Impact of social isolation on behavioral health in elderly: Systematic review

Table 1 Summary of 16 articles on older adults

Ref.	Outcome	Study design	Age in years (n)	Health status	Cultural/ethnic characteristics	Assessment method	Relevant results (significance)
Costa et al[12]	Sleep disturbance	Cross-sectional	≥ 65 (497)	Random selection; Unspecified number of subjects had chronic diseases	Brazilian	Questionnaire, NHP, MLAQ, ISEL	Elderly with sleep problem had lower score on ISEL (P < 0.05)
Jensen et al[13]	Insomnia	Longitudinal	80 (212)	Random selection; Unspecified number of subjects had chronic diseases	Swedish	Questionnaire (graded sociological data)	Severity of insomnia associated with having felt lonely in the past (P < 0.05)
							Severity of insomnia associated with believing that future would bring loneliness (P < 0.01)
Troxel et al[14]	Insomnia	Cross-sectional	≥ 60 (119)	Study Group: Presence of insomnia; Unspecified stable medical and psych condition Control Group: Absence of insomnia	Pittsburg, PA, United States	Questionnaire, Pittsburgh sleep diary, PSQI, actigraphy	Wakefulness after sleep; onset is lower in people with higher social support (P < 0.01)
							In group with insomnia, shorter sleep latency in higher social interaction group (P < 0.01)
Eshkoor et al[15]	Sleep disturbance	Cross-sectional	≥ 60 (1210)	Dementia	Malaysian	SNSL, Mini-mental examination	Social support, marital status, having partner significantly affect sleep disturbance (P < 0.05)
Yao et al[16]	Sleep disturbance	Cross-sectional	65-75 (187)	Random selection; Three-fourths of subjects had chronic illness	Taiwanese	Questionnaire, PSQI (Chinese version)	Good relationship with friends and family is negatively correlated with poor sleep quality (P < 0.001)
McHugh et al[17]	Sleep disturbance	Longitudinal observational	≥ 60 (447)	Random selection; Unspecified number of subjects had chronic diseases	Irish	DJGLS, PSQI	Emotional loneliness (subjective social isolation) rather than social loneliness (objective social isolation) is a stronger predictor of poor sleep quality (P < 0.001)
							Emotional loneliness increases stress (P < 0.001)
							Stress affects sleep quality (P < 0.0001)
Gerst-Emerson et al[19]	Depressive symptoms	Cross-sectional	80-102 (3050)	Random selection; Unspecified number of subjects had chronic diseases	Mexican American in 5 states in the United States (TX, CA, AZ, CO, and NM)	Three-item loneliness scale, 20-item CES-D	Scores on depressive symptoms are positively associated with loneliness (P < 0.001)
Aylaz et al[20]	Depressive symptoms	Cross-sectional	≥ 60 (913)	Random selection; Unspecified number of subjects had chronic diseases	Turkish	GDS, ULS	ULS score and GDS score correlation (r) is 0.608 (P < 0.001)
Theeke et al[21]	Depressive symptoms	Cross-sectional	≥ 65 (60)	All subjects had chronic illnesses	Appalachians	ULS, CES-D, GDS	ULS score and depression has correlation coefficients value (r) of 0.388 (P < 0.01)
Adams et al[22]	Depressive symptoms	Cross-sectional	60-98 (234)	Random selection; Subjects had 1.7 chronic diseases on average	Northeast United States (Retirement community affiliated with Methodist Church)	ULS, GDS	ULS score and GDS score correlation (r) is 0.458 (P < 0.005)
Alpass et al[23]	Depressive symptoms	Cross-sectional	≥ 65 (217)	Random selection; 61% of subjects had chronic illness or disability	New Zealand	ULS, GDS	ULS score and GDS score correlation (r) is 0.625 (P < 0.01)
Bekhet et al[24]	Depressive symptoms	Cross-sectional	65-84 (314)	Random selection; Unspecified number of subjects had chronic diseases	Cleveland, OH, United States (Retirement community)	Questionnaire, CES-D	Elderly who reported feeling lonely had higher depressive symptom (P < 0.001)
Park et al[25]	Depressive symptoms	Cross-sectional	≥ 60 (674)	Random selection; Unspecified number of subjects had chronic diseases	Korean Americans in Tampa and Orlando, FL, United States	SNSL, 20-item ULS, GDS-short form	Loneliness mediates the relationship of social engagement related variables with depressive symptom (P < 0.05)
							Social engagement related variables: not living alone, social network, activity participation
							Exception: the relationship of social network and loneliness in men
Wan Mohd Azam et al[27]	Depressive symptoms	Cross-sectional	≥ 60 (161)	Random selection; Unspecified number of subjects had chronic diseases	Malaysian (Rural/agricultural settlement)	DJGLS, GDS, MOSSS	When perceived social support decreases, feeling of loneliness increases (P < 0.01)
							Social support mediates between loneliness and depression (P < 0.05)
							Social loneliness and depression: Pearson correlation (r) is -0.189 (P < 0.05)
							Emotional loneliness and depression: Pearson correlation (r) is 0.403 (P < 0.01)
Jason et al[28]	Fatigue	Interventional	57.6 on average (30)	Individuals who were diagnosed with chronic fatigue syndrome in the past	Chicago, IL, United States	Buddies (intervention), questionnaire, MOSSF-36, FSS, PSS	After 4 mo of intervention (supportive interaction), severity of fatigue of experimental group decreased compared to control group based on FSS (P < 0.05)
Riemsma et al[29]	Fatigue	Cross-sectional	51-75 (229)	All subjects were diagnosed with rheumatoid arthritis	Dutch	Double-anchored VAS, SSL12-I	Problematic social support and fatigue correlation (r) is 0.28 (P < 0.001)

CES-D: Center for Epidemiological Studies-Depression Scale; DJGLS: De Jong Gierveld Loneliness Scale; FSS: Fatigue Severity Scale; GDS: Geriatric Depression Scale; ISEL: Interpersonal Support Evaluation List; MLAQ: Minnesota Leisure Activity Questionnaire; MOSSS: Medical Outcomes Survey Social Support; MOSSF-36: Medical Outcomes Survey Short Form – 36; NHP: Nottingham Health Profile; PSQI: Pittsburgh Sleep Quality Index; PSS: Perceived Stress Scale; SAST: Short Anxiety Stress Test; SNSL: Social Network Scale of Lubben; SSL12-I: Social Support List Interaction; ULS: University of California-Los Angeles Loneliness Scale; VAS: Visual Analogue Scale.