The Infectious Myth

A Book Project by David Crowe

Measles by the Numbers

Home Updates Email Maps Timelines

David Crowe
March 2019
v2. Section on Canada and increased incidence after 1968
v3. Section on claims that measles vaccine saves more death than the vaccine causes.
v4. Section on claims that the rate of measles deaths is 1 per 1000 infections or more (“Risk of Death”).

Measles has become the favorite illness for inducing infectious disease panic. Which is funny, because it is one of the most benign. But, on the other hand, it is useful, because the symptoms still pop up regularly. It is certainly not AIDS or Mad Cow disease, but then nobody is claiming a vaccine exists for them. SARS was a grand panic for a while, but then completely disappeared (without a vaccine). Nobody wants to claim that Ebola is spreading in the west because that would create an enormous panic. Influenza, well that has a vaccine, but every admits that it is not very good. Measles is a very convenient boogie-man.

The CDC admits that the vast majority of cases of measles have no long-lasting effects (of which death is obviously the most permanent) but it does claim that it has the potential to maim and kill, therefore vaccination is worth the effort, at almost any cost (particularly if you ignore the risk of adverse events from the vaccine).

So, to take the CDC at its word [CDC, 2018], it is useful to see what CDC's own documents say about measles, supplemented by other mainstream sources, where necessary. Their major claims, often repeated are:

Measles Deaths

The biggest worry about measles is obviously the risk of death. The CDC, according to the pro-vaccine blog Vaxopedia, says that the last measles deaths were in 2003, but that there have been death certificates with measles on them since then [Vaxopedia, 2018]. The blog author, Vincent Iannnelli, MD, counts 11 measles deaths since 2000. This is seen as both a huge success for the vaccine, as well as a failure caused by non-vaccinating people still bringing measles back to the United States.

But, let’s go back to the CDC's own page [CDC, 2018]. It notes that there were 6,000 deaths per year in the decade following the start of reporting in 1912, but 400-500 in the decade before the first vaccine in 1963. First of all, this not quite accurate because the CDC's own data shows measles deaths between 1956 and 1962 oscillating between 380 and 552 [CDC, 1966a; CDC, 1967]. This isn't a big fib but it illustrates the simplification of the data that occurs and the lack of attention to detail by vaccine advocates.

It is obvious that the 6,000 deaths in the years at the beginning of the reporting period need to be adjusted for the much smaller population at that time, but also because, back then, death statistics were only kept for a fraction of the total US known as the “registration area”. In 1912, for example, the population of the continental US was 95.3 million, but the population of the death registration area was only 58.2 million (58%).

The big question is how to explain the dramatic drop in deaths between 1912 and the last year without a vaccine, 1962. The death rate in the decade starting in 1912 averages out to 9.02 per 100,000 per year or about 1 death out of 11,000 people. But in 1953-1962, the last decade before a vaccine, it averages to 0.258, or about 1 out of every 350,000 Americans.

The reason cannot be the incidence of measles because, as the following graphs show, measles incidence bears no relationship to mortality. I'm showing three graphs because the first one is from the CDC, and establishes that the second one is legitimate. However, it's such a poor scan that the labels on the axes are not readable. The second graph is from a vaccine skeptic website, and shows similar data to the first, extended out a few more years. The third is from the pro-vaccine website, Vaxopedia, and is noteworthy because it also bolsters the validity of the data, and because of a note in red.

CDC graph [CDC, 1966b]

This graph has a few noteworthy elements. Although the axes are unreadable, due to the poor quality of scanning at the CDC, the title indicates that it has data from 1912 to 1965 (the X axis is therefore marked as one year per tick with a label every 5 years). It shows that the incidence of measles (cases) did not decline over the period of the graph, but that the death rate did. It can be seen that it uses a logarithmic scale, meaning that higher numbers are compressed downwards, but the Y axis labels are completely indecipherable, so it is hard to say exactly what they mean. Luckily the next graph, that shows similar data, is much clearer.

CDC graph of measles incidence and mortality, 1966

Vaccines did not save us graph, [VDNSU, 2010]

The next graph is similar to the CDC graph, but is more readable, and extends out another decade. Also using the same exponential scale, it shows that the measles death rate started out at about 8 per 100,000 and had fallen to about 0.3 in 1962, before the first measles vaccine. This graph extrapolates the death rate out in its decline to estimate an elimination of measles deaths by 2010 (without a vaccine).

