Age-adjusted mortality

October’s Age-Standardised Mortality Rate Was Equal to the Five-Year Average

The ONS announced last week that there were 43,435 deaths registered in England in October, which is about 1,000 less than in September, and 7.1% more than the five-year average.

This is a marked change from last month, when total deaths were 19.4% above the five-year average. Looking at the breakdown by leading cause of death, it is also quite different from September’s:

Last month, several non-Covid causes of death were above their five-year averages, notably dementia and Alzheimer’s, as well as ischemic heart disease. In October, by contrast, all non-Covid causes other than “Symptoms signs and ill-defined conditions” are below their five-year averages.

This suggests that my concerns about the delayed impact of lockdown on mortality may have been misplaced. In other words: last month’s elevated rates of death from non-Covid causes may have been a blip, rather than the start of trend toward rising mortality.

October’s overall age-standardised mortality rate was approximately equal to the five-year average – 0.1% lower, in fact. Again, this is a marked change from last month, when the age-standardised mortality rate was 11.2% higher than the five-year average.

Since age-adjusted excess mortality is the best gauge of how mortality is changing, the fact that October’s value is about equal to the five-year average indicates that any impact of lockdown on mortality must be relatively small. Here’s my updated chart of excess mortality in England since January of 2020:

Various newspapers have reported a large excess of non-Covid deaths in England over the past four months. However, these claims appear to be based on absolute excess deaths, rather than age-adjusted excess mortality.

In October, there were more than 2,000 non-Covid deaths in excess of the five-year average. Yet as I already mentioned, age-adjusted excess mortality was approximately zero – and that includes the Covid deaths. This means that the most of the ‘excess’ non-Covid deaths we’ve seen recently are due to population ageing over the last two years.

All in all, October’s figures are more encouraging than September’s, giving no indication that mortality is unusually high. Let’s just hope it stays that way.

Sweden Has Had *Negative* Excess Mortality Since the Start of 2020

We all remember what happened last year when Sweden’s unflappable state epidemiologist Anders Tegnell announced there wouldn’t be a lockdown. His “trust-based” approach was roundly denounced – not only in the media, but also by some ‘experts’.

As Johan Anderberg notes, Sweden’s pandemic strategy was variously described as “deadly folly” (Guardian), “a disaster” (Time magazine) and “the world’s cautionary tale” (New York Times).

Since the end of the first wave, however, Sweden has been gradually creeping down the list of countries by official Covid death rate. As of 16th October, it was ranked 52nd – well below the European average.

Yet this actually understates how well Sweden has done. As I and others have consistently argued, number of Covid deaths per million is not the best measure of the pandemic’s impact on mortality. Far better is age-adjusted excess mortality.

Thanks to an ONS report published on Thursday, we now have age-adjusted excess mortality numbers for most of the countries in Europe, covering the entire period from January 2020 to June 2021.

As an aside, the report clearly states: “The best way of comparing the mortality impact of the coronavirus (COVID-19) pandemic internationally is by looking at all-cause mortality compared with the five-year average.”

So what do the new numbers show? Sweden has had negative excess mortality. In other words, the level of mortality between January 2020 and June 2021 was lower than the five-year average. If this isn’t a vindication of Anders Tegnell’s approach, I don’t know what is.

The table below (taken from the ONS report) shows age-adjusted excess mortality from 3rd January 2020 to 18th June 2021. As you can see, Sweden is 8th from bottom, with a value of –2.3%.

Interestingly, the bottom six are all small, geographically peripheral countries (three islands, plus Denmark, Norway and Finland). This suggests that geography and border controls were key, and that lockdowns – in the absence of effective border controls – didn’t make much difference.

The top seven are all in Eastern Europe, which again suggests that some geographic factor is at work. What may account for high excess mortality in these countries is the fact that all of them missed the first wave, and hence had even bigger epidemics in the winter. Official Covid death rates are shown below:

If true, this would constitute strong evidence against the House of Commons’ report, which concluded that Britain should have tried to suppress the first wave. As I’ve noted before, this approach always carried the risk of creating an even bigger epidemic in the winter.  

In any event, Anders Tegnell can give himself a well-deserved pat on the back. His country kept civil liberties largely intact, and ended up with one of Europe’s lowest death tolls. Well done, professor.

