A topographic analysis of skin thickness in horses

Abstract

A systematic study of regional skin thickness in different types of horses has not yet been described. Knowledge about regional skin thickness has long been used in human medicine to optimise wound care and skin grafting outcomes and similar knowledge could prove useful in equine wound management. The objective of the current study was to report the topographical variation in skin thickness of Warmblood (WB) and Coldblood (CB) type horses and to compare different methods for measuring skin thickness. Horses free of skin conditions and euthanised for reasons unrelated to this study were included. Skin biopsies were collected post-mortem from 28 locations in 9 WB horses and skin biopsy thickness was measured using a digital calliper. In 6/28 locations, skin fold measurements using a micrometer were also obtained. In another cohort comprising 8 WB and 10 CB horses, skin biopsies were harvested from 6 locations for histologic skin thickness measurements. Descriptive statistics revealed large topographical variation in skin thickness. A mixed effect model assessing the effect of breed and sampling location demonstrated that skin thickness was significantly higher in CBs than WBs (P < 0.001). At the mid-15th rib and between the forelimbs, there was strong correlation between calliper and skin fold measurements (ρ 0.72 and 0.74, respectively), whereas correlation was very strong at the ventral abdomen (ρ 0.83). In conclusion, this study demonstrates large topographical variations in skin thickness in horses, and significant differences between horse types. Skin fold measurements may estimate skin thickness at the ventral abdomen.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11259-026-11124-4.

Introduction

Wounds are encountered by the equine practitioner on a regular basis. In the UK, wounds account for 20% of all cases presented out-of-hours and wounds, injury and trauma accounts for 16.3% of adult horse euthanasia in the USA (Bowden et al. 2020, USDA ). The failure rate of primary wound repair has historically been high (Wilmink et al. 2002), yet outcomes have improved in recent years due to improved wound treatment knowledge and the introduction of new tension relieving suturing techniques (Freeman et al. 2021, Comino et al. 2024). However, for larger wounds not amenable to primary closure, Meek micrograft partial thickness skin grafting represents an optimal treatment option, yielding a faster healing time and superior functional outcomes compared to secondary wound healing (Wilmink et al. 2006, Wilmink and Weeren 2024).

For larger limb wounds, partial thickness skin grafts are preferred compared to full thickness grafts due to their ability to stimulate wound contraction, which leads to faster wound healing (Wilmink et al. 2006, Wilmink and Weeren 2024). The optimal thickness of partial thickness grafts is still undetermined, although excellent results have been achieved when harvesting grafts up to 1.2 mm thick (Wilmink et al. 2006). Despite this, the optimal depth at which the partial skin graft should be harvested could potentially also vary with the nature of the donor site and recipient bed. Indeed, in human medicine, detailed knowledge of topographic variations in skin thickness is used to improve function and cosmesis when performing skin grafting (Lee and Hwang 2002, Kim et al. 2019).

Little objective information exists on topographic variations in skin thickness in horses. In one study, regional skin thickness was reported in 13 horses of unknown breeds (Talukdar et al. 1972), whereas another study provided a detailed description of skin thickness in a single Thoroughbred gelding (Wakuri et al. 1995). Thus, a systematic study of regional skin thickness in different types of horses is lacking. To the authors knowledge, the anecdotal clinical impression of Coldblood (CB) type horses having thicker skin than Warmbloods (WB) has also not been substantiated in literature. Detailing such a difference could impact recommendations for the depth at which partial thickness skin grafts should be harvested in these different types of horses. Therefore, the main objective of the current study was to describe the topographical variation in skin thickness across the body and the extremities of WB and CB type horses in order to identify potential breed related differences. The secondary objective was to compare the thickness of harvested skin samples to a non-invasive skin fold measurement, taken at the same location to identify potential associations.

