Effects of different winter paddock management of Thoroughbred weanlings and yearlings in the cold region of Japan on physiological function, endocrine function and growth

Abstract

Effects of different winter paddock management of Thoroughbred weanlings and yearlings in Hokkaido, Japan, which is extremely cold in winter, on physiological function, endocrine function and growth were investigated. They were divided into two groups; those kept outdoors for 22 hr in the paddock (22hr group) and those kept outdoors for 7 hr in daytime with walking exercise for 1 hr using the horse-walker (7hr+W group), and the changes in daily distance travelled, body temperature (BT), heart rate (HR), HR variability (HRV), endocrine function and growth parameters were compared between the two groups from November at the year of birth to January at 1 year of age. The 7hr+W group could travel almost the same distance as the 22hr group by using the horse-walker. The 22hr group had a lower rate of increase in body weight than the 7hr+W group in January. In addition, lower in BT and HR were observed, and HRV analysis showed an increase in high frequency power spectral density, indicating that parasympathetic nervous activity was dominant. And also, changes in circulating cortisol and thyroxine were not observed despite cold environment. On the other hand, the 7hr+W group had higher prolactin and insulin like growth factor than the 22hr group in January, and cortisol and thyroxine were also increased. Physiological and endocrinological findings from the present study indicate that the management of the 7hr+W group is effective in promoting growth and maintaining metabolism during the winter season.

More than 97% of Thoroughbreds in Japan are bred in Hokkaido [74], which is located in the north of Japan, has mild summers but extremely cold winters. Thoroughbreds in the “weanling to yearling stage” [35], which is from weaning at the autumn of the current year to the start of breaking at yearling in the next autumn, are managed to be turned out in the pasture paddock. As herbivores, horses spend much of their time on pasture foraging and moving around in herds [70, 71], for this reason, day-and-night grazing in which horses are kept on pasture for most of the day, is natural and reasonable management from the perspective of building sociality between horses and promoting healthy body growth [11, 24, 25, 33, 40, 69]. Until the late 1990s in Hokkaido, daytime grazing, in which horses were turned out in the morning and stabled in the evening throughout the year, was common, due to the small size of the paddocks and needed to be prevent from accidents at night [29, 71]. Since the 2000s, as Japanese breeders realized the benefits of such day-and-night grazing, they expanded the size of their paddocks, and day-and-night grazing became more common during the pleasant early summer and autumn [71]. However, during the winter in Hokkaido, it is empirically known that lower temperatures [35] and snow cover make it more difficult for horses to forage for grasses and reduce the distance traveled in the paddock. For this reason, breeding farms in Hokkaido generally manage their horses kept outdoors only daytime in winter. In recent years, several owner-breeder’s farms, who aim to raise and race their own horses, have achieved successful racing results by keeping the weanlings and yearlings outdoors day-and-night during winter in Hokkaido (Fig. 1). Therefore, the management of keeping the young Thoroughbreds outdoors day-and-night in winter has been attracted attention from the breeding and racing industry in Japan, however the physiological function and endocrine function of young Thoroughbreds kept outdoors during the severely cold winter have not been clarified.

Fig. 1.

Winter paddock at Hidaka Training and Research Center, the Japan Racing Association (January).

It has also been reported that growth rates decline in winter at “the weanling to yearling stage” for reasons of cold environment and lack of good grass [34, 62, 77] and dramatically increase following spring at 1 year of age [12, 62]. This compensatory rapid growth has been reported to be one of the risk factors of developmental orthopedic disease (DOD), including osteochondrosis dissecans, epiphysis, angular limb deformities, flexural deformities, subchondral cystic lesions, cervical vertebral malformation and cuboidal bone malformation [19, 37, 65]. DOD is common cause of pain and lameness for young Thoroughbreds which depreciates the commercial value at the yearling sales and limits there future racing performance [39, 48, 65, 68]. Therefore, appropriate winter paddock management for smooth growth of young Thoroughbred is crucial to achieve sound musculoskeletal development and desirable body composition [23, 38, 62, 65].

The present study aimed to compare the two different winter management; those kept outdoors for 22 hr in the paddock (22hr group) and those kept outdoors for 7 hr in daytime with walking exercise for 1 hr using the horse-walker (RN-18R, Ridgeway Co., Ltd., Mukawa, Japan) (7hr+W group), on physiological function, endocrine function and growth parameters of Thoroughbred weanlings and yearlings in Hokkaido, Japan.

