No, you cannot conclude that. You say "surely existed before the introduction of smartphones", but you are also saying that the smartphones have introduced something that did not existed. If it's the case, it means that "new things can be introduced". You still need to prove the real mechanism that really modify the mental health and that is really specific to smartphones. For all we know, plenty of other things were introduced amongst the same time. For example, maybe in a parallel universe where everything is identical but smartphones were not introduced, you would have exactly the same depression curve, because, I don't know, the real source is the social media and people without smartphone will access it on laptop, or because the real source is disillusion from prospects as that generation has noticed they will never have a life as easy as their parents (which was not the case in the past either).

To be clear, it is 100% possible that it is the smartphone, but you cannot "scientifically conclude" it's the smartphone, not with the data that we have. It's not even a question of p-value, the problem is that this conclusion is too much driven by "it feels logical to me" and confirmation bias, rather than honestly trying to drive a conclusion from observations.