The moderating role of personality traits in the relationship between chronotype and depressive symptoms

ABSTRACT

Due to the undeniably morning orientation of the social clock, the evening chronotype can be associated with negative consequences, both at the affective and cognitive levels. Evening-oriented individuals are more susceptible to affective disorders, show poorer educational achievements and consume stimulants more often than morning-oriented individuals. However, little is known about potential factors that may attenuate or amplify these negative emotional consequences of the evening preference. Thus, our aim was to examine whether personality traits interplay with chronotype in predicting depressive symptoms. We assessed the Big Five and the Big Two personality traits, morningness-eveningness and depressive symptoms in an online sample of 913 Polish individuals (468 females, 445 males), aged 18–35 (M = 26.34, SD = 5.15). Eveningness, higher neuroticism, lower conscientiousness and lower alpha-stability were associated with higher depressive symptoms. The magnitude of the association between eveningness and depressive symptoms decreased with higher conscientiousness and alpha-stability, as well as with lower neuroticism. In conclusion, high neuroticism, low conscientiousness and low alpha-stability increase the risk of depressive symptoms, particularly among evening chronotypes. The patients’ chronotypes and personality traits should be taken into account in both the prevention and diagnostics of depression.

KEYWORDS:

• Chronotype
• morningness-eveningness
• personality
• depressive symptoms
• Big Five
• Big Two

Introduction

Our life is organized and bounded by a 24-hour day. The circadian rhythm exerts a prepotent influence on human behavior and psychological functioning (e.g., Jankowski and Ciarkowska 2008; Schmidt et al. 2007). The basic loop of this rhythm is the sleep/wake cycle. Individual differences in this cycle are considered to be an elementary indicator of chronotype and are also referred to as morningness-eveningness preference when analyzed as a continuous dimension. Morning chronotypes (M-types) wake up and go to bed early, showing high regularity in their sleep/wake cycle. Furthermore, they achieve the highest mental and physical effectiveness during the morning hours. In contrast, evening chronotypes (E-types) tend to wake up and go to bed later, achieving the highest efficiency of functioning in the late afternoon or at night. Moreover, their sleep patterns are often irregular. Differences between M-types and E-types’ are not limited to the timing of sleep and waking up but are also associated with a variety of other physiological, psychological and behavioral characteristics (Adan et al. 2012), including susceptibility to affective disorders (Gao et al. 2019). E-types are more prone to developing depressive disorder (Hidalgo et al. 2009; Merikanto et al. 2015; Van Den Berg et al. 2018), anxiety disorder (Antypa et al. 2016), and bipolar disorder (Melo et al. 2017), compared to M-types.

The current literature offers several possible explanations of the association between diurnal preferences and depression, including biological, genetic or neuronal differences (e.g., Archer et al. 2014; Etain et al. 2011). Those approaches are supplemented with the perspective of the misalignment between individuals’ biological and social clocks (Wittmann et al. 2006). People are most alert, energetic and effective at a particular time of day (Stolarski et al. 2016). The period of the most efficient functioning is relatively stable in a given individual (apart from a few systematic shifts throughout life), but at the same time, it varies from person to person, depending on their chronotype (Adan et al. 2012). As a result of how modern society’s schedule is organized (especially in regard to work and school systems), people are rarely at liberty to freely select the time of their activity. Consequently, they often have to work, study and perform at times incongruent with their chronotypes. Operating at a non-optimal time of day is associated with negative consequences, both at the cognitive (Nowack and Van Der Meer 2018) and at the affective (Kivelä et al. 2018) level.

People who have to operate at times of day that are incongruent with their chronotype often manifest diminished alertness, impaired reasoning ability and lesser cognitive effectiveness (Díaz-Morales and Escribano 2015), and show poorer educational achievement (Goldin et al. 2020). They also have an elevated propensity to maladaptive behaviors such as higher consumption of stimulants (alcohol, caffeine), smoking cigarettes (e.g., Patterson et al. 2016; Zhang et al. 2018), and other unhealthy dietary habits that may lead to obesity (Muscogiuri et al. 2020; Ruiz-Lozano et al. 2016). Due to an undeniably morning orientation of the social clock, these issues apply especially to E-types and may be viewed as a way of coping with the affective consequences of operating at a non-optimal time of day (Wittmann et al. 2006). E-types report higher levels of tension and lower levels of pleasure across the whole day compared to M-types (Jankowski and Ciarkowska 2008), and show greater difficulties in self-regulation (Owens et al. 2016), together with a reduced ability to regulate negative emotions (Watts and Norbury 2017).

