The study by Forman et al1 is a welcome addition to the literature exploring the association between hypertension and nonnarcotic analgesic use. However, I am concerned about bias as an explanation for some of the association measured by the authors.
In the “Comment” section, Forman et al1 explore the possibility that men who visited physicians more often may have been more likely to have incident hypertension detected. When the authors limited their analysis to participants who had ever visited their physicians during follow-up, the risk of hypertension associated with acetaminophen use fell to a statistically nonsignificant level. I suspect that this subset analysis may not fully counter the potential bias of physician visits because it does not adjust for frequent vs occasional visits. Patients in pain may be expected to use more analgesics and visit their physician more often (and thus have hypertension incidentally detected). Future similar studies of this issue should measure frequency of physician visits and adjust accordingly.
Montgomery BD. Analgesic Use and Risk of Hypertension: Concern About Bias. Arch Intern Med. 2007;167(21):2368-2369. doi:10.1001/archinte.167.21.2368-b