Under discussion this week is an article by Ralph McBurney. Not Ralph McBurney (1902-2009) the centenarian beekeeper, but Ralph McBurney (1883-1964) the professor of bacteriology, who taught at the University of Alabama from 1921 to 1954. First at the Tuscaloosa campus, then at the Medical College of Alabama when that was founded, and he was chair of the department when the medical campus moved to Birmingham in 1945.
As reported in the Harvard Crimson of October 16, 1930, Professor McBurney was accepted for a Research Fellowship at the Harvard School of Public Health. He used this opportunity to do a study on typhoid carriers and whether climate could affect their excretion of the bacterium (1) (subscription required).
To me it looks like a weird project, and Google Scholar lists no papers citing it, so let’s give it some attention.
* * *
It seems that state governments, or at least the Massachusetts government, had registers of known typhoid carriers, who were sort of the long-term HIV non-progressors of their day. What is it about these people? What makes them special? Are they a menace? How can we make them less menacing?
McBurney starts the paper with three main justifications.
First, you can alter the intestinal flora of dogs by putting them in a “summer room” (95 degrees, 90% humidity, compared to the “ordinary temperature room” of 68 degrees and 40%). In the sauna they end up experiencing more replication of the Bacillus prodigiosus with which you spiked their hamburger. This was demonstrated by Lloyd Arnold of the University of Illinois (2), who made very good-looking graphs by 1929 standards.
Second, an epidemiological study linking environmental factors to diarrhea in Boston (3) found that “the seasonal occurrence depended more on absolute humidity than anything else”. So this sort of thing is affected by climate in humans as well.
Third, with regard to typhoid, everyone knows about “the periodicity of the carrier state”. Nowadays the word “periodicity” is usually used for a disease that produces a regular cycle of illness, followed by recovery, followed by illness again as the immune system wanes or the pathogen changes slightly, followed by recovery again, etc. The big example is malaria. But even with malaria, people use the word “periodicity” to refer to two things: the multi-week intervals between relapses (great review here (4)), or the two- or three-day intervals between fever flareups during a single period of illness (modeled here (5)). I don’t know if McBurney has a specific pattern in mind, where typhoid carriers regularly alternate between excreting bacteria and not excreting bacteria, or if he’s just saying “Sometimes they’re highly contagious, sometimes they aren’t”.
So he decides to move beyond Arnold’s dog studies and do human experiments. Not by keeping people in “summer rooms” for weeks at a time, or introducing acids and bases into their digestive tract through artificial fistulae. In fact most of the work consisted of recruiting a few registered typhoid carriers, convincing them to send stool samples, and getting them to fill out questionnaires about “conditions such a headache, nausea, vomiting, diarrhea, constipation, taking of laxative, intestinal upset, worry, loss of sleep, fatigue, chilling, condition of living and working quarters as regards comfort, presence of colds, etc.” The study started with 8 patients, and 3 more were added in the fourth month
The stool samples were tested for typhoid positivity and those results were correlated to the questionnaire results to see what sort of health conditions were correlated with the times people were excreting the bacteria. McBurney also compared the health and bacterial info to “weather reports kindly furnished by the U.S. Weather Bureau at Boston through the courtesy of G.A. Loveland, meteorologist”, to see if it correlated with heat or humidity or low-pressure systems or dewpoint or whatever.
This went on for a year, from December through November, though I don’t think he ever says what year it was. Some time between 1930 and 1937.
* * *
One problem with the study is that he didn’t think of measuring how many typhoid colonies were in the stool until it was too late. He just looked for positive vs. negative. Since the vast majority of samples were positive, it was hard to correlate the few negative ones to anything. Meanwhile four patients had nothing but negative samples, so those couldn’t be correlated to anything either.
Unfortunately, the percentage of typhoid colonies on weekly platings was not considered until the 29th week; so no comparison can be drawn between the effect of cold and of warm weather on numbers of organisms.
On the other hand, counting the colonies probably would have been unreliable, changing based on how long the sample was in the mail and other factors. Whereas positive or negative status did not change as a result of mail delays.
It is interesting to note here that during the Christmas holiday several of the specimens were delayed in the mails as much as 6 to 8 days. However, from those showing positive stools, this lapse of time did not cause negative findings with the collecting medium used.
So there were not many correlations. Here’s the three conclusions I see:
1. Female carriers had a higher frequency of positive samples (90% compared to 75% in men). Also, it’s known that women are more likely to be carriers. These two factors go together.
2. Two patients, who had produced mostly typhoid-laden samples, had their gallbladders removed, and afterwards all their samples were negative. This isn’t even mentioned in the text, just in the legend of the table. Isn’t that significant?
3. One family was trying to engage in some skulduggery to get off the government list of unclean citizens.
Stools from carrier 7, an 8-year old boy, from March 16 to August 23, (22 weeks), were negative for a period no longer than 2 weeks in succession, whereas following this period all specimens submitted (10) were successively negative. There is reason to believe that there may have been a substitution of this boy’s stool by some other member of the family. This is based upon past history and the fact that, coincident with these negatives, an attempt to reopen the issue with the Massachusetts State Health Department for his admission to public school, to which he had been excluded, was made.
* * *
Now we can move on to the lab experiments.
In March, Experiment A transpired. Carriers #5 and #10 (males aged 29 and 16) were invited to the Harvard campus, having been chosen based on status as an intermittent carrier (67% of stools positive) and negative carrier (ex-carrier?) (0% of stools positive). They were “placed in an automatically controlled hot room for 6 1/2 hours at a constant temperature and relative humidity of 95 F., and 90% respectively.” After 6 1/2 hours, was there an increase in typhoid bacilli among their excretions?
For carrier #5, there was. Then it went down again 24 hours later. Carrier #10 was still negative.
Again, McBurney decided he should have done something more, which would be to continue the measurements beyond the 24-hour mark. So Experiment B was created. This was the same as Experiment A, except that it also included Carrier #3 (94% of stools positive). Carrier #5 once again had increased typhoid excretion, which continued at the 36-hour and 53-hour timepoints. However, Carrier #3 “gave a decrease”. With the expansion of the sample size from 1 to 2, the effect was no longer evident, and it was concluded that heat and humidity had a negligible impact on the ability of typhoid carriers to infect their neighbors.
Heat and humidity would certainly help the bacteria survive once it leaves the host, though. And what about the data from dogs? Climate must still be important.
What was disappointing here was the failure to find a way to predict what conditions made carriers more of a risk, and the failure to see any pattern in the “periodicity” of typhoid shedding. Typhoid carriers still exist (for details, see the 20oo paper by Asma Ismail of the University of Science, Malaysia (6)), but we still don’t know much about why a carrier is more of a risk at some times than at others. If McBurney’s experiment was repeated with a sample size of 200 rather than 2 it might help.
* * *
1. McBurney R (1937). Typhoid carriers: A study of environmental and other factors bearing upon periodicity. J Inf Dis 61(1):122-128.
2. Arnold L (1929). Alterations in the endogenous enteric bacterial flora and microbic permeability of the intestinal wall in relation to the nutritional and meteorological changes. J Hyg (Lond) 29(1):82-116.
3. Grover JI (1916). A study of diarrheas in Boston for 1915. JAMA LXVII(22):1562-1567.
4. White NJ (2011). Determinants of relapse periodicity in Plasmodium vivax malaria. Malaria J 10:297
5. Su Y, Ruan S, Wei J (2011). Periodicity and synchronization in blood-stage malaria infection. J Math Biol 63(3):557-574.
6. Ismail A (2000). New advances in the diagnosis of typhoid and detection of typhoid carriers. Malays J Med Sci 7(2):3-8.