Many of abovementioned cognitive and affective declines may stem from a deterioration in sleep quality, more frequent sleep disturbances and insufficient sleep duration (Müller et al. 2016). However, the relationship between eveningness and depressive symptoms cannot be fully explained by sleep parameters, as recent studies indicate that it is neither direct nor straightforward (Üzer and Yücens 2020) and that a later chronotype does not predict a permanent course of depression (Druiven et al. 2019). In the light of the above, many researchers (e.g., Antypa et al. 2017; Van Den Berg et al. 2018) suggest that additional factors should be taken into account while attempting to untangle the relationship between diurnal preference and depression. These conclusions provide a starting point for the present study in which we aim to look for other factors influencing this relationship.

Given that functioning at a time incongruent with one’s preferences may make individuals more prone to negative affect or maladaptive behaviors, it seems fundamental to ask whether certain variables may attenuate or amplify the association between diurnal preference and depressive symptoms. As individual dispositions are important regulators of behavior (Strelau 1996), we hypothesized that both potential risk and protective factors can be found in the field of personality. State-of-the-art research shows that one’s personality traits can change throughout life due to aging, major life changes or experiences but also through targeted interventions (Robert et al. 2017), thus exploring this area seems even more justified.

One of the most widely used personality conceptualizations is the Five-Factor Model (also referred to as the Big Five) (Costa and McCrae 1992), which assumes the existence of five main personality dimensions: openness to experience, conscientiousness, extraversion, agreeableness and neuroticism. The associations between the Big Five personality traits and diurnal preference have been well-established. Extraversion, agreeableness, and conscientiousness are consistently positively associated with morningness, whereas neuroticism and openness to experience are related to greater eveningness preference (Randler et al. 2017). In addition, conscientiousness is the trait in which E-types and M-types show the highest discrepancy (Tonetti et al. 2009). Personality traits are also associated with depression: neuroticism shows a positive association, while extraversion and conscientiousness typically manifest inverse relationships (Klein et al. 2011).

In the light of the rationale linking the pressure to operate in a non-optimal time of day with depression, various personality traits can be viewed as attenuating and amplifying the undesirable consequences of diurnal preferences. Conscientiousness and neuroticism seem to be of particular importance due to the fact that both are systematically related to chronotype (Randler et al. 2017) and depressive symptoms (Klein et al. 2011) but also because, depending on their level, both show clearly adaptive or maladaptive properties (Bogg and Roberts 2013; Watson and Casillas 2003). Even though E-types on average display lower conscientiousness and higher neuroticism compared to M-types (Drezno et al. 2019), the variability in these traits may still potentially affect how they adapt to functioning in the non-optimal time of day. The protective effects of conscientiousness may stem from the adaptive nature of its facets, such as dutifulness, order or self-discipline, which may support functional and healthy ways of coping with the effects of operating in the incongruent time of day (i.e., decreased mood and lower cognitive capacity) (Preckel et al. 2011). Furthermore, in more conscientious and less neurotic people, the number of substances used decreases (Turiano et al. 2012), which may be due to less impulsive and more thoughtful ways of coping with operating in the non-optimal time of day. In addition, high neuroticism is a strong predictor of various mental and physical disorders (Lahey 2009), whereas low neuroticism is linked with resilience to anxiety or depression (Lipnevich et al. 2017). Therefore, it seems that both lower neuroticism and higher conscientiousness may protect against the development of, as well as support coping with, depression.

Although there is a vast body of evidence for the validity of personality traits as predictors of a wide variety of psychophysiological, behavioral and well-being criteria, it remains debatable if these five factors constitute the simplest and most accurate level of personality description since they are intercorrelated and show high consistency as higher-order factors: alpha and beta, often referred to as the Big Two (DeYoung et al. 2002; Digman 1997). Alpha includes emotional stability (reversed neuroticism), agreeableness and conscientiousness, and as it captures the stability of functioning in the emotional, interpersonal and motivational contexts it is also referred to as stability. The beta factor consists of extraversion and openness to experience and is also known as plasticity because it depicts the flexibility and adaptability of one’s cognition and behavior (DeYoung et al. 2002). Researchers indicate strong biological bases of the Big Two, where plasticity is associated with variability in dopaminergic function and stability is associated with variability in serotonergic function (DeYoung 2006). Serotonin is also strongly involved in the modulation of circadian rhythms (DeYoung et al. 2007), and plays an important role in the regulation of emotions and is associated with depression (Meltzer 1990); therefore alpha-stability may be viewed as a potentially protective factor against the undesirable emotional consequences of eveningness.

Based on THE theoretical considerations and research results discussed above, we hypothesized that lower neuroticism, higher conscientiousness and higher alpha-stability may significantly reduce depressive symptoms. We further predicted that the magnitude of this effect is greater in more evening-oriented individuals, as the adaptive features of low neuroticism (i.e., emotional stability), high conscientiousness and high alpha-stability seem particularly vital for individuals forced to cope with the negative consequences of evening preference. To the best of our knowledge, no previous study has explored the potential moderating role of the Big Five and the Big Two personality traits in the relationship between diurnal preference and depressive symptoms.

Methods

Participants

A total of 982 adults took part in the study. In order to ensure the high quality of the data, we excluded all entries from participants who failed at least one of the several attention checks, did not finish the survey, or were identified as multivariate outliers (based on Mahalanobis distance exceeding the critical value at p < .001 level) (Tabachnick and Fidell 2007). Finally, data from 913 Polish individuals (468 females, 445 males) aged 18–35 (M = 26.34, SD = 5.15) were analyzed. Participants were recruited online from the Polish research panel Pollster and completed the measures described below online, through the Qualtrics platform. Participants were rewarded with points, which they could exchange for small gifts.

To the best of our knowledge, no previous studies investigated interactions between chronotype and personality traits in predicting depressive symptoms. Given that generally, the incremental variance explained by the chronotype*personality interaction terms are rather low (Drezno et al. 2019; Gorgol et al. 2020), we planned the study to have a sample size large enough to detect ΔR² in hierarchical regression equal to .01. Power analyses indicate that this can be achieved with a sample size of 779 (α = .05, power = .80; Faul et al. 2009). We opted to expand the sample size by about 20%, to account for possible missing data, failed attention checks and multivariate outliers.

Assessment

Chronotype was measured with the Composite Scale of Morningness (CSM; Smith et al. 1989) using the Polish adaptation of the scale (Jankowski 2015). The CSM contains 13 questions referring to various aspects of circadian functioning, such as sleep-wake rhythm and preferred hours of physical and mental activity. Higher scores indicate greater morningness, while lower scores indicate greater eveningness. Internal consistency of the CSM measured with Cronbach’s α amounts to .84, showing high reliability of this scale.

The Big Five personality traits were measured with the Polish adaptation (Rowiński et al. 2014) of the International Personality Item Pool of the Revised NEO Personality Inventory (IPIP-NEO-PI-R; Goldberg et al. 2006). IPIP-NEO-PI-R consists of 90 items that examine five scales: openness to experience, conscientiousness, extraversion, agreeableness and neuroticism. Each scale comprises six subscales. This questionnaire has a five-point Likert-type response format, ranging from 1 – ‘it describes me inaccurately’ to 5 – ‘it describes me accurately.’ The Cronbach’s alpha reliability coefficient for the Polish version of this questionnaire ranges between .72 and .85, showing sufficient internal consistency for research purposes. To compute the two higher-order personality factors, we followed Digman’s (1997) and DeYoung et al.’s (2002) instructions: for alpha-stability we summed the scores for the agreeableness, conscientiousness, and inverted neuroticism (emotional stability), while for the beta-plasticity we summed the scores for extraversion and openness to experience.

Depressive symptomatology was assessed with the Patient Health Questionnaire-9 (PHQ-9; Kroenke et al. 2001) using the Polish translation developed by the MAPI Research Institute. The PHQ-9 is a screening tool for determining the risk of depressive disorders. It consists of nine items that refer to the frequency of depressive symptoms described in the DSM-IV diagnostic criteria over the past two-week period. Participants answered questions on a scale from 0 – ‘not at all’ to 3 – ‘nearly every day.’ Higher scores indicate greater depressive symptoms. Psychometric evaluation of the PHQ-9 supports its use as a valid metric with high internal consistency (α = .89).

