You can spend every day surrounded by other people and still feel lonely. You can also spend long periods alone without finding the experience distressing.
That difference may help explain why loneliness affects health differently from social isolation. In a new study, my colleagues and I found evidence that loneliness may contribute to poorer mental health and wellbeing, and to some measures of poorer general health. Social isolation appeared to have a narrower influence, with the clearest evidence involving poorer wellbeing.
Around one in six people worldwide experiences loneliness, according to the World Health Organization. Loneliness is the distress someone feels when their relationships do not provide the closeness or connection they want. Social isolation describes the number and frequency of a person’s social contacts. The experiences can overlap, although a person may experience either one without the other.
Previous research has linked loneliness and social isolation with poorer mental and physical health, including depression and cardiovascular disease. An overview of systematic reviews also found associations with poorer general health.
These patterns cannot establish cause and effect. Loneliness might contribute to ill health, while illness can restrict someone’s activities and relationships. Poverty, disability and childhood circumstances may also influence both health and social connection.
Our study used several statistical approaches with different strengths and sources of bias. Taken together, the results suggest that loneliness may contribute to depression and increase the risk of self-harm or a suicide attempt. It may also reduce life satisfaction and other aspects of wellbeing.
We found evidence that the relationship can operate in the opposite direction. Depression may increase loneliness, while poorer wellbeing may increase both loneliness and social isolation. Evidence that depression contributes to social isolation was less consistent. Previous longitudinal research has also found a two-way relationship between loneliness and common mental health problems.
Loneliness was linked to poorer general health, including multimorbidity, which means living with two or more long-term health conditions. We found no consistent evidence that loneliness or social isolation affected the specific physical-health outcomes examined. These were coronary artery disease, heart failure, stroke, systolic blood pressure, which is the upper number in a blood-pressure reading, and type 2 diabetes.
This does not rule out an effect. Some estimates were too imprecise to support a firm conclusion, and previous longitudinal studies have reported links between loneliness, isolation and physical illness.
Combining different kinds of evidence
In the study, we used an approach known as triangulation. This involves studying the same question using methods affected by different potential sources of bias. When several approaches produce similar results, researchers can be more confident that the conclusion is not driven entirely by the limitations of one method.
First, we analysed information from UK Biobank, which has collected health, lifestyle and genetic information from around 500,000 people in the UK.
We then compared siblings from the same families. Because siblings often share aspects of their childhood environment and genetic background, these comparisons can help account for family influences that might otherwise distort the results.
We also used Mendelian randomisation. This method uses genetic variants statistically associated with a tendency towards loneliness or social isolation as proxies to investigate whether those experiences may affect health.
Mendelian randomisation can reduce some sources of bias that affect conventional observational research. These include reverse causation, where an outcome may influence the supposed cause.
However, the method relies on assumptions that cannot always be fully tested. This is especially important for complex social experiences, which are only indirectly related to the genetic variants used as proxies. We therefore considered these findings alongside evidence from the other methods.
What the study cannot tell us
The research has several limitations. In the UK Biobank analyses, loneliness and social isolation were measured at one point during middle or later adulthood. The findings may therefore apply differently to children, younger adults and people in very old age.
Loneliness was measured using one question asking whether participants often felt lonely. Social isolation was assessed through household size and the frequency of visits from friends or family.
This measure did not include contact through work, group activities or other parts of social life. The findings therefore relate to these particular measures and may not capture every form of loneliness or isolation.
Measuring each experience once also meant the study could not distinguish a short period of loneliness from loneliness lasting several years, or examine how experiences earlier in life might affect later health.
UK Biobank participants tend to be healthier and less economically disadvantaged than the wider population, a pattern described as healthy volunteer selection bias. The genetic analyses were restricted to people of European ancestry, limiting how confidently the findings can be applied to other populations.
The study did not test treatments or services, so it cannot identify which interventions work best. However, it strengthens the evidence that loneliness may contribute to poorer mental health and wellbeing, and that loneliness and social isolation should be examined separately.
This distinction may affect how support is designed. Increasing opportunities for social contact may reduce isolation. Addressing loneliness may require support focused on the quality of relationships and whether people experience them as close, trusting and meaningful.
This article was originally published on The Conversation. Read the original article.
