Research outcomes of the study
Character 1 presents the PRISMA 2020 flowchart illustrating the process of inclusion in the study. A total of 1469 studies were identified through the search conducted in 7 electronic databases. Using EndNote X9 (Clarivate Analytics, Philadelphia, PA, USA), 813 duplicate articles were removed, leaving 656 unique articles. After analyzing the titles and abstracts, 551 articles were excluded because they did not meet the inclusion criteria. Full texts were not available for 3 articles. Eligible full texts of 102 articles were additionally assessed for eligibility. 6 additional articles are included by manual retrieval. A total of 27 studies (31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57) were included in the meta-analysis: 18 studies focused on exercise training, 5 studies examined the effects of vitamin D, and 4 studies investigated the impact of protein supplementation (one study overlapped with exercise).
PRISMA flow chart for the process of selecting included studies
Quality assessment of included studies
Character 2 presents the results of the risk of bias assessment using RoB 2.0 for the 27 included studies (31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57). Among the 27 studies, 26 were judged to have some concerns about bias, while one study was judged to have a low risk of bias. Judgments of some concern for bias were primarily caused by a lack of specific information regarding the generation of the randomization sequence, a lack of clarity regarding blinding procedures, and insufficient information about the randomization allocation process. Moreover, the data from the studies did not indicate the existence of a predetermined design or registration of the clinical trial, which affected the analysis. Overall, the risk of bias for the included studies ranged from low risk to some concern. Two independent reviewers assigned scores to assess the level of consistency of each domain, achieving a score of 96.4% agreement.
Risk of bias summary of individual studies
ASSESSMENT analysis
The outcomes of exercise on lean body mass and knee extension strength, as well as the outcome of vitamin D on handgrip strength, received a moderate level of certainty of evidence. However, the effectiveness of the handgrip strength exercise was assessed as a low level of safety. This downgrading was attributed to uncertainties in the effect estimate, suggesting that the true effect could be significantly different from the estimated effect. The outcomes of vitamin D on handgrip strength and protein on lean body mass gathered a low level of certainty of evidence. This reduction is the result of significant risk of bias, arising from inadequate reporting of allocation concealment methods, and significant imprecision, as evidenced by a confidence interval exceeding the no-effect limit (Table 1).
Characteristics of the study
Table 2 provides a summary of the characteristics of the included studies. Among the 27 studies (31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57), three were conducted in Thailand, Australia, Canada, Spain, and Brazil (55.5%), while two studies were conducted in the United States, China, and Iran (22.2%). Additionally, one survey was conducted in Finland, Korea, Italy, Denmark, Japan and Egypt. Among the included studies, a total of 18 focused on exercise training. A total of 1327 participants took part in these studies. Exercise modalities varied, with 10 studies focusing on resistance training (including resistance training, aquatic exercise, weight training, balance training, functional training, bilateral and unilateral strength training), five studies on aerobic training (including circuit aerobic exercise steps, sprint interval training, multicomponent training, tai chi ball exercise and walking exercise), and four studies on other forms of exercise (including whole body vibration and combined exercise). A total of five studies were included for vitamin D supplementation. 477 participants took part in these studies. Various forms of vitamin D have been used, including vitamin D, vitamin D2, and vitamin D3. Regarding protein supplementation, five studies were included, one of which was also included in the exercise category. 185 participants took part in these studies. Various forms of protein have been used, including soy protein, branched chain amino acids, and isoflavones.
Meta-analytic synthesis of results
For the timing test, two studies examined the effect of exercise (41, 49), while two studies investigated the effect of vitamin D supplementation (30, 44). Regarding muscle mass, two studies reported the effect of vitamin D supplementation (37, 53). One study evaluated the effect of protein supplementation on handgrip strength (33). However, no study provided data on the effect of protein supplementation on physical functional performance, thus preventing a meta-analysis in this regard due to insufficient data.
