John Snow, Asiatic Cholera and the inductive-deductive method - republished
Lecture 4: Competing theories on the origin and spread of infectious disease
Last week, we discussed what was known about the spread of Asiatic Cholera, Snow’s observation of the incubation period, and the problems with defining numerators and denominators. In today’s post, we will describe the competing theories on the origin and spread of infectious diseases that existed in Snow’s time. In the next post, we will learn about disseminated source outbreaks.
The Snow series is an educational course. We hope you will recognise our efforts by donating to TTE or becoming a paying subscriber, as writing the series took a lot of time and effort.
There were two main competing theories as to the origin and communication of cholera and other diseases:
Miasma Theory (Anticontagionism)
Many in the early to mid-nineteenth century felt that cholera was caused by bad air arising from decayed organic matter or miasmata. "Miasma" was believed to pass from cases to susceptibles in diseases considered contagious. Believers in the miasma theory stressed the eradication of disease through the preventive approach of cleansing and scouring rather than through the scientific approach of microbiology.
One prominent supporter of the miasma theory was Dr William Farr, then assistant commissioner for the 1851 census and a career employee of the government's General Register Office (his title was Compiler of Abstracts). For a while, Farr was convinced that cholera was transmitted by air. He reasoned that soil at low elevations, especially near the banks of the River Thames, contained much organic matter which produces miasmata. The concentration of such deadly miasmata would be greater at lower elevations than in communities in the surrounding hills. (The definitions are based on the ones on the John Snow website).
This is clearly represented in Farr’s diagram, which shows a correlation between cholera deaths and altitude (in feet). The higher you lived, the less likely you were to breathe in miasmata.
A vestige of miasma theory is traceable to the modern Italian term “malaria” or bad air, where swamps infested by decaying organic matter were thought to produce miasmata, which caused fevers. Many miasmatists were powerful and influential: Florence Nightingale, Lord Herbert, Sir John Simon, the Medical Officer for London at the time and the great sanitarian Sir Edwin Chadwick, first chair of the General Board of Health.
The historian Ackerknecht, in his landmark Fielding H Garrison Lecture, delivered in 1948, explains some of the socio-political background to the opposition to contagionism:
A variant of the miasma theory is the so-called zymotic (or fermentation) theory, to which both Farr and Simon had migrated from pure miasmatism. Refuse and rotting organic matter were considered substrates on which the poison acted to ferment and become active.
An important aspect of miasma theory was that either by fermentation or stagnation, the vapours were non-specific and could generate and carry all sorts of communicable diseases such as cholera, typhus, malaria and plague. The other consequence was that since they could do nothing about the poison, sanitarians concentrated on clean-ups and draining. Hence, the sanitary movement had a notable beneficial impact in later years, but for the wrong reasons.
He also pointed out the honesty and spirit of sacrifice of anticontagionists who, in “numerous self-experiments” between 1798 and Von Pettenkopfer’s 1892 self-ingestion of “cholerine” (see below), used their own bodies to disprove contagionism. The fact that few went on to develop the disease reinforced their refutation of contagionism.
Contagionism (later Germ Theory)
The alternative theory, supported by John Snow, held that cholera was caused by a specific germ cell not yet identified. He reasoned that this germ was transmitted from one person to another by drinking water and contact with soiled clothes. Gerolamo Fracastoro, a 16th-century physician, philosopher and polymath, first put forward the theory. Fracastoro theorised that infectious disease could be spread by “animaluncoli” or small animals, invisible to the naked eye, thus preceding Hassal, Pacini, Snow, Pasteur and Koch. Fracastoro posited a “contagium animatum” capable of replication.
Contingent Contagionism (sometimes known as localism or centrism)
As always, when two opposing theories (believed to be the truth by their respective proponents) exist, there is a halfway house, so-called contingent contagionism.
This thought that contagion was mediated by external factors like humidity, temperature, vapours, or contact with soil and other material. Contagion was one of the many factors or variables contingent on the onset of disease, so contingent contagionists did not refute the possibility of contagion; they simply considered it one of the many variables necessary for disease onset.
Localists included famous sanitarians like Parkes, Milroy, James Johnson and Riecke. Some miasmatists like Farr and Simon “migrated” to localism in the 1850s, only to subsequently embrace contagion theory. Perhaps the most famous contingent contagionist was the German hygienist Max Von Pettenkofer, who was among those who publicly drank a solution from rice water faeces containing “cholerine” to prove that the faeces were not contagious. He had loose stools from which vibriones were isolated but no cholera (see Lecture 10).
CONTEMPORARY THEMES
Can you spot the confounder in Farr’s diagram?
The confounder is the distance from the primary water source - the Thames. The higher districts of London were supplied by different water companies and had no or fewer river dwellers or workers, likely to draw their drinking water from the river rather than from wells.
