Photo © Simon Rawles Table 1: Predictors of social isolation, loneliness or both (Grenade and Boldy 2007) Both Social isolation Loneliness Household composition “Lower” social class Self-assessed health Morale / self-esteem Number of years widowed Desire for new friends Support network type Excluding age and gender, evidence regarding Kalyan Masih 1990). Similarly, the extent to which the impact of other socio-demographic factors people from ethnic minority backgrounds feel on loneliness tends to vary. Steed et al. (2010) isolated or lonely will depend on several factors suggest that this variability may be related other than their ethnicity, including the extent to the type of data (i.e. cross-sectional data to which they are assimilated and their sense of versus longitudinal) or the measures used (e.g. “belonging” (Sharma 2012). Hence, we need to direct questions versus a scale where the word be wary about blanket statements which ignore “loneliness” is not used), and confounding with individuals’ histories and experiences. other variables. Another group of risk factors relate to health, Evidence for an association between level of both physical (e.g. poor self-assessed physical education, geographical location (e.g. rural health status, chronic illness) and mental health versus urban), material circumstances (e.g. (e.g. reported depression). Although deteriorating limited income), ethnicity and loneliness/isolation physical health (or perceived poor health) is one is inconclusive. For example, there is some of the most consistently identified factors, the evidence to suggest that adolescents in rural direction of causation is still not well understood areas are more likely to experience loneliness (Grenade and Boldy 2008: 471). That is to say, it than their peers in urban areas, but this may is unclear whether poor physical health leads to be due to their socio-economic circumstances feelings of loneliness or vice versa. rather than where they live (Woodward and Isolation and loneliness 10