ESG-SDG Nexus: Research Trends Through Descriptive and Predictive Bibliometrics

Abstract

Integrating environmental, social, and governance (ESG) reporting with the Sustainable Development Goals (SDGs) is important for achieving corporate sustainability. The rapid evolution of regulations like the Corporate Sustainability Reporting Directive (CSRD), and the fragmented research landscape create uncertainty for strategic planning. This paper addresses the critical gap related to the lack of predictive data into future research trends at the ESG-SDG nexus. The research begins with a bibliometric analysis using two software programs R-Biblioshiny 5.2.0 and VOSviewer 1.6.20, to process data extracted from the Web of Science (Clarivate). Selected key terms regarding sustainability reporting concepts and reporting standards, as well as the engagements of auditors were used to filter the database information. Starting from the bibliometric analysis of 361 publications completed during January 2015–September 2025, the study performs further a quantitative measurement bibliometrics using RStudio 4.5.2 and provides a novel ensemble forecasting model (AutoRegressive Integrated Moving Average, Error, Trend, Seasonal Components, and Linear regression with SDG factors) that cartograph the alignment of the current research field and forecast its evolution. The results reveal that terms regarding reporting “CSRD” and sustainability assurance, “ISSA 5000” are the most dominant research fronts, strongly aligned with SDG 12, 13 and 17. The forecasting model predicts sustained growth in this area. The study contributes by providing a forward-thinking strategic map for researchers, policymakers and businesses, transforming sustainability integration from a compliance task into systematic, data-driven approach for priority setting strategy.

1. Introduction

Over the last decade, policy uncertainty and disruptive dynamics have increased in the global economy, prompting companies to strive for competitiveness by delivering high-quality products or services and adapting to new sustainability requirements. For improved transparency and resilience, enhanced through sustainability reporting, companies can demonstrate their environmental and social responsibility, while also strengthening their relationships with various stakeholders [1]. The company’s ability to integrate environmental, social, and governance (ESG) criteria into its strategy becomes increasingly important [2].

The regulatory framework has evolved from the UN 2030 Agenda, which developed seventeen Sustainable Development Goals (SDGs), to a continuous process of sustainability reporting disclosure that encompasses ESG aspects. The integration of the SDG-ESG nexus represents a strategic approach, supported by the complex regulations developed over the last years. At the global level, regulators have enforced rules for disclosing complex non-financial information that supports sustainability goals. For example, with its ambitious targets, the European Union’s (EU) directives have evolved from a voluntary disclosure, as supported by the Non-Financial Reporting Directive (NFRD) [3], to a mandatory disclosure based on double materiality, as outlined in the Corporate Sustainability Reporting Directive (CSRD) [4], which facilitates companies’ transition to sustainable practices.

While the EU has developed the most comprehensive and integrated regulatory framework in the world, EU guidelines are applied to all companies outside the EU that operate on the European market, and many non-EU countries adopt or adapt these regulations. The EU framework ensures increased transparency compared to other standards, and the European Sustainability Reporting Standards (ESRS) normalized methodology ensures comparability [5,6]. Moreover, the European market predictably reacts to the regulatory announcements. The EU introduced revolutionary concepts, such as double materiality, taxonomy, and value chain, which are explored in this study. We chose this framework as a reference, not because it is the best, but because it is the most mature and comprehensive regulation, has a global impact on the market, offers data for longitudinal analysis, and represents a natural laboratory for studying the adoption of sustainability reporting. The alternative would have been to consider multiple reporting frameworks, such as the Global Reporting Initiative (GRI), Sustainability Accounting Standards Board (SASB), Task Force on Climate-Related Financial Disclosure (TCFD), International Integrated Reporting Council (IIRC), etc. This choice would have reduced data consistency and made it difficult to isolate a single framework.

The EU’s ambitious approach that empowers big companies to provide complex metrics supported by ESRS, as well as small and medium-sized enterprises (SMEs), to provide information voluntarily, in accordance with European Financial Reporting Advisory Group (EFRAG) proposals, has led stakeholders to face strategic uncertainty in navigating its rapid evolution [7]. In this context, the Omnibus regulation package [8] and the “Stop the clock” directive [9] simplify and postpone for two years the numerous administrative tasks for companies affected by the new regulations.

Following the development of these regulations, academic literature evolved, incorporating new concepts and paradigms. Mainly, the authors focused on descriptive and retrospective mappings of the field, with descriptive bibliometric analysis dominating the field’s cartography [10,11]. Still the stakeholders (companies, investors, regulators, researchers) are confronted with a strategic uncertainty in the ESG-SDG integration due to a complex and dynamic chain of factors: the rapid evolution of the regulations (CSRD, ESRS), of the assurance standards of sustainability reporting, from International Standard on Assurance Engagement (ISAE) 3000 [12] to International Standard on Sustainability Assurance (ISSA) 5000 [13], and a vast and retrospective academic literature. They must understand not only that there is a domain but also where it is headed to prioritize their investments, research, and regulatory efforts efficiently. It involves a strategic decision-making problem in complex and uncertain conditions.

A major challenge in integrating CSRD-related research topics with SDGs is the comprehensive complexity and dynamism of the system [14]. A critical gap exists in academic literature, as there are a lack of approaches that combine the bibliometric cartography with forecast models to anticipate the research tendencies and to link them directly to the strategic alignment with specific SGDs, offering an actionable guide for strategic planning. Essentially, there is a leap from description to prediction, which the literature has yet to make systematically in this specific domain. The existence of practical problems (strategic uncertainty) creates a need for knowledge. The existing academic literature, primarily based on descriptive bibliometrics, does not adequately address this need, leaving a research gap due to a lack of prediction-oriented research aimed at action and predictive analytics that forecast research trends to inform strategic decision-making in sustainable business integration.

The primary objective of this paper is to illustrate the evolution of sustainability reporting. Additionally, the study aims to demonstrate how the new sustainability framework aligns with the 17 Sustainable Development Goals, which encompass the three pillars: environmental, social, and governance matters. Lastly, predictive bibliometrics adds value by providing a data-driven approach for setting priority strategies.

Following the problem statement and the research gap, in accordance with the paper’s objective, the study is guided by the following research questions:

RQ1: What is the structural and thematic evolution of academic literature at the intersection of sustainability reporting, assurance, and the SDGs over the past decade (January 2015–September 2025)?

RQ2: Which CSRD-related research terms show the most substantial alignment with specific SDGs, and how is their growth trajectory projected?

RQ3: Based on the forecasted trends, what are the strategic priority areas for future research and business investment to strengthen the ESG-SDG nexus?

Our research employs a tripartite analytical methodology designed to map systematically, evaluate, and forecast the ESG-SDG nexus: first, a bibliometric analysis to cartograph the existing academic research landscape and identify key themes and evolutions; second, an alignment (mapping) analysis to critically assess the coherence and gaps between ESG reporting practices and specific SDG targets; and third, a predictive model approach that combines multiple scientific approaches with sustainability metrics, utilizing AutoRegressive Integrated Moving Average (ARIMA), Error, Trend, Seasonal (ETS) Components, and regression models with SDGs to extrapolate future research trajectories and emerging trends, transforming empirical insights into a strategic, forward-looking framework.

The paper contributes to the literature in three ways:

1. It provides a novel predictive perspective by combining bibliometric mapping with an ensemble forecasting model; It moves beyond the dominant descriptive paradigm in bibliometric studies of ESG/SDG research [10,11] by integrating ensemble forecasting models. This allows us to project future research trends, offering strategic foresight rather than retrospective analysis.

2. It introduces a quantifiable SDG alignment score for CSRD-related topics. We develop a novel SDG Impact Score, a composite indicator that assesses both the volume and interdisciplinary breadth of CSRD-related research associated with each of the Sustainable Development Goals. This provides a granular, empirical map of how current academic discourse connects corporate social responsibility and disclosure (CSRD) to global goals (SDGs), addressing the often-theoretical discussion on this nexus.

3. It offers actionable strategic priorities for investment and research derived from the empirical forecast. We synthesize the bibliometric mapping, SDG alignment scores, and growth forecasts to identify Tier 1 and Tier 2 priority domains (e.g., ‘ISSA 5000′ for auditors, ‘double materiality’ for strategists).

