Effects of tai chi on type 2 diabetes
reviewed to identify potentially eligible studies. The detailed
search strategy for PubMed is shown in Table I. treatment of intervention group should be comparable with
those of control group.
Study selection Outcomes
Titles and abstracts of retrieved literature were independently
screened by 2 reviewers (ZL. Z and RZ. Z) according to pre-
defined eligibility criteria. Full-text reading was performed for
studies whose relevance could not be determined from their
titles and abstracts. During study selection, discrepancies were
discussed until consensus was reached.
More specifically, included studies had the following featu-
res (PICOS criteria: Participants, Interventions, Comparisons,
Outcomes, Study design). FPG, HbA1c, FIN, IR, BMI, TC, systolic blood pressure (SBP)
and diastolic blood pressure (DBP) were used to evaluate the
effects of tai chi on biomedical outcomes. Among these indica-
tors, IR should be evaluated by homeostasis model assessment
(HOMA) version 1 or 2. In addition, QoL and balance (measured
with the 36-item Short-Form (SF-36) scale (29) and duration of
single-leg stance, respectively) as well as BMI, were assessed.
Participants Data were independently extracted by 2 reviewers (ZL. Z and
Y. Z) according to a pre-specified data extraction table, which
comprised the following items: first author, publication year,
region, sample size, characteristics of participants (disease
status, age, FPG at baseline, disease duration). Outcomes at
baseline and follow-up were also recorded. Data on intervention
regimens, including type of tai chi, training time, frequency and
duration, were extracted. A mutual check on extracted data was
performed by 2 reviewers. Disputes and disagreements were
solved by discussion or referral to a third reviewer (KP. L).
Patients diagnosed with T2DM age > 18 years.
Study design
Only randomized controlled trials (RCTs) published in Chinese
or English were included in this meta-analysis. Systematic
reviews, meta-analyses or reference papers were excluded.
Studies were also excluded if outcome data were not available
even though attempts were made to contact the authors.
Interventions
Tai chi, regardless of type, must be performed by participants
in the intervention group of included studies. Studies were
excluded if the intervention was tai chi combined with other
treatments. No restriction was imposed on location (community/
hospital, outdoors/indoors), format (group/individual), intensity,
frequency or duration of treatment.
Comparisons
407
The control group received “usual care or exercise”, “sham
exercise”, or “no intervention”. All the conditions except the
Table I. Search strategy for PubMed
Search Query
#1 Diabetes Mellitus, Noninsulin-Dependent [Title/Abstract] OR
Diabetes Mellitus, Ketosis-Resistant [Title/Abstract] OR Ketosis-
Resistant Diabetes Mellitus [Title/Abstract] OR Diabetes Mellitus,
Non-Insulin-Dependent [Title/Abstract] OR Non-Insulin-Dependent
Diabetes Mellitus [Title/Abstract] OR Diabetes Mellitus, Stable
[Title/Abstract] OR Stable Diabetes Mellitus [Title/Abstract] OR
Diabetes Mellitus, Type II [Title/Abstract] OR NIDDM [Title/
Abstract] OR Diabetes Mellitus, Maturity-Onset [Title/Abstract] OR
Maturity-Onset Diabetes Mellitus [Title/Abstract] OR MODY [Title/
Abstract] OR Diabetes Mellitus, Slow-Onset [Title/Abstract] OR
Slow-Onset Diabetes Mellitus [Title/Abstract] OR Type 2 Diabetes
Mellitus [Title/Abstract] OR Noninsulin-Dependent Diabetes Mellitus
[Title/Abstract] OR Maturity-Onset Diabetes [Title/Abstract] OR
Diabetes, Maturity-Onset [Title/Abstract] OR Type 2 Diabetes[Title/
Abstract] OR Diabetes, Type 2 [Title/Abstract] OR Diabetes Mellitus,
Adult-Onset [Title/Abstract] OR Adult-Onset Diabetes Mellitus
[Title/Abstract]
Diabetes Mellitus, Type 2 [Mesh]
#1 OR #2
Tai-ji [Title/Abstract] OR Tai Chi [Title/Abstract] OR Chi, Tai
[Title/Abstract] OR Tai Ji Quan [Title/Abstract] OR Ji Quan, Tai
[Title/Abstract] OR Quan, Tai Ji [Title/Abstract] OR Taijiquan
[Title/Abstract] OR T’ai Chi [Title/Abstract] OR Tai Chi Chuan
[Title/Abstract]
Tai Ji [Mesh]
#4 OR #5
#3 AND #6
#2
#3
#4
#5
#6
#7
Data extraction
Risk of bias assessment
Two reviewers (ZL. Z and RZ. Z) independently assessed the
risk of bias in individual studies by using the Physiotherapy
Evidence Database (PEDro) scale (30), which was commonly
used to judge the quality of RCTs regarding physical therapy.
Studies with a PEDro score over 4 points were regarded as
high quality. Any discrepancies were resolved through discus-
sion. If consensus was not reached, a third reviewer (RS. L)
was consulted.
Data synthesis
Pooled analyses were performed using Stata 12.0 software.
Mean changes from baseline to follow-up were synthesized for
continuous data, and 3 aspects of QoL were evaluated: physical
function, bodily pain and social function. Mean difference (MD)
or standardized mean difference (SMD) with 95% confidence
interval (95% CI) were summarized for continuous outcomes
evaluated, respectively, by identical measurement or different
measurements. Heterogeneity across studies was assessed by I 2
statistic. A fixed-effects pooled analysis was conducted if no or
low heterogeneity (I 2 < 50%) existed, and when there was high
heterogeneity among studies (I 2 > 50%), a random-effects meta-
analysis was performed. Sensitivity analysis was performed to
explore the impact of methodological quality of included studies
on results by removing low-quality studies (PEDro score ≤4).
In addition, data were synthesized after excluding studies in
turn to test the stability of results, if necessary. Funnel plot and
Egger’s test (31) were used to assess the possibility of publica-
tion bias, if appropriate.
RESULTS
Search results
A total of 487 articles were retrieved according to the
search strategy, and no articles were obtained from
J Rehabil Med 51, 2019