Prevalence, risk factors and genetic parameters of cresty neck in Pura Raza Español horses

Prevalence, risk factors and genetic parameters of cresty neck in Pura Raza Español horses M. J. SÁNCHEZ, P. J. AZOR, A. MOLINA†, T. PARKIN‡, J. L. L. RIVERO++ and M. VALERA* Department of Agro–Forestry Sciences, ETSIA, University of Seville, Road Utrera Km 1, 41013 Seville, Spain. †Department of Genetics, University of Cordoba, Road Madrid–Cadiz, Km 396, 14071 Cordoba, Spain. ‡ School of Veterinary Medicine, College of Medical, Veterinary and Life Sciences, University of Glasgow , UK. ++Department of Comparative Anatomy and Pathological Anatomy, University of Cordoba, Road Madrid–Cadiz, Km 396, 14071 Cordoba, Spain. Keywords: crest; heritability; neck adiposity; obesity; body condition *Correspondence e–mail: [email protected] Ethical animal research The study was approved by the Ethical Committee of the University of Seville, Spain. The National Association of Pura Raza Español Horse Breeders (ANCCE) provided the data for this study. Authors’ declaration of interests No competing interests have been declared for any authors. Sources of funding The study was supported by grants of the ANCCE and project P10–AGR–5963, Junta de Andalucía, Spain. Acknowledgements The authors wish to thank the National Association of Pura Raza Español Horse Breeders (ANCCE) for providing the data used in this study. Authorship M. Valera and M.J. Sánchez designed the study. M.J. Sánchez, M. Valera and A. Molina were involved in the study execution and preparation of the manuscript. P.J. Azor was involved in data collection. All the authors participated in the analysis of data, and approved the final version of the manuscript. Word count: 4489 Prevalence, contributing factors and genetic parameters of cresty neck in Pura Raza Español horses Keywords: crest; heritability; neck adiposity; obesity; body condition Summary Reasons for performing study Cresty neck is a relatively common morphological defect in Pura Raza Español horses and other Baroque type horse breeds, which adversely affects breeding industry. Objective: To establish the within–breed prevalence, possible associated factors and heritability of cresty neck in Pura Raza Español horses. Study design: Cross-sectional analysis of a large worldwide database of Pura Raza Español horses. Methods: The database included evaluations of 10,929 Pura Raza Español horses from 24 countries. Cresty neck score, on a nine point scale, girth–to–height ratio, height at the withers, length of neck, head–neck junction and neck–body junction were recorded. A Bayesian genetic animal model included the following systematic effects: girth–to–height ratio, age, sex, coat colour and geographical area of the stud. Animal and residual effects were included as random effects. Results: Within this Pura Raza Español population, 8.91% had a cresty neck score ≥5, which is the threshold for penalty or disqualification in the stud–book of the breed. The age, sex, coat colour and stud geographical area were significantly associated with cresty neck score. The estimated heritability co-efficient for cresty neck score was 0.37 (s.d.=0.034) and genetic correlation between cresty neck score and other conformational traits ranged from –0.06 (height–at–withers) to –0.21 (neck–body junction). Conclusions: Crest neck is a prevalent defect in Pura Raza Español horses, associated with age, sex, coat colour and other conformational traits, with a moderate level of heritability. Breeding to select against this condition may therefore be beneficial in this breed. Introduction Cresty neck [1] affects a large number of Pura Raza Español horses and other “Baroque” type horse breeds with a long neck and abundant mane (e.g., Lusitano, Morgan, Paso Fino). Although causes underlying this condition are currently unknown, it seems to be multifactorial in nature but may relate to dysfunction of the neck muscles and accumulation of localized fat deposits in the dorsal neck [2]. Expansion of adipose tissues within the neck region may be associated with altered metabolic states such as insulin resistance [3,4]. Although there is a clear correlation between obesity and insulin resistance [5], some insulin–resistant animals were not obese, despite having enlarged fat deposits on the neck [4]. The predisposition to obesity and some phenotypic characteristics (e.g., thick necks) of Pura Raza Español horses have been