Clinical Outcomes and Safety of Stem Cell Therapy for Diabetic Foot Ulcers: A Cross-Sectional Analysis of Overlapping Meta-Analyses Authors: Ying Du, Zhendong Xu, Tianlin Wen, Xinxin Meng, Bolun Zhang, Qiantai Gao, Xiyan Zhao, Zhiwei Jia Published: 2026 ABSTRACT AbstractPurposeThe care of diabetic foot ulcers (DFU) continues to be a notable clinical challenge. The efficacy and safety of stem cell therapies versus conventional treatments for DFU are debated. Cross-sectional analyses of meta-analyses enable structured comparisons of overlapping reviews, facilitate the identification of inconsistencies, and support evidence-based decision-making when conclusions differ. This study aimed to conduct a cross-sectional analysis of existing meta-analyses comparing stem cell and conventional treatments for DFU outcomes to identify the highest-quality available evidence and clarify areas of consistency and discrepancy.MethodsWe systematically searched for meta-analyses that investigated stem cell therapies versus conventional treatments for DFU. The complete electronic search strategies for all databases are provided in accordance with PRISMA 2020 recommendations. Methodological quality was assessed using the Oxford Levels of Evidence and original AMSTAR instrument. Data abstraction and quality evaluation were performed independently, and the Jadad decision algorithm was applied to determine the meta-analysis providing the most methodologically robust evidence.ResultsNine meta-analyses were included in this study. Five were categorized as Level II evidence, while the remaining four were classified as Level III evidence. The AMSTAR scores ranged from 7 to 10 (median 9). The Jadad algorithm identified the most robust meta-analysis consistently showing improved ulcer healing rates (P < 0.0001), greater pain-free walking distance (P < 0.00001), lower amputation rates (P < 0.0001) compared with conventional treatments. Furthermore, the incidence of adverse events in the stem cell groups was lower compared to conventional treatments (P < 0.001).ConclusionThe study indicated that stem cell therapy appeared to be a promising therapeutic option for improving clinical outcomes. However, conclusions should be interpreted with caution due to overlapping primary studies, heterogeneity, and the secondary nature of this cross-sectional evaluation. Moreover, no clear evidence of increased serious adverse events was identified in the highest-quality meta-analysis.Keywords: diabetic foot ulcer, stem cell therapy, conventional treatment, meta-analysis, Jadad decision algorithm Introduction Diabetes mellitus stands as a prominent chronic metabolic disorder, with 537 million individuals afflicted worldwide in 2021. Forecasts suggest this population will expand to 853 million by 2050.1 Diabetic foot ulcer (DFU) is a severe and debilitating complication of diabetes mellitus, characterized by neuropathy, ischemia, and infection, often leading to ulceration, infection, and ultimately, a high risk of amputation.2 Affecting a significant proportion of diabetic patients, DFU places an enormous burden on individuals, healthcare systems, and national economies. Previous research has established annual health care costs for DFU patients at approximately $28,000 for Medicare beneficiaries.3,4 The estimated global prevalence of DFU is approximately 6.3%, with lifetime incidence as high as 34% among diabetic patients, underscoring the urgent need for effective therapeutic interventions.5 The profound impact on patient morbidity, mortality, and quality of life, alongside the substantial healthcare costs associated with DFU management, necessitates the development and evaluation of advanced and cost-effective treatment strategies.6 Historically, conventional management, encompassing aggressive wound care, offloading, infection control, glycemic optimization, and revascularization, has formed the cornerstone of DFU treatment.7 While these traditional approaches are crucial, they often yield suboptimal outcomes, particularly in complex or refractory cases, leaving a significant unmet clinical need.8 In recent decades, stem cell therapies have emerged as a promising alternative, leveraging their regenerative, anti-inflammatory, and angiogenic properties to promote tissue repair and improve healing in chronic wounds.9 Despite the growing body of research, the comparative effectiveness and potential risk profile of stem cell therapies versus established conventional treatments for DFU remains a subject of considerable debate and conflicting evidence.10 Numerous systematic reviews and meta-analyses have sought to synthesize the available evidence on these treatment modalities.10–18 However, these studies often present inconsistent findings, with some reporting superior outcomes for stem cell interventions, while others find