Our results demonstrate that SWD was feasible to obtain using commercially available digital scales in a population of neurologically normal dogs ranging from 5 to 20 kilograms in BW. Using two digital bathroom scales (B2) was the simplest and most reliable technique, and resulted in a thoracic limb to pelvic limb SWD of 63% to 37%. The other scale methods might be useful in specific scenarios such as using the kitchen scales for very small dogs, or using one of the four scale methods where capturing left to right asymmetry is important.
We tested four different digital scale combinations and the B2 method was by far the most feasible method across a broad BW range of 5 to 20kg. While the B4 and K4 methods also provided adequate feasibility, they were limited to dogs of higher and lower BW, respectively. This reflects the lower and upper weight ranges of these commercially available scales, reported to be a 1.4kg minimum for the bathroom scales and a 5kg maximum for the kitchen scales. The K2 method was even more limited to smaller dogs weighing less than approximately 9kg. A high level of feasibility of bathroom scales to measure SWD has been previously reported in normal dogs and dogs with osteoarthritis.22 While the prior study evaluated large breed dogs weighing greater than 20kg, our results confirmed that these techniques are straightforward to employ across dogs of different sizes.
We recruited dogs that were amenable to handling and, therefore, encountered only a single dog in which measurements were limited by a behavioral issue (measurements were initially obtained easily but the dog tired of the handling and became aggressive). To facilitate cooperation and tolerance, we allowed all dogs several minutes to acclimate to the procedures and to having their limbs manipulated before starting to collect measurements and took breaks as needed. It is possible that behavioral issues will be more apparent in dogs that might be painful secondary to IVDH or surgery. However, similar measurements were obtained in 41/43 dogs with osteoarthritis with only two dogs unable to participate due to behavioral issues.22 This suggests that feasibility when translated to a clinical population would still be expected to be high especially since all scale methods and the PSW would not typically need to be performed in clinical patients.
Variability between trials made in triplicate was lowest for the two scale methods, B2 or K2. While there was some inter-individual variation (coefficients of variation ranging from 1.7 to 18.2%), the overall variability between trials was less than 10% for both methods. It was easiest to position the dogs appropriately when the thoracic and pelvic limbs were each contained on a single scale and the dogs appeared most comfortable, resulting in minimal cranial to caudal weight shifting while performing the task. Our results are similar to Levine et al. where SWD was evaluated in 10 healthy, large breed dogs.23 The reported mean coefficient of variation across all measurements (also made in triplicate) was 4.0% (0-24%); coefficients of variation for the thoracic and pelvic limbs were reported to be 14% and 13%, respectively.23
With the B4 or K4 methods, dogs appeared to stand squarely; however, any slight shift in position such as mild head movements, likely contributed to greater variation between trials. This variability ranged from approximately 10-21% across dogs and 0-48% between trials in individual dogs. Similarly, left and right pelvic limb SWD measurements have been previously reported in large breed dogs to have an overall reliability of 76% including 61% for normal dogs and 79% for the osteoarthritis group.22 The authors suggested the lower test-retest repeatability in control dogs might be explained by a lack of need to focus on weight bearing between limbs in a normal dog. While we made every attempt to ensure dogs were standing still and squarely, no dogs were specifically trained to stand. Small, random shifting between limbs might be normal in healthy dogs. Considering feasibility and variability together, the B2 method provided the simplest and most robust means to measure SWD in this population of normal, small breed dogs. The other methods were adequate and offer specific situations in which they might be a useful adjunct to the B2 method.
Across all scale methods, there was a mean thoracic limb SWD of 59-63% and mean pelvic limb SWD of 37-41% of total BW. This compares favorably to the distribution reported in normal, large breed dogs, 64% to 36%, respectively.22,23 Interestingly, using the PSW, an established means of measuring dynamic weight distribution in dogs, we found a mean thoracic to pelvic limb SWD of 68% to 32%. This was significantly different from our scale methods with greater weight borne on the thoracic limbs. The reason for this difference is not clear, but it might relate to the manner of testing. For the PSW, dogs were commonly walked and stopped partway across the mat for testing. Limbs were still adjusted as needed to ensure dogs were square, but walking first might have led to artificially increased thoracic limb weight bearing upon stopping, or dogs might have leaned forward slightly in anticipation of starting to walk forward again. Alternatively, since the walkway is a continuous surface, it is possible their stance on the PSW was the most natural and therefore more accurate.
For the four scale methods (B4 or K4), the mean left to right asymmetry in SWD between thoracic limbs was approximately 8% +/-7% and for the pelvic limbs was approximately 4% +/-3%. This is compatible with prior studies in normal large breed dogs showing mild asymmetry between pelvic limbs when standing (3.3% +/- 2.7%) and in limb biomechanics when trotting.22,26 It is possible the left to right difference in our population reflected underlying disease; however, all of the dogs had normal orthopedic and neurologic examinations and no history of prior orthopedic or neurologic abnormalities. The greater degree of asymmetry observed in the thoracic limbs might be directly related to increased distribution of weight on these limbs or be attributed to head and neck position. Other contributing factors might include right or left dominance, similar to handedness in people, or small conformational discrepancies between limbs, which have been reported in dogs.26-28 In general, our data support that mildly asymmetrical SWD occurs in normal dogs and should be taken into consideration when interpreting individual limb values in a clinically abnormal population.
The SWD data obtained in this population of normal dogs will allow for subsequent comparison to dogs recovering from IVDH, the most common cause of acute thoracolumbar SCI in dogs.25 While it has been reported that more than half of neurosurgeons now recommend post-operative rehabilitation for dogs with acute thoracolumbar IVDH, additional validated outcome measures are needed to evaluate the role of rehabilitation in dogs recovering from SCI.3 There are a number of evaluation methods currently available to monitor changes in gait.11-20 Several of these gait scales have been shown to be reliable across raters with broad experience levels.24 A battery of neurologic function tests that broadly assesses motor function called the FINFUN was also recently developed and validated in a group of dogs recovering from SCI.21 However, there is a paucity of objective measures to quantify and track other aspects of functional recovery, such as weight distribution.
Abnormal weight shifting has been suggested in dogs with thoracolumbar SCI.15,18,20 Increased forward loading of weight puts excess strain on cervical and thoracic limb muscles and joints and might cause myofascial pain, exacerbate osteoarthritis or otherwise negatively impact mobility. Decreased loading of abnormal limbs has been suggested to contribute to widespread nervous system and musculoskeletal changes including abnormal coordination, altered proprioception, impaired peripheral nerve health, muscle atrophy and weakness, decreased bone density and decreased overall joint, ligament and tendon health.29-33 We anticipate that quantifying SWD will complement gait analysis tools and provide an additional target that might be useful in the periodic monitoring of neurologically abnormal patients undergoing rehabilitation.