SPEAKER 1: This is a presentation by Human Development Outreach and Extension at Cornell University.
JOHN CACIOPPO: And so the research I'm going to be talking about represents the work of some really wonderful people. And so you'll see a lot of different levels of analysis that we're going to bring to this problem, and it's because I have great colleagues at each of those levels of organization. And at the end, I've got this slide where I'm going to thank them. But it seems fitting in some ways if you're going to study loneliness that you need a crowd to do it, and I have a great crowd.
We've been asking the question, as silly as it sounds, what's fundamental about being human? And if you look at the 20th century-- to address that question, biological, behavioral, cognitive scientists tended to focus on single organisms, organs, cells, intracellular processes, and genes. From the perspective of many scientists during the 20th century, the contributions of the social world were thought to be best considered later if at all. And the reasoning was that social factors were of minimal interest because they played a minimal role in development, structure, or processes of the brain and behavior. Some argue that well, they are relevant, but they're so complicated that they're best considered later-- not considered at that time.
And you can see that in this slide where we have the 50 year celebration of the discovery of DNA. And you have this simple DNA leading to RNA, protein, brain, and everything else with genes and biology thought to determine human behavior. Now, this appeared not that long ago-- just a couple of years ago-- when we also thought there were 100,000 or more genes because it required that many genes to actually explain the complexity of behavior that we see. We now know, of course, that we have half the number of genes as cultivated rice.
Further fueling, though, this focus on the solitary individual and scientific analysis was the dominant metaphor of the mind-- the desktop computer. Complete with input processing, long and short term memory, and output stages, the brain was thought to be analogous to hardware in the mind to the software in the computer. Culture in this context is like the computing operating system-- PC or Mac.
How things have changed just since the dawn of the 21st century and the internet. If you had a computer that was connected only to an electrical outlet, you wouldn't really have a particularly powerful computer by today's standards. To understand computers today, one has to appreciate their capabilities as a connected collective. Culture in this context is not so much about the operating system in a solitary computer as it is the norms, conventions, and practices that have evolved to promote the effective connection and interaction among sets of computers.
Whereas computers have been connected on the order of years, hominids have been mobile and broadband connected for hundreds of thousands of years. Indeed, if you asked a zookeeper to create a proper enclosure for the species Homo sapiens, she would list at the top of her concerns "obligatorily gregarious," meaning that you would no more house a human in isolation than you would an emperor penguin in hot desert sand. It just doesn't make sense to put a creature in an environment that stretches its genetic leash just quite that far.
We like to think of ourselves as rugged individualists, but we're fundamentally social organisms. We're born to the most prolonged period of abject dependency of any species. For the species to survive, human infants must instantly engage their parents in protective behavior, and the parents must care enough for their offspring to nurture and protect them.
Even once grown we're not particularly formidable species. We don't have the natural protection of armor or of flight, and the weaponry at our disposal either in teeth or claw is nothing compared to most mammals we'd have to do battle with. Our major evolutionary advantage is our brain-- our ability to reason, communicate, remember, plan, and work together. That is, our survival depends on our collective abilities not our individual might.
Among humans, ongoing partnerships between mates combined with high male investment in protecting offspring contribute to a tipping point in reproductive success. The theory is that parental teamwork meant not only increasing number of children might survive but that these creatures could afford to be more developmentally and behaviorally complex-- thus, the longer period of abject dependency.
But the greater behavioral latitude that this produced lead to greater diversity, which led to greater innovation, which led to more rapid cultural development. Humans weren't the first bipedal creatures or the first to use tools. But humans apparently, uniquely, contemplate the history of the Earth, the reach of the universe, the origin of the species, the genetic blueprint of life, and the physical basis of their own unique mental existence.
The social nature of human species is not simply an add-on either. It may well be that it has shaped fundamentally the evolution of our biological design, including the rapid increase in neocortical connectivity and the associated increase in intelligence. According to the social brain hypothesis, deducing better ways to find food, avoid perils, and navigate territories has adaptive value for large mammals but the complexities of their ecological demands pale by comparison to the complexities of social living.
These demands include learning by social observation; recognizing the shifting status of friends and foes; anticipating and coordinating efforts between two or more individuals; using language to communicate, reason, teach, and even deceive others; orchestrating relationships ranging from pair bonds in families to friends, bands, and coalitions; navigating complex social hierarchies, social norms, and cultural developments; subjugating self-interest to the interests of the pair bond or social group in exchange for the possibility of long-term benefits; recruiting support to sanction individuals who violate group norms; and doing all this across time frames that stretch to the distant past to multiple possible futures.
Although such hypotheses are inherently difficult to test empirically, some attempts have been made to do so in troops of baboons. Silk and colleagues demonstrated that a composite index of sociality was highly correlated with infant survival. Cross-species comparison by Robert Dunbar and colleagues indicate that the evolution of large and metabolically expensive brains is more closely associated with social than ecological complexity.
And that's actually what's graphed in the right panel here. It seems that the integration into a social group is essential for survival leading to the sculpting of physiological systems capable of integrating in social groups. For many animals, including humans, the absence of social integration presents a threat to survival and therefore elicits potent behavioral and physiological stress responses.