UK Life Expectancy in 2020 Was Still at 2010 Levels and Over 80, OECD Report Shows

A new report from the OECD has shown that the pandemic took life expectancy in the UK in 2020 back to 2010 levels. Life expectancy at birth dropped by one year from 81.4 to 80.4, a level last seen in 2009. In 2008 it was even lower at 79.8.

This has largely been reported as something shocking – “Pandemic wipes out decade of progress on improving life expectancy”, declares the Telegraph – but in fact what it really shows is how limited the impact of the pandemic has been.

Despite all the daily reports of deaths, the running total of over 165,000 Covid deaths, and the repeated lockdowns imposed to protect a health service ever on the brink of collapse, the country has experienced a mortality rate no worse than 2009. I don’t know about you, but I can remember 2009. I don’t recall any lockdowns and panicking, or coerced experimental medicine, or bodies piling up in the morgues. Yet it was a worse year for deaths than the great pandemic year of 2020. Let that sink in.

Why are we destroying people’s lives and livelihoods and dismantling our freedoms to avoid going back to 2009 levels of mortality? Are we that obsessed with extending life at all costs that we regard it as intolerable to return to mortality levels last seen around the time the current party of Government came to power?

2021 is An Average Year for Deaths so Far. But What is Killing the Under-65s?

The latest quarterly report of the Mortality Monitor from the Institute and Faculty of Actuaries, with data to the end of September, brings some encouraging news and some worrying news. The encouraging news is that it shows overall mortality for 2021 so far is about average, meaning Covid has not yet managed to cause more people to die in 2021 than in a typical recent year. The worrying news is that, despite this, mortality in the under-65s is significantly elevated, and it’s not immediately obvious why.

The Mortality Monitor provides graphs and data for the cumulative age-adjusted mortality for England and Wales in 2021 compared to the 2011-20 average. This is a curve that starts at zero on January 1st each year, rises each week that there is above average mortality and falls each week there is below average mortality (see above). It rose at the start of the year during the winter Covid surge then fell as mortality dropped below average in March until, by mid-June, 2021 had below-average cumulative mortality. It started rising again at the end of July, and by the end of September had reached 0.3%, so about average. Note that the 2011-20 average includes the high mortality of 2020, but also the low mortality of 2019. Adjusting the mortality for age ensures that the ageing population is taken into account.

However, while overall mortality is about average, there are big differences by age. Mortality for over-65s is below average (which drives the overall trend, as the large majority of deaths occur in the over-65s), but mortality for under-65s is significantly elevated.

Cumulative mortality is running below average in those aged 65-84 at minus-0.5% and in those aged over 85 at minus-1.6%. But in those aged under 64 it is significantly up at 7.4%.

Age-Adjusted Excess Mortality Since January of 2020

Back in March, the ONS published a chart showing age-adjusted excess mortality since the start of 2020. This chart provides the best guide to the pandemic’s impact on deaths, for reasons I’ve explained before.

In short, it’s based on all-cause deaths, rather than ‘official’ Covid deaths, so it gets around the problem of deciding whether people who died of Covid would have died anyway. In addition, it’s based on age-standardised rates, meaning that it takes into account the changing age-structure of the population.

However, the ONS’s chart stops in February, so it’s a little out of date. Here I’ve produced a similar chart for England that goes all the way up to September of 2021:

Note: my chart looks slightly different from the ONS’s version because theirs is based on weekly data, whereas mine is based on monthly data. But the overall pattern from January 2020 to February 2021 is basically the same.

Clearly, the pandemic was at its deadliest in April of 2020, when the age-standardised mortality rate reached almost 200% of the five-year average. Since then, it has gone above and below the expected value of 100%. Note that the winter peak in January was only 21% higher than the five-year average.

If we take the average value since the start of 2020, it comes out as 5.8%. This is consistent with a recent ONS analysis, which reported an almost identical figure for the relative cumulative age-standardised mortality rate up to July. (The corresponding figure for Wales was only 3.5%.)

Interestingly, the average since the end of the first wave is less than 2%, meaning the pandemic hasn’t caused very many deaths over the past 15 months. This probably owes to a combination of factors: better treatments; better shielding of vulnerable populations; the vaccination program, and the gradual build-up of natural immunity.

