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

by Roenneberg, Till

  The chronotypes of partners, however, are assessed very differently by men and by women (see bottom panel). Men judge the chronotypes of their partners independent of their own chronotype: their partner’s chronotype averages around 3 (intermediate type). This contrasts with how women assess their partners: the later their own MSF-based chronotype, the later they judge their partner’s. This striking difference may be explained by another finding. Men go to bed at different times (in most cases later) when their partner is not home, indicating that men adapt their sleeping habits to those of their partner. That is possibly why women think that the chronotype of their partner is similar to their own, while men have a more realistic notion of their partner’s chronotype. The relationship results in a horizontal line, as one would expect if we don’t choose our partners by chronotype.

  Across most ages, men are on average later chronotypes than women according to our selected sample. The differences decrease as men and women age. Thus, when the man is older than his partner, their chronotypes tend to be more similar.

  Chronotype assessment of self and others: the later their midsleep on free days (MSF), the later both sexes subjectively assess their chronotype, showing no differences between the sexes (above; the gray bars indicate the distribution of chronotypes in the entire database). However, men and women subjectively assess their respective partners very differently (below).

  The next question we approached was whether the way subjects saw their partners influenced how they saw themselves. In an earlier study, we had already shown that the self-assessment of the subjects could be quite inaccurate on an individual basis (compared with their actual MSF-based chronotype) and hypothesized that this could be due to the influence of the partner. For example, if an individual was an intermediate chronotype (based on MSF) but the partner was an extreme owl, then he might consider himself an early type just because the partner’s chronotype is much later. Our results confirmed this hypothesis. Individuals who categorized their partner to be an early type assessed themselves to be later chronotypes than they actually were, according to their MSF-based chronotype, while those who judged their partners to be late types saw themselves as earlier.

  These results show that chronotype should be assessed by the sleep times and not by subjective self-assessments. There are chronotype questionnaires that ask subjects only about their habits and preferences. The problem is that the answers will always be biased by the habits of other people. The chronotype distribution in India, for example, is much earlier compared with that in Central Europe, which must lead to different self-assessments: a midsleep on free days of 4:30 A.M. would be average in Central Europe but would characterize an extreme late type in India.3

  The next set of results makes the strong subjective element involved in assessing chronotype even more apparent (and takes us back to Louise and Bruno). When investigating how age affects people’s judgment about their own and their partner’s chronotypes, we got a surprising result: with increasing age, partners judge each other’s chronotypes to be increasingly different, up to an age of approximately fifty-two, when this difference decreases again.4 This observation is remarkable because, in reality, the actual chronotype differences between men and women continually decrease from twenty to fifty-two (becoming indistinguishable at higher ages). There are at least two possible explanations for this observation, one of which is psychological and the other cultural.

  Let’s first look at the psychological explanation. Remember Bruno’s reflections while having his coffee on the terrace about the “shallow slope from being attractively different, . . . to being so different . . . that it sometimes became a burden.” The longer people are together, the more they seem to occupy different niches in their relationship on all levels, ranging from being responsible for different tasks in their daily lives to developing different tastes or opinions. The reasons for this apparent specialization are complex, are not restricted to heterosexual relationships, and also exist in other long-term relationships (siblings or colleagues, for example). One trivial explanation is that people change with age and, in partnerships, not necessarily in the same direction. It is therefore not surprising that partners also gradually view their partners as different chronotypes. The fact that these apparent discrepancies become less severe at higher ages may have many reasons. Decreased sexuality, for example, might weaken certain rivalries between partners and thereby also reduce the need to be different. Another factor is retirement; partners have to rearrange their lives when they stop working.

  Statistics show that the number of divorces depends on age. Divorce reaches a maximum when spouses are between forty and forty-five. People generally divorce each other because they no longer have anything in common or because they have found another partner, which sort of amounts to the same thing. The decrease of divorce rates at higher ages has in part a very simple reason. Those people who don’t get along with their partners are already divorced by the age of fifty-five and may have found new partners with whom the slow divergence process is only just beginning. This principle could also be behind the decline of subjective chronotype differences. As you can see, I cannot offer solid psychological explanations for our observations on partnership timing and therefore meander through many different possibilities. Scientists call this “hand waving.” Fortunately I can discuss the second, the cultural explanation, with more scientific reasoning.

  We store individual, one-time entries in our database, which makes our analysis a cross-sectional study.5 These studies have the disadvantage of not being able to distinguish between age-related differences and those that are due to the fact that subjects grew up in a different era. The majority of sixty-year-old subjects who filled out the questionnaire probably got married several decades ago, when partners did not live together prior to their wedding. Since they had no experience in sharing life with a different chronotype, this quality of the partner didn’t influence their choice. In contrast, nowadays young people start to live with each other before getting married and might decide that always having breakfast alone or spending most evenings alone while their partner is already asleep speaks against forming a long-term relationship. In that case, we would expect that younger couples are generally more similar in their chronotype than older couples. Our data clearly rule out this possibility: chronotype does not correlate between partners, no matter whether they are young or old. The psychological explanation is therefore more plausible than the cultural: the longer people live together, the more differently they see each other.

