Education and Counseling to Prevent Low Back Pain
Efficacy of Risk Reduction continued...
Thus, most of the trials showed a statistically significant benefit from exercise, but the effects were modest and of uncertain duration. The interventions used may not be relevant to the clinical setting since all allocated work time for exercise, thus greatly increasing the likelihood of compliance. In most of the trials, the authors do not specify whether the control groups continued to work while the intervention groups were allowed time during the workday for exercise.36-39,42 If so, the results may have been biased by the greater exposure time to work conditions associated with back injury in controls compared to the intervention subjects. A common methodologic problem is the lack of precision in specifying whether the goal was to prevent the first episode of low back symptoms, activity limitations, recurrent episodes, work absenteeism, or chronic disability.43 Finally, the inclusion criteria and clinical outcomes of the studies differ and therefore are difficult to compare. Thus, the evidence regarding the effectiveness of exercise in preventing low back pain is inconclusive.
Orthotic devices such as back supports (corsets or back belts) are frequently used to prevent back pain and injuries.44 These devices may be prescribed by physicians, but individuals generally obtain them on their own or from employers. Several studies have attempted to evaluate back belts in the occupational setting. In a controlled trial, 642 airline baggage handlers were randomized to one of four groups: back belt only, training only, back belt and training combined, and control.45 Previously injured workers were included, but the proportion of workers with prior back injury in each group was not reported. Back belts were to be worn only at work. The 1-hour training session included information on back mechanics, proper lifting techniques, and warm-up exercises. At 8-month followup no statistically significant differences in mean rates of work loss were observed between subjects in either the belt only group or the belt and training groups and controls. When all subjects randomized to use a back belt were aggregated and compared to controls, the intervention group showed a trend toward increased frequency of back pain. Results are difficult to interpret, however, because analysis was performed according to intervention received rather than intention to treat, and 58% of subjects assigned to one of the back belt groups who had not yet developed back pain stopped using the supports before the end of the study. In a trial of 90 warehouse workers,22 one third were randomized to a 1-hour educational intervention emphasizing back biomechanics and were provided with a corset for use during work hours; one third received education only; and one third received no intervention. There were no differences in productivity or injury rates, but subjects in the corset plus education group had a significantly greater decrease in days lost from work compared with controls (2.5-day decrease vs. 0.4-day increase). These differences in outcomes occurred only in the subgroup of previously injured workers, suggesting that prophylactic bracing may only benefit those with a history of back injury. A retrospective cohort study assessed 1,316 workers at an Air Force base where policy mandated use of the belt for all employees with a history of back injury whose position required frequent heavy lifting. Those who wore belts were the intervention group while employees in comparable positions who chose not to wear belts were controls, suggesting likely selection bias. The risk of low back injury was reduced 40% among those using back belts, but this difference was not statistically significant. Costs of injury sustained while wearing a belt, however, were substantially higher than if injured without a belt.12 Thus evidence is inadequate to show a benefit from back belts, and suggests possible harms. In addition, poor compliance in these and other studies46 raises the question of whether subjects will routinely use corsets for prevention of back pain.
Epidemiologic evidence suggests that several modifiable risk factors, including smoking, obesity, and certain psychological profiles, predispose subjects to develop low back pain.2,4,6,7,47,48 Risk factors are presumed to exert their influence either by increasing a subject's risk of a precipitating event, or by increasing the chance that such an event will be perceived as painful or disabling. Cross-sectional and prospective studies have consistently shown that smokers have a 1.5-2.5-fold increased risk of back pain compared to nonsmokers.21,23,24,30,49-57 A biologic basis for this risk is suggested by a recent study of identical twins discordant for smoking, showing that smoking increases degenerative changes of the spine.58 Prospective and cross-sectional studies have also associated obesity with back pain, although one study did not support this association.21,23,24,59 The association may be stronger in women.23 Based on these associations some authors have recommended smoking cessation and, for obese persons, weight loss to prevent back pain,6,21 but direct evidence to support these recommendations has not been identified. Psychological risk factors, including depression, anxiety, and perceived high occupational stress, have also been associated with the development of low back pain.23,24,54,60-63 Again, there is no direct evidence that modifying these factors reduces low back pain.