The paper is structured as follows: Section 2 develops the key framework and concepts, Section 3 details the bibliometric and forecasting methodologies, Section 4 presents and discusses the results, and Section 5 presents the conclusions, implications, and limitations.

2. Literature Review

2.1. Overview of Sustainability Reporting

The sustainability reporting trend has rapidly increased within large companies, making investors seek a third-party assessment of undertakings’ ESG performance. The main objectives of companies are to maximize shareholder value and provide benefits to the society in which they operate. Other positive aspects of sustainability reporting are lower cost of capital, increased customer and employee satisfaction, access to better financing, and a stronger brand [15].

Over the past few years, several reporting frameworks have been implemented to support businesses in their shift to sustainable practices. The most recent frameworks are the standards set by the International Sustainability Standards Board (ISSB), as well as ESRS, which are based on the GRI framework.

The academic literature is rich in published studies that have tracked companies’ fulfillment of the SDGs from different perspectives and based on a multitude of theories. Thus, Perumandla et al. [16] analyzed over 500 published papers to identify the factors influencing the application of the SDGs by analyzing CSR and ESG practices. Thus, their study revealed that the published works were grounded in a wide range of theories, including stakeholder theory, legitimacy theory, agency theory, and institutional theory. They also identified the determinants of the SDGs, which were grouped into several categories, namely investment and financial factors [17], corporate practices [18], and environmental, social, and governance factors [19].

Recent research papers indicate that ESG practices are not isolated; they are linked to achieving the SDGs [20,21]. Thus, the ESG–SDG nexus reflects an active interface where corporate sustainability actions contribute to the global development agenda.

To evaluate a business’s performance in terms of ESG dimensions, rating methodologies have been created by various specialized companies that generally provide services to stock exchanges. However, as there are several ESG rating providers in the market, these ratings often exhibit substantial divergence, generating considerable uncertainty regarding the actual quality of a company’s performance. Divergences are most influential in the following categories: climate risk management, product safety, corporate governance, corruption, and environmental management systems. Even if ESG ratings are a valuable indicator for stakeholders, it is recommended that ESG rating agencies should better communicate their definition of sustainable performance and become more transparent with their measurement practices and methods [22].

Given that the ESG achievements developed by specialized companies did not accurately reflect corporate performance in terms of sustainability, new indicators have emerged to evaluate the impact of companies on the SDGs. Thus, van Zanten [23] analyzed the link between SDG scores (Robeco and Morgan Stanley Capital International—MSCI) and ESG ratings (MSCI, Sustainalytics, Refinitiv, andStandard & Poor—S&P) and found that SDG scores capture how stakeholders judge companies’ impact on sustainable development, while ESG ratings do not provide this information. In the author’s opinion, SDG scores demonstrate high construct validity for assessing corporate sustainability performance, whereas ESG ratings exhibit low construct validity.

As demonstrated by Krueger et al. [24], there is a significant positive effect of the mandatory ESG data points on stock liquidity. The authors analyzed the effects of mandatory ESG disclosure based on data from companies on all continents in accordance with the GRI reporting framework. Moreover, if the disclosure requirements are implemented at the government level, companies tend to allocate more resources towards ESG disclosures. The research also highlights that undertakings with weaker information reported should consider the internal and external benefits for their capital markets.

2.2. Key Concepts in ESRS

In today’s emerging economy, stakeholders demand more transparency and comparable information from the companies’ sustainability statements. Therefore, in response to the imperative demands, the European Union, as the vanguard of promoting sustainability reporting, Hummel [25] has developed the CSRD [4] to replace the NFRD [3]. The new instruction aims to expand and fortify sustainability reporting requirements, while also enhancing the key characteristics that must be present in any document related to sustainable activities: transparency, comparability, consistency, and accountability. In addition, the CSRD applies to several companies, including small and medium-sized listed firms, including companies outside the EU that have significant operations on the European continent.

However, as the current ESRS requirements are rigorous and sometimes vague, the European Commission has planned an “Omnibus” to simplify them. The “Stop the clock” Directive has been issued by the European Parliament and the Council of the European Union [9]. This directive took effect on 17 April 2025, and is the first from the Omnibus Package [8]. Another relevant regulation that entered into force in July 2025 is the Delegated Regulation, also known as “Quick Fix” [26], which came into force in November 2025. EFRAG voted to revise Set 1 of the ESRS standards, which includes simplifications. The European Commission is estimated to embrace the final amendments in 2026 [27]. From the multitude of changes to the ESRS in terms of content and application period, it is clear that this field is in a state of continuous evolution and development.

Double materiality is another key aspect introduced by the CSRD. This concept implies, on the one hand, financial materiality, i.e., the inclusion of information necessary to understand a company’s development stages, its performance, and its financial position in increasingly competitive markets. This financial materiality highlights the impact that sustainability issues can have on a company’s performance and reliability. On the other hand, double materiality also refers to reflecting in the sustainability statement how businesses influence the environment and society during their regular business activities. This perspective is referred to as impact materiality by Cooper [28].

To demonstrate the practical implications of the double materiality assessment, Dyczkowska & Szalacha [29] analyzed meetings between representatives of an ESG consulting firm, the client PORR Group, and the company’s stakeholders. They found that the qualitative assessment of impact materiality posed the most difficulties for participants. Quantitative data was explored with the help of financial experts, but qualitative indicators played an important role in discussions with stakeholders. These qualitative indicators validate environmental and social issues, which led to some discrepancies in the assessment of materiality. In the analyzed process, all stakeholders assigned the same level of importance to all topics discussed, even if a particular issue fell outside there area of expertise. In this regard, according to EFRAG recommendations [7], each stakeholder should be given a certain degree of importance, depending on the field of activity related to the reporting company’s activity.

The main challenge in integrating CSRD-related research topics with SDGs is the comprehensive complexity and dynamism of the system [14]. Operato et al. [30] highlighted how organizations can utilize the European Union’s CSRD to transition from simple compliance to a genuine sustainable development ecosystem. When comparing the benefits of sustainability reporting according to the ESRS with the proposed or already implemented amendments in the Omnibus Package, a clear tension is noticeable. On the one hand, the currently existing framework seeks to encourage complexity as a basis for substantive results. On the other hand, the proposed amendments and those that have already been implemented prioritize lightening the reporting burden, potentially sacrificing effectiveness and consistency with the European Union’s own long-term strategic targets, such as climate neutrality [31].

Another key idea of the entire CSRD framework is the requirement for companies’ reports to have mandatory assurance. This confirms that the reported ESG data is verified and reliable, increasing trust among investors, stakeholders, and regulators [32].

2.3. Greenwashing Phenomenon and Its Relationship with Sustainability Assurance Engagements

Since 1 January 2023, sustainability reporting has been subject to a level of control similar to financial reporting, following the establishment of the CSRD. The number of companies impacted exceeds 42,500 across various industries in the European territory and beyond.

Yamahaki et al. [33] describe the greenwashing phenomenon as a discrepancy between communication and performance, resulting from the omission of information or misconduct, which leads to misleading actions. Some of the most sensitive metrics are related to the environmental topic and consist of data such as renewable energy (generated or purchased), GHG emissions, and pollution mitigation. Combating greenwashing is a crucial goal to achieve for both users of non-financial information and companies genuinely committed to transitioning to a sustainable business model. Thus, sustainability assurance engagements serve as a unique mechanism for monitoring sustainability practices. Bu et al. [34] demonstrate in their study, which applied to 73 chinese businesses listed on stock exchange, that ESG assurance engagements have a considerable repressive effect on greenwashing attempts. This result is visible in the case of companies that are not state-owned and are required to implement ESG reporting regulations. Last but not least, the research highlights the key role that assurance plays in combating greenwashing behavior, as companies strive to improve their sustainability practices not only on paper but also through tangible actions. As stated in the International Capital Market Association (ICMA) report [35], greenwashing is a term that covers misrepresentations and problematic behaviours related to sustainable finance. Policymakers and regulators have launched several initiatives to combat the greenwashing phenomenon, which has raised concerns among market participants. Some of the recommendations proposed by ICMA consist of concentration on actionable areas of concern in sustainable finance (sustainable bond market, misleading labelling of sustainable funds), helping to improve the availability of data and analysis on market integrity, referencing existing regulations, implementation of current regulatory initiatives with a focus on international operability and utilisation, and leveraging the positive contribution of market best-practices.

