Date of Award


Document Type


Degree Name

Master of Science (MS)


Biomedical Engineering



Research Advisor

William M. Mihalko


Brooke A. Sanford Kunal Singhal Audrey R. Zucker-Levin


Gait, Low Back Pain, Lumbar Lordosis, Pelvic Tilt, Spine Kinematics, Tight Hamstrings


A direct correlation between hamstring tightness and severity of lower back pain (LBP) has been previously reported. Hamstring contraction creates knee flexion, hip extension and posterior pelvic tilt. Posterior pelvic tilt causes the lumbar spine to flatten (hypolordosis) which places pressure on the anterior structures of the spine, including the intervertebral disc. This increased pressure may predispose individuals to disc degeneration and associated pain. For this reason, hamstring stretching is often prescribed to alleviate LBP, however, the effect of hamstring lengthening on the kinematics of the lumbar spine and pelvis is not well understood. Accurate measurement of lumbar spine movement is difficult to obtain in a motion capture laboratory due to subtle differences in marker placement, tissue interface and distribution. Therefore, the purpose of this study was twofold. The first goal was to develop and validate a method that allows accurate comparison of LL measurements from different testing sessions. Secondly, we aimed to assess the effect of hamstring lengthening on anterior pelvic tilt (PT) and lumbar lordosis (LL) in healthy subjects during gait, sit to stand, forward bend, and picking up a ball from the floor. Two different methods were developed and tested to determine which provided the most accurate LL measure from motion lab data. First, the LL correction method was developed and tested by using EOS bi-planar X-Rays of twenty-one healthy subjects. Two scans were taken of each subject; the first standing normally, and the second wearing a spine altering orthotic. Anatomical LL measurements were taken from the sagittal radiograph and the SterEOS reconstructions of both scans. Marker based LL was also taken from the sagittal radiograph. The first method computed the ratio of the anatomical based measurement to marker based measurement to give a correction factor (CF). The second method computed the difference between the anatomical measurement and the marker based measurement to give a correction constant (CC). The CF and CC were then applied to marker based measurements from the second scan to give a corrected LL. The correction constant resulted in the lowest average error in corrected LL. The average error for this method was 3.2°±2.8° (Mean ± SD) which corresponding to a percent error of 6.2% ± 4.8%. This was found to have less error than the error introduced by inconsistent marker placement between trials, although the use of it is limited to activities with minimal lumbar flexion. Next, nine healthy individuals with tight hamstrings (popliteal angle greater than 25°) were recruited for this study. A physical therapist placed 58 reflective markers by palpation on anatomical landmarks of the torso and lower extremities. Subjects performed walking, sit to stand, forward bend, and ball pick up from the floor, while ten optoelectronic cameras recorded the 3D location of the reflective markers. Subjects were then scanned with the spine markers using an EOS bi-planar X-ray system. Following testing, participants completed a six week stretching program designed to increase hamstring length. All baseline testing was then repeated. PT was defined as the angle between horizontal and the line extending from the posterior superior iliac spine marker to the anterior superior iliac spine marker. LL was defined as the acute angle between the lines connecting the markers on T12 to L2 and S2 to L4. The LL correction constant was applied to walking and standing activities to give corrected LL values. The change in PT and corrected LL was then found for each participant. All subjects had a decrease in popliteal angle (-18.8°±11.6°). It was found that hamstring lengthening increased PT 2.1°±2.6° at heal strike during walking (p= 0.04). The greatest increase of 10.6°±5.4° (p< 0.001) occurred at flexion during forward bend. Ball pick up also saw a similar increase at flexion and knee bend. PT at standing increased 4.4°±2.8° (p=0.001) and 2.0°±2.1° (p=0.05) during forward bend and standing respectively. While increased PT tended to correspond to increased LL, the increased LL was only statistically significant while standing during the forward bend activity (7.0°±7.2°). Hamstring stretching was effective for increasing anterior pelvic tilt during gait, forward bending tasks, and standing. The effect that increased PT has on LL is still a topic for further investigation. Future studies should include increased sample size and improvement of the LL correction method.




Included in

Therapeutics Commons