boys 15-19 that died by suicide: 0.01% -> 0.016%. A 34% increase!

Relative increase percentages (i.e. exponentials) are misleading when they are measuring linear phenomena like suicide rate. You could just as easily say there was a 0.1% decrease in the percent of boys who didn't die by suicide.

It's misleading because the new 0.006% of male suicides are unrelated to the existing 0.01%.

Suicide rate is not a ROI or conversion rate. Use absolute percentages please. I get they're writing to accentuate big impact -- and suicide is a big deal -- but this is disingenuous.