That is still close 800 annual deaths, and now zero measles deaths are being reported most years, so we will have to consider the accuracy of this figure later. Measles Mortality in the USA, 1912-1975

Vaxopedia graph [Vaxopedia, 2016]

Dr. Iannelli at Vaxopedia shows a graph that does not have the logarithmic scale, just the normal arithmetic scale. This allows us to put the pre-vaccine decline in context. The doctor makes the argument that the line in 1963, although visually indistinguishable from zero, was still around 400 deaths a year. This is true, but he appears to be pulling a sleight of hand to avoid asking why the death rate would have declined so much without any effective medical interventions. And, the decline before 1963 is much more important (simply because it represents vastly about 97% of the lives saved) than the decline after 1963.

CDC graph of measles incidence and mortality, 1966

Finally, you can build your own graph of US measles incidence and mortality statistics.

Risk of Death

The mortality rate is normally expressed as the number of deaths per 100,000 people, but the risk of death is also an important statistic, the fraction of measles cases that result in death.

Recently, a number of vaccine advocates have made claims that the risk is quite high, that it is not the benign disease that people used to think and some people still claim. Dr. Natasha Crowcroft of Public Health Ontario claims that, "one out of every 3,000 to 5,000 children who get it will die.” [CTV, 2019]. Robert Booy, Chairman of the Immunisation Coalition, which is funded by vaccine manufacturers GlaxoSmithKline, Pfizer, Sanofi Pasteur and Seqirus Australia, ups the ante, “Without a doubt measles can kill. About 1 in a thousand. That’s in Europe as well.” [ABC, 2019]. The CDC goes even further, “1 or 2 out of 1,000 people with measles will die, even with the best care.” [CDC, 2019]

These claims are clearly nonsense for a number of reasons. US statistics show that the overall measles death rate declined dramatically between 1912 and 1962, but the number of cases (incidence) remained relatively static. That means that the risk of death also declined dramatically. [WHO, 2018] states that, “Severe measles is more likely among poorly nourished young children, especially those with insufficient vitamin A, or whose immune systems have been weakened by HIV/AIDS or other diseases.”. In other words, the risk of death is not a single number, it fluctuates with the health status of the person who has measles.

Risks fluctuated dramatically in the US from year to year. Based on published mortality and reported measles cases, the risk of death was 1/44 in 1912 but 1/28 in 1913 and 1/55 in 1920. It was 1/342 in 1939 and 1/552 in 1946 and 1/1472 in 1958, but had risen to 1/1020 in 1963, the first year with a vaccine. But these are comparing reported deaths to reported cases…

Most measles cases were not reported, so ratios of mortality (almost always reported) to cases (rarely reported) are dramatic overestimates. The under 15 population in 1962 for the US was 57.9 million [OECD] out of a total population of 186.54 million [MULTPL]. Assuming that every person got measles before they were 15, there were probably about 3.86 million cases (57.9 million/15). The number of deaths in 1962 was 408, which would be a rate of one death per 9,460 cases, closer to one out of ten-thousand than one out of one-thousand.


The story about measles becomes more complicated if you consider the problem of diagnosis, and of diagnostic bias. If you, as a doctor, believe that there is a current measles epidemic, and you have a patient who you know has not been vaccinated, you may be more likely to diagnose measles. Conversely, you may find some other diagnosis for a patient with spots and a fever, who you know is fully vaccinated.

Our friend Dr. Iannelli at Vaxpedia helps us out here. He investigates a claim by vaccine skeptics that CNN used a photo of a measles-vaccine injured child as an example of a child with measles [Vaxopedia, 2019]. The doctor clarifies that it was actually the CDC that made the mistake, but that it wasn’t a measles vaccine reaction, but a smallpox vaccine reaction. The important point is that the CDC could not tell the difference between measles and an adverse reaction to a vaccine, illustrating the difficulty of diagnosing a condition based on symptoms that are found in many other conditions (including exposures to toxic chemicals, such as pesticides).