Excessive focus on ‘official’ death numbers has led to a distorted picture of the pandemic’s impact on mortality – both in terms of the total increase and the distribution over time. Tracking age-adjusted excess mortality provides a much-needed corrective.

Are We Starting to See the Impact of Lockdown on Mortality?

The ONS announced on Friday that there were 44,474 deaths registered in England in September, which is about 4,000 more than in August, and 19.4% more than the five-year average.

19.4% is a non-trivial number, which makes this report slightly concerning. Last September, for example, the number of deaths registered in England was only 7% more than the five-year average.

If we look at the breakdown in the chart below, we see that Covid was the third leading cause of death. Interestingly, however, several other causes of death were above their five-year averages. This is in contrast to the situation in August, where eight out of nine other causes were below their five-year averages.

Notably, the age-standardised rates of death from dementia and Alzheimer’s, and from ischemic heart disease, were both above their five-year averages. Given that these are not respiratory conditions, the disparities are unlikely to be due to misattribution of deaths that were really caused by Covid.

The age-standardised rates of death from chronic lower respiratory diseases, from ill-defined conditions, and from colon and rectal cancers, were also above their five-year averages; although in the latter case, the disparity was negligible.

September’s overall age-standardised mortality rate was 11.2% higher than the five-year average, and was approximately equal to the value for March, which coincided with the final part of the second wave. This chart from the ONS shows the age-standardised mortality rate for the first nine months of the year, each year, going back to 2001:

Although the picture is basically the same as last month, cumulative mortality to date was slightly higher, as compared to previous years. However, the first nine months of this year were still less deadly than the corresponding period in 2015.

September witnessed the highest level of mortality since the end of the second wave. More interestingly, it was the first month of the year in which several causes of death other than Covid were above their five-year averages. After months of disruption to healthcare access and provision, are we now seeing the impact of lockdown on mortality?

August’s Age-Standardised Mortality Rate Was 2.5% Higher Than the Five-Year Average

The ONS announced on Tuesday that there were 40,460 deaths registered in England in August, which is approximately the same number as in July, and 9.9% more than the five-year average.

As you can see on this chart, weekly deaths remained above the five-year average for most of the month. Then in week 35, the August bank holiday artificially lowered death registrations:

Deaths being roughly 10% higher than the five-year average sounds like quite a lot. And in fact, the number of deaths registered in August of 2020 was 5.6% less than the five-year average.

Of course, infections were at a local minimum last August, and some of the deaths that would have occurred then had been brought forward by the pandemic. By contrast, August of 2021 coincided with the tail end of the Delta wave, and infections remained elevated throughout the month.

Consistent with this interpretation, COVID-19 was the third leading cause of death in August (a month when mortality is usually low) and deaths from eight of the nine other leading causes were below their five-year averages.

But as I always note in these updates, age-adjusted measures provide a much better guide to changes in mortality than the absolute number of deaths. In August, the age-standardised mortality rate was about the same as in July, and was only 2.5% higher than the five-year average.

This chart from the ONS shows the age-standardised mortality rate for the first eight months of the year, each year, going back to 2001:

As in the preceding two months, cumulative mortality to date was lower than the corresponding figures for both 2015 and 2018. In other words, the first eight months of 2018 – a year with no pandemic – were more deadly than the first eight months of 2021.

Overall then, 2021 is still a fairly normal year for mortality in England. As a matter of fact, it’s the sixth least deadly year on record! This could change, however, if the winter brings a particularly large wave of COVID-19 or seasonal flu.

This post has been updated.

July’s Age-Standardised Mortality Rate Was Equal to the Five-Year Average

The ONS announced on Monday that there were 40,467 deaths registered in England in July, which is 4.8% more than in June, and 7.6% more than the five-year average. In fact, the number of deaths registered in England was above the five-year average in all four weeks of last month.

These increases make sense, given that there has been a small uptick in COVID-19 deaths associated with the ‘Delta wave’. Although COVID-19 was only the ninth leading cause of death in July, deaths from the first eight causes were all below their five-year averages.

