Internal Time: Chronotypes, Social Jet Lag, and Why You’re So Tired

by Roenneberg, Till

  They found Edgar’s body later that morning on the pavement in front of the house. The police had to wake Ingrid, who was still fast asleep when they rang the bell. The inspector came to the conclusion that Edgar must have tripped over his pouch, and so slipped over the rooftop’s edge while looking through his telescope, which was lying not far from his body. All his other instruments were scattered around him, and the empty pouch had, by chance, ended up covering his distorted face.

  Once all the formalities had been settled, Ingrid sold the apartment, returned to Frankfurt, and never went back to Morocco. Every year when daylight saving time came around, however, Ingrid went to church and lit a candle for Edgy, displaying her famous flash of a smile.

  Obsessed people like Edgar are convinced that nature is the only reliable source of information in today’s world of constant make-believe and blatant lies. It is therefore not surprising that they cannot tolerate how society so profoundly messes with the natural course of our lives—for example, when it changes our social clocks twice a year. The population seems pretty much torn between two camps when it comes to daylight saving time (DST)—one considering it a vice, the other a virtue. According to the latest (nonrepresentative) internet polls of spring 2009, the anti-DST faction is gradually growing. DST supporters remind their foes that clocks are changed by one small hour and in the right direction, so that the change supports the natural seasonal progression of the sun. But few studies have properly investigated whether people, or rather their body clocks, adjust to these time changes, and if so, how long this adaptation takes. Most of these studies have looked only at the first week, haven’t separated work days and free days in their analyses, and also have not considered internal time (chronotype).

  Using Germany as an example, I have described how tightly our body clock is tied to the natural light–dark cycle. Despite the overwhelming impact of social schedules on our lives, which are more or less similar throughout any given country, the body clock becomes later from east to west by approximately four minutes per degree of longitude—the same amount of time per degree of longitude that the sun takes to move from east to west.2

  In view of these results, we were curious to know how body clocks coped with the artificial transitions of our social time, namely when we move in and out of DST. We therefore recruited volunteers willing to participate in a study that accompanied individuals for eight weeks around the transitions from and to DST. The study started with the autumn transition from DST to normal zonetime, followed by the inverse transition in the subsequent spring. The volunteers received a fat envelope via mail one week before the study started. The package contained several questionnaires, which had to be filled in before the study started, as well as sleep logs, which they had to fill in every morning. The sleep logs asked questions about their past twenty-four hours.3 The package also contained an actimeter that had to be worn throughout the eight weeks of the experiment.

  In addition to the field study, we investigated how the general population copes with DST transitions and how their sleep–wake habits change over the course of the year. At that time, we had over 55,000 entries in our database—all with a date. To analyze seasonal changes in circadian behavior, we average the wake-up times on free days for each half-month. These are plotted downward, progressing through the year, beginning with December at the top. The same series of twenty-four data points is plotted twice, one below the other, to help us see the changes across the winter months. The horizontal axis represents the time of day from 5:00 to 9:30 in the morning, based on Standard European Time (SET). The edge of the gray area indicates sunrise.

  It is quite remarkable how tightly daily human behavior is coupled to dawn on free days. Following sunrise, people are on average getting up earlier from midwinter into the spring, as we had expected from our east–west study across Germany. What we hadn’t foreseen is that this dawn-tracking stops as soon as Central Europe switches to DST on the last Sunday in March.4 Throughout DST, wake-up times on free days scatter around a stable social time, around 8:45 A.M. (corresponding to 7:45 SET). Yet, when the social clocks are switched back to the “normal” zonetime on the last Sunday in October, dawn-tracking of the wake-up times on free days immediately reappears. Thus, the body clock adapts its phase of entrainment (chronotype) to season. But why does it do so only during the non-DST months of the year?

  An average of the wake-up times on free days for each half-month from the Munich ChronoType Questionnaire database. At the time of the study, the database contained 55,000 entries. The right border of the gray area represents dawn in Central Europe.

  It is unlikely that the body clock would actually track dawn throughout the entire summer—even without DST. Think of people who live close to the poles, where days may never end in mid-summer. In addition, no records from the pre-DST era suggest that they tracked dawn throughout summer. Since they would wake up at around 5 A.M., they would have to be asleep at between 9 P.M. and 10 P.M., unless they would severely shorten their sleep.

  However, the fact that the body clock drops and picks up tracking dawn exactly at the DST transitions indicates that these artificial time changes may well have an influence on our body clock’s natural adjustment to season. The transition to DST is scheduled around the spring equinox, while the release back to SET is usually more than a month later than the autumn equinox.5 These two asymmetrical transitions correspond to quite different day lengths, yet they still elicit the same prompt response in our sleep–wake behavior. If body clocks were to stop tracking dawn at a certain day length in spring, we would expect that they would pick up to do so at the same day length in autumn.

