SABIA ET AL. RESPOND

Abstract

We are grateful for Bozorgmehri’s comments on our article.1 In his letter, Bozorgmehri notes that the 33% reduced mortality risk we found in participants practicing at least 1 hour of moderate activity per week was in accordance with the current literature. He proposes that additional confounding factors ought to have been taken into account in our analyses. In the analyses reported in the article, the models were adjusted for measures that could potentially influence both physical activity and the risk of mortality (see list of covariates included in Table 1, model 1). TABLE 1— Association Between Self-Reported Duration of Moderate Physical Activity and Mortality: Whitehall II Study, 1997–2009 As suggested by Bozorgmehri, we undertook additional analyses that further included the following covariates: systolic and diastolic blood pressure, total blood cholesterol, body mass index, prevalence of cancer, and employment status using unemployment (very uncommon in our cohort [0.7%]) and retirement data. The results from this analysis were in accordance with results reported in our article (Table 1, model 2). Then, in a separate model, we additionally adjusted for lung function measured using forced expiratory volume (Table 1, model 3). This measure was assessed at the 2002 to 2004 screening and only among healthy participants,2,3 thus the start of the follow-up for mortality in these analyses was 2002 to 2004, reducing the sample size to 4708 (number of deaths = 104) participants. In this reduced sample, the association between physical activity and mortality was similar and did not change after adjustment for lung function (Table 1, model 3). A further point made by Bozorgmehri is that the “time frame of analysis did not seem long enough to confidently address questions of mortality in this cohort of mostly healthy, richer men.” We disagree with this statement because our sample was composed of 30% women and a wide range of socioeconomic statuses, with the annual full-time salaries in 1995 ranging from £4995 to £150 000. We agree with Bozorgmehri that a longer follow-up period would lead to more deaths, increasing the precision of our results. To better estimate the effect of physical activity over time, we plan to examine how change in physical activity patterns is associated with mortality in the future when a longer follow-up for mortality is available. Nevertheless, we think that our study, which was based on a 10-year follow-up period with repeated measures on both intensity and type of physical activity, contributes to the research on the association between physical activity and mortality.4–6