Chinese Herbal Medicine for Cervicogenic Dizziness: A Systematic Review and Meta-Analysis

Background Chinese herbal medicines (CHMs) have been widely used in the treatment of cervicogenic dizziness (CGD) based on their empirical effectiveness and safety. Herein, we reviewed and evaluated the clinical evidence of the efficacy and safety of CHMs for CGD. Methods Among the relevant studies published in 11 electronic databases up to December 2021, only randomised controlled trials were included. Methodological quality was assessed using the revised Cochrane risk-of-bias tool for randomised trials, and the strength of evidence for the main outcomes was evaluated using the grading of recommendations assessment, development, and evaluation system. Results All 35 included randomised controlled trials with 3,862 participants were conducted with six types of modified CHM and four types of active controls. More than half of the included studies were of low quality because of the high risk of bias due to deviations from intended interventions. CHM plus active control was more effective in the treatment of CGD than active control alone. CHM plus anti-vertigo drugs, CHM plus manual therapy, CHM plus acupuncture therapy, and CHM plus manual and acupuncture therapy were all effective in treating CGD, with CHM plus manual and acupuncture therapy showing the most reliable effect. All CHMs were effective for specific patterns of CGD when administered with active controls, with Dingxuan Tang and Yiqi Congming Tang demonstrating the most reliable effects. No serious adverse events were reported in any of the included studies. Conclusion The current evidence suggests that CHM may enhance the treatment of CGD when combined with other treatments without serious adverse events. Further high-quality evidence is needed to draw definitive conclusions.

Parlour syndrome. However, these diseases also overlap because they do not have completely distinct mechanisms [15]. Because there are no established diagnostic criteria for CGD, physicians usually diagnose CGD when the patients' symptoms are not related to other neurological or neurootological causes of dizziness [16,17]. e treatment of CGD has not yet been standardised. Previous studies have explored a variety of treatments to improve the severity and frequency of dizziness by relaxing muscles and ameliorating abnormal proprioceptive sensitivity or impaired blood flow in the cervical region. Treatment strategies include physical therapy [1,3,7,10,[18][19][20][21][22], surgery [10,16], topical drug injection [9,23], acupuncture therapy [24,25], and medications, such as muscle relaxants, opioids, nonsteroidal anti-inflammatory drugs, and anxiolytics, in combination with Chinese herbal medicines (CHMs). CHMs have been widely used for CGD, either alone or in combination with other treatments, based on their empirical effectiveness to suppress pain and improve blood circulation in the human body [24,26]. However, there has been no systematic verification of their efficacy and safety in the treatment of CGD based on clinical evidence. erefore, we aimed to review and evaluate the clinical evidence on the efficacy and safety of CHM as monotherapy or adjunctive therapy for CGD, which would promote evidence-based decision-making in clinical practice.

Study Registration.
e study protocol for this systematic review was registered with the International Prospective Register of Systematic Reviews (registration number: CRD42020199222; registration date: October 27, 2020) and the Research Registry (Review Registry Unique Identifying Number: reviewregistry1036; registration date: November 19, 2020). e study protocol was published [27], and there have been no subsequent amendments that could result in a significant change in the study design. is systematic review is reported in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analyses statement [28]. A preprint has previously been published in Research Square (DOI: https://doi.org/10. 21203/rs.3.rs-364098/v1; registration date: March 31, 2021) [29]. one Japanese database (Citation Information by NII). A manual search on Google Scholar was also performed to identify additional eligible studies among those listed in the reference sections of included studies. e search strategies were tailored to the language and search systems of the databases. e search strategies used in the three English databases (MEDLINE, EMBASE, and CENTRAL) are presented in Additional file 1.

Eligibility Criteria
2.3.1. Types of Studies. All randomised controlled trials (RCTs) related to the use of CHMs for CGD were included. All other study designs, including quasi-RCTs, were excluded.

Participants.
All patients with CGD were included as subjects in this study, with no restrictions on ethnicity, nationality, sex, age, or biological status.