Social species create emergent organizations beyond the individuals-- structures ranging from dyads and families to institutions and cultures. These emergent structures evolved hand-in-hand with neural, hormonal, and genetic mechanisms to support them because the consequent social behaviors helped organisms survive, reproduce, and nurture their offspring sufficiently long that they, too, reproduced. These higher organizations have long been apparent, but we're now beginning to understand a little bit more about their neural underpinnings.
So this is from the cover of a book we did in '94 called Emotional Contagion. And it's quite obvious that all these individuals are mimicking the behavior of this particular actor. And so you can represent that fairly easily as some stimulus-- an observed person. There's an implicit monitoring of individuals that's going on. There's a mental representation and a correspondent motor response.
But a little over a decade ago now, Rizzolatti and colleagues in Italy were recording from neurons in the fusiform area and frontal gyrus. And they found that those in the frontal gyrus were fine when the monkey were grabbing objects. And then, accidentally, while the electrodes were still in play, one of the experimenters reached and grabbed an object, and they heard those neurons firing.
And so they pursued that, and they found out-- this is basically the paradigm. They found out that there's the so-called mirror neurons. So-called because it's the same neurons involved in the performance of an action are involved when observing that action on the part of another. And so this is just from a review that they did in the annual review of Neuroscience, I believe. You see actions of these neurons and when you have the same hand movement but there actually is no grasping, you get much less activity out of those neurons.
If the monkey is doing these two actions, by the way, you see the same pattern. That's not the action per se if you-- in fact why I've showed this picture is-- these individuals-- you know not only that they are mimicking that individual but if I asked you about their state of mind, you'd probably say that they were on this person's team. That is, it's not just the observation of the action, it's the observation of an action with a goal in mind. You want this person to succeed. That makes you actually lean forward. If you doubt this, watch a parent when their child is doing some sports activity, and you'll see they're not doing their movements correspondent to just any child out there. It's their child that they're following.
And indeed, the Rizzolatti, et al. Work showed that it isn't the observation of the action that's important, it's really the observation of the action in the context of some goal. That will drive the response strongly as well. And they did that by showing these sets of actions-- one or the other-- and then occluding the area where the block was or was not. And you get the same activity as if the occlusion wasn't there. And so you have this second route, if you will, that's goal directed. Oh, this has been used to talk about simulation, mental simulations, theory of mind simulations, language parity by Rizzolatti and others.
And I'll just point out that it's interesting that mirror neurons in these analyses have been limited, again, to just a single individual when, in fact, it's very simple to take the output of one and that's the input of the other. And then we get a second level of cognition and a more social cognition meaning that it's going across two different individuals. And with that you get a co-regulation.
We see co-regulation all the time. Watch two people on campus walking toward each other. They walk at different paces. Their arms are doing different things. They greet each other and walk off together. Neither has tried, but now they look like a single unit walking off because they're actually regulating one another. And that's basically what's captured in this kind of perspective. So that's the level of looking at these aggregates and looking at what might be the underlying neural pinnings that we were interested in.
And then we started to think about how we typically do experiments. And so whether it's a mouse or a human mouse, we tend to study them in isolation. Well, in the neurosciences if I wanted to understand how important a variable is, like a gene or a piece of the brain, I look at what happens in those with that piece and without that piece. So we do knockout genes or look at Phineas Gage.
So that notion came to mind if we want to look at how important these social organizations are, look at people with and without such organizations. And indeed, if you walk out testing an individual subject and see someone sitting by themselves, we immediately think something's amiss-- something's just not quite right. A variety of studies have shown people spend about 80% of their time each day with others. And, in fact, Danny Kahneman has done a day reconstruction method and finds people prefer that part of their day as well.
So we, basically, started looking at the social isolation work. And this study certainly caught our attention. It's a view of the prospective studies of social isolation to integration. I know a number of people in here are well aware of this work-- not everyone perhaps, though. Basically, you're looking at whether the person has a lot of contact with others or not. And all of the studies-- it was five at this time. I think it's around 17 or so now. They all show that the more isolated the individual the higher rate of morbidity and mortality from all causes-- not from a single. It seems to be all cause, morbidity-- mortality.
Now, the most obvious explanation for this is what's called the social control hypothesis. I call it the spousal nag hypothesis. The notion is that if you are integrated-- you have friends, family, spouse there-- then you have people to tell you, you're getting fat. Go get exercise. You're drinking too much. Stop drinking. Stop smoking. Get some sleep. You have someone to tell you what to do and hope you live a healthier life. Well, that wasn't actually documented particularly well in this study. And subsequent studies have tried to quantify health behaviors. Yeah, it accounted for a little bit, but not very much of the variance. And so that's when we entered this picture to start to look at this.
And I went to the non-human literature and this is just a sampling. But just to give you a quick idea-- isolation in the fruit fly leads to early death in that fruit fly. Isolation promotes the development of obesity and Type 2 diabetes in mice-- delays the positive effects of running on adult neurogenesis in rats. It increases activation of the sympathetic adrenomedullary response to acute stress in rats.
It decreased the expression of glucocorticoid receptor genes in piglets. That's the species here-- decreases open field activity and increased cortisol levels in pigs, larger morning rises in cortisol in monkeys, and higher 24 hour urinary catecholamines and oxidative stress in rabbits. And a variety of different species-- and I doubt we really have spousal nagging going on in these. There seems to be direct biological effects of this isolation of social animals.
