We appreciate the considerate comments from Mahadeva et al. The purpose of our study was to explore the prevalence and comorbidity of dyspepsia, which entailed structured diagnostic assessments for dyspepsia and two mental disorders, namely generalised anxiety disorder and major depressive episodes. As such, in our effort to ensure brevity in a telephone survey, necessary for safeguarding data quality and response rates, there was inevitable omission of otherwise interesting areas deserving separate devoted studies, such as health-related quality of life (HRQOL).
We acknowledge that the cross-sectional design of our study would delimit interpretation of chronological variables such as order of onset, and believe that a longitudinal design would be most suitable to explore the role of HRQOL and other factors related to onset of these related disorders.
The Rome III criteria for functional dyspepsia were restrictive by design to reduce heterogeneity in diagnostic interpretations adopted by clinicians and researchers that former iterations would allow. Our conservative prevalence estimate of 8.0% is consistent with this restrictiveness. We adopted the Rome III criteria as the current globally recognised standard of diagnosis to facilitate comparison, and indeed the literature is still lacking in dyspepsia prevalence estimates based on stringently adopted Rome III criteria.
We did find Rome III dyspepsia to be significantly associated with frequent medical consultations. We do agree with Mahadeva et al. on the possibility that the restrictiveness of Rome III dyspepsia criteria may account for variances in other outcomes, such as the lack of association found with socioeconomic variables, and believe that the matter should be empirically evaluated in further community surveys.
Inclusion of sub-threshold disorders in epidemiological studies may allow a large sample to be captured that facilitates statistical analysis of otherwise insignificant associations. In psychiatric epidemiology, sub-threshold disorders have been usefully included in community and clinic-sample studies in different countries to explore the validity and boundaries of diagnostic entities, such as anxiety and affective disorders.[4, 5]
Similarly, Rome III should benefit from further empirical validation aided by explicitly defined inclusion of sub-threshold conditions to ascertain its internal and external validities – namely socioeconomic associations. To this end, adopting for Rome III the approach previously taken by Mahadeva et al. in examining the ethnic and HRQOL associations[6, 7] are arguably necessary additions to the literature.