As I understand it, the WHO conclusion (linked above) was based on a single study which concluded there was an association (and there were 30 studies showing no such association).
And, what about that single (and unfortunately influential) study? It was based on people’s recall, i.e. “now that you’ve had a brain tumor diagnosed, think back; did you use a cellphone in the past? how much”?
Such studies are next to useless and very often come up with spurious associations. One that comes to mind had to do with multiple sclerosis and dogs. People with MS were asked if they ever had a dog and, if so, did the dog have neurological troubles like seizures. Well, of course people with MS would be more likely to remember that type of thing given they themselves now had neurological problems (they may have had years to reflect on “why did this happen to me? what was I exposed to?”. But, if you’re well, it may slip your mind that Fido once shook all over for five minutes. And even if you do remember it, you may not consider it as being anything significant. If you have MS, you recall it as a seizure. Subsequently, prospective studies showed no association between MS, owning dogs, and/or your dog having neurological problems.