Circulating serum metabolites as predictors of dementia: a machine learning approach in a 21-year follow-up of the Whitehall II cohort study

Table 1 Sample characteristics in 1997–1999 as a function of dementia status at the end of follow-up (31st March 2019)

Data are n (%), unless otherwise specified
^*Data on dementia subtype was as follows: Alzheimer’s disease (N=137), vascular dementia (N=47), Parkinson’s dementia (N=17), mixed Alzheimer’s and vascular dementia (N=21), mixed vascular and Parkinson’s dementia (N=1), mixed Alzheimer’s and Parkinson’s dementia (N=5), other/missing subtype (N=101)
^†p-value for difference in χ² test (categorical data) or Student’s t test (continuous data)

ISSN: 1741-7015