
Low back pain (LBP) is a prevalent and disabling condition that affects individuals of all age groups globally [1, 2]. It is estimated that around 80% of the population will encounter LBP at some point in their lives, leading to significant personal and socioeconomic burdens [3, 4]. Despite the availability of diverse treatment options for LBP, such as medication, exercise, physical therapy, and manual therapy, the effectiveness of various approaches for different stages and subtypes of LBP remains a subject of ongoing debate [5].
For chronic LBP, clinical practice guidelines strongly advocate for the utilization of thrust manipulation – a manual therapy technique involving swift, low-amplitude thrusts administered to specific spinal segments [6-8]. This method is thought to enhance joint mobility, alleviate pain, and restore function. However, the degree of evidence supporting its efficacy for acute and subacute non-specific LBP is less conclusive and has sparked controversy [8-10].
Effectively addressing acute and subacute LBP requires identifying interventions that yield immediate or short-term relief, as these stages often entail substantial pain and functional limitations [11-13]. Given the contradictory evidence and absence of consensus regarding the effectiveness of thrust manipulation for non-specific acute and subacute LBP, there is a clear need for a systematic review and meta-analysis to rigorously assess the available literature and offer a comprehensive consolidation of the current knowledge.
This systematic review and meta-analysis aim to assess the immediate or short-term effects of thrust manipulation on pain in individuals experiencing acute and subacute LBP. Through a methodical search and analysis of pertinent studies, our objective is to provide a robust evaluation of the existing evidence and establish the overall effectiveness of thrust manipulation as an intervention for reducing pain in this specific population.
In this meta-analysis and systematic review, we endeavored to collate and critically appraise the existing research on the effects of thrust manipulation in treating acute to subacute LBP. Our methodology strictly conformed to the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) standards.
We conducted a comprehensive literature review by employing the PICOSD framework, which encompasses Participants [P], Intervention [I], Comparison [C], Outcomes [O], and Study design [SD], as the foundational approach for our database search.
The PICOSDs considered in this review encompassed: (P) individuals reporting instances of LBP within the past 3 months [14]; (I) investigations where lumbar spine thrust manipulation constituted the intervention in the experimental arm; (O) evaluation of pain intensity and disability outcomes; and (SD) inclusion of randomized controlled trials (RCTs) published within the last decade to reflect contemporary research trends.
Exclusion criteria included back pain lasting more than 3 months, studies with manual therapy for the control group, studies not written in English.
In June 2023, a thorough literature search was carried out, with two experienced researchers in meta-analysis methodologies independently collecting the data. The search strategy was developed by integrating terms related to P, I, and SD. Our search was anchored around the following key terms: (low back pain OR nonspecific low back pain NOT chronic low back pain) AND (manipulation OR thrust manipulation OR thrust OR joint manipulation OR joint thrust OR spinal manipulation OR spine manipulation OR spine thrust manipulation) AND (randomized controlled trial). This expansive search spanned several databases, namely MEDLINE, Embase, CINAHL, and PubMed, making use of pertinent index terms.
Upon gathering studies from four global electronic databases, redundant data were eliminated utilizing Reference Manager (EndNote 20, Thomson Reuters, NY, USA). Adhering to the predetermined selection criteria, the titles and abstracts were scrutinized by two researchers for inclusion. Subsequently, any divergences in selection were deliberated upon, and the reasons for exclusions were outlined. Ultimately, the chosen studies underwent categorization, and their attributes were extracted. All stages of database retrieval, including selection and data extraction, were independently executed by the two researchers.
Two researchers with experience in meta-analysis studies evaluated the quality of the selected studies for this review using a seven-item Cochrane Risk of Bias (RoB) tool created by the Cochrane Bias Method Group. The RoB was categorized as low (+), uncertain (?), or high (-). In cases of discrepancies during research selection and data extraction, the original text was revisited and re-evaluated to ensure consistency.
Data synthesis was carried out using ReviewManager (RevMan 5.4, The Cochrane Collaboration, Oxford, UK). In cases where there were identical variables suitable for analysis, or when three or more quantitative variables were available from both pre- and post-intervention tests, they were included in the meta-analysis. The effect size was calculated using the standardized mean difference (SMD) for consistent variables. A random effects model, which recalibrates weights, was employed for the analysis [15]. To ascertain the homogeneity of the chosen studies, the I2 statistic and chi-squared test were employed. The interpretation of I2 results is as follows: less than 40% indicates low heterogeneity, 50% to 75% suggests medium heterogeneity, and over 75% indicates high heterogeneity [16].
Potential publication bias among the analyzed studies was examined through a funnel plot. However, this analysis was omitted if the number of selected studies was fewer than 10 [17].
