 2.1 6.0 ± 1.3 5.6 ± 1.9 5.8 ± 1.6 5.7 ± 3.6 6.8 ± 2.6 Negative phase glucose clearance (mg/dL/min) 1.1 ± 0.6 1.2 ± 0.4 1.0 ± 0.3 1.2 ± 0.5 1.0 ± 0.6 1.4 ± 0.5 Baseline insulin concentration (μU/mL) 11.4 ± 6.5 7.2 ± 3.9 28.8 ± 31.5 11.0 ± 6.4 10.4 ± 10.1 13.8 ± 15.4 75 minutes insulin concentration (μU/mL) 11.0 ± 5.3 10.8 ± 9.2 10.5 ± 6.7 10.6 ± 3.0 9.7 ± 4.3 11.9 ± 7.9 Baseline glucose : insulin ratio 11.0 ± 5.0 16.4 ± 8.1 11.5 ± 12.3 10.7 ± 5.1 15.6 ± 12.8 12.1 ± 7.8 Within rows, values with different letter superscripts are significantly different ( P < .05). On the basis of our definition of an abnormal CGIT (positive phase duration >45 minutes or T = 75 minutes insulin concentration >20 μU/mL), mean CGIT results for all months were normal, but individual horses had abnormal CGIT results. In September, 2/9 horses (1 gelding and 1 mare) had glucose concentrations marginally above baseline (4 mg/dL and 11 mg/dL above baseline) at 45 minutes, resulting in positive phase durations of 49.5 and 60.5 minutes, respectively. Abnormal T = 75 minutes insulin concentrations were seen in 4/36 CGIT tests. These included 1 mare in May (34.3 μU/mL), another mare in June (21.0 μU/mL), and 2 geldings in November (24.2 and 26.8 μU/mL). Discussion The CGIT is a clinically applicable and economical method of assessing glucose and insulin homeostasis in horses. Measures of glucose and insulin responsiveness in dynamic tests such as the CGIT may have less reliable results in insulin‐resistant animals because of changes in the relative contribution of hepatic glucose production, hence understanding extraneous factors that could affect the results of the CGIT in normal animals is essential to interpreting correctly the results in abnormal animals. Given the seasonal changes reported in baseline insulin concentration in the southeastern United States, it was important to assess whether or not season affected the results of the CGIT. No significant differences were detected among months in parameters that are used to define the normal response to the CGIT such as positive phase duration, glucose clearance, and 75 minute insulin concentration. In addition, none of the means for any of the parameters of the CGIT were in the range for an insulin‐resistant horse for any month. There were, however, significant differences among months in the AUCg and glucose nadir. Frank et al demonstrated that normal horses have lower glucose nadirs and increased glucose clearance when compared with insulin‐resistant horses, which results in a lower AUCg (mean of 9.2 × 10 3 mg/dL/min in lean horses versus a mean of 15.5 × 10 3 mg/dL/min in obese/insulin‐resistant horses). Thus, the decreases in AUCg and glucose nadir seen in the fall months in this study may indicate a slight increase in insulin‐mediated glucose clearance in that season rather than the decrease that was expected. Recent literature has demonstrated a seasonal change in the activity of the hypothalic‐pituitary axis in normal and PPID‐affected horses. Some horses with PPID have been diagnosed with insulin resistance that is thought to be secondary to hormonal changes associated with PPID. Theoretically, the increase in pituitary activity in the fall would result in a decrease in insulin sensitivity in these months. This change in insulin sensitivity was not seen in the horses in our study. The horses in our study were tested for PPID by use of the DST in February and November. Although this test is not 100% accurate, the negative results seen in November (late fall), close to the time when false positive results may occur, suggest that the horses in our study were not affected by PPID. In addition, seasonal changes in insulin and glucose homeostasis are likely to be much less pronounced in normal horses than in horses with PPID because the changes in the hypothalamic‐pituitary axis are less extreme in these normal horses. Extraneous factors, as compared with endogenous physiological reasons, also need to be considered for changes in glucose and insulin homeostasis. For example, weight changes could affect glucose and insulin dynamics independent of season. However, no significant changes in BCS or body weight were detected over the study period, and no significant correlations between body weight and AUCg or glucose nadir were identified. With CVs of <15%, especially over months, the CGIT was shown to be repeatable for the parameters AUCg, baseline glucose, and glucose concentrations at each time point. The variation was much higher and the repeatability much lower for positive phase GC, positive phase duration, and negative phase GC. Baseline insulin and G : I were not repeatable over months. The variability in G : I is likely caused by the variability in baseline insulin (discussed below), because baseline glucose was very repeatable. Extraneous factors likely influenced these results because the measurements were taken before the start of the CGIT. Another important factor is the diet of the study population. In the previous studies that evaluated metabolic changes across time, the horses were allowed to graze free choice. Fluctuations in grass non structural carbohydrate concentrations through different seasons could account for changes in glucose and insulin homeostasis in those studies. The horses in our study were given free choice Bermuda grass hay with very limited grazing, similar to the protocol followed by Eiler et al in the