Interventions and Comparisons.
CHMs with any formulation administered orally, such as decoction, capsules, tablets, pills, and powders, were considered experimental interventions.
ere was no limitation on the number or combination of herbs, CHM dose, or the frequency or duration of treatment. If the composition of CHMs used in the included studies differed from the original prescription, "modified" was indicated in front of the CHM name. No treatment and placebo were considered as control interventions to determine the efficacy of CHM as monotherapy. Active controls, such as anti-vertigo drugs, manual therapy, and acupuncture therapy, were also considered as control interventions to determine the efficacy of CHM as adjunctive therapy only when CHMs were equally applied to both the experimental and control groups. Studies comparing different combinations of CHMs or CHM alone with other active controls were excluded because they could not rigorously determine the efficacy of CHMs.

Outcomes.
e primary outcomes were as follows: (1) e change in the patients' overall functional score measured by validated scales (e.g., functional scale for cervical spondylosis of vertebral artery type) (2) e change in the patients' simple score for dizziness (e.g., the numerical rating scale) (3) e change in mean blood flow velocity in the vertebrobasilar artery, as evaluated using transcranial Doppler e secondary outcomes were as follows: (1) e total effective rate, strictly calculated by counting only the number of patients completely cured, to exclude researcher subjectivity and improve the reliability of the results 2 Evidence-Based Complementary and Alternative Medicine (2) e changes in haematological parameters, such as fibrinogen levels, endothelin, total cholesterol (TC), and calcitonin gene-related peptide (CGRP) (3) Adverse events 2.4. Study Selection Process. Two reviewers (HO and SS) independently screened and assessed all retrieved studies for eligibility based on the aforementioned criteria. After duplicates were removed, the titles and abstracts of the remaining studies were screened using EndNote X9 (Clarivate Analytics, London, UK). Next, the full-text review of the eligible studies was conducted for final inclusion. Any divergence in the agreement was resolved through discussion with a third researcher (EL) at each step of the study selection process.

Data Extraction.
Two reviewers independently extracted data from the included studies (HO and SS) using a predefined data acquisition form. is form included four main domains: general information (title, authors, year of publication, country of the study, and study design), participants' characteristics (age, sex, diagnostic criteria, and CGD duration), intervention and comparison details (sample size; CHM formulation and prescription name; number of herbs; CHM dose; CHM daily dose; comparison, frequency, or duration of the treatment; and follow-up information), and outcomes (primary and secondary outcomes and adverse events). Any discrepancies were resolved through discussion with a third researcher (EL).

Quality Assessment.
e methodological quality of the included studies was assessed using the revised Cochrane risk-of-bias tool for randomised trials [30]. e bias domain for risk-of-bias assessment included the following: (1) bias arising from the randomisation process, (2) bias due to deviations from intended interventions, (3) bias due to missing outcome data, (4) bias in the measurement of the outcome, and (5) bias in the selection of the reported result. e risk of bias was independently evaluated by two reviewers (HO and SS) as "low," "high," or "some concerns." Any divergence in the agreement was resolved through discussion with other reviewers (EL and WSC). Studies evaluated as "low-risk" in all domains were defined as highquality studies, whereas those evaluated as "high-risk" in at least one domain were defined as low-quality studies.
Subsequently, the strength of evidence for the main outcomes was evaluated using the grading of recommendations assessment, development, and evaluation system [31]. e risk of bias; inconsistency, indirectness, and imprecision of the results; and publication bias were assessed, and the quality of the evidence was graded on a four-point scale as "high," "moderate," "low," or "very low." 2.7. Data Synthesis. When the included studies were sufficiently homogenous, quantitative synthesis was performed using RevMan software (version 5.3; Cochrane, London, UK) to analyse the efficacy of CHMs in the treatment of CGD. Subgroup analyses were conducted according to (1) the comparison types and (2) the CHM prescription names. Dichotomous outcomes were pooled using risk ratios (RRs), and continuous outcomes were pooled using mean differences (MDs), or standardised mean differences (SMDs), with 95% confidence intervals (CIs). e statistical heterogeneity among studies was assessed by computing I 2 statistics. Data were pooled using a random-effects model, if the included studies had significant heterogeneity (I 2 values ≥50% indicated substantial heterogeneity and I 2 values ≥75% indicated considerable heterogeneity (both were considered significant)). Otherwise, a fixed-effects model was applied [32]. Sensitivity analysis was performed to increase the robustness of the results by excluding studies with a high risk of bias and outliers. If the number of studies was sufficient (n ≥ 10), a visual inspection of the funnel plot was performed to assess publication bias. Data on the safety of CHMs in the treatment of CGD were described qualitatively.

