Ageing, loneliness, and the geographic distribution of New Zealand's interRAI-HC cohort.

Loneliness is a significant negative predictor of ageing well and a contra-indicator for resilience against requiring long-term residential care. Health geographers can contribute to the loneliness and ageing literature through examining how exposures in the physical and social landscape can affect positive and negative health outcomes. As well as improving individual experiences of ageing, spatial analysis may help to contribute to better understandings of loneliness and help reduce the $1.7 billion per annum New Zealand currently spends on publicly-funded aged residential care. Using New Zealand Home Care International Residential Assessment Instrument data from 2012 to 2016, the spatial distribution of the interRAI-HC cohort was examined. Urban and rural distribution, socioeconomic status, and ethnicity within the interRAI-HC cohort was compared against total population and population aged 65 plus. Relative to the socioeconomic status of the 65 plus cohort, those being interRAI-HC assessed were more likely to live in socially deprived areas. Socioeconomic deprivation also positively correlated with loneliness (OR = 1.002). Carer stress was negatively correlated with socioeconomic status (OR = 0.99). Those in rural areas were predicted to be less lonely than urban dwellers (OR = 0.98), and this observation remained similar and significant when socioeconomic status, the experience of negative social interactions or carer stress, and whether they lived alone were included. Living in rural areas had a protective effect against loneliness for all ethnic groups apart from Pasifika. ‘Hot’ and ‘cold’ clusters of loneliness were identified, with the distribution of interRAI-HC assessments in hot clusters less likely to be rural areas (OR = 0.71). Our findings did not diverge greatly from prior research on older people and loneliness in rural areas. Observations of regional differences regarding rurality and socioeconomic status did not show large differences, and this research will benefit in future from analysis at finer geographic scale.