The 3-level model of DISH: Post-publication reflections

Introduction

Diffuse idiopathic skeletal hyperostosis (DISH) has traditionally been defined as a structural disorder characterised by progressive ossification of spinal ligaments and entheses. Although this radiographic description has remained largely unchanged for decades, accumulating evidence suggests that the biological processes underlying DISH extend well beyond the axial skeleton and involve a complex interplay of metabolic, endocrine, inflammatory and osteogenic mechanisms. The recognition of these systemic associations has progressively challenged the classical view of DISH as a homogeneous, slowly progressive condition affecting predominantly older men.

One of the major difficulties in DISH research is the remarkable heterogeneity observed among individuals who share apparently similar clinical and metabolic backgrounds. Some patients exhibit extensive axial ossification with relatively limited vascular involvement, whereas others develop severe vascular calcification despite modest spinal progression. Likewise, a subset of individuals with substantial cardiometabolic burden never develops radiographic DISH despite prolonged follow-up. These observations raise a fundamental question: why do individuals exposed to apparently similar biological pressures develop markedly different structural outcomes?

The answer may lie in the distinction between biological susceptibility, biological activation and structural expression. Although these concepts are frequently discussed together, they are rarely analysed separately. Consequently, factors that predispose to disease, factors that actively drive tissue mineralisation, and the final structural manifestations are often interpreted as if they belonged to a single biological continuum.

The present framework emerged from a progressive research programme conducted within the Camargo Cohort. Rather than being conceived a priori, it evolved through a sequence of observations that gradually revealed different layers of biological organisation.

The first step was the identification of a subgroup of individuals displaying unusually rapid radiographic progression. These subjects, termed Fast Ossifiers (FO), challenged the prevailing notion that DISH invariably develops over several decades. Longitudinal analyses demonstrated that some individuals could progress by multiple Schlapbach grades within a five-year interval, accompanied by distinctive metabolic and skeletal alterations. Importantly, FO was not merely characterised by accelerated ossification. The phenotype also exhibited early deterioration of trabecular microarchitecture, increased insulin resistance, endocrine abnormalities and sex-specific biological signatures. These findings suggested that accelerated ossification represented only one manifestation of a broader systemic process rather than an isolated skeletal phenomenon.

The second step involved a reconsideration of the role of propensity score methodology. Initially developed as a statistical balancing tool, the propensity score unexpectedly demonstrated a remarkable ability to identify individuals sharing a common metabolic background associated with DISH susceptibility. Subsequent analyses showed that individuals occupying the highest propensity score strata concentrated not only established DISH cases but also future Fast Ossifiers and other high-risk phenotypes. This observation led to a conceptual shift. Rather than viewing the propensity score solely as a matching instrument, it became possible to interpret it as an operational projection of a latent biological susceptibility space. This concept subsequently led to the development of the Fast Ossifier Stratification Index (FOSSI), a sex-specific prognostic tool designed to quantify susceptibility to accelerated progression.

The third step was the recognition that susceptibility alone could not adequately explain phenotypic diversity. Individuals with similarly elevated FOSSI values frequently developed divergent outcomes. Some became Fast Ossifiers, others accumulated extensive vascular calcification, and some remained free of axial ossification despite prolonged observation. These discrepancies suggested that an intermediate biological layer existed between susceptibility and structural disease. To capture this component, the Downstream Activation Index (DAI-5) was developed as a multidimensional representation of biological activation integrating insulin resistance, inflammatory balance, endocrine-mineral regulation and bone turnover. Within this framework, DAI-5 serves as an operational representation of what we term the calcific impulse: the integrated biological pressure favouring tissue mineralisation.

The integration of these observations resulted in the proposal of a hierarchical three-level model. In this framework, Level 1 represents latent susceptibility, Level 2 represents downstream biological activation, and Level 3 represents structural trajectories. Importantly, these levels should not be interpreted as independent compartments but as interconnected layers within a dynamic biological system. Susceptibility creates the biological environment in which activation develops. Activation determines the intensity of the calcific impulse. Structural trajectories represent the tissue-specific resolution of that impulse.

A key conceptual feature of the model is the distinction between environment and phenotype. Susceptibility and activation define a biological landscape characterised by different combinations of metabolic burden and osteogenic activity. Structural trajectories emerge within this landscape rather than defin8ing it. Consequently, Fast Ossifier (FO), Low-Load DISH (LLD), Vascular-Dominant DISH (VDD) and Persistent Non-DISH (PND) should not be interpreted as fixed disease categories. Instead, they represent alternative biological configurations observed during a ten-year period of longitudinal follow-up.

Among these trajectories, Persistent Non-DISH may be particularly informative. Individuals belonging to this group exhibit many of the biological features traditionally associated with DISH susceptibility, yet fail to develop axial ossification. This apparent contradiction suggests that high susceptibility is necessary but not sufficient for structural disease. PND therefore introduces the concept of biological uncoupling and provides one of the strongest arguments for separating susceptibility from activation and activation from structural expression.

The purpose of the present review is not to present a definitive mechanistic explanation for DISH. Rather, it aims to integrate the findings generated across the Camargo Cohort research programme into a coherent organisational framework capable of generating testable hypotheses. By distinguishing susceptibility, activation and structural expression as separate but interconnected levels of biological organisation, the model seeks to provide a plausible explanation for the extraordinary heterogeneity observed in DISH and related calcific disorders.

───

STRUCTURE

Document disclaimer

Introduction

Part I. Level 1: Latent Susceptibility – From Propensity Score to the Metabolic–Osteogenic Space

Part II. Level 2: Downstream Biological Activation – The Calcific Impulse

Part III. Level 3: Structural Trajectories – Coupling and Tissue-level resolution

Part IV. Figure 6: The Integrated Biological Landscape of DISH

Part V. What the Model Contributes: From Structural Disease to Biological Organisation

Part VI. Strengths, Limitations and Future Directions

Part VII. Conclusions: A Working Framework for Biological Heterogeneity in DISH