Eligibility criteria

Participants

We will include studies of middle-aged and elderly participants including the following groups: physically healthy individuals, those with hyperlipidaemia, cardiovascular diseases, stroke, hyperlipidaemia combined with hypertension (based on 2010 Chinese guidelines for the management of hypertension18), or hyperlipidaemia combined with diabetes (based on the WHO’s definition of diabetes19).

Interventions

Control group patients will comprise those receiving no treatment, routine treatment or other conventional exercise therapies such as jogging or walking. The intervention group will receive Baduanjin exercises of various duration and frequency based on routine regimens.

Outcome measures

The primary outcome measures will comprise four blood lipid indexes: TC, TG, LDL-C and HDL-C.

The secondary outcome measures will be body mass index, waist-to-hip ratio, Apolipoprotein A1, Apolipoprotein B and adverse events.

Search strategy and identification of studies

We will retrieve articles from the following databases: PubMed, Embase, Cochrane Library, Chinese National Knowledge Infrastructure, Wanfang Data, Chinese BioMedical Database and Chinese Science and Technology Periodicals Database. The publication period will be from inception to March 2017. The search terms will be ‘blood lipid’ OR ‘blood fats’ OR ‘hyperlipidemia’ OR ‘hyperlipaemia’ OR ‘hypercholesterolemia’ OR ‘hypertriglyceridemia’ OR ‘hyperlipoproteinemia’ OR ‘total cholesterol’ OR ‘triglycerides’ OR ‘low-density lipoprotein cholesterol’ OR ‘high-density lipoprotein cholesterol’ OR ‘dyslipidemia’ AND ‘Qigong’ OR ‘Baduanjin’ OR ‘Baduanjin exercise’ OR ‘eight section brocades’. Chinese translations of these terms will be used for the Chinese databases. Table 1 shows the search strategy for PubMed.

Two reviewers (TL and CJ) will independently review the full texts of potential eligible studies. Discrepancies about inclusion in the meta-analysis will be discussed and settled by a third reviewer. A Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow chart will be produced to show the number of articles identified, screened, included and excluded, reasons for exclusion and to ascertain eligible studies. The study selection process will be described in a PRISMA flow chart (http://www.prisma-statement.org) (figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Flow diagram of study selection process. CNKI, Chinese National Knowledge Infrastructure; VIP, Chinese Science and Technology Periodicals Database.

Study selection and data extraction

We will exclude articles for which no data on blood lipid outcomes is presented, relevant information is unavailable or results are duplicated. Two reviewers (TL and CJ) will use the criteria described above to independently review titles and abstracts to select potential references and then to scan all full articles for eligibility. Endnote V.X7 will be used to manage literature and remove duplications.

The reviewers will then extract data about the study population, intervention and outcome measures using a self-developed data extraction form. The data extraction form includes the following items: general information, trial characteristics, participants, interventions and outcomes. Disagreements will be resolved through consultation and if necessary, discrepancies will be discussed with a third author.

Addressing missing data or unclear measurement scales

If necessary, we will contact authors of articles by email or telephone to request missing data or additional information about evaluation scales. If sufficient information cannot be obtained in this way, we will analyse the available data. We will take into account the potential impact of insufficient data on the review results (in the Discussion section).

Risk of bias in included studies

Risk of bias for each included article will be evaluated according to the Cochrane Handbook for Systematic Reviews of Interventions. This recommends the assessment of several sources of bias, including random sequence generation, allocation concealment, blinding of outcome assessments, incomplete outcome data and selective outcome reporting. Bias among participants and investigators will not be considered because the Baduanjin exercises make blinding impossible. We will judge the risks as low, high or unclear (unclear or unknown risk of bias).

Data synthesis and analysis

We will analyse the data using Stata V.13.0. We will calculate weighted mean differences and 95% CIs. The Q and I2 test statistics will be used to assess the heterogeneity of included studies. For the Q statistic, p<0.05 will be considered as indicating significant differences. For the I2 statistic, I²<25% indicates no significant heterogeneity, I²=25%–50% is considered moderate heterogeneity and I²>50% indicates strong heterogeneity. We will use fixed effects models if there is no heterogeneity among studies and random effects models if there is heterogeneity.

Additional analyses

We will perform sensitivity analysis, metaregression and subgroup-stratified analysis based on various study characteristics, such as study type, study location, sample size, study quality, adjustment (or not) for confounders, Baduanjin regimens and other relevant parameters (eg, diet and lifestyle) to explore potential sources of heterogeneity. If data extraction is insufficient, we will create a qualitative synthesis.

Assessment of reporting biases

A funnel plot will be developed to evaluate reporting bias of the included studies. We will use Begg and Egger tests to assess funnel plot symmetry and will interpret values of p<0.1 as showing statistical significance (ie, publication bias).

Quality of evidence

The Grading of Recommendations Assessment, Development and Evaluation approach will be used to evaluate the quality of evidence of the included studies. Reviewers will take into account limitations of the study, inconsistencies, indirect evidence, inaccuracies and publication bias. Four levels of evidence quality will be used: high, moderate, low or very low.