Literature search

We performed a systematic and comprehensive electronic search in PubMed, Embase, and the Cochrane Library from inception through 21 August 2023, without any geographical or linguistic restrictions. The search terms utilized in the literature search were listed as following: (‘osteoprotegerin’ or ‘osteoclastogenesis inhibitory factor’ or ‘tumour necrosis factor receptor 11b’ or ‘follicular dendritic cell derived receptor 1’ or ‘FDCR 1 Protein’), and (‘ischemic stroke’ or ‘ischaemic stroke’ or ‘cerebral infarction’ or ‘brain infarction’). MeSH or Emtree terms and their variations were combined for the search. Detailed searching procedures are shown in S1 Table (Embase), S2 Table (PubMed) and S3 Table (the Cochrane Library). Additionally, a manual examination of the bibliographies of pertinent and incorporated studies was conducted to identify any additional potentially eligible papers.

Selection criteria

Two authors (L.-L. P. and H.-Y. L.) independently undertook the initial search, deleted duplicate records, and subsequently excluded irrelevant literature after reviewing titles and abstracts. The remaining potentially relevant literature was then assessed in full-text format to determine eligibility. We resolved discrepancies or disputes by discussion and consensus.

Studies meeting the following criteria were considered for inclusion: (1) case—control studies, cohort studies, or retrospective studies; (2) studies examining the relationship between circulating OPG levels and poor functional outcome (modified Rankin Scale [mRS] Score of 3–6) or mortality in ischemic stroke; and (3) studies comparing the highest quartile OPG level with the lowest quartile OPG level and providing either the odds ratio (OR) or adjusted odds ratio (aOR) directly, or containing data that allowed for the calculation of OR or aOR. Exclusion criteria included studies that exclusively recruited cases, published abstracts, animal experiments, reviews, comments or letters. The study with the largest sample size or the most comprehensive data was chosen among studies containing duplicate data.

Data extraction

One author (L.-L. P.) extracted essential data from articles included in the study using a standardized data-extraction form, while two other authors (X.-X. W. and W.-X. W.) independently verified the accuracy of the extracted data. The collected data encompassed various elements, such as the primary author, publication year, country of origin, study design, duration of research, inclusion and exclusion criteria, sample size, and outcome data. Data validation and discrepancies were addressed through collaborative discussions with coauthors.

Quality assessment

The quality of the included studies was evaluated utilizing the Newcastle-Ottawa Scale (NOS) [19], a widely recognized and reliable tool for assessing the quality of observational studies. This scale evaluate study quality from the aspect of eight items across three key areas: selection of the participants, comparability of the participants and outcomes. Study quality is categorized as low, moderate, or high based on scores falling within the ranges of 1–3, 4–6, and 7–9, respectively.

Statistical analysis

When both OR and aOR, along with their corresponding 95% confidence intervals are reported simultaneously in a single study, the preference is to utilize aOR for evaluating the pooled effect in meta-analysis due to its adjustment for known potential confounding factors. This adjustment enhances the accuracy and credibility of the pooled results. Heterogeneity among studies was assessed using the I² statistic, with a value greater than 50% indicating significant heterogeneity [20]. Based on the calculated I² value, a fixed effects model was selected when I² was less than 50%, while a random effects model was chosen when I² exceeded 50%, taking into account the variations among the included studies. Sensitivity analysis was performed by systematically excluding each study to evaluate its impact on the pooled outcome. Subgroup analysis, publication bias assessment, and meta-regression analysis were not performed in the present meta-analysis due to the limited number of studies included in the pooled effect estimation for each endpoint. A statistically significant result was determined by a p-value less than 0.05.. All statistical analyses were carried out using Stata software version 12.0 (Stata Corp LP, College Station, TX, USA).