Materials and methods

Macroscopic study

A convenience sample of 9 Dutch WB horses (mean age 8.1 ± 5.15 years; mean body weight 567 ± 61 kg; 2 males, 6 females and 1 unknown) were included, all of which presented for euthanasia for reasons unrelated to this study and free of any significant dermatological conditions as assessed by clinical examination. Horses with obvious dermatological conditions such as granulomatous, fungal or bacterial disease were excluded. Skin fold measurements and skin samples were collected at the same timepoint for each horse and within 4.5 ± 3.2 h of euthanasia. 28 biopsy sampling sites were determined and samples taken from both sides of the body totalling 56 samples per horse (Table 1). Six of these sites were also selected for skin fold measurements (Table 1). Prior to measurements and sampling the sites were shaved to remove all hairs. At those sites for which both skin fold measurements and biopsies were taken, skin fold measurements were performed prior to biopsy sampling, by lifting the skin using rat tooth forceps and measuring the resultant skin fold thickness using a micrometer (Supplementary Fig. 1, Feinmess Suhl GmbH, Suhl, Germany). The in-build pressure gauge of the micrometer emitted a sound signal when resistance was met, which was used as a cue to stop applying pressure on the instrument, ensuring consistent pressure application across measurements. Thereafter, full thickness skin biopsies were taken 5 mm distant to the skin fold measurement site using 8 mm biopsy punches (Stiefel, a GSK Company, Durham, NC, USA). Any attached subcutaneous tissue was carefully removed using a scalpel blade and the whole skin thickness was measured once using a digital calliper (Digital calliper NO. 711 150 mm, Gedore, Remscheid, Germany). These measurements were performed under a magnifying glass for visual control of sample deformation, and care was taken to ensure that the biopsies were not compressed when the measurements were recorded in millimetres.

Table 1.

Descriptive statistics (Mean and standard deviation; minimum, median and maximum) of skin biopsy measurements using a digital calliper across 28 sampling locations in 9 warmblood horses, and skin fold thickness measured with a micrometer across 6 locations. ρ = Spearman’s ρ correlation coefficient; 0.6–0.79 indicates strong correlation whereas 0.8–1.00 indicates very strong correlation. Asterisks denote significant correlation between skin biopsy measurements and skin fold thickness (* P = 0.0005; ** P = 0.0416)

Biopsy Location		Biopsy thickness, mm		Skin fold thickness, mm		ρ
		Mean ± SD	min, median, max	Mean ± SD	min, median, max
Head	1. Axial to medial canthus	1.77 ± 0.21	1.49, 1.79, 2.31
	2. Mid masseter	1.69 ± 0.18	1.42, 1.72, 1,95
Neck	3. 2nd vertebra	2.02 ± 0.21	1.66, 2.03, 2.42
	4. 5th vertebra	2.05 ± 0.23	1.73, 2.00, 2.51
Back	5. Withers	3.02 ± 0.58	1.90, 2.91, 4.02
	6. 5 cm abaxial to L6	4.36 ± 1.08	1.90, 4.44, 5.89
	7. Ileum – tail	3.34 ± 0.62	2.10, 3.55, 4.14
	8. Tuber coxae – tuber ischium	2.55 ± 0.32	2.00, 2.66, 3.02
Side	9. Mid caudal scapula	2.15 ± 0.27	1.78, 2.12, 2.84
	10. Mid 15th rib	3.06 ± 0.74	2.11, 2.77, 4.98	2.80 ± 0.40	2.17, 2.92, 3.24	0.72
	11. 10 cm caudal to mid 18th rib	3.06 ± 0.75	2.00, 3.69, 4.76
Ventral abdomen	12. Mid pectoral	1.86 ± 0.20	1.60, 1.83, 2.47	2.36 ± 0.29	1.99, 2.28, 2.90	0.27
	13. 5 cm abaxial to midline at forelimbs	2.04 ± 0.23	1.68, 2.00, 2.42	2.41 ± 0.30	1.95, 2.40, 2.96	0.74*
	14. 15 cm abaxial level of 15th rib	2.27 ± 0.32	1.76, 2.24, 2.92
	15. 15 cm cranial and 5 cm abaxial to umbilicus	2.02 ± 0.23	1.67, 2.07, 2.51	2.42 ± 0 0.59	1.90, 2.26, 3.27	0.40
	16. 5 cm abaxial to umbilicus	1.94 ± 0.31	1.60, 1.87, 2.73	2.11 ± 0.40	1.45, 2.17, 2.68	0.83**
	17. Over inguinal canal	1.47 ± 0.29	1.06, 1.44, 1.98	2.13 ± 0.20	1.80, 2.19, 2.36	0.22
Forelimb	18. Lateral mid radius	2.00 ± 0.28	1.39, 2.01, 2.51
	19. Medial mid radius	1.96 ± 0.32	1.36, 2.00, 2.55
	20. Dorsal carpus	3.03 ± 0.46	2.30, 3.11, 3.77
	21. Mid dorsal McIII	2.75 ± 037	2.05, 2.73, 3.30
	22. Dorsal, 1.5 cm proximal to coronary band	4.51 ± 0.38	3.90, 4.54, 5.14
Hind limb	23. Middle of semimembranosus/semitendinosus	2.33 ± 0.30	1.90, 2.26, 2.93
	24. Lateral mid tibia	2.47 ± 0.37	1.80, 2.54, 3.05
	25. Medial mid tibia	2.18 ± 0.30	1.66, 2.25, 2.81
	26. Dorsal tarsus	2.54 ± 0.51	1.62, 2.50, 3.55
	27. Mid dorsal MtIII	3.27 ± 0.38	2.65, 3.22, 4.17
	28. Dorsal, 1.5 cm proximal to coronary band	4.57 ± 0.40	3.86, 4.50, 5.34