Oh, that led us to take a slightly different approach to this-- not that we didn't measure health behaviors, but we didn't rely on that as the likely major cause. It was just one of a variety of transduction pathways we've been investigating. And very quickly-- and this will be a surprise to no one here-- we found that objective isolation isn't nearly as important when looking at a population as a whole as subjective or perceived social isolation.
And these three pictures capture that. None of these people are actually physically isolated from others. Yes. Yes. A parent may well have just departed, but I have no doubt there are other adults with them and perhaps even one of the parents. This woman lives in an urban setting surrounded by individuals. This child is sitting there with other children around. And in all three cases, they feel isolated. If you've ever been the last one chosen for a team on which you actually wanted to be-- well, you were still on the team, but that's the perception of isolation that's actually much more impactful and traumatic.
Now, if you are forcibly isolated, that's a different question. But when we're doing population based, we're not dealing with that. So we measure objective and subjective isolation, but in fact it's the perceived social isolation that's going to produce all the effects I'm going to talk about. And, honestly, we've not seen anything objective isolation produces its subject didn't mediate.
Now, it turns out that this perceived isolation has a term in psychology, and the term is loneliness. We didn't coin it. They've been studying it before I got to this area. And it is characterized as the gnawing chronic disease without redeeming features. Loneliness was thought to be depression, negative affect, neuroticism, low social skills, shyness. And I had no reason to disagree with that when we first started. I hope by the end of this, you'll agree with me that that's not the case. It's actually much, much more interesting as a phenomenon.
I see that the right part of the slide is cut off. It's not here, but it is up there. So I apologize-- I may occasionally be describing what you're not seeing. We'll try, at least.
Oh, for our first entry into this-- was we tested about 2,600 undergraduates at Ohio State. These were kids taking Intro to Psych, so they were primarily freshman. And this represents the top and bottom quintile that we followed for a while. And I show this slide because loneliness-- social isolation-- doesn't travel alone. It has all kinds of travel partners. Social support is lower. Shyness is higher. Social skills are lower. Anger is higher. Anxiety is higher. Self-esteem is lower. Fear and negative evaluation is higher. Optimism is lower. Positive mood is lower. Negative mood is higher. Depressed affect is higher. These are the travel partners of loneliness.
And that's a real concern, obviously. I'll show you personality factors in a moment, but this is a concern when doing this kind of research. And so we've tried to address those concerns with two different approaches. One is longitudinal research. Early childhood attachment all of a sudden disappears. There's a potential confounding because we're going to look at changes within that. Genetic constitution disappears as a confounding-- because we're going to look at changes within that. So we looked at both cross-sectional and longitudinal research.
And I will be speaking primarily about a five year-- we're now in year 7-- but we now have five years worth of data that were analyzing from a population-based design in Chicago. In 2000, we recruited 230-- 229 to be specific-- men, women, African-American, Caucasian, Hispanic individuals in the metropolitan Chicago area between the ages of 50 and 67 years. And we've been following them annually since that time.
In addition, we do experiments. This is the results of the first such experiment where we hypnotized. Actually, it was David Spiegel-- an experimental hypnotist-- psychiatrist-- out at Stanford. He had published a study showing that in his high hypnotizable subjects-- in a brain PET study he and Steve Kosslyn did-- that they got activation in these high hypnotizable subjects that was really pushed by what they were told they were seeing whether it was black and white versus color, not based on what they were actually seeing. I found that to be compelling.
So we flew out-- David's a friend-- we flew out and he hypnotized subjects. And we induced in the same individuals, based on what we knew to differentiate these two in terms of loneliness, feelings of high or low loneliness and then gave them the same scales. The ordinate is the same. I made the ordinate big so that all the scales would fit on this. Not all these measures have the same scales so some look shrunk. Well, that's because the standard errors are smaller. Many times the scale points are smaller. But I wanted you to see the profile across. That's why I did this.
And it looks casually and, in fact, statistically these are very similar patterns. Now, the difference is that these are the same individuals induced to feel high or low loneliness. And this was an experiment that made me think for the first time, honestly, that loneliness isn't sitting there with all these other passengers on the bus. It's driving many of these other effects. And it caused me to rethink how central loneliness or social connection to others might actually be.
Well, we worried about clearly there are differences in personality. This is the big five in our study with the-- embedded is the top quintile and low loneliness. The middle quintile and lonely is the highest quintile of loneliness in our undergraduates. And what you see is in all except-- it's openness that differed, but only because the very, very low lonely differ from the middle and the high lonely. Usually one would think the high lonely differs from everybody else. No, it's not. It's the very, very low lonely that differed. And so we enter these variables into our equations as covariates to make sure they're not pushing anything around. And they're not going to modify any of the effects that I'm going to show you.
Now, what I've got here is a variety of levels of organization starting from the demography and going down to genes. And I just want to go through each of these and show you how they map onto this construct. Demographically, we've seen huge and surprising changes. In 1940-- light blue means that there are less than 1% of households with single residence. 1940-- most states were characterized by that. By 1970, it was 15 to 20%. By 2000-- another 30 years-- we're about 25% of the households had a single individual living in that house-- a dramatic increase just over that 60 year period. We're not alone. That is, the United States isn't alone.
These are data from China. And just in the last 10 years, the movement from villages and living in family households to living either in dyads or as individuals has increased dramatically. It's the same kind of increased isolation there.
