For anyone familiar with how things work in America, it shouldn't be surprising that money and sleep go together — for the most part. If you take a quick glance at national health surveys, you might come away with a clear portrait of the poorly slept American: low-income, under-educated and non-white. Steep yourself in the data, however, and that portrait starts to blur.
At this point, we know that poor sleep is linked to all sorts of negative outcomes: diabetes, less occupational success, psychological instability, lower levels of happiness and worse overall functioning, to name a few. And we know that certain groups — particularly those at the low end of the socioeconomic ladder — suffer the plight of crappy sleep at much higher rates than others. But the clarity stops there.
To wade through the fog, we reached out to Dr. Michael Grandner, a psychologist at the University of Arizona who studies the relationship between sleep patterns and health, as well as the environmental, behavioral and social factors that influence these patterns. Based on years of crunching sleep numbers, Grandner has found that it's not income itself that leads to changes in sleep, but what income buys — medical appointments, tuition, nutritious food, single-family houses, therapy — that makes the difference. To improve sleep, Grandner believes, we need to stop framing the quest for shuteye as an issue of personal choice. Because, for the people who need a sleep upgrade the most, it just isn't.
We've seen an increase in efforts to spread sleep-health awareness. In a lot of cases, the private sector is fueling these efforts, and they're aimed at corporate workers who can't seem to disconnect. You're more concerned with people who face less optional barriers to sleep.
What makes it very frustrating is that a lot of people seem to think that if we know that getting seven-to-eight hours of sleep is important, what we should do is just tell people to get that much sleep. Tell that to someone who’s working three jobs and has to get up at 5 o’clock in the morning to catch the bus, and has three kids. It’s more complicated than that. The problem isn’t entirely that people don’t want to sleep. A lot of times it’s that people can’t sleep. What we need to do is understand the social and environmental context in which sleep exists so we can understand what to do and how to solve this problem instead of just pretending that this is an issue of personal choice.
Throughout your work, you've pointed out a number of paradoxes. For example, there's the “immigrant effect," which shows that low-income minorities tend to sleep poorly but that immigrants (who are often low-income minorities) sleep quite well, statistically.
It’s interesting that you bring that up — I literally just came out of a meeting [about this issue]. I’m trying to get funding to go down to the US-Mexico border and understand what’s going on. Most obviously, there's the possibility of measurement error. If you ask someone for sleep complaints outside their native language, then they might not be on the same page about what you’re looking for. It’s not that they don’t understand, but rather that the concept might not translate exactly as you intend it to. Another issue that comes to mind is something called “demand characteristics”: You’re not just asking someone on the street. It's a research study and different cultures have different ways of responding to that environment. Someone might feel pressure not to be a complainer.
Also, we know that the more Americanized you get, the more unhealthy you become. It's something we see over and over. We start exporting culture around the world and all of the sudden, people start getting fat and having more diabetes and heart disease. There are benefits to the typical American culture, but one downside is that our culture is not very healthy.
And immigrants might bring other things with them from their culture, like, for example, a strong sense of family cohesion, which is very different from Americanized individualism. One "pro" to having traditional social roles migh be more balance — i.e., less conflict in terms of people having to do too many things. So could there be things that are protective that they’re doing? Yes, but also, it might be that we're not seeing issues because we’re not asking the right questions. So, we need to figure out which is which.
You also identified an education-sleep paradox when you looked at people with low levels of education and separated out those who didn't finish 9th grade.
Right. It wasn’t that the more education you get, the better off you are in terms of sleep. The finding was if you have a decent education — high-school or some college — you’re more likely to be a short-sleeper than someone who has a college degree (or more). But if you never finished high school, the pattern falls apart. This group is not full of short-sleepers. I would say something similar as I would about the immigrant issue: Who are these people? Who doesn’t get a 9th-grade education? That’s not all that common in the US. One thing that was interesting [in one paper] was the gender and race interaction. Among Hispanic/Latina women, especially, less education was protective in terms of sleep. It could be that people coming from a more traditional Hispanic background have more traditional gender roles. So, a lot of the women don't work and don't need more education. At this point, it’s a guess, which is why I want to dig into these real-world situations.
You found food insecurity — lack of access to enough nutritious food to keep the whole family healthy — to be the most direct predictor of poor sleep. Is that because it's such a strong economic indicator?
I was really interested in food insecurity because it seems to represent a real intersection between metabolic risk and socioeconomic risk. It just goes to show that the people who are struggling to put healthy food on the table are not sleeping. There’s no answer here, but it poses a question I’d love to answer: Does this represent a vicious cycle where you have stressed-out people who can’t maintain health, who then engage in behaviors that make it even harder to maintain health?
In a lot of cases, two people can have the same poor sleep habits but different health outcomes. In parsing public-health data, you see these sorts of unexplained disparities a lot.