Effects of exercise on measures of lean body mass
Fourteen studies (17 datasets) (35,36,37, 39, 40, 43, 44, 46,47,48, 51, 52, 55, 56) with a total of 1053 participants were included in a meta-analysis to assess the effects of exercise on lean body mass. Meta-analysis showed homogeneity among studies (Q= 15.507, p= 0.488, I2= 0%) (Fig. 3– At the table 3). Pooled data analysis showed a significant overall improvement in lean body mass due to exercise (SMD = 0.232, 95% CI: 0.097, 0.366, p= 0.001). A funnel plot indicated no apparent asymmetry, and Egger’s linear regression test showed no evidence of publication bias in a meta-analysis of exercise training on lean body mass (p= 0.073). Subgroup analyzes were performed based on the type of exercise (resistance exercises, aerobic exercises and other exercises), duration of exercise (≤ 12 weeks and > 12 weeks) and frequency of exercise (20-45 min three sessions per week and 60-90 min three sessions per week) . Detailed results can be found in the table 2.
A meta-analysis of the effects of exercise training on lean body mass, handgrip strength, and knee extension strength between two groups
Effects of exercise on measuring hand grip strength
Two studies (52, 57) (three data sets) with a total of 60 participants investigated the effects of exercise on hand grip strength. A meta-analysis showed significant but relatively low heterogeneity among studies (Q= 3.635, p= 0.162, I2= 45%). The overall pooled results indicated a significant improvement in handgrip strength due to exercise (SMD = 0.901, 95% CI: 0.362, 1.441, p= .001) (Fig. 3-B, Table 3). Egger’s linear regression test revealed no indication of publication bias in a meta-analysis of handgrip strength training (p= 0.073).
Effects of exercise on measuring knee extension strength
Six studies (31, 35, 41, 44, 46, 47) (nine datasets) with a total of 250 participants investigated the effects of exercise on knee extension strength. Meta-analysis initially showed significant high heterogeneity among studies (Q = 34.114, p< 0.001, I2= 77%). However, sensitivity analyzes were performed by removing the study (31) with inadequate or unclear data, and the results remained consistent, which indicates the reliability and robustness of the findings. After sensitivity analysis, meta-analysis showed significant but relatively low heterogeneity among studies (I2= 33%). Based on five studies (35, 41, 44, 46, 47) (eight datasets) with 187 participants, the overall pooled results showed a significant improvement in knee extensor strength due to exercise (SMD = 0.698, 95% THIS ONE: 0.384, 1.013, p< 0.001) (Fig. 3-C, Table 3). Egger’s linear regression test revealed no indication of publication bias in a meta-analysis of knee extension strength training (p= 0.446).
Effects of vitamin D on measuring hand grip strength
A meta-analysis of handgrip strength included 4 studies (31, 38, 45, 54) with 425 participants taking vitamin D. The meta-analysis showed significant very low heterogeneity among studies (Q= 0.422, p= 0.936, I2= 0%). The overall pooled results showed a significant improvement in hand grip strength (SMD = 0.267, 95% THIS ONE: 0.065, 0.469, p= .009) (Table 3). Egger’s linear regression test revealed no indication of publication bias in a meta-analysis of vitamin D supplementation on handgrip strength (p= 0.958).
Effects of vitamin D on measures of knee extension strength
A meta-analysis of knee extension strength, based on 3 studies (31, 32, 45) which included 216 participants receiving vitamin D supplements, showed significant and very low heterogeneity between studies (Q= 2.407, p= 0.300, I2= 17%). The overall pooled results did not show a significant effect on knee extension strength with the addition of vitamin D (table 3). Egger’s linear regression test showed no indication of publication bias in a meta-analysis of vitamin D supplementation on knee extension strength (p= 0.247).
Effects of protein on measures of lean body mass
A meta-analysis of lean body mass included 3 studies (49, 53, 56) (4 datasets) with 165 participants with protein supplementation. The meta-analysis showed significant very low heterogeneity among studies (Q= 0.426, p= 0.935, I2= 0%). Overall pooled results without a significant effect were observed in lean body mass (table 3). Egger’s linear regression test revealed no indication of publication bias in a meta-analysis of protein supplementation on lean body mass (p= 0.198).
