OK, Time to poke fun at the 'science'. This is not a prospective study; the subjects already had the outcome (ASD) with a concurrent exposure (mercury).
Their samples were sent to an outside laboratory; so other investigators cannot verify the methods and there is too much test variability due to instrument calibration and different technicians.
Many of the Geiers' references and therefore hypothesis is based upon chronic, low and high level occupational exposure to Hg (primarily vapour) which is fine as long as they don't overreach the data results.
The heme pathway is sensitive to inhibition by both organic and inorganic agents and thus the atypical porphyrinuria pattern is not exclusive to Hg exposure alone and the testing is not standardised to allow for the distinction and certainly cannot differentiate Hg species.
The controls. They are not matched as is claimed and are not robust enough; not matched for exposure and the controls were claimed to be siblings of ASD patients, not siblings of the study patients.
Table 2: I guess they could have achieved statistical significance somehow, somewhere but the table is sloppy and suspect e.g. they report some means with 1 significant figure but the standard deviations with 2. How can this be? The size of the standard deviations as compared to the means suggests more variation within the groups than between the groups.
I think that selection bias was required to even perform this study rather than control for it. Chelated patients had similar porphyrin levels as controls but they don't know or report what the pre-chelation levels were and even if chelation even had an effect.
The discussion was a predictable association with previous Geier 'studies' and the usual suspects with vanity press articles along with cherry-picking select interpretations from some legitimate studies. There is no association with vaccines so I don't know where that notion came from. All in all, a poorly done and predictable study and is nothing more than an entry on their CVs.