The methodology used in the recruitment, assessment and treatment procedures in this study followed the CONSORT statement guidelines for the conduct of randomized control trials
A pre and post-test randomized control pilot study among patients with NSCLBP. The pilot is being carried out in preparation for a large randomized controlled trial in which treadmill walk and abdominal bracing core muscle activation will be the main intervention.
Recruitment and sample size estimation
Non-specific low back pain aged between 18 and 65 years referred from the orthopaedics and spinal surgeons of the National Orthopaedic Hospital Dala who willingly give their consent were recruited to participate in the study.
A sample size of 33 participants was involved in this study, this sample size was based on the pilot study sample size estimation rule of thumb for RCT involving continuous variables; n = 2 (Za + Z1 − β)2σ2/Δ2, Where n = sample size, a two-sided significance of P < 0.05 with 0.80 power . Zα = 1.96 (two-tailed) and Z1-β = 0.84, studies that adopted similar sample size estimation, have been reported [24,25,26,27,28,29]. Thus, the sample size for the present study was calculated to be 2 (0.84) (0.36)/0.64, = 8.82. An attrition rate of 20% was added, thus, a total of 33 were consecutively recruited.
The main inclusion criteria were patients with NSLBP of mechanical origin with no radicular symptoms, and aged 18 to 65 years old. The patients were eligible irrespective of their educational status, in as much as they could read in either Hausa or English languages and consented to participate in the study.
Participants with a history of previous surgery to the spine and/or lower extremity and musculoskeletal conditions associated with pain and reduced mobility affecting the spine and /or extremities were excluded. Participants with medical conditions that might affect the use and correct performance of the treadmill protocol exercises were also excluded.
Ethical clearance was sought and obtained from the Research and Ethical Committee of National Orthopedic Hospital, Dala, Kano State (Ref: NOHD/RET/ETHIC/60), before the commencement of the study. Informed consent was sought and obtained before participants started the intervention. Participants were given enough time to consult with their doctors and make informed decisions of their suitability to participate in the trial.
Screening for inclusion criteria
Before randomization, prospective participants were screened for inclusion criteria and for safe use of treadmill by a research assistant, For safety to use of treadmill exercises Participants were screened for any cardiovascular abnormality such as elevated blood pressure and all acute musculoskeletal injuries or symptoms of physical limitations and cardiovascular complications that might affect the safety of carrying out treadmill based exercises using self-administered physical activity readiness questionnaire ( PAR-Q) as described in the previous study [30, 31].
The individuals that met the eligibility criteria were randomized by an assistant who did not have the knowledge of treatment procedures in groups A, B, and C. A simple randomization procedure was followed by the use of a computer-generated random number to determine the allocation sequence before the allocation of participants to the study group as described by . The participants in group A carried out abdominal bracing while walking on the treadmill together with usual care, while the Participants in Group B carried out treadmill walk with usual care but without abdominal bracing and participants in Group C will only have usual physiotherapy care.
Assessments and outcome measures
Social-demographic demographic variables of the participants such as age, gender history of the recent injury to the spine and or extremities using the researcher-developed data collection form. The weight and height of the participants were measured using a bathroom weighing scale ( Secca Germany) and a height meter (A Charder HM200P Portstad Portable Stadiometer (Charder Medical, Taiwan, ROC, 2007) to the nearest 0.1 kg and 0.1 m, respectively.
Oswestry Disability Index scores of functional limitation were the primary outcome measure in this pilot study while other outcome measures will be secondary. Thus, changes in ODI after the intervention will provide data that will be used for sample size estimation in large RCT. Assessment of outcome measures will be carried out pre and post 6-week exercise intervention.
Pain intensity was measured with box numerical pain rating scale (BNPRS): The Box-21 is identical to the Box-11, but presents 21 boxes in a row, with numbers labelled from 0 to 100 in increments of 5 . A scale with 21 levels has been shown to provide a sufficient level of discrimination for chronic pain patients to describe their pain based on both psychometric properties and preference, it was thus proposed by  that the Box-21 is the instrument of the first choice in a mixed or heterogeneous population of patients such as low back pain. Reliability has been demonstrated even in some other musculoskeletal conditions (r = 0.96 and 0.95, respectively) with a validity of r= 0.86 to 0.95 when compared to VAS .
