Autism spectrum disorders (ASDs) in older adults have been neglected for a long time. As far as we know only five papers have been published. No empirical research in this area was found. Four papers were case studies of men diagnosed with an ASD (James et al., Reference James, Mukaetova-Ladinska, Reichelt, Briel and Scully2006; Naidu et al., Reference Naidu, James, Mukatoeva-Ladinska and Briel2006; van Alphen and Heijnen-Kohl, Reference van Alphen and Heijnen-Kohl2009; van Niekerk et al., Reference van Niekerk, Groen, Vissers, van Driel-de Jong, Kan and Oude Voshaar2011), and the fifth one was an opinion paper concerning the diagnosis of ASD in the elderly and the difficulties arising in this (Heijnen-Kohl and van Alphen, Reference Heijnen-Kohl and Van Alphen2009).

Nevertheless, all publications emphasize the importance of the detection of ASD to reduce patient burden and improve care delivery. Children, adolescents, and adults suffering from ASD benefit, for example, from psycho-education and a more structured day and week program (Volkmar et al., Reference Volkmar, Cook, Pomeroy, Realmuto and Tanguay2000). Furthermore, the diagnosis of an ASD probably also directs the referrals and requirements for the most optimal living arrangements. Empirical evidence concerning older adults is thus urgently needed.

Diagnosing ASD in later life is complicated for several reasons. First, the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) criteria for ASD (American Psychiatric Association, 2000) are developed for children without any information given on age-sensitive interpretations. Second, criteria focusing on the earliest years of life cannot be applied reliably. Informants have usually passed away and the patient's memories about their youth may not be accurate (van Alphen and Heijnen-Kohl, Reference van Alphen and Heijnen-Kohl2009). Third, there are no specific measures for ASD that are suitable for older people. The existing assessment is developed for diagnosing ASD in children, and therefore substantial items refer to early childhood (van Niekerk et al., Reference van Niekerk, Groen, Vissers, van Driel-de Jong, Kan and Oude Voshaar2011). Fourth, diagnosing ASD in general is difficult because the actual assessment relies mainly on self-report, and it has been confirmed that individuals suffering from an ASD can have a distorted view about their own functioning and behavior (van Niekerk et al., Reference van Niekerk, Groen, Vissers, van Driel-de Jong, Kan and Oude Voshaar2011).

In their case-series, van Niekerk and colleagues (2011) stress the importance of informant-based information for diagnosing ASD. If true, diagnosing ASD in older adults may significantly improve if informant-based information can be gathered in more reliable way. For this reason we initiated the research project to detect ASD in later life by structured informant information based on behavioral manifestations of the Hetero-Anamnestic Personality questionnaire (HAP; Barendse and Thissen, Reference Barendse and Thissen2006). This questionnaire consists of 62 items divided over ten different personality behaviors (e.g. uncertain or smug behavior). Individuals suffering from ASD are well known for their deficits in social interactions and social communication. In addition, also suffer from stereotype patterns of behavior, interests, and activities (American Psychiatric Association, 2000). With this in mind, high scores on the HAP scales “unsocial behavior,” “rigid behavior,” and “perfectionist behavior” are expected in the test group.

In this pilot study, at least 40 informants (e.g. spouse, children) of older adults (age ≥ 60 years) with ASD will be compared with 40 informants of patients without ASD and/or Axis II disorders. ASD patients will be recruited by examining the (electronic) medical records of four mental health institutions in The Netherlands. The control group will be recruited among informants of mental healthcare patients without ASD or Axis II disorders. The test and control groups are two independent samples with one measurement in time. Receiver Operating Characteristics (ROC) curves will be examined for the three scales separately, as well as for specific combinations.

As far as we know, this is the first study to examine a specific screening instrument for ASD in older adults. If the results are promising results, next steps are (1) to improve the criterion validity to make the diagnostic process uniform concerning ASD diagnosis, (2) to differentiate between ASD and Axis II disorders, and (3) to take characteristics of informants into account (e.g. quality of relationship between patient and informant).