Author's reply

As Mushtaq & Minn-Din correctly point out, Bijl et al ¹ did not find associations between prevalence rates of any psychiatric disorder (including anxiety disorders) in children and anxiety-related symptoms in parents. However, one must be aware of several methodological differences to our investigation: the results cited by Mushtaq & Minn-Din are based on 12-month prevalence rates and multivariate logistic regression analysis additionally controlling for childhood adversities and socio-demographic characteristics. We would like to clarify that the results of the Bijl et al paper are much more comparable with our study and that the results we are actually referring to are those based on life-time prevalence rates of psychiatric disorders in children without controlling for childhood adversities and reported separately for the various offspring disorders. Here, Bijl et al clearly report associations between anxiety in parents and children.

In addition, it is true that the adult children in the Bijl et al study were considerably older (18–65 years) than the offspring in our study (17–21 years at follow-up). We would like to add that there are other substantial ways in which the studies differ; for example, our use of assessment via direct interviews v. family-history information. ¹ Nevertheless, we do not see why our claim that we confirm and extend the Bijl et al study should be problematic, especially when taking into account the low median for age at onset of anxiety disorders. ²

We would also like to point out that both studies were community-based so that the use of the term ‘patients’ by Mushtaq & Minn-Din is slightly misleading.