From an initial search across international databases, we identified 249 studies. Out of these, 53 were duplicates and were subsequently removed. A further review of titles and abstracts led to the exclusion of 178 studies. Upon detailed examination of the full text of the remaining 18 studies, based on the established eligibility criteria, 12 were further excluded. Consequently, this systematic review and meta-analysis incorporated both qualitative and quantitative analyses of the final 6 RCTs [18-23](Figure 1).
We utilized the RoB tool to gauge the methodological quality of the six studies in question.
For the criterion of random sequence generation, all six studies exhibited a positive rating, with none showing a negative or unclear stance. In terms of allocation concealment, five studies were positively rated, none were negatively rated, while one remained unclear. When it came to the blinding of participants and personnel, there was an even split: three studies received a positive rating and three received a negative rating, with none being unclear.
Regarding the blinding of outcome assessment, three studies were positively inclined, two were negative, and one was marked as unclear. In the domain of incomplete outcome data, the evaluation leaned more towards a negative sentiment, with four studies rated negatively and only two positively; none were left unclear. As for selective reporting, three studies demonstrated a positive score, none had a negative score, but three were categorized as unclear. Lastly, concerning other biases, two studies were positively rated, none were negative, and a larger portion, four in total, remained ambiguous. These findings are further illustrated in Figure 2.
In the review, we selected six studies that encompassed a total of 308 individuals with LBP. Each study exclusively assigned cases involving thrust manipulation to the experimental group. The evaluation focused on pain intensity and disability outcomes, and the specifics of these chosen studies can be found in Table 1.
Six RCTs, comprising 308 patients with acute and subacute LBP, were assessed for pain intensity. The group undergoing thrust manipulation demonstrated significant improvement compared to the control group. When evaluated using the SMD, the results were: SMD = -0.44 with a 95% confidence interval (CI) of -0.80 to -0.08. The heterogeneity was χ2 = 10.72, df = 5, and I2 = 53%. The overall effect, Z, was 2.40 with a significance level of p = 0.02 (Figure 3).
Three RCTs, comprising 212 patients with acute and subacute LBP, were assessed for disability index. The group that received thrust manipulation did not show any notable improvement compared to the control group. When evaluated using the SMD, the results were: SMD = -0.96 with a 95% CI of -2.67 to 0.76. The heterogeneity was χ2 = 57.66, df = 2, and I2 = 97%. The overall effect, Z, was 1.09 with a significance level of p = 0.27 (Figure 4).
The wide prevalence and detrimental impact of LBP on global populations necessitate robust clinical strategies and interventions. With thrust manipulation having garnered attention for its potential in managing LBP, especially in the chronic stages, our systematic review and meta-analysis aimed to understand its immediate or short-term implications on pain intensity and disability in acute and subacute LBP scenarios.
From the comprehensive analysis, it is evident that thrust manipulation has a significant positive effect on pain intensity (SMD= -0.44; 95% CI, -0.80 to -0.08). This result aligns with the general understanding that thrust manipulation facilitates improved joint mobility [24-26]. Joint stiffness, one of the commonly cited reasons for acute pain, can get promptly relieved by re-establishing normal joint motion [27]. The swift, controlled thrusts might play a role in breaking adhesions or restoring the joint's natural movement, which could be the reason behind the immediate reduction in pain intensity post-treatment [28, 29]. The neurophysiological mechanism of thrust manipulation is conclusively explained by spinal cord mechanisms. Pain due to peripheral sensitization is caused by temporal summation of C-fibers, and it has been reported that thrust manipulation can interrupt this summation [30-32].
However, the absence of a significant improvement in disability is an intriguing observation (SMD= -0.96; 95% CI, -2.67 to 0.76). While pain intensity and disability are related, they are distinct entities. It is plausible that even though individuals felt an immediate reduction in pain after receiving thrust manipulation, the underlying biomechanical or neuromuscular deficiencies contributing to functional disabilities might not have been addressed effectively [33].
Another perspective to consider is the psychological aspect of pain and disability. It's established in literature that pain cognition is multi-dimensional, encompassing not only the physical but also emotional, cognitive, and social factors [34, 35]. Thus, even if thrust manipulation managed to address the nociceptive component of pain effectively, it may not have dealt with other underlying issues, like fear of movement or re-injury, that contribute to disability [36-38].
Additionally, it's crucial to note that disability, especially in the context of LBP, is multifactorial. Factors like muscle imbalances, core stability deficiencies, or poor postural habits may not be instantly rectifiable through a singular session or a few sessions of thrust manipulation. Comprehensive rehabilitation approaches, combining manual therapy with tailored exercises and patient education, might yield better results in addressing disability. It's also worth noting that the definition and measurement of 'disability' can vary across studies. Some may use subjective questionnaires while others might employ objective functional tests. This variance might influence the collective understanding of thrust manipulation's effectiveness on disability outcomes.
