Introduction

Oscillations of endogenous biological processes that follow the approximate 24-h, that is circadian, rhythm exist in diverse organisms and arise as an adaptive mechanism to the Earth’s rotation producing the foreseeable transitions of night and day. These oscillations are generated and maintained by endogenous circadian clocks.

In the beginning of life, 3.5 to 3.9 billion years ago, the length of day was about 14 h and ultraviolet radiation was not filtered by the Earth’s atmosphere. Thus, circadian clocks met the night-day transitions with the approximate period of 14 h and were readily involved in protection from ultraviolet radiation. Thereafter, the period has lengthened, along with which circadian clocks have evolved a slowing-down mechanism [1, 2] and adopted periods of longer than 14 h up to the current one of about 24 h as the reference. Circadian clocks are endogenous pacemakers that evolve their properties, when subjected to selection, but have remained conservative during evolution [3].

The individual’s sleep-wake (rest-activity) cycle is a product of the homeostatic sleep process and the circadian process that interact with each other [4] but appears to be dictated mostly by the latter [5]. The endogenous circadian rhythms are normally aligned with the individual’s sleep-wake cycle. Therefore, the behavioral trait of preference to schedule the daily activities for morning or evening hours can be estimated from the timing of sleep phase (bedtime to wake-up time), although the estimates need to be corrected for potential sleep debt. These preferences form a continuum which has the anchorage ends of morning-oriented preference (“early birds”) and evening-oriented preference (“night owls”) [6].

The endogenous circadian period correlates with the subjective report of this behavioral trait, with the circadian phase, and with the wake-up time [7]. The estimates of the endogenous circadian period from 52 women and 105 men, healthy and aged 18–74 years, revealed that the period of the circadian rhythms of melatonin concentration and core body temperature was close to 24 h on average, but shorter in women (24 h 5 min ± 12 min) than that in men (24 h 11 min ± 12 min) [8]. Of note, however, there was no influence of age on the endogenous circadian period [8].

On the basis of twin studies, it is estimated that the genetic effects contribute to the chronotype by about half and the other half is accounted by environmental factors [9, 10, 11•, 12•, 13]. It is suggested that for a given chronotype, the genetic effects may not operate in an additive manner [9], but the underlying liability to chronotype was 38 % on average for the genetic factors due to dominance [10]. Further, there are no gender differences for the magnitude of the genetic and environmental effects [9, 10], nor is the magnitude of these effects influenced by age [10].

The continuum of chronotype diversity is largely mediated by genes, and the schedule of gene expression is fundamentally different, not just merely advanced or delayed, between the “early birds” and “night owls” [14]. In such epigenetic landscape, a single genotype can thereby lead to different phenotypes. Moreover, different genes for the chronotype are expressed in different generations [9]. These phenomena might explain the currently equivocal data on genetic associations with chronotype.

The endogenous circadian clock of “early birds” tracks an earlier hour, as compared with that of “night owls,” with respect to both the social clock time and the individual wake-up time. Most of the individuals have a preference as intermediate with respect to these two extremes. Under such conditions where it is allowed to consider only the individual’s “feeling best” rhythm for planning the day, such as on free days in real life or in constant routine at a sleep research laboratory, the chronotype can be assessed accurately. There are a number of instruments for assessment of the chronotype (see Table 1).

Table 1 Assessment of chronotype
Full size table

The assessment of chronotype has gained attention as working in shifts has become more and more usual, and there is a growing body of literature on health hazards that are related to shift work, especially to night shift work [15]. However, the assessment of chronotype may be highly relevant not only for shift workers but also for others as well. During the past 10 years, health hazards have been attributed to the chronotype itself, being independent of working in shifts, and a summary of potential associated health hazards is presented in Table 2. Therefore, we believe that additional research on the health hazards of “night owls” is needed [16].

Table 2 Health-related hazards among “night owls”
Full size table

Health-Related Behavioral Outcomes

Earlier small-scale studies have demonstrated that “night owls” drink more alcohol [17], smoke more tobacco [18], keep more often irregular or unhealthy diets [19, 20], are less often engaged in physical activity [21], and sleep less or have poorer sleep quality [22–26] than others. Recently, these findings have been verified in a series of studies using a national health examination study dataset on a sample of 6858 adults, aged 25 to 74 years, derived from the general population of Finland [27–31].

Considering all these findings, it is thus not unexpected to find that substance abuse or dependence are more frequent among “night owls” [32], although at least in the presence of bipolar disorder it seems that “early birds” are more prone to substance use disorders [33]. Furthermore, overweight [34, 35•] and sleep apnea [36] are more frequent among “night owls”.

