Materials and Methods

Animals—Twenty-four mixed-breed horses were used to establish reference MNT values of the thoracic limbs. Horses included 11 mares and 13 castrated males with a median age of 3 years (range, 2 to 5 years), mean body weight of 391 kg (range, 318 to 477 kg), and mean wither height of 146 cm (range, 139 to 153 cm). Horses were acquired from a local agent who did not have documented performance or medical histories for them. All horses were broke to lead, although many horses appeared to have limited handling or training experience. Horses were stall-confined, had no turnout, and were in the initial process of being acclimatized to treadmill exercise. Inclusion criteria for a concurrent study required horses between 2 to 5 years of age that had no clinical evidence of thoracic limb lameness during in-hand gait evaluation on a hard surface and had normal radiographic images of both carpi. The lameness examination and review of the carpal radiographs were completed by 2 blinded, board-certified examiners prior to the pressure algometry measurements. All protocols were approved by the Animal Care and Use Committee at Colorado State University.

MNT measurements—All horses were quietly restrained in their stalls with a halter and lead rope. A pressure algometer^a with a 1-cm² rubber plunger tip and a calibrated range from 0 to 30 kg/cm² was used to determine MNT values. Pressure was applied perpendicularly to predetermined anatomic landmarks at approximately 10 kg/cm²/s over 2 to 3 seconds until a local avoidance reaction was observed or until the upper limit of the gauge (ie, 30 kg) was reached. Avoidance reactions included skin twitching, local muscle fasciculations, lifting the thoracic limb, or stepping away from the applied pressure. When a reaction was observed, the applied pressure was stopped immediately and the corresponding value recorded. The examiner did not view the MNT reading during the application of pressure to limit potential bias. The instrument automatically recorded the highest pressure attained and was reset to zero after each measurement. Three consecutive measurements at 3- to 4-second intervals were recorded at each site to assess within-site repeatability. The median of the 3 consecutive measurements at each site was used as the site-specific baseline MNT value for that horse. A single examiner (KKH) performed all measurements.

Pressure algometry was used to establish reference MNT values at 17 anatomic locations within each thoracic limb, which were chosen on the basis of the ease and long-term consistency of identification (Appendix). A fixed-order protocol was used to reduce between-subject variability. All musculoskeletal landmarks were tested in a proximal-to-distal order, beginning with the left thoracic limb and then the right thoracic limb. Measurement sites within each thoracic limb included osseous (n = 12) and soft tissue landmarks (n = 5). Each measurement session lasted approximately 20 minutes.

To assess adaptation or sensitization to the procedure, the 3 consecutive measurements at each site were evaluated for sequential increases (ie, adaptation), sequential decreases (ie, sensitization), or no change or consistent patterns.⁸ The proportions of the 3 patterns and the ranges of the 3 measurements at each landmark were recorded. The mean range (across sites and horses) was interpreted as a measure of overall repeatability. To further assess repeatability of the initial MNT values, a second series of baseline MNT values were obtained an average of 5 days (range, 1 to 7 days) after the initial MNT measurements. The percentage change between the 2 baseline measurements was calculated.

Left-to-right paired comparisons at each site were made to determine whether bilateral MNT values could be pooled into a combined value. For purposes of regional comparisons, the landmarks were grouped into the following 4 anatomic regions: the caudal cervical portion of the vertebral column and scapula, brachium and antebrachium, carpus, and the distal portion of the limb. Mechanical nociceptive thresholds were pooled within each of the 4 anatomic regions and compared within and between each thoracic limb. Mechanical nociceptive thresholds were also pooled for bony and soft tissue landmarks and compared within and between baseline measurements. Effects of age (< 3 years; ≥ 3 years), sex, weight (< 390 kg; ≥ 390 kg), and wither height (< 146 cm; ≥ 146 cm) on MNT values for the initial baseline time point were assessed separately for each site. The relative categoric differences were reported.

A separate nonblinded examiner subjectively graded (on a 0 to 5 scale) the overall willingness or tolerance of horses to stand quietly during the MNT measurements. A grade of 0 corresponded with repeat lifting of the limb and an inability to stand still on the thoracic limb for most MNT measurements. Grade 5 was assigned if the horse stood completely quietly and readily tolerated all MNT measurements. Grades 1 through 4 were intermediate gradations of increasing tolerance to the repeated pressure measurements. Grade 1 corresponded with standing completely still for 20% of MNT values, grade 2 was 40%, grade 3 was 60%, and grade 4 corresponded with standing quiet for 80% of MNT measurements.