At the cultural level, it's actually, initially, often a surprising finding to people. Collectivistic cultures show higher levels of loneliness than individualistic cultures. Now, this is the distribution of loneliness that we find again, and again, and again in those cultures-- in individualist cultures-- with this thing low lonely. Most people are not lonely. The default state is not being lonely. People do something about it when they get lonely. I'm going to argue that loneliness is very much like hunger, thirst, and pain. It is a signal to cue you to do something that's important for your survival and for your genes' survival.
Phylogenetically-- ontogenetically-- we do not survive long on our own. We are effective because we work together. Just like thirst, and hunger, and pain-- that's so important for your survival that we have an aversive signal to alert us when we're losing connection with others around us to motivate us to do something. And that's why we have the default state-- being connected, not lonely.
Why in the collectivist culture-- the levels of loneliness greater? It's because these individuals, when they are in that circumstance, by comparison to those around them into what is normative-- what is valued in their culture-- find themselves to be especially out of sync with everyone else. So if you look at bereavement in Italy and bereavement in Denmark-- in Italy, it's especially salient that everybody else shares a common bond, and I'm outside of that common bond. And that seems to be the reason why you get an even higher level of loneliness in these collectivistic cultures.
If we look at economically-- I worried in part that loneliness might be people like myself-- just funny looking, right? They're short. They're fat. They're, perhaps, not very bright. And they're ugly-- just the opposite of somebody like-- Jordan may have less social capital to bring. And so, obviously, they negotiate fewer relationships and they're left to feel isolated and lonely. Well, very anecdotal cases-- Ben Affleck's a case where he's-- this if just a pop piece talking about how important it is to have relationships. And when he does badly in a movie, he's figured out that isn't the end of the world because he's got-- what's more important are family and friends.
And indeed, you find that what he's saying is typical of other celebrities and typical of the students that we studied. We looked at their physical attractiveness, their household income, the educational status, height, weight, body mass index, age, gender, ethnicity, number of credit hours, scholastic aptitude, motivation, performance in school, number of roommates, and time spent alone and none of that was related to loneliness. That's not that they didn't vary on these dimensions. None of it mapped into loneliness. So it isn't the social capital that one brings into it.
Income-- a few studies have found income to be related to loneliness. That is, the higher the income, the lower the loneliness. The higher the household income, the lower the loneliness. When you put it in marital status, that effect disappears. So here's a shock. Marital status is associated with lower levels of income. And married people on average have higher household income than non-married individuals. So that seems to be why you have the income effect.
Number of voluntary associations, physical health symptoms, and disability-- people who are disabled do, in fact, find themselves isolated and end up lonelier. Chronic work and social stress, small social networks, lack of spousal confidant, and poor quality social relationships-- those are the factors that we found and others have found to be related to loneliness.
At a sociological level-- just in the past 20 years between 1985 and 2004, there's been about a 50% decrease in network size. Perhaps more surprisingly, in 1985 and 2004 these respondents were asked how many confidants do you have? The modal response in 1985 was three. The modal response in 2004 was zero. So we've seen not only structural changes but changes in just the time people are spending with others and the confidants that they have at their access-- at their disposal.
We also know that there are these three, if you will, facets-- dimensions-- of loneliness-- what we call intimate isolation, relational, and collective. Intimate-- this is the extent to which someone else affirms who you are to yourself. Marriage is the best predictor of scoring low on this dimension or high on intimate connectedness. But, in fact, you can be isolated in a marriage.
Linda Waite, a sociologist who's a colleague and collaborator on this research, has demonstrated that marriage has health protective effects. When you put loneliness in those equations, the health protective effect of marriage disappears. In fact if anything, it starts to go the other way. If you're isolated and feel estranged in your marriage, it's an extra stressor.
Relational are all your face-to-face relationships. It's really what people think of when they think of loneliness. They think of just a third of it and this is relational. The best predictor of that's frequency of contact with friends and family.
Collectivist is a social identity-- being a Cornell faculty member or student-- being an American. Right after 9/11, you may recall the harmony that was felt-- the increased cooperation and altruism that even made media. Well, that's because of making salient the collective identity of being American after the tragedy of New York City. The best predictor, at least in our population-based design, is number of voluntary associations. I won't speak much about gender differences, but I will tell you here is where we do get a gender difference. This structure is there in young adults, old adults, African-Americans, Caucasians, Hispanics, men, women.
But the waiting is slightly different for men and women. For women, the relational isolation is relatively important compared to collective-- vice versa for men. The colloquial way of thinking about this-- I joke with my wife-- is that I don't understand her need for girlfriends and to go shoe shopping with her girlfriends. Going with me just doesn't do it. She wants to go with her girlfriends. Whereas she looks at me on a Sunday when I'm watching-- because it's going to be this year that the Chicago Cubs wins that World Series. So I'm sitting in front of my TV. She goes, what an isolate. And I'm sitting there-- I'm with my team. I'm with my crowd. So that's a difference, actually, in perceptions or antecedents between men and women. Otherwise, there aren't a lot such differences.
Let's go down to the behavioral. And, indeed, health behavior does differ between lonely and nonlonely-- not in young adults but in older adults. We find, for instance, that lonely individuals are less likely to engage in vigorous exercise. If they stop vigorous exercise, lonely individuals are less likely to pick it up again.
We looked at whether that was due to social control or self-regulation because as I'll show soon, lonely individuals suffer a decrement in self-regulatory capacity. And what we found was social control didn't predict the extent of this difference. It was self-regulation. It's not limited to exercise. When you look at other comfort activities, lonely individuals also have a higher proportion of calories in fats and sugars.