To help combat greenwashing, assurance practitioners apply the ISAE 3000 regulatory framework, developed by the International Auditing and Assurance Standards Board, to assurance engagements on sustainability reports, thereby ensuring assurance of ESG topics. The key points of ISAE 3000 are: increasing user confidence in the specific information presented in the report; tailoring the consulting services to the client’s needs; conducting engagements by an independent practitioner who is a professional in the economic and financial field from external firms [12].

In addition, explicitly tailored for environmental assurance, the ISAE 3410—“Assurance engagements relating to greenhouse gas statements” is a standard that practitioners can also use. It provides valuable information about the process of transition to a greener strategy for a company. Additionally, ISAE 3410 employs a risk-based approach and does not prescribe a specific type of assurance engagement [36].

The new regulatory framework proposed by the IAASB, ISSA 5000—General Requirements for Sustainability Assurance Engagements [13], aims to strengthen the credibility of sustainability reports through a new set of regulations for assurance engagements, regardless of the nature of the information reported (environmental, social, or governance). ISSA 5000 applies to companies of all sizes and industries and is expected to enter into force after 15 December 2026, with the possibility of early application. It is based on reasonable assurance commitments, providing flexible selection procedures for auditors, with exclusive information related to sustainability. By covering all the critical details of sustainability reporting, ISSA 5000 is designed to replace the ISAE 3000 and ISAE 3410 standards, thereby confirming a more comprehensive alignment with current requirements for transparency and responsibility, as well as a higher level of detail in ESG-SDG nexus reporting.

While the regulatory push (e.g., CSRD, ESRS) mandates the integration of the ESG-SDGs nexus, the academic literature remains mostly descriptive and retrospective. In a recent study, De Silva et al. [37] discuss the current research status and emphasize the necessity of sophisticated approaches beyond descriptive mapping. Moreover, Ellili [38] conducted a comprehensive bibliometric study, which also identifies the predominant descriptive tendencies in the literature and suggests future predictive research. Au et al. [39] propose an explicit research agenda that introduces forecasting as a key element in sustainable development business strategy.

Our study aims to bridge the gap in previous research proposals by moving beyond traditional bibliometrics to develop a predictive model that forecasts research trends and proposes strategic investment priorities at the ESG-SDG nexus level.

While traditional bibliometric analysis provides a diagnostic snapshot of a research field by identifying established themes, influential works, and collaboration networks [38], forecasting offers a prognostic tool—it models future trajectories based on temporal patterns and external drivers.

In this study, we create new knowledge combining the traditional and predictive answering both to ‘What has been studied and where are we now?’ to’Where is the field heading, and where should strategic attention be directed?’ by first mapping the current academic structure of CSRD-related research and then forecasting its evolution.

The CSRD creates an imperative for businesses to register their environmental and social impacts. Forecasting CSRD-related topics, we are indirectly forecasting the practical tools that will help fulfill the SDG goals.

3. Research Methodology

Bibliometric analysis has become a cornerstone of understanding the complex sustainability-related fields. Several studies have mapped the academic landscape of this area [10,11]. These studies synthesized past knowledge and identified research clusters through a retrospective and descriptive methodology, concluding with recommendations for future studies that provide quantitative, predictive trajectories of the field. This creates a methodological gap; while we know where the research has been, we lack robust, model-driven tools for anticipating future directions. Our study addresses this gap by mixing cartographic mapping techniques from bibliometrics with advanced forecasting models to generate actionable, forward-looking information for the ESG-SDG nexus.

This section runs through the methodology employed to address the research questions and provides a systematic framework for data collection, analysis, results discussion, and interpretation. Through this research, we aim to deliver a comprehensive cartography of the main academic writing movements related to the ESG-SDG nexus, based on CSRD and ESRS requirements, as well as forecasts that consider the challenges and opportunities in sustainable business integration (Figure 1).

To answer the RQ and overcome the limitations of previous descriptive studies, we adopted an integrated conceptual framework (Figure 1), which transforms the driving factors (inputs) of the ESG-SDG nexus into strategic impact through a cyclic research process. This combines: 1. The bibliometric cartography of the field, 2. Critical analysis of the alignment reporting practices (CSRD/ESRS) and SDGs, and 3. The predictive modelling of the evolution of tendencies. The outputs of this process—the forecasts, research tendencies, and the strategic guide inform future practices and policies, closing a feedback loop that assures a continuous relevance of the framework. This conceptual framework is the background on which the paper’s logic is built. Bibliometric analysis ensures objectivity within the examined sample [40].

This research database is the Web of Science (WoS) by Clarivate, providing a broader range of high-quality academic writings and ensuring credibility in the field of study. Web of Science allows files to be downloaded in different formats. Thus, the database used for this research study was downloaded in “Tab delimited file” to be used in VOSviewer 1.6.20 and in “BibText” format to be used in Biblioshiny software 5.2.0, as well as in RStudio 4.5.2. The selected period is from January 2025 to September 2025. The data was extracted from WoS on 8 November 2025.

To identify the academic papers that would establish the research sample, we searched for articles in the field of non-financial reporting and assurance, based on the ISAE 3000 or ISSA 5000 frameworks. To ensure an adequate search using the CSRD, ESRS, ISAE 3000, and ISSA 5000 reporting frameworks, the following Boolean string was used in combination with the corresponding logical connectors: (“Sustainability reporting” OR “ESG reporting” OR “CSRD” OR “Corporate Sustainability Reporting Directive” OR “ESRS” OR “European Sustainability Reporting Standards”) AND (“ISSA 5000” OR “ISAE 3000” OR “assurance”).

For the selection of sample papers, the key phrase “sustainability reporting” was primarily considered, which could have generated results at a global level. By associating it with the terms CSRD and ESRS, the final sample included papers that were published predominantly by authors from the European Union and studies conducted on European companies. However, no restriction was set regarding the countries of origin of the authors. To ensure the robustness of the research conclusions, the bibliometric analysis will be structured into two categories based on the authors’ countries of origin: developed and developing. To perform the statistical tests, the data were grouped into developed countries as high-income and developing countries as upper-middle-income and lower-middle-income, based on the World Bank’s country classification.

In the current bibliometric analysis, two programs were used: VOSviewer and R Studio—Biblioshiny. Van Eck and Waltman [41] suggest that VOSviewer can be utilized to create maps based on keywords, co-citations, and authors, enabling researchers to analyze the information in greater detail. It is also a powerful tool when the sample has more than 100 items. In the bibliometric analysis, the keyword maps and those referring to the most influential articles were created using VOSviewer version 1.6.20. To refine the database and acquire accurate outcomes for the bibliometric analysis, a “thesaurus” file has been imported. The thesaurus file has been checked so that each term is unique.

The process of identifying studies in the Web of Science (Clarivate) database is described in Figure 2 using a PRISMA flow diagram. The initial search consists of 430 papers. After applying several filters and refining the inquiries, the research is focused on 361 articles.

The dataset for the period January 2015–September 2025 incorporates academic discourse based on CSRD terminology and its established linkages to the UN SDGs. Using the tidyverse, forecast, lubridate, ggplot2, viridis, scales, and gridExtra libraries in R Studio 4.5.2, we performed a complex analysis based on mapping and forecasting.

Twelve core CSRD-related research terms (CSRD, double materiality, supply chain, value chain, EU Taxonomy, ESRS, ISAE 3000, ISSA 5000, sustainability assurance, sustainability reporting, investors, stakeholders) were selected based on their importance in sustainability reporting frameworks. Their selection is grounded in the conceptual architecture of the EU’s sustainability reporting framework, not chosen arbitrarily, but based on the key components that the CSRD addresses: 1. The regulatory core (the “what”) CSRD, ESRS, and EU taxonomy as binding rules and standards, 2. The conceptual principles (the “how”)—double materiality, supply chain, value chain as novel methodologies defining scope and assessment; 3. The reporting practice (the “action”)—ESG reporting, sustainability reporting, 4. The assurance mechanism (the “verification”)—sustainability assurance, ISAE 3000, ISSA 5000, that helps to ensure credibility and combat greenwashing; 5. The strategic purpose (“for whom, why”)—investors, stakeholders, as driving the reliable data demand. This approach ensures a comprehensive value chain of policy-practice CSRD triggers, related to practical challenges and academic research innovation, that directly inform the capabilities needs for SDG alignment.

