Poor sleep quality, sleep deprivation, and excessive daytime sleepiness are common in medical students, especially during clinical clerkship [1-3]. Sleep disturbances may partly reflect the fact that the normal circadian rhythm clashes with early clinical schedules. Circadian preference is associated with specific personality patterns in young adults [4-7]. The morning-type preference has been reported to be associated with higher conscientiousness and agreeableness and lower openness to experience, extraversion, and neuroticism [6]. In terms of the latter, morningness tends to be associated with low-level neuroticism [4,8], although some studies have reported a negative or no relationship between morningness and neuroticism [7,9]. Eveningness is associated with depression [10-12], eating disorders [13], anxiety [14], addiction [15-17], and personality disorders [4,7,18].

Morningness has been found protect against sleep-related psychological problems in childhood and adolescence [19]. It has been suggested that the chronotype of young adults be considered in developing programs to prevent mental health problems [4]. Thus, it is important to evaluate the chronotype of medical students to determine whether it is related in important ways to personality. As it has been established that a shift toward morningness occurs after early adulthood [20,21], it is also crucial to determine whether significant changes in circadian preference are likely to occur during clinical clerkship. Changes in chronotype may be related to certain defenses, and to changes in defense mechanisms and defense styles during clerkship.

Therefore, we explored the defense styles, defense mechanisms, and chronotypes of medical students before and after clinical clerkship. We expected that certain defenses and/or changes in defenses would be associated with chronotype and/or with changes in chronotype. We hypothesized that meaningful changes in circadian preference would occur through clinical clerkship, and these changes may be associated with the use of certain defenses and/or changes in defenses.

The MEQ score increased (i.e., a change toward morningness) from year 2 to 4 (W=9,085.5, p=0.018, r=0.18). The MEQ scores at year 2 differed among year-groups I–III; post hoc analysis showed that group III had a higher mean MEQ score than did groups I and II (Table 1). The MEQ scores at year 4 did not differ significantly among the year-groups, except that group II exhibited a significantly higher MEQ score than did group I (Table 1). Changes toward morningness from years 2 to 4 were significantly correlated with splitting at year 2 (ρ=-0.29, p<0.001, N=176), changes in splitting from year 2 to 4 (ρ=0.19, p=0.010, N=174), consumption at year 4 (ρ=-0.17, p=0.026, N=177), and use of the immature defense style at year 4 (ρ=-0.15, p=0.043, N=176). Multiple linear regression analysis showed that less splitting at year 2 predicted a greater change toward morningness from year 2 to 4 after controlling for age and sex (Table 2). Another multiple linear regression analysis showed that female sex, less splitting at year 2, and lower use of the immature defense style at year 4 were significantly associated with more change in the direction of morningness over the 2 years of clerkship after controlling for age, depression, and sleep quality at year 4 (Table 3). In terms of individual defenses, less splitting at year 2 and less consumption at year 4 were significantly associated with a greater increase in morningness from year 2 to 4 after controlling for age, sex, depression, and sleep quality at year 4 and with changes in splitting from year 2 to 4 (Table 4). We then explored whether more change in the direction of morningness truly predicted less use of consumption and of the immature defense style at year 4 (Tables 5 and 6). Multiple linear regression analysis showed that splitting at year 2 and differences in morningness significantly predicted consumption at year 4 after controlling for age, sex, depression, and year-4 sleep quality (Table 5). Another multiple linear regression analysis revealed that increases in morningness and decreases in splitting from year 2 to 4 significantly predicted less use of the immature defense style at year 4 after controlling for age, sex, depression, and sleep quality at year 4 (Table 6).

Circadian preference at year 4 was significantly correlated with year-4 use of the immature defense style (ρ=-0.22, p=0.003, N= 180), sublimation (ρ=0.18, p=0.015, N=181), consumption (ρ=-0.19, p=0.011, N=181), and passive aggression (r=-0.23, p=0.002, N=181) and with year-4 sleep quality (ISI scores; ρ=-0.17, p=0.023, N= 178), rising time misalignment between weekdays and weekends (ρ=-0.32, p<0.001, N=179), weekend oversleep (ρ=-0.26, p<0.001, N=181), and mean sunlight exposure from 10:00 to 15:00 (ρ=0.18, p=0.032, N=143). Multiple linear regression showed that female sex, better sleep quality, less use of passive aggression, and greater sublimation at year 4 were significantly associated with morningness after controlling for year-group, depression, and sleep-habit related variables; the model explained 29% of the variance in MEQ scores at year 4 (Table 7).

We found a significant change in favor of morningness by year 4. The within-year-group differences were less marked at year 4 than at year 2. The change toward morningness was significantly related to year-2 splitting after controlling for age and sex. Thus, students who exhibited less splitting at baseline moved toward morningness when experiencing stressful, difficult, but exciting clinical situations. The clinical educational schedule can act as a zeitgeber as clinical work commences 30–60 min earlier than preclinical lectures [34]. Eveningness is usually related to difficulties in managing environmental and social demands, and these problems may worsen when educational schedules become stressful [35,36]. Moreover, eveningness is reportedly associated with poor self-control [37] and impulsivity [38], reflecting brain serotonergic functioning [39,40]. Students who engage in splitting may exhibit less self-control, including poor self-regulation of bedtime; thus, habits compatible with clerkship schedules may be difficult to sustain, limiting any move toward morningness, or even creating a move toward eveningness, during the period of clinical training. Morningness may be socially desirable; morning types manage time cues and arousal/sleep–wake rhythms better [41,42]. Eveningness may be associated with poor attention to time cues and weak entraining of arousal to a sleep–wake cycle appropriate for clinical schedules [41,42]. Therefore, a move toward morningness during clinical clerkship may protect against stressful conditions [43].