Surgically induced osteoarthritis—As part of a concurrent study, all horses underwent bilateral arthroscopic surgery of the middle carpal joint to ensure no preexisting abnormalities. In 1 randomly selected middle carpal joint of each horse, an OCF was created on the distal aspect of the radial carpal bone to induce osteoarthritis.¹² This joint was referred to as the osteoarthritic joint or limb and the opposite sham-operated joint as the control joint or limb. Bandages were applied to the distal portion of the thoracic limbs and changed every 3 to 5 days for 2 weeks. After surgery, all horses received ceftiofur sodium^b (3.5 mg/kg, IM) twice daily for 3 days and phenylbutazone^c (2 g, PO) once daily for 5 days. Throughout the study, postoperative pain was assessed by observation daily in all horses, and if a horse was judged to be in pain, then an appropriate analgesic and antimicrobial was prescribed. Beginning on day 3 and continuing weekly, MNT values were recorded for the control and osteoarthritic limbs and compared to determine whether pressure algometry was able to differentiate sites of pain (ie, osteoarthritic sites) from proximal or distal sites within the same or opposite limb.

In the concurrent study, horses were randomly assigned to 1 of 3 treatment groups involving intra-articular injections into the middle carpal joints on day 14. The control group (n = 8) had 2 mL of saline (0.9% NaCl) solution injected into both middle carpal joints. The other 2 groups had either 2 mL containing 16.3 million nucleated adipose-derived stromal vascular fraction stem cells (n = 8) or 10.5 million bone marrow–derived mesenchymal stem cells (n = 8) injected into the osteoarthritic joint, and 2 mL of saline solution was injected into the contralateral control joint. After intra-articular injection, all horses again received ceftiofur sodium^b (3.5 mg/kg, IM) twice daily for 3 days and phenylbutazone^c (2 g, PO) once daily for 5 days.

Beginning on day 14, clinical examinations by a board-certified surgeon were performed biweekly on both thoracic limbs, which included a lameness evaluation, response to carpal joint flexion, and assessment of carpal joint effusion. Lameness at a trot was graded with the lameness scoring system of the American Association of Equine Practitioners of 0 to 5, with 0 indicating a normal gait and 5 representing nonweight-bearing lameness. Carpal joint flexion was graded on a scale of 0 to 4, with 0 indicating no response and 4 representing a severe response. Carpal joint effusion was graded on a scale of 0 to 4, with 0 indicating a normal amount of joint fluid and 4 representing a large increase. Beginning on day 15 after arthroscopic surgery, horses were exercised on a high-speed treadmill for 2 minutes each at trot (4.5 m/s), gallop (10.5 m/s), and trot again for 5 d/wk until the end of the study to promote the development of osteoarthritis of the carpal joint. Throughout the study, lameness was assessed daily in each horse, and if judged clinically relevant, then treadmill exercise was discontinued. All carpal joints were radiographed again at the conclusion of the study and graded by a board-certified radiologist for bony proliferation at the joint capsule attachment (ie, enthesophyte); subchondral bone lysis; and osteophyte formation with a 0 to 3 scale for each variable, with 0 indicating normal and 3 the most severe change.

All horses were euthanatized by IV administration of sodium pentobarbital^d (88 mg/kg) approximately 70 days after surgery. At necropsy, the middle carpal joints were examined and graded for articular cartilage erosion and synovial membrane hemorrhage on a 0 to 4 scale, with 0 indicating no change and 4 representing a severe change. Middle carpal joint articulations were also evaluated for the absence or presence (0 to 1 score) of contact lesions on the opposing articular surfaces and synovial membrane adhesions. Correlations between MNT values and the clinical, radiographic, and necropsy scores were determined.

