Skip to main content

The Science Behind Sleep Health with Minjee Kim, MD

In this episode, Minjee Kim, MD, shares results of recent studies she has conducted to better understand the role of inadequate sleep on long-term metabolic health in pregnant women and middle-aged adults. She also discusses a new AI-based intervention that could bring a better night of sleep to those who need it. 

Recorded on April 11, 2025.  

 

“What happens in these early parenting years? And if people either develop a new inadequate sleep or continue to have inadequate sleep then would that affect their long-term metabolic risk?” 
— 
Minjee Kim, MD 

  • Associate Professor of Neurology in the Division of Neurocritical Care 
  • Neurologist, Northwestern Medicine 

Episode Notes 

Through her work in neurological intensive care, Kim developed an interest in the biological necessity of sleep, understanding sleep's role in brain function. She then embarked on a series of research projects focused on how lack of adequate sleep and disruptions to natural sleep patterns could contribute to long-term health issues. 

  • While sleep is often studied in older adults, Kim believes that the early and mid-life stages, particularly pregnancy and the years following, should be studied as well because she believes these are critical periods for sleep-related, long-term health outcomes. 
  • Kim shares insight into how modern lifestyle, including excessive artificial light exposure, interferes with our natural sleep-wake cycle and how this contributes to long-term health issues.
  • Her recent study published in JAMA Network Open suggests that persistently short sleep during and after pregnancy may be an important factor associated with the development of cardiometabolic diseases. 
  • Another study published in the American Journal of Obstetrics & Gynecology Maternal-Fetal Medicine focused on the association between light exposure before bedtime in pregnancy and the risk of developing gestational diabetes mellitus. 
  • Kim also introduced sleep as a new component in the MidCog Study, a Northwestern University cohort study focused on identifying modifiable risk factors for cognitive aging beginning in midlife.  
  • She is now focused on using new AI-driven methods to predict and flag patients with sleep problems using electronic health data. In a pilot project, this effort has already led to meaningful diagnoses and treatments in patients who may not otherwise have been flagged for sleep problems, such as sleep apnea. 
  • In the long-term, Kim says her goal is to make healthy sleep more accessible and achievable for everyone — helping people "sleep like it’s 1825," a time when natural sleep patterns better supported human health. 

Additional Reading

[00:00:00] Erin Spain, MS: This is Breakthroughs, a podcast from Northwestern University Feinberg School of Medicine. I'm Erin Spain, host of the show. We all know how bad you can feel after a poor night of sleep, but research shows that the cumulative effect of night after night of short or disturbed sleeping is more than irritating. It has health consequences. My guest today. Dr. Minjee Kim studies the role of sleep disturbance and cognitive impairment, Alzheimer's disease and related dementia and aims to design healthcare interventions for the early detection and treatment of sleep disturbances that could lead to longer, healthier lives for all of us. Dr. Kim is a Northwestern Medicine neurologist and an associate professor of neurology in the division of Neurocritical Care at Feinberg. The research in her lab is driven by the belief that quality sleep is a fundamental human right, and that its preservation is essential for healthy aging. We welcome her to the show today to talk about her research and innovative ways. She's trying to bring a better night of sleep to those who need it. Welcome to the show, Dr. Kim. 

[00:01:16] Minjee Kim, MD: Thank you for having me. Such a pleasure. 

[00:01:19] Erin Spain, MS: Let's start with the basics. More than half of adults do not get sufficient sleep. What constitutes as sufficient sleep and what are some of the top reasons why so many adults are not getting enough? 

[00:01:31] Minjee Kim, MD: We still don't know exactly how many hours of sleep is sufficient for an individual. Studies of lots of people, this epidemiologic studies have characterized sort of how many hours people generally sleep in different life stages. How short of a sleep is bad for various health outcomes, and the societies have come with a guideline to say six to eight hours of sleep is probably what most people need in midlife and into the older age. That being said, certain people who I call the lucky people, don't need to sleep that much, and they can function just fine, and we found some genetic polymorphism associated with that. So it's not universal that you need exact number. Of hours of sleep per night to carry out a healthy life. But I would say generally six to eight hours are probably a good number for most people. 

[00:02:31] Erin Spain, MS: So it's not just hours that are important too, but the quality of your sleep. 

