D. R. Sessions, S. E. Reedy, M. M. Vick, B. A. Murphy and B. P. FitzgeraldPreliminary implications regarding the estrous cycle

Development of a model for inducing transient insulin resistance in the mare:

2004, 82:2321-2328.J ANIM SCI 

http://www.journalofanimalscience.org/content/82/8/2321the World Wide Web at:

The online version of this article, along with updated information and services, is located on

www.asas.org

at Universiteit Utrecht on December 10, 2013www.journalofanimalscience.orgDownloaded from at Universiteit Utrecht on December 10, 2013www.journalofanimalscience.orgDownloaded from

Development of a model for inducing transient insulin resistance in the mare:Preliminary implications regarding the estrous cycle1,2

D. R. Sessions3, S. E. Reedy, M. M. Vick, B. A. Murphy, and B. P. Fitzgerald

Department of Veterinary Science, Maxwell Gluck Equine Research Center,University of Kentucky, Lexington 40546-0099

ABSTRACT: Peripheral insulin resistance is the fail-ure of proper cellular glucose uptake in response toinsulin. Insulin resistance and hyperinsulinemia areassociated with several disease states in the horse andreproductive function disturbances in humans, includ-ing polycystic ovarian syndrome. To test the hypothesisthat insulin resistance (IR) and hyperinsulinemia dis-rupt the estrous cycle in mares, two experiments wereconducted to first develop a model to induce IR and tothen examine the effect of this model on the durationof the estrous cycle. In Exp. 1, a hyperinsulinemic-eug-lycemic clamp (HEC) procedure was performed onseven mares to determine insulin sensitivity before andimmediately following infusion of a heparinized lipidsolution. The HEC procedure was repeated 1 wk afterlipid infusion. Mares developed IR following the lipidinfusion (P < 0.05), and some individuals maintained

Key Words: Equine, Free Fatty Acids, Insulin, Insulin Resistance, Lipid Infusion, Reproductive Cycle

2004 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2004. 82:2321–2328

Introduction

In the horse, obesity is associated with insulin resis-tance and may predispose individuals to the develop-ment of several pathologies, including laminitis (Coff-man and Colles, 1983). In addition, obesity and insulinresistance have been associated with disturbances inthe duration of the breeding season (Fitzgerald andMcManus, 2000) and the duration of the estrous cycle(Fitzgerald et al., 2002).

1The authors thank B. White for assistance with sample collections.We also thank L. Lawrence for the use of the YSI machine and A.True for the use of and technical assistance with the Cobas Ferra II.Finally, we thank G. Thomas and the North Farm crew for the careand maintenance of the mares.

2The research reported in this article is published in connectionwith a project of the Kentucky Agric. Exp. Stn. (03-14-158).

3Correspondence: 333 Gluck Equine Research Center (phone: 859-257-4757, ext. 8-1212; fax: 859-257-8542; e-mail: drsess0@uky.edu).

Received January 7, 2004.Accepted April 21, 2004.

2321

IR for up to 1 wk. Mares also exhibited increased bloodinsulin both immediately following treatment and 1 wklater (P < 0.05). In Exp. 2, induction of insulin resistanceby lipid solution was not accompanied by changes incirculating concentrations of luteinizing hormone, andduration of the luteal phase, compared with the dura-tion of untreated luteal phases. Nonetheless, lipid infu-sion and the resultant insulin resistance were associ-ated with an increased interovulatory period (P < 0.05),and peak concentrations of progesterone (P < 0.05) werehigher during the treated vs. untreated luteal phasesof the estrous cycle. The results from the preliminarystudy suggest that infusion of a lipid solution may in-duce transient insulin resistance and hyperinsuli-nemia. The resulting insulin resistance and hyperinsu-linemia may modify characteristics of the estrous cycle,perhaps at the level of the ovary.

The mechanisms that lead to disruption of the estrouscycle in obese mares remain to be elucidated; however,increased circulating concentrations of insulin disruptgonadotropin secretion and consequently reproductivefunction in many species, including mice (Bruning etal., 2000), pigs (Barb et al., 2001; Mao et al., 2001), andsheep (Bucholtz et al., 2000). In addition, in humanfemales, polycystic ovarian syndrome, a reproductivestate characterized by multiple anovulatory follicles(Pettigrew and Hamilton-Fairley, 1997), is associatedwith insulin resistance and hyperinsulinemia (Nest-ler, 2000).

