It appears that up to 80% of females with autistic spectrum disorder (ASD) have not been diagnosed by the age of 18. This translates to a prevalence of about 5-6%, and if true, has serious implications for female mental health. One way of finding the true value is to use Bayes' Theorem with a comorbid condition as a more easily recognizable flag. An obvious choice is anorexia nervosa (AN), but it transpires that the proportion of women with ASD who develop AN is unknown. This study uses published data in novel ways to provide two methods of estimating a range for this variable, and gives a median value of 8.3% for AN in ASD and, with four other methods, a median prevalence of 6% for female ASD. The clinical implications of the diagnosis and management of ASD and its comorbidities are discussed and, as an example, a solution is provided for the rate of ASD in symptomatic generalized joint hypermobility. It is probable that one in six women with a mental health condition is autistic.
Keywords: Bayes’ theorem; anorexia nervosa; autistic spectrum disorder; diagnosis; female; generalized joint hypermobility; hazard ratio; likelihood ratio; management; prevalence.