[00:02:35] Minjee Kim, MD: Absolutely. Yes, so people have focused on sleep duration for many years, and a lot of research have come about the short sleep or longer sleep duration and what is good and what is bad. But, now people are recognizing that sleep is not just about the duration. The construct of this multidimensional sleep health is rising, to really recognize the various aspect that constitutes sleep, right? So good sleep, what would you think a good sleep should be, 

[00:03:08] Erin Spain, MS: Oh, like you don't wake up several times in the night. You feel rested when you wake up. 

[00:03:13] Minjee Kim, MD: Yes. Yes. So sleep should function to maintain an adequate daytime alertness, right? You should feel good when you wake up, should be refreshing and it should be restorative. All of that should be enough time also should be consolidative that you should not wake up too many times at night. And the other thing that we often forget is it should be regular from day to day. You should probably go to sleep around the same time and wake up around the same time. So you know, there's a weekend catch of sleep you can sleep in. It's probably not a good sleep health. So now we are thinking about these various aspects, alertness regularity, satisfaction, duration, and timing to define a good sleep health. 

[00:03:59] Erin Spain, MS: Tell me a little bit about your background as a neurologist and what led you into sleep and now doing these studies on these different age groups and different folks. 

[00:04:08] Minjee Kim, MD: We spend about a third of our lives sleeping and sleep isn't unique to humans. Virtually all animals have some form of a sleep wake cycle, even though being asleep can make them vulnerable to predators and other natural risk. That's always intrigued me. Like, why would such a risky behavior be so universal across species. That sleep must serve a fundamental role in our biology. And we now know that critical processes occur during sleep, including memory consolidation, clearance of metabolites that can be toxic to brain cells. So sleep isn't just passive rest, it's an active work that's essential for healthy brain and bodily function. I work in the neurological intensive care unit. Caring for patients with severe brain injuries or those experiencing brain dysfunction due to failure of other organs like the heart, the liver, the kidneys. And in this setting, even the definition of sleep becomes murky, some of my critically ill patients are in a coma. They're not awake, but they're also not sleeping in the traditional sense. There are no characteristic brainwave forms and we can't wake them up. The injured brain appears to lose its ability to maintain a normal sleep wake cycle or even generate the state of sleep at all. So we don't yet fully understand whether this breakdown in sleep is just a byproduct of illness or whether it's part of brain's adaptive response, maybe even a mechanism. Interestingly, patients in a coma who do show sleep like activity, brain activity, often have a better chance of recovery. That suggests that sleep or the ability to generate sleep may be important, not just for daily functioning, but also for healing and resilience. But zooming out to a broader view, outside of my clinical work. I think we are facing increasing challenges to maintaining good sleep in modern society. It's probably much harder to consistently sleep seven to eight hours a night in April, 2025 than in 1825 when people use candles to get around and didn't have the constant distraction of emails and, uh, streaming content. So I think in modern society many aspects of the life degrade the quality of our sleep across the lifespan without us actually realizing the harm. And I believe that cumulative toll impacts our long-term health. So that's how my research is driven now, to find ways to improve sleep health in everyday life, to promote overall wellbeing and healthy aging, cognitive aging and aging in general. 

[00:06:56] Erin Spain, MS: Oh, that's fascinating. And you mentioned light and you mentioned candlelight versus the glow of a smartphone, this something you studied in several different populations. Tell me about this work. One study you looked at pregnant women, and this was published in the American Journal of Obstetrics and Gynecology, maternal Fetal Health. Another study was published in the journal Sleep where you looked at light exposure on middle aged folks. Light exposure, there is some impact here on these populations. Can you tell me about that? 