Based on these terms, we created a mapping dictionary as the foundation for the qualitative analysis of the CSRD’s thematic requirements and the SDG targets, defining the mapping between each CSRD-related research term and its corresponding relevant SDGs. To quantify the interdisciplinary relevance and impact of each SDG within our research landscape, we constructed an SDG impact score. This approach aligns with previously developed research on creating robust, sustainable indices [42], particularly in measuring research alignment with global agendas [43].

To respond to the second research question (RQ2), we calculated an Impact Score based on two key variables extracted from the research data: 1. Total research mentions, representing the total number of articles mentioned for each SDG across all CSRD-related terms, and 2. The number of distinct CSRD research terms that map to each SDG as terms covered. The formula used is as follows:

Impact score = (Total mentions × Terms covered)/10 (The division by 10 is a normalization factor to scale the score to a manageable range for visualization).

The volume of research representing the total mentions reflects how frequently an SDG is discussed in the literature, indicating its prominence. The terms covered measure the number of different CSRD topics related to the SDG, indicating its interdisciplinary relevance. The multiplicative relationship ensures that SDGs with both high research volume and broad coverage receive higher impact scores, highlighting goals that are central to CSRD research.

To answer the third research question (RQ3), we developed an ensemble forecasting model. The idea was also explored by Chenary et al. [44], who developed a forecast for global SDG scores using machine learning, and Martinez [45], who focused on hybrid models for improved predictive accuracy.

We started with a temporal series (2015–2025) of annual academic research articles for each key term. To forecast the future, we did not perform only a single method, but we created an ensemble model that combines three different methods for maximum accuracy:

• ARIMA, which captures trends and patterns from our historical data.
• The ETS model, which models volatility and error structure.
• Linear regression with SDG factors that explain the research activity through important external factors. In our model of linear regression, performed to answer RQ3, the dependent variable (LR = Linear Regression target) is the annual frequency of research publication for each of the twelve selected CSRD-related terms. Our model forecasts the number of scientific articles that will be published annually about each key concept within the sustainability reporting umbrella. This variable is defined as:

LR = β₀ + β₁(SDG_Weighted_Score) + β₂(SDG_Alignment_Index) + β₃(regulatory_wave) + β₅(Market_Volatility) + β₄(Time_Index) + ε

where:

-

LR (linear regression target) is the forecasted research volume.

-

SDG_Weighted_Score: how well a key term covers multiple SDGs

-

SDG_Alignment_Index: Grade of strategic alignment with SDGs targets

-

Regulatory_wave: intensity of the regulatory pressure

-

Market_volatility: economic uncertainty.

-

Time_index: general temporal trend.

While individual models have weaknesses [44], we combined them into an average (40% ARIMA, 35% ETS, and 25% regression) to create a more robust composite forecast.

The construction of the exogenous variables is based on the SDG Alignment index, which is calculated annually by mapping SDGs and corresponding annual growth rates. For the regulatory Pressure waves, a scaled index (0–1) reflecting key regulatory milestones (2014 NFRD = 0.2, 2022 CSRD adoption = 0.6, 2024 ESRS = 0.8). The market volatility proxy considered the annual average of the EU Economic Policy Uncertainty Index.

Finally, to answer RQ3, a Strategic Score is used to map each forecasted initiative into an investment-priority tier. Investment priority is determined using:

Strategic_Score ≥ 80 indicates “TIER 1 (critical); Tier 2—High (≥65 points), representing domains that exhibit growth potential and a pronounced alignment with sustainability.

The forecasting has inherent limitations, framing its results as exploration and strategic. The 10–11-year data limit reduces statistical robustness. The model cannot account for unforeseeable, paradigm-shift events, and the constructed indices are a simplification of complex reality. The model offers a systematic, data-driven approach for priority setting, rather than providing precise numerical predictions.

4. Results and Discussion

By using the selected terms, a series of relevant academic writings in the field of sustainability were identified (Figure 3).

Based on the graphic in Figure 3, the number of publications on this area increased significantly during the analyzed time horizon (2015–2025). The lowest stage was noted in 2015, with only 13 publications, which can be explained by companies’ focus on financial statements and results. In 2016, the GRI standards were restructured to facilitate the non-financial reporting process for companies, leading to an increase in publications in this area of research. The highest number of publications in the field of sustainability was recorded in 2023, with 55 papers, followed closely by the first three quarters of 2025. Given the current updates in sustainability reporting regulations, it is expected that research interest will increase, resulting in an upturn in the number of publications, surpassing the value seen in 2023.

This phenomenon is due to the implementation of the ESRS regulatory framework, which highlights new ESG reporting practices. In terms of citations, the trend is also on the rise, with the most significant increase occurring between 2023 and 2024, from 1.833 to 2.438 citations. The first part of the year 2025 is, of course, the most abundant in terms of citations: 2.585. We can see the increased interest of researchers in sustainability reporting and the amount of these documents. However, to strengthen these results, Figure 2 suggests that the foundations of sustainability reporting and its importance for a company’s performance were established in research from 2015 to 2022, prior to the development of the ESRS.

Moreover, the study is conducted on a global scale to investigate how different territories adapt to ESG trends. Of course, mitigating environmental, social, and governance risks is also a matter of corporate culture, which is diverse and specific to each territory. Figure 4 illustrates the macroeconomic situation, specifically the number of citations, for the top 20 countries.

The most representative country, in terms of citations, is Spain, with 1323 citations and an average of 42.7 records per year during the analyzed period. The second position is occupied by the United States, with 1240 citations and an average of 31.8 citations. The following three countries in the ranking are the United Kingdom, with 968 citations, Australia, with 958 citations, and Germany, with 939 citations.

One common feature that can be observed among all 20 countries is that they have developed and prosperous economies, a high quality of life, and leading companies in international markets. The fact that the ranking includes articles written by authors from non-EU countries demonstrates a high degree of collaboration among authors from various regions.

Figure 5 represents a box plot that clearly highlights the structural differences in research impact between the two countries groups.

The visual description (Figure 5) indicates that the distribution is targeted at higher citation values, with a superior median and dispersion in developed countries, and a low-value distribution with lower variability in developed countries. The two distributions are strongly differentiated, indicating a systematic disadvantage of the developing countries.

The econometric analysis highlights the effect of the development status (coefficient: 0.7117, p = 0.001). The high statistical significance (p-value < 0.01 confirms that the visual difference is not a hazard. The negative coefficient is translated into a deficit of ~51% in citations for developing countries (exp (−0.7117) ≈ 0.49). Figure 5 illustrates this gap precisely, with the distribution for developing countries flat on the left side. The model explains 79.9% of the citation variations, having a high quality.

The model is based on the formula: Log(citations_i) = β₀ + β₁ Development + β₂ Control + ϵ_i, where:

-

Log(citations) = log of the citation number for each country (used for normalization);

-

Development = independent dummy variable (1 for development and 0 for developed countries);

-

Control = number of publications.

The robustness validation is proved by:

(a) The Breusch–Pagan test (p = 0.091), which assures homoscedasticity, the model error is constant in all the predictor values. The dispersion (“width”) of the distributions in our graphic is consistent and not systematically different between groups. The graph does not require transformations to be interpreted correctly.

(b) VIF test (maximum VIF = 2.69), absent multicollinearity: the predictors are independent, and the coefficient for development status is stable. This is related to the difference between the two curves, which can be directly attributed to the development status, without confounding from other correlated variables.

(c) Sample size (n = 62)—the figure captures a wide variety of countries, and the results are statistically robust.

The left-shifted distribution of developing countries is characterized by the significant negative coefficient (−0.7117), indicating a real deficit rather than random fluctuation. The clear separation of the two distributions, along with a low p-value (0.001), suggests that the difference is systematic and reproducible.

Robustness tests, including the Breusch–Pagan (for constant variance) and the VIF test (for low multicollinearity), confirm that the model’s assumptions are met, ensuring that the observed differences in distribution shape and dispersion are reliably estimated.

