Don't you know?
Epidemiological studies do nothing to prove or disprove effects in a particular group or sub-group. In fact there's been a court ruling in the US along these lines.
Quite apart from that they are easy to structure and manipulate to get the response "needed".
For example, the Japan study: MMR withdrawn, no change in autism cases. Details: MMR replaced by single vaccines given close together, sometimes a week apart, a month apart, sometimes the same day. Fact: the effects of vaccines take up to a month to emerge. Rashes three weeks later and so on. Fact: Wakefield posited that might be the effects of multiple assaults on the immune system close together that might be a problem. Japan study therefore useless, certainly as a disproof of the MMR / autism theory.
Example: Danish study: three million children, no connection with autism. Fact: a retrospective rather than prospective study, parents not asked to come forward with events, relied on medical records. Fact: the outcomes of only a fraction, hundreds, if that, rather than three million were studied and uses in the statistics. Funded by Merck. Useless as a disproof of MMR-autism.
Example: study of children between approx 1990 and 1999 showing no difference in autism between vaccinated and unvaccinated children. Fact: vaccination around 18 months, autism diagnosis in the study on average just under five years. When the study ended a tranche of children, three and a half years worth, had been vaccinated and were counted as "undiagnosed". When in fact those autism diagnoses hadn't had a chance to be made. Highly misleading. Useless as a disproof of MMR-autism.
These are three of the epidemiological studies that are supposed to reassure us.
Your last sentence is dreadful scaremongering.