Right. A lot of people assume that if you take two people from different backgrounds, then their relationship between sleep and health should be the same. And, if it’s not, then the reason is genetic. Honestly, there's very little evidence that these differences are genetic. It's more likely that they're environmental.
When you look at white people in the US vs. black people in the US vs. Hispanic people in the US, the systematic differences between those groups are mostly not genetic. There’s more variability in skin color among the groups than between them, and if anything, skin color is probably not all that important to the outcomes we’re interested in anyway. The only genetic factors that would be important, for example, are more like facial structure. There’s some data showing that certain Asian groups have a facial structure that predisposes them to sleep apnea at higher rates, for example. But otherwise, you need to look at attitudes, beliefs, behaviors.
Let's talk about Appalachia. You identified insomnia "hot spots" and "cold spots" across the country, and the Appalachian mountain region emerged as a hotbed for, well, hot spots. The demographic profile of the average hot-spot resident was confusing to me: Low-income, white and with health insurance.
The way the algorithm was set up meant that hot spots weren't necessarily areas where there’s a lot of insomnia, in an absolute way, but rather counties where complaints of insomnia are not just high, but unusually high in comparison to surrounding neighborhoods. If you notice, the data’s from 2009 — and the paper wasn’t published until 2013. It took so long to publish because I was trying to figure out why the analysis turned out this way. What is up with this area? I reached out to researchers who study geographic patterns in health and focus on Appalachia. I'd show them the maps and ask them, “How do I make sense of this? Why here? Why is nowhere else like this?”
Everyone seemed to have a different idea. But, one story emerged: [Appalachia] is a very poor area, but it’s not the only poor area. It has high rates of mental health issuess, substance-abuse issues and cardiovascular disease, but it's not the only place with [these high rates]. But, it does seem to be this perfect storm of all the problems of which sleep could be a marker — between the abnormally low socioeconomics and abnormally high amounts of poverty, poor health and psychosocial stress. It just might be that sleep is a main common denominator underlying things as broad as poor mental health and increased stroke risk. If you drew a venn diagram around the hot spots of depression and cardiovascular disease and alcoholism and substance use and all these other things, a lot of them may overlap in the area of sleep, and the hot-spot analysis picked up that overlap. And that’s the best idea that I could come up with.
If I recall correctly, you didn’t find any link to race or ethnic minority status in cold spots or hot spots.
That was somewhat surprising, especially given that when we looked at state-by-state insomnia differences, racial makeup of the state predicted, to some degree, variability in sleep complaints. But, the biggest hot spots weren't necessarily high minority areas. And there are places in the US that are very high-minority areas that were not hot spots. Assuming the findings get replicated, this tells me that race/ethnicity might play more of a role at the individual level than it does at the county level. So a black person living in a white county vs. a black county. Maybe the degree to which you are different plays a role in your sleep.
You looked at other contextual factors and found that perceived racism among middle-aged adults was associated with nighttime worry, and then you referenced another study about the racism-poor-sleep link in a healthcare context.
Racism is a really unique psychosocial stressor because it leaves people feeling powerless, which seems to affect sleep. Feeling like [the target of racism] isn’t something that anyone did. It's a condition to which they're predisposed and it bears out in their health, in ways over and above depressed mood. So, in the study we did, if you parcel out any effect of depressed mood, you still see a relationship between perceived racism and crappy sleep. These people are not sleeping well, even after you adjust for socioeconomics and you adjust for negative mood issues. If i remember correctly, people who perceive racism in healthcare settings (because that was the question that was asked on this particular survey) were 60 percent more likely to have sleep problems at least half the time.
Insufficient sleep is different from insomnia, which describes an inability to initiate or maintain sleep despite having the opportunity to do so. Some health surveys measure insomnia symptoms, whereas others focus on sleep duration, or other relevant metrics. How does the use of slightly different questions affect study outcomes?
The gold-standard way of measuring insomnia is to give someone a sleep diary, have them track their sleep for a week or two, and then do some math. But you can’t really do that in the context of a survey, so we try to come up with these shortcuts that might not be effective or accurate. One reason that sleep deprivation (as a measure) has been a little more successful in epidemiology is because the question is more straightforward. With insomnia, usually the question is “how often do you have difficulty?” Okay, but what does "often" mean? What does "difficulty" mean?
In that paper looking at sleep complaints, not duration, one thing you’ll notice is that when you ask the question “how often do you have difficulty falling sleep,” if you’re a minority, you are way less likely to endorse that item, across the board. But when [the health survey] asks “how long does it take you?” you get a very different result. Asking for a complaint is probably not the right thing to do, if you want to generalize. That’s what signals to me that some of these immigration issues might be a measurement problem. We might be asking the wrong questions. But, I know there are movements to come up with better questions.
