Article Text
PDF
Abstract
Objective Acupuncture is emerging as a potential therapy for relieving pain, but the effectiveness of acupuncture for relieving low back and/or pelvic pain (LBPP) during the pregnancy remains controversial. This meta-analysis aims to investigate the effects of acupuncture on pain, functional status and quality of life for women with LBPP pain during the pregnancy.
Design Systematic review and meta-analysis.
Data sources The PubMed, EMBASE databases, Web of Science and Cochrane Library were searched for relevant randomised controlled trials (RCTs) from inception to 15 January 2022.
Eligibility criteria for selecting studies RCTs evaluating the effects of acupuncture on LBPP during the pregnancy were included.
Data extraction and synthesis The data extraction and study quality assessment were independently performed by three reviewers. The mean differences (MDs) with 95% CIs for pooled data were calculated. We assessed the confidence in the evidence using the Grading of Recommendations Assessment, Development and Evaluation framework.
Main outcomes and measures The primary outcomes were pain, functional status and quality of life. The secondary outcomes were overall effects (a questionnaire at a post-treatment visit within a week after the last treatment to determine the number of people who received good or excellent help), analgesic consumption, Apgar scores >7 at 5 min, adverse events, gestational age at birth, induction of labour and mode of birth.
Results This meta-analysis included 10 studies, reporting on a total of 1040 women. Overall, acupuncture significantly relieved pain during pregnancy (MD=1.70, 95% CI: (0.95 to 2.45), p<0.00001, I2=90%) and improved functional status (MD=12.44, 95% CI: (3.32 to 21.55), p=0.007, I2=94%) and quality of life (MD=−8.89, 95% CI: (−11.90 to –5.88), p<0.00001, I2 = 57%). There was a significant difference for overall effects (OR=0.13, 95% CI: (0.07 to 0.23), p<0.00001, I2 = 7%). However, there was no significant difference for analgesic consumption during the study period (OR=2.49, 95% CI: (0.08 to 80.25), p=0.61, I2=61%) and Apgar scores of newborns (OR=1.02, 95% CI: (0.37 to 2.83), p=0.97, I2 = 0%). Preterm birth from acupuncture during he study period was reported in two studies. Although preterm contractions were reported in two studies, all infants were in good health at birth. In terms of gestational age at birth, induction of labour and mode of birth, only one study reported the gestational age at birth (mean gestation 40 weeks). Thus, prospective randomised clinical studies or clinical follow-up studies were hence desirable to further evaluate these outcomes.
Conclusions Acupuncture significantly improved pain, functional status and quality of life in women with LBPP during the pregnancy. Additionally, acupuncture had no observable severe adverse influences on the newborns. More large-scale and well-designed RCTs are still needed to further confirm these results.
PROSPERO registration number CRD42021241771.
- rehabilitation medicine
- pain management
- complementary medicine
Data availability statement
Data are available on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Strengths and limitations of this study
The studies included in this systematic review and meta-analysis are designed to be randomised controlled trials.
Rather than solely focusing on the effects of acupuncture for chronic pain for pregnant women with low back and/or pelvic pain, this meta-analysis also assessed the safety of acupuncture for pregnancy outcomes and the newborns.
The number of studies included was limited and only two studies reported on the outcome of preterm contractions.
Although we explained statistical heterogeneity through sensitivity analyses to some extent, some factors related to heterogeneity remained uncertain.