This is the first hint we had that there were differences in self-regulatory capacity. This is a dichotic listening task. Under no instructions, you get a right ear advantage. Consonant vowels-- you're trying to mirror back what it is that's being presented. We do better when presented in the right ear than when presented in the left ear. All of these are right-handed individuals. If you're told to listen to what's in the right, everybody does exceptionally well, so I haven't even bothered to graph that there's no differences. It's kind of a ceiling effect.
The interesting thing is when you're told to listen to what's coming in the left because that's kind of at odds with what the default-- your standard-- is. And you see, people can do that. I mean, if you're told to listen to the consonant vowels coming in the left, you do better than if you aren't told to do that. But we have an interaction here so that it's the nonlonely who are especially good at that. The lonely aren't as good when kind of bucking, if you will, what would be the default mode of processing.
I mentioned that loneliness by me and others was thought to be nothing other than negative affectivity in eroticism or depression. So we've paid a lot of attention to that. The CES-D is an epidemiologic measure of depression. And one of the items it is is "I feel lonely." We've done factor analyses in several different studies now where we've looked at the loneliness scale and the CES-D. The "I feel lonely" item loads in the loneliness factor. There are otherwise two peer factors.
This is a cross-lagged panel analysis of five years worth of data from our Chicago Health, Aging, and Social Relations Study. What I have is each year of loneliness, each year of depression, and various covariates. And we load in-- this is just the first pass. We actually load in social support, hostility, objective stressors, perceived stressors, neuroticism-- other personality factors. It doesn't actually change the structure. What you see is loneliness leading to the next-- predicting next year higher levels of depression net all these other effects. There's some hint that depression may contribute to loneliness, but as soon as we start loading in some of the other covariates, that path renders non-significant. The loneliness path remains significant.
This is-- I'll skip this. This is just one of the points of that slide is to say that when clinical psychologists looking at depression are now seeing one of the problems being that it's very inward focused. People start to trim outside activities and think about what is it? If I can go deep enough inside myself, maybe I'll discover what it is. And, of course, the result that I just showed you might suggest that the escape is really outside, not rumination. And that's, actually, Helen [INAUDIBLE] point in that quote anyway in her studies of rumination.
Both in the early studies of ours and in other peoples, this focus on negative affect and on sadness was thought to be the most important. The evolutionary story I've laid out for you says there's an implicit sense of danger particularly from others that's important. Being disconnected from others puts me in danger. I might not survive. My genes might not survive which suggests there may be this threat surveillance that's operative. Now, that's easy to say-- harder to actually look at.
So we thought, well, if that's the case, let's say that we're in evolutionary time. You're sitting there with a stick fending off wild beasts. Well, this is not a great daytime activity to enjoy and probably not a long-term profession. But when would you particularly be vulnerable? It's not in the daytime with the stick although you clearly are vulnerable then. It's when you go to sleep. If you don't have that safe social surround, then you really are at danger-- at risk for predation.
So we thought about doing sleep studies to see if, in fact, there was evidence that they showed less salubrious sleep-- that is, more micro-awakenings. We put them to bed. We awakened them. It was in the hospital. And the reason I wanted to look at sleep was because they're not awake. I mean, they're not conscious. They're not interacting with somebody. It can't be the other person who's at fault. They're asleep. So this was the first test we did of this implicit threat surveillance notion.
These slides are to remind you that sleep does more than just pass time. It's also a quintessential anabolic physiological event. We don't only have wear and tear of our roads. We also repair and maintain our roads-- our bodies. And sleep is a quintessential activity of that kind. So could it be that lonely individuals not only have more terrorized days, in some sense, but also less salubrious nights. And so that would have a health implication.
These are the results of the sleep study. And what we found was more micro-awakenings in our lonely than our nonlonely young adults. We find this in the hospital. We find this in their homes. We used also the Pittsburgh Sleep Quality Index which is a self-report measure. And we find the same things using the Pittsburgh's Sleep Quality. They report greater daytime dysfunction-- less sound sleep.
And when we moved to older adults, we had to reduce the response [? burn ?] in some ways. So we've been using the PSQI in those studies, and we get the same effect. In fact, the effect in older adults is even bigger on the PSQI than we find on the young adults. One of the differences we do see is older adults are also taking more sleep meds. Older lonely adults are taking more sleep meds than older nonlonely adults. We didn't see that in kids-- in our young adult sample. Well, anyone who's ever taken sleep meds knows they work for a short time and then habituate. They no longer work. So they've tried what they can, and they still have greater daytime fatigue.
This is a longitudinal study from the [? Chaser's ?] data. And what you see is-- this is, again, within subject now-- a person's feeling of isolation-- their loneliness-- on day one predicts next day feeling of fatigue. And that's, by the way, net activities of that day. Daytime fatigue on day one doesn't predict how lonely they feel the next day. And that replicates across that three day period. So it seems that not only is this a between subject difference. We get this longitudinally as well.
One other kind of result that we see in young adults is that the frequency of uplifts and hassles is comparable for lonely and nonlonely. When we did an experience sampling method, the hassles and uplifts themselves are comparable for lonely and nonlonely. Most of the hassles and uplifts have to do with other people. That's probably not a surprise to you. What differs here is the rated extremity.
