Mystery Rays from Outer Space

Meddling with things mankind is not meant to understand. Also, pictures of my kids

March 17th, 2014

Influenza subtypes in birds

What subtypes of influenza are found in birds?

Influenza is naturally a disease of wild waterfowl; humans, like dogs, chickens, and whales, are occasional victims of mutated viruses from this vast global reservoir of viruses.  In wild birds, flu viruses reassert and recombine wildly, mixing almost all the known subtypes promiscuously.

There are 18 known hemagglutinin subtypes and 11 known neuraminidase subtypes. Two of each are only known from bats, so there are 16 and 9 that could occur together in birds, for a total of 144 possible combinations.  Some subtypes of each are rare, and some HA and NA types don’t play well together. Which combinations have been found, and how common are each?

Mainly because I was playing with Plotly,1 I tried making a bubble chart of HA and NA subtypes found in waterfowl throughout history.  I used sequences in the influenza databases, sorted by HA and NA, and plotted them with HA on the X axis, NA on the Y, and size of each bubble representing how common they are.

Of course, this is not really a representative look.  It’s wildly distorted by surveillance trends; people intensely test birds for H5N1 (because it’s lethal to humans), and H7N7 was surveyed mainly after multiple people were infected in the 2003 outbreak in the Netherlands, so those are probably over-represented compared to their actual prevalence in birds.  Still, it helps give a sense of the complexity of the viruses circulating out there.

Flu in birds

 


  1. My conclusion: Plotly is pretty cool, but doesn’t offer me very much.[]
March 25th, 2013

On the Influenza Epidemic of 1892 in London

Table I
London.  Weekly Deaths from “Influenza” during Four Epidemics – 1847-48, 1890, 1891, 1891-92
Table V
London.  Causes of the Mortality due to the Influenza Epidemic of 1892

The link between influenza, and deaths from apparently-unrelated causes, has apparently been noted for well over 100 years. It’s still somewhat controversial today.

Effect on the Mortality from other Diseases.
One of the chief characteristics of an influenza epidemic is the effect it produces on the mortality from other diseases. The disorders most conspicuously affected are, as is well known, those of the respiratory system; but others are influenced in an almost equally marked degree. The annexed table gives details of the relation between the mortality due to various causes of death and that attributed primarily to influenza during the course of the 1892 epidemic.

The principal facts are contained in Table III. Besides the rise in the mortality from the respiratory diseases, the augmented fatality of phthisis, diseases of the circulatory system, and whooping-cough is noticeable, as is also the increase in the number of deaths attributed simply to “old age.”

Another diagram (No. 1) exhibits the dependence of the mortality ascribed to the two main respiratory diseases, bronchitis and pneumonia, on the presence of influenza for the whole period from October, 1889, to July, 1892. It will be seen that in every instance of the prevalence of the latter, the curves representing the mortality from, the two former rise far above the average; The peculiar significance of these curves lies in the fact that the presence of influenza as the primary cause of death was not recognised in any of the cases which they represent; otherwise they would have appeared in the Registrar-General’s returns under the head of “influenza,” and not under that of “bronchitis ” or “pneumonia ” as the case may be.

An increased mortality from all or most of the causes that have been mentioned characterises all influenza epidemics, and not that of 1892 alone. It is impossible to believe that the invariable coincidence between the rise of these various causes of death and influenza is accidental, and the conclusion seems inevitable that influenza itself is the determining cause, though it does not so appear in the returns furnished to the Registrar-General. This being the case, in estimating the mortality of any particular epidemic it is necessary to allow for the deaths returned under other heads than that of influenza.

–On the Influenza Epidemic of 1892 in London :: BMJ 2:353-356 (1892)

April 29th, 2010

Influenza variations, part II

Mutation NationAbout 15 minutes after I wrote my last article on influenza variation, I was reading the Journal of Virology  and ran across another paper1 on the same thing, that at least partly addresses some of the missing points in the earlier ones.

To brutally truncate my earlier comments: influenza should generate a huge number of mutants as it replicates; but in the few studies that have been done, not all that many variants have actually been detected.

One of the points I raised was that the influenza variation was sampled at the end-point of the infection — after the patient had died, in the paper I talked about the other day.2 Even though the virus had been through the maximum number of replication cycles, it had also experienced the maximal selection pressure, potentially reducing the number of surviving mutants. Is it possible that more variants arose earlier in the infection, but died off before they were detected?

