Pre-Vaccine Declines in Measles Mortality

On Monday, I discussed "pox parties" and "measles teas", social gatherings where parents get their unvaccinated, nonimmune children together with another kid who has a disease, the purpose being to intentionally infect their kids and make them sick. It is a practice that, for at least a hundred years, has been decried by the scientific and medical community as a wretched idea, with one author describing them as "orgies of death". Indeed, in my opinion, these parties are nothing more than child abuse.

While vaccine preventable diseases are not the killers they once were, as I mentioned in passing in that post, they are still quite dangerous; diseases like diphtheria and measles should be avoided and prevented whenever possible. Anti-vaccine activists seem quite enamored with pre-vaccine mortality data. They like to point to the declining death rates from diseases and declare that vaccines not only did not save us from those diseases, but that we didn't need vaccines anyway. There are a couple of things wrong with this way of thinking. First off, it erroneously focuses on disease mortality and pretends that deaths and incidence are somehow the same thing. The implication is that the incidence, that is, the number of cases, was declining before the vaccines. Or they just come right out and say that death rate and incidence are the same:

Measles cases in all developed countries became much milder than in developing countries mainly due to improved diet. Is it logical that deaths associated with measles declined greatly without any corresponding decline in incidence?

That is, quite simply, false, not to mention sloppy thinking. Second, it ignores the non-fatal, but still quite serious, complications of diseases, such as severe dehydration, pneumonia, deafness, blindness, encephalopathy and permanent mental impairment, among others. And, yes, these diseases can still kill, even in developed nations with good healthcare like the United States.

I will readily admit, the measles vaccine did not contribute to the decline in deaths seen before the vaccine was licensed. (Duh!) But while anti-vaccine activists assert that the disease just got less dangerous on its own, they're wrong. Here's why.

As with many of the claims spewed into being by vocal vaccine deniers, I decided to delve into the literature to see what I could find. It's a practice that, if they adopted it themselves, could save them a lot of time and embarrassment. The graph they like to trot out in support of this particular claim shows measles deaths from 1900 through the 1980s:

U.S. measles mortality per 100,000 population, 1900-1984

Looking at this graph, it's pretty clear that death from measles declined considerably before the introduction of the vaccine. Death rate remained fairly steady through the first 18 or so years, then started to drop until about the 1940s, where it leveled off quite a bit. After introduction of the vaccine, it dropped to almost nothing. A similar graph, though looking at a smaller range and using a logarithmic scale, shows a similar, though more even and gradual, drop in deaths:

U.S. measles cases and deaths per 100,000 population, 1912-1974 (logarithmic scale)
Source: Barkin RM (1975). Measles mortality:a retrospective look at the vaccine era. Am J Epidemiol.

The second graph also shows that although deaths from measles declined, the incidence, that is, the number of cases, remained quite steady until about 1963, when the first measles vaccine was approved. One other thing to note in this graph. Sometimes, anti-vaccine folks will simply draw a straight line following the general slope of the data and extrapolate that by 2010, measles would not cause any deaths and would be gone on its own. But that ignores changes in medical care that occurred over that span of time, as well as ignoring changes in the slope itself. We see, for instance, that there's a gradual decline until the early 1920s, then a slightly steeper slope until the late 1940s, with a leveling off in the 1950s. A perfectly straight line just doesn't fit, at least if one is being honest. We see a precipitous drop in the number of cases around 1963. Notice, also that while death rates had been declining, those also dropped about ten-fold in the space of a few years. Compare that to the pre-vaccine era, where it took decades for an equally great drop in death rates (e.g., 1918-1943 and 1943-1967). That impact is something that you don't see in the graph favored by anti-vaccine folks.

For this post, though, I want to focus on that gradual pre-vaccine decline in mortality. What should be clear is that although the incidence in diseases did not change before the introduction of the vaccine, mortality was decreasing. Was the virus changing; did it becom less virulent? Did the virus stay the same while we just got better at keeping patients alive? Ask an anti-vaccine advocate what happened, and you're likely to get an answer that the virus itself was getting less deadly (i.e., less virulent) or that improved diet, sanitation and hygiene are what did it. Or some combination of those two. Period. End of story. The scientific literature paints a slightly different picture (surprise, surprise).

The early part of the 20th century saw a lot of significant changes affecting the way we live. Diets progressively improved, sanitation got better and overcrowding declined. These did have a big impact on overall health of the population. Malnutrition contributes to more severe courses of disease and increased mortality. Poor sanitation can increase the incidence of diseases spread via contaminated food or water, though not respiratory diseases like measles. And overcrowding creates conditions where there are many opportunities for a disease to find new hosts, spreading rapidly. Of course improvements in these areas would also contribute to improved outcomes following infection with a disease. But they only account for a portion of the measles mortality picture.

In the early 1900s and continuing through the middle part of the century, most measles deaths were not from the virus itself. Rather, secondary bacterial infections contributed significantly to the high mortality rate following measles, the most common of which is bacterial pneumonia (Babbott & Gordon, 1954). Common opportunistic bacteria to cause complications were streptococcus, pneumococcus and H. influenzae. In the first decade, there were no antibiotics that could treat these secondary infections, and for a number of years afterward, the antibiotic arsenal was limited. The only control measures available were imperfect attempts at quarantine to limit outbreaks. But given measles' ability to spread before symptoms appear, such attempts were more or less futile (Curtis, 1916). Outbreaks simply had to run their course, and families had to hope that their children would survive. The advent of antibiotics, however, gave a new tool to drastically reduce the impact of secondary infections. Drugs to wipe out these pathogenic bacteria were a boon to doctors and their patients, allowing for a significant cause of death following measles to be virtually eliminated.

