Sir,Article titled “Age of onset of dependence-Does it help our understanding of opioid dependence by generating categories or by acting as a useful dimension? A critical examination of the classic debate in psychiatry”[1] made an interesting reading. The debate about the categorical versus dimensional approach to diagnosis in psychiatry was termed as the classical debate by Kraemer et al.,[2] the debate that would help us to understand the nature and etiology of psychiatric disorders. Using the case of age of onset of opioid dependence (AOOD) authors have attempted to analyze this debate. As suggested by the authors, AOOD is indeed a very simple and clinician friendly variable. Previously De et al.[3] also used the same variable to classify the patients into two subtypes: Early onset (EO) with a mean AOOD 21 years and late onset (LO) with mean AOOD 27 years. Based on this finding, authors divided the study sample into two groups: EO <20 years and LO >22 years, omitting the year 21 just because the study by De et al.[3] showed that AOOD tended to peak around age 21 years. This omission brings in a methodological issue of creating “categories” out of a continuous variable of age. Recall bias is a common problem, and when two groups have almost no gap, that is, 20 years versus 22 years, one can easily fall into the “other” category. [Table 4] of the article[1] depicts this nicely and also points out the problem of recall bias, that is, age of onset of opioid use and AOOD are not very different, especially for EO group.Further such narrow (age wise) categories were bound to yield no difference. It is very understandable that age gap of 1 or 2 years would not affect the variables under study. Wider age gaps probably would have given us more insight into the issue at hand, as well as a very simple variable AOOD, would have settled the classical debate.