This new paper1  actually looked at exactly that. They used canine influenza as their model, so they could deliberately infect their patients and track through the infections from the beginning through the end. Even though they used a technique that is much less sensitive to mutations (and is probably more error-prone as well) they found tons of variation, and the pattern they found is fascinating:

Mutations arose readily in the infected animals and reached high frequencies in some vaccinated dogs, but they were mostly transient and often were not detected on subsequent days. Hence, CIV populations are highly dynamic and characterized by a rapid turnover of likely deleterious mutations. ((Hoelzer, K., Murcia, P., Baillie, G., Wood, J., Metzger, S., Osterrieder, N., Dubovi, E., Holmes, E., & Parrish, C. (2010). Intrahost Evolutionary Dynamics of Canine Influenza Virus in Naive and Partially Immune Dogs Journal of Virology, 84 (10), 5329-5335 DOI: 10.1128/JVI.02469-09))

(My emphasis) This (assuming it holds true in other studies) beautifully resolves much of the difference between the expected level of variation, and the level that’s observed at any one time point of infection. The explanation is that the variation does indeed appear, but it doesn’t persist.  There is variation is over time as well as at any one time point.

Hoelzer 2010 Fig 1
Figure 1. Variation between challenge influenza virus (yellow) and virus isolated from two naïve dogs 2 to 4 days after infection 1  

There are a lot of very cool things about this study that I’m not going to talk about (differences between vaccinated and unvaccinated animals, evidence for antigenic escape) but there are two things that I thought were particularly exciting.

First is the question of why the mutations seem to be so transient. Part of that could just be chance, part of it is probably selection against deleterious mutants.

But it’s also worth keeping in mind that the viruses are replicating in a dynamic, rapidly-changing environment. The virus enters a host whose immune system is at rest but that immediately recognizes viral infection and ramps up interferons,  then other cytokines, then innate antiviral systems that build up and spill over into an adaptive immune response  …  a whole range of inflammation whose mediators and effectors change from hour to hour. Is this changing environment selecting for mutations that are briefly beneficial, and that then become deleterious as the situation changes a few hours later?

Second – when we think about viruses that are able to jump from one species to another, we think usually of mutants, virus that may be less fit in their “proper” hosts but adequately fit in some other species. (In fact canine influenza itself is a great example of this, a virus that jumped from horses into dogs six or seven years ago.  It is essentially equine influenza, but compared to the equine version it has a half-dozen variants that make it more suitable for replication in dogs.)

If we look at any particular time point we may not find any of these potential emergent mutants. But if we look at all the time points, as in this study, perhaps these potential species-jumping mutants are popping up all the time, but only for a few hours at a time:

This observation suggests that mutations that facilitate adaptation to a new host species might occur transiently in the donor host despite any associated fitness costs and provide a transient reservoir of preadapted mutations1

(My emphasis) There’s also theoretical and experimental work that probably addresses how this sort of pressure could drive population-level robustness.  For example, while heterogeneity is linked to fitness in HIV,3 Claus Wilke says:

Virus strains with a history of repeated genetic bottlenecks frequently show a diminished ability to adapt compared to strains that do not have such a history.4

I don’t know that work as well as I’d like to, but I think it’s probably relevant when considering local and global evolutionary pressures on the virus.


  1. Hoelzer, K., Murcia, P., Baillie, G., Wood, J., Metzger, S., Osterrieder, N., Dubovi, E., Holmes, E., & Parrish, C. (2010). Intrahost Evolutionary Dynamics of Canine Influenza Virus in Naive and Partially Immune Dogs Journal of Virology, 84 (10), 5329-5335 DOI: 10.1128/JVI.02469-09[][][][]
  2. Kuroda, M., Katano, H., Nakajima, N., Tobiume, M., Ainai, A., Sekizuka, T., Hasegawa, H., Tashiro, M., Sasaki, Y., Arakawa, Y., Hata, S., Watanabe, M., & Sata, T. (2010). Characterization of Quasispecies of Pandemic 2009 Influenza A Virus (A/H1N1/2009) by De Novo Sequencing Using a Next-Generation DNA Sequencer PLoS ONE, 5 (4) DOI: 10.1371/journal.pone.0010256[]
  3. Bordería AV, Lorenzo-Redondo R, Pernas M, Casado C, Alvaro T, et al. (2010) Initial Fitness Recovery of HIV-1 Is Associated with Quasispecies Heterogeneity and Can Occur without Modifications in the Consensus Sequence. PLoS ONE 5(4): e10319. doi:10.1371/journal.pone.0010319[]
  4. Novella, I., Presloid, J., Zhou, T., Smith-Tsurkan, S., Ebendick-Corpus, B., Dutta, R., Lust, K., & Wilke, C. (2010). Genomic Evolution of Vesicular Stomatitis Virus Strains with Differences in Adaptability Journal of Virology, 84 (10), 4960-4968 DOI: 10.1128/JVI.00710-09[]
April 27th, 2010

Influenza variations

Mutation comic

Indeed, the amount of HIV diversity within a single infected individual can exceed the variability generated over the course of a global influenza epidemic, the latter of which results in the need for a new vaccine each year. 1

That was said as part of a discussion on HIV vaccines, but let’s think about it from the influenza side.  Why is it true? Why doesn’t influenza have as many variants as HIV?