All the aforementioned results on “night owls,” when compared to early birds,” have a tendency to have less resilience to adversity and less optimism for life [37], worse academic achievement both as school pupils and as university students [38], and having difficulties in their adaptation to an work environment that switches frequently or irregularly between shifts [39, 40•]. But, it may not necessarily need to be this way, as the circadian disruption can be alleviated by adjusted shift schedules [41]. Further, among the “night owls,” there appears to be a subgroup of individuals who have equal cardiovascular risks, sleep well and present similar levels of academic achievement to those of the majority of “early birds” [30, 42].

Mental Health Outcomes

The assessment of chronotype in a sample of healthy persons, aged 18 to 99 years, demonstrated that “night owls” had more severe depressive symptoms, especially those of sadness, inner tension, sleep reduction and pessimism, than others [43]. Among adolescents (aged 12 to 18 years), the chronotype of “night owls” appeared to be a risk factor for depressive symptoms and insomnia, but not for anxiety [44]. This finding is in agreement with that among adults aged 20 to 59 years, “night owls” had increased odds for having depressive states which appear to be independent of sleep parameters [45]. The original report on the association of the “night owls” with depressive disorder compared 39 patients with 39 matched controls [46]. Thereafter, these findings have been replicated and extended by using a national population-based data on adults aged 25 to 74 years [47, 48].

The increased odds for bipolar disorder, either type 1 or type 2, was initially reported from 190 patients and 128 community-driven controls [49]. Thereafter, during the follow-up of 4 years on average, patients with bipolar disorder remained as “night owls” more frequently, regardless of their current mood state, as compared with healthy controls [50••].

Beyond mood disorders, the chronotype of “night owls” appears to be more common among individuals with anxiety and substance use disorders [51]. In addition, studies have shown that “night owls” tend to have a past diagnosis of more than just one mental disorder or, in other words, a higher prevalence of co-morbid conditions [51, 52].

On the admission day to hospital treatment, patients with personality disorder, substance use disorder or anxiety disorder were more often “night owls,” whereas those with psychotic disorder or depressive disorder were more often “early birds” [53]. Even though these findings may seem contradictory to the earlier reports, it has been proposed that “early-bird”-like circadian rhythms might play as a relief factor after the onset of major depressive episode [54]. Currently, it is not known whether this action of relief, if it were to be true, is linked to an attempt to reset the circadian rhythms. The reset hypothesis of mood disorders, claiming therapeutic effects from resetting the desynchronized circadian rhythms, has been put forward and gained experimental support [55–57]. In agreement with this hypothesis are the finding that “night owls” had the increased odds for non-remission of a depressive episode, independent of insomnia [58•] and that, in patients with insomnia, the closer the chronotype was to that of the “night owls,” the less alleviation of depressive symptoms there was after group cognitive behavioral therapy [59].

General Medical Health Outcomes

The influence of the chronotype is not limited to mental health outcomes but is associated with general health outcomes, thus potentially having a high impact on public health. In a national health examination study representative of adults aged 25 to 74 years in Finland, it was found that “night owls” had the increased odds for arterial hypertension, even after controlling for sleep duration and insufficient sleep [60].

Animal models have suggested that individuals with a longer circadian period due to a single deficient circadian clock protein are predisposed to weight gain and disturbances in glucose metabolism, reduced insulin sensitivity, and reduced glucose tolerance, leading to a condition similar to the metabolic syndrome [61]. Non-invasively, housing mice in the 20-h light-dark cycles disrupts their endogenous circadian period and results on a similar condition [62]. Furthermore, human experiments have demonstrated similar effects after sleep deprivation in healthy volunteers [63] and in patients with type 1 diabetes [64], as well as in healthy volunteers after disruption of the circadian clock functions [65].

Considering the aforementioned findings, it is not surprising to see that among patients with type 2 diabetes, independently of sleep disturbances, those whose chronotype was closer to the “night owls” had poorer glycemic control as assessed with the concentrations of glycohemoglobin [66–68]. These findings were extended by studies where “night owls” had the increased odds for type 2 diabetes in a sample of adults aged 25 to 74 years [60] and in another sample of adults aged 47 to 59 years [69].

Beyond cardiovascular and metabolic risks, “night owls” have higher prevalence of wheezy breathing without respiratory infection, a diagnosed or treated bronchial asthma, awoken in shortness of breath, awoken in heavy to breath, wheezy breathing with dyspnea, wheezy breathing as such, and medication for asthma more often than “early birds” [70]. Further, among women aged 25 to 54 years and still menstruating, the closer the chronotype was to the “night owls,” the longer the menstrual bleeding was and the higher the prevalence of infertility was [71•].

