Knowing an individual’s circadian phase is important for designing treatments for adaptation to jet-lag and shift work, and in the diagnosis and treatment of various circadian rhythm sleep disorders (e.g. delayed sleep-phase syndrome). Because measuring circadian rhythm phase using core body temperature or fluctuations in hormonal levels can be a fairly invasive and time-consuming process, questionnaires have been devised to produce an estimate.
The Horne–Ostberg Morningness–Eveningness Questionnaire (MEQ) is a 19-item questionnaire which assesses individual differences in the time of day a person prefers to carry-out various activities, and classifies people as morning-type (M-type), neither-type (N-type) or evening-type (E-type) individuals ( Horne and Ostberg 1976). Several studies have shown that M-type subjects have an earlier sleep schedule ( Horne and Ostberg 1976; Foret et al. 1982 , 1985; Mecacci and Zani 1983; Kerkhof 1991; Kerkhof and Lancel 1991; Carrier et al. 1997 ) and an earlier circadian temperature phase as measured by rectal temperature ( Kerkhof 1991; Lack and Bailey 1994; Kerkhof and Van Dongen 1996; Duffy et al. 1999a ) and oral temperature ( Horne and Ostberg 1976, 1977; Vidacek et al. 1988 ; Kerkhof and Lancel 1991; Neubauer 1992; Gupta and Pati 1994). In accordance with oscillator theory ( Wever 1965) and animal studies (e.g. Aschoff 1965, 1981; Pittendrigh and Daan 1976), it has been hypothesized that humans who have shorter free-running circadian periods would have an earlier circadian phase and more morning tendencies. One recent study revealed that, in fact, individuals with shorter intrinsic circadian periods rated themselves as more morning-like according to the MEQ ( Duffy et al. 1999b ).
Results from studies that have investigated sex differences in circadian phase are somewhat inconsistent, although it appears that women tend to have more morning characteristics than men (see Kerkhof 1985; Tankova et al. 1994 for reviews). With regard to the circadian rhythm of temperature, two studies found that elderly women had an earlier rectal temperature peak than elderly men ( Campbell et al. 1989 ; Moe et al. 1991 ). In young subjects, two studies found that women had a slightly earlier temperature phase than men, but the differences were not statistically significant ( Winget et al. 1977 ; Kattapong et al. 1995 ).
With increasing age, there is a greater tendency toward earlier circadian temperature phase ( Weitzman et al. 1982 ; Nakazawa et al. 1991 ; Czeisler et al. 1992 ; Monk et al. 1995 ; Duffy et al. 1998 ), earlier sleep times ( Costa et al. 1986 ; Czeisler et al. 1992 ; Carrier et al. 1997 ; Duffy et al. 1998 ) and morningness as assessed by the MEQ ( Mecacci et al. 1986 ; Wilson 1990; Drennan et al. 1991 ; Monk et al. 1991 ; Adan 1992; Czeisler et al. 1992 ; Carrier et al. 1997 ).
The Circadian Type Questionnaire (CTQ) was originally developed in an attempt to identify which individuals would adjust readily to shift work, and was based on the premise that variations observed in shift-work adjustment are due to underlying differences in circadian rhythm phase, amplitude and strength ( Folkard et al. 1979 ). Since its initial development, it has been revised and renamed the Circadian Type Inventory (CTI; Barton et al. 1995 ). The CTI is an 18-item questionnaire with two independent factors. Individuals who score high on the first factor (Flexible/Rigid) are ‘flexible’ with the ability to stay awake at odd times of day or night. Those who score high on the second factor (Languid/Vigorous) are ‘languid’ and tend to report difficulty overcoming drowsiness, especially in the morning. One study found that eveningness on the MEQ corresponded to high scores on CTQ flexibility and low scores on CTQ languidity ( Smith et al. 1989 ). However, there have been few published studies about circadian rhythm correlates with the current version of this questionnaire, the CTI. One recent study found a trend toward those with higher CTI flexibility scores having larger temperature rhythm phase shifts during simulated night shift-work ( Martin and Eastman 1998).
The broad personality dimension of extraversion is associated with a later oral temperature rhythm rise and fall ( Blake 1967; Colquhoun and Folkard 1978; Eysenck and Folkard 1980) and eveningness ( Patkai 1971; Horne and Ostberg 1977; Larsen 1985; Mecacci et al. 1986 ; Adan and Almirall 1990; Wilson 1990; Neubauer 1992). However, the literature in this area is somewhat complicated due to the changing nature of the concept of extraversion ( Rocklin and Revelle 1981). Some studies have suggested that impulsivity (which some consider a subcomponent of extraversion) produces the relationship to later circadian phase and evening tendencies ( Eysenck and Folkard 1980; Neubauer 1992). Also, differences in performance in response to caffeine interact with impulsivity and time-of-day in a pattern suggesting that high impulsives have a later circadian phase ( Revelle et al. 1980 ). Conversely, other studies have suggested that it is the sociability subcomponent of extraversion that is associated with eveningness ( Larsen 1985; Wilson 1990).
Individual differences in temperature rhythm amplitude have also been investigated. According to oscillator theory, the strength of a circadian oscillator is reflected by its amplitude ( Wever 1965; Aschoff and Pohl 1978). Thus, it may be easier to shift weaker oscillators (smaller amplitudes) with a shift of the zeitgeber ( Aschoff and Pohl 1978). In other words, weaker oscillators may re-entrain faster and perhaps produce greater adjustment and tolerance to shift work. However, the research in this area is inconsistent. Some studies found that greater tolerance to shift work was associated with a smaller temperature rhythm amplitude, but others found the opposite (see Harma 1993 for a review). A major problem with many studies in this area is that the data tend to be masked by sleep and shift-work schedules. Other findings include that with older age, there is a decrease in the amplitude of the circadian temperature rhythm ( Weitzman et al. 1982 ; Vitiello et al. 1986 ; Czeisler et al. 1992 ), although this relationship may be stronger for men than for women ( Monk et al. 1995 ), and that the amplitude in women is dependent upon the phase of the menstrual cycle ( Lee 1988; Kattapong et al. 1995 ).
The purpose of our study was to relate individual differences in circadian phase and amplitude to the questionnaires and variables described above in a large sample of subjects for whom multiple days of rectal temperature recordings were available.