(Update: Another paper also looks at this question and points to some interesting explanations; I talk about that paper here.)

We know that influenza, like other RNA viruses, is prone to mutation (that is, it has an error-prone polymerase). Depending how you measure it, it’s likely that almost every new influenza genome has at least one mutation in it, meaning that every new infected animal or person should be be generating thousands upon thousands of new influenza variants.

Globally, of course we do see thousands of new flu variants each year.2 But, based on replication fidelity, you’d expect to see a lot more — maybe not quite as many as HIV, but not far from it.

This is also true on a much smaller scale, looking within infected individuals (animals or people).  Even using modern deep-sequencing techniques (like those used in some of the HIV analyses) that should in theory be able to detect large numbers of mutations, there are fewer than you might expect based on the known replication fidelity — far fewer variants than in HIV:

Inasmuch as the mutation rate for type A influenza viruses is estimated at one nucleotide change per 10,000 nucleotide during replication and most infections are caused by as many as 10 to 1000 virions which likely possess varying numbers of nucleotide differences in their genomes, one can expect that each influenza A virion is possibly a quasispecies. However, we identified relatively few quasispecies – probably because the currently available sequence analysis software do not allow robust quasispecies analysis and extensive manual curation is necessary. We believe that with the help of improved bioinformatic tools we would detect more quasispecies populations in our sample sets.  3

H1N1 (swine-origin influenza virus)
H1N1 (swine-origin influenza virus)

I don’t know enough about the computational side to comment on their bioinformatics point.  Another recent paper4 uses a similar approach and (at least at first) seems to reach an more conservative conclusion.  They talk about “quasispecies”, but they seem to be using the term rather loosely, to describe just a handful of distinct genomic sequences. These sequences differ by, for example, a single base (and a single amino acid) in the HA, where one of the sequences was present at about 75% of the sequences, and the other at about 25%. To me that’s not really a “quasispecies” — a quasispecies is something that needs to be defined by an average sequence even though the vast majority of the genomes are different from that average. (Here and here are Vincent Racaniello’s explanations at The Virology Blog.)  Two sequences is just two sequences.

However! The authors do make their data available. I don’t have time to do a detailed look, but from what I think is a very conservative analysis, in one stretch of just 25-40 bases some 5-10% of the genomes have at least one mutation.5 If that’s roughly true across the whole genome, then each genome would have, what, maybe a half-dozen mutations on average. That, to me, really is a quasispecies.

(Do note that this is not the mutation frequency for any individual residue. No single point [with the two or three exceptions that the authors focused on] is mutated at much more than one in a thousand, and most probably more like one in many thousand, which is about what you’d expect. )

There are a myriad of complicating factors separating the error frequency in these genomes from the raw error rate of the viral polymerase. A couple of huge ones: These viruses had undergone a bunch of replications in the host – this isn’t the error rate per replication cycle, it’s the cumulative error rate after many cycles. The virus was from a patient who had died with (and probably of) the virus, and though we don’t know how many time the original infecting virus had replicated it was at least a half dozen cycles, perhaps two or three times that.

Influenza virion
Influenza virion

On the other hand, during those replication cycles, many mutations (quite likely even the great majority of them) would have been deleterious or outright defective, so most of the mutations would have never propagated but would have just silently disappeared and not been counted at the end.

The most interesting point is that these mutations aren’t arising in a vacuum. Thinking now about which mutations survive and get detected, not the baseline rate of mutation formation: The variants are forming are in an environment that’s designed to be very hostile to viruses. Mutations are going to undergo selection by the immune system.

This is one place where influenza is going to experience a very a different set of pressures than HIV. HIV persists in the presence of the adaptive (T cell and antibody-based) immune response, whereas as the adaptive response kicks in for flu the virus gets evicted. HIV therefore not only have a much longer period (years instead of days) to throw out mutations, it also is shaped by the immune response. By comparison flu would probably only have a couple of replication cycles in the presence of an adaptive response.

Changes in the virus that accumulate over the handful of replication cycles would reflect a strong selection pressure. The vast majority of mutations, even those that aren’t completely defective, are going to be less fit than the original virus and won’t accumulate. Knowing which mutations do accumulate should be very interesting because it may tell us what the virus is going through in the host.  That’s what the authors of this paper focused on — the one particular site that had a much, much higher variant  frequency, more like 25% of the genomes.  The assumption is that this arose during the infection and was positively selected for. 6

The variants that replicate best in a host may be quite different from those that are effectively transmitted. That is, there may be multiple sources of selective pressure, of which we have previously mainly only seen transmission pressure (because that’s the main one that will accumulate in a population, because transmission represents a bottleneck in the virus’s evolution [link to The Virology Blog]).  The particular HA variant that was picked up here (that apparently accumulated during the infection) is rare globally. Is that a version of the HA that’s more efficient within a host, but that doesn’t transmit as well?

