"risk their lives"
Unless they're old or vulnerable -- then they're not risking their lives. There's a likelihood that they may catch an incredibly mild virus.
inb4 "it's deadly"
Yes, to the vulnerable and the old it is, to everyone else...not so much.
It's not a literal death ray, unless you consider walking to work/driving in the same realms of risking your life.
This paper does not include or support the assertion you are making. It is not a valid or plausible source for the claim for which you have used it as a citation.
The leap you have made is far beyond the already tortured reading that this poor paper has received throughout social media. In order for the same number of people (leaving aside the "more than twice" claim in your comment) in the 25-44 cohort covered by this paper to die of COVID19 as died of opioid overdoses in 2018 (which is the figure this paper uses), essentially all of the other 62% of the increase in all-cause mortality which does not yet have an available documented cause of death, among a group especially vulnerable to the second-order effects of lockdowns, would have to instead have come from a virus that is known to kill them at vanishingly rare rates and diagnosis of which is now routine nationwide.
Put another way: two-thirds of the lethal COVID19 cases in the 25-44 cohort have to have been completely missed by this nation's medical professionals (and even then, opioid deaths have to have remained at 2018 levels, which is also unlikely). This is just in order for COVID19 deaths to be close to the number of opioid overdose deaths. It is completely impossible for COVID19 to kill twice as many.
To be clear: all of the reliable data - every shred of it - shows convincingly that the assertion you are making is strictly false. The number of opioid overdoses has far outpaced the number of COVID19 deaths (having increased since 2018 rather than remaining flat as you implicitly assert); the same is almost certainly true of car crashes. Depending on what you mean by "young adults" (ie, if you are including the 18-25 cohort in addition to the 25-44 cohort studied in this paper), it will be surprising if COVID19 kills even a third as many will have died in car crashes, as this group dies even more rarely of COVID19 and even more frequently of car crashes.
Moreover, even if your claim were within 60% of being true (and there's no reason, given the available data, to believe that it is) the comment to which you are responding said "old or vulnerable" (emphasis mine), whereas the paper you've provided showing over 4,000 deaths (again, just to be crystal clear, this is less than half of those from opioid overdoses) includes people in this age group with other (sometimes grave) comorbidities.
It is true that people who are neither old nor suffering from one of the three most frequently indicated comorbidities have a risk of adverse effects similar to seasonal coronaviruses.
https://onlinelibrary.wiley.com/doi/10.1111/eci.13423
tl;dr: Around 4,000 people under 45 have died COVID19-related deaths, while more than 10,000 died of opioid overdoses in 2018, the latest year for which a reliable figure is available (but every credible expert believes that the number will be much higher this year).
1) Without having access to the relevant data (and doubting that it is exists in the first place), I think it's reasonable to surmise that this particular group is more likely to either be in a vulnerable sub-population or to live in a multi-generational household than any other group of front-line medical workers.
2) Front-line medical workers have experienced adverse outcomes which have outpaced the risk based solely on their age and comorbidities. There has been widespread speculation (and some research) that this is due to a higher likelihood of receiving a higher initial dose of the virus.