Triglyceride, Insulin, and Cortisol Responses of Ponies to Fasting and Dexamethasone Administration J. F. Freestone, BVSc, K. J. Wolfsheimer, DVM, PhD, R. B. Ford, DVM, MS, G. Church, MApStat, and R. Bessin, MApStat, PhD
Ponies were evaluated for their response to feed withholding and exogenous administration of corticosteroids (dexamethasone 0.04 mg/kg intramuscular [IMJ)in an attempt to reproduce the hyperlipemia syndrome. Because insulin resistance has been associated with hyperlipemia, all ponies were initially evaluated for insulin response to an oral glucose load and normal dexamethasone suppression of serum cortisol. Four ponies were identified as hyperinsulinemic reflecting insulin resistance. All ponies had suppressed cortisol concentrations following dexamethasone administration. Feed withdrawal resulted in hypertriglyceridemia by 48 hours in all ponies. Very low density lipoprotein-triglyceride (VLDL) fraction was primarily elevated. The administration of dexamethasone failed to increase the degree of triglyceridemia. Although insulin resistance has been proposed as the likely cause of the hypertriglyceridemia in ponies, in this study four of eight ponies were considered to have normal insulin responses and yet still developed hypertriglyceridemia. (Journal of Veterinary Internal Medicine 1991;
5~15-22)
and increased endogenous triglyceride synthesis by the liver, resulting in hypertriglyceridemia.g-" Adrenocorticotrophic hormone (ACTH) and cortisol influence lipid metabolism both directly and indirectly. Corticosteroids stimulate hormone sensitive lipase. Hormone sensitive lipase increases the mobilization of free fatty acids from adipose tissue and increases plasma triglycerides. Corticosteroids indirectly influence lipid metabolism by antagonizing the action of insulin." When insulin is functioning normally its release will increase lipoprotein lipase activity. Lipoprotein lipase activity clears the plasma of triglycerides. Therefore stress-induced increases in serum cortisol concentrations may cause hypertriglyceridemia and decreased insulin activity." In dogs, humans, and horses hyperadrenocorticism has been associated with hyperglycemia and hyperin~ulinemia.'~~~~~'~ In hyperadrenocorticoid states, the cellular insulin response is affected at the receptor or postreceptor 1 e ~ e l . I ~ Glucocorticoids have also been implicated in the development of laminitis in ponies and horses.15Hyperadrenocorticoid ponies are predisposed to developing laminitis. In one study, ponies with the most severe hyperinsulinemia were those with a history of la mini ti^.^ In this study the ponies were not screened for hyperadrenocortici~m.~ It is therefore possible that increased circulating cortisol concentrations may play a role in the
HYPERLIPEMIA is a common condition of Hyperlipemia is defined as an increase in plasma triglycerides greater than 500 mg/dl.' Obesity and stressful conditions (such as pregnancy, lactation, anorexia or transportation) have been associated with fat mobilization, hypertriglyceridemia, and clinical signs of the hyperlipemia ~yndrome.*.~,'~' The clinical signs of the hyperlipemia syndrome are nonspecific. Depression, weakness, anorexia, ataxia, and terminal recumbency may be noted.2 Obese ponies, especially those with a history of laminitis are more commonly insulin resistant than are normal ponies5 In obese humans, tissue resistance to insulin is common. Insulin resistance is defined as a state in which increased amounts of insulin are required to produce a biologic response.' In these people insulin resistance results in abnormal lipoprotein lipase activity From Department of Veterinary Clinical Sciences (Freestone and Wolfsheimer), School of Veterinary Medicine, Department of Experimental Statistics (Church and Bessin), Louisiana State University, Baton Rouge, Louisiana, and Department of Companion Animal and Species Medicine (Ford), School of Veterinary Medicine, North Carolina State University, Raleigh, North Carolina. Supported by School of Veterinary Medicine Organized Research Funds, Louisiana State University, Baton Rouge, Louisiana Reprint requests: J. Freestone, Department of Veterinary Clinical Sciences, School of Veterinary Medicine, Louisiana State University, Baton Rouge, LA 70803.
15
16
Journal of Veterinary Internal Medicine
FREESTONE ET AL
TABLEI . Physical, Selected Endocrine, and Fasting Triglyceride Characteristics of 8 Ponies
Pony #
I 2 3 4 5
6 7
8
Age (YrS) 4
5 4 3 14
5 7 4
Sex
Weight (kgs)
Body Score
IIG Indices
F F F F F F F M
I07 167 135 134 130 232 200 I35
Fair Good Good Good Obese Good Fair Good
N ABN N ABN ABN ABN N N
Dex SUPP (t = 24)
TG mg/d (t = 72)
N N N N N N N N
I ,060* 492 170 293 388 626 235 333
N: Normal; ABN: Abnormal: TG: plasma triglycerides after 72 hours of fasting. * Pony # I died between 48 and 72 hours samples.