MATERIALS AND METHODS

Animals

Sixty-four Thoroughbreds (30 colts and 34 fillies) born at four farms in Hidaka region of Hokkaido, Japan were used (Table 1). These consisted of 24 Thoroughbreds (10 colts and 14 fillies) born between 2010 and 2012 at Hidaka Training and Research Center, Japan Racing Association (JRA), 10 Thoroughbreds (5 colts and 5 fillies) born in 2011 at Farm A, 10 Thoroughbreds (5 colts and 5 fillies) born in 2011 at Farm B and 20 Thoroughbreds (10 colts and 10 fillies) born in 2011 at Farm C. They were divided into two groups; the 22hr group of 32 Thoroughbreds (15 colts and 17 fillies) and the 7hr+W group of 32 Thoroughbreds (15 colts and 17 fillies). Changes in daily distance travelled, body temperature (BT), heart rate (HR), HR variability (HRV), endocrine function and growth parameters were compared. The horses were 6 and 9 months old at the initiation of the present study.

Table 1. Horse group information.

Year of birth	Breeding farm	Experiment						Experimental period	22hr (n=32)		7hr+W (n=32)
									Colts	Fillies	Colts	Fillies
2010	JRA Hidakaa)	A	B	C	D			2010 Nov.23–2011 Mar.13	2	2	2	2
2011	JRA Hidaka			C	D	E	F	2011 Nov.21–2012 Feb.29	2	2	1	3
	Farm A					E		2011 Nov.21–2012 Jan.20	0	0	5	5
	Farm B					E		2011 Nov.21–2012 Jan.20	0	0	5	5
	Farm C					E		2011 Nov.21–2012 Jan.20	10	10	0	0
2012	JRA Hidaka			C	D		F	2012 Nov.19–2013 Feb.28	1	3	2	2
Total (n=64)									15	17	15	17

A: Calculation of the distance traveled in the paddock. B: Body temperature (BT) and hematological measurement. C: Electrocardiograms and analyzing heart rate variability (HRV). 2011: (n=6, 22hr: colt 1, fillies 2, 7hr+W: colt 0, fillies 3). D: Growth parameters and average daily gain (ADG). E: Measurement of hormone [prolactine, cortisol and insulin like growth factor (IGF-I)]. F: Measurement of hormone (thyroxine). a) Hidaka Training and Research Center, Japan Racing Association. Months are indicated by their first three letters.

All Thoroughbreds were weaned by the end of September at the year of birth, and they were kept outdoors in the pasture paddock (approximately 40,000–100,000 m²) until the initiation of this study in November. All procedures complied with the guidelines for the use of horses established by the Hidaka Training and Research Center (JRA Hidaka-2010-1).

22hr group

The 22hr group was kept outdoors for 22 hr from 10:30 hr to 08:00 hr the following day. The paddock area for the 22hr group at Hidaka Training and Research Center was approximately 46,000 m² with 4 horses, while Farm A and Farm B, each had approximately 50,000 m² with 5 horses and Farm C had approximately 100,000 m² with 10 horses. The grass in the paddock was mainly Kentucky bluegrass with some white clover. The horses were allowed to forage freely in the pasture until the end of November. After the paddock was covered with snow from December onwards, the horses were fed rolls of timothy-hay in the shelter to enable free foraging and resting on the hay. In the event of heavy snowfall, snow was removed from paddock along the fences and also diagonally in both directions to allow the horses to move to feed, water, and shelter (Fig. 2). For feeding, the horses were fed twice a day, in the stable at 08:30 hr and in the paddock at 16:00 hr. In addition, 1 kg/head of alfalfa hay was fed twice daily at 06:00 hr and 20:00 hr in the corner of the paddock, away from the shelter and manger, respectively, in order to encourage the horses to move. Furthermore, a water manger with a thermal heater was used to prevent the water from freezing. The horses were not wearing rugs all day long.

Fig. 2.

Simplified diagram of turned out padock of the 22hr group and the 7hr+W group. In the event of heavy snowfall, snow was removed from paddock along the fences and also diagonally in both directions to allow the horses to move to feed, water, and shelter (shown in grey).

7hr+W group

The 7hr+W group was kept outdoors for 7 hr from 08:30 hr to 15:30 hr before walking exercise for 1 hr using the horse-walker. The paddock area for the 7hr+W group at Hidaka Training and Research Center was approximately 37,000 m² with 4 horses, while Farm A and Farm B, each had approximately 50,000 m² with 5 horses and Farm C had approximately 100,000 m² with 10 horses. The speed and duration of the exercise were systematically increased to prevent musculoskeletal injuries and to increase physical fitness of the horses (Table 2). The management methods within the paddock were the same as for the 22hr group. The horses were fed twice a day at 06:00 hr and 20:00 hr in the stables. Alfalfa hay was fed 1 kg/head twice daily in the paddock at 09:00 hr and 13:00 hr. The horses were not wearing rugs when outdoors or exercising in the horse-walker, but they wore the rugs on when resting in the stables.