Microscopic study

A convenience sample of 18 horses were included, all of which were euthanised for reasons unrelated to this study and free of skin conditions. Included horses comprised eight WB type horses (4 WB sport horses, 3 Standardbreds and 1 American Quarter Horse) and 10 CB type horses (4 Norwegian-Swedish Coldblooded trotters, 2 Friesians, and 1 of each of the following breeds: North Swedish Horse, Norwegian Fjord Horse, Icelandic Horse and Shetland Pony). There were 14 males (12 geldings, 2 stallions) and 2 females. The mean age was 12.5 years (range 3–30). Samples were collected within 6 h of euthanasia. Six unilateral sampling sites were determined (Table 2); cadavers were sampled on the most convenient side depending on the animal’s orientation. Sampling sites were clipped and cleaned gently with Baktolin^® (Hartmann Group, Heidenheim, Germany) and 70% ethanol. Three full-thickness biopsies were taken per site using an 8 mm diameter biopsy punch (Jørgen Kruuse A/S, Langeskov, Denmark) totalling 18 biopsies per horse. Biopsies were cut from the underlying subcutaneous tissue using a scalpel blade taking care to trim off all subcutaneous tissue and subsequently placed in 10% neutral buffered formalin.

Table 2.

Inter-observer reliability between two observers for histologic skin measurements of biopsies from 18 horses (10 Coldblood type and 8 Warmblood type), measured at three different skin depth levels. An intraclass correlation coefficient (ICC) of 0.5–0.75 indicates moderate reliability

	Measurement level
	Epidermis	Epidermis + Upper dermis	Whole biopsy
Intra-class correlation coefficient (ICC)	0.64	0.70	0.63
Repeatability	37.2%	29.4%	36.5%
Interaction variation	4.4%	1.6%	0.0%
Variance d/t observer	10.2%	0.5%	0.0%
Variance d/t horse	48.2	68.5%	63.5%
Repeatability variation	37.2%	29.4%	36.5%

After tissue fixation, biopsies were bisected perpendicular to the epidermal surface. Following routine processing the biopsies were embedded in paraffin blocks, sectioned at 4 μm and stained routinely with hematoxylin and eosin. The slides were scanned using an automated whole-slide scanner (NanoZoomer-XR, Hamamatsu Photonics K.K., Hamamatsu City, Japan) and visualised using appropriate software (NDP.View 2.9.29, Hamamatsu Photonics K.K., Hamamatsu City, Japan). Skin thickness was measured at three levels using the inbuilt NDP.View software measuring tools following appropriate calibration; epidermal thickness was measured from the top of the stratum corneum to the basement membrane and recorded in µm; the combined thickness of the epidermis and upper dermis was measured from the top of the stratum corneum to the deepest aspect of the follicular units and recorded in mm; and finally, thickness of the whole biopsy was measured and recorded in mm. All measurements were performed perpendicular to the epidermal surface (Fig. 1), and for each biopsy these measurements were performed at 10 consecutive and evenly spaced sites along the tissue, totalling 180 measurements per horse. All samples were measured by the main investigator (HMBO). One randomized biopsy site per horse was also measured in an identical fashion by a European board-certified veterinary pathologist (LW), who was blinded to both breed (WB versus CB type) and biopsy location.

