ABSTRACT: OBJECTIVES:Investigate the relationships between the ability/inability to perform five physical test exercises and the presence or absence of low back pain (LBP). SETTING:Regional Australian council training facility. PARTICIPANTS:Consecutive participants recruited during 39 back education classes (8-26 participants per class) for workers in general office/administration, parks/gardens maintenance, roads maintenance, library, child care and management. Total sample (n=539) was reduced through non-consent and insufficient demographic data to n=422. Age 38.6±15.3 years, range 18-64 years, 67.1% male. METHODS:Cross-sectional, exploratory, observational investigation. LBP presence was ascertained from a three-response option questionnaire: 0=none/rarely (no) 1=sometimes (some), 2=mostly/always (most). Statistical correlation was performed with the number of the five test exercises the individual successfully performed: (1) extension in lying: 3?s; (2) 'toilet squat'; feet flat, feet touched: 3?s; (3) full squat then stand up: 5 times; (4) supine sit-up, knees flexed: 10 times; and (5) leg extension, supine bilateral: 10 times. INTERVENTIONS:Nil. RESULTS:For the group 'no-some', 94.3% completed 4-5 test exercises, while for group 'With', 95.7% completed 0-1 test exercises. The relationship between LBP presence and number of exercises performed was highly significant (?2(10)=300.61, p<0.001). Furthermore, multinomial logistic regression predicting LBP (0=no, 1=some, 2=most) from the number of exercises completed, substantially improved the model fit (initial-2LL=348.246, final-2LL=73.620, ?2(2)=274.626, p<0.001). As the number of exercises performed increased, the odds of reporting 'some LBP' or 'most LBP' dropped substantially (ORs of 0.34 and 0.17, respectively). CONCLUSION:The ability to complete/not complete five test exercises correlated statistically and significantly with a higher LBP absence/presence in a general working population. Training individuals to complete such exercises could facilitate reductions in LBP incidence; however, causality cannot be inferred. Randomised trials are recommended to establish the potential efficacy of exercise-based approaches, considering these five selected exercises, for predicting and managing LBP.