no significant differences. For example, in terms of healing time, Elainein MAA et al17 reported a significant reduction in healing time with stem cell treatment (P = 0.04), while Elsharkawi M et al18 found no distinction in healing time when comparing stem cell treatment to conventional therapies (P = 0.0673). These discrepancies underscore the necessity for a more rigorous and comprehensive evaluation of the existing evidence. A cross-sectional analysis of overlapping meta-analyses offers a powerful methodological framework to compare discordant evidence derived from multiple reviews.19 In medical research, this analytical approach provides several advantages, including the systematic identification of inconsistencies across published meta-analyses, evaluation of methodological quality, and selection of the most reliable evidence when conclusions differ.19–21 This design is particularly useful in areas where several meta-analyses reach conflicting conclusions, as it enables direct comparison of methodological rigor, study overlap, heterogeneity, and evidentiary strength. Such analyses can enhance the transparency of evidence synthesis and support more informed clinical decision-making.20–22 By applying established appraisal tools, such as the Jadad decision algorithm, this approach facilitates identification of the most reliable and methodologically sound synthesis among competing analyses. It has been successfully applied in various medical fields.20–22 There is currently no reported cross-sectional analysis of overlapping meta-analyses specifically comparing stem cell versus conventional treatments for DFU. Therefore, this study aimed to undertake such an analysis to systematically review and critically appraise the available evidence, clarify the relative efficacy and safety profiles of stem cell therapies compared to conventional management in DFU, and ultimately provide meaningful guidance for clinical practice and future research directions. Materials and Methods Search Strategy The design was informed by prior published work,19–22 and structured according to the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).23 A comprehensive search of the PubMed, Embase, and Cochrane Library databases was conducted to identify relevant meta-analyses comparing stem cell therapies and conventional treatments in managing DFU. The search was performed from July 15, 2024 to January 15, 2025. The search strategy incorporated the following terms: diabetic foot, diabetic ulcer, diabetic wound, stem cell, progenitor cell, systematic review, and meta-analysis. The complete electronic search strategies for all databases are provided in Supplementary Table S1 in accordance with PRISMA 2020 recommendations. Furthermore, the reference lists of included papers were manually reviewed to locate any further relevant publications. Eligibility Criteria Studies were deemed eligible for inclusion based on these requirements: (1) meta-analyses written in English; (2) involved DFU patients treated with stem cells; (3) reported at least one predefined clinical outcome. The primary outcomes were complete ulcer healing rate and amputation rate (major or minor). Secondary outcomes included pain-free walking distance and adverse events. Exclusion criteria included: (1) narrative reviews, meeting abstracts, editorials, case reports, and correspondence articles; (2) lacked sufficient information and methodological descriptions; (3) non-English meta-analysis. Study Selection Study selection proceeded in stages, managed by two independent reviewers. The initial Phase Involved assessing titles and abstracts for eligibility. Subsequently, full texts of potentially relevant publications were obtained for a thorough assessment. When reviewers differed on study eligibility, consensus was initially sought through discussion, a third reviewer resolved inclusion disputes where consensus failed. Data Extraction Data abstraction was performed independently by two reviewers for each included study, employing a standardized extraction sheet. The extracted data included: first author, year of publication, journal name, date of the last literature search, number of included randomized controlled trials (RCTs), search databases, language restrictions and primary study design. Conflicts between the two reviewers were settled through discussion. In cases where agreement could not be attained, a third reviewer was engaged for resolution. Quality Assessment The methodological quality of the included meta-analyses were assessed using the Oxford Levels of Evidence and the original Assessment of Multiple Systematic Reviews (AMSTAR) tool,24–26 which is recognized widely for its established reliability and validity as a tool for methodological appraisal. The quality evaluation was performed separately by two investigators. Conflicts