However, because the English population is ageing, the absolute number of people at risk of dying each year is going up. You’d therefore expect to see a greater number of deaths each year, even without a pandemic. What’s more, people who die from COVID-19 tend to be slightly older than those dying of other causes, so the average COVID-19 death is associated with fewer life-years lost.

For these reasons, it’s more informative to track age-adjusted measures of mortality. In July, the age-standardised mortality rate was only 1.3% higher than in May, and was approximately equal to the five-year average. (The exact figure was marginally higher, but the percentage difference was only 0.4%.)

This chart from the ONS shows the age-standardised mortality rate for the first seven months of the year, each year, going back to 2001:

Although 2021’s figure was higher than the figure for 2019, it was 3.6% lower than the figure for 2015 and 2.0% lower than the figure for 2018. This means that – despite higher-than-expected mortality in the winter – the overall level of mortality in the first seven months of 2021 was still lower than three years before.

As a matter of fact, the age-standardised rate from January through July was only 0.8% higher than the five-year average. Another month without many excess deaths and 2021 will officially be an ‘average year’ for English mortality.

Age-Standardised Mortality Rate Comes in Below Five-Year Average for *Fourth* Consecutive Month

The ONS announced on Friday that there were 38,611 deaths registered in England in June, which is 9.1% more than in May, and 0.8% more than the five-year average. However, the increase is relative to an exceptionally low value the month before. What about the age-standardised mortality rate (which is the best overall measure)?

In June, the age-standardised mortality rate was 12.5% higher than in May, but was still 6.1% lower than the five-year average. It was also the second-lowest figure on record for that month. (The only lower figure was observed in June of 2019.)

This means that England’s age-standardised mortality rate has been below the five-year average for four consecutive months. In other words, we’ve had four months in a row of “negative excess mortality”.

This chart from the ONS shows the age-standardised mortality rate for the first six months of the year, each year, going back to 2001:

Although 2021’s figure was higher than the figure for 2019, it was 3.6% lower than the figure for 2015 and 2.4% lower than the figure for 2018. This means that – despite higher-than-expected mortality in the winter – the overall level of mortality in the first half of 2021 was actually lower than three years before.

The past four months have “cancelled out” more than 85% of the age-adjusted excess mortality observed in January and February. Unsurprisingly, COVID-19 was not among the leading causes of death in June. All in all, the first half of 2021 has been pretty normal with respect to the average level of mortality.

What Pandemic? New Figures Reveal Age-Adjusted Mortality in the First Six Months of 2021 is Below the 10-year Average

New figures from the Institute and Faculty of Actuaries (IFA) Mortality Monitor released today show that cumulative mortality in the first six months of 2021 in England and Wales is running 0.4% below the 10-year average, once adjusted for the size and age of the population (see above). This means, despite the surge in winter Covid deaths in January and February that spooked the country into accepting ongoing restrictions, 2021 is officially now a low mortality year. The low mortality since March has entirely cancelled out the initial spike.

The IFA used to report the cumulative age-adjusted mortality figures compared to the 10-year average each week, but controversially changed their baseline in their weekly reports in May from the 10-year average to 2019 (the lowest mortality year on record) just as 2021 was about to go below average. This means we have had to wait for the quarterly report today for the next update in order to be able to announce this milestone.

England and Wales (BMJ)

The past 10 years are the least deadly years in history (see above), so for 2021 to be below the 10-year average (so far) means it too is one of the least deadly years in history. Even the pandemic year of 2020 was one of the least deadly years, having lower age-adjusted mortality than every year before 2009. Not all of the additional deaths are from COVID-19, of course – many are due to lockdowns and other aspects of the Government’s response and the attendant panic.

The new figures raise the obvious question: how can the Government justify continuing with any kind of restrictions or emergency measures for a moment longer when overall mortality is so low? Where is the ’emergency’ that justifies extraordinary measures?

Some will say it is only the restrictions that have prevented things being much worse. But where is the evidence of that? The U.K. has had more Covid deaths per head of population than countries such as Sweden which had fewer restrictions, while U.S. states which had few or no restrictions over the winter fared no worse (and often better) than those which had the strictest measures. Numerous studies have shown no relationship between restrictions and COVID-19 death tolls.

The latest figures, placing 2021 squarely among the least deadly years in history, should leave no one in any doubt that the emergency is well and truly over.