Half heart

In my post 9 December 2024, I wrote about my Bluesky quiz, which turned out to be about time periods for which different countries had reported circulatory disease as underlying cause for at least 50 percent of deaths in their female and male populations (a property denoted X in this post).

I made the tables with some wrangling in Visidata (Pwanson 2026), but recently, I built a function in my R package morr to make similar tables over countries and years where the ratio of two causes of death exceed a certain threshold. By cloning the blog repository, and running the script 2026-04-12-half.R in the subdirectory postdata/2026-04-12-half, one should, provided that the morr package and the data files from WHO (2026) are available, be able to reproduce the tab-delimited files in postdata/2026-04-12-half/data.

These files show, for female (prop_x_f.tsv) and male (prop_x_m.tsv) populations, the lowest (column yr_min) and highest (yr_max) years, and number of years (count) where the ratio between circulatory deaths and total deaths for all ages is at lest 0.5, including only populations with at least 200 total deaths, as in the original quiz. The columns with names ending with _all show the corresponding values for years with available data for which this last condition is satisfied.

Using the latest update of WHO (2026), one can discern certain patterns, both regarding property X and the general levels of reporting during recent years of pandemic and war. Focusing on the female populations, because X has generally been more common for these, one may distinguish between the following categories.

1. Populations newer reporting X for any year with at least 200 deaths, and thus not included in the tables. These may be divided in further subcategories.
   1. Populations which have never reported mortality to WHO, or have done so very sporadically, including many countries in, for example, Africa and South Asia. Some countries have used special, limited codes for subsets of the population, like mainland China (last report in 2000), and have not been included.
   2. Microstates and territories with very small populations.
   3. Populations with long time-series in WHO (2026), but still never reporting X, like Japan, Hong Kong, many Latin American populations, but also some European populations, like Belgium, France, and Netherlands.
2. Populations reporting X before 2000, but not after that.
   1. No longer existing countries and territories, like West Berlin, West and East Germany, USSR, and Yugoslavia.
   2. Countries reporting X for some time during the 20th Century, but then shifting to a different cause pattern. As mentioned in my earlier post, this includes Anglosphere countries (US, Canada, the UK subcountries, Ireland, Australia, New Zealand), Nordic countries (Sweden, Finland, Norway, and, for fewer years, Denmark and Iceland), some Central and Southern European countries (Switzerland, Luxembourg, Italy, Spain, Malta, Slovenia and, just for 1991, Portugal), and some other high-income countries (Israel).
3. Populations reporting X for some years after 2000.
   1. Populations reporting X for some years after 2000, but then shifting to a different cause pattern. This includes some European countries, like Germany, Austria, Greece, Poland, Croatia, Czech Republic, Bosnia and Herzegovina, North Macedonia, Georgia, and also some Asian countries like Turkey, Tajikistan and Kazakhstan, and Mauritius, one of few African countries with long time-series available.
   2. Populations reporting X no more than a few years before their last year with available mortality data but without any data later than 2019 (the last year before covid-19 could affect the cause patterns). This includes Azerbaijan (last report 2007), Albania (2009), Syria (2010), Belarus (2018), Russia, Ukraine, and Egypt (all 2019).
   3. Populations still reporting X 2023–24. This includes all the Baltic countries, i.e. Estonia, Latvia, and Lithuania, some other Balkan and Eastern European countries, including Slovakia, Hungary, Bulgaria, Romania, Serbia, and Moldova, and some former Soviet republics in Asia, including Armenia, Kyrgyzstan, and Uzbekistan.

This clearly shows what I have noted before, that X has gone from a pattern typical for rich Anglosphere or Northern European populations to something mainly seen in former Eastern Bloc countries in Eastern Europe and Asia, and has disappeared even for some of those.

It is also striking how patchy the coverage of WHO (2026) still is. Some of the largest populations, with rapid economic development in recent decades, like India and China, lack any comprehensive data. Other countries, like Syria, Ukraine, and Russia, have relatively long time-series, but have been involved in war or upheaval in recent years, and lack data from the 2020s. Modelings like IHME (2025) may be lacking in transparency, and may, to a large extent, be based on questionable extrapolations from a relatively small subset of all countries.

Both the sparsity of updated data and different practices for ascribing deaths to underlying causes complicate things like assessment of mortality related the covid-19. For many of the countries still having X in recent years, Polizzi et al. (2024) show significant life expectancy losses related to circulatory causes during 2020 and 2021, which may not be very surprising.

References

Webbplatsen drivs av Karl Pettersson och har genererats med Hakyll. Källkod. Om fel uppstår vid besök, kan uppgifter om besökares IP-adress och tid för felet lagras i interna loggar och användas vid felsökning.