[00:07:25] Minjee Kim, MD: I'm following really the footsteps of giants in the field who've made major discoveries about the circadian clock and how the circadian clock is important and what is happening in modern society that's really affecting the natural circadian rhythm in our body. So just to go back, the circadian clock it exists in all cells of our body. And not just humans, but all animals. So there is a natural light, dark cycle that's driven by the sun and the moon to some degree. And we've all evolved to function optimally in that natural setting. With this artificial light and other stimulus, we are getting less of that natural signal that drive and synchronize our circadian clocks in the body. So the beliefs are that we are not getting enough signal and there are other random, non-natural signals that are affecting our circadian clock. So it's getting desynchronized and they're not optimal to maintain the body's functional homeostasis. So that drove some of my research into the light exposure and metabolic dysfunction in people. It is really hard to disentangle the light dark cycle, the circadian rhythms and the sleep wake cycle because they're somewhat associated, right? Then people always thought, okay, the light before sleep is going to affect your sleep and that's why it's bad for your health, and we actually also assume that, but some of this research we've done seem to suggest otherwise. So go back to the light exposure in older adults. This was a cross-sectional study who were recruited a long time ago in their twenties, and they were followed many decades later, now in older age and looked at what kind of chronic disease they have developed and the sleep patterns that they now had. The caveat of that was we didn't know how they slept before. And here we looked at the light exposure during the sleep period. And older adult who had any light actually during their sleep period were much more likely to have conditions like diabetes, high blood pressure, and we can't infer any causal relationship there. And it could be that people with those conditions tend to wake up more and they turn on the light. But that was a one study that started this journey into the light and health research. 

[00:10:03] Erin Spain, MS: Hey listeners, if you're enjoying this episode, please share it with a friend and follow us on Apple Podcasts or wherever you listen to podcasts. Now back to the episode. 

[00:10:13] Minjee Kim, MD: Then thanks to all the collaborators at Northwestern, I had access to this amazing data set of pregnant women who were recruited between 2010 and 2013 during their first pregnancy. And here we looked at their sleep patterns and how they were exposed to light, just like in the older adult. I was initially interested in the light exposure during sleep and how that's bad for them. One major metabolic dysfunction that occurs during pregnancy is gestational diabetes, so your insulin resistance is impaired and that can lead to many adverse outcomes lifelong. So it seems natural to look at sleep patterns and light exposure in early pregnancy and the risk of gestational diabetes, which happens in a second trimester. But luckily for these participants, a lot of these people had no light exposure during sleep. So I say, oh, that's good, but then what happens during the day? Right? Like, you know how much of light should we get at what time of the day? And we still dunno the answer to that, but if we emulate what happens in nature, then we should probably get a lot of light during the daytime, especially early hours and less after sunset. So that's how we thought about okay then, let's look at the hours just before bedtime, because in 1825 there's no light, right? It took a lot of effort to get any light so that, but now it takes a lot of effort to turn off the light. So if you get a lot of light before sleep, is it going to affect your sleep for one and two, your risk of gestational diabetes? So then we didn't know how many hours is enough, so we looked at one hour, two hour, three hour before each night bedtime. And it actually seemed to give pretty consistent message. So we picked three hours, which seemed reasonable based on the guideline of defining the evening hours and women who were exposed to most light in this period, regardless of the light exposure rest of the day seemed to have an increased risk of developing gestational diabetes even after adjusting for potential confounders. But these women actually slept better. So that was very strange. And I remember just looking at the data like, am I making mistake? Are we doing this right? So they were sleeping well. So there was this discordance between the metabolic risk. And then the traditional belief that the effect of light exposure is through the effect on sleeps. It wasn't the case. So it seems to have an independent, adverse, effect on the metabolic regulation, at least in this time period. 

[00:13:01] Erin Spain, MS: And it's important to note, you're capturing this light exposure through kind of like a fancy Fitbit or Apple Watch. Is that right? That they wear? 

[00:13:08] Minjee Kim, MD: That's right. So if we could choose, we would put a light sensor right here next to our eyes, because that's how our brain sees the light, and that's how our circadian clocks perceive the light. But we still don't have a device that's convenient enough to do that. So we capture light at the wrist level as our participants to wear this let's say fancy version of Fitbit for anywhere between seven to 14, sometimes longer days, and capture the light exposure throughout those periods. 

[00:13:42] Erin Spain, MS: So that's interesting. It kind of opens up this whole new world of not only do you have to worry about your duration and quality of sleep, but then your light exposure during the day and before bed. So lots to explore there. You know, what practical steps can people take to reduce their risk of light exposure at night? 

