Sleep and performance

Abstract

I review the path of my career in sleep. My focus has been on the need for sleep and the relationship between sleep and performance. I have done sleep research in the sleep lab setting and have also taken unique opportunities to measure sleep loss effects on real-world performance. My studies have included long and short sleeper studies, evaluations of various sleep aids, sleep loss effects, jet lag effects, naps, and the consequences of being a poor sleeper. Over the course of my career in sleep, I have also taught about sleep in university and professional educational settings. I am a Board Certified Sleep Medicine Specialist with a private practice, providing diagnosis and treatment of sleep disorders in children and adults.

My career in sleep has been focused on human sleep and performance. In the course of my sleep career, I have done many sleep research studies. I have taught about sleep, dreaming, and sleep disorders at universities and in continuing medical education courses (CMEs) and hospital grand rounds. As a Diplomate of the American Board of Sleep Medicine, I evaluate and provide treatment for children and adults with sleep disorders. I have investigated the need for sleep, methods to improve sleep and increase total sleep time, and the effects of sleep loss on human performance. I have been privileged to do my sleep research in highly regarded sleep labs with prominent sleep researchers from the United States and around the world. Much of my sleep research has been done in the sleep lab setting. I have also had unique opportunities to do sleep research in real-world settings. I was one of the first researchers to take ambulatory sleep monitors outside of the sleep lab to study sleep and performance in the real world. Through different approaches, I have been committed to helping people sleep better.

I did my graduate work at Harvard. While I was at Harvard, Ernest Hartmann (Ernie Hartmann) invited me to be a research fellow at Sleep and Dream Laboratory at the Boston State Hospital. I eventually became the Assistant Director. My sleep research was greatly influenced by Hartmann’s theories of the functions of sleep and the biology of dreaming [1, 2]. I was very interested in his research on long and short sleepers and the use of the amino acid l-tryptophan as a natural hypnotic. My first paper with him was a review of what was then known about the effects of various psychoactive medications on human sleep [3].

My Harvard dissertation was on long and short sleepers. I wanted to investigate the individual differences in the need for sleep. Long and short sleepers provided a natural experiment on individual sleep requirement. Most adults can readily and confidently answer the question of how much sleep they need to be alert and feel well rested on the next day. Most adults say that they need 7–8 hours of sleep to be alert and feel rested. There are adults who say they need much more sleep, over 9 hours per night, and they are classified as “long sleepers.” There are also adults who say they need much less sleep, 6 hours or less, and they are classified as “short sleepers.” The need for sleep is normally distributed with long and short sleepers at the extreme tails of the curve. True long and short sleepers do not have medical disorders, sleep pathologies, or complaints about their sleep. What is striking about long and short sleepers is that they are fully aware that they need more or less sleep than other adults. They generally adapt their lives to accommodate their need for more sleep if they are long sleepers or to utilize their extra awake time if they are short sleepers.

In my research projects, the long and short sleepers were carefully screened for participation according to the well-established sleep lab procedures. There were interviews, questionnaires, and sleep logs to confirm that these people were indeed long and short sleepers. I recorded overnight sleep studies in the sleep lab and also collected 24-hour urine samples to measure 3-methoxy-4-hydroxyphenylglycol (MHPG). At that time, there was a great deal of attention focused on the catecholamine theory of depression and the role of catecholamines in the functions of rapid-eye movement (REM) sleep. Joseph Schildkraut and Seymour Ketty at Harvard had determined that urinary MHPG was a valid measure of central norepinephrine turnover in the brain. Their lab did the analyses of the urine samples. The MHPG results showed that long and short sleepers differed in how quickly they turned over norepinephrine in the brain during sleep (see Figure 1).

Figure 1.

Long and short sleepers MHPG.

The graph shows the mean hourly rate of excretion of MHPG ± 1 standard error for the daytime urine collections and nighttime urine collections for long and short sleepers.

