Number one

Last week, National Board of Health and Welfare in Sweden released statistics on causes of death for the whole of 2024 (National Board of Health and Welfare 2025), which was noted in one of Sweden’s largest daily newspapers, by Nilsson (2025). That article mentions that Swedish all-cause death rates have declined with over 50 percent since 1969, which to a large extent has been driven by lower circulatory mortality. As a result of that, the article claims, cancer has become the most common cause of death in several regions. In 2024, neoplastic disease was the most common cause of death among women in Stockholm, Uppsala, and Gotland, and among men in Stockholm, Halland, and Gotland. The heading of the article says that cancer is getting more common as a cause of death.

I have written about this framing before, both in a Swedish and a US context, as in my Swedish 25 August 2016 post. Rankings of causes of death have often been based on the chapters in ICD classification. Sometimes, other partitions are used: e.g. heart disease is often treated as a category in the US. In middle- and high-income countries, circulatory disease, or heart disease, has then often been the number one cause, and neoplasms, i.e. almost entirely cancer, number two. Given this, things may seem simple: if you see a fast decline in mortality rates from circulatory disease, the ratio circulation/cancer will decrease, and eventually will be lower than 1, so that cancer will be the most common cause. And one may also then conclude that cancer must have become significantly more common in terms of its share of all deaths, even though age-standardised mortality rates still may have declined. But does that give an accurate picture of the actual development of the cause pattern in e.g. the Swedish regions?

With my morr R package, cause of death patterns over time can be plotted using the data available via WHO (2025), and also National Board of Health and Welfare (2025), for regional Swedish data, starting with 1997, i.e. the first year ICD-10 was used. Made with the functions from the morr package, fig. 1, fig. 2, fig. 3 and fig. 4 show sex-specific cause of death patterns for all ages for the four regions mentioned by Nilsson (2025).¹

As the figure show, the patterns are somewhat unstable from year to year, especially in Gotland, with its small population. But the overall pattern is clear: diseases in the circulatory chapter, i.e. the cause categories down to and including stroke (excluding vascular dementia), are getting less common, and their ratio to neoplasms has decreased, and was in 2024 lower than 1 for at least one sex in all four regions. But the neoplasm category itself has been rather stable when it comes to its share of all deaths. It is not even hard to find earlier years with somewhat higher share of neoplasm deaths for the sex- and country-combinations with a circulation/neoplasm ratio lower than 1 in 2024.

• In Stockholm, some neoplasm was underlying cause in 25.8/27.6 percent of all deaths among women/men in 2024. That share was highest among women in 2019 (26.5 percent), and among men in 2015 (28.9 percent). Already in 2005, the share was higher among men than in 2024 (27.8 percent).
• In Uppsala, some neoplasm was underlying cause in 26.8/28.8 percent of all deaths among women/men in 2024. That share was highest among women in 2021 (29.0 percent), and among men in 2011 (30.8 percent). Already in 2003, the share was higher among women than in 2024 (26.9 percent), and in 2001, it was higher among men than in 2024 (29.9 percent).
• In Gotland, some neoplasm was underlying cause in 31.3/32.2 percent of all deaths among women/men in 2024. That share was highest among men in 2004 (32.9 percent).
• In Halland, some neoplasm was underlying cause in 25.9/27.8 percent of all deaths among women/men in 2024. That share was highest among women in 2012 (28.8 percent), and among men also in 2012 (32.2 percent). Already in 2001, the share was higher among men than in 2024 (29.4 percent).

In my 4 July post, I described the overall changes in cause of death pattern in Sweden and other countries for a longer period of time, and the trends in Sweden as a whole during the ICD-10 period are, of course, also reflected on the regional level. Myocardial infarction has decreased rapidly, but there has been little change in other heart diseases, taken as a group. Stroke and arterial diseases except for ischemic heart disease and stroke (including generalized atherosclerosis) have also decreased a lot, mirroring a marked increase in dementia, which is not included in the circulatory chapter in ICD-10, even when specified as vascular. And covid-19, of course, disturbed the cause pattern when it arrived in 2020, particularly in Stockholm, where much of the spring wave was concentrated, but has more recently diminished in importance with increasing immunity in the population.

One thing that seems clear is that it is not informative to treat cause of death patterns as elections in single-member districts, where getting to number one is what matters, regardless of other aspects of the distribution of votes.

References

---

1. The figures may be reproduced in R by cloning the blog repository and running 2025-09-28-one.r in the subdirectory postdata/2025-09-28-one. The causes are listed in the capat vector in that file, using the cause definitions with ICD codes from the morr configuration file.↩︎

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.