That stat isn't necessarily causal though. Have they controlled for the fact that 8th graders who start using e-cigs are already more likely to start smoking to

That stat isn't necessarily causal though. Have they controlled for the fact that 8th graders who start using e-cigs are already more likely to start smoking tobacco cigarettes? Which seems like a very reasonably assumption to make