2.1 Search Strategy
A search from the earliest record up to and including June 2019 was carried out using the following electronic databases: PubMed, Scopus and Web of Science. The search strategy employed combined the terms related to the population (e.g. sarcopenia, frailty) with terms for the intervention (e.g. resistance training, strength training). The titles and abstracts of the retrieved articles were individually evaluated to assess the eligibility of studies to be included in the review and meta-analysis. The reviewers were not blinded to the studies’ authors, institutions or journals of publication. Studies with abstracts that did not provide sufficient information according to the inclusion criteria were retrieved for full text evaluation. The authors registered systematic review in PROSPERO, International prospective register of systematic reviews (ref: CRD42019138253) and conducted meta-analysis and systematic review in pursuance of the PRISMA Statement (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) [22].
2.2 Inclusion Criteria
The following inclusion criteria have been applied: (1) research papers released only in English; (2) conference materials, theses, doctoral dissertations and monographs were eliminated; (3) TUG, FFM, HGS, lower limbs strength and gait speed tests were taken into account; (4) RT intervention; (5) older persons (starting at 64 years old); (6) symptoms of pre-frailty, frailty, pre-sarcopenia or sarcopenia must have been diagnosed; (7) measured dynamic changes in muscle strength; and (8) 8 weeks duration.
2.3 Data Extraction
The authors (KT and MS) extracted independently following variables: (1) descriptive data, including the same sample size, age and RT experience; (2) characteristics of the RT training, including training frequency, trial length, number of sets and number of repetitions per set; (3) muscular strength tests used; and (4) the outcomes referred to muscle strength. Any discrepancies were cross-checked and discussed between the authors.
2.4 Methodological Quality
The Downs and Black checklist was used to assess the quality of the methodology [23]. Studies were rated on a scale from “poor quality” ( 11 points) to “good quality” (20–29 points), as in Hacket et. al. [24] and Grgic et al. [25]. Two reviewers (KT and MS) were independently rating these studies. If there are any disputes related to the quality assessment of the research, the third reviewer (EK) helped in obtaining consent.
2.5 Statistical Analysis
The effect sizes (ES) were determined as the standardized mean differences between the resistance training group and the control group. Random-effects meta-analysis was performed using robust variance estimation (RVE) with small-sample corrections which allows for the inclusion of multiple effect sizes from one study, even in the absence of information on the covariance of these effect sizes [26,27]. The RVE method requires a user to specify a value of within-study effect size correlation with the default value of 0.8. A sensitivity analysis that used alternative correlational values found that their choice does not impact the overall results of the meta-analysis. Statistical heterogeneity between trials was examined using the I2 statistics, with values of I2 more than 25%, 50%, and 75% reflecting low, moderate, and high heterogeneity, respectively. All analyses were conducted using packages robumeta (version 2.0) and metafor (version 2.0–0) in R version 3.4.4 (The R Foundation for Statistical Computing, Vienna, Austria).