The apocryphal story I tell is a nonlonely individual walks down the street. And somebody bumps into them. And it's just no big deal. A lonely person-- the same thing happens to and it's another burden in an already overburdened life. It's a feeling you have after staying up all night and doing a grant because of the deadline, right? The next day's tough. Well, this-- they walk around with that mood all the time.
So we thought, OK. If that's the case, we should be able to see this in imaging and get at what might be contributing to it. So we presented a series of pictures that were social or nonsocial, matched on valence-- that is, matched on how positive, and matched on arousal-- matched on visual complexity. These are just illustrations. Then a contrast between those two-- and then convolved that hemodynamic response of each voxel with the loneliness of that person across all the people in the study. Think about it as each voxel's been correlated with the loneliness score. So you have a brain with all the voxels over here with the loneliness score, and the hemodynamic response from that contrast, and then you just do the convolution.
We have false discovery rate, criteria for cluster size, and then we end up with the difference between those two. And you see a very large region of the ventral striatum-- basic limbic reward machinery that in nonlonely individuals is stronger in response to this social picture than nonsocial. Now, I want to remind you these pictures are selected so that lonely and nonlonely rate these two sets of pictures equally positively and equally positively to each other. So they've been selected to be matched on this. And yet the nonlonely are the ones showing the greater reward activity.
I think about this as if you're nonlonely, this is like ice cream. If you're lonely, this is like 24-hour banking. Both are really good, right? But one activates the reward machinery more than the other. Now in truth, both of these pictures are eliciting ventral striatum activity in our lonely and nonlonely. It's relative activation. For the nonlonely, this is showing greater ventral striatal activity than is this set. For the lonely, it's actually the reverse. Now remember, uplifts are less uplifting. They're less quote, "severe." Well, that might well be what you would get by this kind of process.
When we did this with negative pictures, we found a very different pattern of activity. We find in this set of contrasts-- same convolution-- that the lonely show greater visual cortical activity to threatening or negative pictures when they have social content. Now, again, loneliness is associated with this sense of danger and threat, particularly in the social domain. So when you pair that-- something negative and social-- it's particularly attention getting. They drive and ping more visual cortical-- perhaps even eyes focused on it more.
The interesting thing is as you follow the visual stream forward, you get to this temporal parietal junction-- bilateral. Theory of mind is associated with right temporal parietal junction-- perspective taking. Moving your attention around to take the perspective of the other is associated with bilateral TPJ activity. Now, who's showing that tendency is the nonlonely individuals. The lonely individuals are the ones who, in theory, feel threatened-- feel danger. They're the ones who are at risk. If you're talking to someone but you're the one at risk, you don't have the luxury of taking their perspective as much as if you feel safe and secure.
So our thought is that what's going on here is the lonely individuals are staying more egocentric because of this implicit feeling of threat. And the differences we see in social skills expressed by adults isn't because they have different batteries of social skills. In fact, we and others have found they possess the social skills needed. Instead, it's that they're not accessing them because they're not taking the perspective of the individual with whom they're actually interacting as much as the nonlonely individual.
Can go down to the autonomic-- one of the first and most robust findings we see is higher peripheral resistance-- greater vascular resistance to circulatory blood flow in our lonely than our nonlonely. Now, this is baseline-- preparing for a speech-- giving a public speech. This is not an interaction. It's a main effect. These are our young adult. It's a main effect. Loneliness isn't increasing reactivity to the stressor.
I, obviously, had expected that's what we would find. That's the way we developed this paradigm. It's their baseline. Walking around, they feel greater threat. If you threaten somebody they show increased vascular resistance, by the way. I should have mentioned that. That's the vascular signature of threat. And, of course, these are young adults. They're normotensive. Blood pressure's going to be a function of vascular resistance and cardiac output. So if I'm going to be normotensive, I have to decrease something. And, indeed, in our younger adults we see a decrease in cardiac output.
The Framingham study has shown that vascular resistance in young adults predicted who would develop hypertension as they age. So we asked, of course, is loneliness related cross-sectionally to higher blood pressure in older adults? The answer is yes. And then the more difficult question-- is it related longitudinally to increases in blood pressure? And the answer is yes, but it takes a while. We actually expected it. It's not going to be year by year. There's too many steps to actually go through.
By year four, we have a significant association. It's there in year three, four, and five by these links. It takes about two years to start to be seen. And then it really reaches an asymptotic level around year four. This is the model in which we fix the coefficients, but if you look year by year, it really seems to reach an asymptote about year four-- about a three year lag. And that's net. All of these are just covariates having to do with normal biological risk factors for high blood pressure.
What about neuroendocrine activity? Well, remember I talked about morning rise and cortisol differing in non-human animals. We see the same thing in human animals. Now, we weren't the first to observe this. Steptoe and colleagues in the English Longitudinal Study of the Elderly-- they put in our loneliness scale and Steptoe-- Andy-- and his colleagues demonstrated differences in the morning rise. What he didn't demonstrate was whether it was specific. And we worry a lot-- everything I've shown you so far is specific to loneliness. We load in all the other psychological covariates and personality-- we still get the difference.
And so we repeated this study to look at whether it was specific to loneliness. And we found of all the measures we took, two variables-- isolation and, interestingly, feel threatened. We don't always get the [INAUDIBLE] here feeling threatened and isolation. We're so highly correlated, we simply put them together. You can-- it's either one. You get the same thing. But feeling fatigued was also related to this morning rise.