The sleep recordings showed that the sleep stage distribution of long and short sleepers was highly similar during the first few hours of the night. Their sleep for the first several hours showed normal sleep cycles with slow-wave sleep abundant in the first two cycles. The difference in their overall sleep stages was that long sleepers continued the REM and non-REM sleep cycles for additional hours toward morning [4, 5].

In the course of the long and short sleeper research study, I met many long and short sleepers. The longest long sleeper that I have studied needed about 10 h per night. The shortest short sleeper needed only 2 h per night. It was amazing to talk with her about her long days of wakefulness and her short nights of sleep. Originally, her husband brought her to me because he had never seen her asleep, and he was concerned that there was something wrong. On the day that she came to the interview, she had actually painted all of the walls in her kitchen in the early morning hours while her family was still sleeping. Her usual bedtimes were from midnight to 2 am. I recorded her sleep in the sleep lab. She slept for one full sleep cycle that included both non-REM and REM sleep. She then woke up on her own, reporting that she was finished sleeping and was refreshed and ready for the day.

Throughout my research career, I investigated the effects of several different sleep aids. Being able to get enough sleep and get good quality sleep are essential to achieving and maintaining daytime performance. Various sleep aids were of interest and were evaluated in our many sleep lab studies. The goal was to figure out if any of these sleep aids would be useful in promoting more sleep and better sleep without other adverse consequences. One important question was whether the sleep achieved when using the sleep aid was normal and restorative sleep.

Ernie Hartmann and I continued to do research on l-tryptophan. At the time, there was enormous interest in l-tryptophan as a “natural hypnotic.” l-tryptophan is an essential amino acid, regularly ingested in the diet, and is the dietary precursor of serotonin. Serotonin was thought to be involved in the control of slow-wave sleep and in modulation of wakefulness. Ernie Hartmann had been studying l-tryptophan since the 1960s. He was awarded a federal research grant to study the use of l-tryptophan as a hypnotic agent. Many different researchers did studies on l-tryptophan during that time period. The results of the studies differed greatly, depending upon multiple factors. Many dose levels were evaluated in the earliest studies, with doses ranging from 1 g to as high as 15 g. Many different types of subjects were involved and many different protocols were used in the various studies.

We investigated the effects of l-tryptophan in lower doses ranging from 0.25 g to 1 g. These lower doses were of interest because they were within the range of normal dietary intake. The subjects were all sleep-onset insomniacs who took over 30 minutes to fall asleep. Our results established that the 1 g dose was the lowest effective dose for use in sleep onset insomnia [6].

We also did dietary studies on amino acids in collaboration with John Fernstrom at the nutrition program at Tufts. Their researchers created protein and carbohydrate meals for a comparison on the uptake of tryptophan. Our results showed that a carbohydrate meal enhanced tryptophan uptake compared to a protein meal. The carbohydrate meal with l-tryptophan lowered their level of subjective alertness and raised their level of subjective sleepiness [7, 8]. Sleep specialists are frequently asked questions about what to eat at bedtime or what not to eat as a bedtime snack. The findings from this dietary study explained why cookies and milk are the perfect bedtime snack. Cookies provide the carbohydrates, and the tryptophan is well represented in milk.

I was awarded a National Research Council postdoctoral research fellowship. This fellowship provided generous support for a full-time 2-year research position in one of several government-supported research labs. I selected the Naval Health Research Center in San Diego to work with Laverne C. Johnson (Verne Johnson). His research group was already well known for studies on sleep deprivation and the effects of sleep loss on human performance. After my fellowship, I was invited to stay at the Naval Health Research Center. I became the Department Head of the Psychopharmacology Department and Sleep Lab. Our sleep lab was located on the campus of the Naval Hospital, San Diego. Over many years, we had extraordinary support for our sleep research program from the Commanding Officer and from the various clinical departments at the hospital.