Statistical analyses

To uncover the associations between all variables in the study, we calculated partial correlations, controlling for age and gender (see Table 1). To verify our main hypotheses and investigate the interplay between personality traits and morningness-eveningness in predicting depressive symptoms, we conducted a series of regression analyses, with age and gender included in the models as covariates. Next, in each model, we further investigated the potential moderating role of the chronotype and the given personality trait. Finally, we used the Johnson-Neyman technique to identify the regions of significance for the observed interactions.

Table 1. Means, standard deviations, Cronbach’s alphas and intercorrelations between variables included in the present study, controlling for age and gender (N = 913)

	M	SD	α	1.	2.	3.	4.	5.	6.	7.	8.	9.
1. Morningness-Eveningness	34.07	6.94	.86	-
2. Depressive symptoms	17.45	5.64	.86	−.30*	-
3. Openness to experience	3.30	.44	.69	−.09**	.03	-
4. Conscientiousness	3.41	.52	.84	.31**	−.37**	.17**	-
5. Extraversion	3.04	.57	.84	.10**	−.20**	.31**	.24**	-
6. Agreeableness	3.52	.51	.81	.14**	−.16**	.14**	.36**	.08*	-
7. Neuroticism	3.15	.61	.86	−.21**	.51**	−.03	−.38**	−.46**	−.05	-
8. Alpha-stability	9.80	1.15	.88	.31**	−.50**	.16**	.81**	.38**	.62**	−.71**	-
9. Beta-plasticity	6.34	.82	.82	.02	−.13**	.75**	.25**	.86**	.13**	−.34**	.35**	-

^⁎p < .05; ^⁎⁎p < .001

All statistical analyses were conducted using IBM SPSS 26.0.0.1 for Windows, along with Hayes (2017) PROCESS macro for moderation analyses. All the figures were created using Python 3.9.2 software.

Ethics

The investigation was conducted in line with the ethical standards of the Declaration of Helsinki and its amendments. The procedure has been accepted by the Research Ethics Committee at the institution to which the first author was affiliated (approval number: 18/02/2021). All participants gave informed consent to participate in the study.

Results

Correlational analyses

We conducted a correlational analysis to provide initial insight into the nature of the associations between the variables included in the study: morningness-eveningness and depressive symptoms, as well as Big Five and Big Two personality traits, controlling for age and gender. The matrix of partial correlations between the measured variables, as well as the reliability of the psychometric measures, and descriptive statistics are provided in Table 1. Morningness-eveningness was associated with all personality traits. The strongest relationships between diurnal preference and personality traits were observed for conscientiousness (correlating positively with morningness) and neuroticism (correlating positively with eveningness). Moreover, eveningness was positively related to depressive symptoms. Depressive symptoms were positively correlated with neuroticism and negatively with conscientiousness. Furthermore, morningness was positively related to alpha-stability but showed no association with beta-plasticity. Lastly, depressive symptoms correlated negatively with both alpha-stability and beta-plasticity.

Regression models

In order to test our main hypotheses and investigate the interplay of personality traits and morningness-eveningness in predicting depressive symptoms, we conducted a series of hierarchical regression analyses. In step 1, age, gender and morningness-eveningness were introduced to the model. In step 2 a given personality trait was entered. Next, in step 3 we added an interaction term to explore the potential moderating effects. In Table 2, we present two such models in which interaction terms between the personality trait and chronotype were significant: for conscientiousness (model 1) and neuroticism (model 2) (models in which the interactions were nonsignificant are provided in supplementary materials). In model 3, to test for the independence versus redundancy of the two effects, we explored the joint effects of conscientiousness and neuroticism (step 2), as well as their potential moderating effects (step 3). Finally, given that the latter analysis provided evidence for the redundancy of the two interaction effects, and that both neuroticism and conscientiousness are components of one higher-order personality factor, an interaction between chronotype and alpha-stability was tested in model 4 (see Table 2). As recommended by Hayes (2017), interaction terms remained in the model even if they were not significant, as it allows the results to be compared and coherently interpreted.