Introduction
Low back and/or pelvic pain (LBPP) are common during the pregnancy, which consist of generalised lower back pain and dull pains located in the posterior pelvic area.1 The prevalence estimates suggest that range from 24% to 90% of all pregnant women experience LBPP in the world.2 Additionally, LBPP are reported to affect daily functioning, and especially activities that involve weight bearing, and the endurance capacity for standing, walking and sitting is diminished, which affects the quality of life and work.3 Moreover, the costs associated with sick leave are enormous, causing a significant financial burden on women and society.4 LBPP are clinically characterised as pain located between the 12th rib and the gluteal fold, and distal and posterior pelvic region lateral to the lumbar-sacral junction.5 The two types of pain have not been able to be distinguished. Pathologically, pregnant women have increased relaxin produced by both the corpus luteum and the uterine decidua and increased motion in the pelvic joints, which might be the cause of LBPP during pregnancy.6 7 Currently, physical therapy,8 massage therapy9 and special pillows10 are mainly used for LBPP during pregnancy. However, these therapies are not recommended according to the evaluation of results of clinical trials due to their unclear effectiveness.5 Therefore, the investigation of safe and effective therapies for the treatment of LBPP during the pregnancy seems warranted.
Acupuncture originated in China, using thin and solid metallic needles to insert into acupoints along the meridians. The meridian theory was proposed based on ancient physiological system (not western scientific empiricism). This theory suggested that the body’s vital energy (qi) flows along meridians, which are associated with the internal organs. It is believed that internal disharmony of organs is reflected at specific points, known as acupoints.11 12 This intervention as an acceptable complementary therapy is recommended and commonly practised in many countries for the management of dental pain, headache, myofascial pain and pain of many origins.13 Although the analgesic mechanisms of acupuncture remain uncertain, several studies14 15 have shown that it was associated with neurohumoral mechanism—spinal cerebrospinal fluid endorphins and dynorphins increased after acupuncture. Recently, several studies have shown the clinical efficacy of acupuncture for reducing low back pain,16 17 but the evidence is equivocal. Additionally, there are serious concerns about drug use for pain relief in the pregnancy, and the side effects of these treatments. Acupuncture is considered as a safe, non-pharmacological treatment option that is being increasingly used for relieving discomfort during the pregnancy.1 18 Nevertheless, several studies have shown that there was an insufficient clinical evidence for acupuncture for LBPP during the pregnancy, and the results of previous studies were also inconsistent.19–21 Additionally, in view of concerns about using ‘‘forbidden points’’ during the pregnancy and the risk of preterm contractions, one systematic narrative review examined randomised controlled trials (RCTs) of acupuncture using forbidden points prior to 37 weeks of pregnancy to treat pain conditions, with the primary outcome of preterm contractions. It was recommending that due to the concern of high dropout rates from studies that use forbidden points, alternative points should be considered in the treatment of LBPP, which have demonstrated effectiveness. It also suggested that studies collect and report on safety outcomes for trials involving pregnancy women.22 One previous review23 reported the effectiveness of acupuncture in treating LBPP for pregnant and postpartum women. However, the number of studies associated with LBPP during the pregnancy included were small. In addition, this review did not assess the safety of acupuncture for pregnancy outcomes and the newborns. Our study was an update of the previous review, which found high-quality evidence for acupuncture in treating LBPP during the pregnancy and only included pregnant women with LBPP. The objective of this study was to evaluate the clinical efficacy and safety of acupuncture for women with LBPP during the pregnancy.
Methods
Search strategy
This meta-analysis follows the guidelines for Preferred Reporting Items for Systematic Reviews and Meta-analyses and the STRICTA.24 The protocol for this meta-analysis was registered with PROSPERO (CRD42021241771). The following electronic databases were searched for RCTs without language restrictions from the inception dates to 15 January 2022: PubMed, EMBASE, Web of Science and the Cochrane Library. The following search string was used to search: “acupuncture,” “auricular acupuncture,” “low back pain,” “lumbar pain,” “pelvic pain,” “pregnancy” and “Randomized Controlled Trial”. The search strategy is detailed in online supplemental table 1. Two reviewers (JY and YW) independently screened all search results for potentially eligible studies after reading the titles and abstracts. We determined the final inclusion after reading the full text strictly. Disagreements were resolved by a discussion with a third reviewer (ZO).
