Cause (medicine)

Lore & Background

Traditional accounts of the causes of disease may point to the 'evil eye'. The Ancient Roman scholar Marcus Terentius Varro put forward early ideas about microorganisms in a 1st-century BC book titled On Agriculture. Medieval thinking on the etiology of disease showed the influence of Galen and of Hippocrates, with European doctors generally holding the view that disease was related to the air and adopting a miasmatic approach.

Reader's Guide

Etiological discovery in medicine has a history in Robert Koch's demonstration that species of pathogenic bacteria cause specific diseases, summarized in Koch's postulates. In epidemiology, Austin Bradford Hill demonstrated a causal relationship between tobacco smoking and lung cancer, summarized in the Bradford Hill criteria. The infectious diseases are caused by agents falling into five groups: viruses, bacteria, fungi, protozoa, and helminths. The term can also refer to a toxin or toxic chemical. Further thinking distinguished causation from association or statistical correlation, noting that events may occur together due to chance, bias, or confounding. Experimental evidence involving interventions gives the most compelling evidence of etiology. Sometimes several symptoms appear together as syndromes, assumed to have an underlying condition. Other times there is a chain of causation, as with peptic ulcer disease, which may be induced by stress, requires acid secretion, and has primary etiology in Helicobacter pylori infection. Some diseases are heterogeneous, with different etiologies producing the same syndrome, while a single etiology may produce different diseases. An endotype is a subtype defined by a distinct functional or pathobiological mechanism, distinct from a phenotype.

Did You Know?

• The word etiology is derived from the Greek αἰτιολογία (aitiología), meaning 'giving a reason for'.
• Where no etiology can be ascertained, the disorder is said to be idiopathic.

From Myth to Microbe – The Long Road to Etiological Understanding

The concept of disease causation has traveled a remarkably long path from superstition to science. Early human societies often attributed illness to supernatural forces such as the evil eye, reflecting a worldview in which disease was a moral or spiritual event rather than a biological one. A significant early step toward material explanation came from the Roman scholar Marcus Terentius Varro, who in his first-century BC treatise On Agriculture proposed the existence of invisible microorganisms as agents of disease. For centuries after, medieval European physicians leaned heavily on the frameworks of Hippocrates and Galen, and many adopted a miasmatic model that linked illness to corrupted air. The modern era of etiological discovery truly began with Robert Koch, who demonstrated that specific bacteria—Mycobacterium tuberculosis, Bacillus anthracis, and Vibrio cholerae—were responsible for tuberculosis, anthrax, and cholera respectively. His reasoning was later crystallized into what we now call Koch's postulates, a set of criteria for proving that a particular organism causes a particular infection. Yet even this powerful framework has limits: it applies most cleanly to individual cases where experimental evidence can be gathered, leaving broader population-level questions to the tools of epidemiology.

Proving Causation in a World of Correlation

One of the central challenges in medical etiology is separating genuine cause from mere coincidence. In epidemiology, two events can appear linked simply because of chance, measurement bias, or a hidden confounding variable, none of which implies that one event produces the other. Austin Bradford Hill addressed this problem head-on when he established a causal link between tobacco smoking and lung cancer. To systematize the reasoning, he formulated the Bradford Hill criteria—nine principles that collectively help researchers judge whether an observed association is likely to be causal. These criteria complement the more experimental logic of Koch's postulates, which, as noted, are strongest when individual cases can be experimentally manipulated. In practice, the most persuasive evidence for etiology comes from interventions: deliberately introducing or removing a suspected cause and observing whether the disease follows. Hill's framework was later incorporated into a broader proposal for a unified concept of causation, underscoring that establishing etiology in complex, population-level settings demands multiple converging lines of evidence rather than a single decisive test.

Chains, Co-Factors, and the Peptic Ulcer Lesson

Many diseases do not spring from a single trigger. Instead, a chain of causation links an initial event to the eventual clinical picture, and the etiological agent may depend on an independent co-factor or a promoter that amplifies its expression. A classic illustration is peptic ulcer disease, whose true cause was recognized surprisingly late. Stress can act as an inducing trigger, but ulcers also require the presence of gastric acid secretion, and the primary etiological agent is infection with Helicobacter pylori. None of these three elements alone is sufficient; together they form a causal chain that produces the disease. This multi-factorial view also helps explain syndromes—clusters of symptoms that consistently appear together even though no single symptom causes another. In such cases, clinicians infer an underlying condition that accounts for the whole pattern. The chain-of-causation framework is especially valuable for chronic diseases of unknown origin, where multiple epidemiological risk factors may be present, some genuinely causal and others merely associated. By mapping the sequence from trigger through co-factor to clinical disease, researchers can cut through a tangle of correlations and identify the true etiology.

One Disease, Many Causes – Heterogeneity and the Endotype Concept

Not every patient with a given diagnosis shares the same underlying cause. Conditions such as diabetes or hepatitis are defined syndromically by their signs and symptoms, yet they encompass multiple distinct diseases with different etiologies. These are termed heterogeneous conditions. The relationship can also run in the opposite direction: a single etiological agent, like the Epstein–Barr virus, may produce entirely different diseases depending on circumstances—mononucleosis, nasopharyngeal carcinoma, or Burkitt's lymphoma. To navigate this complexity, researchers distinguish between a phenotype, which is any observable trait of a disease (morphology, biochemical properties, behavior) without implying a mechanism, and an endotype, a subtype defined by a specific functional or pathobiological mechanism. Asthma, for instance, is increasingly viewed not as one disease but as a syndrome composed of several endotypes. Similarly, AIDS is understood as the clinical stage IV of HIV infection, where a single viral etiology unfolds through multiple stages. This layered vocabulary—phenotype, endotype, disease entity—gives clinicians and researchers a more precise map of what is actually causing a patient's illness.