Statistical analysis—Eighty-two percent (28/34) of MNT variables were normally distributed on the basis of Komolgorov-Smirnov tests; therefore, parametric tests were used for most analyses. Paired t tests (2-tailed) were used to test all initial baseline differences in MNT values of left versus right limbs, osteoarthritic versus control limbs, and within-horse between session values. Our hypothesis was that osteoarthritic limbs would have lower MNT values, compared with control limbs; therefore, 1-tailed tests were used for after-surgery osteoarthritic versus control limb comparisons of MNT values. The percentage change in MNT values from initial baseline measurements was calculated. Regional differences in pooled MNT values were assessed by a 1-way ANOVA and the Tukey honestly significantly different method for post hoc comparisons of individual means. Differences in pooled MNT values for bony and soft tissue landmarks were tested with 2-sample t tests (2-tailed). Effects of age, sex, weight, and wither height on MNT values were assessed with 2-sample t tests (1-tailed), based on the results of a prior study assessing the interactions of signalment and MNT values.⁸ Signalment interactions were further assessed with a 1-way ANOVA. Tolerance to the procedure, clinical, and radiographic scores were not normally distributed; therefore, the Spearman rank correlation test was used to assess correlations between the MNT values and these variables. Kruskal-Wallis test with post hoc comparisons of mean ranks (A = 0.05) was used to evaluate intra-articular treatment group differences. Significance for all tests was set at P ≤ 0.05.

Results

Tolerance scores—A mean grade of 4.1 (out of 5) for tolerance to the procedure in both thoracic limbs was found for the initial baseline time point. For the second baseline time point, the mean tolerance scores decreased slightly to 3.6 in the left thoracic limbs and to 3.8 in the right thoracic limbs. Reduced tolerance scores for the second baseline time point were the result of 3 horses that refused to stand quietly for most measurements and were assigned a grade of 0. No significant differences in tolerance scores were found between left versus right limbs or between baseline measurements. Tolerance to the procedure was positively correlated with overall limb MNT values for the right thoracic limb at baseline 1 and for both thoracic limbs at baseline 2 (Table 1). After surgery, no significant differences were found in tolerance scores between control versus osteoarthritic limbs. Over the course of the study, 4 of 24 (17%) horses developed a resistance to stand for the procedure (ie, tolerance score of 0) and subsequently did not have a complete data set for all sites.

Table 1—

Spearman rank correlations between the tolerance to procedure gradations and the pooled regional and overall limb MNT values of 24 horses during the initial (1) and second (2) baseline time points.

View Table

Repeatability—No significant differences were found in the proportion of patterns or range of 3 consecutive MNT measurements between left versus right limbs or between baseline measurements. Across baselines, the 3 consecutive measurements sequentially increased in 20 ± 11%, sequentially decreased in 8 ± 8%, and had no change or consistent pattern in 72 ± 11% of measurements. The mean range of 3 consecutive measurements across baselines was 2.0 ± 1.4 kg/cm². Across all weeks after surgery, the 3 consecutive measurements sequentially increased in 16 ± 10%, sequentially decreased in 8 ± 7%, and had no change or consistent pattern in 76 ± 13% of measurements. The mean range of 3 consecutive measurements across all weeks after surgery was 1.8 ± 1.3 kg/cm².

Differences between left versus right limbs—For the initial baseline time point, the mean within-horse difference between left versus right limbs was 0.6 ± 0.9 kg/cm². Across sites, left thoracic limb MNT values were lower than the paired right thoracic limb MNT values at 15 of 17 (88%) sites (Figure 1); however, differences between left versus right limbs were not significant at 14 of 17 (82%) paired sites. Two sites had significant differences between left versus right limbs of < 1.1 kg/cm², which was within the measurement error of the instrument. Differences between left versus right limbs of > 2.0 kg/cm² were measured at the metacarpophalangeal (fetlock) joint (3.4 kg/cm²). For the second baseline time point, the mean left versus right limb difference in MNT values was 0.6 ± 0.6 kg/cm². Left limb MNT values were again lower than right limb MNT values at 16 of 17 (94%) paired sites, although no significant differences were found between left versus right limbs at 12 of 17 (71%) sites. Four sites had significant differences between left versus right limbs of < 1.4 kg/cm². Differences between left versus right limbs of > 2.0 kg/cm² were measured at the accessory carpal bone (2.3 kg/cm²) during the second baseline time point. Within each session after surgery, significant differences were found between left versus right limbs at 25% (3/12) to 53% (9/17) of sites, with left limb MNT values less than right limb MNT values. Differences between left versus right limbs had minimal effects on the after-surgery portion of the study because the left and right limbs were randomly assigned to control or osteoarthritis groups; therefore, the direction of limb bias was evenly distributed between the 2 groups.

View Full Size

Mean ± SD left-and-right pooled baseline MNT values organized anatomically from the proximal-to-distal aspect of the limbs of 24 horses. Vertical lines demarcate the pooled regional landmarks. See Appendix for an anatomic description of musculoskeletal landmarks 1 through 17.