Supplemental material
Selection criteria
The eligibility criteria of the studies were formulated according to the participants, interventions, comparison, outcomes and study design criteria.25 (1) Participants: Women with LBPP (pain located between the 12th rib and the gluteal fold, and distal and posterior pelvic region lateral to the lumbar-sacral junction) during the pregnancy without contraindications to acupuncture or significant risk factors; (2) Intervention: acupuncture, including traditional acupuncture (using thin and solid metallic needles to insert into specific parts), needle and auricular acupuncture (auricular pressure needles, 1.5 mm long and 0.20 mm in diameter,) acupuncture combined with other treatments; (3) Comparators: acupuncture versus other treatments, acupuncture versus no intervention, acupuncture versus placebo acupuncture, acupuncture combined with other treatments vs other treatments; (4) Outcomes: primary outcomes were pain (Visual Analogue Scale shows pain intensity where 0 describes ‘no pain’ and 10 ‘worst possible pain’.), functional status (Disability Rating Index (DRI) contains 12 common daily activities and the participants were required to rate their level of disability in each activity where 0 describes ‘no disability’ and 100 ‘maximum disability’.), and quality of life (Short Form-12 Health Survey shows quality of life). Secondary outcomes were overall effects, analgesic consumption (using analgesic drugs as an adjunct for pain treatment during the study period), Apgar scores >7 at 5 min (Apgar scores were used to assess the physical condition of the newborns, if scores were ≥7, the newborns were in good condition), and adverse events (adverse pregnancy outcomes and other effects from acupuncture); (5) Study design: RCTs. Studies met any of following criteria would be excluded: (1) pain caused by reasons other than pregnancy; (2) conference abstracts, repeated publications, animal experiments, case reports, observational studies or reviews.
Data extraction
Two reviewers (JY and YW) independently extracted the following data from the included studies. These included participant characteristics (eg, age, gestation weeks), study characteristics (eg, author names, publication year, region, study design, sample size, intervention type, intervention characteristics) and study outcomes. If necessary, the corresponding authors would be contacted for additional information.
Quality assessment
The methodological quality of included studies was assessed by two reviewers (JY and YW) according to the Cochrane Handbook for Systematic Reviews of Interventions.26 Each study was evaluated from seven items: random sequence generation (checking for possible selection bias), allocation concealment (checking for possible selection bias), blinding of participants and personnel (checking for possible performance bias), blinding of outcome assessment (checking for possible detection bias), incomplete outcome data (checking for possible attrition bias due to the amount, nature and handling of incomplete outcome data), selective reporting (checking for reporting bias), and other bias (checking for bias due to problems not covered by (1) to (5) above). Each item was assessed as low, high or unclear risk of bias. According to the Grading of Recommendations Assessment, Development and Evaluation System (GRADEprofiler, V.3.6),27 the quality of evidence was also assessed by two reviewers (JY and YW). Disagreements were resolved by a discussion with a third reviewer (ZO).
Statistical analysis
Review Manager (V.5.3) was used to conduct statistical analyses. All continuous variables were pooled by mean differences (MDs) with 95% CIs. Heterogeneity was evaluated by using Higginss I2 statistic, which ranges from 0% to 100%. I2 describes the percentage of the variability in effect estimates that is due to heterogeneity rather than sampling error (chance). A rough guide26 to interpretation in the context of meta-analyses of randomised trials is as follows: 0%–40% might not be important; 30%–60% may represent moderate heterogeneity; 50%–90% may represent substantial heterogeneity; 75%–100% may represent considerable heterogeneity. In terms of the overlapping I2, we did note the overlapping ranges and the equivocation (‘‘may’). After all, this was a rough guide according to Cochrane Handbook for Systematic Reviews of Interventions. Inevitably, studies brought together in a systematic review will differ. Any kind of variability among studies in a systematic review may be termed heterogeneity. A χ2 test p<0.1 or I2>50% indicates statistically significant heterogeneity.28 Fixed-effect meta-analyses ignore heterogeneity. The summary effect estimate from a fixed-effect meta-analysis is normally interpreted as being the best estimate of the intervention effect. However, the existence of heterogeneity suggests that there may not be a single intervention effect but a variety of intervention effects. Thus, the summary fixed-effect estimate may be an intervention effect that does not actually exist in any population, and therefore, have a CI that is meaningless as well as being too narrow. A random-effects meta-analysis may be used to incorporate heterogeneity among studies. This is not a substitute for a thorough investigation of heterogeneity. It is intended primarily for heterogeneity that cannot be explained. The random-effects method and the fixed-effect method will give identical results when there is no heterogeneity among the studies. Fixed effect models would be applied to combine results if statistical heterogeneity is low; If p<0.1 or I2 >50% we combined results using a random-effect model.29 Additionally, funnel plots were constructed, if possible, to assess publication bias. The difference was considered statistically significant when p<0.05.
