backtop


Print 33 comment(s) - last by torpor.. on Jul 1 at 1:35 PM

Infections are mildly correlated to population, with little evidence supporting warming hypothesis from non-experts

Naegleria fowleri is one nasty microorganism.  A member of the amoeba family, it resides in inland bodies of water in North America.  It attacks humans -- particularly young men.  An alarmist report, backed by a non-expert, is proclaiming that global warming may drive infections.  Is this the case?  Let's examine the facts.

I. Background

The amoeba enters via the nose and works its way into the brain where it multiplies, slowly killing the victim.  The time between infection and onset of symptoms is nearly a week and the symptoms are rather nondescript at first -- headaches, vomiting, and confusion.

Once infected, the host's brain is literally eaten alive.  The disease is referred to as PAM (primary amebic meningoencephalitis).
PAM infection
[Image Source: Trevor Haggie]

Since 1962 only one patient (out of 128 confirmed infections) has survived and infection by the deadly parasite -- a 99 percent fatality rate.  In most cases, the infection isn't even caught.  Vanderbilt University medical school professor William Schaffner describes diagnosis as more luck than any other factor, remarking, "It’s not something that’s necessarily touched on in medical school.  You have to really probe what patients were doing in the last several days, you have to ask if they were swimming. Honestly, an accurate diagnosis is basically serendipity."

N. Fowleri
N. Fowleri is a deadly amoeba. [Image Source: CDC]

Given the low rate of infections -- and even lower rate of infections that are recognized before death -- treatment is more guesswork than science.  Patients are administered a cocktail of antibacterial and antifungal compounds.  Jonathan Yoder, an expert with the Centers for Disease Control’s division of parasitic diseases, describes in a recent interview with The Verge, "That’s very frustrating for us.  Did the treatment only work that one time? Does it work at all? Would other treatments work better? We just don’t know."

The potentially bad news is that PAM incidents -- at least correctly diagnosed ones -- are on the rise.  Here's a graph with the # of deaths and state of the death by year:

PAM incidents by year
PAM incidents by year [click to enlarge] [Image Source: CDC]

Notice that there were big jumps in confirmed infections in the 1970s and in the 2000s.  So what's to blame?

II. Hypothesis: Warming is To Blame

The Verge has run a rather alarming piece entitled "Brain-eating amoebas thrive in US lakes as global warming heats waterways".  To be fair they have a source for this -- Travis Heggie, a Bowling Green State University (Ohio) Sports and Leisure professor.  Now Professor Heggie has no professional degrees in epidemiology or microbiology, but he tracks N. fowleri infections as a hobby.

Yet, he speaks authoritatively, telling The Verge, "The climate is changing, and let me tell you, so is this.  If warm weather keeps up, I think we’ll see N. fowleri popping up farther and farther north."

Warming
Is warming to blame for the rise in infections? [Image Source: Tehran Times]

The Verge's Katie Drummond writes, "That speculation seems to be reinforced by recent cases of PAM, once a health woe confined to fresh water in southern states like Texas and Arizona. In Minnesota, public health officials were stunned to see two fatalities caused by N. fowleri — both young children — in 2010 and 2012. The cases are the only in state history, and occurred about 550 miles farther north than any previous reported PAM fatality in the US."

The author is correct, two infections have occurred in Minnesota which are the only infections reported in Northern states.  But note that nowhere in the piece do the real experts -- Dr. Schaffner and Professor Yoder -- suggest that the rise in infections is related to warming (a hypothesis fielded by Professor Heggie alone).  Let's dig into whether the rise in infections is truly correlated with warming.

III. The Boy Who Cried Warming (Hypothesis Invalidated)

Let us consider factors that could drive rates of confirmed infections up (along with a suggestion on how it might impact infection rates):
  1. Warming (most N. fowleri resides in water 86 degrees and up)
  2. National population (more people would mean more infections, you'd expect)
  3. Urban population (people use more crowded water resources)
  4. Papers on the disease (more interest means more awareness, more confirmations)
  5. Mutations (hardy strains survive in new habitats)
Let's acknowledge, but toss out #5, as we don't have the data to track that.