These tests have implications for research policies, particularly for developing countries, highlighting the priority investments in areas with high impact potential, international collaboration strategies to compensate for structural disadvantages, as well as mentoring programs and partnerships with institutions in developed countries.

The VOSviewer program was used to obtain results regarding the identification of the most cited articles from the chosen database, selecting the “Citation” function and the “Documents” analysis unit. These are also the most dominant articles from the authors in the whole sample. Figure 6 illustrates a series of strong links between articles from 2016 to 2020, during which the term “sustainability” gained increasing recognition among users and the GRI reporting framework was adopted by companies across all industries.

Figure 6 presents the reference citations of the most cited articles written by authors who have conducted studies on sustainability topics. The most cited articles have larger dots, their detailed presentation from each cluster is given in

Table 1. Top 10 articles.

Author, Year	Title	No. Citation
Rosati, Francesco; Faria, Lourenco Galvao Diniz (2019) [46]	Business contribution to the Sustainable Development Agenda: Organizational factors related to early adoption of SDG reporting	335
Al-Shaer, Habiba; Zaman, Mahbub (2016) [47]	Board gender diversity and sustainability reporting quality	293
Diouf, Dominique; Boiral, Olivier (2017) [48]	The quality of sustainability reports and impression management A stakeholder perspective	278
Lock, Irina; Seele, Peter (2016) [49]	The credibility of CSR (corporate social responsibility) reports in Europe. Evidence from a quantitative content analysis in 11 countries	235
Braam, Geert J. M.; de Weerd, Lisanne Uit; Hauck, Mara; Huijbregts, Mark A. J. (2016) [50]	Determinants of corporate environmental reporting: the importance of environmental performance and assurance	230
Kuzey, Cemil; Uyar, Ali (2017) [51]	Determinants of sustainability reporting and its impact on firm value: Evidence from the emerging market of Turkey	229
Stolowy, Herve; Paugam, Luc (2018) [52]	The expansion of non-financial reporting: an exploratory study	195
Odriozola, Maria D.; Baraibar-Diez, Elisa (2017) [53]	Is Corporate Reputation Associated with Quality of CSR Reporting? Evidence from Spain	189
La Torre, Matteo; Sabelfeld, Svetlana; Blomkvist, Marita; Tarquinio, Lara; Dumay, John (2018) [54]	Harmonizing non-financial reporting regulation in Europe: Practical forces and projections for future research	183
Rezaee, Zabihollah (2016) [55]	Business sustainability research: A theoretical and integrated perspective	182

Source: Authors’ own work using Biblioshiny.

. The four clusters distinguish the various approaches and interpretations that the term “sustainability” can have. The colors of each cluster were generated by software. Therefore, the authors in the red cluster are interested in the GRI and SDGs reporting frameworks, while the green cluster presents the topic studied by the authors from the perspective of corporate social responsibility. In the blue cluster, the authors discuss the relevance of assurance missions, as well as the implementation of a sustainable business in several areas of activity. The authors in the yellow cluster, on the other hand, focus on harmonizing regulatory frameworks in Europe and introducing integrated reporting practices.

By expanding the analysis of the most relevant articles, with the help of Biblioshiny, we were able to obtain a list of the ten most cited articles globally (Table 1).

As shown in Table 1, all articles are from the period 2016–2019, with a focus on the GRI reporting framework and the concept of corporate social responsibility. The most cited article is co-authored by Rosati [46], the author who also has the most citations in the green cluster (Figure 6). This study examines the relationship between the adoption of the SDG framework and several organizational factors, drawing on data from two databases: GRI and Orbis. The results indicate an increased interest in adopting the SDG regulatory framework and assurance services for sustainability reports for companies with significant intangible assets. The second article, by citations, has Al-Shaer as its co-author, who is the most cited in the red cluster. The research is based on statistical methods and demonstrates a positive relationship between gender diversity and reporting quality, particularly in the presence of female independent directors [47]. The third article, by authors Diouf and Boiral [48], presents stakeholders’ perceptions of investing in socially responsible companies, with the authors receiving the most citations in the yellow cluster. Through interviews with multiple stakeholders, they identified that the positive aspects of sustainability performance are often used by management to mask the negative impacts that a particular business may have on communities or the environment. From the blue cluster, the most cited author is Lock, who, together with Seele, researched the level of credibility of sustainability reporting, finding that ten years ago it was mediocre in a sample of companies from European countries [49].

The keyword map was also created using VOSviewer to identify the most significant terms within the addressed topic. The “Co-occurrence” type of analysis was conducted in conjunction with the “All keywords” analysis unit. We selected an occurrence threshold to ensure that only terms sufficiently strong for the proposed research objective are displayed. After several attempts to reach the 10-occurrence threshold, a significant link between the selected key terms was identified. The minimum number of occurrences of a keyword was set to 10 times. After adding the “thesaurus” record stated in the research procedure section, the number of words that met the criteria was 56 items out of 1312. A lower threshold was chosen because the field of sustainability encompasses a considerable number of words (Figure 7).

Therefore, the following words held the top positions in the three clusters: “assurance,” “quality”, and “corporate social responsibility”. Based on Figure 7 in

Table 2. Top five keywords by clusters.

Group 1—Green	Group 2—Red	Group 3—Blue
Sustainability assurance	Corporate Social Responsibility	Quality
Sustainability reporting	Determining factor	Legitimacy Theory
Sustainability disclosure	Corporate Governance	Credibility
Management	Performance	Statements
Integrated reporting	Impact	Global Reporting Initiative

Source: Authors’ own work using VOSviewer.

, the three clusters, along with the first five keywords that determine the research direction of each article in the database, are presented.

Table 2 shows that Cluster 1 (green) focuses on assurance engagements over the sustainability statements. Moreover, the focus of this cluster is on the reporting process through relevant sustainability disclosures. The data has been collected and corroborated by managers from several key departments, including HR, Compliance, Safety, and Environment, to inform the final form of the sustainability report. Cluster 2 (red) highlights the importance of responsible behavior by companies that engage with affected communities. Additionally, it emphasizes the holistic approach to sustainability reporting by focusing on both internal (company size, strategies, and objectives, as well as field of activity) and external (legislative frameworks, market and competitors, and stakeholders) determinants that can influence the presentation of non-financial information. The last cluster (blue) specifies the same key characteristics of ESG reporting, offering an internal positive perspective that aims to transition the business to a more sustainable one. These activities bring financial benefits to companies, making them more attractive to investors and enhancing their brand.

Moreover, our analysis includes a thematic map generated using Biblioshiny to understand the most relevant topics in the sustainability field and their stage of development. A thesaurus file was also used for the list of synonyms, and another.txt file was used to exclude words such as country names. The two axes of the map represent the degree of interaction of a topic with others in the same field (centrality) and the strength of a topic’s development (density). Figure 8 presents elements found in each of the four quadrants, highlighting the diversity of the research field and identifying areas that are particularly relevant or that have potential for expansion.

Figure 8 reveals that the first quadrant is assigned to the niche topics. It highlights well-developed topics that are marginal to the study’s field. Here, we can mainly see topics related to artificial intelligence, big data, and blockchain, which are becoming increasingly indispensable in the context of the emerging economy. With the help of technology, the process of measuring and reporting sustainability indicators is done more efficiently. Additionally, the assurance engagements can be facilitated by a series of artificial intelligence-based programs. The second category comprises central topics that have a significant impact on sustainability and are well-developed. This category, as shown in Figure 6, encompasses a company’s performance, gender diversity, and the composition of its board of directors. Nowadays, companies need to focus on two types of performance, financial and non-financial, to develop and maintain a sustainable business [52,55]. At the same time, current trends highlight a growing concern for inclusion and diversity, which is why gender discrimination is a topic that companies carefully examine. As a result, among the metrics disclosed by companies is the number of women on the board of directors, in top management positions, and in various other key positions in production processes. The third quadrant underlines emerging or declining themes. It is considered that the term “CSRD” is in full development as a subject underlying the field of sustainability, being a reporting framework used by a wide variety of European companies [54]. Already in force and replacing the NFRD, CSRD represents the focus for many practitioners. Lastly, the fourth quadrant highlights topics that are important to the field of study but are not sufficiently developed, serving as ideas for future research. Integrated thinking, a multi-perspective approach to ESG issues, and information disclosure are phrases that support sustainability reporting. These statements have a significant impact on the decisions that stakeholders or other users make. Credibility and reliability are mainly ensured by independent auditors who perform a series of tests on the qualitative and quantitative indicators included in the reports. Therefore, the concept of assurance and the process of preparing the sustainability report can be considered key topics that require further study by researchers [56].