Patient and public involvement
Patients and public were not involved in this study.
Results
Study selection
A total of 228 potentially eligible studies were identified by searching electronic databases, and none were included via other sources. All studies were imported into EndNote V.X9 (Bld,12062) for detection of duplicates. Following that, 188 were removed through title and abstract screening, and the remaining 20 studies were screened through full-text review, and 10 further studies were removed. Finally, 10 RCTs21 27 30–37 were included in the final analysis (figure 1).
Flow chart of study selection. RCT, randomised controlled trial.
Study characteristics
The countries in which the studies conducted were varied. Six studies were conducted in Sweden,21 27 30–32 35 one each in the UK,34 the USA,36 Spain37 and Brazil.33 All studies were published in English between 2000 and 2020. In total, 1040 participants were examined, with study sample sizes ranging from 18 to 130. All participants were healthy pregnant women (on average, 17–30 weeks of gestational age) with LBPP. Moreover, the two types of pain were not distinguished clearly in all studies. All studies recorded that the research team was multidisciplinary and the therapists involved were trained for acupuncture intervention. In three studies, acupuncture intervention was delivered by acupuncturists,21 33 36 in four by physical therapists,27 30 34 35 in two by midwives32 37 and unclear in another study.31 Seven studies reported that participants in acupuncture group received body acupuncture,21 27 30–34 and three studies used auricular press needles.35–37 All studies reported the acupuncture points for treatment, needle retaining time and intervention dose. However, seven studies used ‘forbidden points’ (GB 21, BL 60, BL 67, BL 32 or BL 33) during the pregnancy.21 27 30–34 Seven of the studies were supported by government or a professional organisation.21 27 30 32 34–36 Three studies did not report the funding.30 33 38 Two studies had greater than 20% drop-out rate.30 35 The characteristics of included studies were summarised in table 1.
Characteristics of the included studies
Risk of bias in included studies
The methodological quality of these studies was assessed by two reviewers (JY and YW) according to the Cochrane Handbook for Systematic Reviews of Interventions (figures 2 and 3).26 All subjects of these studies were randomly allocated to groups using adequate allocation procedure. Six studies stated that the allocation was concealed,30 31 34–37 while four studies were unclear.21 27 32 33 Two of the studies used subject blinding,30 34 three studies blinded assessors21 27 37 and none of the studies blinded the therapists. In addition, all studies provided an adequate summary of drop-outs. The drop-out rates are specifically reported fully in table 1. However, only one study employed intention-to-treat analysis.
Risk of bias graph.
Risk of bias graph.
Meta-analysis
Primary outcomes
All studies recorded outcomes on pain, but one of these studies provided insufficient data. We did not receive response by contacting authors. Meta-analysis results suggested that acupuncture significantly relieved pain during pregnancy (MD=1.70, 95% CI: (0.95 to 2.45), p<0.00001). High heterogeneity was detected between studies (p<0.00001, I2=90%, figure 4). However, the heterogeneity was still high heterogeneity after excluding extreme outliers,32 and the overall effect remained unchanged.
Forest plot of meta-analysis for pain. IV, inverse variance.
