Table 1 Descriptive characteristics of the included studies

Edgecombe et al. [19] (2015)

• A self-controlled cross-over design.

• 8 anesthetized felines.

• A VRFD device operating was used to deliver a PA SMT at four anatomic sites: SP (L6 and L7), lamina (right L6), and mammillary process (right L6).

• Control intervention: contact load alone.

• The VRFD device was used to measure PA spinal stiffness at L6 SP immediately before and after each SMT application site. Stiffness at the maximal indentation load (TIS) and slope of the linear regression line best fitting the data (k) were calculated.

• ↑ in k following SMT at L6 SP compared to the control intervention.

• ↓ in TIS following SMT at L6 SP and the lamina compared to the control intervention.

Wong et al. [10] (2015)

• Nonrandomized controlled study.

• 48 participants with LBP.

• 59 participants with no current LBP (no intervention).

• 32 participants with LBP: A SMT consisting of the application of a PI thrust to the patient’s pelvis (both sides) in a supine position. Two treatments and one follow-up session.

• 16 participants with LBP and the participants without LBP received no intervention.

• Spinal stiffness was measured by a mechanical indentation device at L3 SP with the participant in the prone position before and after the intervention and at follow-up.

• Lumbar multifidus thickness ratio (LMTR) and intervertebral disc diffusion (IVDD) were also measured.

• ↓ spinal stiffness after each SMT in the SMT responders only.

• ↑ in LMTR and IVDD after the first SMT in the SMT responders only.

• ↓ in spinal stiffness was significantly associated with ↑ in LMTR and IVDD.

• ↑ in LMTR was significantly related to the ↑ in IVDD.

Shum et al. [20] (2013)

• An experimental between-group study.

• 20 participants without LBP.

• 19 participants with LBP.

• A PA MOB at the L4 involving 3 cycles of large amplitude of oscillatory PA forces (grade III). One session.

• The bending stiffness of the lumbar spine was measured by a single application of a PA load (250 N) at L4 SP by a physiotherapist.

• Pain intensity, ROM and tolerance to load were also assessed.

• ↑ in curvature change, in tolerance in load application before onset of pain, and in active flexion and extension ROM following intervention in LBP.

• ↓ spinal stiffness and pain intensity following intervention in LBP.

Fritz et al. [9] (2011)

• Prospective case series.

• 48 participants with LBP.

• A physical therapist or chiropractor provided a SMT using a PI thrust at the subject’s pelvis (right and left side) in supine position. Two treatments and one follow-up session.

• A mechanical indentation device was used to measure spinal stiffness (global [GS] and terminal [TS] coefficient) at L3 SP with the participant prone. Spinal stiffness was assessed before, after the intervention and at follow-up.

• LM recruitment, ODI and pain intensity were also evaluated.

• ↓ in GS and TS following SMT at session 1, and in TS following SMT at session 2.

• Less initial TS and greater immediate ↓ in GS were associated with greater ODI improvement.

• Less initial TS was associated to greater ↑ in LM recruitment.

• Being a likely SMT responder correlated with greater immediate ↑ in LM recruitment.

Stamos-Papastamos et al. [21] (2011)

• Same-subject, repeated-measures, crossover study

• 32 students without back pain.

• SMT: A rotational manipulative thrust on both sides of the L4/5 segmental level.

• MOB: A 3 sets of 1 min of grade IV+ PA MOB.

• Each intervention was delivered by a therapist.

• Bending stiffness was calculated by quantifying the PA force applied to L4 during 5 central PA pressures using a force platform with the participant lying prone.

• Lumbar flexion and extension ROM were also measured.

• No effect of SMT and MOB on bending stiffness or ROM.

• No significant correlation between Δ in bending stiffness and ROM.

Haussler et al. [22] (2010)

• Randomized clinical trial

• 24 actively ridden horses with no history of acute back problems or lameness.

• 12 horses: A single application of manually applied, HVLA, PA thrusts were applied bilaterally at T14/15, T17/18, L1/2, L3/4 and L6/S1. Once a week for 3 weeks.

• 12 horses: no intervention (control group)

• Spinal stiffness was measured by the manual application of a cyclic load perpendicular to the dorsal midline over the intervertebral site of interest.

• Greater ↑ in displacement and force amplitudes during spinal stiffness assessment following SMT compared to control.

• A trend of an ↑ in spinal stiffness following SMT compared to a ↓ with the control group.

Ferreira et al. [23] (2009)

• A randomised clinical trial.

• 191 participants with non-specific LBP.

• 71 participants received SMT: Joint MOB or manipulation techniques (at the discretion of the physiotherapist) were applied to the subject’s spine or pelvis.

• 60 participants received general exercises.

• 60 participants received motor control exercises.

• Subjects attended 12 sessions over 8 weeks.

• Spinal stiffness was manually measured before and after 8 weeks of treatment by two physiotherapists using a stiffness reference device to anchor judgements of stiffness on an 11-point scale.

• Pain, disability, GPE were also measured.

• ↓ mean stiffness following treatment in all groups.

• No significant between group differences in Δ stiffness.

• Weak correlation between Δ stiffness and Δ GPE.

• Weak correlation between Δ stiffness and Δ function for the SMT group only.

Haussler et al. [24] (2007)

• A randomized crossover study

• 10 healthy adult horses.

• One HVLA PA thrusts applied to T14/15, L5/6 SPs, and left and right L1/2 articular processes using a reinforced hypothenar contact.

• Control intervention: no SMT.

• Spinal stiffness was measured by manually applying a vertical force perpendicular to the dorsal midline over the intervertebral site of interest.

• ↑ in displacement and force amplitude following SMT but not following control intervention.

• No change in stiffness following any intervention.

Allison et al. [25] (2001)

• A within-subjects, repeated-measures design

• 24 participants without LBP over the past 6 months.

• A standardized PA MOB (average load = 146 N, frequency = 1.5 Hz) was applied by one physiotherapist to the L3 SP for a period of two minutes.

• The SPAM apparatus was used to measure PA stiffness at L1, L3 and L5 SPs with the participant lying prone immediately before and after the intervention.

• No change in stiffness.

Goodsell et al. [26] (2000)

• A self-controlled cross-over design.

• 26 participants with LBP.

• A PA MOB manually applied (load at the discretion of the therapist) to the SP of the most symptomatic spinal level.

• Control intervention: patient lying prone for 3 min.

• Spinal stiffness was measured using a device applying a PA force to the SP of the vertebra receiving the intervention with the participant lying prone before and after the intervention.

• Pain and ROM were also measured.

• No difference between control and MOB on the change in spinal stiffness, overall pain and ROM.

• ↓ pain on the worst movement following MOB when compared with control.