Of course, you've probably already guessed-- yes, we did a longitudinal. This is, again, across days. And what we found is the day before perceived isolation is what predicted the next morning rise in cortisol. That morning rise in cortisol did not predict that day's level of loneliness. On the other hand, feeling fatigued was predicted by that morning rise in cortisol. But the day before feeling of fatigue didn't predict that next morning's rise in cortisol.
Now, this makes perfect sense. Cortisol is a big steroid that breaks down carbohydrates and fats to give you energy. That's one of its major functions. And so the fact that the more cortisol you have, the lower the daytime fatigue you feel makes perfect sense. And this suggests it's loneliness not anything else that's driving that morning rise.
Gene transcript-- About 12 years ago, we did our first gene transcript study. It took a huge effort. We were able to do one. Steve Cole is at UCLA. Steve is really the one responsible for this technology that was brought into this study. With these high density arrays now, we were able to do 3,300 gene transcripts on a chip.
And what we found, after making sure there wasn't false discovery, was that 79 transcripts were overexpressed and 131 transcripts were under expressed in our lonely individuals. Now, that alone doesn't tell you a lot except that there's differences in gene expression. This is not genotype. This is given the genes, what's being expressed? This is inside leukocytes, I should mention. So it's an immune cell. And we're looking at gene transcripts within that immune cell, and we get this distribution.
But Steve has also developed a bioinformatics tool called TELiS. And TELiS allows you to look at the transcripts, and treat the transcripts like letters, and determine what words are being spelled by those letters. They're looking at these transcripts and saying what function are these families of transcripts actually performing? So it's a way to do cluster analyses of your transcripts to see, is there anything coherent? If there's no function that they actually spell out-- word that these letters create-- then you've got-- maybe it looks like it's significant, but it's a spurious effect.
What we found-- actually a surprisingly clear story. Remember the lonely individuals have higher levels of morning rise in cortisol. These leukocytes are being buffeted by higher levels of cortisol. The highest level of cortisol, as you probably know, is in the morning. So the cortisol is higher in lonely than nonlonely. These are transcripts that are related to glucocorticoid receptors signaling from the cell surface to the nucleus of the cell. That's what these transcripts are performing, and they're down-regulated in our isolated individuals.
It's a classic receptor adaptation process. Higher levels of stimulation-- the receptor down-regulates. What does that mean that it down-regulates? It means that the receptor activity is actually not getting signaled from that receptor to the nucleus of the cell. That's what these transcripts have done. They've been down-regulated. They're now not being expressed as much as in our non-isolated individuals.
One of the things that's being manufactured in immune cells is pro-inflammatory materials. Now, this is both good and bad. You need pro-inflammatory materials in order to bring appropriate chemicals to sites of abrasion or injury. But if it's overproduced-- if it's unchecked-- then you promote various diseases including cardiovascular disease and diabetes.
What you see here are the transcripts that travel from the nucleus out to produce those pro-inflammatory cytokines-- this NF-kappa betas set of transcripts. And those are up-regulated in the lonely. Why might that make perfect sense? Well, I mentioned that cortisol breaks down carbohydrates and fats in order to produce sugar.
A second critically important function of cortisol-- and anybody who follows baseball knows this about pitchers-- is it decreases inflammation. That's why pitchers get cortisol injections in their elbows when they swell up. It's because it shuts down this transcription process. Well, if that signal is muted, then of course this goes less checked which is why you get this rise in pro-inflammatory cytokines.
Well, finally we looked at whether there are differences in genes in our lonely and nonlonely. And we did that with the help of Dorret Boomsma and her colleagues. She runs the Netherland Twin Register. And she was able to look at over 8,300 family members and twins and determine that we had about 50% heritability of loneliness in adults-- 50% noncommon environmental variance. In adults, there's no common environmental variance. It's all noncommon environmental variance.
There's also one study we've done with Dorret to try to identify at least where on the chromosomes these genes might differ. Now, this is just a first study. We're just gearing up now to do another such study. But at least in this initial study, we found a region on chromosome 12 called q23-24. It's just a locus or a region on that chromosome. And that particular region has been associated with bipolar disorders-- that is, poor regulation of emotions. So we find that to be interesting. And it raises a specific hypothesis about what might incline one to live a lonely life and that is this impaired self-regulatory capacity in the face of emotional stimuli.
There's been just a variety of other replications. Perhaps the most interesting in my reading of this is the one by Wilson and Bennett at West University in Chicago. They actually find loneliness associated with faster progression of Alzheimer's disease, and it's not because of differences in plaques and tangles. They don't know what it's due to, but it seems to be an interesting finding worthy of follow-up.
So let me just try to put this together real quickly. One of the regrettable features of being human or being any kind of organism is that we get the law of thermodynamics operating. We get wear and tear, and so over time there's a decrease in our resilience. The slope, however, can be affected by factors like social isolation. Health behaviors tend to be slightly poor, especially as you get older, and that speeds, if you will, the dissent in terms of your less resilience.
Exposure to stressors-- young adults-- we don't get a difference in exposure to stressors. By the time we look at our older adults, we do. Lonely individuals are in more stressful circumstances, and of course that's going to contribute to a steeper slope in this decline as a function of aging. Social isolation is related to perceived stress. I showed you even in our young adults-- even when the events are the same as measured by experience sampling methods-- our lonely individuals find stressors to being more intense and uplift to be less uplifting.