in assessment were settled following discussion with a third reviewer. Risk of Bias and Evidence Quality Assessment To summarize the methodological quality and potential bias of the included meta-analyses, a structured Risk of Bias (ROB) summary table was constructed.24 The table included primary study type, overlap of primary studies, overall risk of bias of the primary studies as reported in the original reviews, and GRADE certainty of evidence when available.24 Relevant information was extracted directly from the included meta-analyses, and previously reported assessments using tools such as the Cochrane Risk of Bias tool or modified Jadad scale were summarized to reflect the overall quality of the evidence.27 In addition, the degree of overlap among primary studies across the included meta-analyses was quantified using the Corrected Covered Area (CCA). The CCA was calculated using the formula proposed by Pieper et al28 where CCA = (N-r)/(rc-r), with N representing the total number of included studies across meta-analyses, r the number of unique primary studies, and c the number of meta-analyses. According to established thresholds, CCA values of 0–5% indicate slight overlap, 6–10% moderate overlap, 11–15% high overlap, and values above 15% indicate very high overlap. Application of Jadad Decision Algorithm The Jadad decision algorithm was used to identify the highest-quality evidence from the meta-analyses under consideration.29 The analysis addressed heterogeneity across several key areas: the specific clinical questions posed, criteria for study eligibility, procedures for data abstraction, approaches to quality appraisal, techniques for data pooling, and the statistical methods used for synthesis.14 Utilizing this algorithm, the most robust evidence was identified, thereby enabling structured comparison of discordant meta-analyses and facilitating identification of the most methodologically reliable synthesis.14 Three independent reviewers applied the algorithm, and through discussions, they reached a consensus on the meta-analysis offering the most dependable findings. Results Literature Search Our initial systematic exploration across major academic databases yielded a total of 216 titles and abstracts. Following the screening process, nine meta-analyses10–18 were ultimately chosen for inclusion, adhering to our predefined eligibility criteria for this cross-sectional investigation (Figure 1). A detailed summary of the core characteristics of these selected meta-analyses is provided in Table 1. These studies were published from 2017 to 2025, and each integrated between 4 to 20 individual primary research studies.10–18 Altogether, 92 RCTs were included in these analyses, the details of each RCT from these meta-analyses are itemized in Table 2. The flow chart of study selection. Table 1Characteristics of the Included Meta-AnalysesFirst AuthorDate of PublicationJournalDate ofLast Literature SearchNo. of Included StudiesNo. of Included RCTsGuo J112017Stem Cell Research & TherapyJanuary 7, 201764Shu X122018Endocrine JournalDecember 20, 201676Dai J132020Stem Cells InternationalMay 30, 201988Sun Y102022Stem Cell Research & TherapyApril 10, 20221411Huang L142023Advances in Skin & Wound CareNovember 30, 20211616Elsharkawi M182023International Journal of Lower Extremity WoundsNovember, 202144Mudgal SK162024International Journal of Lower Extremity WoundsSeptember 30, 20232018Tong L152025Frontiers in GeneticsOctober, 20242419Elainein MAA172025Diabetology & Metabolic SyndromeDecember, 202366 Characteristics of the Included Meta-Analyses Table 2Primary Studies Included in Meta-AnalysesFirst Author (Year)Guo J(2017)11Shu X(2018)12Dai J (2020)13Sun Y(2022)10Huang L (2023)14Elsharkawi M (2023)18Mudgal SK(2024)16Tong L(2025)15Elainein MAA (2025)17Huang P (2005)30+++++Zhang Z (2007)31+Debin L (2008)32++++Dash N (2009)33++++Han S (2010)34+++++++Lu D (2011)35+++++++Jain P (2011)36+++++Ozturk A(2012)37+++++Kirana S (2012)38+++++++You H (2012)39+++Mohammadzadeh L (2013)40++++++Lavery L (2014)41+++He YQ (2014)42++Albehairy A (2016)43+Moon K (2019)44++++Lonardi R (2019)45++++Park K (2018)46+++Qin H (2016)47++++Smith O (2020)48+++++Liu PY (2023)49+Uzun E (2021)50+++++Tettelbach W (2019)51+Tettelbach W (2022)52+Tanios E (2021)53+Thamm OC (2023)54+Wang XF (2024)55+Wang GY (2024)56+ Primary Studies Included in Meta-Analyses Search Methodology The methodological approaches to literature searching varied among the included meta-analyses. Six studies restricted their scope exclusively to English-language publications,10–12,16–18 whereas the remaining two broadened their search to encompass both English and non-English sources.13,14 One study did not explicitly report any language restrictions.15 While PubMed consistently served as the primary search platform across most analyses, there was notable