Table 2. Horse-walker exercise program for 7hr+W group.

Week	Date (2010)	Speed (km/hr)	Exercise time (min/day)	Total distance (km)
1	Nov.22−Nov.28	5.0	30	2.5
2	Nov.29−Dec.5	5.0	45	3.8
3	Dec.6−Dec.12	5.5	45	4.1
4	Dec.13−Dec.19	5.5	45	4.1
5	Dec.20−Dec.26	5.5	60	5.5
6	Dec.27−Jan.2	5.5	60	5.5
7	Jan.3−Jan.9	6.0	60	6.0
8	Jan.10−Jan.16	6.0	60	6.0
9	Jan.17−Jan.23	6.0	60	6.0
10	Jan.24−Jan.30	6.0	60	6.0

Months are indicated by their first three letters.

Feed

All Thoroughbreds were fed oats, Sugar-beet pulp (Standlee Premium Products, Kimberly, WI, USA) and compound feed (Stamm 30, Hallway Feeds, Lexington, KY, USA) twice daily. The energy content was adjusted according to the Japanese Feeding Standard for Horses (2004 edition) [23], however, in the present study, they were fed approximately 25% more digestible energy (DE) than recommended. This is because of the lower ambient temperature at night in the 22hr group [15] and the exercise load in the 7hr+W group (Table 3). The body condition score [32] of the two groups of horses were generally controlled between 4.5 and 5.5.

Table 3. Daily feed management.

Month	Oatsa) (kg)	Sugar-beet pulpa) (kg)	Compound feeda) (kg)	Alfalfa hayb) (kg)	Timothy hayc) (kg)	Digestible energy (DE: Mcal)
Nov.	1	0.3	1	2	5	22.12
Dec.	1.5	0.3	1	2	5	23.58
Jan.	2	0.3	1	2	5	25.05

a) Twice a day, at 08:30 hr in the stable and 16:00 hr in the paddock for the 22hr group, at 6:00 hr and 20:00 hr in the stable for the 7hr+W group. b) Twice a day, at 06:00 hr and 20:00 hr in the paddock for the 22hr group, at 9:00 hr and 13:00 hr in the paddock for the 7hr+W group. c) Both the 24hr and the 7hr+W groups are allowed to eat freely in the paddock or in the stable. Months are indicated by their first three letters.

Calculation of the distance traveled in the paddock

Eight Thoroughbreds born at Hidaka Training and Research Center in 2010 were used (Table 1). Portable GPS logger (Trip Recorder 747 Pro, Transystem Inc., Hsinchu, Taiwan) was attached to the headcollar and a dedicated software (GPS Photo Taggaer ver. 1.2.4, Transystem Inc.) was used to measure the distance traveled while being kept outdoors [70]. Since horses kept outdoors behave in herds, the distance traveled by each group was measured using the average of one to four horses randomly selected from the 22hr group and the 7hr+W group. Distance traveled during being kept outdoors were measured once a month after weaning from August 26 at the year of birth to early November. Then the distance traveled by the 22hr group and the 7hr+W group were each measured twice a week from November 22 at the year of birth to January 31 at 1 year of age. The distance of the 7hr+W group was calculated by adding the walking distance shown in Table 2.

BT and hematological measurement

Eight Thoroughbreds born at Hidaka Training and Research Center in 2010 were used (Table 1). Rectal temperatures were measured as BT with an electronic thermometer (Medical Electronic Thermometer MC-6740, Omron Health Care Co., Matsusaka, Japan) at 08:30 hr in the stable. Blood was also collected from the jugular vein at 08:30 hr once a week, and white blood cell count (WBC) and hematocrit (Ht) were measured once a week, from November 22 at the year of birth to January 31 at 1 year of age, using a multiparameter automated blood cell calculator for animals (Automated hematology analyzer KX-21NV, Sysmex Corp., Kobe, Japan).

Daily changes of snow depth, mean temperature and minimum temperature

To analyze the impact of weather on distance traveled, BT and hematological measurement, the data from 2010 November 1 to 2011 February 28 of Urakawa District Meteorological Station, Hokkaido, Japan, where Hidaka Research and Training Center is located (142°12’E-42°51’N), were cited. The changes in daily maximum snow depth, daily mean temperature and daily minimum temperature were shown in Fig. 4.