Study selection and characteristics

The PRISMA statement flowchart illustrates the steps of identification, screening, and selection of eligible studies (Fig 1). We identified 148 citations from the initial electronic database search. Following removing duplicates and screening the titles and abstracts, 28 articles were deemed potentially suitable for inclusion. After conducting a comprehensive full-text review, five studies [9, 13–16] were ultimately included in the meta-analysis. The quality assessment of the included studies was performed using NOS (see Table 1; detailed scores provided in S4 Table). One study received a score of 8 stars [13], while the remaining four studies received scores of 7 stars [9, 14–16], indicating a high level of quality for each study.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Flowchart of screening for literature selection.

https://doi.org/10.1371/journal.pone.0303832.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristic of included studies that evaluated OPG levels and poor functional outcome and mortality in the meta-analysis.

https://doi.org/10.1371/journal.pone.0303832.t001

Study characteristics

The primary attributes of the included studies are outlined in Table 1, encompassing research conducted in China, Korea, Poland and Denmark. These studies were published between 2009 and 2023, with sample sizes ranging from 240 to 3,490, totaling 4,506 patients. The majority of enrolled patients exhibited symptoms of acute ischemic stroke within 7 days, including those who underwent endovascular thrombectomy in the investigation conducted by Park et al [14]. One study provided aOR and 95% confidence intervals for both poor functional outcome and mortality [9], while two studies reported only poor functional outcome [13, 14], and two other studies reported only mortality [15, 16].

Outcomes

While individual studies indicated a positive correlation between OPG and poor functional outcome [9, 13, 14], the pooled analysis revealed no significant association between OPG and poor functional outcome in patients with ischemia stroke (aOR 1.29, 95% CI 0.90–1.85). The high I² statistic of 79.5% indicated a considerable level of heterogeneity among the studies (Fig 2A). During sensitivity analysis, each individual study was sequentially excluded, leading to the identification of Song et al’s study [13] as a significant contributor to heterogeneity and a chang in the outcome regarding concomitant use (aOR 1.53, 95% CI 1.09–2.15) (Fig 3A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2.

Forest plot illustrating the association between osteoprotegerin (OPG) and poor functional outcome (Fig 2A) and mortality (Fig 2B) in ischemic stroke patients.

https://doi.org/10.1371/journal.pone.0303832.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Sensitivity analysis for the robust assessment of poor functional outcome (Fig 3A) and mortality (Fig 3B).

https://doi.org/10.1371/journal.pone.0303832.g003

Similarly, a lack of significant correlation was observed between OPG levels and mortality in patients with ischemic stroke (aOR 1.57, 95% CI 0.90–2.74) (Fig 2B), despite the individual findings of three included studies indicating a positive association between OPG levels and mortality [9, 15, 16]. Sensitivity analysis revealed that the study by Wajda et al. [13] contributed to heterogeneity and resulted in a reversal of the outcome when considering concomitant use (aOR 2.16, 95% CI 1.30–3.58) (Fig 3B).

Main outcome

After reviewing the existing literature, our analysis indicates that there is no significant association between circulating OPG levels and poor functional outcome or mortality in patients with ischemic stroke. Further examination of this finding is warranted, as it may not be consistently supported in sensitivity analyses.