Study Selection.
A total of 8,746 studies were identified through the database searches, and 1 additional study was identified through other sources. After removing 305 duplicates, 8,442 studies were excluded by screening the titles and abstracts. rough a review of the full texts, a further 659 studies were excluded: 17 studies with unavailable full texts, 31 nonclinical studies, 21 case reports, 164 noncomparative studies, 13 nonrandomised controlled trials, 258 studies not related to CGD, 49 studies not related to eligible intervention, and 106 studies not related to the clinical question. Finally, 35 RCTs with 3,862 participants were included in the analysis ( Figure 1).
In addition, 10 types of outcome measurements were identified: 5 studies evaluated overall functional scores, 22 studies evaluated simple scores, 17 studies assessed the mean blood flow velocity in the vertebral arteries, 18 studies assessed the mean blood flow velocity in the basilar artery, 33 studies evaluated the total effective rate, three studies measured endothelin levels, and four studies measured CGRP, fibrinogen and TC levels. e incidence of adverse events was reported in three studies. e study characteristics and the main outcomes are summarised in Table 1.
Each CHM prescription was applied to a specific pattern of symptoms in traditional Chinese medicine: BBTT to the wind-phlegm type or phlegm stasis type; BYT to qi and blood deficiency type; DXT to spleen deficiency and dampness type, qi deficiency and blood stasis type, or hyperactivity of liver yang type; GGT to wind type with disharmony between ying and wei; GJT to collateral stasis type; and YCT to qi and blood deficiency type or qi deficiency and sputum silting up type. All modified CHMs included at least one-third of the original prescriptions. e duration of administration ranged from 10 days to 8 weeks, with 2-and 4-week regimens being the most frequent. e details of the CHMs prescribed in the included studies are summarised in Tables 2 and 3.

Risk-of-Bias Assessment.
For bias arising from the randomisation process, 18 studies were evaluated as "lowrisk" because the randomisation process for the allocation sequence was clearly described. e remaining 17 studies were evaluated as "some concerns" because insufficient relevant information was provided. For bias due to deviations from intended interventions, 21 studies, most of which included manual or acupuncture therapy as active controls, were evaluated as "high-risk" because it was unclear whether blinding of participants and trial personnel had been sufficiently performed using sham-massage or sham-acupuncture. e remaining 14 studies were evaluated as "some concerns." For bias due to missing outcome data, 30 studies

NR
Evidence-Based Complementary and Alternative Medicine 5 Evidence-Based Complementary and Alternative Medicine

NR
Shang [52] 82        (2) 10 Evidence-Based Complementary and Alternative Medicine    Evidence-Based Complementary and Alternative Medicine     14 Evidence-Based Complementary and Alternative Medicine    16 Evidence-Based Complementary and Alternative Medicine were evaluated as "low-risk," and 1 study was evaluated as "high-risk" because there were missing data (only the results of the per-protocol analysis were reported). e remaining 4 studies were evaluated as "some concerns" because insufficient relevant information was provided. For bias in the measurement of the outcome, 20 studies were evaluated as "low-risk," and the remaining 15 studies were evaluated as "some concerns" because it was difficult to judge whether the outcome measures used in the studies were affected by the awareness of the outcome assessors. For bias in the selection of the reported result, 3 studies were evaluated as "low-risk" because there was no suspicion of deliberate nonreporting, and 3 studies were evaluated as "high-risk" because selective outcome reporting was suspected. e remaining 29 studies were evaluated as "some concerns" because there was no basis for bias assessment (e.g., study protocols). Finally, for the overall risk of bias, 23 studies assessed as "high-risk" were considered low-quality studies; 2 were considered high-quality studies; and the remaining 10 studies were evaluated as "some concerns" (Figure 2). e risk of bias was evaluated as "low," "high," or "some concerns," represented by the following symbols: "L," "H," and "C," respectively. D, bias due to deviations from intended interventions; Me, bias in the measurement of the outcome; Mi, bias due to missing outcome data; O, overall risk of bias; R, bias arising from the randomisation process; and S, bias in the selection of the reported result.