Electrocardiograms and analyzing HRV

Twenty-two Thoroughbreds born at Hidaka Training and Research Center in 2010, 2011 and 2012 were used (Table 1). Measurements of electrocardiograms (ECGs) were taken in the third week of each month from November at the year of birth to January at 1 year of age at 08:30 hr while the horses of both groups were resting in the stable. A Holter-type ECG (SM-60, Fukuoka Denshi Co., Ltd., Tokyo, Japan) was used to obtain heart rate (HR) by base-apex leads with horses for 60 min, and the recorded ECGs were analyzed with an ECG processor analyzing system (Softron Co., Ltd., Tokyo, Japan) which has been previously reported [45, 61]. The program first detected R waves and calculated R-R interval (RRI) tachogram as the raw HRV in sequence order. From this tachogram, data sets 512 points were resampled at 200 msec [61]. We applied each set of data to a Hamming window and a Fast-Fourier Transform to obtain the power spectrum of the fluctuation [61]. Then recording results were classified RR interval variability into low frequency (LF) and high frequency (HF) components and to evaluate autonomic function by calculating the power spectrum density (PSD: msec²) of each [3, 45, 61, 66]. The frequency settings were 0.01–0.07 Hz for LF power and 0.07–0.6 Hz for HF power [61]. In the present study, HR, LF power, HF power, and the LF/FH ratio were obtained from each recording.

Growth parameters and average daily gain (ADG)

Twenty-four Thoroughbreds born at Hidaka Training and Research Center in 2010, 2011 and 2012 were used for the measurement of growth parameters (Table 1). Body weight (BW), height at withers (height), girth circumferences (girth) and cannon circumferences (cannon) were measured at the end of each month from September at the year of birth to March at 1 year of age, and the monthly rate of increase of the four parameters were calculated.

And, eight Thoroughbreds born at Hidaka Training and Research Center in 2011 were used for measurement of ADG (Table 1). They were weighed every morning at 08:00 hr by scale (TRU-TEST S3, Datamars, Auckland, New Zealand) from October 25 at the year of birth to February 1 at 1 year of age. ADG was calculated by week.

Measurement of hormone

Forty-eight Thoroughbreds born in 2011 at Hidaka Training and Research Center, Farm A, Farm B and Farm C (Table 1) were used for the measurement of prolactin, cortisol and insulin like growth factor (IGF-I) to evaluate exercise stress and growth. Sixteen Thoroughbreds born at Hidaka Training and Research Center in 2011 and 2012 (Table 1) were used for the measurement of thyroxine to evaluate metabolism in cold weather. Blood was collected from the jugular vein into a heparinized vacutainer at 08:30 hr. Plasma was harvested and stored at −20°C until assayed. For prolactin, cortisol and IGF-I, the blood samples were taken on November 21, the initiation of the present study. Thereafter, blood samples for prolactin and cortisol were collected twice per month, in December at the year of birth and January at 1 year of age; for IGF-I, blood samples were collected on Dec.12 and twice in January. For thyroxine, blood samples were collected once a month from November at the year of birth to February at 1 year of age.

Hormone assays

The plasma concentration of prolactin was determined by homologous double-antibody equine radioimmunoassay (RIA) methods as described previously [18]. Plasma concentrations of prolactin were measured using an anti-equine prolactin serum (AFP-261987) and purified equine prolactin (AFP-8794B) for radioiodination and the reference standard. The intra- and inter-assay coefficients of variation were 7.1% and 9.8% respectively. The plasma concentrations of cortisol were determined by RIA methods as described previously [4] using rabbit anti-cortisol (HAC-AA71-02RBP85) and anti-rabbit gamma globulin goat serum (#42-1) for radioiodination and the reference standard. ¹²⁵I-labeled cortisol (07121126 MP Biomedicals. LLC, Irvine, CA, USA) was used as the labeled antigen. The intra- and inter-assay coefficients of variation were 9.6% and 1.3% respectively. The plasma concentration of IGF-I were measured by RIA as previously described [16] using anti-sera against human IGF-I (AP 4892898) and purified IGF-I (Lot#090701) for radioiodination and for the reference standard. The intra- and inter-assay coefficients of variation were 2.7% and 14.8% respectively. The plasma concentration of thyroxine was determined by time-resolved fluoroimmunoassay using dissociation-enhanced flluoroimmunoassay (DELFIA, Thyroxine 1244-030 kit, PerkinElmer, Waltham, MA, USA). The intra- and inter-assay coefficients of variation were 6.3% and 7.0% respectively.