In the clinical practice guidelines for non-invasive treatment of acute to subacute LBP reported by the Annals of Internal Medicine in 2017 [39], superficial heat, nonsteroidal anti-inflammatory drugs, and skeletal muscle relaxants were recommended based on moderate-quality evidence. Meanwhile, thrust manipulation was categorized as having low-quality evidence; therefore, our systematic review and meta-analysis aimed to elevate its clinical significance.
This study has several limitations worth noting. The analysis was based on six studies with 308 participants, potentially lacking in diversity and comprehensive representation of the broader LBP patient population. Variations in thrust manipulation techniques and measurement tools for 'disability' across studies could introduce inconsistencies in outcomes. Therés also a potential for publication bias, where studies with unfavorable results might be underrepresented. Additionally, the review's emphasis on short-term outcomes leaves the long-term efficacy of thrust manipulation unexplored. External factors, placebo effects, and inherent biases from original studies might also influence our findings. Consequently, while the study offers essential insights, caution is advised in its interpretation.
The clinical implications of our findings are profound. Thrust manipulation provides tangible relief from pain intensity in patients with acute and subacute LBP, making it a valuable intervention for immediate pain management. However, when it comes to disability outcomes, thrust manipulation alone may not be sufficient. Clinicians should consider integrating thrust manipulation with other therapeutic modalities and comprehensive rehabilitation approaches to address both pain and functional limitations effectively. This holistic strategy can ensure not only a reduction in pain but also an enhancement in the overall functional capacity of LBP patients. Future research should delve deeper into combining manual therapy with other treatments to optimize patient outcomes.
The Research has been conducted by the Research Grant of Gwangju Health University in 2023 (2023006).
The authors declare no potential conflicts of interest with respect to the authorship and/or publication of this article.
Characteristics of included studies
Study | Sample size | Duration | Therapeutic intensity | Outcomes |
---|---|---|---|---|
Bialosky, et al. 2014 [18] | EG = 28 CG = 27 |
2weeks |
Lumbar spine thrust manipulation: The researcher passively side bent the patient toward the side to be manipulated and asked the subject to interlock hands behind his/her head. The researcher then passively rotated the subject away from the side to be manipulated and delivered a posterior and inferior thrust to the opposite ASIS. Sham: placebo spinal manipulative therapy |
Pain intensity: PPT |
Fagundes Loss, et al. 2020 [19] | EG = 12 CG = 12 |
Immediately |
Lumbar spine thrust manipulation: HVLA lumbar manipulation was performed by positioning hypomobile vertebrae during thrusting. Sham: The position was held for approximately 20 seconds without receiving HVLA thrust. |
Pain intensity: NPRS |
Schneider, et al. 2015 [20] | EG = 35 CG = 34 |
4weeks |
Lumbar spine thrust manipulation: Participants were given high-velocity low-amplitude thrust manipulation in the side posture position by a licensed chiropractor. Segmental levels where manipulation was applied were determined using standard chiropractic methods of static and motion palpation. Sham: These participants were told that most new episodes of back pain are typically self-limiting, were prescribed over-the-counter analgesic and NSAID medications, given advice to stay physically active and avoid prolonged bed-rest. |
Pain intensity: NPRS Disability: ODI |
Selhorst and Selhorst, et al. 2015 [21] | EG = 18 CG = 16 |
1weeks |
Lumbar spine thrust manipulation: The patient was passively side-bent away from the therapist. The therapist passively rotated the thoracic spine and then delivered a quick posterior and inferior thrust through the anterior superior iliac spine. The manipulation was performed on the side, which the patient reported to be more symptomatic. Sham: The manual therapist performed the sham lumbar manipulation technique with the patient sidelying. |
Pain intensity: NPRS Disability: PSFS |
Vining, et al. 2020 [22] | EG = 55 CG = 54 |
4weeks |
Lumbar spine thrust manipulation: Spinal manipulation consisted of high-velocity thrust-type manipulation directed toward the thoracolumbar or pelvic regions. Spinal manipulation involving other spinal regions or extremities was also allowed when clinically indicated. Wait-list: Wait-list group participants were free to seek any health care, except chiropractic or spinal manipulation from any provider during the trial timeframe. |
Pain intensity: NPRS Disability: RMDQ |
Younes, et al. 2017 [23] | EG = 10 CG = 7 |
Immediately |
Lumbar spine thrust manipulation: It consisted of 45 minutes of spinal manipulation (various types of HVLA spinal manipulation or passive mobilization) and muscle manipulation. Sham: The Sham intervention simulated these techniques, but with improper patient positioning, deliberately misdirected movements, and diminished treatment provider force. |
Pain intensity: NPRS |
ASIS: anterior superior iliac spine, CG: control group, EG: experimental group, HVLA: high-velocity low-amplitude, NPRS: numeric pain rating scale, NSAID: nonsteroidal anti-inflammatory drug, ODI: Oswestry disability index , PPT: pressure pain threshold, PSFS: patient-specific functional scale, RMDQ: Roland Morris disability questionnaire.