I think a major reason for the difference between HIV and influenza variant accumulation is the difference between within-host and between-host (transmission) selection.  HIV spends long, long periods within a single host, thousands of replication cycles, accumulating mutations.  The transmission bottlenecks come at much longer intervals and have a much larger accumulated population to work with.  Influenza has a comparatively brief period within the host, only a handful of replications before a new transmission bottleneck hits. 7

This sort of deep sequencing experiment on influenza will probably be improved over the next few years, and I’ll be very interested to see just how much variation there really is within on flu-infected host.


  1. Walker, B., & Burton, D. (2008). Toward an AIDS Vaccine Science, 320 (5877), 760-764 DOI: 10.1126/science.1152622[]
  2. More correctly, I suppose, we infer the presence of thousands of new variants based on the hundreds of them that we see, and knowing that we are only examining a tiny fraction of all the flu cases that are out there.[]
  3. Ramakrishnan, M., Tu, Z., Singh, S., Chockalingam, A., Gramer, M., Wang, P., Goyal, S., Yang, M., Halvorson, D., & Sreevatsan, S. (2009). The Feasibility of Using High Resolution Genome Sequencing of Influenza A Viruses to Detect Mixed Infections and Quasispecies PLoS ONE, 4 (9) DOI: 10.1371/journal.pone.0007105[]
  4. Kuroda, M., Katano, H., Nakajima, N., Tobiume, M., Ainai, A., Sekizuka, T., Hasegawa, H., Tashiro, M., Sasaki, Y., Arakawa, Y., Hata, S., Watanabe, M., & Sata, T. (2010). Characterization of Quasispecies of Pandemic 2009 Influenza A Virus (A/H1N1/2009) by De Novo Sequencing Using a Next-Generation DNA Sequencer PLoS ONE, 5 (4) DOI: 10.1371/journal.pone.0010256[]
  5. I extracted the FASTQ data containing the short sequence reads matching influenza sequences from the supplemental PDF, converted it to FASTA, and used xdformat to move it into a BLAST database. Then I grabbed 40 bases from the genbank sequence CY045951.1, the PB2 segment of the closest-match influenza strain, choosing a region (positions 2151-2190) with very high coverage, and BLASTed this sequence against the short sequence data, using parameters such that I retrieved sequences that match at least 25 of 40 positions.  Of the  ~2050 hits I retrieved, about 120 had at least one internal mismatch. I can’t distinguish these from sequencing errors, but I think it’s much higher than you’d expect from sequencing error.  And I hope that my conservative approach (for example, I would have discarded mismatches at the ends of the hits) would balance out that source of confusion. []
  6. One point, by the way, that the authors didn’t cover was the possibility that this patient had actually been initially infected with more than one viral sequence.  We do know that a significant number of flu cases are doubly infected. The fact that the minor variant is a very unusual strain makes this less likely, but not impossible.[]
  7. And I think it’s fair to say that the global population-based HIV variation — the transmission-selected amount of variation, as opposed to the vast within-individual variation — is rather more comparable to that of influenza.[]
February 16th, 2010

On predictability of influenza pandemics

Despite the fact that the recent pandemic was the best studied and recorded to date, the knowledge gained will probably have little predictive value for the next pandemic, either in qualitative or quantitative terms.

Communicable Diseases and Epidemics
Martin M. Kaplan
Bulletin of the Atomic Scientists, June 1960

Kaplan  was referring to the 1957-58 influenza pandemic, but the sentence could have been taken from many places in 2009. In any case, Kaplan was pretty much correct; the next three pandemics after that were quite different from the 1957-58 pattern, and quite different from each other.

January 13th, 2010

How is avian influenza evolving?

Carrel et al 2010 PLoS ONE Fig5
Geographic distribution of H5N1 highly pathogenic avian
influenza viruses (HPAIVs) used in this study. Darkened
provinces indicate locations of virus isolation.

“We found several patterns that suggest one general model of evolution in this viral system: 1) within regions, viral mixing in poultry moves toward heterogeneity and the emergence of local types; 2) differentiation was centered around regional viral hubs located at centers of human and bird population density; and 3) evolution occurs because of relative isolation of the hubs, most likely fed by the abundant supply of domesticated poultry (and people) at the hubs. The analysis thus suggests that at the scale of neighboring city hubs and the intervening hinterland, evolution of H5N1 follows the pattern described by classical theory of genetic differentiation due to isolation by distance.” 1

Carrel et al 2010 PLoS ONE Fig1
Genetic versus geographic distance of HK821-like HPAIVs in Vietnam.