Statistical analysis

All results are presented as means ± standard errors of the mean (SEM). Statistical analyses were performed by use of statistical software JMP (Ver 16.0). Differences among sex, year of birth and breeding farm in each examination were evaluated by the two-way factorial measure analysis of variance (ANOVA). After confirming that there were no significant differences among them, the one-way ANOVA was used to detect significant differences between dates in BT, hematological measurement, HRV, growth parameters, ADG and hormonal changes in the same group. When the analysis was significant, difference between specific dates in each experiment were analyzed by Tukey-Kramer test. Statistical comparisons between the two groups were performed by Student’s t-test when uniformity of variance was confirmed by the F-test. When the variance was not uniform, an unpaired t-test with Welch’s correction was used. P-values less than 0.05 were considered statistically significant.

RESULTS

Distance traveled in the paddock, BT, Ht and WBC

There were no significant differences between 8.3 ± 0.2 km of the 22hr group and the 8.0 ± 0.3 km of the 7hr+W group in the mean distance traveled in the paddock (Table 4). However, the mean distances traveled of the 22hr group and the 7hr+W group were significantly higher than the 3.0 ± 0.1 km of being kept outdoors for 7 hr without walking exercise (P<0.05: Table 4). The mean values of BT of the 22hr group were significantly lower than those of the 7hr+W group in the present study (P<0.05: Table 4). However, no significant differences were observed between the two groups in the mean values of WBC and Ht (Table 4).

Table 4. The mean values of distance traveled, body temperature (BT), white blood cell (WBC) and hematocrit (Ht).

Horse group	22hr (n=4)	7hr+W (n=4)
Distance traveled (km)	8.3 ± 0.2a	8.0 ± 0.3a (7hr+W), 3.0 ± 0.1b (7hr)
BT (°C)	37.5 ± 0.1*	37.8 ± 0.1
WBC (/μL)	10,502 ± 3	11,219 ± 3
Ht (%)	36.9 ± 0.6	37.9 ± 0.8

(7hr): Distance traveled without walking exercise. Results are expressed as means ± SEM. Different alphabets represent significant differences at P<0.05. *Significant differences between 22hr group and 7hr+W group at P<0.05.

Figure 3 shows the changes in the mean values of BT from November 1 to January 31. The values of BT in the 22hr group were significantly lower than the 7hr+W group on November 26, November 29, December 1, December 2, December 10, December 24, January 7, January 25, and January 31 (P<0.05: Fig. 3).

Fig. 3.

Mean body temperature (BT) at 08:30 hr in the 22hr group (● n=4) and the 7hr+W group (□ n=4). Results are expressed as means ± SEM. Grey fill indicates the duration of the present study. Months are indicated by their first three letters. *Significant differences between the two groups at P<0.05.

Figure 4 shows the daily changes of snow depth, mean temperature and minimum temperature from November to February. In January, the daily mean temperature was below 0°C on most of the days and the daily minimum temperature was below −10°C on some days. In addition, the pastures were deeply covered with snow and frozen from December 25 to February 14.

Fig. 4.

Changes in daily maximum snow depth (■), daily mean temperature (●) and daily minimum temperature (○) from 2010 November 1 to 2011 February 28 at Urakawa District Meteorological Station, Hokkaido, Japan. Grey fill indicates the duration of the study. Months are indicated by their first three letters.

HR and HRV

The changes in HR and HRV were shown in Fig. 5. The mean HR of the 22hr group decreased from November to January, and was significantly lower in January than in November (P<0.05: Fig. 5A). The HF power, which reflects parasympathetic activity, showed an increasing trend from November to January in the 22hr group, whereas the 7hr+W group showed a decreasing trend. HF power in January was significantly higher in the 22hr group than in the 7hr+W group (P<0.05: Fig. 5B). LF power, which reflects both sympathetic and parasympathetic nervous activity, showed a trend of increase from November to January in the 22hr group, while there was little change in the 7hr+W group (Fig. 5C). The LF/HF ratio showed a trend of increase from November to January in both groups, and in January, the 22hr group tended to have a lower LF/HF ratio than the 7hr+W group (Fig. 5D).

Fig. 5.

Changes in heart rate (HR: A), high frequency power spectrum density (HF PSD: B), low frequency power (LF PSD: C) and LF/HF ratio (D) in the 22hr group (● n=11) and the 7hr+W group (□ n=11) from November at year of birth to January at 1 year old. Results are expressed as mean ± SEM. Months are indicated by their initial letters. Different letters indicate significant differences in the same group (P<0.05). *Significant differences between the two groups at P<0.05.

Growth parameters and average daily gain (ADG)

The monthly rate of increase in BW, height, girth and cannon from October to March in the 22hr group and the 7hr+W group were compared in Fig. 6. The 22hr group showed lower monthly rate of increase in BW in January and February than the 7hr+W group (P<0.05: Fig. 6A), whereas the monthly rates of increase in BW and girth were significantly lower in the 7hr+W group than in the 22hr group in November (P<0.05: Fig. 6A and 6C). No significant differences were observed in height and cannon between the 22hr group and the 7hr+W group (Fig. 6B and 6D).

