cognition who were followed for 4.60 (0.9-10.8) years and underwent annual/biannual neuropsychological testing and expert clinical consensus 4 to determine clinical conversion to mild cognitive impairment or dementia. Therefore, because NfL cut-points may differ based on choice of cognitive outcome measure or cognitive "event," it is uncertain how to compare findings. We note, moreover, that Buhmann et al 2 quote a median value of 18 pmol/mL for serum NfL in their study. Because NfL has a molecular weight of 70 kD, 18 pmol is equivalent to 1260 ng, which would mean that their median serum NfL value is 1,260,000 pg/mL, extremely high compared to values reported by many groups, including our own. 3,5,6 Despite differences, findings from both studies support the value of plasma/serum NfL in PD and underscore the importance of establishing absolute quantitation measures and more reliable cut-points for use across cohort studies. Given agerelated increases in NfL over time in both PD patients and healthy controls, 6,7 we agree that age-adjusted NfL percentiles may provide more consistent results in predictive models. Further studies are needed to explore the use of age-adjusted NfL cut-points in predicting cognitive decline in PD.Acknowledgments: We thank our patients and their families for their generosity in contributing to this research.