This is in Vietnam, and the basic finding was that H5N1 viruses isolated in Vietnam show signs of local evolution, in that the viruses cluster into local sub-strains in different areas of the country.  I’m not all that knowledgeable about H5N1 spread, but I had thought that infection of wild, especially migratory, birds would be an important factor in spreading H5N1 between chickens.  If I’m interpreting this paper right, it looks as if H5N1 is mostly circulating within local regions, within the chicken population, and distant spread isn’t a major factor. That has obvious implications for control of the virus.


  1. Carrel, M., Emch, M., Jobe, R., Moody, A., & Wan, X. (2010). Spatiotemporal Structure of Molecular Evolution of H5N1 Highly Pathogenic Avian Influenza Viruses in Vietnam PLoS ONE, 5 (1) DOI: 10.1371/journal.pone.0008631[]
December 14th, 2009

Influenza before 1918, part II: 1872

In 1872, a pandemic influenza outbreak brought the US to its knees:

“The streets are almost deserted.” –Washington, D.C.

“A Sunday quiet prevails upon the streets.” –Springfield, OR

“The streets yesterday looked deserted.” –San Francisco, CA

“The street cars have stopped.” – Erie, PA 1

And yet, if you look at the mortality rates for influenza in 1872, it’s not a particularly impressive year — if anything, the influenza death rates were exceptionally low that year.  At least, they were low in humans.  1872 brought a pandemic equine influenza, laying low almost every horse in North America.

On the evening of October 21st only a few animals were affected, but on the morning of the 22d there was scarcely an animal of the equine species that was not affected.  Horses, mules, and even a zebra.  More than twenty thousand were suffering in different degrees. 2

An estimated 3-4% of the tens of thousands of horses in New York died. 2 But the deaths weren’t the biggest problem:

The actual money losses, in an epizootic of influenza, are more in the way of the loss of work and the complete stagnation of trade in all departments, than in the number of deaths.  Yet even in this sense it may prove more ruinous than would a disease having a less universal away though far more fatal to the animals attacked. 3

Without horses, business slammed to a halt; the mail didn’t run, groceries didn’t reach the cities, crops weren’t harvested or transported.  After a few weeks, most of the horses recovered and business followed, but the epizootic swept across the country1  (intensely tracked by the newspapers of the day, warning each city in turn that it was going to be attacked), finally fizzling out the following summer in British Columbia.

Equine influenza map, 1872


  1. Adoniram B. Judson, MD (1873). History and Course of the Epizootic Among Horses Upon the North American Continent in 1872-1873. Public Health Papers and Reports. American Public Health Association. Hurd and Houghton, New York, 88-109[][]
  2. Annual report of the Department of Health of the State of New Jersey. By The New Jersey State Dept. of Health, 1877 (“Epizootic influenza”, p. 160)  [][]
  3. Text book of veterinary medicine, Volume IV.  By James Law, F.R.C.V.S.  1906 []
December 7th, 2009

Influenza before 1918

The huge 1918 influenza pandemic, caused by the great-grandfather of today’s swine-origin pandemic H1N1, wasn’t the first time influenza was seen in people — not by a couple of thousand years. 1 Seasonal flu was around before it, just as it has been since; and epidemics and pandemics regularly swept through the world before 1918.

The charts below, published in 1921,2 show  completely modern-looking patterns of influenza.

Note the 1890/91 pandemic of “Russian flu”: 3

Within the space of a few weeks in 1890 this disease prostrated hundreds of thousands in Europe and America, enormously increasing the death rate, and leaving many of its surviving victims in a condition of pronounced debility for many months.  For a time it closed factories and workshops, it checked business, and obstructed the prosecution of many enterprises. 4

Influenza - England and Wales, 1845 (Vaughn)

But also note the classic seasonal peaks before and after 1890:

Influenza - Massachusetts, 1887 (Vaughn)

In spite of all we’ve learned about influenza since 1845, we haven’t been able to do much to change its patterns.