There has been limited research on the role of obesityand insulin in the regulation of reproduction in mares.For this reason, the preliminary studies described inthis paper were designed to test the hypothesis thattransient insulin resistance and the resulting hyperin-sulinemia disrupt the estrous cycle. To better under-stand the significance of insulin resistance in reproduc-tive function in the mare, we developed a model for theinduction of transient insulin resistance. This model isbased on studies in several species, including humans,

at Universiteit Utrecht on December 10, 2013www.journalofanimalscience.orgDownloaded from

Sessions et al.2322

that demonstrated development of insulin resistancefollowing an intravenous infusion of a heparinized lipidsolution (Lee et al., 1988; Boden et al., 1995; Paolissoet al., 1995).

Materials and Methods

Horses

Light horse mares (5 to 15 yr of age) were selectedaccording to weight, lean body condition score (Hennekeet al., 1983), and low percentage of body fat as estimatedusing ultrasound of tailhead fat thickness (Kane et al.,1987). The Institutional Animal Care and Use Commit-tee approved all experimental procedures.

Experiment 1

Hyperinsulinemic-Euglycemic Clamp (HEC). Owingto high repeatability (Soop et al., 2000) an HEC proce-dure, validated for the horse (Powell et al., 2002), wasused to determine peripheral insulin sensitivity in allmares (n = 7, mean BW = 532 ± 51.41 kg; mean percent-age of body fat = 13.1 ± 4.7%). Briefly, following a 12-h fast, circulating concentrations of glucose were deter-mined at the beginning of the HEC procedure by useof a hand-held glucose meter (One Touch; Johnson andJohnson, New Brunswick, NJ). Hand-held meters werevalidated for use in the horse by demonstrating thatblood glucose values were similar to those measuredusing a glucose auto-analyzer (YSI 3000 STAT Plus;Yellow Springs Instrument Co. Inc., Yellow Springs,OH). A bolus injection of insulin (0.4 mU/kg crystallinebovine insulin; Sigma-Aldrich, St. Louis, MO) was ad-ministered (i.v.) and followed immediately by a constantinfusion of insulin (1.2 mU�kg−1�min−1) for 120 min. Twominutes following the start of insulin infusion a 50%(wt/vol), dextrose solution was infused simultaneously(30 mL/min), and the infusion rate was adjusted tomaintain euglycemia. Circulating concentrations ofglucose were determined every 5 min throughout theperiod of infusion. As all mares maintained euglycemiawithin the 120-min period, the rate of glucose infusionduring the final 30 min of the HEC procedure was usedto determine insulin sensitivity.

Experimental Design. One week after the initial HECprocedure (control period), each mare was infused (i.v.)with a heparinized (0.2�IU�kg−1�min−1) 20% (wt/vol)lipid emulsion (Liposyn II; Abbott Laboratories, NorthChicago, IL) at a rate of 2 mL/min for 4 h. Concurrentwith the start of the infusion, a 200-IU bolus of heparinwas also administered to facilitate lipolysis into freefatty acids (Orme and Harris, 1997). Immediately fol-lowing lipid infusion, another HEC procedure was exe-cuted to determine insulin sensitivity (treatment pe-riod). To confirm the anticipated transient insulin resis-tant condition, an additional HEC procedure wasperformed 1 wk after infusion of the lipid solution (re-covery period).

Collection of Blood Samples. Before each HEC proce-dure, blood samples were collected at 10-min intervalsfor 20 min before the HEC to determine physiologicalconcentrations of insulin and glucose. Thereafter, bloodsamples were collected at 10-min intervals to determineconcentrations of insulin. Additional blood sampleswere collected at 20-min intervals into evacuated tubes(Vacutainer Systems; Becton Dickson, Franklin, NJ)containing EDTA and used for determination of freefatty acids. Sampling for insulin and determination offree fatty acids occurred for 120 min, encompassing theduration of the HEC procedure. Blood samples wereimmediately centrifuged and the plasma harvested forsubsequent determination of FFA concentrations or al-lowed to clot overnight at 4°C. The next day, bloodsamples were centrifuged at 1,900 × g and the serumharvested and stored frozen for subsequent analysis ofinsulin concentration.

Experiment 2

Experimental Design. Mares were selected at randomfrom the general herd, and the HEC procedure wasperformed to determine insulin sensitivity. Mares thathad a glucose infusion rate of greater than 100 mL/hwere considered the most insulin sensitive and weretherefore selected for the experiment (n = 7). All maresserved as their own controls. Estrous cycles were syn-chronized (Loy et al., 1981), and mares completed onecontrol cycle (n = 7) approximately the first week ofMay. To decrease the duration of the experiment andthereby minimize the effects of environmental influ-ences on insulin sensitivity, mares were administeredprostaglandin F2α to initiate premature luteolysis. Dur-ing this “short cycling,” mares again underwent an HECthe first week of June, to ensure sustained insulin sensi-tivity. Ovarian follicular development was determinedat 2- to 3-d intervals by palpation per rectum and ultra-sonography. On identification of an ovarian follicle of30 mm or greater, each mare received a heparinizedlipid infusion. This infusion occurred approximately 2d (2.00 ± 1.5 d; n = 7) before ovulation. Immediately aftercessation of the lipid infusion, an HEC was executed toconfirm induced insulin resistance. An additional HECwas performed 1 wk later to determine duration of tran-sient insulin resistance.

Collection of Blood Samples. During the estrous cycle,before and after lipid infusion, blood samples were col-lected three times per week (Monday, Wednesday, andFriday). Samples were subsequently assayed for con-centrations of progesterone and LH. Changes in circu-lating concentrations of progesterone were used to iden-tify the occurrence of ovulation, the interval betweensuccessive ovulations (interovulatory interval) and theduration of the luteal phase in each estrous cycle.

Hormone and FFA Determination

Circulating concentrations of insulin were deter-mined by RIA (Coat-A-Count; Diagnostic Products

at Universiteit Utrecht on December 10, 2013www.journalofanimalscience.orgDownloaded from

Insulin resistance and equine estrous cycle 2323

Figure 1. Changes in mean glucose infusion rates (±SEM; n = 7) during a hyperinsulinemic-euglycemic clampprocedure performed for the control (�), immediately following treatment (�), and recovery (�) periods for Exp. 1.The period immediately following treatment differed from both the control and recovery periods (P < 0.05).

Corp., Los Angeles, CA), as described elsewhere (Powellet al., 2002). Intra- and interassay CV of pooled sampleswere 3.8 and 11.0%, respectively (n = 5 assays). Detec-tion limits for the insulin assays were approximately2.19 �IU/mL. Progesterone was determined using aRIA described previously (Silvia et al., 1992). Intra-and interassay CV of pooled samples were 2.9 and 6.1%,respectively (n = 4 assays). The limit for detection forthe progesterone assay was 0.04 ng/mL. Concentrationsof LH were determined by a double antibody RIA asdescribed by Thompson et al. (1986). Intra- and in-terassay CV for pooled samples were 4.46 and 18.7%,respectively (n = 4 assays). The detection limit of theLH assays was approximately 0.90 ng/mL. Free fattyacids were determined by an in vitro enzymatic colori-metric method using a nonesterified fatty acid kit(NEFA C; Wako Chemicals Inc., Richmond, VA)adapted to a Cobas Fera II, a semiautomated spectro-photometer (Eisemann et al., 1988). Intra- and in-terassay CV on for the prepared pools were 2.3 and6.9%, respectively. The detection limit for FFA analysiswas 62.03 �mol/L.

Statistical Analyses

Data are presented as means ± SEM. For Exp. 1,mean glucose infusion rates (GIR), and pre-HEC con-centrations of insulin and free fatty acid were analyzedusing repeated measures ANOVA employing the proce-

dure of Satterthwaite for degrees of freedom using SAS(SAS Inst. Inc., Cary, NC). For Exp. 2, glucose infusionrates and concentrations of insulin were also evaluatedusing repeated measures ANOVA. Interovulatory in-tervals, luteal phase durations, peak and mean concen-trations of progesterone, and mean concentrations ofLH were compared between control and treated cyclesby Student’s paired t-test. In all instances with the useof the repeated measures ANOVA, time was the fixedeffect, and the mares were the random effect; conse-quently, these were the only sources of variation in themodel. There was no interaction specified in the modelbecause there is only one fixed effect. The error termused to test the main effects was residual error.

