I cannot view the article but from the graph it seems "young" means those aged 18-44. They should have been more granular here because variations within this range would have been interesting to see as well.