development of laminitis, insulin resistance, and hyperlipemia. Our objectives in this study were 1) to identify ponies with hyperinsulinemia by monitoring the insulin response following oral glucose; 2) to identify ponies with hyperadrenocorticism by the dexamethasone suppression test; 3 ) to measure total and fractional triglycerides, glucose, cortisol and insulin in normal and hyperinsulinemic ponies during a 72-hour acute feed withdrawal period; and 4) to measure the changes in glucose, insulin, cortisol, and triglycerides following dexamethasone administration at time 72 hours during 120 hours of fasting. Materials and Methods
Ponies Eight ponies, consisting of seven females and one male, with a mean age of 5.7 _t 3.5 yrs and weight of 155 -t 41.9 kgs were studied. The studies were done during December and January. All ponies were normal on physical exam. The hemograms were normal, and they were negative for equine infectious anemia. The ponies were wormed with fenbendazole 5 mg/kg orally before the start of the study. The ponies’ body condition was subjectively scored as obese, good, and fair. Five ponies were in good condition, two were fair, and one was obese (Table 1). Ponies were kept on pasture and were supplemented by a high concentrate pelleted ration” and grass hay. Ponies were removed from pasture before each experiment and kept in stalls. All ponies, except #7, gained weight from the start of the study ( 1 7 1.8 -+ 40.8 kgs) to the fasting trial. Two ponies ( 3 , 7) developed a respiratory infection during the study and were treated with oral trimethoprim-sulfamethazine. They recovered before endocrine evaluations and fasting. One of these ponies (#7) lost weight. A recovery period of 14 days was allowed between each endocrine study and before fasting. * Pure Pride 200 Purina Mills Inc, St Louis, MO.
Endocrine Evaluation Oral Glucose Tolerance Test
The insulin response to oral glucose administered at 1 g/kg as a 20% solution was measured. Feed was withheld from ponies overnight ( 1 7 hrs) before glucose administration. Serum and EDTA anticoagulated blood was collected from the jugular vein. EDTA tubes were spun immediately and plasma retrieved for glucose determination.? Serum insulin was measured by a commercially available kit$ by methods previously validated in the horse. Samples were collected at 0, 15,30,45,60,90, 120, 150, 180, and 210 minutes post-glucose administration as previously d e ~ c r i b e dNo . ~ sedatives were administered. Based on insulin response, ponies were classified into normal or abnormal groups by evaluating 1) the insulin to glucose (I/G) ratio at 90 minutes; 2) summing insulin and glucose concentrations at times 60,90, and 120 minutes and comparing the ratio (I/G sum); 3 ) calculating total insulin secreted (TIS); 4) determining insulin peak response (IPR) and glucose peak response from baseline (GPR); 5) calculating the insulinogenic index; and 6) comparing the total insulin secreted to total glucose ratio (TIS/GT) (Table 2). The IPR was defined as the highest concentration of insulin achieved above the baseline insulin concentration. The insulinogenic index was calculated as the ratio of the IPR to the greatest glucose increment above the fasting value.I6 Total insulin secretion (TIS) and glucose (GT) was determined by calculating the area under the curve from baseline to 2 10 minutes. Dexamethasone Suppression Testing
A dexamethasone suppression test was performed on all ponies as previously described.” Ponies were removed from pasture at 9.30 A.M. and not fed until the completion of the study 29 hours later. Water was available ad libitum. Dexamethasone (0.04 mg/kg) was administered
t Encore Chemistry System, Baker Instruments, Allentown, PA. $ Coat-a-Count, Diagnostic Products, Los Angeles, CA.