Functional status was recorded in four studies. Two studies evaluated functional status using DRI. Meta-analysis results suggested that acupuncture significantly improved functional status for pregnant women with LBPP (MD=12.44, 95% CI: (3.32 to 21.55), p=0.007) High heterogeneity was detected between studies (p<0.0001, I2=94%; figure 5). In addition, one study used Oswestry Disability Index and the other used Roland Morris Disability Questionnaire to measure the outcome on functional status. The results of two studies also indicated that acupuncture significantly improved functional status for pregnant women with LBPP, although were not included in the meta-analysis.
Forest plot of meta-analysis for functional status. IV, inverse variance.
Quality of life was recorded in five studies. Meta-analysis results suggested that acupuncture significantly improved quality of life for pregnant women with LBPP (MD=−8.98, 95% CI: (–11.90 to –5.88), p<0.00001). Moderate heterogeneity was detected between studies (p=0.06, I2=57%; figure 6). However, there was no heterogeneity after excluding extreme outliers,37 while the overall effect remained unchanged.
Forest plot of meta-analysis for quality of life. IV, inverse variance.
Secondary outcomes
Overall effects were recorded in four studies. Meta-analysis results suggested that there was a significant difference in overall effects when acupuncture was compared with other interventions or no intervention (OR=0.13, 95% CI: (0.07 to 0.23), p<0.00001). Low heterogeneity was detected between studies (p=0.36, I2=7%; figure 7).
Forest plot of meta-analysis for overall effects. M-H, Mantel-Haenszel.
Analgesic consumption was recorded in two studies. Meta-analysis results suggested that there was no significant difference in analgesic consumption during the study period when acupuncture was compared with no intervention (OR=2.49, 95% CI: (0.08 to 80.25), p=0.61). Moderate heterogeneity was detected between studies (p=0.11, I2=61%; figure 8).
Forest plot of meta-analysis for analgesic consumption. M-H, Mantel-Haenszel.
In addition, Apgar scores >7 at 5 min were recorded in four studies. Meta-analysis results suggested that acupuncture is a safe treatment and there was no significant difference in Apgar scores of newborns when acupuncture was compared with other interventions or no intervention (OR=1.02, 95% CI: (0.37 to 2.83), p=0.97). No heterogeneity was detected between studies (p=0.62, I2=0%; figure 9).
Forest plot of meta-analysis for Apgar scores. M-H, Mantel-Haenszel.
In terms of gestational age at birth, induction of labour and mode of birth, only one study reported the gestational age at birth (mean gestation 40 weeks). This study was unable to be included in the Meta-analysis. Thus, prospective randomised clinical studies or clinical follow-up studies were hence desirable to further evaluate these outcomes.
Adverse events
All studies examined adverse events. Preterm birth from acupuncture during study period were reported in two studies. Although preterm contractions were reported in two studies, all infants were in good health at birth. Additionally, seven studies recorded other expected minor adverse effects from acupuncture. Bishop et al reported that all were expected minor adverse events with the most common being pain and slight bleeded at the needle sites. Slight bruising, drowsiness and slight soreness which occurred in less than 4% of treatment sessions. Guerreiro et al reported that two women mentioned small bruises, three ecchymosis at one or two points. Kvorning et al reported that one or more of the following symptoms were reported by or found in 38% of women given acupuncture: local pain, sweating, local haematoma, tiredness, nausea and weakness. Jorge et al reported that two women suffered pain and redness at the acupuncture site. Wang et al reported that one woman experienced transient ear tenderness that resolved spontaneously. Wedenberg et al reported that most women reported a feeling of tiredness or sleepiness but also a sensation of well-being, and two women experienced small subcutaneous haematomas in the ear. Elden et al reported that the following symptoms: Headache plus severe drowsiness (n=1); headache (n=1); rash developed on the place of the needles at (n=2); severe nausea with feeling faint, sweating and dizziness (n=4). However, these responses disappeared spontaneously with no impact on continuation of treatment. Nevertheless, in spite of the minor adverse events, the women rated acupuncture favourably and a majority of them were willing to use the same treatment in the future if needed (table 2). Additionally, two studies had greater than 20% dropout rate.30 35 Nevertheless, in the study reported by Lund et al,30 13 women (19%) were excluded from analysis due to noncompliance in submitting pain diaries not due to the adverse events rate. In the study reported by Wedenberg et al,35 20% drop-out rate occurred in the control group.
