A rather interesting paper by Beth Oakley and colleagues  (open-access might be available here) appeared recently providing a "cautionary note on the Reading the Mind in the Eyes Test" , one of the premier assessments thought to offer a performance-based measure "involving mental state attribution and complex facial emotion recognition from photographs where only the eye region of the face is available."
Most people with some knowledge about autism research history will have heard about the proposal that Theory of Mind (ToM) - a term often used to cover that "mental state attribution" - might be affected in cases of autism (see here) and indeed, how careers and reputations have been made on such a generalisation. These days ToM is less and less being talked about as the heterogeneity of the autism spectrum becomes better understood and how specificity in particular, has proved to be an Achilles' heel for the concept (see here).
One of the emerging ideas to account for some of the results obtained using the Reading the Mind in the Eyes test (RMET) in cases of autism is that alexithymia - a construct characterised by an inability to describe or understand emotions - might actually be the more important issue than autism per se. The idea being that alexithymia can co-occur alongside some autism and that for those presenting with that combination, ToM and assessments like RMET might be problematic.
So Oakley et al delved a little deeper into some of the hows and whys of some of the RMET results obtained with autism in mind and whether "the RMET indexes emotion recognition, associated with alexithymia, or ToM, associated with ASD [autism spectrum disorder]." They did it on the basis of examining a small group of participants diagnosed with an ASD (n=19) alongside 24 participants without autism. Alexithymia was assessed using "the 20-item Toronto Alexithymia Scale (TAS–20)." Autism symptoms severity was measured using the Autism Spectrum Quotient (50) (oh dear..) and "current functioning" in the autism group was assessed using the gold-standard that is the Autism Diagnostic Observation Schedule (ADOS) (more like it).
Results: well bearing in mind the small participant numbers and the need for further independent replication of the findings, "Reading the Mind in the Eyes Test (RMET) performance was unaffected by autism spectrum disorder... but was negatively impacted by alexithymia." Indeed, we are told that: "Six ASD and eight control participants met the criterion for severe alexithymia, with a score of 61 or above on the 20-item Toronto Alexithymia Scale (TAS–20)." Further: "in individuals with ASD and comorbid alexithymia, it is alexithymia, rather than ASD per se, that impairs emotion recognition performance."
I probably don't need to say too much more about this line of research and its important implications outside of perhaps the requirement to screen for alexithymia as and when autism is diagnosed. Insofar as the idea of an "alexithymia hypothesis of emotion-related deficits in ASD", this does sound like a tantalising option but again, I'd be slightly reluctant to go all-in with yet another grand theory for autism given the trials and tribulations that psychological theories in particular have faced over the years with autism in mind. As to other potential impacts from work such as this, well, assertions that a lack of empathy might the root of all evil (see here) also made by proponents of ToM might do well to take on board a role for alexithymia in any future judgements...
Minus any charges of plagiarism, I leave you with the general summary from Oakley and colleagues:
"This study suggests that a highly popular test of the ability to detect what someone else is thinking—the Reading the Mind in the Eyes Test—is instead a test of the ability to recognize another person’s emotional expression. This is important because it suggests that patients who perform badly on this test may still be able to understand another person’s mental state and that, conversely, patients who perform well on this test may still have difficulties in mental state understanding."
 Oakley BF. et al. Theory of mind is not theory of emotion: A cautionary note on the Reading the Mind in the Eyes Test. J Abnorm Psychol. 2016 Aug;125(6):818-23.
 Baron-Cohen S. et al. The “Reading the Mind in the Eyes” Test: Complete Absence of Typical Sex Difference in ~400 Men and Women with Autism. PLoS ONE. 2015; 10(8).
Oakley BF, Brewer R, Bird G, & Catmur C (2016). Theory of mind is not theory of emotion: A cautionary note on the Reading the Mind in the Eyes Test. Journal of abnormal psychology, 125 (6), 818-23 PMID: 27505409