As another commenter said, you put the bar arbitrarily high by requiring it to be "more of a risk of death than any other possible event or disease that can happen to healthy people in that age group", because of course, once one is "unhealthy" (whatever that means, as healthy and unhealthy is not binary but a spectrum), then their death doesn't count for data. But here is an article with sources saying it was the number four cause of death for those aged 25-34 in 2021 [0].

[0] "And for those ages 15 to 24 and 25 to 34, COVID-19 wasn't in the top five in 2020 but ranked as the fourth leading cause of death in both age groups in 2021." - https://arstechnica.com/science/2022/07/covid-was-the-leadin...