Vol. 5
. NO. 1 , 1991
17
TRIGLYCERIDE AND ENDOCRINE RESPONSE IN PONIES
TABLE 2. Insulin Response to Oral Glucose (lg/kg) Administration in 8 Ponies Pony ## Hyperinsulinemic 2 4
5 6 Normal I 3 7 8
IPR (uIU/ml)
GPR (mg/dU
Insulinogenic Index
Istl/G~o
Iso-tzo/G~o-lzo
Ratio
Ratio
TIS uIU/ml
GT (mg/dl)
TlS/GT Ratio
90.0 55.5 149.5 65.0
I I9
89 1 00 84
0.76 0.62 1.5 0.77
0.60 0.37 0.87 0.33
0.54 0.37 0.79 0.38
492.5 304.0 769.9 345.6
1094.0 946.3 1 1 10.2 959.8
0.45 0.32 0.69 0.36
39.9 17.2 38.0 19.7
I26 78 I13 I25
0.32 0.22 0.34 0.16
0.23 0.09 0.24 0.08
0.19 0.09 0.22 0.10
284. I 109.7 191.5 128.8
1298.5 1061.7 978.5 1145.2
0.22 0.10 0.20 0.1 I
IM at 2.30 P.M., 5 hours later. Blood was collected from the jugular vein before and at 15, 19, and 24 hours after dexamethasone adminjstration. Triglyceride (TG) and cholesterol (CH) and the concentration of the four lipoprotein fractions (total [TI, very low density [VLDL], low density [LDL], and high density [HDL]) were measured on the heparinized plasma.5 Total cholesterol was determined using a standardized modification of the Liebermann-Burchard reaction and read spectrophotometrically at 640 nm. Total cholesterol was considered adequate to illustrate the overall cholesterol trend in this study. Triglycerides were determined by a totally enzymatic method (hydrolysis by microbial lipase) and read spectrophotometrically at 530 nm. Separation of the lipoprotein fractions was performed by ultracentrifugation. Chylomicrons were recovered from the top layer following ultracentrifugation at 100,000 RPM for 10 minutes. Plasma samples were then ultracentrifuged at 100,000 RPM for a further 2.5 hours and the supranatant recovered for VLDLs. High density lipoproteins were assessed after initial precipitation of LDL and VLDL fractions with a nonmetallic polyionic precipitating agent, moderate centrifugation, and supernatant recovery. For chylomicrons, VLDL and HDL purity was verified by single band electrophoresis. Low-density lipoproteins were derived mathematically subtracting HDL from the combined HDL-LDL fraction. Serum cortisol determinations were made by radioimmunoassay using a commercially available kit7 validated for use in the horse. Fasling Studies: Experimental Design All ponies were weighed and kept in stalls on the day
before the start of the study. A complete pelleted feed ration11 was given at midday then withdrawn 3 hours later (3 P.M., and baseline samples were collected. Blood samples were collected 24,48,72 hours postfasting. One pony (# 1 ) died between 48 and 72 hours of fasting, and was not included in the fasting data analysis.
At 72 hours the ponies were divided into two groups. Group 1 (4 ponies) received dexamethasone IM (0.04 mg/kg) and group 2 (3 ponies) received 0.9% saline (placebo). Each group contained two hyperinsulinemic ponies. Blood samples were then collected at 80, 88, 96, 104, 112, and 120 hours of fasting. At each sampling time (0, 24, 48, 72, 80, 88, 96, 104, 112, 120 hours) during fasting plasma triglyceride and total cholesterol, plasma glucose, serum insulin, and serum cortisols were determined. In addition cholesterol and triglyceride concentrations in each of three major lipoprotein fractions (VLDL, LDL, HDL) were determined.
Statistical Analysis The data of the oral glucose tolerance test were analyzed using a repeated measures one-way analysis of variance and the Huyhn-Feldt test statistic to adjust for sphericity. A hypothesised abnormal group was compared with a normal group. Means were compared univariately using least square In the feed withholding experiment, groups were compared with a two-way repeated measures analysis of variance using a Huyhn-Feldt test statistic to adjust for sphericity.” During the first 72 hours of feed, withdrawal ponies with normal insulin responses (n = 3) were compared with hypothesised abnormal ponies (n = 4). In addition, repeated measures analysis of covariance using the Huyhn-Feldt test statistic to adjust for sphericity was performed on the drug versus placebo group during all 10 experimental times and during the six post-drug administration times (80 h-120 h). A covariate of the insulin to glucose sum was used to remove variation of hyperinsulinemic ponies. Means were compared univariately for groups within time and times within a group using least square Significance was reported at the P 5 0.05 level.”
ResuIts Oral Glucose Tolerance Test
0 Lipid ultracentrifugation,
Beckman Airfuge, Fullerton, CA. 7 Coat-a-Count, Diagnostic Products, Los Angeles, CA. 1) Horse Chow 200 Purina Mills Inc, St Louis, MO.
Ponies resting plasma glucose concentrations (range, 66-95 mg/dl) and serum insulin concentrations (range,
18
Journal of Veterinary Internal Medicine
FREESTONE ET AL.
3.8- 1 1.2 uIU/ml) were within reference ranges. Oral glucose absorption was normal in all ponies and the plasma glucose concentration doubled over baseline within 90 minutes. Four of eight ponies were hypothesised to be hyperinsulinemic in response to the administration of glucose orally based on calculated indices (Table 2). The I/G sum used to compare the normal versus the hyperinsulinemic ponies was significantly different (P < 0.03). In addition the I/G sum of the normal versus the abnormal ponies changed across time (P < 0.0002) and the groups changed differently over time (P < 0.02). Insulin concentrations in the abnormal ponies were significantly increased versus normal ponies at times 15, 30, 45, 60, 90, 120, 150, 180, and 210 minutes (Fig 1). In the hyperinsulinemic ponies, there was a rise in peak insulin concentration over baseline of 1,18496. In contrast, the normal pony peak insulins increased 522% above fasting concentrations (Fig I). There was no significant difference between the normal and abnormal ponies for glucose concentration (Fig 2).