CHMs plus Manual erapy versus Manual erapy
Alone. In the subanalysis of the seven studies using manual   Evidence-Based Complementary and Alternative Medicine MD, 16.48 (95% CI: −0.34-33.31); I 2 � 98%) did not show significant differences between the intervention and control groups.

CHMs plus Manual and Acupuncture erapy versus
Manual and Acupuncture erapy Alone. In the subanalysis of the one study using manual and acupuncture therapy as an active control, CHMs plus manual and acupuncture therapy significantly reduced the overall functional scores (one study: MD, 7.06 (95% CI: 6.27-7.85)) and improved the total effective rate (one study: RR, 1.40 (95% CI: 1.02-1.94)), compared with the active control alone.

GJT plus Active Controls versus Active Controls Alone.
In the subanalysis of the one study using GJT as CHM, compared with the betahistine alone, GJT plus betahistine significantly increased the simple scores (one study: MD, 2.00 (95% CI: 1.75-2.25)). However, the total effective rate (one study: RR, 2.19 (95% CI: 0.99-4.86)) was not significantly different between the intervention and control groups.

YCTplus Active Controls versus Active Controls Alone.
In the subanalysis of the eight studies using YCT as CHM, compared with the active controls alone, YCT plus active controls significantly reduced the overall functional scores        Summarizing the results of the subanalysis according to CHM prescription names, BBTT, DXT, GGT, and YCT showed significant treatment effects on various primary and secondary outcomes and had relatively more clinical evidence compared with the remaining CHM prescription names. GJT was investigated in only one RCT and demonstrated a significant effect on only one primary outcome (change in the simple scores), without statistically significant effects on the other outcome (total effective rate). In the two RCTs investigating BYT, there were significant effects on two primary outcomes (change in the simple scores and blood flow velocity in the left vertebral artery), while the effects on the remaining outcomes were either not significant (blow flow velocity in the basilar artery and total effective rate) or were significant in the control group (the blood velocity for the right vertebral artery). e results of the total analysis and the subanalyses of the efficacy of CHMs are shown in Table 4.

Safety.
ree of the thirty-five included studies reported adverse events. ere was one case of gastrointestinal discomfort in the BBTT plus manual therapy group; one case of abdominal pain; one case of fainting during acupuncture therapy in the BBTT plus acupuncture therapy group; one case of rash, diarrhea, and gastrointestinal discomfort each, and two cases of fatigue in the DXT plus anti-vertigo drugs (flunarizine and betahistine) group. All reported adverse events were mild and transient and were evaluated as "not serious" (Table 1).

Quality of Evidence.
In the comparison of CHMs plus active controls versus active controls alone, the quality of evidence for the primary outcomes ranged from "very low" to "low." For the secondary outcomes, the quality of evidence for the total effective rate was graded as "moderate," while that for the other outcomes was graded as "very low" or "low." e overall quality of evidence in the total analysis was graded as "low." In the subanalysis based on the type of active control, the overall quality of evidence was graded as "moderate" for CHMs plus manual and acupuncture therapy and as "low" for CHMs plus any other active control (antivertigo drugs, manual therapy, or acupuncture therapy). In the subanalysis based on the CHM prescription name, the overall quality of evidence was evaluated as "low" for all CHM prescriptions. However, its quantitative and qualitative levels were highest for DXTand YCTand lowest for BYT and GJT, respectively. e main reason for the downgrade was the high risk of bias in the included studies, the imprecision of the results due to the small sample size, and the inconsistency of the results due to the high heterogeneity among them (Table 4).

Sensitivity Analysis.
For the outcomes with considerable heterogeneity among studies, we performed sensitivity analysis and adjusted the quality of evidence based on the results. After heterogeneity was eliminated by removing one to two outliers considered to have a high risk for selection and reporting biases, the quality of evidence for the efficacy of CHMs for CGD was similar to that obtained before the sensitivity analysis. erefore, the findings in this systematic review are considered robust to the decisions made in the process of obtaining them (Table 5).