[00:14:00] Minjee Kim, MD: I would say just be mindful first of all of what lights are around you, and I'm actually gonna share this constant kind of battle with my husband. I have two young children, there are three and seven. There are more research coming out demonstrating how sensitive children are to this evening light. So to adult we generally say dim light less than 5 to 10 watts is probably okay. Recent research suggested children even at a very dim light. Their melatonin goes now it's not good. So then it's very hard to turn up all the light because night light we have stairs and the safety. So we try to dim most light and then turn up screen for one. I actually have glasses that block blue light. The evidence is, you know's, coming up, it's we'll see it's jury's still out there, but at least that's something that I am doing that I think people can easily adopt. It's not as easy to reduce the light level at home, but you can turn off the devices. And, you know, turn up the blue light mode and then use warmer light nightlight if you have to. Just dim everything and be mindful of it. And, you know, those glasses are inexpensive and it probably wouldn't hurt. I don't know if there are benefits yet, but the jury's out there hopefully will have more evidence on that. 

[00:15:25] Erin Spain, MS: So there's been two studies recently you've done with either pregnant or postpartum women looking at metabolic health. Tell me about that population and why it is that you were so interested in studying them . 

[00:15:36] Minjee Kim, MD: Generally in the field, a lot of studies have been done looking at the link between sleep and cognitive function and aging. What people are realizing now is that it might be chicken and egg situation. Poor sleep likely leads to poor brain function, but the poor brain also cannot generate good sleep. So then are you actually looking at the byproduct of a brain changes and calling it as a risk factor because these changes happen at the same time? So then can we do something about it if it's actually in your brain unable to generate good sleep? So the field is moving to an earlier life stage, midlife. And for women specifically, the challenges are unique to them, including pregnancy and menopause. Right? We know that short sleep duration is linked to many adverse health outcomes. Cardiovascular disease called decline, even mortality. And this relationship seems to vary by age and sex. Men and women separately and also cross life stages. So that's how I was interested in this cohort study that followed the pregnant women. Who were pregnant between 2010 and 2013, it's ongoing now, ten year follow up. Now they're entering middle age and they have been characterized for all kinds of biological and psychosocial changes. So then can we look at some of the changes that happen in their early or just before the midlife. And there must be something that can be intervenable. The previous research have focused on pregnancy and menopause, but there have been gap between those two life events. There are parenting years, it's hard to maintain good sleep. There's so many challenges, and I may even say that sometimes harder than the pregnancy itself and the people also have multiple pregnancies. 

[00:17:25] Erin Spain, MS: The middle of the night feedings and bad dreams, and getting up to tend to children, that can be very disruptive. 

[00:17:30] Minjee Kim, MD: Very disruptive, right. So that the research that have been done in the immediate postpartum period, six month, one year, even up to two years and how women sleep, even feeding schedule can affect overall wellbeing. But then the study really drops out. We don't know what happens after that when most children would sleep through the night, but still there are challenges. So that's what we were interested in, like what happens in these early parenting years, and if people either develop a new inadequate sleep or continue to have inadequate sleep then would that affect their long-term metabolic risk? So we leveraged the data set of pregnant women who were enrolled during their first pregnancy and followed two to seven years after the index delivery. And about 4,000 women reported both sleep duration during the pregnancy, and again, a follow up that was about 3.1 years of interval The median sleep duration decreased from eight hours to seven hours, so people sleeping less. About quarter of them had less than seven hours of sleep during pregnancy, which then rose to 35% at follow up. But there was significant shift. Some people now have new short sleep and some people have resolved short sleep. So there are changes, and what we found was that about 13% of the participants developed a metabolic syndrome at follow-up, which is a constellation of conditions like high blood pressure, dyslipidemia, obesity, and insulin resistance. And this is thought to be a significant risk factor for cardiovascular morbidity and mortality later. And people who had the persistent short sleep, meaning slept less than seven hours, both during pregnancy and at follow up, were at significantly higher risk of developing this metabolic syndrome. Just over three, little over three years of follow up. And when we define the short sleep as less than six hours, so it's more severe short sleep, the risk seemed even greater. Now, having short sleep at either time point or both time points were all associated with the greater risk. So there might be a dose dependent relationship between short sleep and metabolic risk. I think this is important to know because metabolic syndrome is considered a modifiable risk factor. So this might be the time to intervene before these women go on to develop a full-blown cardiovascular disease. Also interesting is that other longitudinal studies have shown that sleep duration is relatively static across midlife. So people who sleep little, continue to sleep a little for 20 years from age forties and on. But we saw a lot of shift in sleep duration. So this is a time that people are vulnerable to establishing suboptimal sleep habits and maybe we should do something about it. And the other thing is that the burden of inadequate sleep duration was not uniform across racial and ethnic groups. The non-Hispanic Black mothers were much more likely, almost twice as likely to develop unusual short sleep and more than twice as likely to continue to have short sleep at follow up. We don't know what's driving this, there are probably multiple factors including structural racism and other things that we can address in different ways. So this is going likely to expand and widen the health disparity over the lifespan. So think the next step for the field is really to think about developing an intervention in this life stage that's culturally appropriate that can modify the long term cardiovascular and also that leads to neurologic risk over time. 