  According to the results from the large database, arguably the DST transitions merely supported a natural tendency in spring to wake up earlier and in autumn to wake up later, and that the body clock should therefore adjust quite easily to these social time changes. However, the results of our field study indicate that this is not so. The sleep–wake behavior in the relatively small group of volunteers is very similar to that found in the large database. During the four weeks before social clocks were switched from DST back to SET in October, they woke up (on free days) at approximately the same social time—their body clocks did not track dawn. Once the DST transition occurred, however, wake-up times immediately went back to tracking dawn.

  At first inspection, the sleep–wake behavior of our volunteers around the spring transition indicated that body clocks easily adapt to DST. Yet, when we separated the results according to chronotype, it became clear that only early types adjusted perfectly in their sleep–wake patterns. Late types had still not fully adapted by the fourth week after the time change. When we then analyzed the volunteers’ activity–rest rhythms recorded by the actimeters, the inadequate adaptation became even more apparent.

  The daily activity recordings of different chronotypes mirror their respective sleep–wake behaviors. The activity profiles (on free days) shown in the next graph are averaged over the four weekends before the DST change in autumn. The typical free-day activity profile of an early chronotype is drawn in front and that of a late type in the back. While the former begins to be active at around six, the latter doesn’t start before 10:30. At the other end of the day, the two profiles show approximately the same time difference. In this specific example, judged by the times of inactivity, the late type appears to be a slightly longer sleeper than the early type.6

  The next graph illustrates the weekly averages of activity onsets (on free days) of early and late chronotypes and shows that they parallel dawn (the right borders of the gray areas) before the DST transition in the spring.7 Onsets are a bit scattered for the early types but almost form a line parallel to dawn for the late types. At the DST transition, early types jump one hour in their onset times while late types do not seem to respond at all. Yet four weeks into DST, neither of the two groups had fully adjusted to the social time advance of one hour. The early types had shifted only by approximately forty-five minut
es, and the late types went back to where they had started eight weeks before.

  The free-day activity profiles of an early chronotype (front) and a late chronotype (back), averaged over the four weekends before the daylight saving time change in autumn.

  But why do we have such difficulties in adjusting our body clocks to the regular time changes when moving in and out of DST? After all, it is only one small hour! The answer becomes quite obvious when we look at what happens at these time changes in more detail. When we emerge from the long winter nights, we obviously adjust our body clocks to the gradually advancing dawn times. This adjustment is only apparent on free days. On the more frequent work days, we get up more or less at the same local time—in most cases earlier than on free days. At first we get up before the sun, then with her, and then after her rise. This continues for approximately three months until the clocks are switched to DST in March.8 From one day to the next, we are thrown back by approximately three weeks and now get up again before the sun rises. The closer the DST change to the equinox, the shorter is the time we are thrown back in our natural seasonal progression. In the autumn, this scenario is reversed. With the days getting shorter again, we first get up after the sun, then with her, and then before dawn. When the clocks are set back (in Europe approximately a month after the equinox in October), we are thrown back by approximately four weeks (the exact amount depends on latitude). Follow the thick black line (sunrise) in the next graph from left to right (a hypothetical wake-up time of 7 A.M. is indicated by the dashed horizontal line). The spring time-change that throws the seasonal progression back by three weeks (if one lived like Ingrid and Edgar in Frankfurt) is equivalent to travelling fifteen degrees of longitude to the west; the autumn DST transition, which throws our seasonal progression back even more, reverses this virtual travel, transporting us back fifteen degrees to the east. In addition to the abrupt time changes, the seasonal amplitude of dawn is greatly reduced in relation to our daily work life. (The white dotted curve indicates the progression of dawn without DST.) Reducing the sun’s seasonal amplitude is equivalent to a virtual journey toward the equator. For people living in Central Europe, the interference of DST therefore corresponds to a virtual journey from Frankfurt to the south of Morocco and back.

  Weekly averages of the calculated maximum activity times on free days of early and late chronotypes. Both parallel dawn (the right border of the gray areas) before the DST transition in the spring. Early types have not advanced their activity by a full hour even after four weeks, while late types appear not to adapt to the clock change at all.

  In reality, DST is nothing but a collective decision to start work an hour earlier or to work for a company situated one time zone farther to the east, without ever leaving our hometown (similar to what Timothy did in working for LayIn&Out). A cynic might argue that we are made to change our clocks only so that we don’t notice this collective decision, making us believe that we go to work at the same time. Since the results of our field study indicate that body clocks do not properly adjust to the time change (especially those of the later chronotypes), this social intervention is bound to increase our social jet lag and decrease our sleep duration.9

  The broad smile that Ingrid produced at every DST change while lighting a candle for Edgy did not only express her relief at no longer having to cope with his measuring mania. She had also long realized the irony that lay in their annual exodus to southern Morocco. Didn’t it essentially amount to the same changes between their daily behavior and sunrise created by the DST transitions that Edgy abhorred? They might just as well have stayed in Frankfurt.

