AIM: To systematically collate, appraise and summarise systematic reviews that assess the effects of exercise on non-specific low back pain (NSLBP). METHOD: A systematic review of electronic databases was carried out from January 1965 to April 2008. Hand searching and screening of reference lists was also undertaken. Two reviewers worked independently throughout all stages of the review and agreement was achieved by consensus. Strict inclusion/exclusion criteria were applied based on standard definitions of key terms. Data from included studies were abstracted onto a pre-tested standardised form. Quality was assessed using the Scottish Intercollegiate Guidelines Network (SIGN) checklist for systematic reviews. RESULTS: Four reviews met the inclusion criteria. Three of these achieved a low bias rating indicating a moderate/good quality review in which some/most of the SIGN criteria were fulfilled. These studies provide strong prima facie evidence that exercise programmes reduce sick leave and improve pain and disability in people with non-acute NSLBP. The clinical utility of this evidence is, however, compromised by diversity in the types of exercise interventions, cointerventions and comparison groups and the current lack of consensus on NSLBP sub-group classification and core outcomes. DISCUSSION: Results reflect developments in the methodology of systematic reviews over the last decade. Synthesis of the findings of reviews may be facilitated by back pain research which focuses on issues relevant to contemporary theories of pain, such as strategic approaches to exercise, rather than further structural-pathological sub-classification of NSLBP.

Full text (sometimes free) may be available at these link(s):      help

• DOI
• PDF locator
• publisher