It's a cross-sectional study and makes no mention of the timing of vaping vs asthma diagnosis. Which suggests they did not look at timing, a pretty important element of any claims which hint at possible causal effects.

It's not necessarily the same story but it's very reminiscent of this retracted study, which claimed to have found a link between vaping and heart attacks without checking which came first. In the case of heart disease, an obvious interpretation is that a lot of people quit smoking after they run into heart troubles, and vaping is a very effective way to quit for many people. The authors' failure to satisfactorily address that obvious point is what caused the retraction.

There was a systematic review of the asthma question published last year: Association Between E-Cigarettes and Asthma in Adolescents: A Systematic Review and Meta-Analysis. The results of this study are consistent with that review but all of the included studies are cross-sectional so they all have the same weakness when it comes to considering causality. Which is kind of weird because it shouldn't be that hard to establish the timing of a diagnosis of asthma.

One possible causal pathway isn't vaping itself but the use of flavour chemicals called diketones (the best known being diacetyl), which are known to cause COPD in food factory workers. Asthma isn't COPD but the diagnostic criteria are a bit woolly (asthma is intermittent, COPD is permanent but can still fluctuate in severity of symptoms). Diketones are one of the very real risks of vaping if manufacturers are careless but it is dismissed by too many vapers (and manufacturers) who are jaded by the avalanche of scare stories based on ridiculously bad science. It's also barely been addressed by regulatory authorities (because Big Food is still fighting off compensation claims).

I have no idea what the 'correct' answer is here but this study, and others like it, is not well-designed for causal claims and it is strange that there don't seem to be any studies which include timing as a variable. It's always difficult to prove causality with observational designs but cause happening before event is pretty fundamental and not particularly difficult to include.