The Public Health Image Library confirms that image 13322 is a smallpox vaccination reaction “resembling measles”:
PHIL image 13322, smallpox vaccine reaction resembling measles

My own story with diagnosis is that I, like all my siblings and friends, came down with all the childhood diseases in my first years of life, in the late 1950s and early 1960s. I recently asked my mother, a woman who later trained as a scientist, by the way, about my recollections. She confirmed that we never went to see the doctor because she, like most other mothers, saw measles as a self-limiting, short term disease, that was just an inconvenience. I'm sure she would have taken me or my siblings if things had got out of hand. I asked her how she knew it was measles and not rubella ('german measles') or chicken pox, all characterized by spots. She replied that “Measles was going around then”. I also asked whether she would have diagnosed measles if I had previously had it, and she said what I thought she would say, that once you had measles once, you could not get it again.

I have three pieces of evidence that I was not vaccinated against measles. (1) My mother who, despite being a firm believer in the infectious measles theory, was not an early adopter, and told me recently that she saw no need for vaccination against such a benign illness, and that, in any case, we had already had measles before the vaccine was available. (2) I also have clear memories of receiving the oral polio vaccine on a sugar cube, but no recollection of being vaccinated against measles. (3) I have a vaccination record book from when I moved to Canada in 1966, that has only three lonely entries, for two smallpox vaccinations in 1969 and 1973, and what looks like a combined Tetanus/Polio vaccine in 1978.

My experience with dealing with measles completely at home, like everyone else I knew, raises the mystery of how governments obtained accurate statistics of measles cases (incidence) particularly given the vast number of cases that were being reported. One answer is that they were just estimates based on the belief that all children got measles. Perhaps public health officials filed monthly estimates for each region, and they would reflect the belief that this was a good year or a bad year for measles.

In 1962, for example, the measles case rate was 259 per 100,000. The US population was then 186.54 million, so that would be over 480,000 cases of measles. It's inconceivable that, in the pre-computer era, they could have all been diagnosed with a viral test and filed with the federal government. And without a viral test, misdiagnosis is easy, as the CDC illustrated by picking a photograph of vaccine damage as an example of measles.

Even if cases were diagnosed by a doctor (rather than a mother), they probably would have been diagnosed based on symptoms, not with an actual measles virus test, and the doctors would have used similar assumptions to my mother.

Following the commercialization of a measles vaccine, first in 1963, and then the currently used vaccine in 1968, doctors would have believed that vaccinated children were unlikely to have measles, and that unvaccinated children were. This could create a bias towards diagnosing spots as measles in unvaccinated children, and diagnosing them as something else in vaccinated children. With the majority of children vaccinated, this could have caused a collapse in the rate of diagnosis of measles, not necessarily a real change.

Vaccine Effectiveness

Although the 1968 vaccine is described by [CDC, 2018] as better than the original, 1963 vaccine (“an improved and even weaker measles vaccine;”) it does not provide good immunity. The CDC notes that, “a 1989 measles outbreaks among vaccinated school-aged children prompted the Advisory Committee on Immunization Practices (ACIP), the American Academy of Pediatrics (AAP), and the American Academy of Family Physicians (AAFP) to recommend a second dose of MMR vaccine for all children”.

But even this was not enough. A significant epidemic occurred in 2011 in Quebec, Canada, in a population of school children of which 97% had one dose of a measles vaccine by 28 months and 90% had two doses, and vaccine coverage “is even higher by school entry” [De Serres, 2013].

Right at the very beginning of the use of the 1968 vaccine there was a massive surge in measles cases for 1969/70 and 1970/71, the latter exceeding 1966/1967 [CDC, 1971]. They reported, “In the measles epidemiologic year (EY) 1970-71, 77,682 cases of measles were reported in the United States. This represents a 73.8% increase over the total reported for EY 1969-70 and a 225% increase over that for EY 1967-68”. The associated graph (go to link to view) actually shows that in the peak 4-week period in 1971 about 14,000 cases were reported compared to less than 4,000 during the same period in both 1968 and 1969.

In the Ukraine there appears to have been an accidental experiment showing an inverse correlation between vaccination and protection from measles. In 2016, due to difficulty obtaining vaccines, vaccination rates were 31%, but there was no measles epidemic. However, after a 'catch-up' campaign in 2017, an epidemic of 12,000 or more cases of measles occurred [Gabrielle, 2018]. This article notes that only a minority of diagnoses are based on laboratory tests, most being based on the detection of a rash, fever, and cough, cold or red eyes, possibly leading to over-diagnosis of measles (including mis-diagnosis of vaccine adverse reactions as measles).

