It is interesting to see both this effect and the surprise at it. Wuhan noted this after they knocked down the local infections and opened up a bit. They started getting infections being introduced from other parts of China and reported it as less severe (fewer severe cases, shorter hospitalizations).
I think people need to separate (1) getting infected, (2) an infected individual having a severe/fatal case, and (3) communities having severe/fatal cases.
As far as I can tell, there is no dramatic age difference for testing positive for the infection FOR ADULTS. (See
SCC case demographics and
age demographics) However, that might be affected by different practices of avoiding exposure by different ages, as well as the consequence of mild cases not being detected. At-risk-of-infection is less a function of your health characteristics and more a function of your social characteristics (congregate housing, exposures to large number of people, etc.) For instance, firemen would be at-risk-of-infection.
It has long been known that individual outcomes are worse for classes of individuals, associated with various factors: age, gender, comorbidities, race, blood type, etc. For the disease to be milder, these risk numbers would need to change. At-risk-of-severe-outcome is more a function of your health characteristics and less of your social characteristics. For instance, an 80yo person living at home would be at risk-of-severe-outcome.
Severe/fatal cases for a community SHOULD be dropping everywhere as a result of general prophylactic measures, and for measures applied to each of the at-risk-of-infection and at-risk-of-severe-outcome groups. For instance, at the
SCC dashboard page, you can see that the number of infections at long-term care facilities (LTCF) spiked early but presumably even tighter controls were applied there to keep cases way down after that initial gut punch. Compare the temporal distribution of cases at the LTCF (bottom of page) with the general population (top of page). The effect of the tighter controls at those facilities reduced the infections in a at-risk-of-infection community, and simultaneously reduced the infections in a at-risk-of-severe-outcome community. As a result, I suspect that if you looked at the historical demographics of SCC age distribution of COVID-19 infections, you will see reductions in average age (or, at least, older age), and probably reductions in percent for many of the at-risk comorbidities.
As a result, the "average" current case now should be less likely to have severe outcome and be hospitalized less.
-----
I agree there is
no selection basis for the virus to spread better without a severe outcome for the patient. By the time a severe outcome happens (or doesn't happen), the viral load levels have dropped.
-------
If there truly would be a less severe outcome for an identical twin with the same comorbidities on June 1 than there was for the twin on April 1, maybe it isn't the virus that has changed. I could conceive that as even 0.5% of the population gets the disease, that there are little bits of the viral RNA, perhaps inactivated virus, in the environment where healthy people get exposed to small doses. Maybe our bodies are reacting to it in small ways, perhaps becoming sensitized (to kill it off faster) or desensitized (avoiding the massive immune reaction that does the real damage) or both.
I'm not a physician so I may be way off base. But, perhaps naively,
I'd believe people's immune systems have changed as more likely than the virus has changed.