  1. Stephen Dando-Collins, in “Caesar’s Legion: The Epic Saga of Julius Caesar’s Elite Tenth Legion and the Armies of Rome“, cites claims that influenza kept many of Julius Caesar’s legions from sailing to his assistance during his civil war with Pompey the Great, in 49 BC[]
  2. Warren T. Vaughn (1921). Influenza: An Epidemiologic Study The American Journal of Hygiene Monographic Series[]
  3. So-called in Western Europe; but the Russians called it “Siberian Fever”, and Siberians called it “Chinese Distemper”.  It supposedly originated in Bokhara, in present-day Uzbekistan, according to Encyclopaedia Medica, Volume V. By Chalmers Watson. New York, Longmans, Green, & co. 1900 []
  4. The Cottage Physician. For Individual and Family Use.  King-Richardson Publishing Co., Springfield, MA 1897[]
November 17th, 2009

Swine-origin influenza virus risk factors

My friend Lauredhel, at the Hoyden About Town blog, made an interesting point about risk factors for swine-origin influenza virus (SOIV), and the perception of those risk factors in the press.   The press has made a big deal of the putative link between obesity and risk of severe SOIV.  But, as she pointed out back in June, the data1 at that point showed no such link — in fact the percent of obese people with severe SOIV was if anything lower than the frequency of obesity in the general population.  However, the press picked up on this because of a reply to a question at the May CDC press briefing, and the headlines were all over this (actually non-existent) link.

Lauredhel made the same point when a recent paper in the Medical Journal of Australia reviewed cases of SOIV. 2  Once again, Lauredhel points out, in this series of patients, obesity was actually lower in severely affected people than in the general population:

Around 18% of the Australian population is obese. Around 7% of people severely ill with H1N1 flu are obese. 3

However, I don’t think the story is that simple.  I made a few comments there, which I’ll review here.  The summary of the exchange, I think (Lauredhel might disagree) is that some, but not all, studies have found a connection between obesity and risk of severe SOIV; the largest studies do show a connection, but when looking at the overall picture, it’s not a strong connection.  The press has, however, vastly overstated this link, focusing on it apparently because of one comment in a press briefing, but ignoring several attempts to clarify and downplay the observation.

My comments:

The largest series I’m aware of is the recent JAMA study. 4 (The eMJA paper came about about simultaneously, so they can be forgiven for claiming theirs is the largest study.) Here 48% (of hospitalized or fatal cases, where data were recorded) were obese. The rate was slightly higher in fatal cases (66%, of 110 cases total) than non-fatal (52%, of 212 cases; these are cases over 18 years, and the non-fatal rate would be lower if we included younger people). The fatal cases, especially, were disproportionately in the highest BMI cases — 50% of fatal cases had a BMI over 40, if I’m reading their Table 2 right.

I don’t know what the relevant population rates of obesity are, so we don’t know relative risk. In the US generally, I believe obesity is in the 30% range. The authors say “Of adults with BMI data available, more than half were obese and one-quarter were morbidly obese. As a point of reference, the percentage of adults who are morbidly obese in the United States is 4.8%”.5

(An important concern is that this may be distorted. It looks as if data weren’t recorded for obesity on the majority of patients. I would worry about a recording bias, with information on obese patients being recorded more readily than for non-obese. Still, even if not one of the non-recorded patients had BMI over 40, the case rate is higher than the 4.8% background.)

In smaller studies, there seems to be a similar picture. In a Michigan survey (June ‘09) 9 of the 10 patients with swine-origin H1N1 hospitalized with ARDS were obese; in a European survey, 8 of 13 fatal cases were obese.

The numbers are still quite small, and they’re not all consistent, but from what I see here, I wouldn’t dismiss obesity as a risk factor.

Also, I see a survey in Australia6 where the relative risk of obesity and “morbid obesity” (BMI > 40) is worked out. Just as you note, the relative risk of “obesity” for death is less than 1 (0.6), but probably 1 is easily in the 95% confidence interval. But the RR for obesity of ICU admission is up (1.7) and for morbid obesity is 4.4; death RR for the latter is 2.4.

These data aren’t entirely consistent with the eMJA data, but I don’t have time to try to resolve the differences. One point that’s raised by the JAMA study I quoted above is whether BMI is specifically recorded in these cases. The eMJA study only notes 8 patients that were obese, but unlike the JAMA study they don’t explicitly give a denominator — that is, they don’t specifically say that the remaining 104 patients had a BMI recorded at all. Could the BMI simply not be available for some of these remaining patients? I don’t know, I’m asking.

Lauredhel said in the comments:

I saw the CDC and other ‘experts’ at the beginning deciding that obesity was obviously a major risk factor based on early data that appeared to show the complete opposite

I replied:

For what it’s worth, neither the CDC MMWR article from May, nor the eMJA paper that just came out,7 say that obesity is a factor. The CDC report includes obesity last in a list of underlying medical conditions, and never says that it’s a risk factor per se. The eMJA paper only mentions obesity when they define it, show the rate in a table, and don’t mention it in their discussion at all.

So I’m not sure you can blame the scientists here. The overwrought press coverage, as far as I can see, entirely arose out of a comment by Anne Schuchat in a press briefing. (Schuchat would not be one of the authors of the article.). Her comment certainly was misleading, but it’s not quoting the scientists who did the work; it seems to come out of nowhere. Not excusing her here, but I would bet her comment was in response to a specific question from the press, not something she raised herself, and she seems to have only been referring to “severe cases” (not all the cases in that MMWR report), which at that time would have been a tiny subset of a tiny subset of cases.