Results

Experiment 1

Glucose Infusion Rates. As depicted in Figure 1, therewas a significant difference in glucose infusion ratesbetween treatment periods (P < 0.05). Analysis of differ-ences of least squares means indicated that changeswere observed between the control and treatment HECprocedures (2.30 ± 0.29 mg�kg−1�min−1 vs. 1.41 ± 0.25mg�kg−1�min−1; P < 0.05). Although lipid infusion sig-nificantly modified group insulin sensitivity, the GIRin one mare was greater than pretreatment values.

at Universiteit Utrecht on December 10, 2013www.journalofanimalscience.orgDownloaded from

Sessions et al.2324

Figure 2. Changes in mean concentrations of insulin(±SEM; n = 7) in blood samples collected immediatelybefore a hyperinsulinemic-euglycemic clamp procedureperformed during the control, treatment, and recoveryperiods during Exp. 1. Means for each period with differ-ent letters differ (P < 0.05).

Overall, mean glucose infusion rates did not differbetween control and recovery periods (2.02 ± 0.54 vs.2.29 ± 0.30 mg�kg−1�min−1). However, 1 wk after infu-sion of the lipid solution, three of six mares that werepreviously identified as insulin resistant immediatelyfollowing treatment were found to have maintained aninsulin-resistant state compared with pretreatment.One mare that did not display decreased insulin sensi-tivity immediately following lipid treatment exhibiteddecreased insulin sensitivity 1 wk after treatment. Col-lectively, therefore, infusion of a heparinized lipid solu-tion was accompanied by development of insulin resis-tance in all seven animals between 1 to 7 d aftertreatment.

Concentrations of Insulin. Concentrations of insulinincreased following treatment as illustrated in Figure2 (P < 0.01). Mean concentrations of insulin tended tobe higher for the treated vs. control period (10.11 ± 1.46vs. 7.15 ± 2.73 �IU/mL; P = 0.18). Circulating insulinvalues at the time of the recovery were markedly higherthan pretreatment concentrations (18.78 ± 3.20 vs. 7.15± 2.73 �IU/mL; P < 0.01).

Concentrations of Free Fatty Acids. Mean (± SEM)concentrations of free fatty acids were significantlyhigher following treatment compared with control val-ues (1.97 ± 0.24 vs. 0.57 ± 0.22 mmol/L; P < 0.001; Figure3A). However, at the time of the recovery period, HECmean concentrations of free fatty acids did not differfrom control values (0.61 ± 0.11 vs. 0.57 ± 0.22 mmol/L). Mares that maintained decreased insulin sensitivityat the time of the recovery HEC also showed a decreasedability of insulin to stimulate FFA uptake, whereasmares that were sensitive at the time of the recoveryshowed an increased response to insulin (Figure 3B).

Figure 3. Changes in mean concentrations of plasmaFFA during Exp. 1 (n = 7). Panel A illustrates mean(±SEM) concentrations during the hyperinsulinemic-eug-lycemic clamp for the control (�), treatment (�), andrecovery (�) periods. The treatment period differed fromboth the control and recovery periods (P < 0.001). PanelB illustrates concentrations of FFA in mares identified tobe insulin sensitive (�, n = 2) or resistant (▲, n = 5)during the recovery period during Exp. 1.

Experiment 2

Glucose Infusion Rates. Glucose infusion rates for thecontrol cycle, treated cycle preinfusion and treated cycleHEC 1 wk posttreatment were not different (Figure 4).The control cycle, treated cycle pre-infusion, and 1 wkafter heparinized lipid infusion glucose infusion rateswere 2.04 ± 0.19 mg�kg−1�min−1, 2.12 ± 0.26 mg�

kg−1�min−1, and 2.10 ± 0.39 mg�kg−1�min−1 respectively.However, as expected, the GIR immediately followingtreatment was lower than preinfusion values (1.49 ±0.29 vs. 2.12 ± 0.26 mg�kg−1�min−1; P < 0.05).

Concentrations of Insulin. Pre-HEC concentrations ofinsulin were different following treatment with hepa-rinized lipid (P < 0.05). However, this varied from theresults observed in Exp. 1, in that the greatest increasein concentrations of insulin was observed immediatelyfollowing treatment compared with control values (4.92± 0.90 vs. 3.53 ± 0.73 �IU/mL) rather than at the time

at Universiteit Utrecht on December 10, 2013www.journalofanimalscience.orgDownloaded from

Insulin resistance and equine estrous cycle 2325

Figure 4. Mean glucose infusion rates (±SEM) duringthe hyperinsulinemic-euglycemic clamp for the controlcycle (�), treated cycle prelipid infusion (�), treated cycleimmediately following treatment (�), and 1 wk followingtreatment (▲) for Exp. 2. Glucose infusion rates for theperiod immediately following treatment differed fromthe control period, the treated cycle preinfusion, and theperiod 1 wk following treatment (P < 0.05).

of the recovery. Circulating insulin values at the timeof the recovery had decreased to control levels (3.15 ±0.43 vs. 3.53 ± 0.73 �IU/mL).