3% is Better Than Nothing

One argument in favor of measles vaccination is that even 3% of the 1912 rate of deaths represents almost 700 deaths per year (assuming a rate of 0.2, as in 1962, and a 2018 US population of 329 million). As one anonymous person (@VaccineUK) on Twitter wrote to me, “your lack of empathy for the remaining 3% of deaths is shocking”. However, there are a number of counter-arguments against assuming that 700 or so deaths are prevented every year by widespread measles vaccination:

Environmental Causes

There are several pieces of evidence pointing to environmental factors that at least change the severity of the disease, and increase or reduce the risk of death. This is strongly indicated by the massive decline in deaths between 1912 and 1962 without any vaccine or specific medical treatments. There is a strong association with Vitamin A deficiency, and also an interesting seasonality and regionality.

Vitamin A Deficiency

The World Health organization recognizes a strong association between Vitamin A deficiency and death from measles [WHO, 2018]:
“Severe measles is more likely among poorly nourished young children, especially those with insufficient vitamin A…Measles outbreaks can result in epidemics that cause many deaths, especially among young, malnourished children…All children diagnosed with measles should receive two doses of vitamin A supplements, given 24 hours apart. This treatment restores low vitamin A levels during measles that occur even in well-nourished children and can help prevent eye damage and blindness. Vitamin A supplements have been shown to reduce the number of deaths from measles by 50%.”

Seasonality and Regionality

Measles was much more prevalent early in the year, with a peak before the middle of the year. From the following graph it is not possible to identify the individual month, but it is possible to see that it is in the first half of each year [CDC, 1966a].

Measles Seasonality, 1956-1966

Another graph in the same issue of the CDC publication MMWR shows that the peak from 1954-1963 was in week 20 of the year (mid-May), and a little earlier in 1964-1966 [CDC,1966a].

Measles cases per week, 1954-1966

Measles also showed different patterns regionally, with about triple the death rate in the US South (despite about the same rate of cases), compared to the Northeast and North Central regions in 1965, for example:
Measles cases and mortality in different US regions, 1965 and 1966

Infectious or Environmental?

Putting this together we could see that malnutrition, particularly Vitamin A deficiency, could be at least a contributing factor to measles deaths in the United States. Today we can get fresh fruits and vegetables from supermarkets at any time of the year. But in the 1960s and before, May would have been at the end of the season of limited amounts of fresh fruits and vegetables. Many fruits and vegetables were obtained from canned produce, where nutrients are reduced by high temperature processing. Regionality also supports this hypothesis. The South is the poorest part of the United States, and is where outbreaks of Pellagra were most serious, Pellagra being a Vitamin B deficiency.

The infectious theory would predict that measles would be most prevalent in September, when children return to school and the opportunity for infection is highest. But this is not when measles epidemics occurred. It is hard to think of a reason why an infectious epidemic would be most prevalent in late spring or early summer.

Ironically another strongly seasonal disease was polio that was observed to be strongly correlated with the fruit harvest season early in the twentieth century, which is after the middle of the year (usually July, August, September, October in the Northern Hemisphere). This may have been due to greater exposure to fruit contaminated with pesticides, an opposite problem to a vitamin deficiency.

The infectious theory would also predict that the incidence of the disease is related to the mortality. But while the incidence stayed high from the start of monitoring in the USA, in 1912, to the start of the vaccination period, mortality dropped dramatically.

If we estimate the role of infectivity versus environmental factors in the death rate of measles, we see that at least 97% of the death toll from measles is due to environmental factors, based on the trends of mortality in the US, and at most 3% are due to the infection alone. We cannot eliminate the possibility that measles deaths would have been eliminated without a vaccine, given that the trend in the US was still downwards.

Canadian Statistics

The CDC, in 1976, provided an insight into Canadian measles statistics [CDC, 1976], and even a graph of measles incidence from 1924 to 1975:

Measles incidence in Canada, 1924-1975

They explain the obvious gap by reporting that, “Measles has been officially notifiable in Canada since 1924, with the exception of a 10-year hiatus from 1959 to 1968.”. Why did they stop reporting measles in 1959? Measles incidence had dropped the last four years, but it had dropped even lower in the past, and then risen again. Could it be that, by 1959, it was considered a mild illness that was not of public health significance? Certainly, if it was considered a threat, they would have continued to count cases.