diversity in the supplementary databases utilized, which included Embase, Web of Science, Medline, Ovid, and the Cochrane Library. Comprehensive specifics regarding the individual search strategies and methodologies employed are delineated in Table 3. Table 3Search Methodology of the Included StudiesFirst Author (Year)Restriction of Publication LanguageRestriction of Publication StatusSearch DatabasePubMedEmbaseCochrane LibraryScopusBIOSISOthersGuo J (2017)11YesNA+++––+Shu X (2018)12YesYes+++––+Dai J (2020)13NoYes+++–++Sun Y (2022)10YesNA+++––+Huang L (2023)14NoYes+++––+Elsharkawi M (2023)18YesNA++––––Mudgal SK (2024)16YesYes++++–+Tong L (2025)15NAYes++++–+Elainein MAA (2025)17YesYes+–++–+ Search Methodology of the Included Studies Methodological Quality Consistent with the Oxford Levels of Evidence framework, of the nine meta-analyses included in our review, five were categorized as Level II evidence, while the remaining four were classified as Level III evidence. The software platform RevMan emerged as the predominant tool for statistical analysis across these studies (Table 4). 10–18 The AMSTAR scores, indicative of methodological quality, ranged from 7 to 10, with a median value of 9. Notably, the meta-analysis conducted by Huang L et al14 received the highest quality assessment score (Table 5). Table 4Methodological Information for the Included StudiesFirst Author (Year)Design of Included StudiesLevel of EvidenceSoftwareGRADE UseSensitivity AnalysisGuo J (2017)11RCTsLevel IIRevManNoYesShu X (2018)12CCTs, RCTsLevel IIIRevManYesYesDai J (2020)13RCTsLevel IIRevMan, STATANoNoSun Y (2022)10CCTs, RCTsLevel IIIRevManNoYesHuang L (2023)14RCTsLevel IIRevMan, STATAYesNoElsharkawi M (2023)18RCTsLevel IIStatsDirectNoNoMudgal SK (2024)16CCTs, RCTsLevel IIIRevMan, Open MetaYesYesTong L (2025)15CCTs, RCTsLevel IIISTATANoYesElainein MAA (2025)17RCTsLevel IIRevManYesYesAbbreviations: CCT, controlled clinical trial; RCT, randomized clinical trial. Methodological Information for the Included Studies Abbreviations: CCT, controlled clinical trial; RCT, randomized clinical trial. Table 5AMSTAR Scores for the Included StudiesItemsGuo J (2017)11Shu X (2018)12Dai J (2020)13Sun Y (2022)10Huang L (2023)14Elsharkawi M (2023)18Mudgal SK (2024)16Tong L (2025)15Elainein MAA (2025)171. Was an a priori design provided?0000101112. Was there duplicate study selection and data extraction?1111111113. Was a comprehensive literature search performed?1111111114. Was the status of publication (ie grey literature) used as an inclusion criterion?0010000005. Was a list of studies (included and excluded) provided?1110101116. Were the characteristics of the included studies provided?1111111117. Was the scientific quality of the included studies assessed and documented?1111111118. Was the scientific quality of the included studies used appropriately in formulating conclusions?1111111119. Were the methods used to combine the findings of studies appropriate?11011010110. Was the likelihood of publication bias assessed?10111111011. Was the conflict of interest stated?111111111Total scores99981071099 AMSTAR Scores for the Included Studies Risk of Bias and Overlap Assessment A summary ROB table was constructed for the included meta-analyses, incorporating methodological quality, primary study design, overlap of included primary studies, overall risk of bias of the primary studies, and certainty of evidence where available (Supplementary Table S2). Among the included meta-analyses, only Mudgal et al and Elainein et al explicitly reported GRADE certainty of evidence, whereas the remaining reviews did not provide formal GRADE assessment. The degree of overlap among primary studies was substantial across reviews, consistent with the high CCA (30.1%), indicating multiple meta-analyses relied on a largely similar pool of primary RCTs. Heterogeneity Assessment The I2 statistic was universally applied across all selected meta-analyses as the primary metric for quantifying the degree of statistical heterogeneity present among the included primary studies (Table 6). Several outcomes demonstrated substantial heterogeneity (I2 > 75%), which may reflect variations in study populations, intervention protocols, and definitions of conventional treatment across the included trials. Furthermore, sensitivity analyses, specifically tailored to evaluate the impact of methodological quality, were performed in six of the studies (Table 4).10–12,15–17 Table 6I2 Statistic Values for Each Variable in Meta-AnalysesItemsGuo J (2017)11Shu X (2018)12Dai J (2020)13Sun Y (2022)10Huang L (2023)14Elsharkawi M (2023)18Mudgal SK (2024)16Tong L (2025)15Elainein MAA (2025)17Healing time75.2%99%96%Ulcer size area35%26%ABI33%53%0%31.1%TcPO295%65%Healing rates4%12%10%0%41%19%0%Amputation rates0%0%0%0%15%0%New vessels formation65%80%Pain-free walking