To analyze this, let's first compile the rate of infections per decade which peaked from 2002-2011, thus far [Sources: 12].  For #1 let's take the average rise in surface air temperatures over the U.S. on a decade averaged basis [Source].  For #2 we can use widely available census data [Source].  For #3 we can use data from a world clearinghouse [Source].  For #4 we can use Thomson Reuters "Web of Knowledge" Product.

With this data in hand, we simply analyze the correlation of cases per decade versus influence and here is what we find:
PAM Correlation
Actually climate change has the lowest correlation of any examined factor.  The strongest correlations are with the population itself (around 0.86), and more specifically with the urban population (around 0.85).  

I'd be inclined not to discount the relatively high correlation to the amount of peer-reviewed research published per decade, though it's hard to say whether this is a cause or an effect.  Ironically The Verge report hints at this.  Mr. Yoder comments, "I'll swear my life to that — this is being wrongly diagnosed [for bacterial meningitis].  If someone died of meningitis, I’d suggest looking into that case a little more closely ... and looking at where the brain was eaten away at."

A 1970 study shows that for every 16,000 cases of bacterial meningitis, one was actually a misdiagnosed case of PAM.

But even these correlations are highly speculative when one looks at the graph of change in the various variables, scaled to change in PAM infections per decade:

PAM Correlation

Is there an outside possibility that warming is having some affect on PAM?  Perhaps, but it seems unlikely, given that most cases remain in southern states, some of which have actually cooled (like Georgia):
NOAA warming by state
[Image Source: EPA]

Furthermore, while Minnesota is among the states to have warmed more over the past century, Michigan and the Eastern states have warmed at least as much -- yet had no cases.  Michigan and Minnesota have roughly the same number of lakes [Source: 12] over 100 acres -- yet there were no infections in Michigan.

In other words, while N. fowleri is certainly a poorly understood disease and frightening public health threat.  But to blame warming for this disease (or publicize such claims) is highly dubious.  Quite simply put, the correlation is weak, as is the supporting evidence for such a claim.  It appears this is yet another case of crying wolf.

Sources: CDC via Epidemiology and Infection [abstract], The Verge



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

RE: hmmmm
By ShieTar on 6/28/2013 9:19:50 AM , Rating: 2
Furthermore, the whole "article" debunked itself by trying very hard to produce a meaningful trend out of the statistical variation of data. If you group the data a little different, you get to the following:

62-66 5
67-71 13
72-76 5
77-81 20
82-86 9
87-91 16
92-96 10
97-01 14
02-06 12

Apparently there are always more cases at the end of an decade. A strange oscillation has been found. Any attempt to correlate this rather obviously random string of numbers with any kind of more or less linear development like global warming or population increase should be punished by confiscating the high school diploma of the responsible person.


RE: hmmmm
By Solandri on 6/28/2013 4:11:44 PM , Rating: 3
The sun is a variable star with an oscillation period of about 11 years. The data set you present is rather small, but your data matches almost exactly with sunspot counts (higher solar output = more sunspots).

https://spark.ucar.edu/sites/default/files/styles/...

If you look at the sunspot minima - 1964, 1975, 1986, 1996, 2007 - you see it falls right in the middle of the 1962-1966 and 1972-1976 periods. At the edge of the 1982-1986 and 1992-1996 period. And outside the 2002-2006 period. So in your data, the lowest years (5 cases) are when the minimum fell in the middle of the 5 years, moderate years (9 and 10 cases) are when the minimum fell right on the edge. So your data set actually matches almost exactly what you'd expect if you viewed an 11 year cycle in 5 year chunks.

Next up, we'll compare to phase of the moon...


RE: hmmmm
By sorry dog on 6/29/2013 10:54:43 AM , Rating: 2
....speaking of high school.... your high school English teacher would be proud of the last sentence you wrote.


"Mac OS X is like living in a farmhouse in the country with no locks, and Windows is living in a house with bars on the windows in the bad part of town." -- Charlie Miller














botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki