Consistent with an Australian analysis of oral and injectable TRT prescription rates, which showed that introducing a Tubeimoside-I mandatory phone call to authorize a TRT prescription resulted in only a partial and temporary curtailment. Finally, our study shows that TRT, which carries potential safety concerns, is being prescribed to elderly men with a variety of comorbidities who differ substantially from younger, healthier men generally included in clinical trials of these products. Of particular concern is the large proportion of TRT users being treated for cardiovascular risk factors such as hypertension, diabetes and dyslipidemia considering the new evidence of the potential association between testosterone replacement and myocardial infarction and stroke. Several factors may have contributed to the increasing use of TRT observed in this study and the lack of sustained effect of the drug reimbursement policy. First, TRT has been aggressively marketed to both clinicians and patients for non-approved indications such as ‘andropause’ and male sexual dysfunction. In fact, in the United States, where direct to consumer advertising is legal, the rates of TRT use are three times higher than those observed in our study. This is consistent with our observation that a low proportion of men had a documented diagnosis of testicular dysfunction. In addition, topical TRT is more convenient and produces more stable drug levels than oral or injectable formulations, properties that may increase patient and clinician acceptance of testosterone. In addition, a TRT manufacturer recently published a study urging clinicians to improve adherence to topical testosterone, promoting long-term treatment for unofficial conditions such as age-related androgen decline. Indeed, in our study we found that rates of topical TRT use have increased consistently since its introduction in 2005 as compared to rates of oral and injectable TRT use, which have remained relatively stable since 2006. Finally, publications related to testosterone use are given considerable exposure in the lay and medical press, which further drives medicalization of aging, drug campaigns and development of new products. Our study has some limitations that merit emphasis. First, our data do not include patients Acetylcorynoline younger than 65 years of age who may also use TRT and so may not be generalizable to the entire population of men in Ontario. A recent study of IMS data by Handelsman et al. found high volumes of TRT sales in Canada that exceed the prescription rates reported in our study. It is likely that this discrepancy is driven by the different study populations in these two studies. Handelsman et al. included all prescriptions sales for TRT in Canada, which capture prescriptions dispensed to younger men, as well as those paid for through private drug plans and cash payments. However, the authors not distinguish between testosterone sold to be used in Canada and the large proportions of sales from internet pharmacies intended for export. As a result, it is unsurprising that we observed lower rates of use in our study, as prescribing was restricted to older men in a public drug program where some prescribing restrictions are place. Second, although the government plan notionally reimburses TRT only for men who meet pre-specified diagnostic criteria, no objective serum testosterone level or formal application process is required and physicians are not subject to review or auditing processes. Due to this, and our lack of access to laboratory data that precluded verification of patients’ serum.