Citation: American Journal of Veterinary Research 68, 11; 10.2460/ajvr.68.11.1167

• Download Figure
• Download figure as PowerPoint slide

Baseline measurements—Between baseline measurements, MNT values increased 4.0 ± 5.6% (1.2 ± 2.4 kg/cm²), 3% within the left thoracic limb and 6% within the right thoracic limb. No significant baseline measurement differences in MNT values were found at 26 of 34 (76%) sites. Baseline measurement differences of > 2.0 kg/cm² were measured at the following 5 sites: the left intermediate carpal bone (2.7 kg/cm²), third metacarpus (3.0 kg/cm²), fetlock joint (6.1 kg/cm²), right third metacarpal bone (3.2 kg/cm²), and coronary band (2.3 kg/cm²). Significant increases in pooled MNT values were detected between baseline measurements for the carpal and distal limb regions (Table 2). The pooled MNT value for bony landmarks was significantly higher than for soft tissue landmarks for both baseline measurements. Between baseline measurements, a significant increase in pooled MNT values was found for bony landmarks but not for soft tissue landmarks.

Table 2—

Mean ± SD left and right pooled thoracic limb regional MNT values and pooled bony and soft tissue MNT values of 24 horses for the initial (1) and second (2) baseline time points

View Table

The timing of the second baseline measurement appeared to have an effect on the percentage change in MNT values between sessions. When baseline measurements were recorded closer together (ie, within 1 to 3 days; n = 7 horses), an overall 10 ± 7% increase in MNT values was found between baseline measurements; 24 of 34 (71%) sites had differences of > 2.0 kg/cm², of which 7 of 34 (21%) sites had significant differences in MNT values. In contrast, when baseline measurements were recorded further apart (ie, within 5 to 7 days; n = 17 horses), an overall 1 ± 8% increase in MNT values was found between baseline measurements; 12 of 34 (35%) sites had differences of > 2.0 kg/cm², of which 10 of 34 (29%) sites had significant differences in MNT values.

Only differences in MNT values of greater than ± 2.0 kg/cm² were judged important for signalment comparisons (Table 3). No significant effects of age or weight were found on MNT values. Geldings had significantly higher MNT values than mares at 17 of 34 (50%) sites. Mechanical nociceptive thresholds were significantly higher in tall horses at 10 of 34 (29%) sites. The proportion of tall horses did not differ significantly by sex.

Table 3—

Comparison of thoracic limb MNT values within signalment categories of 24 horses.

View Table

Differences between control versus osteoarthritic limbs—At both baselines, no significant differences in MNT values were found between control and osteoarthritic limbs. No significant intra-articular treatment differences in the clinical, radiographic, or necropsy scores or MNT values were found between the bone marrow– or fat-derived stem cells and the saline control groups. Therefore, for the purposes of this study, all treatment group data were pooled into control and osteoarthritic limb groups. We were not able to measure MNT values at the carpal and distal portion of the limb landmarks at day 3 and week 1 because of the presence of bandages after surgery in the region. On day 3, 2 of 10 (20%) sites had significantly higher MNT values within the osteoarthritic limb, compared with the control limb. During weeks 1 to 4, 18% (3/17) to 42% (5/12) of sites had significantly lower MNT values within the osteoarthritic limbs, compared with the control limb. During week 1, 3 sites within the caudal cervical portion of the vertebral column had significantly lower MNT values within the osteoarthritic limb, possibly resulting from the presence of limb sensitization from the arthroscopic surgery. Sites within the carpal and distal portion of the limb regions (ie, from the intermediate carpal bone to the suspensory ligament) had significantly lower MNT values within the osteoarthritic limb that persisted from weeks 2 to 5, indicating increased pain at the carpal and distal portion of the limb landmarks. From weeks 6 to 9, few significant differences in MNT values were found between control versus osteoarthritic limbs. One horse had increased lameness, heat, and swelling in the osteoarthritic carpal joint between week 4 and 5, which was successfully treated with antimicrobials and phenylbutazone for 5 days. For this horse, treadmill exercise was discontinued for 7 days to allow resolution of the lameness. Data from this horse were removed for further analysis from week 4 to the end of the study.