Duration of the Luteal Phase and Circulating Concen-trations of Progesterone. A trend toward a lengthenedluteal phase was observed in the estrous cycle followinginfusion with a heparinized lipid solution comparedwith the untreated cycle. The duration of the lutealphase of the cycle was lengthened in three of sevenmares after treatment compared with the control cycle(mean durations 19.57 ± 2.66 vs. 15.57 ± 0.97 d, P =0.09, n = 7). However, of the seven treated estrous cy-cles, three displayed lengthened interovulatory inter-vals (IOI) compared with the control cycles and themean duration of treated cycles was longer than controlcycles (26.0 ± 2.41 vs. 20.29. ± 0.78 d, n = 7; P < 0.05).

Mean concentrations of progesterone are depicted inFigure 5. There was no significant difference in meanconcentrations of progesterone between treated andcontrol cycle (7.46 ± 1.99 vs. 5.64 ± 1.39 ng/mL; n = 7).Also depicted in Figure 5 is peak progesterone duringthe luteal cycle. During the estrous cycles of treatmentmares, significantly higher peak progesterone valueswere demonstrated compared with control cycles (12.87± 4.03 vs. 8.78 ± 2.23 ng/mL; n = 7; P < 0.05).

Luteinizing Hormone. Mean concentrations of LH didnot differ for the duration of the estrous cycle that fol-lowed lipid infusion compared with the control cycle(3.95 ± 0.34 vs. 3.29 ± 0.13 ng/mL). Variation existedamong mares as to differences in concentrations of LHbetween the control and treated cycles (Figure 6). Onemare (Mare 59) exhibited a markedly higher rise in LHbefore ovulation during the cycle following treatment.

Figure 5. Comparison of overall mean (±SEM) and peakconcentrations of progesterone during the luteal phaseof the estrous cycle for control (n = 7) and treated (n =7) estrous cycles during Exp. 2. Mean concentrations ofprogesterone were not different. Means for each treat-ment group with different letters differ for peak concen-trations (P < 0.05).

Discussion

The current study demonstrates that infusion of aheparinized lipid solution resulted in decreased insulinsensitivity in adult horses. In agreement with findingsin other species (Boden et al., 1995; Paolisso et al.,1995), development of insulin resistance was accompa-nied by reduced glucose infusion rates during the HECprocedure and increased insulin concentrations imme-diately preceding the HEC. However, the timing of thedevelopment of insulin resistance appeared variablebecause one mare displayed insulin resistance onlyafter 7 d after treatment and not immediately aftertreatment. In human studies, a 3- to 4-h delay in thedevelopment of insulin resistance has been reported(Boden et al., 2001). To our knowledge, this is the firstreport in any species that the development of insulinresistance may occur several days after infusion of alipid solution and, furthermore, that resistance may besustained for 1 wk after treatment. An explanation forthe latter observation is not forthcoming; however, itis unlikely that the maintenance of insulin resistancewas caused by increased circulating concentrations offree fatty acids 1 wk after lipid infusion. At this timepoint, circulating concentrations of FFA were similarto control values. It has been suggested that in humansthe timing of insulin resistance following infusion oflipids coincides with an increase in skeletal musclehexo-phosphates, whereas there is a decrease of glu-cose-6-phosphate formation, indicating that the FFAeffect on peripheral insulin action is through an in-creased flux of UDP-N-acetyl-hexosamines into the glu-cosamine pathway (Hawkins et al., 1997). It is conceiv-

at Universiteit Utrecht on December 10, 2013www.journalofanimalscience.orgDownloaded from

Sessions et al.2326

Figure 6. Luteinizing hormone profiles for the control(�, n = 7) and treated (�, n = 7) estrous cycles for Exp.2. Luteinizing hormone was determined in blood samplescollected every 2 to 3 d. The date of ovulation is desig-nated time 0. Infusion of the lipid solution occurred 2 dbefore ovulation.