distance0%0%0%Rest pain score0%Adverse events0%22% I2 Statistic Values for Each Variable in Meta-Analyses Results of Jadad Decision Algorithm The Jadad decision algorithm was systematically applied to discern the most robust evidence derived from the nine selected meta-analyses. The summarized findings from these meta-analyses are visually presented in Figure 2. As illustrated in Figure 3, meta-analyses were sequentially assessed according to similarity of research questions, degree of primary trial overlap, methodological quality, language restrictions, and consistency of inclusion criteria. According to the Jadad algorithm, the selection of optimal evidence is predicated upon several key criteria, including but not limited to publication status, the methodological robustness of the underlying primary studies, and any language-based restrictions. Upon thorough final evaluation, the meta-analysis by Huang L et al14 was identified as possessing the highest methodological quality (Figure 3), which comprehensively synthesized the included studies, reported notable improvements in ulcer healing rates (P < 0.0001) and pain-free walking distance (P < 0.00001), alongside significant reductions in amputation rates (P < 0.0001). These results indicate an improvement in DFU patients’ physical condition and overall quality of life following stem cell therapy.14 However, the best stem cell type and treatment regimen cannot be determined. Results of included meta-analyses. The flow chart of Jadad decision algorithm. Discussion DFU represents a significant global health concern, frequently leading to amputations, and even fatal outcomes if not managed effectively.57,58 The substantial psychological and economic burdens associated with DFU highlight the urgent need for evidence-based and economically sustainable interventions.59,60 While conventional treatments have long been the cornerstone of DFU management, the emergence of stem cell therapy has spurred debate over its comparative efficacy, with meta-analyses yielding conflicting conclusions.60 Cross-sectional analyses of overlapping meta-analyses provide a robust framework for synthesizing evidence by examining the consistency and quality of various systematic reviews on a particular subject.61 This approach helps identify the most reliable evidence and resolve discrepancies among studies.61 Our study represents the first cross-sectional evaluation of overlapping meta-analyses comparing stem cell therapy with conventional interventions for treating DFU. To better understand the sources of these inconsistencies, we examined the characteristics of the included meta-analyses. Most of the included meta-analyses, though conducted during overlapping timeframes and employing broadly similar search strategies, exhibited significant heterogeneity in their selection of primary trials and conclusions regarding DFU treatment. These discrepancies highlight the challenges of synthesizing evidence when studies diverge in methodology and interpretation.10–18 To address these issues, our research utilized Jadad’s decision algorithm,62 a recognized framework designed to pinpoint the highest-quality evidence from conflicting systematic reviews. This method prioritizes methodological rigor, transparency in selection criteria, and analytical coherence, enabling researchers to systematically navigate overlapping but contradictory meta-analyses.10–18 By adopting this approach, we aimed to ensure that clinical recommendations for DFU management were grounded in more robust available evidence, consequently helping to reduce any potential bias arising from methodological variability across studies.29 The Jadad decision algorithm identified the meta-analysis by Huang L et al14 as providing the most robust evidence among the included studies, revealing notable benefits of stem cell therapy compared to conventional interventions for several outcomes, including reducing amputation rates, improving ulcer healing rates and pain-free walking distance.14 This aligns with previous studies suggesting that stem cell therapy can enhance revascularization, reduce inflammation, and promote tissue regeneration, all of which are essential in managing DFU complications.10 To better interpret these findings, it is important to consider the characteristics of both stem cell therapy and conventional management used in the included trials. In the included studies, conventional therapy generally referred to standard diabetic foot management, including glycemic control, infection control, wound debridement, dressing care, pressure off-loading, and revascularization when necessary.7,59,60 These approaches aim to control infection and improve tissue perfusion but often target only limited aspects of the complex pathophysiology of DFU.2,7 Stem cell therapy in the included trials was derived from