Percentage change from initial baseline measurements—Significant decreases in pooled MNT values were found at day 3 for control and osteoarthritic limbs and at week 1 for the osteoarthritic limbs, compared with initial baseline measurements, which indicates possible detection of pain associated with arthroscopic surgery (Table 4). At most sites within the control and osteoarthritic limbs, significant increases in MNT values were found at week 1, which reached a plateau and remained increased throughout the remainder of the study. The increased MNT values likely represent early adaptation to the applied pressure. However, many of the variables did have the highest MNT values recorded during the last session of the study, which was indicative of continued adaptation over time.

Table 4—

Percentage change from the initial baseline MNT values at sites within the control and osteoarthritic limbs of 24 horses (expanded version of data is available from the authors upon request).

View Table

For the control limb, significant increases (up to 35%) were found at all 4 sites within the brachial and antebrachial regions. Decreased (ie, negative) MNT values recorded at the accessory carpal site from week 2 to week 6 were likely the result of bandage sores over the caudal aspect of the accessory carpal bone. From week 2 to week 9, significant increases were detected sporadically within the control limb at the carpal sites, as well as individual sites within the caudal cervical portion of the vertebral column and scapula region and distal aspect of the limb region. The coronary band had consistent increases in MNT values that remained 16% to 25% above initial baseline measurements.

Within the osteoarthritic limb, significant decreases in MNT values were detected at sites within the carpal region from week 2 to week 6, which suggests the detection and gradual resolution of bone and joint pain associated with the induced OCF (Table 4). Fewer significant increases were found at the distal portion of the limb sites, compared with the control limb, which indicates less adaptation to the pressure algometry within the osteoarthritic limb. Across all weeks, MNT values were similar for the control and osteoarthritic limbs, except at the carpal sites.

Clinical, radiographic, and necropsy scores—Lameness, flexion test, and joint effusion scores all had significant increases within the osteoarthritic limb to a grade 2 (out of 4 or 5) from week 2 to the end of the study. Radiographic and gross necropsy variables were also significantly increased within osteoarthritic joints, compared with control joints. In general, MNT values were poorly correlated with lameness, flexion test, and joint effusion scores. However, during weeks 2 and 4, significant negative correlations between lameness scores and MNT values were found at the third carpal, third metacarpal, fourth metacarpal, and fetlock joint sites within the osteoarthritic limb. No significant correlations were found between MNT values and the radiographic or necropsy scores for either the control or osteoarthritic limbs.

Discussion

Reference MNT values for the thoracic limb were established by use of pressure algometry. Reference MNT values increased from proximal-to-distal sites of the thoracic limbs. Baseline MNT values were affected by sex, wither height, and the bony or soft tissue composition of measured sites. Although long-term repeatability of MNT measurements was good, short-term adaptation to the pressure algometry procedure was evident. Mechanical nociceptive thresholds were decreased significantly in the carpal region of limbs with osteoarthritis, compared with control limbs; however, MNT values were poorly correlated with clinical examination, radiographic findings, and necropsy scores.

The tendency for increased MNT values within distal versus proximal limb sites supports previous research. In humans, a similar increasing gradation in MNT values was observed from the cervical portion of the vertebral column to the distal portion of the thoracic limb; MNT values over the cervical portion of the vertebral column are significantly lower than MNT values of the thoracic limb.^14,15 These regional differences might be related to the local concentration of nociceptors or the result of increasing length of the spinal nerves of the thoracic limb.^7,16 In the current study, MNT values may also have been affected by the mechanics of limb withdrawal and the temperament of the horses. Thresholds were highest within the carpal region, possibly as a result of the mechanical application of pressure along the dorsal or extensor surface of the thoracic limb, which hampered carpal joint flexion and subsequent limb withdrawal, producing a delayed avoidance reaction. Nociceptive thresholds were comparatively lower at the accessory carpal bone, fourth metacarpus, and lateral branch of the suspensory ligament, which were located along the lateral and palmar (ie, flexor) surfaces of the distal portion of the limb. The caudal cervical portion of the vertebral column and scapular sites in the current study were, on average, 15% lower than equine MNT values at the same sites reported within the axial skeleton.⁸ These differences are likely the result of the unbroken and unridden horses used in the current study, compared with frequently handled or ridden horses used in the prior study.