initial description of the CGIT. The hay was purchased in large quantities from the same vendor to reduce variability, but carbohydrate analysis of the hay was not performed, hence it is unknown if starch content in the hay affected the results of the CGIT. However, Eiler et al reported that free‐choice water and grass hay did not affect glycemia during the CGIT. Baseline insulin concentrations were ≤20 μU/mL in 30/36 (83%) measurements and ≤40 μU/mL in 32/36 (89%) measurements in our study. No significant differences were detected across months in mean baseline insulin concentrations. Increases in baseline insulin concentrations in June occurred in 3 horses that had baseline insulin concentrations of 56.2, 56.5, and 91.5 μU/mL. All 3 horses had normal insulin concentrations by 75 minutes (≤20 μU/mL) and had normal glucose curves during the June CGIT. The baseline insulin concentrations in these horses may have been affected by hay feeding as recently reported. The findings of Chumbler et al were not reported before the performance of our study. Thus, hay was given free‐choice to all animals before and during the test as previously recommended. The glucose and insulin parameters found in this study are similar to those reported by Eiler et al, but the baseline G : I and AUCg were higher overall in our study (Table ) than those previously reported by Frank et al. Our study demonstrated a baseline G : I of 10.7–16.9 over the months compared with a mean of 6.4 (range: 3.2–11.6) at a single time point reported for normal horses by Frank et al. Our study also had higher AUCg concentrations over the months (10.4–13.2 × 10 3 mg/dL/min) than those reported by Frank et al at a single time point (7.7–10.1 × 10 3 mg/dL/min). The use of lithium heparin tubes in our study versus sodium fluoride tubes used by Frank et al likely explains most of the difference between studies, because sodium fluoride decreases blood glucose concentrations by approximately 7–12%. Lithium heparin tubes were chosen in this study because they are readily available in most veterinary hospitals and provide accurate glucose measurements if analyzed within 30 minutes. Eiler et al reported that insulin concentrations should return to baseline by 75 minutes during the CGIT. In our study, 32/36 insulin concentrations at 75 minutes (89%) were ≤20 μU/mL. Thus, most horses in most months had insulin concentrations return to baseline by T = 75 minutes. The 2 stallions used in the study had an apparent increase in insulin‐mediated glucose clearance compared to the geldings and mares. In fact, 1 stallion showed clinical signs of hypoglycemia (eg, muscle fasciculations, sweating, tremors) during the nadir portion of the CGIT when blood glucose concentrations were very low (21–33 mg/dL). The other stallion also had very low glucose concentrations during the nadir portion of the CGIT (23–36 mg/dL), but did not show clinical signs. Both horses had BCS of 6/9, and they had similar amounts of turnout and diet as the remaining horses in the study. Thus, a possible sex difference may exist in insulin sensitivity, and additional care may need to be taken when performing the CGIT on stallions. In humans, lower serum testosterone concentrations are associated with insulin resistance in men. Studies have shown a significant difference in insulin sensitivity (although within reference ranges) in mares depending on the stage of estrous cycle or pregnancy. Additional research is needed to fully evaluate the influences of sex hormones on the results of the GCIT in horses. In conclusion, our research demonstrated no seasonal change in the majority of parameters used to assess glucose and insulin dynamics in the CGIT using similarly aged horses maintained on a static diet, and none of the parameters were within the ranges for insulin‐resistant horses. The significant differences that were identified in the AUCg and glucose nadir were suggestive of a slight increase in insulin sensitivity in the summer and fall months. Taken as a whole, season does not cause clinically relevant changes in the results of the CGIT in normal horses in the southeastern United States. Acknowledgment We acknowledge Auburn University technicians and students Brittany Chaddick, Amy Bley, Michelle Brown, Ann Busch, Charles Smith, Heather Edwards, and Kathleen O'Donnell for technical assistance. Footnotes Schreiber CM, Stewart AJ, Behrend EN, et al. Seasonal variation in diagnostic tests for pituitary pars intermedia dysfunction in normal aged geldings. J Am Vet Med Assoc. In press Dexamethasone (2 mg/mL), Phoenix Pharmaceutical Inc, St Joseph, MO Dextrose 50%, Abbott Laboratories, Abbott Park, IL Humulin R (100 U/mL), Eli Lilly and Company, Indianapolis, IN Boehringer Mannheim/Hitachi 911 system, Boehringer Mannheim Corp, Indianapolis, IN Coat‐A‐Count Cortisol In‐vitro Diagnostic Test Kit, Siemens Medical Solutions Diagnostics, Los Angeles, CA Coat‐A‐Count Insulin In‐vitro Diagnostic Test Kit, Siemens Medical Solutions Diagnostics GraphPad Prism version 5.00 for Windows, GraphPad Software, San Diego, CA Chumbler NS, Toth F, Elliott SB, Frank N. Effects of sampling time and hay feeding on blood glucose, insulin and adrenocorticotropin hormone (ACTH) concentrations in horses. 27th Annual Forum of the ACVIM 2009 (abstract)

Journal

Journal of Veterinary Internal Medicine – Wiley

Published: Jul 1, 2012

Keywords: ; ;

There are no references for this article.