COMMENT & RESPONSEIn Reply We thank Cheung et al for their thoughtful letter and the opportunity to discuss some of the limitations of our study. 1 Cheung and colleagues suggest that determining optimism based on initial levels of optimism may be biased as levels of optimism may change over time. Prior research indicates that optimism is largely stable over time, and initial levels can have long-term health implications even when it subsequently changes. Other work shows that optimism levels often return to baseline even after major life events like cancer diagnosis and treatment. 2,3 Although optimism can be modified through interventions, this requires intentional effort and targeted interventions. 4 Cheung and colleagues question the psychometric quality of the optimism measure, noting that the Life Orientation Test-Revised (LOT-R) was published in 1994, whereas the measure was administered between 1993 and 1998. We reviewed the initial items administered and confirmed that although the LOT items were administered in the first version, the form also included items in the LOT-R in that assessment. We used only the items from the LOT-R to construct our measure.Cheung and colleagues recommend a sensitivity analysis removing those who were assigned the minimum and maximum values if outcome observations were missing due to safety or health-related concerns or because tasks were attempted, but could not be completed, as this may bias results. We had conducted this analysis and found it did not change results for grip strength and chair stands, although models did not converge for walking speed. As results were similar for 2 outcomes, and to remain consistent with other studies, 5 we maintained this imputation method and did not include the sensitivity analyses due to space constraints. Cheung and colleagues mention that Rosso et al 5 removed only the top 2.5% of implausible assessments. However, their study states, "Participants who…assigned a 0 for grip strength and chair stands and the lowest observed value (0.1 m/s) for walk speed." 5(p.211) Finally, Cheung and colleagues mention that the average body mass index (BMI) of participants met the obesity criteria. In fact, this is not true, as the mean BMI of enrolled participants was 29 (calculated as weight in kilograms divided by height in meters squared), which meets criteria for overweight but not obesity. 6 The authors were also concerned about different muscle mass compositions and sarcopenic obesity and how these could confound the results. We agree that these are important issues to consider. However, as indicated in the additional analysis section, we conducted sensitivity analyses to assess interaction by age and also stratified analyses by age group. We found no significant interaction by age for any