|Lundström et al. 2015. BMJ
As well as looking at the annual numbers of children with a reported diagnosis of autism spectrum disorder (ASD), researchers also looked at data that had assessed for "the autism symptom phenotype" over each year based on parental responses to the Autism-Tics, ADHD and other Comorbidities inventory (A-TAC). I've talked about the A-TAC quite recently (see here) and it's usefulness in assessing the 'ESSENCE' of autism. A-TAC by the way, should not be confused with the ATEC (see here).
Results, and a long quote coming up: "Using unique, large Swedish population based resources, we found that the annual prevalence of the autism symptom phenotype was stable over a 10 year period when investigating 9 and 12 year old children, while simultaneously the annual prevalence of clinically diagnosed autism spectrum disorder in a service based register steadily increased."
What this translates into is that the numbers of children presenting with a profile that looks like the autism phenotype based on the A-TAC results remained pretty steady year on year over the investigated period. When it came however to the numbers of children being diagnosed with autism or an ASD, there was "an almost linear increase over the examined years." The authors translate this as providing support for the idea "that administrative factors that affect the registered prevalence may account for much of the rise in the reported prevalence of autism spectrum disorder" quashing the notion of any major actual increase in the rate of autism. Not every piece of previous research has wholeheartedly agreed with that last sentence (see here and see here).
These are quite compelling data insofar as the numbers of participants included for study and the comprehensiveness of the various Scandinavian patient registers. The authors do however point out that the A-TAC "does not have perfect sensitivity or specificity, meaning that some degree of “diagnostic misclassification” should be expected" albeit probably not greatly affecting the results obtained given also it's equal use over the study years. I might also reiterate that this study looked at prevalence between birth years of 1993 and 2002 so subsequent study using the same methods should provide a clearer picture of more contemporary times also.
Without wishing to argue with the conclusions reached by Lundström et al, I did decide to have a closer look at some of the data presented in the paper. As per the graph presented above based on the information presented in Table 2 of the paper showing the calculated prevalences of registered diagnoses from the NPR data over the period of study and the "clinically validated cut-off for autism spectrum disorder of ≥8.5 to define the autism symptom phenotype" for the same period, there are some interesting details. I've not included the confidence interval (CI) data included in the original figures so one has to be a little guarded about drawing too many conclusions but the point raised by the authors about A-TAC (autism symptom phenotype) scores being "stable over a 10 year period" is not something necessarily as straight-forward as one might first think (follow the blue line).
"Research and clinical practice need to refocus on the child’s overall clinical situation and to acknowledge that autism is but one of the many Early Symptomatic Syndromes Eliciting Neurodevelopmental Clinical Examinations (ESSENCE)." I would very definitely agree with the authors' sentiments here and how the presentation of autism as part of a wider spectrum of issues (some perhaps at times even more 'disabling' than core autism symptoms) is something requiring much greater prominence in these days of overlapping comorbidity (see here). That some of that comorbidity might also extend into more 'somatic' features (see here) is also an important point to make with various implications for identification and appropriate management.
Update: 30 April 2015: I'll also draw your attention to the paper from Van Naarden Braun et al  (open-access) also providing some rather relevant data.
 Lundström S. et al. Autism phenotype versus registered diagnosis in Swedish children: prevalence trends over 10 years in general population samples. BMJ 2015; 350.
 Van Naarden Braun K. et al. Trends in the Prevalence of Autism Spectrum Disorder, Cerebral Palsy, Hearing Loss, Intellectual Disability, and Vision Impairment, Metropolitan Atlanta, 1991–2010. PLoS ONE. 2015; 10(4): e0124120.
Lundstrom, S., Reichenberg, A., Anckarsater, H., Lichtenstein, P., & Gillberg, C. (2015). Autism phenotype versus registered diagnosis in Swedish children: prevalence trends over 10 years in general population samples BMJ, 350 (apr28 2) DOI: 10.1136/bmj.h1961