Why did they start reporting measles again in 1968? Could it be that now that they thought they had a good vaccine (it is still in use today), and needed the statistics to be able to generate an epidemic of fear to drive parents to vaccinate their children. I await notice of more innocuous reasons to stop reporting, and then start again.

CDC also reports that, “Since the mid-1960s, when measles vaccine became generally available, measles incidence had been reduced to approximately 20% of the levels of the 1950s”. This is impossible to say, unless there is data that they are not sharing, because measles incidence was not reported until 1968 which clearly is ‘late” 1960s and not ‘mid-’. Furthermore, there was a rise in measles incidence after 1968, before falling again – the application of the 1968 measles vaccine was associated with a rise in cases for at least one year (we don’t know what the incidence for 1967 or 1968 was).

Finally, the CDC reports that, “The number of measles deaths in Canada has declined in the past 25 years as the incidence of measles has been reduced”. Given that they are reporting 1975 statistics, this presumably means that measles deaths have been declining since 1951. If the pattern followed the US, which is likely, they probably had been declining for at least half a century at that point. Additionally, this means that they are admitting that measles deaths declined from about 1951 to 1962, without a vaccine being in place (or 1951-1968 if you agree that the 1963 vaccine was not very effective). And, over the period 1951-1975 (if that is the period they are referring to), there was a rise in cases in 1955, compared to 1954, and then the blank between 1959 and 1968. If they are interpolating a straight line from 1959 to 1968 (which would be stupid) then they would be saying that measles incidence was declining from 1955 to 1963 – without a vaccine.


[ABC, 2019]Drive: Discussion with Richard Booy.
[CDC, 1966a]Measles - 1966. MMWR. 1966 Sep 3; 15(35): 302-303.
[CDC, 1966b]Measles mortality - 1965. MMWR. 1966 Nov 5; 15(44): 384.
[CDC, 1967]Reported incidence of notifiable diseases in the United States, 1966: Measles (Rubeola), 1965-1966. MMWR. 1967 Nov; 15(53).
[CDC, 1971]Surveillance Summary, Measles - United States, 1970-71. MMWR. 1971 Nov 27; 20(47).
[CDC, 1976]Measles - Canada, 1924-1975. MMWR. 1976 Dec 17; 25(49).
[CDC, 2018]Measles history. CDC. 2018 Mar 19.
[CDC, 2019]Measles: It Isn’t Just a Little Rash Infographic. 2019 Mar [accessed].
[De Serres, 2013]De Serres G et al. Largest measles epidemic in North America in a decade-Quebec, Canada, 2011: contribution of susceptibility, serendipity, and superspreading events. J Infect Dis. 2013 Mar 15; 207(6): 990-8.
[Gabrielle, 2018]Gabrielle M et al. Outbreak of over 12,000 cases of measles in Ukraine is caused by recent vaccination campaign?! Stichting Vaccin Vrij. 2018 May 12.
[HSUS]Historical statistics of the United States: Colonial times to 1970. Chapter B: Vital Statistics and Health and Medical Care. US Census. 1975 Sep.
[MULTPL]US Population by Year.
[OECD]Historical population data and projections (1950-2050). OECD.
[PHIL]Public Health Image Library. CDC.
[Vaxopedia, 2016]Iannelli V. Graphs That Show Vaccines Don’t Work. Vaxopedia. 2016 Oct 1.
[Vaxopedia, 2018]Iannelli V. When was the last measles death in the United States. 2018 Apr 15.
[Vaxpoedia, 2019]Iannelli V. Did CNN Apologize for using a Fake Measles Photo? Vaxpoedia. 2019 Feb 1.
[VDNSU, 2010]Vaccines did not save us – 2 centuries of official statistics.
[VSUS]Vital statistics of the United States. CDC.
[HSUS]Historical statistics of the United States: Colonial times to 1970. Chapter B: Vital Statistics and Health and Medical Care. US Census. 1975 Sep.
[WHO, 2018]Measles. WHO. 2018 Nov 29.

Copyright March, 2019 by David Crowe