associated with insulin resistance [2,6], and insulin sensitivity is lower in PRE than in Standardbreds independent of obesity [7]. Pura Raza Español horses are predisposed to equine metabolic syndrome [7] and laminitis [8]. Cresty neck decreases the value of horses at sales and high scores are a disqualifying morphological defect in the stud–book [9]. Equine health and value is influenced by conformational traits [10,11]. Cresty neck score has been monitored in the whole in Pura Raza Español horse population since 2010 [9]. However, there are no previous studies on its prevalence, potential associated factors and genetic parameters of cresty neck in Pura Raza Español horses. We hypothesized that both environmental and genetic factors play a significant role in the development of this condition. A better understanding of how these factors contribute to the development of cresty neck is crucial to improving both the management and breeding selection of horse breeds at risk of developing this problem. The specific aims of this work were to: 1. Determine the within–breed prevalence of cresty neck in a worldwide population of Pura Raza Español horses. 2. Establish whether the development of cresty neck in this breed is associated with girth–to– height ratio, age, sex, coat colour and geographical location of the stud where the horse lives. 3. Estimate heritability of cresty neck and genetic correlations with specific conformational traits. Material and methods Description of the traits and database All the Pura Raza Español horses recorded as attending the basic cresty neck grading between 2010 and 2014 from 24 countries worldwide were included in the study. The database included records of cresty neck score, girth–to–height ratio, height at the wither, length of neck, head– neck junction and neck–body junction) of 10,929 animals (4,004 stallions and 6,925 mares) with a mean age of 4.42 years (range, 3–21). The cresty neck trait was analysed on a linearized scale by using an adaptation of the scoring system proposed by Carter et al. [1]. This scale includes nine scores in which the extremes represent the biological limits for this morphological trait and the 1-9 cresty neck score scale was believed to approximate to a continuous linear scale (Supplementary Table 1). Cresty neck scoring was carried out by twenty independent and previously trained appraisers, familiar with the routine basic grading of the linear score evaluation of Pura Raza Español horses, and a mean of two scores was used in each horse. Scores 5, 6 and > 7 were considered by the breed authorities to be serious, very serious and disqualifying defects, respectively. The following morphological traits were also examined on a horse–by–horse basis:  The girth–to–height ratio was estimated according to Carter et al. [1] and used as an estimator of body condition, since it is strongly associated with the Henneke’s body condition scoring system [1,11].  The height was measured as the vertical length between the highest point of the withers and the ground with the horse in standing position.  The length of the neck was measured as the distance between the tuber of the spina scapulae and the base of the auricular region.  Head–neck junction score (type of insertion of the head in the neck ranging from very distinct to very indistinct) and neck–body junction score (distance between the point of insertion of the neck to the trunk (the ventral part) and the line connecting the two shoulder joints ranging from <2cm to >23cm) were estimated by a linear scale of 9 classes according to Sánchez et al. [12] (Supplementary Table 2). All measurements were taken from the left side of the horse while it was standing on a hard surface and flat ground, assuming a natural position. The horses were positioned for measurement with the front legs and hind feet parallel and as near to perpendicular as possible; the toes were in line. No sedatives were used. All these morphological traits were determined by two independent and experienced appraisers and a mean of both scores was used. Statistical and genetic model The descriptive statistics for each trait (cresty neck score, girth–to–height ratio, height, length of neck, head–neck junction score, neck–body junction score) were analysed using the package Statistica 8.0 for Windows [13]. To estimate the association of girth–to–height ratio and conformational traits, individual horses were classified as