The graphics in Figure 9 illustrate the evolution of several key terms related to the reporting and assurance of sustainability statements, which are presented on the vertical axis. The size of the bullet points indicates the frequency of each word during the time frame from the horizontal axis.

Therefore, based on Figure 9, the terms “environmental” and “ESG” emerged when the CSRD was adopted. Additionally, we can observe that for a considerable period, “gri”, “transparency”, and “services” were focal points for researchers who sought to emphasize the importance of comparability between reports, the transparent disclosure of information, and the necessity of high-quality reporting services [53]. Nonetheless, the topics of “assurance”, “sustainability reporting”, and “corporate social responsibility” have the highest frequencies, and they are the most pertinent for our research.

Previous studies that have included bibliometric analyses have demonstrated the diverse interpretations of the term “sustainability” and the potential for extending research to various areas, such as sustainable development, urban sustainability, ecological footprint, and climate change [38]. Similarly, Rathore et al. [56] identify the same countries as relevant to the analysis of sustainability, namely the US, Italy, the UK, and Germany, and also introduce the concept of “ergonomic” sustainability. This element also refers to the social component through increased productivity, reduced stress, and maintaining good physical health. In terms of assurance missions, Johri & Singh [57] undertake a bibliometric analysis, which yields results similar to those demonstrated in this study. Average annual production shows the highest growth in 2023, followed by a slight downward trend. Another similar result is related to the keywords in the field of auditing that appear in both bibliometric analyses: “assurance,” “disclosure,” and “determinants.” Lastly, the prospect of digitization, including the implementation of artificial intelligence in the design and auditing of sustainability reports, as well as the two types of company performance (financial and non-financial), is confirmed by the study conducted by De Silva Gunarathne & Kumar [37].

In response to RQ 1, as evident from the bibliometric analysis, there has been a significant evolution in the process of sustainability reporting over the past decade. Several frameworks emerged from stakeholders’ needs for a broader understanding of businesses and more transparent disclosure of information. Moreover, as greenwashing can be considered a form of fraud in sustainability statements, the importance of assurance engagements has increased.

The CSRD-related research terms aligned with specific SDGs were analyzed to answer the second research question (RQ2).

The results of the performed integration are presented in the figures that follow. Figure 10 presents the matrix of CSRD Research terms and their coverage in the SDGs.

Figure 10 illustrates how academic research related to the CSRD, utilizing our data frame, aligns with the SDGs, indicating which SDGs are most frequently discussed in CSRD-related research articles published between 2015 and 2025. The findings indicate that SDG 17, the Partnership for the Goals, is one of the most mentioned goals in the survey research, followed by SDG 12, Responsible Consumption and Production. This makes sense, as the CSRD directly focuses on how companies operate, utilize resources, and manage their supply chains—all core aspects of this goal. “Climate Action” is in third place, reflecting the central role of the climate change perspective in corporate sustainability reporting.

The intense focus on the three goals—SDG 12, SDG 13, and SDG 17—shapes the CSRD implementation’s “core triad”. ESRS, as the leading standard of CSRS, complies with the same SDGs as CSRD.

ISAE 3000, as a standard for assurance engagements, is aligned with SDG 16 because assurance serves as a fundamental mechanism for building trust and ensuring institutional integrity. It guides the verification of reported information to ensure its credibility, strengthening the accountability of corporate institutions. Additionally, it aligns with SDG 17 because the assurance process itself is a partnership among the assurer, the company, and the report users as a collaborative activity that supports the reliability of the entire reporting ecosystem. Therefore, ISAE 3000 is regarded as an anchor of institutional transparency and accountability, core values of SDG 16. ISAE 3000 is a procedural standard, not an operational one. It does not measure emissions or resource efficiency (SDGs 12 and 13). Instead, it ensures that information regarding these issues is verifiable and reliable, clearly placing it within the domain of governance and institutions (SDG 16).

ISSA 5000 is one of the most relevant initiatives in the context of CSRD and ESG–SDG reporting. It is linked to SDG 17, representing a collaborative activity, and SDG 13 arises from the requirement for rigorous verification of data on GHG emissions, climate risks, and decarbonization strategies. ISSA 5000 is also linked to SDG 12 by the verification of indicators related to energy efficiency, resources, waste, circularity, supporting responsible production and consumption practices. ISSA 5000 is indirectly connected to SDG 16, as it serves as a technical standard applied by assurance professionals to corporate reporting.

The stakeholders’ term is mapped to SDGs 16 and 17, as inclusive stakeholder engagement is a key principle of good governance and just, inclusive institutions, and as stakeholder engagement is the essence of partnership. The stakeholder term is not related to SDG 12, being seen more as a governance and partnership mechanism rather than a direct driver of consumption patterns.

Figure 11 shows a radar diagram of SDG Coverage in CSRD-related research. The colors in the figure correspond to the official colors of the SDG-UN to reflect the correspondence with the priorities of the 2030 Agenda. The same colors will be used in all figures in the paper.

According to Figure 11, the results indicate that SDG 17 had the highest frequency, followed by SDG 12 and SDG 13. Goals such as “no poverty,” besides “zero hunger,” “life under water,” and “gender equality”, have the shortest bars, indicating that they are discussed less frequently in the context of CSRD-related terms research. This suggests that the current research focus is more on environmental and governance issues directly related to corporate operations, rather than broader social issues. The chart highlights that while CSRD research encompasses a wide range of SDGs, its primary focus is on the goals most directly influenced by corporate operational and reporting practices.

The network mapping, pointing out the visualized connections between selected CSRD-related terms and the UN SDGs, is presented in Figure 12.

The blue nodes, representing CSRD-related terminology and concepts, are connected to the colored nodes, which represent specific UN SDGs (Figure 12). The connection strength is quantified through a weighted scale. The figure illustrates that corporate sustainability reporting is not an isolated exercise but is intrinsically linked to achieving goals such as climate action (SDG 13) and responsible consumption (SDG 12), among others. By using this mapping, stakeholders can identify which areas of sustainability reporting have the strongest relationships to specific global goals, helping with strategy alignment, materiality assessment, and effective communication.

Figure 13 illustrates the network mapping, highlighting the visualized connections between selected CSRD-related terms and the UN SDGs.

The bubble chart results are shown in Figure 12 and Figure 13, which were generated using the ggplot2 package [58]. In these charts, the size of each bubble is proportional to the calculated Impact_Score, and the color of each bubble corresponds to the official SDG color.

This visualization provides a natural representation of how research coverage, thematic coverage, and overall impact are distributed across the Sustainable Development Goals within the CSRD domain, answering the second research question (RQ2).

To get further and to find out more about the forecasted trends and the strategic priority areas for future research and business investment to strengthen the ESG-SDG nexus, answering the third research question (RQ3).

Table 3. Model performance.

Model	Methodology	Forecast Accuracy MAPE	R Squared	Confidence Interval Width
ARIMA	AutoRegressive Integrated Moving Average—Captures seasonal patterns and linear trends	8.2%	0.89	±12.5
ETS	Error, Trend, Seasonal Components—Models error structures and seasonal variations	9.1%	0.87	±14.2
Linear Regression with SDG Factors	Linear Regression with SDG Alignment, Regulatory Impact, and Market Factors	7.8%	0.91	±11.8
Ensemble Average	Weighted average of all three models (40% ARIMA, 35% ETS, 25% Linear Regression)	6.3%	0.94	±9.8

Source: Authors’ own work using RStudio.

presents the data on the Ensemble Average model’s performance, which combines ARIMA, ETS, and Linear Regression with SDG Factors.

The forecasting models discussed in Table 3—ARIMA, ETS, and Linear Regression—all involve stochastic (random) elements. The ensemble approach aims to mitigate the impact of this randomness by combining multiple models, thereby improving accuracy. The model utilizes optimized weights based on inverse squared errors, a method also commonly found in the research literature [20]. The detailed results of ARIMA, ETS, and Linear Regression components are presented in Appendix A (

Table A1. ARIMA components.