Fig. 1.

Whole-slide image sections showing the three levels that were measured histologically. (A) The epidermis was measured from the top of the stratum corneum to the basement membrane (a). (B) The upper dermis was measured from the top of the stratum corneum to the deepest follicular units (b), whereas the whole biopsy was measured from the top of the stratum corneum to the dermal-subcutaneous junction (c). All measurements were done perpendicular to the epidermal surface

Statistical analyses

Statistical software (JMP Pro 16) was used for data analyses. For the calliper measurements, a mixed model was used to assess effects of horse and sampling side (right versus left) using sampling side, sampling location and their interaction as fixed effects and horse as random effect. As there was only a significant effect of the sampling location (P < 0.0001) and sampling side was not significant, data from all horses and from both sides of the body were pooled for further analysis. Descriptive statistics (mean, standard deviation, minimum, median and maximum) were calculated for all locations. Measurements taken from the same horse at similar locations on the fore – and hind limbs were compared using the Wilcoxon’s signed rank test. Spearman’s ρ correlations were calculated to examine associations between skin fold measurements and biopsy calliper measurements for each of the 6 locations in which skin fold data was available.

For the histologic measurements, the inter-rater reliability between the observers was assessed using intra-class correlation coefficients (ICC) and Bland-Altman plots. As there was moderate reliability between the two observers (Table 2), data from the main investigator (HMBO) was used for further analysis. A mixed effect model was used to assess the effect of breed (WB versus CB type horses) and sampling location, using breed and sampling location and their interaction as fixed effects and the 10 measurements as repeated measures.

Using the WB data only, the histologic measurements were compared to the biopsy calliper measurements taken at similar locations (neck; pectoral; rib; ventral abdomen; mid metacarpal III (McIII) and mid metatarsal III (MtIII)) and assessed using the Wilcoxon’s two sample test . A similar comparison was not possible in CB horses due to the lack of biopsy calliper skin thickness measurements in these animals.

Results

Macroscopic study

Skin biopsy calliper measurements across all 28 locations are found in Table 1, ranging from approximately 1 mm in the inguinal region to almost 6 mm on the dorsum as shown in Fig. 2. At the locations most commonly used for harvesting full- and partial thickness skin grafts (ventral abdomen and pectoral region), biopsy measurements ranged from 1.60 mm to 2.92 mm. Skin thickness across comparable locations in the fore- and hindlimbs are detailed in Fig. 3. In both fore – and hindlimbs, skin thickness generally increased from proximal to distal. The skin was significantly thicker over the lateral mid tibia versus the lateral mid radius (2.47 ± 0.37 mm and 2.00 ± 0.28 mm respectively, P = 0.0002); over the dorsal carpus versus the dorsal tarsus (3.03 ± 0.46 mm and 2.54 ± 0.51 mm respectively, P = 0.0002); and over the mid dorsal MtIII versus the mid dorsal McIII (3.27 ± 0.38 mm and 2.75 ± 0.37 respectively, P  < 0.0001).

Fig. 2.

A graphical heatmap presentation of the topographical variations in skin thickness along the equine body and extremities based on biopsy calliper measurements in 9 Warmblood horses, across 28 locations. Figure created using gimp.org and biorender.com. (Olsen, H. 2026; https://BioRender.com/ktb9hg1)

Fig. 3.

Box plot of skin thickness across multiple locations in forelimbs (blue boxes) and hind limbs (orange boxes). In both forelimbs and hind limbs, skin thickness generally increased from proximal to distal. Asterisk denote significant differences between fore- and hindlimbs (* P = 0.0002; ** P = 0.0002; *** P < 0.0001)

Correlations between skin fold thickness and biopsy calliper measurements are detailed in Table 1; there was strong correlation between the two methods at the mid 15th rib and between the forelimbs (ρ 0.72 and 0.74, respectively), whereas there was very strong correlation at the ventral abdomen abaxial to the umbilicus (ρ 0.83).