And she corrected herself, at least partially, later on. If you look at a subsequent press briefing (in July) she specifically says that the difference is not there, especially accounting for other underlying conditions, and “They [obese people] would not be a targeted group.” That didn’t get any press, as far as I can find.

So, not surprisingly, the press has done a poor job of covering this, jumping on the comment from Schuchat without checking the figures. The experts, at least those who are actually doing the work, aren’t making the connection except in those studies that actually do see a disproportionate number (the JAMA study and others).

Lauredhel said:

Schuchat’s original statements weren’t a single throwaway remark

I replied:

That’s true, but in fact the context of her explanation was exactly the point you are making — that the frequency of obesity in SOIV patients wasn’t necessarily higher than that in the population (”So it’s hard for us to say at this point to say whether the number of patients with reported obesity is significantly higher than we would expect”). In other words, I’d say that “her perceptions of obesity and risk” were pretty much what you’re saying.

That was the May press briefing, the one that led to the press rampage. Now, reading the press conference transcript, the point she tried to make didn’t come across very well, because she started off sounding as if she agreed with the obesity issue and didn’t make the qualifications until several questions later. She clearly recognized that she hadn’t been clear, because she tried to clarify the point in each of the subsequent briefings (in June and in July). But I’m not seeing this as the experts making assumptions — quite the opposite, in fact. The expert was carefully not making the assumption, but the press didn’t pick up on the qualifiers that she explicitly presented. She could have presented this better, but I’m inclined to put it down to imperfect communication, not jumping to conclusions.


  1. Centers for Disease Control and Prevention (CDC) (2009). Hospitalized patients with novel influenza A (H1N1) virus infection – California, April-May, 2009. MMWR. Morbidity and mortality weekly report, 58 (19), 536-41 PMID: 19478723[]
  2. Justin T Denholm, Claire L Gordon, Paul D Johnson, Saliya S Hewagama, Rhonda L Stuart, Craig Aboltins, Cameron Jeremiah, James Knox, Garry P Lane, Adrian R Tramontana, Monica A Slavin, Thomas R Schulz, Michael Richards, Chris J Birch, & Allen C Cheng (2010). Hospitalised adult patients with pandemic (H1N1) 2009 influenza in Melbourne, Australia The Medical Journal of Australia, 192, 1-3[]
  3. Obesity Still Dramatically Decreases Risk of Severe H1N1 Flu?[]
  4. Louie JK, Acosta M, Winter K, Jean C, Gavali S, Schechter R, Vugia D, Harriman K, Matyas B, Glaser CA, Samuel MC, Rosenberg J, Talarico J, Hatch D, & California Pandemic (H1N1) Working Group (2009). Factors associated with death or hospitalization due to pandemic 2009 influenza A(H1N1) infection in California. JAMA : the journal of the American Medical Association, 302 (17), 1896-902 PMID: 19887665[]
  5. Louie JK, Acosta M, Winter K, Jean C, Gavali S, Schechter R, Vugia D, Harriman K, Matyas B, Glaser CA, Samuel MC, Rosenberg J, Talarico J, Hatch D, & California Pandemic (H1N1) Working Group (2009). Factors associated with death or hospitalization due to pandemic 2009 influenza A(H1N1) infection in California. JAMA : the journal of the American Medical Association, 302 (17), 1896-902 PMID: 19887665[]
  6. New South Wales public health network (2009). Progression and impact of the first winter wave of the 2009 pandemic H1N1 influenza in New South Wales, Australia. Euro surveillance : bulletin europeen sur les maladies transmissibles = European communicable disease bulletin, 14 (42) PMID: 19883546[]
  7. Justin T Denholm, Claire L Gordon, Paul D Johnson, Saliya S Hewagama, Rhonda L Stuart, Craig Aboltins, Cameron Jeremiah, James Knox, Garry P Lane, Adrian R Tramontana, Monica A Slavin, Thomas R Schulz, Michael Richards, Chris J Birch, & Allen C Cheng (2010). Hospitalised adult patients with pandemic (H1N1) 2009 influenza in Melbourne, Australia The Medical Journal of Australia, 192, 1-3[]
November 16th, 2009

Pandemic patterns: Is the influenza pandemic going away?

