This post is about loneliness and social isolation, especially with regard to seniors. Presented are noteworthy views, both conceptually and empirically, as well as some puzzlement.
First it’s necessary to clarify terms and I also propose a new word. Loneliness can be distinct from social isolation. Loneliness is usually defined as perceived isolation. You might feel lonely in a crowd or even while married, while someone else living in solitude might not feel lonely at all. It’s your feeling lonely that makes it so. Social isolation, on the other hand, is more extrinsic. Think of an elderly person living alone, perhaps with a visitor once a week, and not being able to get about unassisted. It doesn’t have to be that extreme but that’s the idea of social isolation.
Although loneliness and social isolation are often viewed as distinct, there are times, as in this post, when you want to treat them as a unit, summarized in a single word. I couldn’t find such a word so I manufactured one. That word is soloinity. For this essay, soloinity refers to being alone without others around, either figuratively or literally.1
Soloinity and Health
Chronic soloinity is not only bad for quality of life but also for health. It can have the same mortality risk as smoking 15 cigarettes a day or 2-3 times the risk of obesity.2 Compared with those who feel connected, chronic soloinity over time is associated with significant loss in ability to do normal activities of daily living3 and increased levels of cortisol and c-reactive protein, biomarkers for disease and disability. 4
Soloinity and its effects occur at all stages of life. Let me share an experience from when I was an undergraduate. I was interested in a part-time position working with psychologists who were doing research on babies. During the interview I was told one of the studies involved snapping the soles of infants’ feet with a rubber band and recording their cries. (This was during the 1960s.) Some of the infants in the hospital ward had mothers who would pick them up and hold them for feeding and care while other infants had been abandoned. Of the latter, babies without caring mothers, some would laugh when snapped with the rubber band, almost surely because they sensed they were getting attention.
Soloinity apparently creates social inefficiency. A friend, a professional in health policy, shared with me that during a meeting with public emergency medical service (EMS) personnel, such as Medic One responders, he was told that 40% of their calls were not emergencies but brought about from loneliness. I found that estimate hard to believe. A few days later I asked a local police lieutenant about false calls related to loneliness or isolation and mentioned the 40% figure. He said he couldn’t speak for EMS but based on police work 40% seemed reasonable. I made a few phone calls enquiring about recorded facts but it appears data on the likely causes of false EMS calls is not regularly collected. In summary, soloinity is linked with anecdotal undocumented societal cost.
A Puzzling Graph
Eliciting my interest in soloinity and related was a recent research project on successful aging. The project was for Empire Health Foundation, the same organization for whom I estimated county health care expenditures.5 An offshoot of the research was the following chart.
The 39 data pairs represent seniors in Washington Counties. I find this graph intriguing. It shows a strong linear trend between a measure of disability and population density: as proportionally more seniors live in a county the prevalence of a disability decreases. The linear trend is significant (p < 0.001), as is a similar trend with senior poverty. The best fitting model of those I tried was where population density is both a standalone independent variable and additionally as a variable that interacts with poverty estimates:
difficulty independent living = 17 – 0.5*density + 0.04*density*poverty, adj R^2 = 0.66.
Including poverty in the model makes sense since many poor outcomes are related to poverty. But the significant relationship of difficulty with independent living and population density is surprising. I think it reflects declining soloinity, that more seniors in a county means increased opportunity to connect and therefore less soloinity and its bad effects. But that’s only conjecture. Furthermore, it’s not supported by looking at a few other states. It would gain credence, however, if there were data or a study that linked loneliness or social isolation to population density. I searched the web but unfortunately couldn’t find either data or a study to support the soloinity conjecture.
What did appear during the search was a paper showing another remarkable empirical relationship with population density, this time across the 50 states. The paper’s Figure 1. shows a strong trend of the number of suicides per 1,000 people decreasing as the number of people per square mile increases (log transformation of density, adj R^2 = 0.72).6 It makes sense as higher density presents greater opportunity to connect either for support or to alleviate feelings of aloneness or worthlessness, which in turn implies fewer suicides. Of course, that reasoning is also conjecture. Still, it’s a pretty strong correlation and may relate to the graph above and the effect of density. According to the paper, the single best environmental predictor of suicide in the U.S. is population density.
What About Other States?
I looked at similar data from four other states. Here’s an infographic highlighting salient features of the five states, including Washington.
