In Reply We appreciate the interest of Sivakumaran in our article.1 We agree with the author that several confounders not included in our analyses may play important roles in defining the late-onset Alzheimer disease (LOAD) risk profile. This point was acknowledged as a limitation of the study in the Discussion section.
Nevertheless, 2 important aspects have to be taken into account. First, collecting a large number of covariates for which to adjust the models might address some statistical questions but not necessarily biological ones. In general, cardiovascular risk factors are considered associated with increased risk for dementia, but the underlying mechanism remains uncertain and there are inconsistencies. For example, a well-established risk factor as body mass index has shown conflicting results, leading to the “obesity paradox.” Being overweight and obese are associated with cardiovascular diseases, but a recent publication found that elderly and other types of patients can have a surprisingly lower mortality risk.2 In the context of LOAD, higher midlife body mass index can increased the risk for LOAD,3 although others have reported no findings regarding midlife and late-life weight and risk for conversion to mild cognitive impairment (but only the rate of weight loss being significantly associated).4