SLR neurodynamic testing range of motion is highly variable, ranging from approximately 15° to over 90° with a moderate association with multiple demographic characteristics, such as sex, weight, BMI and activity level. Specifically, heavier and less active individuals had lower SLR range of motion bilaterally compared to more active individuals who weighed less, just as women had more SLR range of motion bilaterally compared to men. The correlations between these demographic characteristics and overall SLR range of motion were similar bilaterally suggesting that the influence of these factors is equivalent in each limb. Previous studies have found similar variability in SLR range of motion [1, 3, 6, 9] and that females have more SLR range of motion compared to men . Establishing a cutoff for normal SLR range for motion is problematic with such a high degree of variability and with so many demographic characteristics related to mobility.
In contrast, variability in inter-limb differences was much smaller and was independent of these demographic factors. For purposes of generalizability to the greater population, we can use the upper limit of a tolerance interval. Based upon this calculation, we can be 95% certain that “normal” inter-limb differences would be no greater than 10.9° for PF/SLR and 9.4° for DF/SLR in 90% of the general population of healthy individuals. Findings above these ranges could be considered non-normal and potentially important if found in a patient experiencing unilateral lower extremity pain. Further validation for this threshold comes from two previous studies that examined the inter-limb difference in symptomatic individuals. One study found an average of 12° less mobility on the symptomatic side in people with low back pain with or without lower extremity pain with a positive SLR test . The other study found an average of 30° (SD 10°; range 10° to 55°) less range of motion in people with unilateral lumbar radiculopathy .
Utilizing intra-individual, inter-limb differences as the normative standard provides added value because this measurement is independent of various demographic characteristics that commonly impact overall SLR range of motion. In contrast, comparing group means between limbs of healthy, asymptomatic individuals to establish the normative standard for asymmetry in SLR range of motion does not tell the whole story of normal responses to SLR testing. If equal percentages of individuals have greater SLR range of motion on the left (above the y=x line in Figure 1) as do have on the right (below the y=x line in Figure 1), the group averages will equal out and appear to be no different. In fact, we found that considerable intra-individual asymmetries can be present even in healthy, asymptomatic individuals (Figure 2) despite nearly identical group means (Table 2). This is consistent with a previous study where greater than 5° inter-limb differences in ankle range of motion has been documented despite no difference in group mean comparisons . Clinically, intra-individual, inter-limb comparisons are valuable to help determine if neurodynamic involvement is present, which reinforces the need for normative values for this inter-limb difference. Recently, mean inter-limb differences of 7° (6.6° SD) between the dominant and non-dominant limb were documented during upper limb neurodynamic testing . While a threshold level was not presented in this study, one can be calculated from their data using a similar tolerance level upper limit such that we could be 95% certain that 90% of healthy individuals would have no more than a 18.4° inter-limb difference during upper limb neurodynamic testing. This range of “normal” inter-limb differences is higher than in the SLR. We speculate that this difference reflects how asymmetrical use of the upper limbs is more common than for the lower limbs, but further research is necessary to substantiate this hypothesized rationale for the differences noted.
Phase 1 aimed to control the confounding variable of ankle positioning by strict fixation of the ankle position as has been done in previous studies [1–3, 6]. It is equally important to test the reliability and validity of manual fixation of ankle positioning during SLR testing, as was the aim of Phase 2. Previous research has suggested that ankle dorsiflexion to 10° with the knee in full extension and during SLR testing is difficult to achieve and dorsiflexion may be limited to only 4.3-4.8° (SDs: 3.6-4.8°) in this position [6, 19]. For this reason, a neutral ankle position was targeted with DF/SLR in the present study. Repeatability of ankle positioning had good reliability (ICC2,1: 0.78-0.89), but tended to be in 1.1° to 2.0° degrees shy of neutral dorsiflexion at the beginning of testing. On average, the ankle position changed by between 1.4° and 2.7° from the beginning to the end of SLR testing. This suggests that there was a slight shift in ankle position during manual fixation of the ankle, but that the change averaged less than 3° and represents a potential confounding variable that may have influenced the outcome measures. Since there were no significant differences in inter-limb measurements between test phases (Figure 2) and reliability of measuring SLR range of motion was equivalent between phases, the threat to the overall study conclusion is minimal.
The question remains as to why healthy, asymptomatic individuals are not perfectly symmetrical. It is unlikely that sub-clinical nerve injuries are responsible for the asymmetries documented, as all participants had normal lower extremity segmental neurological exams and quantitative sensory testing within normal ranges. Despite considerable efforts to exclude individuals with injuries to the musculoskeletal system, it is possible that some individuals had sub-clinical injuries that were not apparent at the time of enrollment. In the current study, variability in individual activity levels on the MBQ was considerable. According to these results, recreational activities ranged from no primary mode of exercise to running, biking, weight training and participating in group exercise classes. Habitual asymmetrical use of the limbs during daily function and recreation may create asymmetries in the tolerance of the neural tissues to movement. There is considerable evidence that habitual use of our limbs is not symmetrical during activities such as gait initiation , walking [26, 27], turning , jumping [29–31], kicking , and crossing our legs . While 85% of participants in the present study were right hand dominant (for writing) which is similar to proportions presented in previous literature,  a limitation to the present study is that lower limb dominance was not characterized in these individuals. Previous literature using various methods for determining limb dominance has shown a strong association between being right hand dominant and being right foot dominant (75.5%-93.5%), with a slightly lower association between left hand and foot dominance (56.9-79.4%) [34–36]. Lower limb dominance may have influenced the magnitude and direction of inter-limb asymmetries found in this study and further research is necessary to characterize the specific effects of lower limb dominance and asymmetrical activities on SLR range of motion.
Additional limitations include the small number of male participants, as equal distribution of men and women were not sought in this sample of convenience. It should be noted that the impact of sex that has been demonstrated in previous studies  was still evident in the present study despite unequal numbers of males and females. We did not account for the menstrual cycle in women participants, nor did we have participants perform a warm up prior to testing which are additional limitations to the present study, although it is hypothesized that the effect on SLR range of motion would be equal bilaterally and thus not affect inter-limb differences. Additionally, the high reliability demonstrated in the present study is limited to intra-rater, intra-session and cannot be extrapolated to comparisons between raters or between sessions measurements. Lastly, it is possible that small but clinically relevant correlations exist between demographic characteristics and range of motion measures that we were unable to detect due to inadequate power of the present study to detect correlations of 0.35 or less.