Publication Bias.
For seven outcomes included in more than ten studies, we examined publication bias using funnel plot analysis. For the comparisons of CHMs plus active controls, anti-vertigo drugs, or acupuncture therapy versus active controls, anti-vertigo drugs, or acupuncture therapy alone, respectively, the funnel plots of the total effective rate were symmetrical for all (Figures 3-5). Conversely, for the comparison of CHMs plus active controls versus active controls, the funnel plots of the simple scores and the blood flow velocity in the vertebrobasilar arteries showed

26
Evidence-Based Complementary and Alternative Medicine asymmetry. In the funnel plot of the blood flow velocity in the left vertebral artery, the asymmetry was presumed to be due to considerable heterogeneity. e asymmetry for the remaining outcomes suggested potential publication bias; thus, there may be negative results not published in the literature (Figures 6-9).

Summary of Findings.
In this study, we reviewed and evaluated the available clinical evidence on the efficacy and safety of CHM as monotherapy or adjunctive therapy in the treatment of CGD to promote evidence-based decisionmaking in clinical practice. As none of the 35 included RCTs  assessed the efficacy of CHM as monotherapy for CGD, we evaluated its efficacy as adjunctive therapy in combination with other active controls. e included studies were conducted with 6 types of modified CHMs and 4 types of active controls. In the risk-of-bias assessment, more than half of the included studies were considered to be of low quality because of the high risk of bias due to deviations from intended interventions. e results of the efficacy analyses of CHMs plus active controls indicated the following. First, CHMs plus active controls were more effective in treating CGD than active controls alone (the duration of administration ranged from 10 days to 8 weeks). Second, CHMs plus anti-vertigo drugs (flunarizine/betahistine/flunarizine and  betahistine/diphenidol/nimodipine), CHMs plus manual therapy, CHMs plus acupuncture therapy, and CHMs plus manual and acupuncture therapy were all effective in treating CGD. Among all, CHMs plus manual and acupuncture therapy showed the most reliable effect. ird, BBTT, BYT, DXT, GGT, GJT, and YCT were effective for specific patterns in patients with CGD, when administered with active controls. Among the CHM prescriptions, DXT and YCT exhibited the most reliable effects, when combined with active controls. Regarding the safety of CHMs plus active controls in the treatment of CGD, no serious adverse events were reported in any of the included studies.

Implications for Clinical Practice.
In traditional Chinese medicine, CHMs are prescribed to match the specific pattern of the patients' signs and symptoms. It is reasonable to select and prescribe the most appropriate CHM for a specific pattern in each patient with CGD, as opposed to consistently prescribing one CHM to all patients with CGD, even if it is the most evidence-based prescription for CGD. us, although DXT and YCT had the highest level of clinical evidence for the treatment effect on CGD in this review, it may be more effective to use other CHMs for specific patterns in some patients with CGD. In traditional Chinese medicine, wind, fire, phlegm, blood stasis, and deficiency are considered the main pathogenetic factors for CGD [25]. DXT is usually prescribed for CGD syndromes of spleen deficiency and dampness, qi deficiency and blood stasis, or hyperactivity of liver yang. DXT has the effect of removing a pathogenic mass as the original prescription, and it can be prescribed for both deficiency and excess syndromes by modification of the original prescription. CHMs can be modified for better efficacy and fewer side effects [68]. In cases of combined excess and deficiency syndromes, such as spleen deficiency and dampness type or qi deficiency and blood stasis type, DXT was modified by the addition of herbs that have effects on invigorating the qi and spleen (Codonopsis pilosulae Radix and Atractylodis Rhizoma Alba), regulating qi-flowing (Citri Reticulatae Pericarpium), enriching the blood (Angelicae Gigantis Radix), and soothing the nerves (Fossilia Ossis Mastodi), but with subtraction of other herbs from the original prescription, which have effects on suppressing hyperactive liver for calming endogenous wind (Uncariae Ramulus Cum Uncis and Scorpio) and promoting blood circulation while removing blood stasis (Salviae miltiorrhizae Radix) [43,47]. Conversely, in cases of excess syndrome only, such as hyperactivity of liver yang type, DXT was modified by adding Puerariae Radix, which has the effect of dispelling wind-heat [57,66]. For the combination of DXT and other treatments, quantitative clinical evidence has been reported for the use of DXT with manual therapy [47,57,66]. Both YCTand BYT are usually prescribed for CGD syndromes of qi and blood deficiency, while YCT is also used for the sputum silting up type. e clinical evidence for YCT is better than that for BYT because the latter showed low precision for outcomes. For the combination of YCT and other treatments, the majority of quantitative clinical evidence was reported for the use of YCT with acupuncture therapy [36,38,45,51,54,59,62]. BBTT, which has the effect of dispelling pathogenic wind and eliminating phlegm, is usually prescribed for CGD syndromes of wind-phlegm or phlegm stasis [37,39,56]. GGT is usually prescribed for CGD syndromes of wind with disharmony between ying and wei [52]. Both BBTT and GGT were used with various active controls and showed reliable treatment effects. GJT was used for the collateral stasis type with betahistine [40]. rough this review, we gain a clue about the relationship between specific patterns of CGD and CHM prescriptions; however, it remains unknown which CHM prescription is most effective for specific patterns of CGD because all included studies used only one CHM prescription with one specific pattern of CGD. Furthermore, studies are needed to confirm which CHM prescription is most effective for specific patterns of CGD.

