I think the "on average" was clearly implied.

Mathematically, it might be close to average but I don't think that's a great way of putting it either. If you click through to the study, it was actually only men that died earlier. The lifespan of women was unchanged.