  Bruno, by the way, rejected the curtain dividing their matrimonial bed and confessed to his wife that he rather enjoyed falling asleep in the guest room. Eventually they reached the very amiable compromise of sometimes sleeping apart and sometimes together.

  22

  A Clock for All Seasons

  It was 7 P.M. on the twenty-second of September when Gerry’s cell phone started to play Pink Floyd’s “Time” from the album Dark Side of the Moon. Barbara’s phone went off simultaneously, playing Nat King Cole’s “You Are My Sunshine”—she loved sentimental 1940s music. The two phones produced a rather cacophonous concert but Barbara and Gerry seemed not to mind. They were driving home to their suburban row house after having visited a friend who worked as a scientist at the local hospital’s psychiatric unit. The sun had almost completely set when the phones went off, and Gerry handed Barbara a rather peculiar-looking pair of pink sunglasses, which he had produced from her handbag. Before putting on the glasses Barbara winked at Gerry, who smiled back and reached into his coat pocket for his own pair, identical to hers. “Are we really sure we want to do this for Tom?” said Gerry, who looked a bit ridiculous with his pink glasses.

  When Barbara had finished her maneuver, overtaking another car, she looked at Gerry and replied, “Of course we’ll pull this off. We promised him, and I think it’s going to be rather exciting—challenging, but exciting. You look ridiculous, by the way.”

  “Speak for yourse
lf,” Gerry replied.

  Tom had asked them to participate in a study he conducted that involved several drastic changes in their daily life. Gerry and Barbara had led so far an almost boringly normal life together—that of many young professionals in the suburbs. But now they had to wear special spectacles at specific times of the day, and they had to expose themselves to as little light as possible once the sun had set and to as much light as possible while the sun was up. Tom’s technician had provided them with two phones on which he had installed a special application that produced an alarm whenever they should be keeping away from the light and another one to remind them when to seek light. The users could, of course, choose their own ringtones.

  Barbara and Gerry had no difficulty in complying with their prescribed protocol during the first weeks. Thanks to daylight saving time, sunset was around the time they got back home from work and dawn occurred about the time they had to get up anyhow. Although they slept without the curtains drawn, they also used one of those new bedside lamps that simulate dawn—“wouldn’t hurt on a rainy day,” Tom had said. The dawn simulator started out with almost undetectable light levels, increasing the intensity to its maximum over approximately the same duration as the natural dawn outside. An alarm went off when the light reached maximum intensity, just in case the user hadn’t already woken up. Of course, Gerry and Barbara also had their programmed phone reminders. Barbara had banned Gerry’s phone from the bedroom because she couldn’t tolerate the combination of ringtones in the morning. She had selected a gentle start to the day—Keb Mo’s “Every Morning,” which didn’t harmonize with Santana’s “Put Your Lights On.” So Carlos had to sing on his own in Gerry’s study, except when Gerry was away on a business trip.

  The last week of September was unusually beautiful—a true Indian summer. They got up at sunrise and had breakfast on their terrace before going to work. They were surprised to find how early they were able to fall asleep once they retreated into their dark bedroom. Tom’s technician had provided them with special bulbs that gave off a warm light, similar to that of an open fireplace. They had installed these light bulbs throughout their house—it was the only type of light they were allowed to use. The technician had also given them adhesive pink acetate sheets that now covered their television and computer screens.

  The lamps were essential for the upcoming season of short days and long nights. When they were not at home after sunset, Barbara and Gerry wore their pink glasses to avoid contamination by “unsafe” light. Complying with the protocol in midwinter was quite a challenge, but they somehow survived the dark season. They had slept much longer in the past two months than ever before in their grownup lives. During the long evenings, they did a lot of talking. Often they went to bed as early as eight. During those long nights, they sometimes slept in two parts. After the first, “real” sleep, they would hover for some hours in a state somewhere between sleep and wake before going back to a second round of “real” sleep. At breakfast they would then exchange the strangest tales they had experienced when their brains had gone on fairyland journeys between their two sleep episodes.

  When spring finally came, they were filled with a hitherto unknown surge of energy. One morning, looking at the trees outside, Gerry mused that he now knew what it must feel like to shoot leaves. When Tom called and said that it was time to get back to their normal, unnatural life, they seriously contemplated continuing the experiment. In the end they did go back to their normal lives, but they kept their indoor lighting and the acetate sheets over the screens. The only things they didn’t miss were those silly-looking glasses.