The number of influenza cases this year seems to have peaked and started to drop in the last few weeks, according to both the CDC surveillance data and Google Flu Trends (which updates more in real-time).  Does that mean swine-origin influenza virus is gone for good? We don’t know, of course, but I was struck by the resemblance of this year’s caseload to charts I’ve seen of the 1918 influenza mortality rates:

Pandemic influenza by week, 2009 vs 1918

Weekly influenza cases for 2009 (red) vs. weekly influenza mortality from 1918 (black)

I don’t have the original data that were used to make the 1918 data1 (black traces; these are mortality data, rather than cases; some sources suggest that this would underestimate the case number for the spring/summer peak of 1918, if the virus increased its virulence before the fall outbreak) , but I think I have the chart aligned with the CDC’s weeklies (red) for this year’s pandemic. 2 They’re not identical, but they’re similar so far.  In particular, there was a surge in the summer, a big peak early in the fall — much earlier than standard seasonal influenza, which doesn’t usually get going until January or February — and then the fall cases dropped dramatically.  The 1918 influenza was followed by a third, smaller, peak, in winter, around the usual period for influenza.

In 1957, when a new pandemic influenza struck the US, the pattern of three waves — smallish in summer, very large in fall, followed by a slightly smaller but still major wave in early winter — was also broadly similar.3 Again, there was a spring/summer wave throughout the US, starting in June and peaking, maybe, in August.4  The fall outbreak started in September and peaked in late October, there was a lull, and then there was a new surge early in 1958, this time perhaps peaking a little later in the winter than the 1918 version.

1957 influenza pandemic, by week
1957/1958 influenza cases, by week5

The next pandemic was 1968/1969, when H3N2 moved in and supplanted the H2N2.  The pattern here seems different: There was little or no spring/summer wave,6 and while the outbreak did start a little earlier than usual, it was only by a few weeks:

The first outbreaks in the civilian population developed in Puerto Rico and Alaska in late September and early October. The first outbreak in  civilian population in the continental USA did not develop until the third week of October, when the small desert city of Needles, Calif., reported an influenza-like illness involving 35%-40% of the of the population. 7

(By comparison, by the third week in October this year, the pandemic seems to have just about peaked.)  Along with the late start in 1968, there was no lull; the influenza kept building to its peak, in December, and then quickly dropped down and stayed down. It was really more like a very large seasonal epidemic, very different from the patterns of 1918, 1957, or 2009.  (There’s actually a more impressive bump in the following summer, 1969, so this might be simply because the virus didn’t reach the US in time, and carried over into the summer rather than presaging the pandemic.)

1968/69 pandemic cases, by week
1968/1969 influenza cases, by week, compared to previous years7

The next pandemic was 1977-1978, when H1N1 returned.  I don’t think we have good data for that, because the main measure was death, and that pandemic strain didn’t impact mortality significantly. 8 ,9  However, as far as I can tell, that was mainly a winter-only wave. So the common claim that pandemics come in several waves is not universally true — just two of the four 20th century pandemics acted that way — but it does seem to be true for the 2009 pandemic.

In any case, based on previous pandemic patterns where the disease started this early, I’m guessing that we’re already past the very worst of the 2009 SOIV outbreak, and we’re going to enter a bit of a pause; but it’s going to come back early in 2010 — perhaps not to the same levels as we’ve been seeing in late October/early November, but not too far off.


  1. Taubenberger JK, & Morens DM (2006). 1918 Influenza: the mother of all pandemics. Emerging infectious diseases, 12 (1), 15-22 PMID: 16494711[]
  2. http://www.cdc.gov/flu/weekly/index.htm[]
  3. This was the introduction of an H2N2 strain that replaced the previously-circulating H1N1 virus[]
  4. I haven’t seen the data for the summer of 1957, so this is based on comments in the descriptive papers, and I don’t know how big the summer wave was[]
  5. D. A. Henderson, Brooke Courtney, Thomas V. Inglesby, Eric Toner, & Jennifer B. Nuzzo (2009). Public Health and Medical Responses to the 1957-58 Influenza Pandemic Biosecurity and Bioterrorism, 7 (3), 1-9[]
  6. You can see a little bump in the summer of 1968, which could conceivably have been the summer wave, but it’s definitely much less dramatic than the 1957 or the 2009 summer.[]
  7. Robert G. Sharrar (1969). National Influenza Experience in the USA, 1968-69 Bull Wld Hlth Org, 41, 361-366[][]
  8. Lui, K., & Kendal, A. (1987). Impact of influenza epidemics on mortality in the United States from October 1972 to May 1985. American Journal of Public Health, 77 (6), 712-716 DOI: 10.2105/AJPH.77.6.712[]
  9. The 1977-78 H1N1 was almost certainly a laboratory accident, a release of an earlier strain from pre-1957.  Accordingly, most people over their mid-20s were already immune to the pandemic virus.  Since older people are usually the ones with the highest mortality rates, the 1977-78 pandemic didn’t trtanslate into increased mortality, and doesn’t show up in mortality rates.[]
:)