Adverse events of the included studies
Sensitivity analysis
We conducted sensitivity analyses of primary outcomes using the leave-one-out approach39 (online supplemental table 2). The analysis results indicated that the meta-analysis results for pain outcome did not alter when each study was removed in turn, and the findings were robust. For example, meta-analysis results suggested that acupuncture significantly relieved pain during pregnancy (p<0.00001), and p value was associated with test for overall effect. In addition, high heterogeneity was detected between studies (I2=90%), and I2 >50% indicates statistically significant heterogeneity. However, the meta-analysis results for pain outcome did not change when the study by Bishop et al was removed, suggesting that this study was not responsible for the between-study heterogeneity (p<0.0001 (p value was associated with test for overall effect), I2=90%). Similarly, the analysis results for pain outcome did not change when other studies were removed in turn, which suggested that maternal age and gestational weeks might be the primary source of heterogeneity. The heterogeneity of functional status outcome was very high. Except for the difference in pregnant woman’s age, gestational age and the number of children. The outcome of functional status was measured by different methods, which might lead to the heterogeneity. There was no heterogeneity for the outcome of quality of life after excluding the study by Lund et al, which suggested that this study could be the potential source of heterogeneity.
Publication bias
Funnel plot for pain was constructed to examine publication bias and revealed a degree of asymmetry (figure 10). Funnel plots are the primary visual tool for investigation of publication bias in a meta-analysis, yet the use of a funnel plot is subjective to researchers and may be interpreted differently by different researchers. This is a scatter of the estimate of the effect size (binary: logOR or logRR) against the measure of precision for each study that is included in the meta-analysis. In the absence of publication bias, the plot resembles a symmetrical inverted funnel; otherwise, it is asymmetrical. It is important to realise that publication bias is not the only source that contributes to asymmetry. There are at least two main sources that can contribute to asymmetry of a funnel plot. One is publication bias: small and unfavourable treatment effects are more likely to be missing, and small studies with large effect sizes are likely to be published. Another one is the heterogeneity, for example, smaller studies may select women who are more likely to benefit from the intervention. Visual inspection of funnel plots revealed that it was asymmetrical. This might be related to the fact that some trials showing adverse effects from acupuncture may not be published and are not included in this meta-analysis.
Funnel plots of nine RCTs recording pain. MD, mean difference; RCTs, randomised controlled trials.
Reporting quality assessment
The reporting quality of included RCTs according to the STRICTA checklists40 was shown in table 3. There are 17 subitems included in 6 items. Item 1: for style of acupuncture, treatment theory and the extent to which treatment was varied, three studies21 27 31 used segmental acupuncture, one study34 used the principles of Western acupuncture and trigger point acupuncture, and five studies21 27 30 31 33 provided ‘the extent to which treatment was varied’. Item 2: number of needle insertions per subject per session, names (or location if no standard name) of points used, depth of insertion, responses sought, needle stimulation, needle retention time and needle type were generally well described. Item 3: number of treatment sessions and frequency and duration of treatment sessions were well reported in all studies. Item 4: details of other interventions administered to the acupuncture group were reported in four studies.33–35 37 Setting and context of treatment, including instructions to practitioners, and information and explanations to patients were not described in any study. Item 5: All studies recorded that the research team was multidisciplinary and the therapists involved were trained for acupuncture intervention. Item 6: the rationale for the control or comparator in the context of the research question were described in six studies.27 33 34 37 A precise description of the control or comparator was described in five studies.30 31 34 35 38
Quality of reporting according to stricta guideline
Quality of evidence
