NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Institute of Medicine (US) Forum on Microbial Threats; Knobler SL, O'Connor S, Lemon SM, et al., editors. The Infectious Etiology of Chronic Diseases: Defining the Relationship, Enhancing the Research, and Mitigating the Effects: Workshop Summary. Washington (DC): National Academies Press (US); 2004.

Cover of The Infectious Etiology of Chronic Diseases

The Infectious Etiology of Chronic Diseases: Defining the Relationship, Enhancing the Research, and Mitigating the Effects: Workshop Summary.

Show details


The traditional standards for establishing a microbial or bacterial cause of disease are those that were developed for acute infections. Known as “Koch's postulates,” they state that the causal organism must be:

  1. present in diseased tissue;
  2. isolated and grown in pure culture outside the animal host;
  3. shown to induce the same disease when injected into a healthy animal; and
  4. isolated from the experimentally inoculated animal in pure culture and shown to be the same as the original agent.

In almost all cases, identifying and confirming an infectious cause of a chronic disease using Koch's postulates is complicated by several factors, including:

  • Disease etiology may be multifactorial, including environmental, host genetic, and microbial genomic factors. Michael Dunne (Chapter 1) surveys a number of pathogens proposed to contribute to atherosclerosis and cardiovascular disease. Chlamydia pneumoniae, cytomegalovirus, and herpes simplex virus are among the suggested bacterial and viral pathogens, with the greatest body of evidence surrounding C. pneumoniae. Yet in all of these cases, when these agents are considered in the context of well-established risk factors for cardiovascular disease—family history, high-fat diet, inactivity—it is less clear how much infection would truly contribute to the condition and the outcome. Eduardo Franco (Chapter 1) describes the association that has been found between human papillomavirus and cervical cancer. Even years after discovery of this link, however, questions remain about the roles of cofactors, as only some of the many people infected develop malignancy.
  • Microorganisms may act in a hit-and-run fashion, striking and then disappearing from the host by the time the disease process becomes apparent. This form of attack appears to be the case in Reiter's syndrome, Guillain-Barré syndrome, and rheumatic heart disease. As another example, Robert Yolken and Fuller Torrey (Chapter 1) report that a retrospective study of schizophrenics found that their mothers' blood at the time of birth exhibited elevated IgG, IgM, and certain cytokines, suggesting that an ongoing inflammatory process may possibly be associated with Toxoplasma infection. Antibodies directed against endogenous retroviruses, including Herv-W, have also been found at elevated levels. Infection in the perinatal period may have set the stage for later neurological disease, although there may be little or no evidence of active infection at the time of diagnosis.
  • Acute, chronic, latent, or recurrent infections may be involved in pathogenesis, and coinfections may play a critical role in disease manifestation. Richard Johnson (Chapter 1) describes postinfectious encephalomyelitis in which patients develop fever, become obtunded, and develop multifocal neurological signs several days after resolution of an acute rash caused by a virus. This is an example of an acute systemic disease with a postinfectious immune response leading to demyelination within the central nervous system. In other cases, a long period of active viral replication may precede the onset of disease. William Mason (Chapter 1) reports that the risk of developing chronic hepatitis B virus infection is greater than 90 percent when a person is infected at birth or in early childhood, but drops to less than 10 percent when a person is infected as an adult. In such chronic infections, the risk of fatal liver disease (cirrhosis or liver cancer) rises to approximately 25 percent, with a 30 year to 50 year interval between the onset of the infection and the consequent pathological outcome. A similar picture occurs with hepatitis C virus infection, although in this case there is virus persistence when the infection occurs in adulthood. Viral infections may also be latent at the time of diagnosis. For example, there are several viruses for which patients with multiple sclerosis (MS) exhibit higher antibody levels than control patients. Johnson reports on one study which revealed that 23 percent of MS patients had antibodies to two or more viruses present within their central nervous systems, with one patient presenting with 11 viruses. It is not yet clear whether infection(s) triggers MS or whether elevated markers of infection are secondary to the underlying inflammatory processes of the disease. Such findings emphasize the complexity of directly attributing chronic disease to one or more specific infectious agents.
  • Detecting and/or isolating microbes that are present in a variety of tissues may pose significant technical difficulties. Current methods to identify novel or rare microorganisms may be inadequate. During the workshop, David Persing reported on the deficiencies and weaknesses of conventional methods for identifying and subtyping microorganisms. However, newer molecular technology, such as broad-range amplification of ribosomal targets directly from tissue or culture, can complement conventional systems, and these tools have helped in identifying several new species and pathogenic subtypes. For example, the infectious agent strongly suspected of causing Whipple's disease remained elusive for years. Applying broad-range polynuclear chain reaction techniques enabled scientists to amplify and categorize the etiologic Tropheryma whipplei bacterium. Patrick Moore (Chapter 3) recounts the development of a technique called representational difference analysis to identify Kaposi's sarcoma-associated herpesvirus as a cause of AIDS-associated Kaposi's sarcoma. These discoveries exemplify the diligent effort required to move from identification of a new DNA sequence to confirming causality in a specific disease. During the workshop, Persing also described the potential for gene expression arrays (microarrays), proteomics, and other technologies to identify patterns of host response to an infection(s) that might explain the pathogenic processes from exposure to chronic disease and lead to the development of diagnostic tools for these entities. Phylogenetic analysis can relate new pathogens for which there are no effective diagnostic assays to known agents through conserved epitopes and other properties, facilitating the evaluation of new infectious causes of disease.

Given the various reasons why it may often prove difficult to satisfy Koch's postulates in linking a particular infectious agent to a particular chronic disease, alternative sets of criteria may need to be developed for determining causation. Such criteria must take into account the more complex relationships that are being observed between microbial agents and chronic disease, and they likely will require collection of more challenging types of experimental data, especially molecular data, that can help clarify discrete causal links. Toward this goal, several promising avenues of research are being pursued, including extending various genetic technologies and modifying animal and cell culture models of human disease to make them more immediately relevant to microbial disease causation.

Ensuing discussions highlighted gaps in scientific knowledge and in the translation of research data to health care interventions for both well-accepted and more speculative causal associations. Participants noted the complexity of these issues, as well as the importance of strengthening the critical linkages among clinicians, researchers, epidemiologists, and public health officials.

Copyright © 2004, National Academy of Sciences.
Bookshelf ID: NBK83697


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (4.4M)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...