Microscopic study

Intra-class correlation coefficients (ICC) of histologic measurements at the three different skin depths are detailed in Table 2, demonstrating moderate inter-observer reliability. Bland-Altman plots and data table are found in Supplementary Fig. 2 and Supplementary Table 1. The mixed model demonstrated a significant effect of breed and location (P < 0.001), and as illustrated for the whole biopsy measurements in Fig. 4, the skin was systematically thinner in WB type horses versus CB type horses at all six locations measured. Measurements detailed per breed across the 6 locations and skin depths are found in Table 3. Notably, the thinnest skin for both types was found at the ventral abdomen measuring 2.92 ± 0.32 mm and 3.67 ± 0.73 mm in WBs and CBs respectively. The thickest skin for both horse types was found over the dorsal MtIII measuring 3.64 ± 0.34 mm in WBs and 4.57 ± 0.55 mm in CBs.

Fig. 4.

Box plot of whole biopsy histologic skin thickness measured at 6 different locations. Blue boxes: Coldblood (CB) type horses (n = 10). Orange boxes: Warmblood (WB) type horses (n = 8). Asterisk denote significant differences between the two breeds (P < 0.001)

Table 3.

Descriptive statistics (mean and standard deviation (SD); minimum, median and maximum) of skin biopsy histological measurements across 6 sampling locations, measured at three different skin depth levels (epidermis; epidermis + upper dermis; whole biopsy) in 8 Warmblood type horses (WB) and 10 Coldblood type horses (CB)

Biopsy Location	Horse type	Epidermis, µm		Epidermis + Upper dermis, mm		Whole biopsy, mm
		Mean ± SD	min, median, max	Mean ± SD	min, median, max	Mean ± SD	min, median, max
1. Mid neck, 5 cm below nuchal crest	WB	31.0 ± 5.4	20, 30, 47	1.30 ± 0.34	0.8, 1.2, 2.3	3.79 ± 0.42	3.0, 3.8, 4.8
	CB	43.3 ± 11.2	20, 44, 70	1.60 ± 0.48	0.8, 1.6, 3.0	4.22 ± 0.74	3.0, 4.1, 5.5
2. Pectoral region	WB	31.0 ± 4.8	18, 31, 46	1.13 ± 0.24	0.7, 1.1, 1.8	2.64 ± 0.38	1.9, 2.6, 3.7
	CB	40.9 ± 8.7	22, 40, 64	1.42 ± 0.29	0.8, 1.5, 2.1	3.13 ± 0.57	2.1, 3.0, 4.2
3. Costochondral junction of 15th rib	WB	31.2 ± 7.4	17, 31, 51	1.06 ± 0.22	0.7, 1.1, 1.9	3.86 ± 0.39	3.0, 3.9, 5.0
	CB	41.4 ± 12.9	17, 41, 84	1.35 ± 0.38	0.4, 1.4, 2.1	4.50 ± 0.59	3.3, 4.6, 5.7
4. 1/3 between linea alba and 12th rib	WB	32.5 ± 6.4	22, 31, 52	1.07 ± 0.24	0.7, 1.0, 1.7	2.92 ± 0.32	2.2, 2.9, 3.8
	CB	41.5 ± 12.9	14, 41, 74	1.28 ± 0.37	0.6, 1.3, 2.2	3.67 ± 0.73	2.5, 3.6, 5.4
5. Mid dorsal McIII	WB	62.7 ± 9.1	43, 63, 85	1.47 ± 0.23	1.1, 1.4, 2.2	3.19 ± 0.49	2.3, 3.0, 4.2
	CB	72.6 ± 20.7	39, 71, 133	1.96 ± 0.33	1.3, 1.9, 2.9	3.92 ± 0.54	2.7, 4.0, 5.0
6. Mid dorsal MtIII	WB	64.1 ± 10.5	40, 63, 85	1.51 ± 0.27	1.0, 1.5, 2.1	3.64 ± 0.34	2.8, 3.7, 4.5
	CB	75.2 ± 20.3	35, 73, 138	1.86 ± 0.34	1.2, 1.9, 2.6	4.57 ± 0.55	3.4, 4.6, 5.7

In WB type horses, there were significant differences in histologic and calliper skin thickness measurements of biopsies obtained from comparable locations (P < 0.001), where the histological measurements consistently yielded a thicker measurement than the calliper method (Fig. 5). The largest difference was seen at the neck (3.79 ± 0.42 mm and 2.05 ± 0.23 mm for the histological and calliper measurements, respectively, P < 0.001) whereas the smallest difference was seen at the MtIII (3.64 ± 0.34 mm and 3.27 ± 0.37 mm, respectively, P < 0.001).