multiple sources, including bone marrow, peripheral blood, adipose tissue, and umbilical cord.11–14,30,37 Differences in cell sources, delivery routes such as intramuscular or intra-arterial administration, and treatment protocols may partly contribute to variations in healing outcomes across studies.10–14,16 Regarding safety, the lower incidence of adverse events reported in stem cell groups may seem counterintuitive, considering that conventional therapy is typically non-invasive. However, most reported adverse events were ulcer-related complications such as infection or progression to amputation, rather than procedure-related side effects.10,11,14 By promoting angiogenesis and tissue repair, stem cell therapy may accelerate wound healing and thereby reduce these ulcer-related complications.9,30,37,52 Importantly, the therapeutic potential of stem cells has also been demonstrated in several non-DFU clinical settings, including critical limb ischemia, chronic lower-extremity ischemic wounds, and peripheral vascular diseases. Previous studies have shown that stem cell transplantation can improve limb perfusion, stimulate angiogenesis, and accelerate tissue repair in ischemic tissues.30,32,37 These regenerative effects are largely attributed to paracrine signaling, immunomodulation, and enhanced neovascularization, mechanisms that are also highly relevant to the pathophysiology of diabetic foot ulcers.9,30,32,37,52 The consistency of these biological effects across different ischemic disease models supports the mechanistic plausibility of stem cell therapy as a therapeutic strategy for DFU. Moreover, the use of stem cell treatment was not associated with an increased incidence of serious adverse events in the selected highest-quality meta-analysis. However, it should be noted that these findings are based on secondary syntheses of randomized controlled trials included in meta-analyses, rather than on newly conducted primary studies.14 In addition, considerable variability across studies in stem cell sources, processing techniques, dosages, and routes of administration may have contributed to heterogeneity in treatment effects. Although the available evidence suggests potential clinical feasibility and revascularization benefits, the optimal cell type and delivery strategy remain uncertain and warrant further well-designed investigations. These findings highlight the potential role of stem cell therapy in DFU management.10,16 Based on the synthesized evidence, stem cell therapy may be considered in selected patients with DFU, particularly when conventional management yields suboptimal results, rather than being universally recommended as a first-line replacement.14 To better understand the robustness of these findings, the methodological characteristics of the available evidence should also be considered. To ensure we drew conclusions from the most reliable evidence, we employed the Jadad decision algorithm to identify the most robust data based on publication status, methodological quality, and statistical analysis.14 Overall, the included meta-analyses demonstrated moderate to high methodological quality based on AMSTAR scores. Sensitivity analyses within some of these studies further illustrated how changes in methodology can influence outcomes.10–12,15–17 The ROB summary provides useful methodological context for interpreting the findings of this overview. However, a substantial overlap of primary studies was observed (30.1%), indicating that several reviews were based on largely shared pools of randomized controlled trials. Such overlap is common in fields with a limited number of available trials and may partly explain the similarity of findings across reviews. In addition, the certainty of evidence assessed using the GRADE framework varied across outcomes, suggesting that further well-designed randomized controlled trials would help strengthen the current evidence base. Our results carry significant implications for clinical practice and future research. Clinically, this study suggests that stem cell therapy may be considered as a potential therapeutic option for DFU management. For clinicians, the consistent superiority of stem cell therapy across various critical outcomes suggests that it offers a valuable therapeutic option, particularly in cases where conventional treatments are insufficient.14 This approach aligns with efforts to improve patient quality of life and potentially reducing the substantial economic burden associated with DFU, as evidenced by the high rates of morbidity and mortality.2,3 For future research, our analysis underscores the need for more rigorous and consistent methodological approaches in meta-analyses.27 This includes the use of comprehensive search strategies, explicit language inclusion criteria, and robust heterogeneity assessments.27 