Table 1.

Published Studies on Loneliness and Dementia

Study	Cohort	Country	Sample size	Loneliness measure	Dementia classification	Follow-up interval	Time- to-event analysis	Statistic
He et al., 2000	--	China	1,203	Unknown	Diagnosis	10 years	Yes	1.63 (0.93–2.86)
Wilson et al., 2007	Rush	United States	823	5 items	Diagnosis	4 years	Yes	1.51 (1.06–2.14)*,a
Holwerda et al., 2014	AMSTEL	The Netherlands	2,173	Single item	Structured interview	3 years	No	1.96 (1.33–2.89)*a
Rafnsson et al., 2017	ELSA	United Kingdom	6,677	3-item UCLA	Self, proxy, performance	6 years	Yes	1.44 (1.11–1.88)*a
Zhou et al., 2017	CLHLS	China	7,867	Single item	Self, proxy	3 years	No	1.22 (1.10–1.35)*,a
Poey et al., 2017	ADAMS	United States	663	Single item	Diagnosis	5 years	No	1.06 (0.44)
Rawtaer et al., 2017	SLAS	Singapore	1,601	Single item	Diagnosis (included MCI)	8 years	Yes	1.19 (0.78–1.81)

Note: The Table includes prospective cohort studies that measured loneliness among cognitively health participants at baseline who were subsequently tested or evaluated for dementia at a future follow-up assessment. PubMed and Web of Science databases were searched up to February 2018 with inclusion criteria: (1) longitudinal study of cognitively healthy participants at baseline, (2) loneliness assessed at baseline, (3) dementia assessed at follow-up, (4) report of the association (e.g., hazard ratio) associated with loneliness, and (5) written in English. All loneliness scales were included and cognitive status could be determined by a clinical diagnosis, self/proxy report, and/or a performance-based measure of cognition.

ADAMS = Aging, Demographics, and Memory Study; AMSTEL = Amsterdam Study of the Elderly; CLHLS = Chinese Longitudinal Healthy Longevity Survey; ELSA = English Longitudinal Study of Aging; MCI = Mild cognitive impairment; Rush = Rush Memory and Aging Project; SLAS = Singapore Longitudinal Ageing Study.

^aRemains significant when social isolation is included as a covariate.

*Significant association.