Compared with mares, geldings had higher MNT values of the thoracic limb, which is consistent with reported findings of MNT values within the axial skeleton.⁸ Potential hormonal and behavioral differences between geldings and mares may account for these differences in MNT values. Similar sex differences in nociception have been documented in humans.¹⁷ Specific etiologies for the sex differences in MNT values in humans have not been clearly defined, although, psychosocial, physical, neural, and hormonal factors have been investigated.^10,18 In the current study, taller horses had higher MNT values, which could not be explained by sex differences because tall horses were equally distributed among both mares and geldings. We are not able to provide a biological reason why taller horses would have increased MNT values of the thoracic limbs. Age and weight differences in the baseline MNT values were not found, presumably because of the selection of a uniform population of young and lightweight horses. In a prior study⁸ on horses with a wider range of ages and body weights, higher MNT values were detected in older and heavier horses. In the current study, MNT values were higher at bony landmarks than soft tissue landmarks, which may be the result of differences in mechanical stimulation or variations in the concentration or depth of nerve fibers.^7,19 Similarly, within the axial skeleton, most bony landmarks have higher MNT values than adjacent, paired soft tissue sites, except for the cervical portion of the vertebral column, where ep- axial muscle sites have higher MNT values than the adjacent transverse processes.⁸ In humans, no consistent differences have been reported between muscle and periosteal sites within the same anatomic regions.⁷

Repeatability of 3 consecutive MNT measurements was lower in the current study than in a previous report⁸ on MNT values of the equine axial skeleton. Differences in the physical processes of responding to applied pressure within the axial versus appendicular skeletons may account, in part, for differences in repeatability of consecutive MNT measurements.⁸ Withdrawal or avoidance reactions of the thoracic limbs are more difficult because of the effects of weight bearing. During pressure algometry of the appendicular skeleton, horses are required to repeatedly shift their body weight away from the tested limb in order to unweight and then withdraw the limb. Within the axial skeleton, MNT measurements evoke a primary avoidance reaction and do not require the prerequisite redistribution of body weight, thereby producing more consistent avoidance reactions. This was supported by reduced variability in MNT values at proximal sites within the proximal portion of the thoracic limb, compared with distal sites. The increased variability within the distal portion of the limb may have also contributed to the reduced ability to measure significant differences between the control and osteoarthritic limbs.

The right limb had significantly higher MNT values than the left limb at several sites. Although most left versus right limb differences were not significant at the start of the study, differences between left versus right limbs increased over the course of the study. These differences could be the result of the fixed-order design of the study or right-side dominance in the studied horses. In the current study, we used a fixed order to measure MNT values, with the left limb always evaluated prior to the right limb. Adaptation within the right limb caused by the prior pressure application within the left limb may explain the consistently higher MNT values within the right limb. If the order of the sites and the limbs had been randomized, we would have had a better measure of overall variability of the procedure but would have lost the ability to specifically measure the order bias. Alternatively, differences between left versus right limbs may have been attributable to rightside dominance. In most human studies, left-side MNT values are typically lower than the comparable rightside landmarks as a result of the influence of right-hand dominance.²⁰ Left- or right-side dominance was not addressed in this study but warrants further investigation as a potential contributing factor to MNT asymmetries in horses.

Mechanical nociceptive values can be affected by examiner, rate of pressure application, and adaptation of the subject to the procedure.²¹ In the current study, all MNT measurements were made by a single examiner. Although interexaminer repeatability has been reported to be good in human subjects,²² future studies are needed to assess interexaminer repeatability in horses. The current study used a high rate of pressure application, which was not objectively monitored. Slower rates of pressure application are likely to produce lower MNT values.²¹ Adaptation to the procedure in the current study is similar to that reported for other species and appears to be affected by time between algometry sessions. In humans, short-term repeatability over several days, at 1-week intervals, or over several weeks is good.^7,19,23 In the current study, no significant differences were found in MNT values between the first and second baseline time points at most sites; however, pooled regional MNT values of the carpus and distal portion of the limb for the second baseline time point were significantly greater than for the initial baseline time point, likely caused by mild adaptation to the pressure measurements. The timing of the second baseline measurement did have an effect on MNT values. In horses where the 2 baseline measurements were recorded closer together at 1 to 3 days, large increases in MNT values were found between baseline measurements, which suggest short-term adaptation to the procedure. It is possible that when MNT values were recorded further apart at 5 to 7 days, the learning response or adaptation to the procedure was reduced or lost, as evidenced by a smaller difference between the 2 baseline measurements. After surgery, adaptation continued gradually for 10 days (ie, 4 measurement sessions), and at most sites, MNT values remained at a consistent level throughout the remaining 8 weeks (ie, 7 measurement sessions) of the study.