Nine studies (n=815) contributed data for the outcome of pain. Adjusted pooled results showed that acupuncture significantly relieved pain during pregnancy (MD=1.70, 95% CI 0.95 to 2.45, p<0.00001). There is low evidence for pain outcomes regarding the GRADE confidence in this estimate. Two studies (n=151) contributed data for the outcome of functional status. Adjusted pooled results suggested that acupuncture significantly improved functional status for pregnant women with LBPP (MD=12.44, 95% CI 3.32 to 21.55, p<0.0001). There is moderate evidence for functional status regarding the GRADE confidence in this estimate. Five studies (n=337) contributed data for the outcome of quality of life. Adjusted pooled results showed that acupuncture significantly improved quality of life for pregnant women with LBPP (MD=−0.98, 95% CI −1.35 to −0.61, p<0.00001). There is low evidence for quality of life regarding the GRADE confidence in this estimate. Four studies (n=284) contributed data for the outcome of overall effects. Adjusted pooled results suggested that there was a significant difference on overall effects when acupuncture was compared with other interventions or no intervention (OR=0.13, 95% CI 0.07 to 0.23, p<0.00001). There is moderate evidence for overall effects regarding the GRADE confidence in this estimate. Two studies (n=177) contributed data for the outcome of analgesic consumption. Adjusted pooled results showed that there was no significant difference on analgesic consumption during the study period when acupuncture was compared with no intervention (OR=2.49, 95% CI 0.08 to 80.25, p=0.61). There is low evidence for analgesic consumption regarding the GRADE confidence in this estimate. Four studies (n=471) contributed data for the outcome of Apgar score of 7–10. Adjusted pooled results suggested that there was no significant difference on Apgar scores of newborns when acupuncture was compared with other interventions or no intervention (OR=1.02, 95% CI 0.37 to 2.83, p=0.97). There is low evidence for Apgar scores regarding the GRADE confidence in this estimate (online supplemental table 3).
Discussion
Summary of main results
This meta-analysis included 10 studies, reporting on a total of 1040 women with LBPP during the pregnancy. Results of this study demonstrated that acupuncture significantly relieved LBPP and improved functional status and quality of life during the pregnancy. Moreover, there was a significant difference of acupuncture in overall effects. There was no significant difference in analgesic consumption during the treatment and Apgar scores of newborns after the treatment.
For the adverse events, no preterm labour or other adverse pregnancy outcomes were reported during acupuncture. In addition, ‘forbidden points’’ were used in seven studies. Preterm birth from acupuncture during the study period were reported in two studies. But all infants were in good health at birth despite preterm contractions were reported in two studies, which might result in women withdrawing from the research studies. Other expected minor adverse effects from acupuncture were recorded in seven studies, but these responses disappeared spontaneously with no impact on continuation of treatment. Therefore, acupuncture could be considered a relatively safe and effective intervention in the treatment of LBPP during the pregnancy.
Comparison with other reviews
In recent years, only a few studies have systematically investigated the effects of acupuncture for LBPP during pregnancy. One systematic review revealed that acupuncture was effective in treating LBPP during pregnancy using limited evidence.41 However, the authors did not conduct meta-analyses because of the paucity of additional high-quality evidence, which limited the comparison with our study. Koukoulithras et al 42 conducted a study to compare the effectiveness between various methods and typical care on low back pain during pregnancy. However, only one study used auricular acupuncture in the intervention group from all the studies included in this meta-analysis, the rest were exercise, manipulation, transcutaneous electrical nerve stimulation and neuroemotional technique in the intervention group, which precluded comparison with our study. In a recent review and meta-analysis,23 the results suggested that acupuncture significantly relieved pain for both pregnant and postpartum women, but the number of studies associated with LBPP during the pregnancy included was small and the safety of materal acupuncture on the outcomes for the newborn was not assessed. One existing literature tested the effects of adding acupuncture to standard care for pregnancy-related back pain, but it was not a review. In addition, this study mainly assessed the feasibility of a future large, multicentre RCT through surveying midwives, physiotherapists and pregnant women.19
How the intervention might work
LBPP, always difficult to distinguish between the two types for pregnant women, are a common and serious ailment worldwide.43 Pregnant women with LBPP often suffered from insomnia, depression, anxiety, disturbance of activity and low quality of life as well.10 41 43 Pregnancy was a time where interventions were necessarily minimal, and should not include pharmaceuticals unless absolutely required. This was due to concerns of teratogenicity for the fetus and harm to the mother.44 Thus, non-pharmacological therapies are potentially attractive, such as exercise, physical therapy, massage, yoga and acupuncture, which could reduce pain and relieve some comorbidities.45–47 Acupuncture is often thought to be an alternative form or complementary therapy in pain. It has been identified that acupuncture can promote the release of endogenous opioids, which could be one of the analgesic mechanisms for pain management.48 It also significantly increases blood flow of local skin and muscle to relieve various pain conditions.49 Moreover, acupuncture appears to be effective in reducing the severity of depression,50 which also might alleviate pain caused by emotional problems.