Relationship and potential mechanisms between OPG and IS

Osteoporosis and cardiovascular disease, despite appearing as distinct conditions, are intricately linked due to shared risk factors and a common feature of low grade chronic inflammation [8]. OPG, originally identified by Simonet et al. as a pivotal modulator in bone metabolism [21], has garnered increasing attention for its involvement in vascular disease and calcification in recent years. Nevertheless, the precise mechanisms by which OPG operates within the vascular system remain elusive, and the findings of pertinent investigations are subject to debate. Numerous clinical researches have demonstrated a positive correlation between elevated levels of circulating OPG and the incidence of various cardiovascular diseases, such as coronary artery disease, heart failure, unstable angina, peripheral artery disease, symptomatic carotid stenosis, and vulnerable carotid plaques [22]. OPG was found to be independently correlated with traditional risk factors of atherosclerosis, subclinical peripheral atherosclerosis, and clinical atherosclerotic disease including ischemic heart disease and ischemic stroke [23]. Findings from the 4^th Copenhagen City Heart Study [24] and the Tromsø Study [25] indicated that elevated levels of circulating OPG were predictive of increased risk for ischemic heart disease, ischemic stroke, and all-cause mortality in general population. Similarly, studies have demonstrated elevated levels of circulating OPG in IS patients compared to the general population [11, 26, 27]. Moreover, elevated levels of OPG have been found to be associated with incident [10, 11, 28], severity at admission [10, 27, 28], poor functional outcome [9, 13, 14] and mortality [9, 15, 16] following an IS event. This association is supported by evidence linking elevated plasma OPG levels to vulnerable carotid plaques [29], larger infarctions [13], higher National Institutes of Health Stroke Scale scores at admission [13] and the occurrence and severity of hemorrhagic transformation [30]. The potential mechanisms by which OPG may be linked to the development of inflammatory diseases such as atherosclerosis are multifaceted. Firstly, OPG has been shown to upregulate the expression of intercellular molecule-1, vascular adhesion molecule-1, and E-selectin on endothelial cells, leading to increased leukocyte adhesion, a critical early event in endothelial dysfunction [31]. This suggests that OPG palys a important pro-atherosclerotic role in vascular diseases. Secondly, OPG has been found to significantly enhance the RANKL-stimulated effect on matrix metalloproteinase (MMP) activity, particularly at specific OPG/RANKL molar ratios [32]. Additionally, high concentrations of OPG have been shown to directly induce MMP activity directly [32]. MMP is widely recognized for its role in promoting extracellular matrix degradation, potentially contributing plaque instability destabilization and increasing the incidence of vascular diseases. Thirdly, OPG has been shown to facilitate the anchoring of ultra large von Willebrand factor (vWF) multimers at the site of vascular injury, thereby promoting localized thrombus formation [33].

In contrast to the aforementioned perspective, there are studies that offer a divergent viewpoint, as demonstrated by the results of our research article. To date, only three studies have presented divergent perspectives on the clinical implications for IS patients. One such study, which focused on a cohort of 490 white women aged 65 years and older, found no significant association between OPG levels and baseline bone mineral density or with subsequent strokes or fractures [7]. Additionally, Nybo and colleagues conducted a nested case-control study within a large population of 57,053 individuals in Denmark, concluding that plasma OPG concentrations were not associated with an increased risk of ischemic stroke [12]. Another separate investigation examing the impact of serum osteopontin, OPG and RANKL in individuals with acute ischemic stroke indicated that OPG did not demonstrate any significant association with radiological or clinical parameters and scores [17], although detailed data was regrettably unavailable. There are two potential explanations for the adverse outcomes observed in these study. Firstly, elevated serum OPG levels in humans may represent a reactive response rather than a causal factor in the development of atherosclerosis or vascular calcification [7]. Secondly, cardiovascular disease may potentially benefit from OPG due to its inhibitory effects on vascular calcifications at higher concentrations [12]. Research involving animals has also shown that OPG exerts a protective influence against vascular calcification by potentially inhibiting alkaline phosphatase activity [34–37]. Hence, the absence of correlation between OPG and IS probably repesents the underlying circumstances.

Strengths and limitations

There are some limitations in our study. First of all, the final result lacks robustness in sensitivity analysis, potentially due to the inclusion of rare included literature, variations in study populations, sample sizes, study designs and adjusted confounding factors. Secondly, we had not performed subgroup analysis, publication bias and meta-regression analysis due to the scarcity of included literature, preventing further identification of factors influencing result robustness. Furthermore, the establishment of a causality relationship between OPG and IS is hindered by the reliance on observational studies in the existing literature, as well as the lack of clarity regarding the exact pathophysiological mechanisms of OPG. Therefore, additional well-designed researches are necessary to investigate and address these unresolved issues.

Conversely, our study possesses several strengths. To the best of our understanding, the present meta-analysis is the inaugural publication to thoroughly assess the relationship between OPG and IS. We employed adjusted ORs from each incorporated literature to aggregate the findings in order to mitigate the influence of confounding factors, thus enhancing the accuracy and credibility of our results.