Fig. 6.

Comparison of the monthly rates of body weight (BW: A), height at withers (height: B), girth circumference (girth: C) and cannon circumference (cannon: D) between the 22hr group (■ n=12) and the 7hr+W Group (□ n=12) from October at the year of birth to February at 1 year old in Thoroughbreds reared at Hidaka Training and Research Center, Japan Racing Association. Results are expressed as means ± SEM. Months are indicated by their initial letters. *Significant differences between the two groups at P<0.05.

Weekly changes in ADG from November 2 to February 1 were shown in Fig. 7. The 22hr group was significantly lower than 7hr +W group during the weeks of December 2, January 11 and January 18 (P<0.05). On the other hand, the 7hr+W group was significantly lower than the 22hr group during the week of November 24, when this study was initiated (P<0.05).

Fig. 7.

Comparison of the average daily gain (ADG) between the 22hr group (■ n=4) and the 7hr+W group (□ n=4) from November 2 at the year of birth to February 1 at 1 year old in Thoroughbreds reared at Hidaka Training and Research Center, Japan Racing Association. Grey fill indicates the duration of the present study. Results are expressed as means ± SEM. Months are indicated by their first three letters. *Significant differences between the two groups at P<0.05.

Hormones

Changes in the plasma concentrations of prolactin, cortisol, IGF-I and thyroxine were shown in Fig. 8. The plasma concentration of prolactin increased slowly in the 7hr+W group, and was significantly higher on December 26, January 10, and January 23 than on November 21, the initiation of the present study (P<0.05: Fig. 8A). The plasma concentrations of prolactin in the 22hr group increased slowly after December 26, and was significantly higher on January 23 than on November 21 (P<0.05: Fig. 8A). From November until December 12, the 22hr group remained higher than the 7hr+W group, and the 22hr group was significantly higher than the 7hr+W group on December 12 (P<0.05: Fig. 8A). However, the 7hr+W group showed higher trend than the 22hr group after December 26, and was significantly higher than the 22hr group on January 10 and January 23 (P<0.05: Fig. 8A).

Fig. 8.

Changes in the plasma concentrations of prolactin (A), cortisol (B) and IGF-I (C) in Thoroughbred weanlings of the 22hr group (● n=24) and the 7hr+W group (□ n=24), born in 2011 at Hidaka Training and Research Center, Japan Racing Association, A Farm, B farm and C Farm. And monthly changes in thyroxine (D) in Thoroughbred weanlings of the 22hr group (● n=8) and the 7hr+W group (□ n=8), born in 2011 and 2012 at Hidaka Training and Research Center. Japan Racing Association. Results are expressed as means ± SEM. Months are indicated by their initial letters. Different letters indicate significant differences in the same group (P<0.05). *Significant differences between the two groups at P<0.05.

The plasma concentrations of cortisol in the 7hr+W group increased slowly, and was significantly higher on January 10 and January 23 than on November 21 (P<0.05: Fig. 8B). On the other hand, no significant changes were observed in the 22hr group during the present study period.

The plasma concentrations of IGF-I increased in the 7hr+W group, and was significantly higher on January 10 and January 23 than on November 21 (P<0.05: Fig. 8C). And the plasma concentrations of IGF-I in the 22hr group increased slowly after December, and was significantly higher on January 10 and January 23 than on November 21 (P<0.05: Fig. 8C). Furthermore, the 7hr+W group was significantly higher than the 22hr group on January 10 and January 23 (P<0.05: Fig. 8C).

The plasma concentrations of thyroxine in the 7hr+W group increased, and was significantly higher in December, January and February than in November (P<0.05: Fig. 8D), whereas no significant changes were observed in the 22hr group. The plasma concentrations of thyroxine in the 7hr+W group tended to be higher from December to February than in the 22hr group, however no significant differences were observed between the two groups.

DISCUSSION

In the present study, the two different winter management of the 22hr group and the 7hr+W group on physiological function, endocrine function and growth parameters were compared during “the weanling to yearling stage” of Thoroughbreds in Hokkaido, which is cold region of Japan. Both the 22hr and the 7hr+W groups were reared in a healthy manner without any clinical abnormalities, based on WBC and Ht findings, despite some severely cold days in January when the minimum temperature was below −10°C.

