Rh Sensitization and Induced Abortion—Reply

Abstract

In Reply We thank Dr Hecht for his comment on our recent research article examining sedentary behavior and incident dementia. 1 We agree that the possibility of ascertainment bias is an important issue that warrants careful consideration. As we discussed in the Limitations section of our article, 1 the use of hospital records and death registry data for dementia diagnoses may provide less accurate diagnoses or may underestimate cases in this cohort, leading to the potential for ascertainment bias related to links between physical activity and overall health. To help address this limitation, we included presence of chronic disease (cardiovascular disease, cancer, or diabetes) as well as self-reported general health in our fully adjusted models to account for health conditions other than dementia in this sample. 1 To further determine whether the likelihood of hospitalization or death impacts our results, we have now included the number of hospitalizations after the start of follow-up, as well as death events, as additional covariates in our fully adjusted model. Inclusion of these covariates did not substantially affect our results showing that sedentary behavior has a nonlinear association with incident dementia; however, the hazard ratios for dementia were somewhat attenuated. In a model including these new covariates, hazard ratios for dementia relative to median sedentary behavior (9.27 hours per day) were 1.06 (95% CI, 1.02-1.10; P < .01) for 10 hours per day, 1.46 (95% CI, 1.22-1.76; P < .001) for 12 hours per day, and 2.51 (95% CI, 1.61-3.92; P < .001) for 15 hours per day. In a second analysis, we restricted our sample to only individuals with a hospital record or who died during follow-up. Again, results were similar to our original findings: hazard ratios for dementia relative to median sedentary behavior (9.27 hours per day) were 1.06 (95% CI, 1.01-1.11; P = .03) for 10 hours per day, 1.40 (95% CI, 1.08-1.80; P = .01) for 12 hours per day, and 2.21 (95% CI, 1.20-4.08; P = .01) for 15 hours per day. While these analyses cannot fully rule out the possibility of ascertainment bias, we believe our original results controlling for health status, combined with these new analyses controlling for hospitalizations and death, further support our overall finding that engaging in large amounts of sedentary behavior is associated with greater risk of dementia. Future work that can include other sources of information for dementia diagnoses will allow us to further test the association between sedentary behavior and all-cause dementia.