It's amazing how hard it still is to get properly analysed numbers, or indeed complete datasets susceptible to analysis. Here's what I've come up with merging the ONS data with the recently-released NHS stats.
296,603 people died in the UK up to the end of May. 50,107 (17%) of them had Covid19 on their death certificate. 44,796 of those (89%) were 65 or over. 4,789 were between 45 and 64 (just under 10%). 517 were between 15 and 44. 5 were under 15.
To give context, 8,018 people under 45 have died this year; only about 6.5% of them had Covid on their death certificate.
33,287 people between 45 and 64 have died this year. About 14% of them had Covid on their death certificates.
Consider co-morbidities (being sick, with something other than Covid). Let's ignore, shall we, the nosocomial nature of Covid apart from noting that 20% of eventual Covid fatalities in hospital didn't have Covid when they were admitted.
88% of people under 60 who died with Covid on their death certificates had an underlying condition, according to NHS data. The number goes up for older people - 97% of 80+ year-olds who died had an underlying condition.
Applying the 88% to the under 65 group, that means 574 people between 45 and 64 have died this year with a Covid diagnosis who weren't already ill. 63 people under 45 have died in similar circumstances.
We'll gloss over the fact that obesity on its own isn't counted in the NHS statistics, even though we know from ICNARC that it's a significant contributor to Covid fatality risk.
So 637 people have sadly died while neither old nor sick.
In the same time period 35,994 people under 65 have also sadly died, but without Covid on their death certificates.
We don't seem terribly exercised about them.