A Magic Shield That Prolongs Life?

Vaccines Reduce All Causes of Death…

Magic Shield

Yes, we need a magic shield! Coronavirus is evil in its own right: it wants to clog your lungs and kill you. However, it brings many more symptoms with it, such as fever, cough, fatigue, muscle aches, loss of taste and smell, … Moreover, even a mild case of COVID can leave the victim with long-term problems. These include the effects already mentioned, plus “brain fog,” pounding heart and difficulty sleeping.

However, just as COVID appears to be a many-headed monster, COVID vaccines are emerging as a magic shield against many medical conditions, not just coronavirus. The vaccines seem to protect us from non-COVID causes of death, for reasons that are not at all clear. And this is not a new effect: it’s also been seen to almost the same degree from influenza vaccines.

Today’s blog summarizes the surprising and controversial evidence for this many-faceted protection, and offers possible reasons why this can occur. Here are links to the major sections of this blog:
– Unexpected Results from a Study of Vaccine Safety
– Flu Vaccine is Also a Magic Shield
– How Could Vaccination Reduce Overall Mortality?
        – 1. The “Good Behavior” or “Healthy-User” Hypothesis
        – 2. Access to Health Care
        – 3. Risky Behavior
        – 4. Boosting the Immune System
        – 5. Co-Infection
        – 6. Complex Interactions Within the Human Body
– Conclusions

Read on for some happy news. Happy, that is, if you received COVID or any other kind of vaccination.

Unexpected Results from a Study of Vaccine Safety

Researchers at the Centers for Disease Control wanted to see whether the COVID vaccination increased death rate due to other diseases. That is, the vax clearly protects against coronavirus; however, that would not be very helpful if the vax also made the recipient susceptible to another fatal medical problem.

            Surprise: Vaccinated People Had Lower Death Rate, Both From COVID and From All Other Causes

The researchers found an unexpected result: vaxed people had lower death rate from all causes. And not just lower, but three times lower!

The following three graphs show my presentation of data obtained from a CDC article by Xu et al (see Credits at the end of this blog).

Non-COVID mortality by age group

In this graph, I expanded the “total for all ages” bars by 10 times to make them easier to read. Death rates go up with age, naturally, but at every age the vaccinated folks have several times fewer deaths than the unvaccinated ones have.

Non-COVID mortality by gender

It’s generally known that men have higher mortality than women. In addition to that, both have much higher chance of death if they are unvaccinated.

Non-COVID mortality by race / ethnicity

This graph shows that every racial and ethnic group benefits from several times less death rate when they are vaccinated.

In addition, it shows that Hispanics and Blacks seem to derive slightly more benefit from the Pfizer vaccine than from Moderna. That could be an artifact of the data, or a difference that is not statistically significant.

The following notes apply to the above graphs:

            The Analysis is Difficult

In graduate school several of my courses involved statistics. I came away with profound respect for those who specialize in this field. And I have noticed how often professional publications of research results credit a statistician either as co-author or contributor.

The book “How to Lie with Statistics” has a catchy title that captures the truth that statistics can be mis-used. And, they are! During the pandemic, many people claiming to quote “the science” to support a political statement were in fact purposely or unintentionally lying. This mis-use of science happens so often that it unfairly discredits the experts who thoughtfully and carefully use statistics to tease out subtle truths that are often buried in data.

And when we try to interpret demographic data, statistical skill and insights are critical. In the present case, we are discussing data collected after the fact, not as part of a carefully designed and controlled experiment. Therefore, if the data seems to point to some conclusions, our analysis must consider all possible confounding factors and try to eliminate them or correct for them.

When we discuss the mortality data we will point to some of these factors and consider whether they may or may not call the conclusions into question.

            Lower Mortality Means Longer Life

The results above show that mortality, that is, risk of death, is one-third as high after you receive one of these two vaccines. Vaccination seems to be a magic shield against the Grim Reaper.