Another regression analysis showed that less splitting at year 2, less use of the immature defense style at year 4, and female sex were significantly associated with a change toward morningness after controlling for age and year-4 depression and sleep quality. Decreased use of an immature defense style was associated with less neuroticism [44] and less depression, as well as a decreased rate of eating and personality disorders [27,45,46]. We found that students who moved toward morningness during clinical clerkship tended to employ an immature defense style less frequently. Females have been reported to be more morning-oriented than males during puberty and early adulthood [7,47], and the shift to morningness during puberty occurs earlier in females than in males [47]. However, other studies found no sex differences or even greater eveningness in females [20,48].

In terms of specific defense mechanisms, less splitting at year 2 and less consumption at year 4 were significantly related to a shift toward morningness over time. Interestingly, consumption at year 4 was predicted by a change toward eveningness over time, as was less splitting at year 2. Students who use consumption as a defense may employ alcohol or tobacco to resolve or relieve inappropriate impulses or needs. Eveningness has been associated with substance use and addiction in adults [15-17]. As the circadian preference, circadian gene polymorphisms, and brain reward circuitry may be interrelated, eveningness may directly increase substance abuse [49]. As those who favor eveningness tend to act out and criticize traditional standards [41], a move toward eveningness may increase consumption and substance abuse, especially in those who employ splitting less prior to clinical training. Our previous cross-sectional results from a subpopulation of the present study showed that year-2 eveningness was associated with greater consumption, less sublimation, and male sex [5]. Furthermore, a greater decrease in splitting and a move toward morningness during clinical clerkship were significantly related with less use of the immature defense style at year 4 in this study. Splitting is a typical image-distorting defense [27] used primarily by those with borderline personality disorder [50,51]. Students who gradually abandon splitting as clinical learning proceeds may increasingly integrate their self-image with other images, directly reducing use of the immature defense. Moreover, students who attain increasingly better control of the sleep–wake cycle, which then becomes compatible with their social schedules, may gradually forsake use of the immature defense style.

Morningness at year 4 was associated with greater sublimation, less passive aggression, better sleep quality, female sex, and year-group (i.e., group I, II, and III) after controlling for depression; it was also associated with rise time misalignment between weekdays and weekends, weekend oversleeping, and mean sunlight exposure between 10:00 and 15:00. Sublimation is a sophisticated way to express instinctual urges that reduces intrapsychic conflict while minimizing any marked loss of pleasure. In sublimation, aggressive impulses are expressed via pleasurable games, sports, or hobbies, and libidinal impulses are expressed in romantic relationships. Unlike neurotic defenses, such as repression, the instincts are channeled rather than dammed or transformed [52,53]. Sublimation is a mature defense style associated with the Big Five personality traits of openness, emotional stability, and extraversion [54]. High levels of openness [8,55] and emotional stability [55] are associated with morningness. Although no relationship between extraversion and morningness has been described [34,36], several studies have found significant associations between lower levels of extraversion and morningness [56,57]. Tankova et al. [34] indicated that morningness and extraversion may be related through impulsivity, which is closely associated with serotonergic functioning [38,39]. As sublimation is both an ability and a mental activity controlling diverse internal impulses, an association between greater sublimation and morningness would support a relationship between low impulsivity and morningness. Less use of passive aggression was also associated with morningness in our fourth-year students. Passive aggression is a defense against aggression whereby the latter is expressed indirectly and ineffectively. Individuals who are hostile toward another person exhibit certain behaviors, including procrastination, failure to communicate, and provocation, as expressions of aggression [52]. The DSQ considers passive aggression an immature defense style that is negatively correlated with conscientiousness, agreeableness, and emotional stability [54]. Higher levels of conscientiousness [8,56-58], agreeableness [39,58-60], and emotional stability [55] are associated with morningness. Sublimation and passive aggression are defense mechanisms that seek to deal with instincts that must be self-controlled and self-regulated, both of which are enhanced by morningness [37]. Consistent with previous studies [10,61,62], we found that morningness was associated with better sleep quality. As greater sublimation, lower levels of passive aggression, and better sleep quality may play adaptive roles in terms of enhancing clinical activity, encouraging medical students to develop morning-type sleep–wake cycles may be crucial. We found no significant associations between year-4 circadian preference and rise time misalignment between weekdays and weekends, weekend oversleeping, or mean daily sunlight exposure in a multiple regression analysis. A future study should explore these topics in a larger population.

Our study had several limitations. First, as chronotype is regarded as a relatively stable trait in adulthood, the changes in chronotype during the 2-year study period might have occurred by chance. Thus, future prospective studies with larger samples are needed to explore whether there are real changes in chronotype during early adulthood. Second, the influences of environment were not considered in this study. Future studies considering other factors that may have an impact on changes in chronotype are needed. Third, the relationship between consumption/immature defense style and increased morningness may have been bidirectional in nature, but cause-and-effect inferences cannot be drawn.

Despite these limitations, our study has several strengths. First, this is the first study to explore the effects of personality change on chronotype in medical students. To the best of our knowledge, we here describe, for the first time, the relationship between changes in defense mechanisms and styles throughout the course of clinical clerkship with increased morningness. Second, we suggest that natural chronotype modification using social synchronizers might reduce the use of consumption and other immature defense styles.

In conclusion, increased morningness and a gradual reduction in splitting during clinical clerkship may reduce the use of the immature defense style. We suggest that this finding be considered when attempting to reduce medical students’ immaturity via changes in chronotype.