Table 1.

Integrated results

Authors	MMAT score (%/4)	Study aim	Study design	Participants age, gender and ethnicity	Outcome measures (no. of items)	Loneliness and social isolation results
Quantitative studies
Alpass & Neville 24	75%	Investigate the relationships between loneliness, health and depression	Correlational Cross‐sectional survey. Non‐probability sampling	Men Urban dwelling, small city Aged ≥65 (65–89 years) 28% live alone (n = 217)	Self‐rated physical health (score 1–7), SSQ6 (6) R‐UCLALS (12) GDS (30)	Loneliness was more strongly related to depression than all other factors, including living alone and network size Greater loneliness was significantly associated with higher reported depression
Alpass & Neville 25	75%	Investigate the correlates of suicidal ideation in non‐clinical sample of older men	Correlational Cross‐sectional survey	As above	Self‐rated physical health, TRHS (53) THS (20) SSQ6 (6) GDS R‐UCLALS (12) TSIQ (30)	Loneliness was significantly associated with suicidal ideation; but number of, or satisfaction with, social supports were not Those who lived alone were lonelier, and were more depressed
La Grow, Alpass, & Stephens 26	75%	Test the assumption that those diagnosed with a visual impairment would have less social support and be more socially isolated (lonely) than those who had not	Sub‐sample of HWR Study_2006 participants 53% response rate to postal survey	NZ electoral roll representative sample Aged 55–70 (n = 5975 included) Visually impaired (n = 411) Sighted (n = 5564)	ELSI SF‐36 SPS	Those with visual impairments had statistically significantly less social support available, and felt more lonely and socially isolated than those without visual impairment
Stephens, Alpass, & Towers 27	75%	Test the prediction that economic living standards are related to social support and loneliness, and these factors are predicted to affect mental health	Cross‐sectional survey, sub‐sample of HWR Study_2006 (as above)	Representative population sub‐sample (as above) Aged 55–70 (n = 1720) Māori (n = 131) Non‐Māori (n = 1589)	SF‐36 ELSI‐SF SPS SWS (single item)	Women were more likely than men to report greater loneliness and lower living standards Māori were more likely than non‐Māori to report greater loneliness, poor mental health, & lower living standards. Perceived low social support, and more loneliness were associated with poorer mental health Loneliness and social isolation explained 15% of variance in mental health scores, with loneliness having the strongest effect on mental health
Stephens, Alpass, Towers, & Stevenson 28	75%	Use an ecological model of ageing to examine the effects of social networks on health. Network types: Wider community (WC), local integrated (LI), private restricted (PR), family dependent (FD) and local self‐contained (LS)	Postal questionnaire HWR Study_2006 (as above) Wave 1	Representative population sample (as above) Aged 55–70 (n = 6662) 36% aged 55–59 29% aged 60–64 25% aged 65–70	SF‐36 ELSI‐SF PANT SPS SWS (single item)	Non‐Māori perceived stronger total social support and felt less lonely than Māori Those with higher living standards were less loneliness and perceived more support Compared with younger ages, the older group were more likely to perceive less social support, and less likely to report loneliness Women reported more social support and were more likely to report loneliness Loneliness was moderately associated with total social support, yet contributed more strongly than social support to physical and mental health variance Loneliness and social provisions were positively related to WC and LI networks, and negatively to PR, FD and LS networks
La Grow, Towers, Yeung, Alpass, & Stephens 29	75%	Investigate the rate and degree of loneliness, and contribution to perceived quality of life for visually impaired older adults	Secondary analysis of survey data, wave 2 of NZLSA	Older adults aged ≥65 (n = 2683) Visually impaired (n = 315) Sighted (n = 2368)	Scored visual impairment (single item) dJGLS PQOL ELSI‐SF SF‐12v2	Over a half of the visually impaired, and over a third of the sighted group felt lonely The visually impaired were significantly lonelier, and almost twice as likely to be severely lonely, than the sighted group Increasing loneliness was directly, negatively, and significantly associated with economic well‐being, mental health, satisfaction with life and perceived quality of life (PQOL) Social loneliness made a unique, significant contribution to PQOL, while emotional loneliness did not
Robinson, MacDonald, Kerse, & Broadbent 30	75%	Explore how the psychosocial effects of companion robot, Paro, compared with a control group over 12 weeks	Randomised controlled trial Experimental group	Aged care residents Aged 55–100 (n = 40) 13 men 19 scored ≤6 on AMT	R‐UCLALS (V3) GDS QoL‐AD	Significant between‐group change in loneliness scores, after adjusting for baseline scores Experimental group (companion robot) mean loneliness score decreased (−5.38); control group (usual activities) mean loneliness score increased (2.29) over time No significant between‐group changes in depression or quality of life, after adjusting for baseline scores in both measures
La Grow, Neville, Alpass, & Rogers 31	50%	Identify the rate and degree of loneliness, and determine the impact on self‐reported mental and physical health	Cross‐sectional survey	Community‐dwelling older adults Aged ≥65 (65–98 years) 57% women 62% married or partnered (n = 332)	dJGLS SF‐36	Over half were lonely, including 44% moderately and 8% severely lonely No significant difference for sex, marital status or age Loneliness was significantly related to poorer physical and mental health
Qualitative
Park & Kim 32	75%	Explore the immigrant experiences of older Koreans and their intergenerational family relationships	Qualitative methodology, phenomenological based interviewing	Korean immigrants, older adults on arrival in NZ Aged 71–88 (n = 10) Key informants (n = 20), including co‐ethnic community members and health professionals	Semistructured interviews. Concept mapping to identify themes	Others’ discriminatory attitudes intensified feelings of not belonging and loneliness, like ‘living in an invisible prison’ English language was the most difficult life barrier, limiting social networks Men more likely to feel demoralised, with nothing to do inside or outside the home Intergenerational relationships became complex and difficult Less participation in host society due to language and cultural unfamiliarity, becoming isolated and lonely

Key for quality evaluation: MMAT, mixed methods appraisal tool. Key for studies: HWR Study, Health, Work and Retirement Study; NZLSA, New Zealand Longitudinal Study of Ageing. Key for measures: dJGLS, de Jong Gierveld Loneliness Scale (11 items); ELSI/ ELSI‐SF, New Zealand Economic Living Standards Indicator/Short Form; GDS, Geriatric Depression Scale (15 or 30 items); PANT, Practitioner Assessment of Network Type (social engagement measure); PQOL, Perceived Quality of Life; QoL‐AD, Quality of Life for Alzheimer's Disease; R‐UCLALS, University of California Los Angeles Loneliness Scale; SF‐12v2, Health Survey‐Short Form, Volume 2 (12 items); SF‐36 Health Survey (36 items; 8 subscales); SPS, Social Provisions Scale (single item of one's feelings of isolation); SSQ6, Social Support Questionnaire (6 items); SWS, NZ Social Wellbeing Survey 2004, Question 9 (single item); THS, The Hopelessness Scale (20 true/false); TRHS, The Revised Hassles Scale (53 items); TSIQ, The Suicidal Ideation Questionnaire (30 items).