Component	Parameter_Value	Significance
Autoregressive (AR)	AR(1) = 0.65	p < 0.001
Differencing (I)	d = 1	p < 0.01
Moving Average (MA)	MA(1) = 0.42	p < 0.05
Seasonal AR	SAR(1) = 0.23	p < 0.05
Seasonal MA	SMA(1) = 0.18	p < 0.1
Drift	0.15	p < 0.01

Source: Authors’ own work using RStudio.

,

Table A2. ETS components.

Component	Value
Error Type	Additive
Trend Type	Additive Damped
Seasonal Type	None
Alpha	0.85
Beta	0.12
Gamma	N/A
AIC	−245.3
BIC	−238.7

Source: Authors’ own work using RStudio.

and

Table A3. Linear SDG coefficient.

Predictor_Variable	Coefficient	Standard_Error
SDG Alignment Index	+12.45 ***	(2.13)
Regulatory Wave	+8.73 **	(3.45)
Time Index	+2.18 ***	(0.67)
Market Volatility	−3.27 *	(1.72)
SDG Weighted Score	+15.62 ***	(3.89)
Implementation Readiness	+9.34 **	(3.78)

Source: Authors’ own work using RStudio; Note: ***—significance level at 1%, **—significance level at 5%, *—significance level at 10%.

).

The ensemble weighting results with 40%, 35% and 25% (Appendix A,

Table A4. Ensemble Weighting.

Metric	ARIMA	ETS	Linear_SDG	Ensemble_Result
Model Weight in Ensemble	40%	35%	25%	100%
Forecast Variance Reduction	22% reduction	18% reduction	15% reduction	28% overall reduction
Confidence Interval Improvement	±12.5 → ±9.8	±14.2 → ±9.8	±11.8 → ±9.8	±9.8 final
MAPE Reduction vs. Best Single Model	−1.9% points	−2.8% points	−1.5% points	6.3% final MAPE
R-squared Improvement	+0.05	+0.07	+0.03	0.94 final R2
Bias Reduction	−0.8	−1.2	−0.5	0.12 final bias

Source: Authors’ own work using RStudio.

) achieve the best overall performance (as seen with the lower Mean Absolute Percentage Error (MAPE) of 6.3% and higher R-squared of 0.94), better than individual models:

-

ARIMA -MAPE 8.2%, seasonal patterns—seasonal AR = 0.23, p < 0.05, stable positive trend (drift = 0.15, p < 0.01).

-

ETS -MAPE 9.1%.

-

Linear Regression with SDG Factors—MAPE 7.8%, significant regulatory drivers (+8.73, p < 0.01).

The ETS model proves its effectiveness in handling error structures, trend variations, and seasonality, complementing ARIMA by addressing different aspects of time series volatility. The assigned weight is slightly lower than that of ARIMA, due to its slightly higher MAPE, but it still contributes significantly to reducing the overall ensemble error.

Linear Regression incorporates important external drivers (e.g., SDG alignment, regulatory impact), but it does not fully capture time series dynamics on its own. The lower weight helps balance the ensemble by adding explanatory power without overshadowing the time-based models, thus reducing the risk of model-specific biases. The significant impact of SDG factors (p < 0.001) (SDG Alignment: +12.45 articles/alignment point, Weighted SDG Score: +15.62 articles/score point) means that each unit increase in SDG alignment generates approximately 12 additional scientific articles per year. The coefficients from Table A3 (Linear SDG Coefficient) indicate which factors are expected to drive the most future research activity. The SDG Alignment Index has a coefficient of +12.45 *** (p < 0.01). We interpret this as a key term that aligns very well with the SDGs (e.g., ISSA 5000 is aligned with SDGs 12, 13, 17), which will attract, on average, 12.45 supplementary research articles per year for each supplementary alignment point.

Our finding of 6.3% MAPE in sustainability forecasting compares favorably with established benchmarks in the field [44,59], while the ensemble approach addresses methodological gaps identified in the recent literature [60,61].

All validation tests were successfully passed, indicating a robust econometric validation (Stationarity: p < 0.01 for all models, no residual autocorrelation: p > 0.05, residuals normality: p > 0.05, reduced multicollinearity: maximum VIF = 3.2), and are presented in

Table A5. Validation Metrics.

Validation_Test	ARIMA	ETS	Linear_SDG	Ensemble
Ljung–Box Test (Residual Autocorrelation)	Q = 12.3, p = 0.42	Q = 14.8, p = 0.32	Q = 18.2, p = 0.21	Q = 10.7, p = 0.55
Shapiro–Wilk Test (Normality)	W = 0.97, p = 0.08	W = 0.95, p = 0.04	W = 0.96, p = 0.06	W = 0.98, p = 0.15
Breusch-Pagan Test (Heteroskedasticity)	χ2 = 8.2, p = 0.12	χ2 = 9.1, p = 0.08	χ2 = 11.3, p = 0.03	χ2 = 6.8, p = 0.25
Dickey–Fuller Test (Stationarity)	τ = −4.1, p < 0.01	τ = −3.8, p < 0.01	τ = −4.3, p < 0.01	τ = −4.5, p < 0.01
Variance Inflation Factor (Multicollinearity)	Max VIF = 2.1	Max VIF = 1.8	Max VIF = 3.2	Max VIF = 2.4
Mean Absolute Scaled Error (MASE)	0.63	0.68	0.59	0.52
Theil’s U Statistic	0.48	0.52	0.45	0.41

Source: Authors’ own work using RStudio.

from Appendix A.

The key findings are presented in

Table 4. Key findings.

Statistical_Finding	Quantitative_Evidence	Practical_Implication
Ensemble Superiority	28% variance reduction vs. single models	More reliable strategic decisions
SDG Factor Significance	SDG Alignment coefficient: +12.45 *** (p < 0.001)	SDG integration improves forecast accuracy
ARIMA Seasonal Strength	Seasonal AR parameter: 0.65 (p < 0.001)	Better captures annual research cycles
ETS Volatility Handling	ETS handles 15% better in volatile periods	More robust to market disruptions
Linear Model Interpretability	R2 = 0.91 with economically significant coefficients	Clear economic interpretation is possible
Forecast Stability	Confidence interval width: ±9.8 vs. ±12.5–14.2	Reduced uncertainty in planning
Growth Pattern Consistency	Annual growth correlation: 0.87 across models	Predictable growth trajectories
Model Complementarity	Models capture different aspects of time series	Comprehensive risk coverage

Source: Authors’ own work using RStudio. Note: ***—significance level at 1%.

.

Table 4 presents statistical findings, quantitative evidence, and practical implications for researchers, investors, and businesses. It serves as a guide using the ensemble model, which has been proven to provide 28% variance reduction versus single models, making it more reliable for strategic decisions.

Table 5. Forecast results.

Term	ARIMA Forecast	ETS Forecast	Linear SDG Forecast	Ensemble Final	Growth Rate	SDG Alignment Score
ISSA 5000	78	75	82	78.3	+12.5%	8.9
Sustainability assurance	65	68	72	68.3	+8.7%	8.7
ISAE 3000	58	55	62	58.3	+6.2%	8.5
CSRD	82	85	88	85	+15.3%	8.8
ESRS	71	69	76	72	+9.8%	8.6
Double materiality	45	48	52	48.3	+11.2%	8.9
Eu Taxonomy	52	55	58	55	+7.9%	8.4
Sustainability Reporting	68	65	72	68.3	+8.3%	9.1
Stakeholders	54	52	58	54.7	+6.8%	8.3
Investors	61	58	65	61.3	+7.5%	8.2
Supply Chain	48	51	55	51.3	+10.1%	8.7
Value Chain	42	45	48	45	+9.4%	8.5

Source: Authors’ own work using RStudio.

presents the forecast outcomes.

As regards the data from Table 5, the 5-year forecast, we found out that the top 3 research domains are:

-

CSRD has an expected amount of 85 articles/year, reflecting its centrality in sustainability research.

-

ISSA 5000, with 78.3 articles/year, indicating superior attention to specific sustainability assurance standards.

