Meta-analysis is upheld as a higher order analysis but it is not without fault as is evident in the conundrums raised by Hegeman et al’s article. ^{Reference Hegeman, Kok, van der Mast and Giltay1} Although the methodology was rigorously applied, even the final list of 11 studies included in the meta-analysis were not entirely comparable, particularly in the area of medical comorbidity, a point highlighted by the authors. Our own clinical experience and findings (possibly shared by other psychiatrists) have been that the elderly do have significant medical comorbidity that affect clinical presentations. In one study, all but one elderly patient had a medical condition and 60% had two medical conditions. ^{Reference Ko, Kua and Chow2} The meta-analysis also does not take into account the significant role socioeconomic and cultural factors have in depressive symptom development and progression in the elderly. Socioeconomic issues play an important part given the changes in occupation, lifestyle and other roles in the elderly. This paper also serves to highlight the recently resurfaced distinction between clinimetrics and psychometrics. ^{Reference Fava, Rafanelli and Tomba3} The relevance and applicability of psychometrically driven research is sometimes difficult to translate for the psychiatrist in clinical settings. Clinicians cannot rely entirely on rating instruments to arrive at a diagnosis and to devise management care plans. Clinicians will sieve through the history and presentations in detail and make global judgements on information presented, a process that goes beyond the one-dimensional nature of checklists and rating scales. Fava et al alluded to the ‘sophisticated thinking that underlies clinical decisions’ and that is a point that deserves consideration even as we review journal articles on research which is largely psychometrically driven to glean benefit for our clinical practice.