International Journal of Molecular Medical Science, 2025, Vol.15, No.4, 185-194 http://medscipublisher.com/index.php/ijmms 187 and linking them with logical words such as "and" and "or". Additional search methods may include manually flipping through the reference list, relevant journals, and unpublished materials. This is done to reduce the phenomenon of only publishing good results and to ensure that as complete evidence as possible is found (Zhang and Vijayakumar, 2016; Gurevitch et al., 2018; Jeyaraj and Dwivedi, 2020). To enhance transparency and enable others to follow suit, it is necessary to record in detail the search process, including which databases were used, which search terms were entered, the time range of the search, and whether there were any language restrictions or whether it was officially published. Using both electronic search and manual search methods simultaneously can lead to more research and also help identify potential gaps in the literature (Gurevitch et al., 2018; Jeyaraj and Dwivedi, 2020). 3.2 Clearly define which studies are included and which are not Establishing clear criteria to determine which studies can be included (inclusion criteria) and which should be excluded (exclusion criteria) is crucial for ensuring that comprehensive analysis addresses core issues and maintains methodological rigor. Inclusion criteria usually specify the type of study (for example, it must be a randomized controlled trial), characteristics of the study subjects (elderly), details of the intervention measures (vitamin D supplementation, alone or with calcium), and the outcome of concern (incidence of fractures). Exclusion criteria may include studies with incomplete data, incomparable intervention measures, or duplicate datasets (Field and Gillett, 2010; Jeyaraj and Dwivedi, 2020). Careful implementation of these standards can ensure that the selected research collection represents past research and reduce bias. It is suggested to adopt some strategies, such as eliminating studies with overlapping data, properly handling multiple effect values (effect sizes) produced by the same study, and separately recording different datasets when necessary, to ensure the independence of each piece of data and ultimately make the results of the comprehensive analysis more reliable (Jeyaraj and Dwivedi, 2020). 3.3 Data extraction, quality evaluation and statistical analysis methods Data extraction refers to the systematic collection of important information from each included study, such as the number of participants in the study (sample size), the characteristics of the participants, the specific content of the intervention, the method of result measurement, and the effect value (effect size). This process is usually completed independently by two people respectively to reduce errors and inconsistencies. The use of quality assessment tools, such as the Cochrane Risk of Bias tool, to evaluate the rigor (quality) of the methods in each study is also helpful for subsequent sensitivity analysis (Zhang and Vijayakumar, 2016; Mikolajewicz and Komarova, 2019; Jeyaraj and Dwivedi, 2020). In the statistical part of comprehensive analysis, fixed-effect models or random-effect models are usually employed to combine the effect values of various studies. The specific choice depends on the magnitude of the differences among different studies. More in-depth methods, such as meta-regression and group analysis, can be used to explore where these differences come from. Sensitivity analysis and checking for publication bias (such as drawing funnel plots) will also be conducted to ensure that the research results are tenable and reliable (Zhang and Vijayakumar, 2016; Gurevitch et al., 2018; Balduzzi et al., 2019; Mikolajewicz and Komarova, 2019). 4 The Overall Effect Of Vitamin d Supplementation on the Risk of Fractures in the Elderly 4.1 Evaluation of the overall effect size and statistical significance Recent large-scale summary analyses (meta-analyses) and high-quality studies (randomized controlled trials) have consistently found that healthy elderly people taking vitamin D alone do not significantly reduce the total number of fractures. For instance, a summary analysis involving more than 71 000 people showed that the total incidence of fractures was similar between those taking vitamin D and those taking a placebo (RR 1.03; 95% ci 0.93-1.14; p=0.56) (LeBoff et al., 2022; De Souza et al., 2024; De Souza et al., 2024). Similarly, another summary analysis also reported that among elderly people living at home, there was no significant association between vitamin D intake and the risk of hip, spinal or other (non-vertebral) fractures (Figure 1) (Zhao et al., 2017; Khatri et al., 2023).
RkJQdWJsaXNoZXIy MjQ4ODYzNA==