225
T
504
:
:
:
:
:
0 15 30 45 60
:
: 90
:
: 120
:
: 150
:
: 180
:
I
210
TIME (rnins)
FIG.2. Glucose response to oral glucose administration ( 1 g/kg as a 20% solution) by hyperinsulinemic (n = 4) and normal (n = 4) ponies. Data expressed as mean SD.
*
between the normal and hyperinsulinemic ponies for insulin, glucose, cortisol, cholesterol, or plasma triglycerides.
Dexamethasone Suppression Test
Baseline cortisol for all ponies ranged between 4.4-9.8 ug/dl. In all ponies, the exogenous administration of dexamethasone-suppressed serum cortisol concentration and the cortisol concentration was still suppressed by 24 hours postadministration (Table 3). There were no changes in the lipid profile after dexamethasone administration (Table 3). Fasting During 72 Hours: Normal vs. Hyperinsulinemic Ponies
Fasting resulted in decreased glucose concentrations across time (P < 0.0001) and increased Total - TG (P < 0.0001), VLDL - TG (P< 0.0005), and Total - CH (P < 0.01) (Table 4). There were no significant differences
*r
***
125
hyperinsulinemic normal
9-9
Fasting-0
The withholding of feed followed by administration of either dexamethasone or a placebo at time 72 hour caused significant decreases in glucose concentration (P < 0.04) and increases in TG-T ( P < 0.000 l), VLDL-TG (P < 0.004) and total CH (P < 0.003) during the study. Glucose concentration was significantly different between the groups (P < 0.03), increasing after the administration of dexamethasone. Cortisol showed significant group (P < 0.03), group X time (P < 0.0008) and time (P < 0.04) effects, as dexamethasone decreased cortisol concentrations in the treated group (Table 5). Baseline glucose and insulin concentrations were elevated, associated with feeding 2 hours before the collection of the samples. Feed withdrawal during the first 72 hours resulted in a decline in plasma glucose and serum insulin concentration (Table 5). As glucose decreased serum insulin also decreased, while VLDL-TG was increased by 48 hours (P < 0.05) and remained increased (Table 5). Pony #I developed clinical signs of depression and ataxia between the 48- and 72-hour sampling and was found to have severe hypoglycemia (20 mg/dl) and hypertriglyceridemia (TG-T I060 mg/dl). No treatment was administered. The pony died before the 72-hour sample. Necropsy showed centrolobular congestion, diffuse periportal fatty change, hemosiderin accumulation in the liver, and catarrhal sinusitis. The liver changes were consistent with the hyperlipemia syndrome. The affected pony was in fair body condition and had a normal oral glucose tolerance test and dexamethasone suppression test (Table I).
25t 2 04
f : :
:
:
0 15 30 45 60
:
:
90
:
:
120
:
:
150
:
:
180
:
4
210
TIME (mins) FIG.I . Insulin response to oral glucose administration ( 1 g/kg, as a 20% solution) by hyperinsulinemic (n = 4) and normal (n = 4)ponies. Data expressed as mean f SD. Significant differences from normal ponies: * (PI 0.05), ** (PI0.00I), *** ( P I0.0001).
to 72 Hours: Dexamethasone vs. Placebo
VOl. 5
. No. 1. 1991
19
TRIGLYCERIDE AND ENDOCRINE RESPONSE IN PONIES
TABLE3. Alteration in Cortisol Concentration and Lipid Profile Following Dexamethasone (0.04 mg/kg, IM) in 8 Ponies Over Time Time (hr after dexamethasone) Np. of Ponies
0
15
19
24
0.28 f 0.1
0.29 f 0.1
0.3 f 0.1
8 Cortisol (ug/dl) Triglyceride (mg/dl) Total VLDL LDL H DL Cholesterol (mg/dl) Total
* Data expressed as mean
6.2 f 1.8* 78.5 f 11 .O 5.5 f 3.1 15.0 f 7.9 58.9 f 8.8
83.1 f 11.8 3.6 f 3.1 17.9 f 3.8 62.0 f 9.4
85.9 6.5 17.6 62.4
f 11.5 f 6.6 f 2.3 f 9.0
86.3 f 9.6 6.9 f 4.9 21.8 f 16.0 57.9 18.6
17.9 f 6.1
33.9
41.4
f
20.9
52.1 f 18.9
+_
14.9
*
f sd.
