#197 Loneliness: Association or Causation?

Loneliness

In previous Newsletters I have discussed loneliness. In particular its impact on mental and physical health. Researchers are still exploring these relationships.

A Five Minute Read.

The number of “households” in Europe grew by 7% in the ten years to 2023. The number of households with only a single adult grew by 21%. Of the 200m households in Europe today, 73m consist of a person living alone. Over one third of the over 65s now live alone, the same percentage as Japan. Social Isolation must be increasing.

In previous Newsletters (#95 “Loneliness”) I have shown the distinction between social isolation and loneliness. The former is physical state. It reflects the day-to-day contacts between an individual and other people. Loneliness is a psychological state. It is the feeling of a gap between your current level of social interaction and some desired state. The two are related but it turns out only weakly. Individuals can be lonely when surrounded and interacting with many other people. People can be socially isolated and have little contact but not feel lonely.

Social isolation and particularly loneliness are associated with a multitude of negative consequences. These include death. A recent paper looked at 70 different studies and 3.4m participants. It showed that loneliness was associated with a 26% increased chance of dying. Social isolation alone was associated with a 29% and living alone 32%.

It is claimed that loneliness can also have a major impact on quality of life. Studies have shown a 50% increased association with dementia. Heart failure patients have a much higher chance of ending up in hospital if they live alone. Social isolation and loneliness appear to increase your chances of a heart problems. Your chances go up by 30%.

Association and Causation

The problem with these studies is that they measure association. Epidemiologists will take large samples and find powerful patterns. These might show an association between two variables. This does not necessarily imply that A causes B. It easy to slip in to talking about "cause". This is particularly the case in the popular media. It could be that some other variable causes both A and B to respond. A number of studies have shown precisely this problem. A vitamin supplement seems to be associated with a lower risk of a heart problem. A subsequent controlled experiment demonstrates no causation. There could be many reasons.

Reverse causations asks whether B in fact causes A? The pattern would be the same. Does poor health cause depression or depression cause poor health? The only true way of proving a causation is with a controlled experiment. Classically these are the drug trials. Individuals are randomly assigned to receive the drug to be tested and a placebo. The participants, and even the doctors, do not know who got what until the study finishes. Such experiments are impossible when studying the relationship between loneliness and mortality. They would completely unethical. Researchers are continuously working on the problem. All solutions are approximations but they are “triangulating” on the problem. The evidence is still strong.

Large scale longitudinal studies measure the same person over many years. They can be used to refute reverse causation. Loneliness can be seen to come before depression. There are natural experiments that can approximate the gold standard. For example, there have been large scale initiatives to improve loneliness. They often break down social isolation. A “before and after study” can offer strong evidence for causation. Many such studies report positive changes in health-related indicators. These include mental issues such as depressive symptoms. Sleep, blood pressure, inflammation, health-related quality of life, and primary care attendance all fall. Of course, the sample is not randomized. The latest set of studies use the genetic data of respondents. These are used to create a pseudo randomised study.

Small scale simulations try to replicate the impact of isolation. Individuals are denied social contact for between eight and ten hours. Fatigue increases and happiness is reduced. Other experimental studies have focused on the impact of ostracism. The impact of being isolated and ignored. “Cyberball” is a virtual ball-tossing game. It is used in social psychological experiments. In this game the participant supposedly plays with two (or more) other participants. They are in fact part of the computer program. The program varies the degree to which the participant is passed the ball. Ostracized players are passed the ball less often. They behave in line with everything we know about the impact of loneliness.

The Medical Input

There are strong associations between loneliness and heart disease. This includes heart attacks and strokes. A recent study however did not find the association. They could not prove it did not exist. However medical research, published this week, showed that lonely people have higher levels of certain proteins. Higher levels are associated with higher level of heart problems.

There is now a remarkable UK data base of over half a million people. Each has allowed their medical records including blood tests to be inputted anonymously. Each has provided a blood sample for genetic testing. Only with data like this can scientists start to explore protein production and subsequent heart disease. How the exposome and the genome interact together (See Newsletter #107 “The Genome and the Exposome”). The joint impact of genetics and our lifetime experience. Loneliness and social isolation are dangerous to mental and physical health and well worth fighting.

If you enjoy the Newsletter please recommend it to friends and send them to Substack to subscribe. It is free. Alternatively try the button below. All the Newsletters and lots more background is on my website thebusinessofage.com