moderate (girth–to–height ratio ≤1.26), overweight (>1.26 and <1.29) or obese (≥1.29) [1]. Potential risk factors (girth–to– height ratio, age, sex, coat colour and stud geographical area) were analyzed using a General Linear Model followed by a post–hoc test of least squared means except for assessing the stud geographical area effect, where a Dunnett’s post–hoc test was performed taking horses belonging to Spanish studs as the control group. Countries with fewer than 20 horses were grouped by continent. Pedigree information was collected from the official Pura Raza Español stud–book. A minimum of four generations were considered in this analysis giving a total of 41,055 horses (12,546 stallions and 28,509 mares). Heritability coefficients (h2) and genetic correlations between cresty neck score and other morphological traits were estimated using BLUP evaluation [14] based on a multiple-trait animal model using a Bayesian approach with TM software [15]. Currently, most national genetic evaluation systems for several domestic livestock species are based on animal or random regression models using BLUP [16]. With the BLUP method, the effects of systematic factors and the breeding values are estimated simultaneously by solving a large set of equations with equally as many unknown solutions [17]. The equation in matrix notation to solve the mixed model was: y= Xb+Zu +e, and had: 2 2 u e N                    u0 A0 , e0 0I where y is the vector of observations, X the incidence matrix of systematic effects, Z the incidence matrix of animal genetic effects, b the vector of systematic effects, u the vector of direct animal genetic effects, e the vector of residuals, u2 the direct genetic variance, e2 the residual variance, I an identity matrix, and A the numerator relationship matrix. The fitted model included the following systematic effects for each specific trait under analysis; age as a linear co-variable; sex (male or female); the stud geographical area (1 to 20), and the coat colour (grey, brown, chestnut, black, other). Marginal posterior distributions of all parameters were estimated using the Gibbs sampling algorithm [18]. Prior distributions for systematic effects were assigned as bounded uniform prior distributions and the variance components were scaled using inverted Chi–squared distributions (v=2 and S=0) to perform a flat prior distribution. Total Gibbs chain lengths of 1,000,000 samples for each analysis were defined, with a burn–in period of 100,000 and a thinning interval of 100. Results Descriptive statistics for cresty neck score and morphological traits are shown in Table 1. There were 9,955 horses out of 10,929 PRE (91.09%) that were considered to have a satisfactory cresty neck score (i.e. < 4). A total of 974 horses (8.91%) were penalized or disqualified by the breed authorities; 911 horses (8.33%), were penalized (cresty neck score 5 or 6) and 63 horses (0.58%) were disqualified (cresty neck score ≥ 7) due to cresty neck. Factors associated with cresty neck score are shown in Table 2. Girth–to–height ratio was not significantly associated with cresty neck score. Males had a significantly higher mean cresty neck score than females and horses above six years old had higher cresty neck scores than both horses below three years old and four to six year-old horses. Horses with grey coats had a higher cresty neck scores than horses with brown, chestnut, black and other coats. There was also a significant association with cresty neck score and country of in which the stud was located (Supplementary Table 3). In general, cresty neck scores were lower in studs located in Spain than in other geographical areas, and there were no countries with cresty neck scores significantly lower than Spanish studs. The biggest significant difference was between the mean cresty neck score in Spain (1.41 with 7,565 horses) and the Other South American countries group (3.66 with 21 horses). Heritability coefficients (h2) and genetic correlations between cresty neck score and other morphological traits are listed in Table 3. Discussion This study has addressed for the first time the within–breed prevalence of cresty neck within a large worldwide population of Pura Raza Español horses, and examined risk factors for this condition in this breed. Cresty neck was a prevalent morphological defect in this breed and Nearly 9% of this population had this defect, ranging in severity from a noticeable cresty