210 Scientific Reports: Original Study JAVMA, Vol 222, No. 2, January 15, 2003 EQUINE D uring the spring of 2001, an epidemic of abortions and stillbirths occurred in central Kentucky. This epi- demic of abortions was identified as mare reproductive loss syndrome (MRLS). 1 Although the magnitude of mor- bidity and mortality of the epidemic has not been com- pletely described, the experience of 2 of the authors (SEB, TWR) indicated that many of these abortions were in early gestation (ie, the first 90 days of gestation). The cause of MRLS remains unknown. Identification of fac- tors predisposing mares to having MRLS would provide clues of potential causal factors and a basis for developing preventive strategies. The objective of the study reported here was to describe results of a case-control study in which mares that had MRLS-associated early-term abor- tions (ETAs; early fetal losses) were compared with 2 groups of control mares to identify factors that could pre- dispose mares to having MRLS-associated ETAs. Materials and Methods Study population—Veterinarians from 2 large equine clin- ics a,b in Lexington, Kentucky, were asked to provide a list of farms to which they provided equine reproductive services. Horses that received care from veterinarians from these 2 clin- ics were the reference population for a study 2 of MRLS-associ- ated late-term abortions; although there was some overlap, the study populations and methods for the late-term abortion study and the study reported here were distinct. For each farm, vet- erinarians from the 2 clinics also provided rates of pregnancy losses for the farm (defined as the number of mares that were verified as pregnant at 28 days after breeding that subsequently lost the fetus divided by the number of mares verified as preg- nant at 28 days after breeding) during 2000 and 2001. The dis- tribution of loss rates for farms was examined, and farms were characterized as high-impact or low-impact. Farms in the upper 33rd percentile of pregnancy loss rate for 2001 or farms that had more than a 3-fold increase in pregnancy loss rate dur- ing 2001 relative to 2000 were considered severely affected by MRLS and categorized as high-impact farms. Farms in the lower 33rd percentile of pregnancy losses for 2001 were cate- gorized as low-impact farms. Categorization of farms was made prior to obtaining any data regarding predictors of outcome. At each high-impact farm, 5 mares that had ETAs during the peri- od between February 1 and July 1, 2001, were randomly select- ed as ETA cases (ETA group). Farm managers were instructed to identify mares that aborted during the first 90 days of preg- nancy and that were bred during the period from February to April 2001. When fewer than 5 such mares were available, mares bred during February, March, or April that aborted fetus- Case-control study of early-term abortions (early fetal losses) associated with mare reproductive loss syndrome in central Kentucky Noah D. Cohen, VMD, MPH, PhD, DACVIM; James G. Donahue, DVM, MPH, PhD; Vincent J. Carey, PhD; Janyce L. Seahorn, DVM, MS, DACVA, DACVIM, DACVECC; Donna Piercy, DVM; Judith K. Donahoe, BS; Deborah M. Williams, DVM; Stuart E. Brown II, DVM; Thomas W. Riddle, DVM Objective—To identify factors associated with abor- tions during early gestation classified as mare repro- ductive loss syndrome (MRLS). Design—Case-control study. Animals—324 broodmares from 43 farms in central Kentucky, including 121 mares from 25 farms that had early-term abortions (ETAs) associated with MRLS (case horses), 120 mares from the same farms but that did not abort, and 83 mares from 18 farms that were not severely impacted by MRLS. Procedure—Farm managers were interviewed to obtain data on various management practices and environmental exposures for the mares. Data for case and control horses were compared to identify risk fac- tors for mares having MRLS-associated ETAs. Results—Several factors increased the risk of MRLS- associated ETAs, including feeding hay in pasture, greater than usual amounts of white clover in pas- tures, more eastern tent caterpillars in pastures, abor- tion during a previous pregnancy, and sighting deer or elk on the premises. Conclusions and Clinical Relevance—Analysis indi- cates that certain characteristics of pastures predis- posed mares to MRLS-associated ETAs. Methods for limiting exposure to pasture (keeping mares in stalls longer) during environmental conditions similar to those of 2001 (ie, sudden freezing in mid-April follow- ing warmer-than-usual springtime temperatures and larger-than-usual numbers of eastern tent caterpillars in and around pastures) should reduce the risk of mares having MRLS-associated ETAs. It was not pos- sible to determine whether exposure to white clover or caterpillars were causal factors for MRLS or were merely indicators of unusual environmental conditions that resulted in exposure of mares to a toxic or infec- tious agent. (J Am Vet Med Assoc 2003;222:210–217) From the Department of Large Animal Medicine and Surgery, College of Veterinary Medicine, Texas A&M University, College Station, TX 77843-4475 (Cohen); 3640 Sequoia Trail, Verona, WI 53593 (Donahue); Channing Laboratory, Harvard Medical School, Boston, MA 02115 (Carey); the Livestock Disease Diagnostic Center, University of Kentucky, Lexington, KY 40511-1280 (Seahorn, Piercy, Donahoe, Williams); Hagyard-Davidson- McGee Associates, 4250 Iron Works Pike, Lexington, KY 40511-8491 (Brown); and Rood & Riddle Equine Hospital, PO Box 12070, Lexington, KY 40580-2070 (Riddle). Supported by the Grayson-Jockey Club Research Foundation, the Commonwealth of Kentucky, and the Link Equine Research Endowment. The authors acknowledge Drs. Summer Helbert, Shelly Lenz, Christopher Smith, and Michelle Jobert for technical assistance. Address correspondence to Dr. Cohen.