able, therefore, that the maintenance of insulin resis-tance in some mares 1 wk after lipid infusion reflectsmaintenance of hexosamine end product build-up. Analternative explanation is that it might reflect differ-ences in concentration of intramuscular triglyceride.The accumulation of intramuscular triglyceride variesamong subjects and is independent of total body fat(Phillips et al., 1996). It has been demonstrated thatin fit men with similar body mass indexes, subjects withhigher concentration of intramuscular triglycerides,more specifically intramyocellular lipid, had decreasedwhole-body glucose uptake in response to insulin (Vir-kamaki et al., 2001). These subjects also displayed aninability of insulin to decrease circulating concentra-tions of FFA, consistent with the response of the maresthat maintained insulin resistance 1 wk following treat-ment with a lipid solution. Therefore, mares that areobese may not have high concentrations of intramuscu-lar triglycerides, whereas mares in a lean body condi-tion may have higher levels, thereby accounting for thelack of a consistent relationship between the body fatpercentage and the maintenance of insulin resistance.

In several species, the model of induction of transientinsulin resistance described in this study may providean opportunity to investigate the association betweeninsulin resistance and several disease states. In thisregard, current animal models of insulin resistancedemonstrate disturbances in cytokine and hormoneproduction in association with insulin resistance. Con-ceivably, similar studies in the horse may provide newinsight into the relationship between insulin resistanceand the development of laminitis, osteochondrosis des-sicans lesions, and other disorders. However, additionalstudies should be conducted to determine more specifi-cally the duration of the induced insulin resistance.

Increased circulating concentrations of insulin havebeen shown to affect reproductive activity by modifica-tion of GnRH-mediated LH release (Bruning et al.,2000), and also by direct affects on the ovary (Diamanti-Kandarakis and Bergiele, 2001; Mao et al., 2001). Ex-periment 2 was a preliminary study conducted to testthe hypothesis that insulin resistance and associatedhyperinsulinemia affect the estrous cycle in the mare.Following induction of insulin resistance by infusion ofa heparinized lipid solution, the mean duration of lutealphase of the succeeding estrous cycle was marginallylengthened compared with the control cycle. Addition-ally, the time between successive ovulations followinglipid infusion was lengthened significantly. The mecha-nisms that underlie the increased interval between suc-cessive ovulations are unknown, but it is likely to reflecta combination of a marginal increase in the duration ofthe luteal phase of the treatment estrous cycle, togetherwith a nonsignificant increase in the duration of thefollicular phase of the succeeding estrous cycle.

In human females, long-term insulin resistance andhyperinsulinemia is frequently associated with an in-crease in the duration of the follicular phase of themenstrual cycle and is particularly well documented in

at Universiteit Utrecht on December 10, 2013www.journalofanimalscience.orgDownloaded from

Insulin resistance and equine estrous cycle 2327

patients with polycystic ovarian syndrome. The mecha-nism by which this occurs may reflect an arrest of follic-ular development. Insulin can play a role in this phe-nomenon by enhancement of the effects of FSH on folli-cles that have acquired LH receptors. The prevailinghypothesis is that small follicles that have just attainedLH receptors, in the presence of insulin, display en-hanced estradiol production equal to mature follicles,thereby inhibiting further growth and arrest of the folli-cles in the immature stage (Diamanti-Kandarakis andBergiele, 2001). It is possible that some mares showlonger follicular phases because follicular developmenthas been slowed via this mechanism. In the currentstudy, however, circulating concentrations of LH werenot significantly different between the follicular andluteal phases of treated and untreated cycles. This lat-ter observation might suggest that in the short-termany action by insulin is unlikely to involve a change inthe hypothalamic-pituitary axis.

An alternative site of action by insulin might be theovary. In this regard, hyperinsulinemia accompaniedby increased steroid secretion by the ovary is observedin humans as noted in women with polycystic ovariansyndrome (Nestler, 2000). Observations from Exp. 2support these findings because concentrations of pro-gesterone were slightly increased following treatment.Peak concentrations of progesterone were significantlyhigher during the cycle following treatment with FFAand development of insulin resistance compared withthe control cycle. However, because samples were takenonly three times per week, the values representing peakprogesterone may not represent the true peak, as wouldhave been detected by sampling daily. Because thesestudies were preliminary, it was determined to be suffi-cient to collect samples three times per week. Anotherexplanation for the differences observed in concentra-tions of progesterone could be due to environmentalconditions, such as photoperiod, varying from the con-trol to treated cycle. However, the administration ofprostaglandin F2α to “short-cycle” the mares was em-ployed to minimize these effects. The observation thatinsulin resistance and hyperinsulinemia was accompa-nied by increased steroid secretion by the ovary is con-sistent with observations made in other species. Onesuch study was conducted in pigs. In gilts, feed restric-tion during follicular development depressed plasmaprogesterone; however, when animals were supple-mented with insulin during feed restriction, the de-creased concentrations of progesterone did not occur(Mao et al., 2001). Another study conducted (Cox et al.,1994) found that estradiol was decreased in diabeticpigs. Similar to the findings of Exp. 2, there was nochange in LH secretion, indicating that the absence ofinsulin decreased steroidogenesis. This observation inpigs and other species support the proposal that insulinexerts a stimulatory effect on progesterone productionby the corpus luteum, and this effect is mediated bychanges that occur in the developing follicle before ovu-lation.