  Daylight saving time drastically affects the relationship between sun time and our daily habits, similar to traveling southeast in spring and northwest in autumn.

  20

  Light at Night

  Marco Gonzales took the call and looked at his computer screen that showed him all the details of the weather in Manchester. “Good morning, Mrs. Taylor. How are you today? Isn’t it great that it finally stopped raining?” Marco had just started work and Mrs. Taylor’s call to the bank was the first of an endless series he would be taking for the next twelve hours. “My name is Marco. How can I help you?” Mrs. Taylor asked him to pay a doctor’s bill from her account. She was elderly and handicapped and wouldn’t have known how to bank online even if she had owned a computer. “Certainly, that’s what we are here for,” answered Marco cheerfully. “Why don’t you first give me your PIN and then the name of the doctor and his or her account details.” After typing the information into the mask on his computer screen, he said, “and now I will need an identification number for this transaction, Mrs. Taylor.” She slowly spelled out the six-digit number. “I am sorry, Mrs. Taylor, but this number has already been used. Why don’t you take the next one on your list?”

  For the next three hours, Marco dealt with about thirty more customers. Between calls, he rolled his chair back to cast a glance at Maria, who worked three stations away in an endless line of cubicles. It was a particularly busy day and he had had no luck in making eye contact with her. Whenever he tried, she was concentrating on a client. They had met six months ago during the introductory training program that had taught them all the necessary skills for their job, including communication skills and intensive language coaching.

  Besides several big rooms crammed with rows of three-walled cubicles, each of which contained only a computer and screen on a tiny desk, the facility included several small rooms with two bunk beds each. Employees could take a nap in these whenever they felt tired. The frequency and length of such recuperations was strictly regulated. Marco took off his headphones and stood up to look into the row of cubicles parallel to his. Cubicles A04 and A11, normally occupied by Roberto and Miguel, were empty. They both must have gone to take a nap or have a smoke outside. Marco noticed that they had recently synchronized their breaks and wondered why. Today was Miguel’s birthday, and a dozen or so of their friends had arranged to meet in a bar after work to celebrate. Marco sat down again and rearranged his headphones to take a call. Having dealt with the customer, who was complaining about an erroneous transaction, he rolled back his chair again and finally caught Maria between two calls. She routinely turned around after finishing with a client to see if Marco was looking at her, and her face lit up when she finally managed to make contact. Using their usual sign language, they arranged to meet in their accustomed bunkroom in twenty minutes.

  When Maria and Marco were finally done for the day, they walked out of the building and were dazzled by the bright day outside. Their eyes took quite some time to adjust. Most of the rooms of the facility had no windows, so when they emerged they often felt like miners coming back to the earth’s surface. They had arranged to meet the others near the hangout area, where they usually went during their breaks for strong coffee and cigarettes. When everyone had gathered, they walked for less than a mile to their favorite bar. While walking to the bar, Marco turned to Maria and said, “Did you hear a typhoon is on its way?”

  Maria looked at him, slightly worried. “Oh no, not another one. I hate typhoons, and we have had enough already this season.”

  Marco put his arm around her. “Maybe it will turn another way. We’ll be all right, you’ll see.”

  The bar had designed itself around the needs of the young workforce of the newly developed industrial park. There were no neighbors to complain about the noise that surrounded the place at the oddest hours. The party went on for almost five hours and everyone had a lot to drink. Fortunately, they all had the next three days off work. Everyone was in an excellent mood when the party began, but the atmosphere went downhill after José, who was quite drunk already, started to report the latest gossip about who had broken up with whom and couldn’t stop himself revealing rather confidential information about some of the new liaisons that had been the causes of the break-ups. His insensitive disclosures sparked several rows and a fair bit of shouting; in the end, three women were in tears and five
men left the party in a sulking rage. It was already around lunchtime by the time the party was finally over, but Maria and Marco didn’t want to let each other go. Like most of their friends, they still lived at home with their parents, and so they would not see each other for three endless days. They and all of their friends loved their work, and not only for the good salary. It also gave them the opportunity to create their own subculture, away from the strict rules of their families.

  One of the most blatant assaults on the body clock in modern society is shift work. At present, close to 20 percent of the workforce in the industrialized world is engaged in schedules that deviate from traditional work hours and, in most cases, these involve rotations. Several decades of epidemiological research have clearly shown that shift workers develop more health problems than day workers. These include sleep problems, depression, cardiovascular pathologies, digestive tract issues, diabetes and other metabolic diseases, and obesity. Health risks related to shift work even include several types of cancer. As a consequence, the World Health Organization has recently classified “shift work that involves circadian disruption” as a potential cause of cancer. In 2009, the state of Denmark started to pay compensation to women who have worked in night shifts and subsequently developed breast cancer. In the Netherlands, shift workers have started to file legal suits for compensation for various health problems potentially linked to their long-term shift rotations.