Fig. 5.

Box plot of whole biopsy histologic skin thickness and macroscopic skin thickness measured at 6 different locations in WB horses, corresponding to locations 1–6 in Table 3 (histologic measurement), and locations 4, 12, 10, 14, 21 and 27 in Table 1 (calliper measurements). Blue boxes: macroscopic (calliper) measurements from 9 Dutch Warmblood horses. Orange boxes: histologic measurements from 8 Warmblood type horses. Asterisk denote significant differences between the two measuring methods (P < 0.001)

Discussion

This study provides an objective assessment of topographic variation in skin thickness in WB and CB type horses. Although the skin was significantly thicker in CBs in comparison to WBs, the regional variation was similar across breeds, with thicker skin in areas typically exposed to more frequent rubbing and impact, such as at the dorsum and extremities.

The epidermal thickness was fairly uniform across the whole body, ranging from 31.0 to 32.5 μm in WB and 40.9 to 43.3 μm in CB, coinciding well with a previous report on WB horses (Jørgensen et al. 2018). The epidermal thickness at the extremities, measured at the dorsal aspect in our study, was found to be double that of the body, differing from previously reported results where measurements were obtained from the lateral aspect (Jørgensen et al. 2018). These discrepancies point towards locoregional variation in extremity skin thickness, where thicker skin is found at the dorsal aspect. Our CB data on epidermal thickness across the body was similar to what has been previously reported in 13 horses of unknown breeds (Talukdar et al. 1972); however, skin thickness measurements of the extremities were not performed in this report. In addition, the variation in dermal thickness reported in the current study follows the same pattern as described previously although the definitive thickness does not directly match (Talukdar et al. 1972, Wakuri et al. 1995); both thinner skin and thicker skin has been reported while our data falls somewhere in the middle (Talukdar et al. 1972, Wakuri et al. 1995). The observed differences are likely due to variations in breed, sampling and processing techniques. The method of sampling was not described in the two aforementioned studies; however, measurements were performed only on histological sections in both studies, although the processing techniques varied slightly between the reports.

Across mammalian species, dermal thickness of the dorsum is often positively correlated to body weight (Wada et al. 2024). The differences in skin thickness between WB and CB horses in the current study cannot be explained by body weight, as the CB type group contained a variety of breeds of different sizes, some of them smaller, than the more homogenous WB type group. The observed differences may instead be a result of evolution and adaptation, as CB type horses are generally originating from colder climates and historically worked in harsher environments than WB type horses, perhaps necessitating a thicker skin in order to provide greater insulation and resistance to wear as an adaptation to these conditions. The practical consequence of these observed breed differences may be of importance, for instance, when harvesting partial thickness skin grafts, as harvesting at a standardized graft thickness will yield a graft cut at a proportionally deeper dermal layer in WBs versus CBs. The dermal depth at which partial thickness grafts are harvested may impact not only the mechanical and cosmetic properties of the graft itself, but may also influence the overall healing characteristics of the recipient bed (Chan et al. 2015). The mechanism responsible for the observed induced wound contraction when using partial thickness skin grafts compared to full-thickness grafts has yet to be elucidated (Wilmink et al. 2006, Wilmink and Weeren 2024). If there are certain dermal components that need to be either included or excluded in the harvested graft to trigger this effect, it would be beneficial to ascertain at what dermal depth such components reside in the different regions of the horse. Preliminary results from our research group indicate that the total tissue secretome from the upper aspect of the dermis induces increased contractility in vitro compared to the secretome from the deeper dermal layers. Therefore, when harvesting partial thickness skin, it could be prudent to harvest at a more superficial level to allow greater contractility of the resulting graft bed; however, thin grafts (< 0.5 mm) are not recommended as they lack strength, durability and have sparse to no hair follicles and exocrine glands, resulting in an inferior functional and cosmetic result (Frankland 1979). Greater graft durability and hair coverage is seen when increasing graft thickness to between 0.63 and 0.75 mm (Frankland 1979). According to our microscopic measurements of epidermal/upper dermal thickness, harvesting at a maximum depth of 1.1 mm at the ventral abdomen in WB type horses and at maximum depth of 1.3 mm in CB type horses should ensure harvesting at the upper dermis, at the level of the deepest aspect of follicular units, which could potentially stimulate wound contraction.