Implications

Infusion of a heparinized lipid solution is a safe, reli-able means for inducing insulin resistance and hyperin-sulinemia in mares. The resulting hyperinsulinemiamay modify follicular development and luteal functionat the ovarian level rather than the hypothalamic pitu-itary level, as evidenced by no changes in luteinizinghormone. However, further studies are needed to deter-mine whether a critical period exists in follicular devel-opment, during which increases in insulin are influen-tial, and by what mechanism insulin exerts its effects.Finally, this model of inducting transient insulin resis-tance provides an opportunity to investigate the associ-ation between insulin resistance and many diseases.Current animal models of insulin resistance demon-strate cytokine and hormone production disturbancesassociated with insulin resistance. Similar studies inhorses may provide insight into the relationship be-tween insulin resistance and development of laminitis,osteochondrosis dessicans lesions, and other disorders.

Literature Cited

Barb, C. R., R. R. Kraeling, and G. B. Rampacek. 2001. Nutritionalregulators of the hypothalamic-pituitary axis in pigs. Reprod.Suppl. 58:1–15.

Boden, G., X. Chen, J. Rosner, and M. Barton. 1995. Effects of a48-h fat infusion on insulin secretion and glucose utilization.Diabetes 44:1239–1242.

Boden, G., B. Lebed, M. Schatz, C. Homko, and S. Lemieux. 2001.Effects of acute changes of plasma free fatty acids on intramyo-cellular fat content and insulin resistance in healthy subjects.Diabetes 50:1612–1617.

Bruning, J. C., D. Gautam, D. J. Burks, J. Gillette, M. Schubert, P.C. Orban, R. Klein, W. Krone, D. Muller-Wieland, and C. R.Kahn. 2000. Role of brain insulin receptor in control of bodyweight and reproduction. Science 289:2122–2125

Bucholtz, D. C., A. Chiesa, W. N. Pappano, S. Nagatani, H. Tsuka-mura, K. I. Maeda, and D. L. Foster. 2000. Regulation of pulsatileluteinizing hormone secretion by insulin in the diabetic malelamb. Biol. Reprod. 62:1248–1255.

Coffman, J. R., and C. M. Colles. 1983. Insulin tolerance in laminiticponies. Can. J. Comp. Med. 47:347–351.

Cox, N. M., K. A. Meurer, C. A. Carlton, R. C. Tubbs, and D. P.Mannis. 1994. Effect of diabetes mellitus during the luteal phaseof the oestrous cycle on preovulatory follicular function, ovula-tion and gonadotrophins in gilts. J. Reprod. Fertil. 101:77–86.

Diamanti-Kandarakis, E., and A. Bergiele. 2001. The influence ofobesity on hyperandrogenism and infertility in the female. Obes.Rev. 2:231–238

Eisemann, J. H., G. B. Huntington, and C. L. Ferrell. 1988. Effectsof dietary clenbuterol on metabolism of the hindquarters insteers. J. Anim. Sci. 66:342–353.

Fitzgerald, B. P. and C. J. McManus. 2000. Photoperiodic versusmetabolic signals as determinants of seasonal anestrus in themare. Biol. Reprod. 63:335–340.

Fitzgerald, B. P., S. E. Reedy, D. R. Sessions, D. M. Powell, and C.J. McManus. 2002. Potential signals mediating the maintenanceof reproductive activity during the non-breeding season of themare. Reprod. Suppl. 59:115–129.

Hawkins, M., N. Barzilai, R. Liu, M. Hu, W. Chen, and L. Rossetti.1997. Role of the glucosamine pathway in fat-induced insulinresistance. J. Clin. Invest. 99:2173–2182.