Here’s what I derived from the CDC data:

Life Expectancy from CDC Data:

AgeUnvaccinatedPfizer VaxedModerna Vaxed
12819191
188191.591.5
45819191
65839292
75859393
85899696

Derived from: CDC, Xu et al (see Credits)

These increases in life expectancy are not negligible. It’s as if a benevolent fairy waved her wand and granted you an additional 7 to 10 years of life.

Flu Vaccine is Also a Magic Shield

And here’s another surprise: This is not the first time that vaccination seemed to confer protection against sickness and death. Several large studies concluded that influenza vaccine is also a “magic shield” of sorts. It reduces the death rate from non-flu causes by one-half. For a typical set of results, see for example the large study conducted at Kaiser Permanente.

It’s true that there are researchers who find no mortality benefit from vaccines. However, there are also critics who disagree with those articles. It’s so complicated to sort through the possible confounding effects that experts may legitimately disagree on the results.

It’s interesting to note that the CDC study by Xu et al found that the J&J vax reduced death rates to one-half rather than the one-third seen in the Pfizer and Moderna vaxes. I did not quote this data in detail above because it uses a different comparison population, and that becomes confusing rather than illuminating.

Why would the J&J vaccine have a different effect on mortality than the other vaxes? One possible reason is that the Pfizer and Moderna vaxes are mRNA vaccines; they operate under different principles than other vaxes. Perhaps that difference would account for the different vaccine recipient mortality reduction (one-third versus one-half).

            Not Everyone Believes the Results

The findings that vaccines reduce all sources of mortality are controversial. Why?

  • Because there’s no general agreement as to a mechanism by which this could occur.
  • There’s no controlled double-blind experiment. The studies are a demographic analysis of correlations on a filtered but basically uncontrolled population.
  • Because the study looks at correlations, the statistical analysis is sufficiently complex that experts disagree on how to interpret the results.

How Could Vaccination Reduce Overall Mortality?

Magic Shield

The observed data certainly suggests that both flu and coronavirus vaccinations are a magic shield. They reduce overall mortality, that is, improve life expectancy. Let’s consider ways in which this interpretation may be wrong, or misleading, or in fact true.

Here are some possibilities:

        1. The “Good Behavior” or “Healthy-User” Hypothesis

Perhaps people who choose to get the COVID vaccine are the sort of people who take better care of their health. But Xu tried to correct for this by only counting people who had taken the flu vaccine. Thus, each of them had already demonstrated some commitment to their own health.

        2. Access to Health Care

The fact that people were able to get the COVID vaccine suggests that they had at access to health care. And this access could have reduced their death rate.

However, that alone does not seem to explain the data. We often hear that non-white groups have less access to health care in the US; but since all ethnic groups show the same benefit in vaxxed mortality, this does not look like a good explainer.

Here’s another example: the Netherlands has universal health care so that everyone has access to treatment; however, in that country data shows that a flu vaccination provides 30% reduction in total mortality risk.

Don’t focus on 30% being different from 50%, because that may be due to differences in the comparison groups in different studies. My point is that even with universal access to health care, vaccination seems to significantly prolong life.

        3. Risky Behavior

Another possibility is that people who choose not to get the COVID vaccine have other characteristics that the analysis could not correct for. Perhaps non-vaxxers engage in activities that tend to shorten their lives. Snowmobiling? Skydiving? Violent arguments? We don’t know, we can only speculate what non-vax people might do that is different and that might increase their chance of death.

        4. Boosting the Immune System

Perhaps receiving the COVID vaccine stimulates the immune system so that it can better fight off unrelated diseases. If this is the case, that might explain the difference in mortality benefit (2 to 1) from flu and J&J COVID vaccines, compared with the 3 to 1 seen in the mRNA vaccines, which operate under a different mechanism.

        5. Co-Infection

Instead of the vaccine protecting, what if this is a case of COVID secretly damaging? That is, the opposite of the situation just described.