[00:21:16] Erin Spain, MS: What do you that could look like? 

[00:21:17] Minjee Kim, MD: Even something simple like awareness and counseling of sleep habits. Turn off the light and try to maintain consistent time for bed. So if you think about I need to go to bed at nine o'clock to maintain seven hours of sleep, then start turning off the light at seven and then winding down at eight so that you are more likely to fall asleep at nine. So just be mindful of that, not like you constantly doing something up until nine o'clock and I gotta go to bed. It's much harder to turn that off. 

[00:21:46] Erin Spain, MS: So you're currently working with other Northwestern investigators on the MidCog study. Explain this study to me and how it contributes to your research on sleep and cognitive function. 

[00:21:56] Minjee Kim, MD: Yes, so my mentor and collaborator, Dr. Michael Wolf, is a phenomenal researcher and he has carried out a longitudinal cohort study of older adults over almost two decades. We've learned a lot about what happens in older age, but now we are realizing that a lot of risk factors actually start developing in earlier life stage in midlife, as we talked about earlier. So this is his new cohort study, that is looking at potentially modifiable risk factors of cognitive aging in this lifespan.When I started talking t o him back in 2021 we clicked because I've always been interested in sleep, but not in this general population, and he's carried out successfully this cohort studies for many years, but he has not investigated sleep. So I say, let's collaborate on this. Let's add a sleep aspect to this. I think this is a major gap. So this cohort study recruits about 1200 people from the greater Chicagoland area from primary care. And we are characterizing potentially modifiable risk factors, including physical activity, diet, and chronic conditions, vital signs plus sleep and rest patterns. We're asking them to report their sleep quality using multiple validated questionnaires. We're also asking them to wear this fancy version of Fitbit to characterize their rest activity patterns. What we are trying to achieve with this cohort is to better understand what happens in midlife. Like how do people sleep? There are major gaps in that knowledge and what aspect of sleep is particularly bad for your health, not just now, but also at a lot later life stage, that we can do something about it. 

[00:23:54] Erin Spain, MS: And you're leveraging AI and the electronic health records as well, to develop tools to maybe detect people who have sleep issues but aren't being treated for the problem. First, tell me about this problem of sleep issues being overlooked in primary care and how you plan to tackle this problem with AI and electronic health records. 

[00:24:13] Minjee Kim, MD: Yes. Actually, this is completely new areas that I'm delving into. Many years that I've used and leveraged this core studies, I was always a little bit flustered by the fact that there are all these observational studies suggesting that poor sleep is bad for you. There have been relative palsy of intervention studies showing that what we can do about it, and I think one of the major roadblocks there is that it's hard to find these people with poor sleep, no matter how you define it, because it's not on your chart and people don't advertise it. So even though one third of people sleep not enough, you have to put in significant effort to identify those people and screen them, make sure that they're eligible for studies. It's very labor intensive. And once you find an intervention that works in this group of people, how can you implement and all these people, you again have to find them as I was sort of struggling with this gap between the knowledge they we're accumulating and what we can do about it. Now there is this major advances in artificial intelligence and also widespread use of electronic health records, and this seems to be really an optimal combination that can change and tackle these roadblocks that we've had. What we are trying to do is utilizing this functionality. So let's say this MidCog study is recruited through primary care, and because of that we actually have access to the participant electronic health records. So. What we are trying to do is, okay, so we identify these people with sleep problems of the different types. Can we now predict who's gonna have these problems without asking them to wear all this fancy watch and answer questions and questionnaires using artificial intelligence? Then if we do that, then can we nudge the clinicians when the patients come to see them for annual visit or anything else? You are at risk of sleep problem, that's bad for you and it might trigger a simple question. And if it seems to be the case, then okay, then there's something that we can send you to, we can refer you or we can provide some sleep hygiene things that they can do something about. We don't know whether that's gonna work, but at least that's the idea of the next steps. 

