Non-specific low back pain is defined as a pain with no specific vertebral-related cause, such as infectious disease, tumor, osteoporosis, fracture, structural abnormality, inflammatory disease, radicular compressive syndrome or cauda equine syndrome. Non-specific low back pain is a common disease in many countries. This musculoskeletal disorder is costly to public health systems. Therefore, the use of manual therapies is important in the treatment of this disease and studies show the effectiveness of this type of therapy.

Spinal manipulation is applied in manual therapies such as Osteopathy, Chiropractic and Physical Therapy and is widely used for acute and chronic non-specific low-back pain. There is moderate evidence that spinal manipulation is superior to sham spinal manipulation for improving short-term pain and function in chronic and acute non-specific low back pain. However the therapeutic mechanisms involved in this procedure are not well understood. Furthermore, while the high velocity and low amplitude spinal therapy has been shown to be effective in reducing pain and improving functional capacity in subjects with non-specific low back pain, the effect on postural variables have not been investigated. Therefore, the aim of this study is to evaluate the acute effect of lumbar manipulation on pain and postural variables.

Twenty-four individuals with non-specific low back pain will be randomly allocated to two groups. The intervention group will receive high velocity and low amplitude spinal therapy, while the control group will receive sham manipulation. Immediately before and after the respective manipulation protocol, both groups will be evaluated regarding pain level, using a visual-analogue scale and algometer, and postural variables, using center of pressure displacement measured with the aid of a force plate. While the patient and therapist manipulator will be aware of the protocol applied in each case, the evaluator will be blind. A statistical treatment will be used to compare the results.

Mean and standard deviation are calculated for each variable of postural control, as well as the values obtained in the measurement of pain using the numeric pain rating scale and algometry data.

t-test or Mann-Whitney test will be used to compare data before and after the manipulation protocol.

Multiple One-Way ANOVAs will be used to compare the manipulation protocol, for each dependent variable: effects of pain and center of pressure displacement.. Statistical analysis will be conducted with a confidence interval of 95%, α value of 5%, p value of ≤ 0.05.

Inclusion Criteria:

• recurrent low back pain in the last three months;
• medical diagnosis of non-specific low back pain, daily or almost daily;
• subjects that will present at least four variables which encompass the clinical prediction rule: symptoms duration less than 16 days, Fear Avoidance-Beliefs Questionnaire score lower than 19 in the work subscale, segmental lumbar hypermobility test in prone decubitus, internal hip rotation higher than 35 degrees and symptoms that don't pass the knee.

Exclusion Criteria:

• subjects presenting symptoms that pass the knee;
• impairment of neurologic signs like sensibility, muscle force and deep tendon reflex;
• a prior history of spondylolisthesis, spinal stenosis, inflammatory disease, cancer, musculoskeletal degenerative disease, pregnancy, disease and drugs that impairment the balance;
• subjects who received a HVLA manipulation of the lumbar spine at least three months earlier;
• will be excluded women older than 50 years, that don't will present a bone mineral density exam.