Feed Withdrawal Over 120 Hours: Post-Dexamethasone Response (t 80-120)
The administration of dexamethasone led to differences between the groups for only cortisol concentration (P < 0,008). Serum cortisol was suppressed after dexamethasone administration and differed between the groups at times 80 ( P < O.Ol), 88 (P < 0.001), 96 ( P < 0.001), and 104 ( P < 0.01) hours. Glucose concentration was increased over time ( P < 0.04). This increase was noted at the first sample collected 8 hours (t = 80 hours) after sample, following the administration of dexamethasone at 72 hours ( P < 0.05). The glucose concentrations for the two groups also changed differently across time ( P
Ponies were evaluated for their response to feed withholding and exogenous administration of corticosteroids (dexamethasone 0.04 mg/kg intramuscular [IMJ)in an attempt to reproduce the hyperlipemia syndrome. Because insulin resistance has been associated with hyperlipemia, all ponies were initially evaluated for insulin response to an oral glucose load and normal dexamethasone suppression of serum cortisol. Four ponies were identified as hyperinsulinemic reflecting insulin resistance. All ponies had suppressed cortisol concentrations following dexamethasone administration. Feed withdrawal resulted in hypertriglyceridemia by 48 hours in all ponies. Very low density lipoprotein-triglyceride (VLDL) fraction was primarily elevated. The administration of dexamethasone failed to increase the degree of triglyceridemia. Although insulin resistance has been proposed as the likely cause of the hypertriglyceridemia in ponies, in this study four of eight ponies were considered to have normal insulin responses and yet still developed hypertriglyceridemia. (Journal of Veterinary Internal Medicine 1991;
5~15-22)
and increased endogenous triglyceride synthesis by the liver, resulting in hypertriglyceridemia.g-" Adrenocorticotrophic hormone (ACTH) and cortisol influence lipid metabolism both directly and indirectly. Corticosteroids stimulate hormone sensitive lipase. Hormone sensitive lipase increases the mobilization of free fatty acids from adipose tissue and increases plasma triglycerides. Corticosteroids indirectly influence lipid metabolism by antagonizing the action of insulin." When insulin is functioning normally its release will increase lipoprotein lipase activity. Lipoprotein lipase activity clears the plasma of triglycerides. Therefore stress-induced increases in serum cortisol concentrations may cause hypertriglyceridemia and decreased insulin activity." In dogs, humans, and horses hyperadrenocorticism has been associated with hyperglycemia and hyperin~ulinemia.'~~~~~'~ In hyperadrenocorticoid states, the cellular insulin response is affected at the receptor or postreceptor 1 e ~ e l . I ~ Glucocorticoids have also been implicated in the development of laminitis in ponies and horses.15Hyperadrenocorticoid ponies are predisposed to developing laminitis. In one study, ponies with the most severe hyperinsulinemia were those with a history of la mini ti^.^ In this study the ponies were not screened for hyperadrenocortici~m.~ It is therefore possible that increased circulating cortisol concentrations may play a role in the
HYPERLIPEMIA is a common condition of Hyperlipemia is defined as an increase in plasma triglycerides greater than 500 mg/dl.' Obesity and stressful conditions (such as pregnancy, lactation, anorexia or transportation) have been associated with fat mobilization, hypertriglyceridemia, and clinical signs of the hyperlipemia ~yndrome.*.~,'~' The clinical signs of the hyperlipemia syndrome are nonspecific. Depression, weakness, anorexia, ataxia, and terminal recumbency may be noted.2 Obese ponies, especially those with a history of laminitis are more commonly insulin resistant than are normal ponies5 In obese humans, tissue resistance to insulin is common. Insulin resistance is defined as a state in which increased amounts of insulin are required to produce a biologic response.' In these people insulin resistance results in abnormal lipoprotein lipase activity From Department of Veterinary Clinical Sciences (Freestone and Wolfsheimer), School of Veterinary Medicine, Department of Experimental Statistics (Church and Bessin), Louisiana State University, Baton Rouge, Louisiana, and Department of Companion Animal and Species Medicine (Ford), School of Veterinary Medicine, North Carolina State University, Raleigh, North Carolina. Supported by School of Veterinary Medicine Organized Research Funds, Louisiana State University, Baton Rouge, Louisiana Reprint requests: J. Freestone, Department of Veterinary Clinical Sciences, School of Veterinary Medicine, Louisiana State University, Baton Rouge, LA 70803.
15
16
Journal of Veterinary Internal Medicine
FREESTONE ET AL
TABLEI . Physical, Selected Endocrine, and Fasting Triglyceride Characteristics of 8 Ponies
Pony #
I 2 3 4 5
6 7
8
Age (YrS) 4
5 4 3 14
5 7 4
Sex
Weight (kgs)
Body Score
IIG Indices
F F F F F F F M
I07 167 135 134 130 232 200 I35
Fair Good Good Good Obese Good Fair Good
N ABN N ABN ABN ABN N N
Dex SUPP (t = 24)
TG mg/d (t = 72)
N N N N N N N N
I ,060* 492 170 293 388 626 235 333
N: Normal; ABN: Abnormal: TG: plasma triglycerides after 72 hours of fasting. * Pony # I died between 48 and 72 hours samples.