Long before my time at the Naval Health Research Center, Verne Johnson was conducting various studies of the performance effects of extended periods of sleep loss. Information about the effects of a long vigil were provided in part by the science fair project of a teenager named Randy Gardner about 60 years ago. I consider Randy Gardner to be one of the folk heroes of sleep research. He was a San Diego high school student who stayed awake for 11 days and 24 minutes for his science fair research project. His vigil became newsworthy. Verne Johnson quickly organized sleep researchers to come to San Diego to study Randy during his vigil and to record his recovery sleep in the sleep lab. William C. Dement (Bill Dement) was one of the researchers who came to San Diego for this project. At the time, Randy set the Guinness World Record for time awake without sleeping. As expected, he became increasingly sleepy as the days went by. It was hard for him to stay awake during the vigil, and it was harder still to keep him awake as the vigil progressed. As soon as the vigil was completed, Verne Johnson had Randy transported to the sleep lab so that his recovery sleep could be recorded. There was actual concern that maybe Randy might not wake up or that he would never be the same. His recovery sleep was about 14-hour duration. When he woke from sleep after his first recovery night, he felt quite restored and back to his normal self. We learned from this project that he did not have to make up for his lost sleep on a minute-by-minute basis over many nights, even after such a sustained period of wakefulness. His recovery sleep showed a rebound effect in both slow-wave sleep and REM sleep. I was intrigued by this story and also impressed that the sleep researchers were committed to studying Randy using traditional sleep lab recordings [9].

When I moved to San Diego for my research fellowship, one of the very first things that I did was to try to find Randy. I was able to locate him. He was still living in the San Diego area. I invited him to come back to the sleep lab to talk with us about his experience. At the time, he was certain that being awake for 11 days had not changed him or harmed him. He did describe to us in vivid detail the overwhelming sleepiness that he felt during his 11-day vigil.

At the Naval Health Research Center, I continued to do research on the effects of various sleep aids on sleep and performance. We evaluated the effects of l-tryptophan on daytime nap sleep and the waking electroencephalogram (EEG). The subjects were full-time day workers who did not ordinarily nap during their workday. The subjects were scheduled to come into the sleep lab to take a nap during their usual work hours, either in the morning or in the afternoon. Lights out for the naps were at 9:50 am in the morning or at 1:50 pm in the afternoon, once after administration of l-tryptophan 4 g and once after taking a matching placebo. Blood samples were collected to measure plasma tryptophan levels. The protocol was a model for jet lag, since the subjects were scheduled to take a nap at times that were not their regular sleep times. In addition, we were evaluating whether l-tryptophan would have sleep-enhancing effects in subjects who were not significantly or intentionally sleep-deprived. One of the most surprising findings was that every subject fell asleep under both the placebo condition and the l-tryptophan condition. l-tryptophan significantly reduced daytime sleep latency without altering nap sleep stages. l-tryptophan administration did effectively elevate plasma total and free tryptophan levels. The waking EEGs showed that l-tryptophan increased alpha and theta activity in the awake EEGs. The EEG changes and the reduction of sleep latency were consistent with the theory that l-tryptophan modulates wakefulness, lowering the level of arousal during the awake state to permit more rapid sleep onset. It was also important that l-tryptophan could promote more rapid sleep onset during an unusual time for sleep [10, 11].

Detrich Schneider-Helmert from Switzerland and I collaborated on a review article of research studies that evaluated use of l-tryptophan in the treatment of insomnia [12].

We also did several sleep studies on triazolam. At the time, triazolam was a newer benzodiazepine hypnotic. Triazolam was of interest because it was already known to be an effective hypnotic and because of its shorter half-life. It was possible that this medication would effectively promote sleep but not leave any residual performance decrements the next morning. This study evaluated the effects of triazolam 0.5 mg compared to placebo. Subjects were poor sleepers who had sleep onset insomnia. They qualified as poor sleepers on the basis of an all-night polysomnographic sleep study according to our usual screening procedures. This study was a long research protocol, 11 nights. Performance testing included the Wilkinson 4-Choice Reaction Time Test, the Digit Substitution Test, the Williams Word Memory Test, and the Card Sorting Test. Performance was assessed every morning at 8.25 hours post-drug administration. There were no morning performance impairments caused by triazolam use. On one study night, subjects were awakened from sleep at 1.5, 3, and 5 hours post-drug administration for performance testing. Subjects had to get up out of bed and walk down a hallway to the testing room to do their performance tests. After testing, they returned to their bedroom, and all of them returned to sleep. Performance was significantly impaired by triazolam on all performance measures during nighttime test sessions. As an example of the findings, Figure 2 shows the results for one performance test, the Wilkinson 4-Choice Reaction Time Test. Reaction time was slower during the nighttime test sessions after taking triazolam. We also found that triazolam produced anterograde amnesic effects on a test of word pair recognition of word pairs learned during the nighttime test sessions [13].