Implications for Research.
In this review, we identified fibrinogen, endothelin, TC, and CGRP as haematological parameters used in clinical studies on CGD. Endothelin and CGRP were used as indicators to determine the efficacy of   Evidence-Based Complementary and Alternative Medicine CHMs plus anti-vertigo drugs and CHMs plus manual therapy. Fibrinogen and TC were used to determine the efficacy of CHMs plus acupuncture therapy. Endothelin is an endogenous vasoconstrictor that reduces the perfusion of brain tissues by constricting the blood vessels in the brain [69,70]. CGRP is a vasodilator, mainly distributed in the central nervous system [71]. In a previous study, endothelin and CGRP were reported as important factors affecting the development of CGD with vertebrobasilar arteriospasm [72]. Fibrinogen also promotes the formation of artherosclerotic plaques [73], and TC accelerates atherosclerosis and causes lipid metabolism disorders [74]. In summary, control of endothelin and CGRP levels improves the prognosis of patients with CGD, and evaluation of fibrinogen and TC levels helps predict CGD progression. erefore, it is recommended to use them as outcomes when conducting further clinical trials of CGD. is research is valuable as the first systematic review to comprehensively evaluate the efficacy and safety of CHMs in treating CGD, to guide clinicians in selecting and prescribing suitable CHMs for specific patterns of CGD based on evidence-based decision-making. Furthermore, it provides knowledge of which treatments will be effective in combination with CHMs. is review may contribute to the development of effective strategies for the treatment and management of an increasing number of patients with CGD due to population ageing. Nonetheless, further high-quality evidence from rigorously conducted clinical studies, preferably conducted outside China, is required to support the clinical recommendations regarding the use of CHMs for CGD. In addition, placebo-controlled RCTs are needed to evaluate the efficacy of CHMs as monotherapy for CGD. Furthermore, experimental studies of the mechanism of action and the dose-response relationship of CHMs are necessary to determine the optimal dose.

Limitations
is review has some limitations. First, some of the major Chinese databases, such as Wanfang and VIP, were not included in the search process. Additionally, grey literature was not considered. us, there is a possibility that relevant studies were omitted. Second, the quality of the included RCTs was generally poor, in particular, because of the high risk of bias due to deviations from intended interventions.
ird, most meta-analyses showed high heterogeneity among studies. Fourth, potential publication bias could not be ruled out because the assessment of publication bias was not conducted in the meta-analyses in which the number of included studies was less than 10, and all RCTs were conducted in China and published in Chinese. Fifth, there is the possibility of attrition bias because few studies presented dropout or withdrawal statistics. Sixth, it is unknown whether the treatment effect of CHMs plus active controls was maintained after completion of the intervention because most studies did not perform follow-up assessments. Finally, the safety of CHMs in patients with CGD remains unknown because few studies clearly reported that there were no adverse events.

Conclusions
Current evidence suggests that CHMs may have the potential to enhance the treatment effect on CGD when combined with other treatments without serious adverse events. As the overall quality of the studies included in this review was generally low, additional high-quality evidence is needed to draw definitive conclusions.