Chronotypes over the Course of Life

First, in early childhood, circadian preferences that are close to the “early birds” seems to best characterize the preference in healthy children aged 30 to 36 months, whereas none were definitely “night owls” [72].

During the puberty, among adolescents aged 10 to 16 years, the buildup of sleep pressure that increases sleepiness over the course of day was slower among those whose puberty was more advanced than that of the others [73]. Further, among adolescents aged 9 to 19 years, the dissipation of sleep pressure did not change across adolescent development nor correlated with the circadian phase [74]. However, after puberty, among persons aged 19 to 34 years with intermediate circadian phases, both the baseline level and decay rate of slow-wave activity during the night-time sleep were higher in “early birds” than that in “night owls,” but the extreme chronotypes did not differ from each other [75]. The latter one of these findings [75] suggests that chronotype may originate from some fundamental differences in the dissipation of sleep pressure, whereas the first one [74] appears to contradict this notion. Thus, the sleep and circadian processes influence the sleep-wake schedule preference independently, but the roots of chronotype may lie in the circadian process rather than in the sleep process. This assumption agrees with the proposed three-oscillator model [5].

Finally, in late adulthood among seniors aged 65 to 94 years, circadian preferences that are close to the “early birds” are again more common than other chronotypes [76]. Such findings have suggested that the chronotype changes with aging toward a more morning-type preference. However, these data were based on cross-sectional samples, and as the longitudinal data on the chronotype of the same persons as assessed with the same methods were missing, until recently there was no answer to the question whether the chronotype changes with aging.

To answer this aforementioned question, a longitudinal dataset on individuals aged 57 to 97 years was analyzed and revealed that there was a shift from “night owls” to another chronotype with aging from the years of 1985 to 2008 [77••]. The closer to the “early birds” the chronotype was, the more persistent it remained with aging [77••]. Alarmingly, the relative risk of all-cause mortality was on average 1.3-fold (95 % confidence interval of 1.0–1.6) higher for “night owls” as compared with “early birds,” these premature deaths starting to emerge after the age of 55 years [77••]. It is of note here that these data concerned only men and that such data on women are still missing. However, this finding supports the view that with aging, the decreased proportion of “night owls” is partly due to premature mortality in this group.

Recently, a view on the distribution of chronotypes among older age groups proposed that “night owls” tend to die younger than “early birds” [78]. The finding of the increased all-cause mortality among the “night owls” is new and supports the view, but currently there is no clear explanation to it. However, it holds that “night owls” do have greater deviations of their circadian period from the 24-h period than others [7]. It is not known how this may link to their greater morbidity or mortality or whether their greater morbidity or mortality is due to a sustained drive for delays and subsequent circadian disruption. This view might not be far-fetched though, because currently all the health hazards that associate with chronotype are more frequent among “night owls,” and this is likely to contribute to subsequent morbidity and thereby potentially to all-cause mortality.

Future Directions

There are still many research questions on the chronotype that await elucidation. One of these “scientific coconuts” that await breakthrough experiments for explanation is the change in chronotype with aging. During the puberty, there seems to be a subgroup of adolescents who gravitate toward the preference of the “night owls” and who remain as such. It will be not only of importance to health promotion but also of clinical relevance to find out the key mechanisms that favor this gravitation of chronotype, identify persons who remain at clinically meaningful risk and provide them a protective intervention early enough.

Another goal is to collect more data from big-scale population-based studies to have valid estimates of the prevalence rates of chronotypes and their associations with outcomes of health status after controlling for potential confounding or masking. Ideally, such studies need to provide data that are representative of the whole population, or persons aged 0 to 120 years. These actions will initiate a series of cohort studies and eventually help in solving the problem of chronotype gravitation. In addition, prospective design for these studies will clarify the causal relationships between the chronotype and the attributed conditions. This call is out for all research groups on the topic.

As “night owls” have equal sleep durations and dissipation rates of sleep pressure to others, it remains to be seen instead whether “night owls” have a stronger slowing-down mechanism in their circadian clocks than others. Such property may lead to a longer circadian period and a slower buildup of sleep pressure during wakefulness, which tends to characterize the “night owls”.

Conclusions

So far, without any exception, of those health hazards that do differ between the chronotypes, all have been more common among the “night owls” as compared with the “early birds.” Therefore, the routine assessment of chronotype as part of health status examinations in health-care settings on every level and early enough seems not only relevant but also urgent.