The DISCO series of diagnostic interviews, which I hadn't heard of before seeing this article, focus on early developmental history and are conducted between the diagnostician and the potentially autistic person's parent or other caregiver. It is used not so much to diagnose autism as to uncover "the pattern over time of the skills and impairments that underlie the overt behaviour."
In Chapter 28 of the Handbook of Autism and Pervasive Developmental Disorders, which describes and evaluates all the different tools available for diagnosing autism, Catherine Lord and Christina Corsello say:
[T]he first version of the DISCO was developed to assess the pattern of development in individuals with ASDs and their individual needs (Wing et al., 2002). The primary purpose of the DISCO is not to provide a diagnostic classification. Rather, the instrument was designed to obtain information on behaviors relevant to autism for the purpose of assisting clinicians in determining a child's development in different areas as well as his individual needs (Leekam, Libby, Wing, Gould, & Taylor, 2002). It is based on the concept of a spectrum of disorders rather than categorical diagnoses.So, while other tests might be better suited to answer the question "Does this person have autism?" the DISCO seems to be more geared toward answering the question "What does this person's autism look like?"
The DISCO is an investigator-based interview in which the interviewer asks questions designed to elicit descriptions of behavior and makes coding decisions based on the information provided. ... The DISCO includes items covering behavioral manifestations of the deficits associated with ASDs, including social interaction, communication, imagination, and repetitive activities. In addition, it includes items designed to assess developmental levels in a variety of domains. Many of these items are based on the Vineland Adaptive Behavior Scales [link] (Sparrow et al., 1984). There is also a section on atypical behaviors that are not specific to autism. These include unusual responses to sensory stimuli, differences in attention and activity level, challenging behaviors, and other psychiatric disorders. Items relating to developmental delay are rated on a 3-point scale, as "delay," "minor delay," or "no problem." An actual age is coded for some of the developmental items. Atypical behaviors receive codes for "current" and "ever" and are rated as "severe," "minor," or "not present."
Anyway, the researchers did a full psychodiagnostic workup on those children and teens who turned up at the clinic without having been evaluated for autism before, and quizzed the parents of those kids who had previously been evaluated in the above-described manner.
They found that 6% of the 233 young people referred to the clinic (14 people or so) had some kind of ASD. This rate --- which the LB/RB blogger point out is about six times the rate at which autism occurs in the general population --- remained consistent regardless of gender-identity outcome: kids who ended up with a diagnosis of full-blown Gender Identity Disorder were no more or less likely to be autistic than the kids diagnosed with GID-NOS or the kids discharged without any diagnosis.
Most interestingly, the people who did end up with dual diagnoses of ASD and GID included both sexes, which, if that finding is replicated in larger studies of autism prevalence among trans or otherwise gender-variant people, could come into conflict with Simon Baron-Cohen's Extreme Male Brain theory of autism, which hypothesizes a direct relationship between fetal testosterone levels, masculinity, and autism. A high proportion of autistic trans men and butch women would thus be in accordance with theory, but not trans women or feminine men.