The Lyme Disease Controversy: An AI-Driven Discourse Analysis of a Quarter Century of Academic Debate and Divides

Abstract: The scientific discourse surrounding Chronic Lyme Disease (CLD) and Post-Treatment Lyme Disease Syndrome (PTLDS) has evolved over the past twenty-five years into a complex and polarised debate, shaped by shifting research priorities, institutional influences, and competing explanatory models. This study presents the first large-scale, systematic examination of this discourse using an innovative hybrid AI-driven methodology, combining LLMs with structured human validation to analyse thousands of scholarly abstracts spanning 25 years. By integrating LLMs with expert oversight, we developed a quantitative framework for tracking epistemic shifts in contested medical fields, with applications to other content analysis domains. Our analysis revealed a progressive transition from infection-based models of Lyme disease to immune-mediated explanations for persistent symptoms. This study offers new empirical insights into the structural and epistemic forces shaping Lyme disease research, providing a scalable and replicable methodology for analysing discourse, while underscoring the value of AI-assisted methodologies in social science and medical research.

• The paper introduces a novel AI-driven methodology combining large language models and expert validation to classify 8,829 abstracts on Lyme disease.
• The paper reveals significant shifts in academic stances, with increased support for PTLDS post-2010 alongside persistent advocacy for CLD.
• The paper discusses practical implications for research funding, clinical practice, and health communication in addressing the Lyme disease controversy.

An AI-Driven Discourse Analysis on Lyme Disease Controversy

Introduction and Objectives

The paper "The Lyme Disease Controversy: An AI-Driven Discourse Analysis of a Quarter Century of Academic Debate and Divides" presents a rigorous exploration of the contentious scientific discourse surrounding Chronic Lyme Disease (CLD) and Post-Treatment Lyme Disease Syndrome (PTLDS) over the past 25 years. Utilizing an innovative hybrid AI methodology, the study systematically analyzes thousands of scholarly abstracts to track epistemic shifts within this polarized debate. The paper aims to provide a quantitative framework for understanding how Lyme disease research has evolved and suggests potential implications for policy and practice.

Methodology and Implementation

This research employs LLMs combined with expert validation for comprehensive content analysis. Through data acquisition and pre-processing, a robust dataset of 8,829 academic abstracts was curated, focusing on human studies relevant to the CLD/PTLDS controversy. The methodology consists of multiple classification steps:

1. Pre-Screening Classification: LLMs are leveraged to eliminate irrelevant abstracts, categorizing them based on their relevance to CLD/PTLDS.
2. Stance-Framing Classification: Abstracts are analyzed for implicit and explicit stances on PTLDS and CLD. This involves sentiment and frame detection where studies are classified into categories like "Supports PTLDS," "Supports CLD," or "Neutral."
3. Self-Reflection Classification: An advanced self-reflection prompting technique allows the LLMs to reassess initial outputs, enhancing classification accuracy.
4. Human Validation: Inter-Rater Reliability (IRR) analysis among multiple LLMs and human experts ensures classification consistency and validity.

   Figure 1: Overview of the steps comprising the proposed hybrid AI-driven content analysis methodology.

Key Findings

Evolution of Discourse

The research identifies significant shifts in the academic discourse:

• Increased Alignment with PTLDS: Post-2010, there is an observable increase in studies supporting PTLDS, reflecting a growing consensus around immune-mediated explanations for persistent symptoms.
• Persistent CLD Advocacy: Despite a decline, CLD-supporting research maintains a presence, indicating continuous advocacy within certain epistemic communities.
• Neutral Stance: A substantial portion of the literature remains neutral, underscoring ongoing uncertainties and the complex nature of Lyme disease diagnostics.

  Figure 2: Number of relevant studies by year (2000-2004) and their classification. Classifications consisted of Neutral (blue), Supports PTLDS (orange), and Supports CLD (green).

Thematic Structures

The thematic analysis reveals eight primary themes, including "Diagnostic Complexity and Uncertainty," "Therapeutic Controversies and Antibiotic Efficacy," and "Sociocultural and Ethical Factors." These themes highlight:

• Diagnostic Challenges: The ongoing difficulty in establishing clear diagnostic criteria and reliable biomarkers is a major barrier to consensus.
• Therapeutic Uncertainty: Debates over prolonged antibiotic treatments reflect deeper epistemological divides within the medical community.
• Sociocultural Dimensions: The influence of advocacy groups and media continues to shape public and scientific perceptions, adding layers of complexity to the clinical discourse.

  Figure 3: Top 20 journals by volume representing our dataset and depicting their potential bias. Values are the difference in percentage points between published papers classified as supporting PTLDS or CLD.

Discussion and Implications

The analysis provides critical insights into the structural dynamics of scientific debate, emphasizing the influence of editorial biases and journal specialization on the dissemination of research findings. The results suggest a need for:

• Equitable Research Funding: Balanced funding strategies are crucial to explore diverse scientific inquiries, particularly those challenging prevailing paradigms.
• Patient-Centered Care Models: Integrating patient narratives into clinical frameworks can enhance diagnostic and therapeutic approaches.
• Strategic Health Communication: Addressing public mistrust and improving communication strategies are essential for bridging the gap between scientific consensus and patient experiences.

Conclusion

This study underscores the complexity and evolving nature of the Lyme disease controversy. The novel AI-driven approach demonstrates the potential of combining computational techniques with social science frameworks to dissect intricate medical debates. The findings have significant implications for policymakers, clinicians, and researchers, advocating for a triangulated approach that integrates empirical evidence with patient-centered and sociocultural insights. This research sets a foundation for future studies to further explore contested conditions and utilize AI methodologies in discourse analysis.

Figure 4: Study classifications on Lyme disease from 2000 to 2025. The yearly count of abstracts on Lyme disease from 2000 to 2025 was classified into three classifications: Neutral, Supports PTLDS, and Supports CLD, which are depicted as blue, orange, and green, respectively.