Moreover, additional high-quality, large-scale, randomized, double-blind multicenter trials are required to better understand the optimal cell types, therapeutic parameters, and long-term efficacy of stem cell approaches in various patient groups and anatomical locations. The current study has several limitations. First, while this cross-sectional analysis incorporated searches across multiple databases, its scope was restricted to studies published exclusively in English. Consequently, non-English studies meeting inclusion criteria were excluded, potentially introducing language bias and overlooking relevant research from non-Anglophone regions. Second, the methodological framework prioritized high-quality evidence by including RCTs in meta-analyses. However, the certainty of evidence ranged from very low to moderate across different outcomes, indicating varying degrees of limitations in the available evidence. Consequently, this study cannot establish clinical recommendations grounded in Level I evidence for all outcomes, which represents the highest tier of empirical support. Third, certain clinical heterogeneity exists across the included studies. Differences in stem cell sources, delivery methods, and treatment protocols, as well as variability in the definitions of conventional therapy, may have contributed to inconsistencies in reported outcomes. Additionally, overlap of primary RCTs among included meta-analyses may have influenced effect estimates, and substantial clinical heterogeneity limits direct comparison across studies. Conclusion This cross-sectional analysis is the first study to compare stem cell therapies with conventional treatments for DFU, leveraging the Jadad decision algorithm to identify the most robust evidence. Findings consistently demonstrated that stem cell therapy appears to provide improved outcomes in ulcer healing rates, pain-free walking distance, and reducing amputation rates compared with conventional treatment, all while maintaining a favorable safety profile. Evidence from the highest-quality meta-analysis indicates that mesenchymal stem cells and mononuclear cells represent the most frequently studied and potentially effective cell types for DFU treatment, with bone marrow-derived cells being the most commonly applied source. In most randomized trials, stem cells were delivered through local administration, particularly intramuscular injection, which may enhance regional perfusion and tissue repair. Although current evidence supports the therapeutic potential of stem cell therapy, heterogeneity in cell sources, dosing strategies, and delivery routes highlights the need for further well-designed randomized trials to determine the optimal stem cell type and treatment regimen for DFU management. These conclusions are based on Level II and Level III evidence derived from overlapping meta-analyses and should be interpreted cautiously pending further high-quality, large-scale randomized controlled trials. Future large-scale, rigorously designed RCTs with long-term follow-ups are crucial to address these gaps, standardize protocols, and fully integrate stem cell therapy into evidence-based clinical practice. Funding Statement National Natural Science Foundation of China (82205089), Beijing Health Promotion Public Welfare Foundation (320.2745.2024102501), National Key R & D Program of China (2022YFC3502100), Fundamental Research Funds for the Central Public Welfare Research Institutes (ZZ13-YQ-030), Scientific and Technological Innovation Project of China Academy of Chinese Medical Sciences (CI2021A01614), Clinical Research Center Construction Project of Guang’anmen Hospital, CACMS (2022LYJSZX13), High Level Chinese Medical Hospital Promotion Project (HLCMHPP2023009),The Escort Project of Guang’anmen Hospital, China Academy of Chinese Medical Science-Backbone Talent Cultivation Project (9323002). Abbreviations DFU, Diabetic Foot Ulcer; RCTs, Randomized Controlled Trials; CCTs, Controlled Clinical Trials; AMSTAR, Assessment of Multiple Systematic Reviews; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; CCA, Corrected Covered Area; ROB, Risk of Bias. Data Sharing Statement Data are available on reasonable request to the corresponding author. Author Contributions Ying Du and Zhendong Xu: Data curation, methodology, investigation, validation, formal analysis, writing—original draft; Xinxin Meng, Bolun Zhang, and Qiantai Gao: Methodology, investigation, visualization, formal analysis; Tianlin Wen, Xiyan Zhao, and Zhiwei Jia: Conceptualization, supervision, writing—review and editing. All authors took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Disclosure All authors declare no conflicts of interest in this work. References Associated Data Data Availability Statement Data are available on reasonable request to the corresponding author.