Of these, only Oregon showed the same relationship as Washington regarding difficulty and density. The Oregon two-variable regression model was similar to the Washington model above (statistically the same coefficients, all significant at p <= 0.006, but with an adj R^2 = 0.27).7 However, poverty as a standalone variable was not a significant predictor of difficulty (p = 0.07). This is indicated in the infographic by the lack of a red trend line in Oregon’s plot of difficulty v. poverty. Overall, the three other states, New Mexico, South Carolina and West Virginia, had disability averages around 19%, were more variable, and shifted to the right. The national average for independent living disability is 16.2%, so Washington and Oregon, both with means of 14.1%, are below average.8 (The national average is from the same source given in the charts. Note that the states other than Washington were convenience selections with an eye to having between 35 and 55 counties within a state. Over two-thirds of the states don’t fit that criterion. Georgia and Texas have more than 150 while the New England states range between 0 and 16 counties.) The result: there’s no confirmation looking at a handful of other states.
What’s to Be Done
To help resolve whether or not soloinity underlies the relationship between impairment and population density requires linking data or finding studies that demonstrate a relationship between the two. We already know that soloinity is correlated with impairment; is it also associated with population density?
What about interventions to alleviate soloinity and its effects? If you search the web, you’ll find more than enough literature on loneliness, social isolation and related interventions, but not much on intervention efficacy.
Interventions can be viewed as either increasing connectedness or fostering meaningfulness, sometimes both. Accomplishing either of these is frustrated by our society’s consumer ethos which emphasizes the immediate and individualism, both contrary to one’s outlook later in life. Ignoring that hurdle for now, I’d like to touch on meaningfulness as an intervention.
There’s this adage about aging (and meaningfulness):
The young man said to the old man, “What is your greatest burden as you
grow old?” The elder replied, “That I have nothing to carry.”
While doing the research for EHF I was especially struck that some organizations in the EU were using the idea of generativity as a conceptual model for helping seniors overcome soloinity.9 The term, generativity, was used by Erik Erikson beginning in the 1950s though in another context.10 Relevant to this post it means being engaged in furthering the next and future generations. Here’s an example. A couple years ago while living in Seattle I participated in a group of mostly senior citizens actively working to ensure that Social Security would be there for our children. I don’t think anyone in the group–and some had difficulty just getting to the meetings–were concerned about losing their Social Security: the concern was that it should be available for the next generation. That’s how I see the concept of generativity as an intervention. Seems like a promising idea.
- Soloinity was derived from the latin prefix, solus, meaning alone, by oneself with no others around (http://en.wiktionary.org/wiki/solus). It was made into a noun by adding inity and then rendered, hopefully, easier to pronounce as soloinity rather than solusinity. Shorter versions like solinity were rejected because of confusing or negative connations found using a Google search. ↩
- Perissinotto, Carla M., et al. 2012. Loneliness in Older Persons: A Predictor of Functional Decline and Death. http://blogs.umb.edu/sylviaguillory001/files/2014/03/death_nurture-1f0ofy9.pdf ↩
- Ibid, and many other references readily found via a search. ↩
- Rueggeberg, Rebecca, et al. 2012. Associations Between Health-Related Self-Protection, Diurnal Cortisol, and C-Reactive Protein in Lonely Older Adults. http://sites.northwestern.edu/foundationsofhealth/files/2013/03/Psychosom-Med-2012.pdf. ↩
- http://www.lettingthedataspeak.com/county-health-care-costs-washington-state-example/ ↩
- Benson, Kirk. 2009. Predictors of Suicide in the United States. http://isites.harvard.edu/fs/docs/icb.topic668880.files/RegressionReportKirkBenson10910.pdf. ↩
- The data used for this analysis will be at this link for about a year: https://dl.dropboxusercontent.com/u/4207674/DisabilityDensity.xls. ↩
- It’s possible that states with a lower disability average or less variability than NM, SC and WV might exhibit a significant relationship between disability and density, that the three states’ higher average or variability is caused by confounding factors which obscures such a relationship. ↩
- See, e.g., Villar, Feliciano. 2012. Successful Ageing And Development: The Contribution of Generativity in Older Age, http://diposit.ub.edu/dspace/bitstream/2445/44303/1/614149.pdf, or Kruse, Andreas and Eric Schmitt 2012. Generativity as a Route to Active Ageing, http://www.hindawi.com/journals/cggr/2012/647650/. ↩
- http://en.wikipedia.org/wiki/Erikson’s_stages_of_psychosocial_development. ↩