The caveat to this recommendation is that the different measuring techniques used in this study produced different results where the biopsy calliper measurements were consistently smaller than the histologic measurements. This discrepancy is most likely explained by passive recoiling of dermal elastin fibres causing immediate primary biopsy contraction, whereas biopsies may re-expand slightly when placed in formalin for fixation (Kerns et al. 2008). Although we believe the histological measurements to be more representative of the true skin thickness than the biopsy measurements, all measurements reported in this study must be interpreted in light of the respective measuring method and cannot be regarded as a definitive assessment of skin thickness.

The high correlation between the calliper and skin fold measurements at the ventral abdomen suggests that the skin fold technique could be used as a non-invasive method of estimating skin thickness in normal condition horses. However, this method is only suitable for certain locations in which a skin fold can easily be manipulated and is therefore restricted to areas with relatively loose skin. When employing this method, one should also keep in mind that the resultant measurement equals twice the skin thickness, with entrapment of subcutaneous tissue caught in the skin fold. Varying amounts of subcutaneous tissue and fat is most likely the reason why there was poor correlation between the skin fold method and the calliper method at the pectoral region. Indeed, in human medicine, skin fold measurements using callipers is an established method for estimation of body fat percentage (Nolte et al. 2016, Amaral et al. 2011), building on knowledge of regional skin thickness. However, in our cases, in order to pick up a skin fold the skin was stretched, resulting in a considerably thinner skin thickness measurement in all cases. Over the 15th rib where picking up a skin fold can prove challenging, the skin fold measured 2.80 ± 0.4 mm, whereas the biopsy measured 3.06 ± 0.75 mm. This shows that even though the skin fold thickness may correlate to skin biopsy measurements it cannot be used as a method to measure the definitive thickness of the skin. Another non-invasive method of measuring skin thickness is ultrasonography, where ultrasonography of up to 25 MHz has been reported as a reliable way of measuring skin thickness in both humans as well as veterinary species (Zanna et al. 2012, Kleinerman et al. 2012, Brown et al. 2000, Bendeck and Jacobe 2007). At present, this modality has not been evaluated in horses and was beyond the scope of this study.

Limitations of the current study include a small study population and the lack of an established gold standard method of measuring skin thickness. However, as no systematic assessment of skin thickness in different horse breeds has been undertaken to date, this study serves as an important step in order to guide skin graft thicknesses and achieve better functional and cosmetic repairs to skin wounds in horses. As such, the results reported in the current study must be interpreted in light of the methods used.

In conclusion, there are large topographical variations in skin thickness in horses, ranging from approximately 1 mm over the inguinal area to almost 6 mm at the dorsum. There are significant differences between horse types, where the skin thickness of CB type horses is thicker than that of WB type horses. Estimation of skin thickness by skin fold measurement is reliable at the ventral abdomen, abaxial to the umbilicus. Estimation of skin thickness in other locations may be better served by using other, non-invasive techniques such as ultrasonography.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

H.M.B.O., J.M.W., M.V. and C.T.F. contributed to conception and study design. Data collection was performed by H.M.B.O. and M.V. Data analysis was performed by H.M.B.O., L.W., C.T.F, and M.V. H.B.M.O. and C.T.F. wrote the manuscript. All authors contributed to critical review of the manuscript and all authors approved the final manuscript and are accountable for all aspects of the work. C.T.F. confirms full access to all data and takes responsibility for data integrity and accuracy of the data analyses.

Funding

Open access funding provided by Norwegian University of Life Sciences.

Data availability

The datasets used and analysed for this study are stored at the NMBU Open Research Data Repository (https://dataverse.no/dataverse/nmbu).

Declarations

Ethical animal research and informed consent

The study was approved by the Norwegian University of Life Sciences Ethical Committee for approval of studies with animal patients (Approval number 14/04723 − 118) and was in accordance with Norwegian legislation regarding use of animals in research (FOR-2015-06-18-761). A signed consent form was obtained from owners of all horses included in the study.

Competing interests

The authors declare no competing interests.