Cold school
The Swedish autumn semester started a few weeks ago, and, just like previous years, reported rhino- and enterovirus infections have started going up, according to tests from Stockholm healthcare (Karolinska Universitetslaboratoriet 2026). Infections with these non-enveloped viruses tend to increase rapidly during early autumn, remain high for months and then decrease somewhat during winter, but with relatively small seasonal variation: the ratio between highest and lowest detections over a year may be 2–4.
There has been a lot of discussion about the seasonality and transition to endemicity for SARS-CoV-2. The older human coronaviruses (OC43, NL63, HKU1, 229E), all follow a pattern with peaks in late autumn or winter, similar to other enveloped respiratory viruses (Price, Graham and Ramalingam 2019). SARS-CoV-2 cases did indeed increase already during the first autumns, 2020 and 2021, but the interpretation of this was unclear due to the complex interaction between changes in testing, contact patterns, vaccinations, and new variants. In 2022, cases fell after the initial omicron wave, but then started increasing during early summer, which led to a long autumn plateau, followed by an early winter wave, peaking in December. I discussed this in my Swedish 5 February 2023 post, where I also asked whether SARS-CoV-2 after that winter would settle into a seasonal pattern similar to the established human coronaviruses.
Later in 2023, there was a small late summer increase followed by an autumn wave, peaking in November, and then declining during the winter. In 2024, cases started increasing again in late summer, but there was no significant further increase during late autumn or winter. After a winter and spring decline, there was a new increase in late summer 2025, followed by an autumn plateau at a lower level than in 2024. It seemed that the seasonal pattern for SARS-CoV-2 had become similar to rhinoviruses, but that would be a surprising long-term endemic state for a betacoronavirus. In early 2026, SARS-CoV-2 cases gradually fell to very low levels, and have remained there. According to Karolinska Universitetslaboratoriet (2026), there has been about 10 cases per week for the weeks 31–35, which is similar to influenza A, with the same testing panels being used for both viruses. Will we see a distinct wave during late autumn or winter, with SARS-CoV-2 cases increasing rapidly and then decreasing? If such a pattern emerges and persists the coming seasons, this will mean that SARS-CoV-2, at last, has entered an endemic state similar to other coronaviruses. It is not known how long the corresponding transition took for these viruses: we do not know when they emerged, even though it has been suggested that OC43 was (partially) responsible for the respiratory pandemics around 1890 (Brüssow and Brüssow 2021), and modern virological surveillance was obviously not established back then.
In parallel with decreasing SARS-CoV-2 cases, mortality from covid-19 as underlying cause of death has also continued to decrease in Sweden (Socialstyrelsen 2026). During the season 2025/26 (where seasons are defined as 1 July one year to 30 June the next year), 116 women and 151 men have been reported dead from covid-19, and these figures are unlikely to rise much further, given the low number of cases during spring. This is a large decline, not only relative to the first seasons, when the majority of the population was naive, but also relative to 2022/23 (1080 deaths in women, and 1522 in men) or 2023/24 (591 deaths in women, and 798 in men), despite most people having both been vaccinated and had at least one infection. Back then, I was somewhat pessimistic about the future trajectory of covid-19 mortality, as I discussed, for example, in my Swedish 2 September 2022 post, where I compared SARS-CoV-2 with influenza A(H3N2), which has continued to cause rather high mortality for decades, mainly in older cohorts who were not exposed to that subtype at a young age.
For influenza as underlying cause of death, the season 2025/26 saw the highest number of deaths so far during this decade, with 320 deaths in women, and 315 in men. It was a mixed A(H3N2) and A(H1N1) season, and cumulative ICU admissions with influenza is 345 for the season (Svenska intensivvårdsregistret 2026), which is slightly below most seasons after 2015, excluding those where pandemic-related contact reductions led to decreased influenza circulation. The pattern of susceptibility versus age differs between the different A subtypes and influenza B, which may explain this discrepancy, because only 39 of the 635 influenza deaths, but 198 of the 345 ICU admissions, were in people younger than 70 years. A season dominated by influenza types with high susceptiblity in the old may thus have a high number of deaths, but relatively few ICU admissions.
Influenza B is generally most common in younger age groups, with some exceptions, for example the probably last B/Yamagata season 2017/18. This is probably is related to its lack of antigenic shifts by reassortment with animal viruses, as occurs with influenza A. Wall to Wall (1998), which was also aired on Swedish television during the early 2000s, tells the story about a young female bicycle courier and singer, played by Sharon Duncan-Brewster, who gets infected with influenza B (the species is explicitly mentioned in the episode, but not how it differs from influenza A), and contains a rather detailed description of the development of adaptive immunity.1 It could have been added that the immunity she gained probably still would protect her against severe disease form circulating B strains into her old age, because the strain infecting her will not be replaced by any new pandemic B type.
References
The mechanisms of fever are also discussed in detail. When the singing courier experiences headache and feverishness, she searches for pain relievers in a medicine cabinet, but does not find any, which is claimed to be for the better, as their fever-reducing effects may prolong the illness, according to the narrator. But the thought of checking her temperature never occurs to her. I suspect that this would have been different in, for example, a Scandinavian production, and I wonder whether not bothering with temperature-checking, or not even owning a thermometer, may be a British thing.↩︎