Exploring social isolation: insights from five British longitudinal studies

Despite an increase in policy interest, there is little evidence documenting the associations between social isolation, loneliness and subjective wellbeing across our lives and between generations. 

Led by Professor Praveetha Patalay, our Social isolation and loneliness across the lifecourse project aims to address this evidence gap, while also generating a range of comparable measures of social isolation for future research. This work forms part of a secondary data analysis initiative (SDAI) to address evidence gaps in our understanding.

Here we explore key insights from the first report, which used data from five British longitudinal cohort studies, covering people born between 1946 and 2001, to explore social isolation trends over time.

Longitudinal, population-based studies have the unique advantage of providing insights based on the same individuals across time, allowing us to observe experiences within and between generations.

The research used longitudinal data from five successive generations (born 1946, 1958, 1970, 1989-90 and 2000-01) to explore social isolation and connectedness:

1. within different contexts (household; partnership, family and friends outside the household; education and employment networks; community engagement)
2. over the course of people’s lives
3. between generations.

By investigating social isolation, we shift the focus from individual to structural factors that include broader economic and social conditions. This is a necessary first step in identifying concrete policy, community and societal actions we can take to reduce isolation and loneliness.

What is social isolation?

Social isolation is an objective condition that can be quantifiably assessed by the number and frequency of social connections and interactions, for example how often someone meets up with their friends, or how many people they live with. 

This is distinct from loneliness, which is a negative subjective feeling defined by the individual’s perception that the quantity or quality of their social relationships is inadequate. 

It is possible to feel lonely despite being socially connected, and to feel content in solitude. 

Key findings

The research highlights that social isolation is a multi-dimensional concept, with a variety of dynamic experiences across contexts, generations and life stages.

Some people’s social contact is focused on interactions with family and/or friends, while for others there may be more of a mix. For example, engaging with colleagues at work, taking part in community activities or interacting with those they live with.

Notable insights include:

• Younger generations are less likely to belong to a club or engage with religious activity, but are more likely to volunteer.
• Younger generations are more likely to live alone than older generations, particularly in their 20s.
• For the oldest three generations (b. 1946, 1958, 1970), most people experienced isolation in one or more contexts during midlife.
• The oldest generation (b. 1946) show an increase in the likelihood of engaging in regular social activities around age 65. This corresponds with retirement age and being less likely to be in education or employment.
• Across generations, the probability of females being neither in education nor employment aged 16-31 has decreased, indicating changing patterns of family formation and labour market participation.

There is a need to focus on a range of social isolation indicators across contexts and generations to:

1. understand how people compensate for specific types of isolation
2. understand how broader economic and social factors can affect the timing of life transition points – such as retirement, and trigger social isolation.

For a more detailed synthesis of the report’s findings – alongside a discussion on the implications for researchers, policymakers and practitioners and recommendations for how this research can be put into action – see our briefing.

Accelerating access to datasets

Questions regarding social isolation were not asked in the same way for each longitudinal study, making it difficult to compare generations. 

A core part of the research was deriving comparable measures of social isolations for each context. The groundwork has now been laid, and will enable future researchers carrying out life course work to easily use these variables for future cross-generation comparisons in UK longitudinal data.