development of laminitis, insulin resistance, and hyperlipemia. Our objectives in this study were 1) to identify ponies with hyperinsulinemia by monitoring the insulin response following oral glucose; 2) to identify ponies with hyperadrenocorticism by the dexamethasone suppression test; 3 ) to measure total and fractional triglycerides, glucose, cortisol and insulin in normal and hyperinsulinemic ponies during a 72-hour acute feed withdrawal period; and 4) to measure the changes in glucose, insulin, cortisol, and triglycerides following dexamethasone administration at time 72 hours during 120 hours of fasting. Materials and Methods
Ponies Eight ponies, consisting of seven females and one male, with a mean age of 5.7 _t 3.5 yrs and weight of 155 -t 41.9 kgs were studied. The studies were done during December and January. All ponies were normal on physical exam. The hemograms were normal, and they were negative for equine infectious anemia. The ponies were wormed with fenbendazole 5 mg/kg orally before the start of the study. The ponies’ body condition was subjectively scored as obese, good, and fair. Five ponies were in good condition, two were fair, and one was obese (Table 1). Ponies were kept on pasture and were supplemented by a high concentrate pelleted ration” and grass hay. Ponies were removed from pasture before each experiment and kept in stalls. All ponies, except #7, gained weight from the start of the study ( 1 7 1.8 -+ 40.8 kgs) to the fasting trial. Two ponies ( 3 , 7) developed a respiratory infection during the study and were treated with oral trimethoprim-sulfamethazine. They recovered before endocrine evaluations and fasting. One of these ponies (#7) lost weight. A recovery period of 14 days was allowed between each endocrine study and before fasting. * Pure Pride 200 Purina Mills Inc, St Louis, MO.
Endocrine Evaluation Oral Glucose Tolerance Test
The insulin response to oral glucose administered at 1 g/kg as a 20% solution was measured. Feed was withheld from ponies overnight ( 1 7 hrs) before glucose administration. Serum and EDTA anticoagulated blood was collected from the jugular vein. EDTA tubes were spun immediately and plasma retrieved for glucose determination.? Serum insulin was measured by a commercially available kit$ by methods previously validated in the horse. Samples were collected at 0, 15,30,45,60,90, 120, 150, 180, and 210 minutes post-glucose administration as previously d e ~ c r i b e dNo . ~ sedatives were administered. Based on insulin response, ponies were classified into normal or abnormal groups by evaluating 1) the insulin to glucose (I/G) ratio at 90 minutes; 2) summing insulin and glucose concentrations at times 60,90, and 120 minutes and comparing the ratio (I/G sum); 3 ) calculating total insulin secreted (TIS); 4) determining insulin peak response (IPR) and glucose peak response from baseline (GPR); 5) calculating the insulinogenic index; and 6) comparing the total insulin secreted to total glucose ratio (TIS/GT) (Table 2). The IPR was defined as the highest concentration of insulin achieved above the baseline insulin concentration. The insulinogenic index was calculated as the ratio of the IPR to the greatest glucose increment above the fasting value.I6 Total insulin secretion (TIS) and glucose (GT) was determined by calculating the area under the curve from baseline to 2 10 minutes. Dexamethasone Suppression Testing
A dexamethasone suppression test was performed on all ponies as previously described.” Ponies were removed from pasture at 9.30 A.M. and not fed until the completion of the study 29 hours later. Water was available ad libitum. Dexamethasone (0.04 mg/kg) was administered
t Encore Chemistry System, Baker Instruments, Allentown, PA. $ Coat-a-Count, Diagnostic Products, Los Angeles, CA.
