Author Affiliations: University Hospital, Basel, Switzerland (Dr Conen) (email@example.com); and Division of Preventive Medicine, Brigham and Women's Hospital, Boston, Massachusetts (Drs Glynn and Albert).
In Reply: We agree with Dr Parakh and colleagues that the accurate presentation of study characteristics in populations using time-varying models over long periods of follow-up is challenging. We acknowledge that our presentation of baseline characteristics in Table 1 did not capture the changing characteristics of the population over time. As an additional method, Parakh et al advocate for the presentation and analysis of characteristics during periods of exposure. A concern about this approach is that changes in characteristics triggered by the onset of the exposure (eg, smoking cessation in response to an episode of AF) would be related to antecedent exposures, instead of our approach that measured all characteristics before exposure onset.
Conen D, Glynn RJ, Albert CM. Mortality in Women With New-Onset Atrial Fibrillation—Reply. JAMA. 2011;306(11):1200–1201. doi:10.1001/jama.2011.1323