You mentioned one study that found black residents in the NYC metro saw sleep apnea (the sleep-breathing disorder marked by snoring) as a type of insomnia, a natural part of getting older and/or a condition caused by bedtime activities. Can you explain these misconceptions?
Twenty years ago, no one had heard of sleep apnea. Now, most people have heard of it and know someone diagnosed with it, but people still don’t know what it is. I talk to people on the street who have sleep apnea — and are being treated for sleep apnea — and still don’t understand what’s going on. People still think they’re going to suffocate, and that the doctor told them they’ll have a heart attack because of a lack of oxygen, and it’s not because of a lack of oxygen – they’re going to wake up, but it’s the intermittent hypoxia and fragmentation and oxidative stress and all of these things no one ever talks about, and people say, “Oh, why wasn’t I told that?” Well, mostly because, under the current [healthcare] model, reimbursement comes from running a diagnostic test, not talking to people and understanding what’s going on. [Doctors] are paid to give patients machines, not help them learn how to use them.
People are more likely to adhere to treatment when they understand their condition, and whether or not they adhere to treatment predicts their health. Sleep labs are closing down all over the country because reimbursement’s been made tougher. So if you’re stressed about keeping your doors open, you might not sit there and talk with your patients about what’s going on, but you need to. Unfortunately, it's not incentivized.
Shouldn’t there be better sleep-health education at a basic level? Should primary care providers assume more responsibility?
Yes and no. Yes it would be great, but primary care providers are already stretched so thin. The number of sleeping pill prescriptions in the US is ridiculous, unreasonable and, frankly, a problem, because the data show that sleeping pills should not be the first-line treatment for insomnia. They carry inordinate risks and because of that, you should do other things first, but no one’s ever heard of CBT-I (Cognitive Behavioral Therapy), and they don’t know anyone that does it. Primary care providers certainly can’t do it — there’s no time for that. And it's much easier for them to write an Rx because they get reimbursed and it’s easy, so what’s the incentive to do anything else?
But, it's a complicated question. Should more sleep medicine be put on primary care? In the real world, primary care doctors should ask about sleep at every visit. That should be part of the review of systems: "How are you sleeping?" I can’t think of one medical specialty for which sleep, and sleep disturbance, is not relevant to understand.
Right. Should “how are you sleeping” and “do you have a history of sleeping issues” be questions that are part of every medical checkup, across specialties?
Well, I have two things to say about that. Number one, my personal opinion is yes. Number two: We take temperature as a vital sign. When is temperature relevant? If it’s high, it’s important. But if you’re going to an endocrinologist, your temperature is not relevant to your diagnosis. Blood pressure is not relevant for a lot of different office visits, but it’s seen as important enough to do.
I would argue that sleep quality is at least as relevant as either of these. And there’s data to back that up. We now ask about smoking at every visit, because it has such a wide-ranging impact on health. It doesn’t matter if you’re going in for a musculo-skeletal visit; the doctor’s supposed to ask about smoking, because it's so imporant. And I think primary care providers should ask about sleep, and get the training to do so, for the same reason, but with the recognition that getting more things for primary care to do might not be the answer. It's more about giving them options. If all they know is how to write a prescription, that’s what they’ll do. But if they know about other things out there, they might be able to do something else.
Going back to CBT, in January, the American Academy of Sleep Medicine announced a mobile platform for sleep telemedicine. This seems like a great way to make CBT available to people in rural communities, and more affordable to everyone. But, I've found that people who know about CBT tend to be those who could afford to visit sleep-medicine specialists anyway. It doesn't seem like CBT is reaching populations for whom effective, cost-effective and virtually accessible treatment might make make the biggest difference in quality-of-life.
It’s a shame. It’s hard for me to think of a treatment modality that is this well-supported by the scientific literature that's so poorly taken up by the clinical establishment. So, why is that? There’s no conspiracy, but people don’t know about it. There aren't a ton of practitioners out there because there aren't a ton of referrals. There aren't a ton of referrals because doctors don't know a lot of people to send patients to. It becomes this chicken-egg problem. And, the other thing is, a lot of the time, CBT is practiced by psychologists. And right now there’s a crisis with reimbursments for psychologists, where insurance companies don’t want to pay for doctoral-level providers when they can pay for something less.
But it takes someone with more experience and a higher level of training to do [CBT] successfully in complex patients, so you run into this problem where being able to do [CBT] and also get paid for it is a nightmare. So, just today, I had a referral from the department of medicine for a patient with insomnia. Their insurance covers all their medical visits. They’re referred to me and they’re going to have to pay the full price [of therapy] in cash. Same hospital, but ...oh, that’s not real medicine. We can’t pay for that. So, it’s not covered. What’s funny is that their insurance probably covers things like chiropractic services but it won’t cover [insomnia treatment] because, since I’m a psychologist, this counts as mental health, and as a country, we haven’t decided that mental health is worth treating.
This story has been updated.