Figure 2.

Effect of triazolam on reaction time.

The graph shows mean reaction time on the Wilkinson 4-Choice Reaction Time Test. Subjects were awakened from sleep for performance testing at 1.5, 3, and 5 h post-administration of triazolam 0.5 mg or placebo. The morning test session (am) was conducted at 8.25 h post-administration.

We wanted to know if taking triazolam would impair response to important signals during sleep. We evaluated response to a smoke detector alarm. Subjects in this study were sleep-onset insomniacs. They took triazolam 0.25 mg, triazolam 0.5 mg, or placebo at bedtime. A standard home smoke detector alarm was sounded during stage 2 sleep, slow-wave sleep (stages 3 and 4), and at the time of the morning awakening. The smoke detector alarm was 78 decibels sound pressure level (dB SPL) at the pillow. Almost 50% of the triazolam subjects failed to awaken from slow-wave sleep even when the alarm sounded for up to 3 minutes. All subjects awoke readily to the alarm in the morning. These results did show that triazolam was effective in maintaining sleep even when there was significant environmental noise. But, it was of concern that the hypnotic effect could also significantly reduce response to an important relevant auditory signal, a smoke detector alarm, raising safety concerns [14].

There was considerable debate about whether shorter half-life benzodiazepine hypnotics caused rebound insomnia. Rebound insomnia is a worsening of sleep compared to baseline that occurs upon discontinuation of the medication. J. Christian Gillin (Chris Gillin) was on the faculty at the University of California in San Diego and the La Jolla Veteran’s Affairs (VA). He was an officer in the reserves and was assigned to our sleep lab to do his required active duty for training (ACDUTRA). During his time in our lab, we collaborated on an important review of the research papers that discussed the factors related to rebound insomnia [15].

In related investigations, we measured the effects of sleep medications on EEG activity during the waking state as well as during sleep and on sleep stages. Benzodiazepine hypnotics increased the frequency of sleep spindles and stage 2 sleep. In slow-wave sleep, delta amplitude was reduced, and there was a related decrease in minutes of slow-wave sleep stages 3 and 4 [16, 17]. The amplitude of auditory evoked potentials (AEPs) was reduced by triazolam.

Of interest, I found that just lying awake in bed for 10 minutes prior to lights out for the collection of AEPs resulted in more rapid sleep onset after lights out [18].

In the real world outside of the sleep lab, there are sleep loss issues associated with the requirements of various occupations. Sleep loss effects are of concern in commercial aviation. One group of interest was commercial airline pilots who fly long-haul flight patterns with multiple flight legs and multiple layovers. We collaborated with an international group of sleep researchers coordinated by the National Aeronautics and Space Administration (NASA) on a study of sleep of commercial pilots who flew long-haul flight patterns between their home countries and the United States. We traveled on the flights with these pilots. We recorded their sleep in their hotel rooms during the layovers. We also recorded sleep during preplanned 40-minute rest periods en route; the pilots were scheduled for a rest period in their seat in the cockpit while wearing the ambulatory recorders. The rest periods were carefully scheduled in view of safety concerns, and the naps were approved for this research study. The pilots fell asleep in about 5–6 minutes and achieved about 25–26 minutes of sleep during these rest periods. There was improved alertness after the naps. The study showed that there was significant sleep loss associated with these long-haul flight patterns. In addition, the results showed that the pilots could nap readily in their seats in the cockpit [19].

