Table 2.

Dementia risk factors and dementia risk scales (adapted from Ranson et al., 2021²¹ and Frisoni et al., 2019¹⁹).

Risk factor	Relative riska	Assessment method	Dementia risk scalesb
			CAIDES1	ANU-ADRIS2,S3	BDSIS4
Genetic
APOE-ε4 heterozygous	1.9	Real-time protein chain reaction	●
APOE-ε4 homozygous	5.3
Potentially modifiable without measured brain pathology
Early life (age <45 years)
Less education (primary school only)	1.6	I—International Standard Classification of EducationS5	●	●	●
		P—Years of educationS5
Midlife (age 45–65 years)
Hearing loss	1.9	I—Pure tone audiometryS6 P—Whispered Voice Test,S7 speech-in-noise paradigms or self-report
Traumatic brain injury	1.8	I—Ohio State University traumatic brain injury identification methodS8 P—Medical history, informant- or self-report		●
Hypertension (>135–140/85–90)S9,c	1.6	I—Ambulatory devices, physician measurement P—Domestic devices, patient self-measurement	●
Alcohol consumption (>21 units per week)	1.2	I—Quantity-frequency measures with beverage-specific assessment of time frames and binge-drinking episodesS10 P—Self report		●
Obesity (body-mass index ≥30)	1.6	I—Waist circumferenceS11 and measurement of height and weight P—Body mass index based on self-report	●		●
Late life (age >65 years)
Smoking	1.6	Self-report of smoking status (pack years, i.e. number of daily packs multiplied by number of years smoking; or current smoking status, i.e. current versus former/never smoker)		●
Depression	1.9	I—Rating scales e.g. Patient Health Questionnaire (PHQ)S12 or the Hospital Depression and Anxiety ScaleS13 P—Self-report of feeling depressed or having history of diagnosed depression		●	●
Social isolation	1.6	I—Rating scales, e.g. the Lubben Social Network ScaleS14 or the Duke Social Support IndexS15 P—Self-report of social isolation		●
Physical inactivity	1.4	I – Accelerometers,S16 heart rate counters,S16 smart phone,S16 or smart watch appsS16 P – Self-reported measures/questionnaires	●	●
Diabetes	1.5	I—Fasting plasma glucose levels (≥7.0mmol/l) or HbA1c (≥6.5%), or oral glucose tolerance test to diagnose impaired glucose toleranceS17 P—Medical history, informant or self-report		●	●
Air pollution	1.1	Further research is needed to establish a practical and clinically relevant measureS18
Potentially modifiable, brain pathology
Amyloidosis	5.0 at 65 y 2.4 at 85 yd	I—Automated or semi-automated assay of Aβ in the CSF or amyloid PET using visual reading or centiloid quantificationS19-S23 P—Ultra sensitive assay of Aβ in plasmaS19
Amyloidosis and tauopathy	2.8–9.1	I—As above plus p-tau in the CSF or tau PETS19,S20,S24,S25 P – As above plus p-tau in plasmaS19
Neurodegeneration	1.6–3.1	I – Ultra sensitive assay of NfL in the CSF or plasmaS26 P—Qualitative or quantitative assessment of ventricular dilatation and medial temporal atrophy on MRIS27
Amyloidosis and neurodegeneration	21.4 at 65 y 4.9 at 85 yd	See above
Subcortical cerebrovascular disease	1.7–3.0	I—Volumetry of white matter changes and standardized scales for microbleeds on MRIS28 P—Visual rating of white matter changes, lacunes, and microbleeds on MRIS29

Potentially modifiable risk factors without measured brain pathology are those 12 identified in 2020 by The Lancet Commission.²² Assessment methods are summarized, categorized into ideal (I) and practical (P), and referenced. The assessment method may differ when the risk factor is measured in the context of dementia risk scales; in this case, please refer to the original scale reference. References cited in the table can be found in the Supplementary material.

^aThe relative risks of potentially modifiable risk factors without measured pathology are taken from (Livingston et al., 2020²²); these were computed taking into account communality (the variance in observed variables accounted for by common factors). The relative risks of APOE is taken from Rasmussen et al. (2015)S30 in the Danish population; for estimates in different ethnicities, please refer to Raichlen and Alexander (2014).S31 The relative risks of amyloidosis and tauopathy are taken from Yu et al. (2019)S32 and Ebenau et al. (2020)S33; of neurodegeneration (neurofilament light) from Kern et al. (2019) S34 and de Wolf et al. (2020)S35; of subcortical cerebrovascular disease from Inzitari et al. (2009),S36 Kitagawa et al. (2015),S37 Sigurdsson et al. (2017),S38 and Inzitari et al. (2007)S39. The relative risk of amyloidosis and amyloidosis and neurodegeneration is computed from the 10-year risk reported in Brookmeyer & Abdalla (2018).²³ The relative risks of genetic risk factors and potentially modifiable risk factors of brain pathology are generally adjusted for each other but communality with potentially modifiable risk factors without measured pathology is not taken into account.

^bCAIDE: Cardiovascular Risk Factors, Aging and Dementia Score. ANU-ADRI: Australian National University Alzheimer's Disease Risk Index. BDSI: Brief Dementia Screening Indicator. All scales include age, CAIDE also sex and cholesterol, ANU-ADRI also sex, cognitive stimulating activities, and fish intake, and BDSI also difficulty with instrumental activities of daily living and previous stroke.

^cThe threshold for the definition of hypertension differ according to the monitoring device and setting.S9.

^dRelative risks are reported for the 10-year risk of dementia in women, based on Brookmeyer & Abdalla (2018).²³ Relative risks for men are only marginally different.