I don't think anyone should be too alarmed at the findings reported by Gavin Stewart and colleagues . But their observation that "autism traits as measured by the BAPQ [Broad Autism Phenotype Questionnaire] may confer additional risk of cognitive decline in aging" represents something that requires quite a bit of further investigation.
Some twenty 'older' adults who were questioned and deemed to have met criteria for the broader autism phenotype (BAP) were tested on skills related to executive function alongside episodic memory. Their results were compared with twenty 'older' adults who did not reach criteria for the BAP. Authors reported that: "Despite no differences in age, sex ratio, educational history or IQ, the BAP group demonstrated poorer performance on measures of executive function and episodic memory compared to the COA [control older adults] group." They interpret this in the context of that 'additional risk of cognitive decline in aging'.
The numbers of participants in the Stewart study were low and imply that one has to be quite careful about making any sweeping generalisations as a result. Bear also in mind that the BAP does not necessarily equal autism or autism spectrum disorder (ASD) as a function of it describing sub-threshold autistic traits (sub-threshold for a diagnosis). Traits, I might add, that are seemingly not just potentially representative of autism (see here for one example).
But set within a 'gap' in the research base looking at autism in older adults (see here), there is a scheme of work to follow. If for example, the Stewart 'trend' does overlap with the experiences of older adults on the autism spectrum in terms of 'cognitive decline', there is a whole barrage of potentially important implications to consider. More so when one considers that the autism prevalence data continues to head in only one direction (see here) and what this means for societal financial and resource planning.
Just before I go, one more detail was revealed in the Stewart paper: "Older adults who met the BAP criteria also reported higher levels of depression and anxiety." Continuing a theme on this blog that various over-represented issues/diagnoses in relation to autism might not be best described as just being 'comorbid' (see here and see here and see here), I believe that this finding adds further weight to the notion that autistic traits (clinical and sub-clinical) might have some important 'direct' relationships with other psychopathology. Not necessarily a welcome opinion in some quarters, but something that also requires a lot more investigation.
 Stewart GR. et al. Aging with elevated autistic traits: Cognitive functioning among older adults with the broad autism phenotype. Research in Autism Spectrum Disorders. 2018; 54: 27-36.