We also find an interesting difference in coping I didn't delineate for you. But very, very briefly-- nonlonely individuals engage in more active coping. The lonely individuals-- more passive. It fits this notion that it's everybody against them. Nonlonely individuals are also more likely to seek emotional and social support from other people than are lonely individuals.
The stress response itself-- stress response isn't what Selye and Cannon suggested. It isn't actually uniform and homogenous across everything. We've shown some pretty clear, specific stress responses-- loneliness related, for instance, to vascular but not myocardial activity in the cardiovascular system. It's related to cortisol but not catecholamines. So it seems to be related to very specific aspects of the stress system. But we do get exaggerations of those aspects, and of course those contribute to a faster descent.
And finally, these are all catabolic processes. The anabolic processes themselves seem to differ and most notably sleep-- with the sleep amount in young adults and older adults being comparable. It's not the hours in bed. It's not even the hours of sleep. It's the efficiency or the salubrity of that sleep that seems to differentiate. So even for equally toxic days, the lonely individuals aren't detoxifying those days as completely as the nonlonely individuals.
To put it in an evolutionary context because I've been hinting that that's the story I'd like to be telling here and why it tells us, I think, something about human nature-- it's basically a simple Darwinian explanation with a twist. Genes that promote behavior which increase the odds of the genes surviving are perpetuated. This usually means the organism is also more likely to survive to procreate.
So here I have sardines being predated by a swordfish. They create this really dynamic looking fish ball. And it looks like it's chaotic, but you can program and see the same thing on computer. Just program a lot of little dots, and program them all to swim to the middle, and you'll get the same thing. A very selfish behavior-- swim to the middle where they're safest and you have this collective action. But it makes perfect sense. Dawkins would love this with the selfish gene notion producing selfish organisms.
But we're not sardines. In fact, even penguins aren't sardines. Most of you have probably seen or at least heard something about march of the emperor penguins. Or if you go online and read about emperor penguins, you find that they did a reasonably good job. Penguins-- these are male penguins-- create huddles. So male and female penguins mate. They march an ungodly distance into the middle of the Antarctica where the temperatures get to be minus 75 and wind speeds get to be 100 miles per hour.
The mothers birth the egg-- moves it with some reluctance to the father's feet, places it under his abdominal fat. The mothers go back to the ocean where they can feed. The fathers sit in these huddles to try to survive this godforsaken winter. The chicklings-- baby penguins-- hatch. They remain under the abdominal fat because they'd freeze in a microsecond. And unlike the sardines, these penguins apparently cooperate allowing the ones on the outer rim to rotate to the inside. They spend a lot of their time sleeping, but they don't resist. They, in fact, allow this movement in and out of the outer rim.
Now, if you think about it, the penguin is not the element of survival here. It's not the vehicle for genetic transmission. These have an abject dependent period of dependency of about six months. And they need the collective. They need the huddle as a vehicle for genetic transmission. It's not their ability to reproduce. For a species with long periods of abject dependency, it's not the ability to reproduce that's sufficient. Is the ability of their offspring to have children.
So it's your ability to have grandchildren that determines whether your genes make it into the gene pool. If they were to do the sardine game, the outer rim would die, then the next rim would die, then the next rim would die, and very few if any would actually survive. They need that group to stick together to have the warmth necessary to survive the long winter. So the vehicle goes from being the organism to, in this case, a very simple collective.
Humans have more complex collectives. But if we were to be in evolutionary time marching around hunting together because we're more likely to kill big game, and I can't eat it all anyway so it's no loss to me to work with you-- that you can see adaptive, selfish, collective action. But most times, we wouldn't manage to kill that big game. We'd be barely foraging and surviving on the little that we could find. Why share? Why go to risk for the help of others? If, in fact, you did not do that, you might well live long enough to procreate much more than those who were by some odd mutation willing to sacrifice oneself to try to help others.
But your genes are not going to survive because those offspring that you do create aren't going to have the protection needed. So the vehicle, again, starts to not be the individual. Because of the long abject dependency, it's these odd collectives. And so you get a different mode of structure. With that you also have to have some aversive signal, and that's essentially what we think loneliness is.
And we think loneliness is misconstrued. Men are less likely to say they feel lonely than women, because there's this stigma. I would like to suggest to you that if hunger had the same stigma as loneliness in this culture, we'd have a lot more anorexic looking people walking around not wanting to admit that they, actually, were hungry.
And with that I'd like to actually show you all the people who helped in this research. It's really a great group. And the persons whose name you will see most often on these is Louise Hawkley. She was a biopsychology graduate student of mine at Ohio State. She started this work when she was a graduate, and then stayed with me since that time, and is really just a wonderful colleague and friend. So I'd like to thank her and the rest for this. And thank you for your attention.
SPEAKER 1: This has been a presentation by Human Development Outreach and Extension at Cornell University.
We've received your request
You will be notified by email when the transcript and captions are available. The process may take up to 5 business days. Please contact firstname.lastname@example.org if you have any questions about this request.
Watch the full presentation with speaker footage and slides.
University of Chicago psychologist John Cacioppo provides a fascinating overview of his research on how social isolation or perceived social isolation (loneliness) effects social cognition and emotions, personality processes, the brain, biology, and health.