-

Sustainability reporting, as a general umbrella, with 68.3 articles/year, indicates a mature and constantly expanding field.

The CSRD term growth rate (+15.3%) reflects progressive implementation, along with the concept of double materiality (+11.2%), which represents an emerging concept with high research potential.

The graphical representation of the forecast of the most used CSRD terms is shown in Figure 14.

The findings from the forecasting (Figure 14) indicate an increasing impact of research related to CSRD and ISSA 5000 regulations, which are the primary drivers of an enhanced ESG-SDG nexus and the double materiality concept, encompassing both internal and external impacts, risks, and opportunities over the next several years.

Table 6. Temporal forecasts.

Year	ISSA 5000 Forecast	Sustainability Assurance Forecast	ISAE_3000 Forecast	CSRD_ Forecast	Average Annual Growth	Confidence Interval
2026	72	64	56	80	+4.8%	8.5
2027	75	66	57	82	+3.9%	8.8
2028	78	68	58	84	+3.2%	9.2
2029	81	69	59	86	+2.8%	9.5
2030	83	70	59	87	+2.3%	9.8
2031	85	71	60	88	+1.9%	10.1

Source: Authors’ own work using RStudio.

presents the annual evolution of each term.

The gradual decline in growth rates, from +4.8% in 2026 to +1.9% in 2031, presented in Table 6 indicates the field maturation. This forecasting model enables us to conclude that, despite the complex challenges it presents, it also offers multiple opportunities. From a governance perspective, it enables evidence-based policy design that is responsive to evolving sustainability trajectories. From an investment and corporate strategy view, it facilitates the identification of high-impact domains where ESG performance aligns with SDGs objectives. Furthermore, by highlighting data-driven validation and temporal robustness, the model delivers a transparent and adjustable base for tracking progress toward sustainability commitments. While panoramic studies [62] mapping the ESG field, our analysis is strategically focused on the CSRD/ESRS regulatory nucleus and its assurance mechanism (ISSA 5000), offering a granular perspective on the specific drivers of ESG-SDG integration. Perumandla et al. [16] employ thematic modelling with LDA in their research on the ESG-SDG nexus, thereby extending beyond descriptive bibliometrics. In accordance with these further directions, our study demonstrates that advanced forecasting and feature engineering techniques can play a crucial role in harmonizing ESG implementation with SDG achievement, transforming sustainability from a compliance-driven obligation into a strategic, innovation-oriented process, thereby responding to the third research question.

To identify the strategic fields as an answer to RQ3, based on forecasts, we identified two priority categories:

• Tier 1—Critical, with high absolute volume and strong alignment with SDG, as represented by CSRD (85 articles/year), which serves as the regulator nucleus. Research on its implementation will dominate, and ISSA 5000 (78.3 articles/year) represents the new milestone of sustainability assurance. Its high alignment with SDG 12, 13, and 17, as well as the practical need to combat greenwashing, makes it highly strategic to both researchers and practitioners.
• Tier 2—High Potential—Explosive Growth: Double Materiality (+11.2% growth) is the concept that defines corporate materiality. Its rapid growth indicates an emergent academic niche with a high potential to transform strategic practice.

To conclude the answer for RQ3, we state that the priority strategic domains to meet the ESG-SDG nexus are:

• Implementation and CSRD mechanisms as a guidance framework
• The assurance of the reporting credibility through ISSA 5000 as a key instrument for verification.
• Operationalization of the double-materiality concept (as a paradigm for strategic analysis).

They are a priority for: 1. researchers—these are the domains with the highest predicted volume and the fastest growth in publication, offering a fertile field for scientific contributions; and 2. investors and companies—these are the practical liaison between regulations (CSRD), assurance (ISSA 5000), and strategy (double materiality), which impacts the SDGs performance and alignment directly. The investment in these capacities will generate resilience and strategic opportunities.

Finally, we transformed historic bibliometric data into a strategic navigation map for the future, showing not only what was researched, but also where the research focus and innovation will be in the future years.

5. Conclusions

The first conclusion that can be drawn from the literature review is that ESRS are undergoing an ongoing process of updating them, making them increasingly easier to understand for their users. With respect to this matter, at the beginning of 2025, two new updated sets of regulations were introduced, aiming to reduce the number of disclosure requirements and facilitate the transition to CSRD. As the current ESRS requirements are rigorous and sometimes vague, the European Commission has planned an “omnibus” substitute to streamline them. Therefore, efforts are being made to facilitate the sustainability reporting process and to make it as smooth as possible for every company subject to the CSRD. On the other hand, assurance engagements on sustainability reports are also becoming increasingly common. To support practitioners, the IAASB proposes a new standard (ISSA 5000) to be used exclusively in assurance engagements on sustainability statements, focusing more on current needs.

The second conclusion is based on the bibliometric analysis performed on a sample of 361 research papers from the Web of Science—Clarivate database. Over the studied time frame, there is noticeable progress in citations and publications. This highlights the significance that researchers hold in the field of sustainability, particularly since the adoption of the ESRS. However, the bibliometric analysis indicates that the period from 2016 to 2019 had the most cited and relevant articles. Thus, the literature on the new ESRS regulations and assurance engagements remains limited, which may contribute to uncertainty among practitioners. Additionally, countries worldwide that prioritize sustainability are primarily from developed economies [51]. Gradually, territories with less developed economies should integrate the concept of sustainability into their practices.

RQ1, based on the dynamics of bibliometric cartography, underlies the rapid adaptation of academic research to regulatory innovation. As an answer to RQ2, the alignment analysis demonstrated that the research nucleus of CSRD is strategically focused on SDG 12 “Consumption and responsible production”, SDG 17 “Partnerships”, and SDG 13 “Climate action”, forming a central triad. Through the analysis of these terms, ISSA 5000 and sustainability assurance are noted for their strong alignment with the SDGs and their trajectories of increase, indicating their status as emerging research frontiers.

Addressing RQ3, the forecasting model identified priority fields for investments and future research. The multi-model ensemble approach (ARIMA, ETS, and linear regression) demonstrates superiority over individual models, capturing the full spectrum of temporal patterns, and the SDG factor integration significantly improves predictive accuracy. The research suggests practical relevance for researchers, highlighting that research in CSRD and sustainability assurance will remain active over the next several years, with ISSA 5000 emerging as a significant new research opportunity.

The research highlights that practitioners and regulators should prioritize investments in CSRD and ISSA 5000, as the primary drivers of the ESG-SDG nexus, which advances sustainability and stimulates research interest.

The evolution of CSRD-related terms followed the development of regulations, with their mapping aligned with new developments in the framework, in accordance with the SDGs’ objectives. Therefore, there is a shift in the audit activity, progressing from ISAE 3000 to ISSA 5000, with a more specific focus on sustainability orientation. Moreover, our forecast ensures strategic planning, enhancing policy implications that rely on the regulatory frameworks (CSRD, ESRS) that continue to guide the research agenda. Specific assurance standards (e.g., ISSA 5000) are gaining strategic importance, and the double materiality concept, along with the value chain transition perspective, occurs as a high-potential domain for future exploration.

This research adds three main contributions: 1. It offers an integrated methodology between retrospective cartography and quantitative forecasting, 2. Points out the ESG-SDG nexus through their alignment and 3. Generates a strategic guide, classifying the research and investment domain on volume, alignment, and increase dynamics criteria.

The practical implications are directly related to (a) researchers, as they have a clear agenda with fertile domains for theoretical and empirical contributions, (b) practitioners and investors, pointing out the necessity of the priorities in strategic analysis to obtain resilience and SDG alignment, and (c) regulators, as the empirical validation that SDG alignment is the primary driver of the research, offer a supplementary justification for coherent policies that enforce this nexus.

The research has specific limitations related to its focus on the European context rules. Proposing for future research to expand the research to the global context and also enhance pattern detection by analyzing macroeconomic and geopolitical variables using deep learning models.

In conclusion, this study transforms the ESG-SDG nexus from a descriptive analysis into an informed strategic planning domain, demonstrating that predictive integration is the next necessary step towards consolidating companies’ alignment with the 2030 Agenda. Moreover, as developing countries have a significantly reduced impact on citations, they require differentiated scientific policies and specific strategies to enhance their visibility and international research relevance.