Prof. Michael Fuhrer

Evolutionary adaptation to global change reduces sustainable fisheries yields Global warming is altering the fisheries that underpin food security, but projections of these impacts generally exclude evolutionary processes. We describe a model that forecasts how fish will adapt ...

Out today in @science.org our paper on how evolutionary adaptation to warming will reduce fisheries yields. Please let me know if you need me to send you a copy. Summary of the paper coming soon. www.science.org/doi/10.1126/...

Are you using age-standardised data? Without age standardisation, I don't think you could possibly extrapolate accurately from pre-pandemic to today.

Looking at % excess, the excess was roughly 10% in 2022, but in 2024 was 0.9% (-17%, 3.8%). This seems *very* far off from your estimates. I think an excess of 10% in 2023 and 2024 can be completely ruled out statistically.

A simple linear extrapolation of ASMR from StatCan gives an excess which is fairly consistent with the covid-19 age-standardised mortality rate (red).

I am confused by this chart. StatCan has age-standardised mortality data available through 2024, and 2024 ASMR is lower than 2019.

www150.statcan.gc.ca/t1/tbl1/en/t...

Yes, that's very nice! I haven't seen many countries that do hospital surveillance of pathogens beyond covid/flu/RSV. I don't think the US tracks hospital admissions beyond those 3.

Australia has a sentinel hospital network that monitors several pathogens:

(report www.cdc.gov.au/resources/co...)

Also covid hospitalizations in Germany are the lowest they've been since 7 March 2020. So there's some reason to believe it's not an aberration!

infektionsradar.gesund.bund.de/en/covid/hos...

Germany posted its lowest wastewater level last week (and this week's is second lowest).

bsky.app/profile/mich...

Oh yes, I don't read too much into the Infectieradar number; it corresponds to ~3 individuals so error bars are big. But also things are still dropping so next few weeks will be interesting.

(Will be funny if there's a week with 0 infections in the 30,000+ cohort)

Here's the estimated excess drug overdose daily death toll due to interaction with the pandemic.

An interesting perverse consequence: The three Economic Impact Payments are each associated with a significant rise in the drug overdose excess (the known "check effect" of lump-sum payments).

2/2

Interesting paper here:

The drug overdose epidemic and Covid-19 pandemic have each claimed >1M lives in the US. Covid deaths are declining but overdose deaths are still increasing exponentially.

This paper find that the two combined were worse than on their own, by ~67,000 excess drug deaths.

1/

And that's the stuff we test for. If you've got cold symptoms right now in California, and you test negative for flu and covid, there's still <10% chance it's HMPV.

So this is cool and educational and fun, but, not really meaningful to the average person with a cold.

3/3

bsky.app/profile/did:...

Here's syndromic testing in the US:

Light blue is HMPV, just above the bigger dark blue which is all rhinoviruses/enteroviruses.

RSV (purple), and the 4 other non-covid human coronaviruses together (orange) CoVs, are detected a bit more often than the 4 co-circulating HMPV strains right now.

2/

While I think it's awesome that people are learning that there are many more respiratory viruses out there than flu and covid, HMPV isn't doing anything special this year.

(shown below is positivity for HMPV for CDC Region 9 incl California)

1/

Yes, I think T-cells are the under-rated heroes here.

In this thread I've shown replotted data from the following sources:

www.rivm.nl/en/coronavir...

www.rivm.nl/en/coronavir...

4/4

Just exactly *why* BA.3.2 hasn't done anything epidemiologically, despite all expectations, is one for the virologists and immunologists - I can only guess that it's to do with the fact that circulating antibodies aren't the sum total of the immune system.

3/

I've been watching the Netherlands as they're an early adopter of BA.3.2 (and they have fantastic surveillance!)

But BA.3.2 (red line on lower chart) hit 40% variant share 3 months ago, and ... nothing has happened since.

2/
bsky.app/profile/mich...

The Netherlands posted its lowest Covid-19 wastewater and prevalence ever last week:

WW concentration of 40 million RNA copies/person is down >99% from peak and lowest since recording started in Oct 2021

Infectieradar prevalence 0.01% (!) down 99.8% from peak, lowest since started Oct 2020

1/

Smallpox Champion (Albini Session), by Fugazi from the album Albini Sessions (Benefit for Letters Charity)

Gen-X time capsule unearthed.

(Fugazi's 1992 session recorded at Steve Albini's studio, just released 🤯)

fugazi.bandcamp.com/track/smallp...

Separate analysis of life expectancy for men and women would be interesting as deaths of despair disproportionately affect men.

Updating this - it looks to me like this divergence from the trend in the US is the "deaths of despair" phenomenon. Which is fairly complicated and ongoing in many countries; I wonder how this can be disentangled from covid mortality?

bsky.app/profile/mich...

4/ Rather than a uniform rebound, countries diverged into four distinct mortality patterns, with varying levels of overall life expectancy loss and recovery patterns. #demography #covid #episky

1/ Has life expectancy fully recovered from the COVID-19 pandemic? In a new pre-print, we find that 31 of 34 high-income countries had still not returned to their expected life expectancy trajectories five years after the onset of the COVID-19 pandemic. www.medrxiv.org/content/10.6... #demography

(P.S. don't take this too seriously as I am not actually Reviewer 2! And like Reviewer 2, I am asking for a pile of additional work that I understand is quite difficult and beyond the scope of the paper. I do appreciate any qualitative insights you can offer though!)

I don't know if the 2010 shock in the US is understood at all. But is seems to cast some uncertainty over whether the Lee and Carter model applies. (Forecasts work, until they don't!)

I had a look at the paper from Lee and Carter, and the US follows their forecast quite well, before a sharp turn in 2010. So I would conclude that there already is an unexplained shock in the US in 2010.
(Black is their figure, blue is observations for US from CDC.)

The reason I ask is that US life expectancy does seem to deviate from the (naive!) trend well before 2020. That seems to be seen in Canada (below), UK, and perhaps some other developed countries too.

By that I mean, for example, if you used 1995-2014 as the baseline, would you find that there is no significant systematic deviation in 2015-2019? A kind of negative control experiment.

Also, for @drjenndowd.bsky.social: If I may play "Reviewer 2" for a moment, I am curious whether, were the analysis run on a different set of years, could you accurately predict the *absence* of a shock?