Henneke, D. R., G. D. Potter, J. L. Kreider, and B. F. Yeates. 1983.Relationship between condition score, physical measurementsand body fat percentage in mares. Equine Vet. J. 15:371–372.

at Universiteit Utrecht on December 10, 2013www.journalofanimalscience.orgDownloaded from

Sessions et al.2328

Kane, R. A., M. Fisher, D. Parrett, and L. M. Lawrence. 1987. Estimat-ing Fatness in Horses. Pages 127-131 in Proc. 10th Equine Nutr.Physiol. Symp.

Lee, K. U., H. K. Lee, C. S. Koh, and H. K. Min. 1988. Artificialinduction of intravascular lipolysis by lipid-heparin infusionleads to insulin resistance in man. Diabetologia 31:285–290.

Loy, R. G., R. Pemstein, D. O’Canna, and R. H. Douglas. 1981. Controlof ovulation in cycling mares with ovarian steroids and prosta-glandins. Theriogenology 15:191–200.

Mao, J., B. K. Treacy, F. R. Almeida, S. Novak, W. T. Dixon, and G.R. Foxcroft. 2001. Feed restriction and insulin treatment affectsubsequent luteal function in the immediate postovulatory pe-riod in pigs: progesterone production in vitro and messengerribonucleic acid expression for key steroidogenic enzymes. Biol.Reprod. 64:359–367.

Nestler, J. E. 2000. Obesity, insulin, sex steroids and ovulation. Int.J. Obes. Relat. Metab. Disord. 24(Suppl 2):S71–S73.

Orme, C. E., and R. C. Harris. 1997. A comparison of the lipolyticand anticoagulative properties of heparin and pentosan polysul-phate in the thoroughbred horse. Acta Physiol. Scand.159:179–185.

Paolisso, G., A. Gambardella, L. Amato, R. Tortoriello, A. D’Amore,M. Varricchio, and F. D’Onofrio. 1995. Opposite effects of short-and long-term fatty acid infusion on insulin secretion in healthysubjects. Diabetologia 38:1295–1299.

Pettigrew, R., and D. Hamilton-Fairley. 1997. Obesity and femalereproductive function. Br. Med. Bull. 53:341–358.

Phillips, D. I., S. Caddy, V. Ilic, B. A. Fielding, K. N. Frayn, A. C.Borthwick, and R. Taylor. 1996. Intramuscular triglyceride andmuscle insulin sensitivity: evidence for a relationship in nondia-betic subjects. Metabolism 45:947–950.

Powell, D. M., S. E. Reedy, D. R. Sessions, and B. P. Fitzgerald. 2002.Effect of short-term exercise training on insulin sensitivity inobese and lean mares. Equine Vet. J. 34:81–84.

Silvia, P. J., L. Johnson, and B. P. Fitzgerald. 1992. Changes in thehypothalamic-hypophyseal axis of mares in relation to the wintersolstice. J. Reprod. Fertil. 96:195–202.

Soop, M., J. Nygren, K. Brismar, A. Thorell, and O. Ljungqvist. 2000.The hyperinsulinaemic-euglycaemic glucose clamp: reproduc-ibility and metabolic effects of prolonged insulin infusion inhealthy subjects. Clin. Sci. (Lond.) 98:367–374.

Thompson, D. L., Jr., L. Johnson, R. L. St George, and F. Garza,Jr. 1986. Concentrations of prolactin, luteinizing hormone andfollicle stimulating hormone in pituitary and serum of horses:effect of sex, season and reproductive state. J. Anim. Sci.63:854–860.

Virkamaki, A., E. Korsheninnikova, A. Seppala-Lindroos, S. Vehka-vaara, T. Goto, J. Halavaara, A. M. Hakkinen, and H. Yki-Jarvinen. 2001. Intramyocellular lipid is associated with resis-tance to in vivo insulin actions on glucose uptake, antilipolysis,and early insulin signaling pathways in human skeletal muscle.Diabetes 50:2337–2343.

at Universiteit Utrecht on December 10, 2013www.journalofanimalscience.orgDownloaded from

Referenceshttp://www.journalofanimalscience.org/content/82/8/2321#BIBLThis article cites 26 articles, 12 of which you can access for free at:

Citationshttp://www.journalofanimalscience.org/content/82/8/2321#otherarticlesThis article has been cited by 6 HighWire-hosted articles:

at Universiteit Utrecht on December 10, 2013www.journalofanimalscience.orgDownloaded from