Sleep is also an important factor in military readiness and mission success. The military became concerned about sleep loss effects and wanted to consider sleep logistics in mission planning. The military often engages in large-scale troop movements that interfere with sleep. Military missions may require sustained performance resulting in sleep loss. Adequate sleep and good quality sleep were essential to military readiness and to mission success. Military personnel may need to sleep at odd and unusual times to take advantage of available rest periods. My association with the Navy offered unique opportunities to take sleep research and pharmacology research into real-world military operational settings.

The office of the Commandant of the Marine Corps contacted me because he was interested in my research on sleep loss, jet lag, and l-tryptophan. He offered me a unique opportunity to do a large-scale field study of jet lag, sleep loss, and performance with Marines. We deployed with a battalion of Marines from Camp Pendleton in San Diego to Okinawa, Japan. The Marines wore ambulatory monitors. All sleep was recorded, including a baseline period, during the flight, and after arrival. The Marines arrived sleep-deprived and very tired. Target shooting scores were collected at Camp Pendleton and after arrival at Okinawa. Target shooting accuracy was reduced by 33% upon arrival in Okinawa. Sleep before they deployed and after their arrival was disrupted, and total sleep time was reduced. Although there were many hours on the long flight that could have been used for sleeping, the Marines slept very little, averaging about 2 hours of sleep en route. On our second deployment with the Marines, we tested interventions designed to promote sleep. I controlled the environment aboard the plane. Caffeinated beverages were not served on the flight. The level of lighting, including the times for the opening and closing of window shades, was all preplanned. I arranged the scheduling of meals and showing movies only at specific times to provide uninterrupted periods of time for sleeping. These interventions resulted in greatly increased sleep time en route. Performance testing showed improvements on reaction time tests and memory tests compared to the baseline study [20].

I established a research program on psychopharmacological agents for possible use in operational settings. I was invited to present our research program at a North Atlantic Treaty Organization (NATO) Defense Research Group meeting in 1983 [21]. The research program first emphasized evaluation of sleep aids [22]. The research program was eventually expanded to evaluate stimulants, including caffeine.

We also wondered how being a poor sleeper affected job performance long term. We had an opportunity to follow good and poor sleepers in their military careers. The military keeps detailed records on each person throughout their years in the military. We tracked a cohort of Navy corpsmen from the time of their training through their performance in the fleet. We wanted to find out how their perceived sleep quality might be related to their real-world performance in the fleet. Corpsmen are trained to provide medical care similar to paramedics, nurse practitioners, and physician assistants. In certain environments, the corpsman may be the only medical provider on a ship or in a clinic. We learned that these young adults could readily categorize themselves as good sleepers, average sleepers, or poor sleepers. We followed good and poor sleepers throughout their Navy careers. We found that there was a clear consequence to being a poor sleeper. Poor sleepers were overall less successful in their Navy careers. As a group, they were less likely to be promoted, and therefore they remained in lower pay grades; they were less likely to be recommended for reenlistment, and their attrition rate was higher. Poor sleepers were also more likely to be hospitalized during their career. Good and poor sleepers did not differ in the number of psychiatric diagnoses related to these hospitalizations or in the number of medical discharges. The poor sleepers were not disciplinary problems. There were no differences in desertions, unauthorized absences, court marshals, or dishonorable or bad conduct discharges from the service [23, 24].

I was asked to determine if sleep problems were a factor in attrition from military special forces training. Special forces training is a highly demanding training program. These trainees selected for special forces are exceptionally fit and able. When they began their training, as a group, the trainees reported being good sleepers, and they did not have complaints about their ability to sleep. Some of them developed sleep problems as the training progressed and became more rigorous and stressful. We found that special forces trainees who developed insomnia as the training course progressed were less likely to complete the training program successfully [25]. As a result of these findings, I was asked to create and teach a class on sleep that was added as a regular component to the training program.

In related research, we studied the sleep of university undergraduate students who were honors students and compared them to students who were not honors students in the same program. We found that honors students had better sleep hygiene overall and slept more on a regular basis than did students who were not honors students [26].