  The American psychiatrist Tom Wehr asked himself what would happen to our sleep behavior if we weren’t surrounded by all those modern light sources, which allow diurnal beings like us to see the world around us despite the sun having long set.1 He persuaded people to participate in his experiment: he asked them to lead a normal life after sunrise and before sunset. But as soon as our astronomical light source had gone, he made them retire into pitch-dark apartments: no lights, no television, not even a refrigerator light that would go on if its door was opened.2 So what is a sighted person to do when suddenly dumped into complete darkness? The best strategy to avoid too many bruises would be to go to bed, which the subjects usually did. They reported falling asleep after lying awake in the dark for a while. But they hardly ever slept through the entire night, which was longer than twelve hours at the time of year when the experiment was performed. The subjects reported waking up several times during the night, often without gaining full consciousness, thus lingering between sleep and wakefulness. I personally believe that this state must have been the perfect time for the birth of fairy tales and sagas, but I don’t remember reading about this in Tom Wehr’s papers.3 What I do remember is that the subjects reported after the experiment was over that whoever thought they knew what “rested” meant had no idea what they were talking about.

  During my first two postdoctoral years with Jürgen Aschoff I investigated human annual rhythms. You may think that two years is not enough to scrutinize annual rhythms, and if I had done experimental work you would be absolutely right. But in fact I managed to include more than 5,000 years into my study—or rather more than 60,000 months. I collected monthly rates of vital human statistics from 166 countries, focusing on the annual ups and downs of mortality, suicide, and births.

  Aschoff had already analyzed many different statistics and had found that most of them show seasonality (they include exotic statistics, like the number of books taken out of public libraries or crime rates involving assaults). After I had finished my doctorate, I took over this project and worked full time on finding sources, writing to statistical agencies worldwide and compiling a database of monthly human statistics as well as climate data.4 We wanted to find out whether the rhythms in human statistics correlate with environmental factors or whether they are merely a product of our seasonal social life (workload in agriculture, for example, is clearly seasonal). Since agriculture also depends on environmental factors, it was difficult to distinguish between environmental and social influences. That’s why I collected as many years from as many countries as possible. The oldest statistics I found date to 1669, and the most recent to 1981.

  Throughout this book you’ve read that practically everything in our body and in our lives follows a daily cycle. This is more or less also true for our seasonal existence. Seasonality increases the farther humans live from the equator, and it is six months out of phase between the northern and the southern hemispheres. In regions that experience drastic changes in photoperiod (including real winters with temperatures below freezing point and possibly also with a lot of snow), the population’s entire life is dominated by the seasons—or at least it was in the preindustrial era. It’s therefore not surprising that many statistical aspects of human life show seasonality—some for quite trivial reasons. I have already mentioned that people use libraries differently according to season—borrowing more books in winter than in summer, since reading is an indoor activity during long winter evenings. There are, of course, also more skiing and skating accidents in winter. In Europe in the summer, there are more accidents on the way to school (because more students are on the roadways, riding their bikes to school), and more people drown in swimming accidents or suffer from sunstroke. You already know that we sleep longer in winter than in summer. The difference is not lengthy—only around twenty minutes—but it is very systematic and statistically highly significant.

  Other seasonalities in statistics are less trivial: children grow more rapidly in spring than in autumn but put on more weight in autumn than in spring; people eat more carbohydrates and less protein in winter and are more depressed than in summer. “Depressed” is a rather grand word for the seasonality in mood of the general population. Researchers have used questionnaires that are normally used to diagnose depression and its severity in potential patients in the general population at different times of the year and found that depr
ession ratings go up in the autumn and come down in spring. The poll was carried out in response to a syndrome called seasonal affective disorder (SAD). SAD patients become severely depressed every autumn and recover during spring. Unlike other forms of depression, in which patients often feel too low to eat a proper meal, SAD patients develop a carbohydrate craving when the days get shorter.5 In their seasonal moods and eating habits, SAD patients show a pathological extreme of what the normal population goes through in the short days of winter.

  Another seasonal rhythm evident in human statistics concerns the rates of suicide. If I asked you to guess in which month the annual rhythm of suicides had its peak, you might guess November or December—that, at least, is what most people think. Given that the seasonality of mood reaches its low point around that time of year, it’s a fair assumption. The data, however, show something else. The maximum number of suicides worldwide occurs around the summer solstice.6 The hypothesis that tries to explain this counterintuitive fact is only indirectly linked to the seasonality of mood. It assumes that when everyone is feeling low, the drive to die by suicide is also lower than at times when everyone else is in excellent spirits. In addition, patients who suffer from bipolar depression and die by suicide do so during their manic phase and rarely during their depressed phase.7 The actual act of suicide (and not merely the thought about it) takes a level of energy that depressed individuals cannot muster during their most depressed times of year. Thus those individuals who are desperate enough to put an end to their lives do so with higher probability in midsummer, when their energy is highest.