Opinion: Loneliness in Serious Mental Illness – Why Does it Matter?

Even prior to the COVID-19 pandemic, there has been academic interest in the study of loneliness, as an increasing number of Americans, at all ages, are living alone. It turns out that this “loneliness epidemic” exacts not only a psychological price but appears to be a risk factor for premature death, comparable to smoking 15 cigarettes per day (1). Why should loneliness literally kill people? One explanation is that humans are biologically hard-wired for connection. Not having these connections creates physical stress that, when prolonged, takes its toll (2). Loneliness is a signal that something important is missing. 

Social distancing during the time of the pandemic has caused me to reflect on just how much my day-to-day activities involve other people, and how unpleasant loneliness can be. The current pandemic has also made me realize that it is not only strong connections (family, true friends, colleagues) whose absence we note, but also so-called “weak ties” (3): people we superficially know and may say ‘hi’ to, without engaging them much further. A person we recognize every morning on the bus would be an example. 

An important point to remember is that loneliness is in the eye of the beholder. It is a subjective, emotional experience, and it is not the mere physical absence of other people. For example, people may feel lonely in a group because they have difficulties makings the emotional connections they desire with members of the group. Eventually, some people even give up trying and avoid other people completely, which compounds the problem. Loneliness begets loneliness. 

“…[loneliness] is a subjective, motional experience, and it is not the mere physical absence of other people…”

Psychiatrists have traditionally concerned themselves with treating symptoms of an illness to relieve suffering. As a profession, we would do well to be more attuned to loneliness as an important concern for many of our patients. In appreciation of this, I recently started asking patients about loneliness in addition to the usual questions about sleep, appetite and anxiety for example. Patients of course have always known about loneliness. Some of my patients with serious mental illness have noted, “This has been my life all along, before COVID-19,” when I asked them about the loneliness-inducing effects of social distancing.

What can be done to reduce loneliness? Reaching out to people who are lonely is a good starting point. Peers and mutual support groups can play a crucial role in this. During the pandemic, telemedicine (including a simple phone call) has become an important way to remain connected to people. However, helping people feel less lonely does not mean to simply increase contacts. It is the quality of connections that matters. Increasing the number of meaningless social contacts will not do: it really doesn’t matter how many social media friends one has. What matters are friends or acquaintances who can help when there is a need, like a pick-up from the train station or moving a piece of furniture.

All members of society, including those with a serious mental illness, need to feel a sense that they belong to their community, with both strong and weak ties. Without connection, we risk feelings of alienation, psychological decline, and quite possibly also an earlier death. Being part of a community not only makes one feel psychologically better, it also makes one physically healthier. We need to pay attention to a worsening epidemic of loneliness within the current pandemic, particularly for a group of people who may have been lonely to begin with.

Oliver Freudenreich, MD, FACLP
Co-Director, MGH Schizophrenia Program


  1. https://www.hrsa.gov/enews/past-issues/2019/january-17/loneliness-epidemic
  2. Holt-Lunstad J. The potential public health relevance of social isolation and loneliness: prevalence, epidemiology, and risk factors. Public Policy & Aging Report. 2017;27(4):127–130.
  3. Granovetter MS. The strength of weak ties. Am J Sociology. 1973;78(6):1360–1380.