The algometry procedure was well tolerated by most horses on the basis of subjective grading of each horse's willingness to stand for the procedure. Similar findings have been reported for humans, in whom pressure algometry is generally well tolerated and even described as pleasant.⁷ Intolerance to pressure algometry appeared to be associated with behavioral issues or nociception. Tolerance to procedure scores were positively correlated with most overall pooled MNT values for each thoracic limb, which suggests a clinical relationship between observed pain behavior (as defined by the tolerance scores) and MNT values. Four horses refused to stand for the MNT measurements; refusal appeared to be related to behavioral responses that developed subsequent to repeated procedures directed at assessing osteoarthritis of the carpal joint in the concurrent study and did not appear to be associated with pain produced by the pressure algometry procedure itself, although this could not be confirmed.

Although MNT values were decreased significantly in the carpal region of limbs with osteoarthritis, compared with control limbs, MNT values were poorly correlated with clinical examination, radiographic findings, and necropsy scores. Poor correlation may have been caused by differences among the diagnostic methods being compared. The characterization and differentiation of joint pain are difficult and often require a multimodal approach to diagnose conditions and manage affected patients.²⁴ The clinical assessment of musculoskeletal pain by digital palpation is often subject to observer bias and interexaminer variability.²⁵ Pressure algometry provides a quantitative assessment of mechanical pain by the application of blunt pressure.⁴ In humans, myofascial pain and fibromyalgia are frequent causes of chronic pain and are characterized by low mechanical pain thresholds.²⁵ Pressure algometry has been useful in the diagnosis and management of these conditions by identifying painful and nonpainful tissues within local and regional landmarks. Degenerative changes within joints are frequent causes of pain; however, these changes are not always painful. Pain experienced in the vicinity of an osteoarthritic joint may be referred from periarticular or distant structures.²⁴ Similar to the findings of the current study, correlation between MNT values and other subjective measures of pain intensity (ie, visual analogue scale and McGill Pain Questionnaire) is poor in humans with osteoarthritis of the knee joint.¹³ In humans and horses, different test methods may measure different aspects of the pain experience. In humans, MNT repeatability was reported to be better than other subjective measures of pain and MNT values could differentiate patients with osteoarthritis from healthy control subjects with moderate reliability.¹³ Musculoskeletal structures affected by adaptation or sensitization within adjacent or distant regions (eg, compensatory lameness) have the potential to be identified and localized with pressure algometry, which is currently difficult to quantify with most clinical examination techniques.²⁶

Nociceptive thresholds may provide a more objective and consistent long-term assessment of the presence and resolution of thoracic limb pain than more subjective methods of lameness evaluation (ie, flexion tests or joint effusion).²⁷ Alternatively, poor correlation between MNT values and other diagnostic methods may have been the result of the nature of the induced injury and the relatively short time frame of the current study. More advanced osteoarthritis or severe soft tissues injuries may have produced stronger correlations between MNT values and the lameness, flexion test results, and joint effusion scores.²⁸ The relatively mild clinical disease and short duration of the study reduced the likelihood of the development of peripheral or central sensitization in the current study. Naturally occurring osteoarthritis is likely to have regional and systemic effects on nociceptive thresholds.²⁴ Reduced pain thresholds have been reported for human patients with unilateral osteoarthritis of the hip joint in the affected hip and the unaffected contralateral hip.²⁹ A similar central mechanism may have affected the results in the current study.

The nature of the study design impeded the ability to assess acute pain after surgery. A decrease in MNT values was expected in osteoarthritic and control limbs after surgery because both received arthroscopic surgery. Unfortunately, MNT values at the surgical sites could not be measured until 2 weeks after surgery because of the presence of bandages. In addition, systemic effects of phenylbutazone administration on MNT measurements obtained on day 3 and week 2 are unknown; however after week 2, the acute effect of the arthroscopic procedure within the control limbs could be differentiated from the additional presence of the OCF within the osteoarthritic limbs. In the current study, pressure algometry was able to localize a known source of pain of the thoracic limb but was poorly correlated with other clinical measures of osteoarthritis. Pressure algometry may prove to be an important adjunct to the traditional orthopedic or lameness examination. Further studies are needed to measure MNT values in naturally occurring osteoarthritis and in the treatment of other chronic musculoskeletal conditions.