Implications for clinical practice and further research
Given the poor methodological quality of the included studies and the limited strength of evidence, future trials should use more rigorous and robust methodology. Future trials should reduce the risk of bias by ensuring intention-to-treat analysis. To minimise publication bias, the researchers should register their studies in the clinical trial registry before recruiting participants. Moreover, in our study, the average age of the participants was rather young (mostly <35 years of age), the possibility of serious adverse effects to acupuncture in pregnant women older than 35 years of age could not be excluded despite that no serious complications (abortion) were observed. Further trials are hence needed to further evaluate the efficacy or adverse effects of acupuncture in pregnant women older than 35 years of age. Additionally, some important outcome measures were not included in these studies, including: mode of birth (vaginal birth, instrumental vaginal birth, caesarean section), pain relief used during labour (opioid, epidural, sterile water injections) and onset of labour (induced, spontaneous). These were vital components that needs to be included in future studies. With regard to Apgar scores, if researchers were able to capture this data, presumably by accessing patient records following birth as part of routinely collected data, it could feasibly be included in the outcome data. Considering that acupuncture has potential therapeutic value, therefore, it is necessary to further evaluate the optimal intervention programme of acupuncture (eg, acupoints, frequency and duration).
Limitations
The limitations of this study are similar to most systematic reviews. First, the number of studies included was limited. Second, the results of the study should be interpreted with caution given the quality of the included studies. Third, some factors related to heterogeneity remained uncertain despite we explained statistical heterogeneity through sensitivity analyses to some extent. However, considerable heterogeneity remained unclear. This might associate with the differences in characteristics of participants (eg, age, gestational weeks and numbers of children), acupuncture intervention characteristics (eg, acupoints, frequency and duration) and the measuring method of outcomes. Moreover, no study reported cost of acupuncture treatment. Therefore, the cost-effectiveness of acupuncture could not be evaluated.
Conclusion
Acupuncture significantly improved pain, functional status and quality of life in women with LBPP during the pregnancy. Additionally, acupuncture had no observable severe adverse influences on the newborns. More large-scale and well-designed RCTs are still needed to further confirm these results.
Data availability statement
Data are available on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
Acknowledgments
The authors thank Dehui Fan and Yingzi Huang of the Fifth Clinical Medical College of Guangzhou University of Chinese Medicine, Guangzhou, China, for carrying out the searches.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Press Release
Footnotes
JY and YW are joint first authors
Contributors JY, TW and ZS conceived and designed the study. WD and YL served as scientific advisors. ZS and YL critically reviewed the study proposal. JY, YW, JX and ZO performed the literature search. JX, ZM and TY performed the data extraction. JY, TY and ZM analysed and interpreted the data. JY, WD and ZS did statistical analyses. JY, YW, ZO and TW drafted the review. WD and ZS did the language editing. All authors revised the article for important content and approved the final version for the article. JY is responsible for the overall content as guarantor.
Funding This work was supported by Scientific Research Project of Traditional Chinese Medicine Bureau of Guangdong Province, project number 20213002.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.