"Our sample comprised over 1,000 Top 40 recordings from 25 years spanning five decades. Over the years, popular recordings became longer in duration and the proportion of female artists increased. In line with our principal hypotheses, there was also an increase in the use of minor mode and a decrease in average tempo, confirming that popular music became more sad-sounding over time...."

https://psycnet.apa.org/record/2012-12935-001

I like the methodology in general, but boy howdy does "25 years spanning five decades" set off alarm bells. Which 25? Why not ~50?

(It's not necessarily a problem; every other year for instance probably wouldn't be a cherry picked slice.)

Do you have any data to support that more generally?

No. I threw out all my CDs years ago.