[00:26:35] Erin Spain, MS: Okay, so this is really new. As you mentioned, you're doing a pilot of this?  What can you tell me about that? 

[00:26:39] Minjee Kim, MD: So the pilot study that we just completed was trying to find sleep apnea in older adults in primary care and send them to diagnostic studies than treatment. It was a minimal effort for the clinicians or the patient we send a screener, that's a four item questionnaire to patient who are 65 and above who come for wellness visit, and if they answer those questions in a way that seem to be at risk of sleep apnea, then the clinicians are automatically nudged to ask them about sleep habits and order diagnostic study. Then what we wanted to know in this pilot study was, is this gonna work? Like, do people find this burdensome or do people actually follow through with the recommendation? We were actually surprised that these older adults, over 40% of them actually did complete the screener and, a significant number of them followed through with the recommendations and were diagnosed and started treatment for that. So minimal effort was not costly but made some differences in people. So that was the pilot study that we're going to also utilize for the next step in different settings and also in middle age adults. 

[00:27:54] Erin Spain, MS: You've been really busy, you've been putting out a lot of studies. You're working on these pilot projects, you're working with all these investigators, you know, what do you hope happens with your lab and your work in the next 5 to 10 years? 

[00:28:04] Minjee Kim, MD: I have a lot of ideas, but what I would like to focus on is this pilot study of utilizing AI and electronic health record functionalities to intervene or detect a bad sleep patterns or suboptimal sleep patterns and intervene on yeah. I think that's where I would really like to focus my efforts on, because that's scalable, that can be modified in different settings, and that's my dream and the goal, like, you know, can we all sleep better? That's a fundamental human right that we don't know about. And then, you know, but I think we should all sleep like it was 1825. 

[00:28:39] Erin Spain, MS: I love it. Well sleep like it's 1825. That's great. Thank you so much for all this insight and sharing the incredible work that you're doing. I really appreciate it. 

[00:28:48] Minjee Kim, MD: Thank you. 

[00:28:52] Erin Spain, MS: You can listen to shows from the Northwestern Medicine Podcast Network to hear more about the latest developments in medical research, health care, and medical education. Leaders from across specialties speak to topics ranging from basic science to global health to simulation education. Learn more at feinberg.northwestern. edu slash podcasts. 

Physicians who listen to this podcast may claim continuing medical education credit after listening to an episode of this program.

Target Audience

Academic/Research, Multiple specialties

Learning Objectives

At the conclusion of this activity, participants will be able to:

  1. Identify the research interests and initiatives of Feinberg faculty.
  2. Discuss new updates in clinical and translational research.

Accreditation Statement

The Northwestern University Feinberg School of Medicine is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to provide continuing medical education for physicians.

Credit Designation Statement

The Northwestern University Feinberg School of Medicine designates this Enduring Material for a maximum of 0.50 AMA PRA Category 1 Credit(s)™. Physicians should claim only the credit commensurate with the extent of their participation in the activity.

American Board of Surgery Continuous Certification Program

Successful completion of this CME activity enables the learner to earn credit toward the CME requirement(s) of the American Board of Surgery’s Continuous Certification program. It is the CME activity provider's responsibility to submit learner completion information to ACCME for the purpose of granting ABS credit.

Disclosure Statement

Minjee Kim, MD, receives grant or research support the National Institutes of Health. She previously received grant or research support from Genetech, Inc. but the relationship has ended. Course director, Robert Rosa, MD, has nothing to disclose. Planning committee member, Erin Spain, has nothing to disclose. FSM’s CME Leadership, Review Committee, and Staff have no relevant financial relationships with ineligible companies to disclose.

All the relevant financial relationships for these individuals have been mitigated.

Claim your credit