The purpose of this project was to apply established principles of feedback training to occupational lifting, with the aim of reducing lumbar spine compression during symmetrical lifting tasks. The project consisted of several phases:​ development of a modified linked segment model suitable for estimation of the dependent variable (lumbosacral compression);​ a pilot project to assess the feedback protocols together with the various modes of feedback available;​ and a larger main study to measure the effectiveness of the feedback training statistically.

The model development was a minor portion of this project, by virtue of starting with an existing model that had already been validated. Some modifications to the lumbosacral portion of the model were made to bring it up to date with current research, and then the data acquisition system was validated by comparing ankle forces computed with the linked segment analysis to those measured directly from a force plate.

Next, a pilot study was performed to assess the feasibility of using feedback training to reduce the compression force developed at the lumbosacral junction (L5/S1) during occupational lifting. Volunteer subjects (n = 17) were divided into 4 feedback treatment groups and one control group. Feedback groups received one of the following types of information:​ (1) real-time EMG feedback;​ (2) a computed acceleration index representing the difference between static and dynamic computations of the lumbosacral compression force;​ (3) the dynamic compression force developed at L5/S1;​ or (4) the horizontal distance between the load and L5/S1. Kinematic data were collected during symmetric sagittal plane lifts from pallet height to standing knuckle height. Following a baseline session, two feedback training sessions were given to the treatment subjects, with each session separated by a one-week period. Control subjects were given training sessions without the benefit of feedback, but otherwise the protocols were identical. A post-training session one week after the final feedback session was used to assess pre-topost changes in the dependent variable:​ peak lumbosacral compression. Of the four types of feedback training, three were comparable in terms of the mean effect (acceleration, horizontal distance, and compression force feedback);​ the EMG feedback was less effective.

The final phase of this project was the Main Study, using two feedback methods chosen from the pilot study, with additional subjects added to each group to achieve statistical power. The protocols for the Main Study were identical to those used in the pilot study. Volunteer subjects (n = 28) were divided into 2 feedback treatment groups and one control group. Two distinct types of feedback were compared:​ (1) real-time electromyographic feedback;​ and (2) an acceleration index delivered verbally post-lift. Kinematic data were collected during sagitally symmetrical lifts. All three groups showed reductions in peak compression from pre-to-post:​ on average the control group improved by 11.2%, the EMG group by 16.7%, and the acceleration group by 25.3%. A repeated measures ANOVA found a significant interaction between groups (p = 0.023), and a post-hoc Bonferroni test found a difference between the improvement in the control group and that in the acceleration group (p < 0.05). These reductions in lumbosacral compression persisted after a 7-day interval without feedback training, suggesting that this approach could provide sustained risk-reduction during manual materials handling.

Given the established relationship between manual materials handling and lumbar compression, training of this sort has the potential for reducing the incidence of job disability and work related low back pain.