Confounders are variables (say, age and smoking habits, but in this case, exposure to contaminated river water) which change the relationship between exposure and outcome. In Farr’s view, exposure was to miasma, and the outcome was cholera, having stratified his analysis by altitude in feet. This hypothesis was wrong as he had not considered the actual exposure - the distance from the Thames. Running in a valley or alluvial plain like all rivers do, the banks of the Thames would all be higher than the shoreline. Farr is considered one of the fathers of epidemiology. His work was meticulous and brilliant (see Lecture 13); still, at this stage in his life, he tried to fit his observations into miasma or zymotic theory, as Milroy and Parker did.
Learning objective: What methods can be applied to control potential confounding factors?
The most inferentially dangerous confounders are the ones which bear a relationship to exposure and to the outcome but are unknown and so cannot be adjusted for. As we will show in Lectures 7 to 10, casual allocation (randomisation) is the only process by which we can be sure that known and unknown confounders are spread evenly across the exposed and unexposed arms of a study, thereby nullifying their confounding effects. By sheer chance and hard work, Snow had hit on a “great” or “natural” experiment, which came near to mimicking a comparison based on random allocation.
If science were serious about ascertaining the mode of transmission of respiratory viruses, it would randomise parts of the community (or individuals, schools, homes, workplaces, etc.) to behaviour modification and other interventions and observe their effects on the influenza-like illness (ILI) syndrome with standard protocols ready in advance of the likely periods of maximum circulation. This would not necessarily test the mode of communication of a particular agent. Starting with the syndrome would do for starters. Then, layer by layer, like an onion, each agent and its peculiarities could be observed in repeated experiments. Rules of adjudication would also have to be agreed in advance. However, today, case definition and ascertainment are the main difficulties in dealing with two conditions as diverse as cholera and ILI.
A further consideration is that quarantines or isolation should not be applied in a blanket fashion to all infectious diseases and in all types of outbreaks. Quarantines were re-introduced in the 19th century to deal with epidemics of the “big three” (cholera, yellow fever and plague). Cholera, for example, had hit Europe four times: in 1831-2, 1848-49, 1852-55 and 1865-67. This protracted period of infection had caused great devastation, killing millions of people. In England and Wales alone, the estimated combined mortality was 150,000. Quarantines as a protracted period of closure do not make much sense in acute respiratory disease (see Lecture 6) with its rapid onset and equally rapid disappearance nor in short duration point source outbreaks (See Lecture 13). In fact, “influenza” epidemics were considered a periodic nuisance in the 1800s, not worthy of legislation.
Learning objective: Why is confounding so important in epidemiology?
“Confounding, the situation in which an apparent effect of an exposure on risk is explained by its association with other factors, is probably the most important cause of spurious associations in observational epidemiology.” [Elm E 2004]
Readings
Paneth N et al. A rivalry of foulness: official and unofficial investigations of the London cholera epidemic of 1854. Am J Public Health. 1998 Oct;88(10):1545-53.
Morabia A. Epidemiologic interactions, complexity, and the lonesome death of Max von Pettenkofer. Am J Epidemiol 2007;166:1233–8.
Ackerknecht EH. Anticontagionism between 1821 and 1867: The Fielding H. Garrison Lecture. 1948. Int J Epidemiol. 2009 Feb;38(1):7-21.
Elm E. The scandal of poor epidemiological research. BMJ. 2004 Oct 16;329(7471):868-9.
Susser M, Adelstein A. An introduction to the work of William Farr. Am J Epidemiol. 1975 Jun;101(6):469-76. doi: 10.1093/oxfordjournals.aje.a112117. PMID: 1098451.
Thanks, again, for this lecture.
For me, the most important sentence is "[Farr] tried to fit his observations into miasma or zymotic theory [...]". In my humble opinion - humble because I'm not a medical doctor - this sadly widespread attitude*) leads at best to the failure of explaining 'what is' by trying to fit it into 'what I think it should be'. At worst it leads to suppression of observed data which 'don't fit', or worse, to the falsification of the observed data.
*) For other, non-medical examples, simply take a look at the 'models' for 'climate emergency catastrophe' ...
Dear Both - you possibly have come across this posting already, but in case not I thought I should share it. Given your republishing of the excellent 'Transmission Series', is the study referred to below a revival of the MRC Cold Unit experiments but with more up to date diagnostic technology? Potentially, it appears it could well provide greater understanding in who gets COVID and who doesn't. Or, am I missing something? Certainly, I would value your comments. See the following links - https://theconversation.com/we-finally-know-why-some-people-got-covid-while-others-didnt-233063? Published: June 28, 2024 4.46pm BST. Authors. Marko Nikolic Principal Research Fellow/Honorary consultant Respiratory Medicine, UCL. Kaylee Worlock Postdoc Research Fellow, Molecular and Cellular Biology, UCL. The full results of the study can be found at https://www.nature.com/articles/s41586-024-07575-x