Let’s suppose that when flu, or COVID-19, is circulating, people who are fighting a serious medical problem may also be fighting a low level of viral infection. The infection might be so mild that any symptoms it causes are hidden behind the more serious evidence of the primary disease. If this is true, then many deaths that appear to be completely independent of COVID might be made more serious, and more deadly, by the presence of coronavirus nearby.

How could we tell whether this is happening? We could analyze the death rate by primary disease and study the causes separately: specialists (cardiologists, oncologists, hepatologists, …) could consider how their particular disease might be affected by an accompanying weak viral infection. Although every organ of the body seems vulnerable to COVID attack, it’s likely that some types of disease should show a greater or lesser mortality among unvaccinated patients. Differences between causes of death could help explain these effects.

The study just described might lead to revisions in reporting. Thus, physicians might start reporting more deaths as having multiple causes, one of the causes being a virus.

        6. Complex Interactions Within the Human Body

It’s also possible that the way vaccines affect their recipients is much more complex than suggested above.

For example, a review of literature concluded that the flu vaccine reduces cardiovascular mortality (death) by 18%. However, mysteriously, the vaccine did not show a reduction of myocardial infarction (heart attack). To me, this confusing result implies that there are mechanisms at work that are not well understood.

Here’s another example: It’s well known that depression increases the risk of early death from all causes, both in men and in women. And at least one article links depressive symptoms to the immune response from a vaccine (meningococcal conjugate) commonly advised for teenagers. If vaccines interact with depression, they could either help or hurt mortality in a complicated way.

Conclusions

Among the six explanations above, the first two seem less likely for the reasons discussed. However, we don’t have enough information to know the relative importance of the others. And it’s always possible that someone will come up with a more insightful explanation, one that has predictive value in saving lives.

Bottom line, people are very complex, and we are far from understanding all the ingredients that affect our death rate and more importantly, our high-quality “living rate.” Nevertheless, it does seem to be clear that getting vaccinated is more than a magic shield against one virus: it’s also a boost in our total ability to survive.

We all need good news in these times, when there’s so much bad press about coronavirus. The “magic shield” health-giving effects of the COVID vaccines, if they are confirmed by additional studies, are a source for hope. We may emerge from the pandemic not only healthy, but healthier than we were before it infected us. And that’s something to look forward to!

Credits:
– Xu S, Huang R, Sy LS, et al. COVID-19 Vaccination and Non–COVID-19 Mortality Risk — Seven Integrated Health Care Organizations, United States, December 14, 2020–July 31, 2021. MMWR Morb Mortal Wkly Rep 2021;70:1520–1524.
– On openclipart.org: Magic Mirror by GDJ, and Magic Doll by betmariss
– On pixabay.com: Shield Escutcheon by Clker-Free-Vector-Images

Comments

A Magic Shield That Prolongs Life? — 2 Comments

  1. This is really positive news. A question: How can an increase in life expectancy be estimated if the vaccines have only been around for around 10 months?

    • Good question, Roger. CDC looked at the death rate per year for people in all the different categories (age, gender, race). And that part of the data is clear, although it’s still possible to argue about what factors might lead to that data. I made an extrapolation: we know the death rate for people who are 46, 47, 48, 49 years old now. So I’m assuming that a person who’s 45 and has a certain probability of surviving one year, will then experience the death rate of people who are 46 now, and if he lives another year, the death rate of folks 47 now. Thus we forecast the future probability of survival at various ages as being the same as the survival rate of other people who have those ages now. If you believe that survival rates we see now will continue in the future, you can extrapolate life expectancy. (Specifically what I did was to use 1 minus the death rate as probability of surviving a year, I had Excel compute random numbers for 10,000 people each year and saw how many years passed until the number of survivors had dropped by half. Even with 10,000 people there’s noise in the data, but the numbers I showed in the expectancy table are correct to the accuracy shown.) Sorry if this explanation is TMI or unclear! – Art