Another medical intervention which had a big impact on measles mortality was the use of gamma globulin. There were essentially two ways to use gamma globulin: temporary prevention of disease or modification of disease. In both cases, it was given to those who had been exposed to the virus. For prevention, it had to be given within 6 days of exposure (Janeway, 1945). If a sufficient dose was given soon enough, there was a decent chance of preventing the infection from taking hold. The good side is, the child would not get sick and be at risk for complications, like death. The bad side is that the immunity would only be short-lived; eventually, if re-exposed, the child could be infected again. For that reason, the dominant recommendation for gamma globulin was to use it to modify, rather than prevent, the illness. A smaller dose or one given too late for prevention could make the course of measles milder, while also allowing the patient to develop immunity to the virus. Although this could make the incubation period for the virus longer, the recognizable disease would have a shorter duration and much lower risk of complications. (Janeway, 1945). Other serum products, predating gamma globulin use, also had potential to prevent or modify the disease, though not as safely or effectively as gamma globulin (Babbott & Gordon, 1954). Modified disease, however, was still transmissible. These products, though, were not particularly cheap and did carry a risk of some serious side effects. They had to be used judiciously.

Another aspect that likely contributed to the declining death rates from measles were fertility and family size. From the beginning of the 20th century until the mid-1930s, birth rates steadily declined, leading to fewer children per family (Guyer, Freedman, Strobino & Sondik, 2000). This pushed the median age of infection off until children were slightly older. The shift in age of infection was hypothesized to have played a significant role in the reduction in infectious disease mortality (Reves, 1985), considering how dangerous measles was for those under 3 years of age (Henry, 1921).

A final factor that likely played a large role in the reduction of deaths from measles, particularly in the first few decades of the 1900s, was the establishment of local health departments and improved access to health care (Guyer et al., 2000). Local health departments could help to educate the populace about diseases and how to minimize the risks of exposure and infection. Over this period, as well, improvements in neonatal and postneonatal care reduced mortality and improved the overall health of infants. Increased access to health care for infants and young children could address conditions that might have otherwise contributed to complications of measles, just as measles could negatively affect the course of other diseases, such as tuberculosis (Babbott & Gordon, 1954).

A couple things that did not change were the virulence of the virus (Babbott & Gordon, 1954), nor the risk of encephalitis, at about 1 in 1,000 cases (Babbott & Gordon, 1954; Perry & Halsey, 2004; Griffin, 2010). And while the risk of death has dropped significantly, it is still somewhere around 1 in 1,000 cases to 1 in 3,000 cases here in the U.S.

The history of measles mortality in the pre-vaccine era is rather complex; it's not simply a matter of better food and sanitation. While improvements in overall nutrition, sanitation and housing/overcrowding did contribute to improved health outcomes early in the 20th century, they are not the only factors that played a role in the declining mortality following measles infection. Advances in medicine and healthcare, the use of gamma globulin and other serum products, the advent of antibiotics, improved education and decreases in family size also helped to make measles a less fatal disease. However, it is difficult to quantify the exact degree each of these factors played, since we are dealing with historical data of oftentimes varying quality. I would argue, though, that if I had to choose one intervention that has had the most impact on measles deaths throughout history, I would pick vaccination. After all, if you never get infected with it, you can't die from it.
___________________________
References and Additional Reading:

• Babbott, FL Jr. & Gordon, JE. (1954). Modern measles. Am J Med Sci. 1954 Sep; 228(3):334-61.
• Barkin, RM. (1975). Measles mortality: a retrospective look at the vaccine era. Am J Epidemiol. 1975 Oct; 102(4):341-9.
• Curtis, FG. (1916). Is the control of measles and whooping-cough practicable?. Am J Public Health (N.Y.). 1916 Mar; 6(3):265-68.
• Griffin, DE. (2010). Emergence and re-emergence of viral diseases of the central nervous system. Prog Neurobiol. 2010 Jun; 91(2):95-101.
• Guyer, B, Freedman, MA, Strobino, DM & Sondik, EJ. (2000). Annual summary of vital statistics: trends in the health of Americans during the 20th century. Pediatrics. 2000 Dec; 106(6):1307-17.
• Henry, JE. (1921). A brief statistical study of recent experience with measles and whooping cough in Massachusetts. Am J Public Health (N Y). 1921 Apr; 11(4):302–306.
• History of Vaccines. Timelines: Measles. Retrieved online June 19, 2014 from http://www.historyofvaccines.org/content/timelines/measles
• Janeway, CA. (1945). Use of concentrated human serum γ-globulin in the prevention and attenuation of measles. Bull N Y Acad Med. 1945 Apr; 21(4):202-22.
• Perry, RT & Halsey, NA (2004). The clinical significance of measles: a review. J Infect Dis. 2004 May; 189 Suppl 1:S4-16.
• Reves, R. (1985). Declining fertility in England and Wales as a major cause of the twentieth century decline in mortality. The role of changing family size and age structure in infectious disease mortality in infancy. Am J Epidemiol. 122(1):112-26.