The monthly rates of increase in BW and girth were lower in the 7hr+W group than in the 22hr group in November, and the values of ADG in the 7hr+W group were lower than the 22hr group during the first week of the present study in November. The reason for this was assumed to be a temporary increase in physical activity of the weanlings to investigate the new environment due to the change of paddock, and a decrease in grass forage due to shortened grazing hours from 22 hr to 7 hr. Since the value of BW and girth are reported to be directly correlated [14], as a result of the weight reduction, girth also decreased. Therefore, it was suggested that reductions in grazing time should be phased in to prevent rapid weight loss in young Thoroughbreds.

The ADG of Thoroughbred is reported to be highest in the first month at 1.4 to 1.7 kg/day, and then declines linearly to 0.4 to 0.6 kg/day at 8 to 11 months of age [12, 23, 38, 64, 77], which is the month of January in the present study. In this study, the 22hr group showed lower ADG and the monthly rate of increase in January compared to standard values, despite feeding and travel distances comparable to the 7hr+W group. It was suggested that the 22hr group may not have met the maintenance energy requirements to withstand the severe cold [14].

The lower critical temperature (LCT), which is the end of the thermoneutral zone, is defined as lowest temperature below which the horse must increase metabolic heat to maintain body core temperature. LCT varies with age, breed, cold acclimation, feed and coat condition [51], however has been reported as −15°C [49] and −11°C [15]. On the other hand, it has been reported that horses can be kept healthy even in colder temperatures than LCT by installing shelters to protect them from wind and cold, because their coats grow densely and their length increases when they are kept outdoors during the severely cold winter [11, 14, 50]. Another reason for the cold tolerance of horses in winter is thought to be the heat generated by bacterial fermentation during digestion of grass fiber in large colon and cecum [14, 26].

In the present study, the 22hr group, which was kept outdoors in cold environment for a longer period, showed lower BT values than the 7hr+W group. Homeothermic animals can maintain body core temperature by heat production in cold environments. Low ambient temperature information is transmitted from cold receptors in the skin to the hypothalamic preoptic area (POA), the thermoregulatory center, via the spinal cord and lateral parabrachial nucleus in the brainstem [56, 57]. POA activates sympathetic nerves, constricts peripheral blood vessels, and inhibits heat dissipation, while simultaneously stimulating skeletal muscle shivering and heat production [55]. Sympathetic nerve stimulation also promotes thyroxine secretion from the thyroid gland via the hypothalamus-pituitary-thyroid axis, which stimulates basal metabolism and produces heat [10, 49, 63]. Furthermore, in homeothermic animals, including humans, heat production by the brown adipose tissue (BAT) is important for maintaining body core temperature in cold environments [47]. This response produces heat by activating uncoupling protein 1 (UCP1) in the mitochondria of the BAT by noradrenaline released from sympathetic nerve terminals [6, 13, 55]. In addition, leptin secreted from white adipose tissue (WAT) by cold stimuli activates sympathetic nerves via the hypothalamus and increases heat production of UCP1 in the BAT [30, 31]. However, it has been reported that leptin secretion is decreased in mice on restricted diets or starvation [2] and that the functions of the hypothalamic-pituitary-adrenal and hypothalamic-pituitary-thyroid axes are impaired [2]. And, in rats on starvation, ghrelin [43, 59] is released from the stomach into the blood and activates neuropeptide Y (NPY) neurons in the hypothalamic arcuate nucleus [42]. NPY activates GABAergic neurons in the reticular formation in the medulla oblongata, inhibits sympathetic promoter neurons in the medullary suture nucleus, and reduces heat production of UCP1 in the BAT [21, 58]. These responses are thought to help protect the organism from starvation by reducing metabolism during cold [58]. Furthermore, Przewalski horse [5, 67], Shetland pony [8, 9], sheep [22], moose [27] and red deer [76] raised in a cold natural environment have been reported to exhibit responses that reduce metabolic rate and decrease BT and HR with the aim of defending the organism against reduced food intake.

HF power is derived from the parasympathetic tone associated with breathing and is considered a quantitative index of parasympathetic activity [3]. On the other hand, the low-frequency (LF) component, LF power, is derived from the Meyer wave, a systolic blood pressure variation that occurs in approximately 10-sec cycles and reflects both sympathetic and parasympathetic nervous activity [3]. The LF/HF ratio is also an indicator of the balance between sympathetic and parasympathetic autonomic nervous activity [66]. In the present study, the HR was lower in January than in November in the 22hr group. And, the 22hr group had significantly higher in HF power than the 7hr+W group, and in addition, had lower trend of LF/HF ratio, indicating that parasympathetic nervous activity was dominant. Brinkmann et al [10] also reported that in Shetland ponies, blood thyroxine levels increased in the group that did not restrict foraging during the cold winter, but not in the restricted group. Furthermore, HRV analysis has reported an increase in root mean square successive difference (RMSSD), a measure of parasympathetic activity, in Przewalski horses reared for long periods of time in cold environment [67]. Based on these reports and the results of the present study, it was suggested that the 22hr group may have had insufficient DE requirements in the cold environment, and had a response to reduce energy expenditure by decreasing BT and metabolism to sustain life and energy consumption [5, 8].