Vol. 5
. NO. 1 , 1991
17
TRIGLYCERIDE AND ENDOCRINE RESPONSE IN PONIES
TABLE 2. Insulin Response to Oral Glucose (lg/kg) Administration in 8 Ponies Pony ## Hyperinsulinemic 2 4
5 6 Normal I 3 7 8
IPR (uIU/ml)
GPR (mg/dU
Insulinogenic Index
Istl/G~o
Iso-tzo/G~o-lzo
Ratio
Ratio
TIS uIU/ml
GT (mg/dl)
TlS/GT Ratio
90.0 55.5 149.5 65.0
I I9
89 1 00 84
0.76 0.62 1.5 0.77
0.60 0.37 0.87 0.33
0.54 0.37 0.79 0.38
492.5 304.0 769.9 345.6
1094.0 946.3 1 1 10.2 959.8
0.45 0.32 0.69 0.36
39.9 17.2 38.0 19.7
I26 78 I13 I25
0.32 0.22 0.34 0.16
0.23 0.09 0.24 0.08
0.19 0.09 0.22 0.10
284. I 109.7 191.5 128.8
1298.5 1061.7 978.5 1145.2
0.22 0.10 0.20 0.1 I
IM at 2.30 P.M., 5 hours later. Blood was collected from the jugular vein before and at 15, 19, and 24 hours after dexamethasone adminjstration. Triglyceride (TG) and cholesterol (CH) and the concentration of the four lipoprotein fractions (total [TI, very low density [VLDL], low density [LDL], and high density [HDL]) were measured on the heparinized plasma.5 Total cholesterol was determined using a standardized modification of the Liebermann-Burchard reaction and read spectrophotometrically at 640 nm. Total cholesterol was considered adequate to illustrate the overall cholesterol trend in this study. Triglycerides were determined by a totally enzymatic method (hydrolysis by microbial lipase) and read spectrophotometrically at 530 nm. Separation of the lipoprotein fractions was performed by ultracentrifugation. Chylomicrons were recovered from the top layer following ultracentrifugation at 100,000 RPM for 10 minutes. Plasma samples were then ultracentrifuged at 100,000 RPM for a further 2.5 hours and the supranatant recovered for VLDLs. High density lipoproteins were assessed after initial precipitation of LDL and VLDL fractions with a nonmetallic polyionic precipitating agent, moderate centrifugation, and supernatant recovery. For chylomicrons, VLDL and HDL purity was verified by single band electrophoresis. Low-density lipoproteins were derived mathematically subtracting HDL from the combined HDL-LDL fraction. Serum cortisol determinations were made by radioimmunoassay using a commercially available kit7 validated for use in the horse. Fasling Studies: Experimental Design All ponies were weighed and kept in stalls on the day
before the start of the study. A complete pelleted feed ration11 was given at midday then withdrawn 3 hours later (3 P.M., and baseline samples were collected. Blood samples were collected 24,48,72 hours postfasting. One pony (# 1 ) died between 48 and 72 hours of fasting, and was not included in the fasting data analysis.
At 72 hours the ponies were divided into two groups. Group 1 (4 ponies) received dexamethasone IM (0.04 mg/kg) and group 2 (3 ponies) received 0.9% saline (placebo). Each group contained two hyperinsulinemic ponies. Blood samples were then collected at 80, 88, 96, 104, 112, and 120 hours of fasting. At each sampling time (0, 24, 48, 72, 80, 88, 96, 104, 112, 120 hours) during fasting plasma triglyceride and total cholesterol, plasma glucose, serum insulin, and serum cortisols were determined. In addition cholesterol and triglyceride concentrations in each of three major lipoprotein fractions (VLDL, LDL, HDL) were determined.
Statistical Analysis The data of the oral glucose tolerance test were analyzed using a repeated measures one-way analysis of variance and the Huyhn-Feldt test statistic to adjust for sphericity. A hypothesised abnormal group was compared with a normal group. Means were compared univariately using least square In the feed withholding experiment, groups were compared with a two-way repeated measures analysis of variance using a Huyhn-Feldt test statistic to adjust for sphericity.” During the first 72 hours of feed, withdrawal ponies with normal insulin responses (n = 3) were compared with hypothesised abnormal ponies (n = 4). In addition, repeated measures analysis of covariance using the Huyhn-Feldt test statistic to adjust for sphericity was performed on the drug versus placebo group during all 10 experimental times and during the six post-drug administration times (80 h-120 h). A covariate of the insulin to glucose sum was used to remove variation of hyperinsulinemic ponies. Means were compared univariately for groups within time and times within a group using least square Significance was reported at the P 5 0.05 level.”
ResuIts Oral Glucose Tolerance Test
0 Lipid ultracentrifugation,
Beckman Airfuge, Fullerton, CA. 7 Coat-a-Count, Diagnostic Products, Los Angeles, CA. 1) Horse Chow 200 Purina Mills Inc, St Louis, MO.
Ponies resting plasma glucose concentrations (range, 66-95 mg/dl) and serum insulin concentrations (range,
18
Journal of Veterinary Internal Medicine
FREESTONE ET AL.
3.8- 1 1.2 uIU/ml) were within reference ranges. Oral glucose absorption was normal in all ponies and the plasma glucose concentration doubled over baseline within 90 minutes. Four of eight ponies were hypothesised to be hyperinsulinemic in response to the administration of glucose orally based on calculated indices (Table 2). The I/G sum used to compare the normal versus the hyperinsulinemic ponies was significantly different (P < 0.03). In addition the I/G sum of the normal versus the abnormal ponies changed across time (P < 0.0002) and the groups changed differently over time (P < 0.02). Insulin concentrations in the abnormal ponies were significantly increased versus normal ponies at times 15, 30, 45, 60, 90, 120, 150, 180, and 210 minutes (Fig 1). In the hyperinsulinemic ponies, there was a rise in peak insulin concentration over baseline of 1,18496. In contrast, the normal pony peak insulins increased 522% above fasting concentrations (Fig I). There was no significant difference between the normal and abnormal ponies for glucose concentration (Fig 2).