neck (score 5) to a permanently droop (score 9) in a standardized cresty neck scoring system adapted for Pura Raza Español horses [9]. There are no previous data of within–breed prevalence of a cresty neck in horses but, in horses of moderate body condition, cresty neck scores were significantly higher in Pura Raza Español –cross horses than in Standardbreds [7]. Although the Pura Raza Español horse is considered a light breed, phenotypic and body scoring evidence support the observation that this breed has an intrinsic predisposition to accumulate large fat deposits [2,19]. These horses commonly exhibit generalized obesity or increased regional adiposity [20]. The current results indicate that, in Pura Raza Español horse, cresty neck score was independent of girth–to–height ratio. Although it has long been recognized that adipose tissue associated with the nuchal ligament is increased in obese horses [20,21], the contribution of neck adiposity to overall adiposity is controversial in horses. In a cohort of horses grouped towards the lower end of the expected range of total body fat content, crest fat thickness was a good indicator of total empty carcass fat [22]. Similarly, in a group of Pura Raza Español stallions with moderate body condition, cresty neck score also correlated strongly with the girth–to–height ratio [23]. However, in a study of mature pony mares with highly variable body conditions, among superficially accessible fat depots, nuchal fat was the poorest predictor of total body fat, [24]). These discrepancies were related with a loss of sensitivity of subjective body condition scoring systems in overweight/obese subjects. Some authors have suggested that tissue within the nuchal region in horses and ponies is akin to visceral abdominal fat in humans, so the type of fat stored in a cresty neck may be structurally and functionally different to that of fat stored elsewhere [25]. Cresty neck score was higher at the end of winter than at the end of summer, which was the opposite pattern seasonal variation to that observed for general obesity [25]. However, seasonal effects, which also suggest that cresty neck score is not dependent of obesity, were not examined in the current study. In the current study Pura Raza Español stallions above six years old had higher mean cresty neck scores. Previous studies have not reported specific differences in cresty neck score between horses of different ages and sex within the same breed. The increased neck fat adiposity associated with age was expected, since fat content tends to increase with ageing, resulting in an unequal distribution of fat between different anatomical regions, as demonstrated in humans [26]. Furthermore, in young horses subcutaneous fat tissue has a slow growth rate 18. Geman, S. and Geman, D. (1984) Stochastic relaxation, GIBBS distributions, and the bayesian restoration of images. IEEE Trans. Pattern Anal. Mach. Intell. PAMI-6, 721– 741. 19. Potter, S.J., Bamford, N.J. and Harris, P.A. and Bailey, S.R. (2013) Comparison of weight loss, with or without dietary restriction and exercise, in Standardbreds, Andalusians and mixed breed ponies. J. Equine Vet. Sci. 33, 339. 20. Frank, N. (2011) Equine metabolic syndrome. Vet. Clin. North Am. Equine Pract. 27, 73–92. 21. Johnson, P.J., Wiedmeyer, C.E., Messer, N.T. and Ganjam, V.K. (2009) Medical implications of obesity in horses - Lessons for human obesity. J. Diabetes Sci. Technol. 3, 163–174. 22. Znamirowska, A. (2005) Prediction of horse carcass composition using linear measurements. Meat Sci. 69, 567–570. 23. Díez de Castro, E., Pineda, C., Martin, M., Quintero, S., Mendez, N., Aguilera, E. (2012) Valoración de obesidad y resistencia a la insulina en el Caballo de Pura Raza Española (PRE). Equinus 34, 58–75. 24. Dugdale, A.H.A., Curtis, G.C., Harris, P.A. and Argo, C.M. (2011) Assessment of body fat in the pony: part I. Relationships between the anatomical distribution of adipose tissue, body composition and body condition. Equine Vet. J. 43, 552–561. 25. Giles, S.L., Nicol, C.J., Rands, S.A. and Harris, P.A. (2015) Assessing the seasonal prevalence and risk factors for nuchal crest adiposity in domestic horses and ponies using the Cresty Neck Score. BMC Vet. Res. 11, 13. 26. Villareal, D.T., Apovian, C.M., Kushner, R.F. and Klein, S. (2005) Obesity in older adults: Technical review and position statement of the American Society for Nutrition and NAASO, the Obesity Society. Obes. Res. 13, 1849–1863. 