Seasonal changes of circulating prolactin in Thoroughbred foals, weanlings and yearlings were also observed in our previous studies [18, 44, 52], and these results demonstrated the prolactin levels were low in the short-day period and high in the long-day period. However, in the present study, circulating prolactin and IGF-I in the 7hr+W group increased significantly from December to January, which is short-day period, and higher than the 22hr group. And also, the prolactin levels were higher in the 7hr+W group than in the 22hr group despite the similar daily distance traveled by both groups, suggesting that the exercise load caused by horse-walker was more stressful than the natural movement of the 22hr group. The HR during walking exercise in the present study was not measured, however the HR of adult Thoroughbred horses in walking is approximately 60 to 70 beats/min [53], it was suggested that the exercise in the present study of the HR increased to approximately 1.5 to 2 times the resting HR, lasted 60 min to the extent. Prolactin promotes cortisol secretion from the adrenal glands [36] and has anti-stress effects that increase during various stresses to protect the organism from stress [73]. It was suggested that prolactin may have been involved in the increased cortisol levels in the 7hr+W group. However, it is unclear the reason why prolactin levels were higher in the 22hr group than in the 7hr+W group on December 12. But it was suggested that little snow cover on the pastures allowed horses to move and graze freely in the 22hr group. Therefore, the stress level of the 22hr group might have increased for some reason in the outdoor environment. In addition, a previous studies demonstrated that chronic elevation of prolactin levels within the brain results in reduced neuronal activation within the hypothalamus, specifically within the paraventricular hypothalamic nucleus, in response to an acute stressor [20, 75]. Thus, the prolactin acting in various relevant brain regions exerts profound anxiolytic and anti-stress effects and might contribute to the attenuated physical and emotional stress responsiveness in the 22hr group [41].

IGF-I is secreted from the liver in response to growth hormone [46] and promotes bone growth [46] and muscle hypertrophy [1, 17]. In our previous study [41], the stress of incremental treadmill exercise in adult Thoroughbred horses rapidly increase the secretion of prolactin and growth hormone levels, and that secretion of them persisted for a specific period after exercise. In the present study, the 7hr+W group had higher IGF-I levels in January than the 22hr group, suggesting that the 7hr+W group had a higher rate of BW increase as a result of walking exercise stimulating IGF-I secretion after growth hormone. The results suggested that exercise of horse-walker in the 7hr+W group may be an effective winter management to reduce winter growth suppression [34, 62, 77] and to promote healthy growth.

Low ambient temperature promotes the secretion of thyroxine [7, 49, 72] and cortisol [63] to maintain BT and metabolism, resulting in increased metabolic rate and heat production [22]. The results of the present study suggest that paddock management of the weanlings and yearlings kept outdoors most of the day during the winter, is not recommended for market-breeders whose goal is to sell horses, from the viewpoint of metabolic and growth suppression in winter. However, there are aspects that cannot necessarily be ruled out, as there are advantages from the viewpoint of keeping horses healthy in a natural environment and the cost savings for owner-breeders whose goal is to run the horses by themselves. In addition, increasing feed to meet nutrient requirements for severe coldness, as well as wearing rugs to keep the horses warm and applying walking exercise by horse-walker may improve metabolic and growth suppression. Furthermore, our previous studies have shown that an extended photoperiod from December at the year of birth to May at 1 year of age promotes prolactin secretion, gonadal function and winter haircoat replacement [28]. Thus, wearing a light mask [54, 60] may also be effective in promoting proper growth during winter. Further research is needed to develop optimal winter paddock management techniques in Hokkaido, which is severely cold in winter.

In conclusion, it was suggested that the 22hr group had insufficient DE for the lower ambient temperature, which may have resulted in lower BT, lower HR, higher HF power, and lower metabolism, to protect the organism from the severe cold environment. On the other hand, the 7hr+W group was able to travel almost the same daily distance as the 22hr group by using the horse-walker, and no metabolic or growth suppression was observed in the cold environment. The physiological and endocrinological findings of the present study indicate that keeping weanlings and yearlings outdoors in daytime during winter and providing them with walking exercise is effective in promoting growth and maintaining metabolism.

CONFLICT OF INTEREST

The authors declare no conflicts of interest associated with this manuscript.