225
T
504
:
:
:
:
:
0 15 30 45 60
:
: 90
:
: 120
:
: 150
:
: 180
:
I
210
TIME (rnins)
FIG.2. Glucose response to oral glucose administration ( 1 g/kg as a 20% solution) by hyperinsulinemic (n = 4) and normal (n = 4) ponies. Data expressed as mean SD.
*
between the normal and hyperinsulinemic ponies for insulin, glucose, cortisol, cholesterol, or plasma triglycerides.
Dexamethasone Suppression Test
Baseline cortisol for all ponies ranged between 4.4-9.8 ug/dl. In all ponies, the exogenous administration of dexamethasone-suppressed serum cortisol concentration and the cortisol concentration was still suppressed by 24 hours postadministration (Table 3). There were no changes in the lipid profile after dexamethasone administration (Table 3). Fasting During 72 Hours: Normal vs. Hyperinsulinemic Ponies
Fasting resulted in decreased glucose concentrations across time (P < 0.0001) and increased Total - TG (P < 0.0001), VLDL - TG (P< 0.0005), and Total - CH (P < 0.01) (Table 4). There were no significant differences
*r
***
125
hyperinsulinemic normal
9-9
Fasting-0
The withholding of feed followed by administration of either dexamethasone or a placebo at time 72 hour caused significant decreases in glucose concentration (P < 0.04) and increases in TG-T ( P < 0.000 l), VLDL-TG (P < 0.004) and total CH (P < 0.003) during the study. Glucose concentration was significantly different between the groups (P < 0.03), increasing after the administration of dexamethasone. Cortisol showed significant group (P < 0.03), group X time (P < 0.0008) and time (P < 0.04) effects, as dexamethasone decreased cortisol concentrations in the treated group (Table 5). Baseline glucose and insulin concentrations were elevated, associated with feeding 2 hours before the collection of the samples. Feed withdrawal during the first 72 hours resulted in a decline in plasma glucose and serum insulin concentration (Table 5). As glucose decreased serum insulin also decreased, while VLDL-TG was increased by 48 hours (P < 0.05) and remained increased (Table 5). Pony #I developed clinical signs of depression and ataxia between the 48- and 72-hour sampling and was found to have severe hypoglycemia (20 mg/dl) and hypertriglyceridemia (TG-T I060 mg/dl). No treatment was administered. The pony died before the 72-hour sample. Necropsy showed centrolobular congestion, diffuse periportal fatty change, hemosiderin accumulation in the liver, and catarrhal sinusitis. The liver changes were consistent with the hyperlipemia syndrome. The affected pony was in fair body condition and had a normal oral glucose tolerance test and dexamethasone suppression test (Table I).
25t 2 04
f : :
:
:
0 15 30 45 60
:
:
90
:
:
120
:
:
150
:
:
180
:
4
210
TIME (mins) FIG.I . Insulin response to oral glucose administration ( 1 g/kg, as a 20% solution) by hyperinsulinemic (n = 4) and normal (n = 4)ponies. Data expressed as mean f SD. Significant differences from normal ponies: * (PI 0.05), ** (PI0.00I), *** ( P I0.0001).
to 72 Hours: Dexamethasone vs. Placebo
VOl. 5
. No. 1. 1991
19
TRIGLYCERIDE AND ENDOCRINE RESPONSE IN PONIES
TABLE3. Alteration in Cortisol Concentration and Lipid Profile Following Dexamethasone (0.04 mg/kg, IM) in 8 Ponies Over Time Time (hr after dexamethasone) Np. of Ponies
0
15
19
24
0.28 f 0.1
0.29 f 0.1
0.3 f 0.1
8 Cortisol (ug/dl) Triglyceride (mg/dl) Total VLDL LDL H DL Cholesterol (mg/dl) Total
* Data expressed as mean
6.2 f 1.8* 78.5 f 11 .O 5.5 f 3.1 15.0 f 7.9 58.9 f 8.8
83.1 f 11.8 3.6 f 3.1 17.9 f 3.8 62.0 f 9.4
85.9 6.5 17.6 62.4
f 11.5 f 6.6 f 2.3 f 9.0
86.3 f 9.6 6.9 f 4.9 21.8 f 16.0 57.9 18.6
17.9 f 6.1
33.9
41.4
f
20.9
52.1 f 18.9
+_
14.9
*
f sd.
Feed Withdrawal Over 120 Hours: Post-Dexamethasone Response (t 80-120)
The administration of dexamethasone led to differences between the groups for only cortisol concentration (P < 0,008). Serum cortisol was suppressed after dexamethasone administration and differed between the groups at times 80 ( P < O.Ol), 88 (P < 0.001), 96 ( P < 0.001), and 104 ( P < 0.01) hours. Glucose concentration was increased over time ( P < 0.04). This increase was noted at the first sample collected 8 hours (t = 80 hours) after sample, following the administration of dexamethasone at 72 hours ( P < 0.05). The glucose concentrations for the two groups also changed differently across time ( P