27. Martin-Rosse, W., Vernet, J., Dubroeucq, H., Arnau, G., Picard, B., Vermorel, M. (2008) Variation of fatness and energy content with body condition score in sport horses and its prediction. In: Nubrition of the Exercising Horse. European Association for Animal Production., The Nether., Ed: W. Saasamoinene, M.T., Martín-Rosset, Number 125., Wageningen. pp 167–176. 28. Kearns, C.F., Mckeever, K.H., Kumagai, K. and Abe, T. (2002) Fat-free mass is related to one-mile race performance in elite standardbred horses. Vet. J. 163, 260–266. 29. Fonseca, R.G., Kenny, D.A., Hill, E.W. and Katz, L.M. (2013) The relationship between body composition, training and race performance in a group of Thoroughbred flat racehorses. Equine Vet. J. 45, 552–557. 30. Finno, C.J., Spier, S.J. and Valberg, S.J. (2009) Equine diseases caused by known genetic mutations. Vet. J. 179, 336–347. 31. Brosnahan, M.M., Brooks, S.A. and Antczak, D.F. (2010) Equine clinical genomics: A clinician’s primer. Equine Vet. J. 42, 658–670. 32. Bellone, R.R. (2010) Pleiotropic effects of pigmentation genes in horses. Anim. Genet. 41 Suppl 2, 100–110. 33. Duensing, J., Stock, K.F. and Krieter, J. (2014) Implementation and prospects of linear profiling in the warmblood horse. J. Equine Vet. Sci. 34, 360–368. TABLE S1: Description of the cresty neck linear scale used in the population of Pura Raza Español horses under consideration. TABLE S2: Description of head–neck and neck-body junction traits. TABLE S3: General Lineal Model and a post–hoc Dunnett’s test analyzing the stud geographical area in cresty neck score, taking Spain as the control geographical area, included in this study and the percentage of animals with a Cresty Neck Score considered as a morphological defect by the breeder association (≥5) for the population of Pura Raza Español horses TABLE S1: Cresty neck linear scoring system Class Description 1 Crest is absent. There is no visual appearance, nor can any crest filling be felt with palpation 2 There is no visual appearance of a crest, but a slight filling can be felt with palpation 3 Neck is not straight; a slight deviation is visually detected. Crest is not easily cupped in one hand and not bent from side to side 4 Neck is not straight but it is not very noticeable. Crest can be cupped in one hand and bent from side to side 5 Noticeable appearance of a crest, but fat is deposited fairly evenly from poll to the wither. Crest is easily cupped in one hand and bent from side to side. 6 Crest is enlarged and thickened. Crest filling does not fill cupped hand and keeps side to side flexibility 7 Crest enlarged and thickened, so fat is deposited more heavily in middle of the neck than toward poll and the wither, giving the appearance of a mound. Crest fills cupped hand and starts to lose side to side flexibility 8 Crest is very large but it does not permanently droop to one side 9 Crest is so large that it permanently droops to one side *Adapted from Carter et al. (2009) TABLE S2: Head-neck and neck-body junction scoring systems. Trait Definition Class 1 Class 9 Head–neck junction Type of insertion of the head in the neck ranging from very distinct to very indistinct (subjective morpholological linear trait) Neck–body junction Distance between the point of insertion of the neck to the trunk (the ventral part) and the line connecting the two shoulder joints ranging from <2cm to >23cm) were estimated by a linear scale of 9 classes (objective morphological linear trait) 23 cm 2 cm TABLE S3: Associations between country of residence and cresty neck score in 10,929 Pura Raza Español horses from 24 countries. Country P value LSM Country P value LSM Spain (n=7,565, 4.89%) referent 1.41 Holland (n=35, 5.71%) >0.9 1.34 Australia (n=22, 13.64%) 0.07 2.18 Honduras (n=44, 4.55%) <0.001 2.20 Belgium (n=80, 6.25%) >0.9 1.61 Italy (n=84, 14.29%) <0.001 2.09 Colombia (n=55, 10.91%) 0.04 1.93 Mexico (n=720, 5.69%) <0.001 1.89 Costa Rica (n=228, 32.89%) <0.001 2.92 Morocco (n=20, 0%) >0.9 1.00 Denmark (n=25, 4%) >0.9 1.32 Nicaragua (n=136, 11.03%) <0.001 2.34 Ecuador (n=77, 12.99%) <0.001 2.16 Panama (n=32, 3.13%) >0.9 1.50 Other European countries group (n=66, 39.29%) <0.001 2.73 Other South American countries group (n=21, 57.14%) <0.001 3.66 Germany (n=229, 54.15%) <0.001 2.96 Sweden (n=42, 11.90%) 0.8 1.76 Guatemala (n=192, 6.77%) <0.001 1.98 USA (n=1,256, 21.58%) <0.001 2.20 LSM, Least squared mean.