Functional Disability was measured with the modified Oswestry Disability Index (ODI), The modified Oswestry Disability Index which is also known as the Oswestry Low Back Pain Disability Questionnaire is an important tool that researchers and disability evaluators use to measure a patient’s permanent functional disability. ODI is the most commonly used and cited tool for this purpose, followed by the Roland Morris disability questionnaire (RMDQ) This test is considered the gold standard of low back functional outcome tools. This will be used to assess the level of disability among the participants. It has an internal consistency of 0.75 and test-retest reliability of 0.91 as reported by 
Before walking on the treadmill, participants were shown the location of the core stability muscles as described by . In crook lying position, they were trained on how to carry out abdominal bracing manoeuvres to activate core muscles as described by . Only those who can brace abdominal muscles effectively in crook lying, standing and walking positions were allowed to progress to treadmill walk combined with abdominal bracing exercises. Abdominal bracing for the activation of the core muscle was confirmed with palpation of the transversus abdominis muscles. Effective abdominal bracing for activation of core muscles in this study is described as the ability to achieve at least 10mmHg deflection on the scale of pressure biofeedback unit (PBU) in crook lying position while it was inflated to 70 mmHg initial pressure, the participants must be able to sustain abdominal bracing for 10 s in 10 successive attempts [37, 38]. Participants were asked to maintain abdominal bracing while walking on the treadmill by very gently contracting or tensing the abdominal wall, especially in the area below the belly button, keeping the inward curve in their lower back (avoid rounding their lower back), breathing normally throughout and not to hold their shoulders in elevated positions [39, 40].
Participants in this group carried out treadmill exercises using the modified Bruce protocol as previously reported in previous chronic low back pain intervention studies [21, 41,42,43,44,45]. The detail of the protocol is as listed below;
Warm-up: Speed: 3.3 miles per hour (mph), 0% inclination and duration of 5 min
Speed: 3.3 miles per hour (mph), 0% inclination and duration of 5 min
Weeks 1: Speed: 3.3–5.0 mph (increase 0.5mph/ minute for 10 min), at 1 % inclination
Week 2: Speed: 3.3–5.0 mph (increase 0.5mph/ minute for 15 min), at 2 % inclination and total duration of 20 min
Week 3–6: Speed: 3.3–5.0 mph (increase 0.5mph/ minute for 20 min), 3 % inclination with subsequent 1% increase till 6 weeks
The patients received Infra-red radiation for 15 min before performing lumbar conventional physiotherapy exercises such as Flexibility exercise which involves single and double knee to the chest for stretching and flexibility of the back extensor and strengthening of the rectus-oblique muscles. In prone lying with pillow support with one leg sliding for strengthening the back extensor muscles, cycling in supine for strengthening the abdominal muscles and coordinating anterior and posterior lumbar muscles. Bridging exercises was also done for strengthening back extensor muscles as described by .
GROUP A: treadmill walking combined with abdominal bracing and usual care (TWAB)
Participants in this group were taught how to carry out abdominal bracing before they were asked to combine it with walking on the treadmill as described above.
Group B: treadmill walk without abdominal bracing and usual physiotherapy care (TWW)
Participants in this group performed treadmill walk using the modified Bruce protocol as described by  with usual care as described above, twice a week for 6 weeks.
Group C: usual physiotherapy care group (UC)
Some of the best evidence in conventional or usual care physiotherapy as described by some scholars which were based on meta-analysis and systematic reviews as described above was applied to participants in this group [25, 47, 48].
The physiotherapist evaluating the outcome measures was blinded to the participants' assigned intervention groups. The therapists providing the intervention were blinded to pre-treatment and post-treatment assessment scores before the analysis of the results.
The data analysis procedure was carried out using the SPSS 20th version software ((IBM Co., Armonk, NY, USA). Shapiro-Wilk test showed that all the data generated from the independent variables were normally distributed (P>0.05). Means and standard deviations were used to summarize the data such as age, height, weight, BMI and duration of back pain.
One-way analysis of variance (ANOVA) was used to analyse the difference between the scores obtained BPNRS, ODI, and FABQ pre and post-intervention. Paired t test was used to analyse pre-intervention and post-intervention differences between the independent variables within the individual groups.
Post hoc analysis was carried out to ascertain the clinical significance of the statistical difference observed between the independent variables. The level of significance was set at 0.05.