Over the years, the general public became increasingly aware of the need for good sleep. As a sleep researcher, I was often asked to consult on patients with sleep disorders even many years ago. The knowledge obtained from sleep research became the basis of the new specialty of sleep medicine. The American Board of Sleep Medicine was created by William C. Dement (Bill Dement). Sleep Medicine Board examinations were developed and standardized. Formal requirements for Sleep Medicine training and experience were established. I passed the Sleep Medicine Board examinations and became a Board Certified Sleep Medicine Specialist, a Diplomate of the American Board of Sleep Medicine. Early in the development of diagnosis and treatment of sleep disorders, the practice of sleep medicine was generally hospital-based. I established the first military sleep medicine program at Naval Hospital, San Diego. I was the first Board Certified Sleep Medicine Specialist at Naval Hospital, San Diego. Later, working with the pulmonologist Lucien Jassy, I established the Sleep Disorders Center at Scripps Mercy Hospital, San Diego. It was approved as an accredited center by the American Academy of Sleep Medicine. I continue to be a strong supporter of accreditation of sleep disorders centers. For several years, I was on the accreditation committee of the American Academy of Sleep Medicine.

As a Board Certified Sleep Medicine Specialist, I have provided clinical sleep medicine services in my private sleep medicine practice for many years. I evaluate and treat sleep disorder patients of all ages, including both children and adults.

I have always been eager to teach courses about sleep, dreaming, sleep disorders, and how to sleep better. I believe that it is important to teach about good sleep and sleep problems to as many different people as possible, so that they would know the importance of sleep. Even many years ago, there was support for expanding the course offerings at colleges and universities to include classes on sleep and dreaming and on sleep and human performance. I taught the first undergraduate class on sleep at Harvard College. At Harvard, I also taught a multidisciplinary seminar on sleep and dreaming with the Harvard anthropologist Even Z. Vogt. Over the years, I did much of my teaching as an Adjunct Professor at the University of California, San Diego (UCSD). I taught the first class on sleep at UCSD and a course on psychopharmacology that focused on sleep medications.

At the Naval Health Research Center, I always had adequate research space and resources to host other sleep scientists to visit for extended periods of time to collaborate on various studies and projects. These sleep scientists included Reidun Ursin from Norway, who worked with us on the daytime nap study, and Detrich Schneider-Helmert from Switzerland, who collaborated with me on the review article that evaluated use of l-tryptophan in the treatment of insomnia. As previously described, we were fortunate to have Chris Gillin working with us at the sleep lab on a regular basis for his ACDUTRA. We were also fortunate to have an active-duty Navy Officer researcher assigned to our sleep lab, Schuyler C. Webb (Sky Webb). He was a Navy Lieutenant when he was with us. Ultimately, he was promoted to Captain during his Navy career.

When I first became a sleep researcher, we knew very little about sleep and sleep need. We asked very basic questions and sought to quantify sleep loss effects. Surprisingly, there was a time when I was frequently asked how could a person get by on less sleep. Or how could we somehow do away with sleep entirely? Over the years, how to sleep better became the most important question. Various interventions to promote and enhance sleep became the focus of both sleep research and clinical sleep medicine practice.

I very much appreciate the dedication and accomplishments of my contemporaries. For many of us, our careers have been fully committed to sleep. I value my friends in sleep research and in sleep medicine. I am impressed with the sleep researchers and sleep clinicians who have spent their entire careers trying to answer the important questions about sleep and adding to the ever-expanding knowledge about human sleep. I am grateful for my mentors, my sleep colleagues, my sleep research collaborators, and my students.

This paper is part of the Living Legends in Sleep Research series, which is sponsored by Idorsia Pharmaceuticals and Jazz Pharmaceuticals.

Author Contributions

Cheryl L. Spinweber (Conceptualization, Resources, Writing—original draft)

Disclosure Statement

Financial disclosure: No financial arrangement or connections to declare.

Non-financial disclosure: No conflicts of interest to declare.