Table 2. Summary of hierarchical regression analysis for variable predicting depressive symptoms

Model 1. Moderator: conscientiousness
	Variable	B	SE	β	t	LLCI	ULCI	F	R^2
Step 1	Gender	−1.70	.34	−.15**	−4.94	−2.37	−1.02	56.93	.158
	Age	−.16	.04	−.15**	−4.66	−.23	−.09
	M-E	−.24	.03	−.30**	−9.39	−.29	−.19
Step 2	Gender	−1.65	.33	−.15**	−5.02	−2.29	−1.00	69.26	.234
	Age	−.11	.03	−.10**	−3.29	−.18	−.04
	M-E	−.17	.03	−.20**	−6.37	−.22	−.11
	C	−3.21	.34	−.30**	−9.47	−3.88	−2.54
Step 3	Gender	−1.63	.33	−.15**	−5.00	−2.28	−.99	57.52	.241
	Age	−.11	.03	−.10**	−3.31	−.18	−.05
	M-E	−.17	.03	−.21**	−6.47	−.22	−.12
	C	−3.18	.34	−.30**	−9.42	−3.85	−2.52
	M-E x C	.12	.04	.08*	2.89	.04	.19
Model 2. Moderator: neuroticism
	Variable	B	SE	β	t	LLCI	ULCI	F	R^2
Step 1	Gender	−1.70	.34	−.15**	−4.94	−2.37	−1.02	56.93	.158
	Age	−.16	.04	−.15**	−4.66	−.23	−.09
	M-E	−.24	.03	−.30**	−9.39	−.29	−.19
Step 2	Gender	−.16	.32	−.01	−.51	−.78	.46	121.21	.348
	Age	−.10	.03	−.09**	−3.21	−.16	−.04
	M-E	−.16	.02	−.20**	−6.99	−.21	−.12
	N	4.41	.27	.47**	16.27	3.88	4.94
Step 3	Gender	−.16	.32	−.01	−.49	−.78	.46	98.73	.352
	Age	−.10	.03	−.09**	−3.31	−.16	−.04
	M-E	−.16	.02	−.20**	−6.99	−.21	−.12
	N	4.37	.27	.47**	16.14	3.84	4.90
	M-E x N	−.08	.03	−.07*	−2.46	−.15	−.02
Model 3. Moderators: conscientiousness and neuroticism
	Variable	B	SE	β	t	LLCI	ULCI	F	R^2
Step 1	Gender	−1.70	.34	−.15**	−4.94	−2.37	−1.02	56.93	.158
	Age	−.16	.04	−.15**	−4.66	−.23	−.09
	M-E	−.24	.03	−.30**	−9.39	−.29	−.19
Step 2	Gender	.31	.31	−.03	−.98	−.92	.31	104.71	.366
	Age	−.08	.03	−.07**	−2.59	−.14	−.02
	M-E	−.13	.02	−.16**	−5.54	−.18	−.09
	C	−1.67	.33	−.16**	−5.06	−2.31	−1.01
	N	3.91	.29	.42**	13.76	3.36	4.47
Step 3	Gender	−.30	.31	−.03	−.95	−.91	.32	76.91	.373
	Age	−.08	.03	−.07**	−2.65	−.14	−.02
	M-E	−.13	.02	−.16**	−5.62	−.18	−.09
	C	−1.64	.33	.42**	−5.02	−2.29	−1.00
	N	3.89	.28	−.15**	13.70	3.33	4.44
	M-E x C	.08	.04	.06*	2.00	.00	.17
	M-E x N	−.04	.04	−.03	−1.10	−.12	.03
Model 4. Moderator: alpha-stability
	Variable	B	SE	β	t	LLCI	ULCI	F	R^2
Step 1	Gender	−1.70	.34	−.15**	−4.94	−2.37	−1.02	56.93	.158
	Age	−.16	.04	−.15**	−4.66	−.23	−.09
	M-E	−.24	.03	−.30**	−9.40	−.29	−.19
Step 2	Gender	−1.30	.31	−.12**	−4.22	−1.91	−.70	109.07	.325
	Age	−.09	.03	−.08*	−2.82	−.15	−.03
	M-E	−.13	.02	−.16**	−5.30	−.18	−.08
	S	−2.17	.15	−.44**	−14.96	−2.45	−1.88
Step 3	Gender	−1.29	.31	−.11**	−4.19	−1.89	−.68	90.54	.333
	Age	−.09	.03	−.08*	−2.90	−.15	−.03
	M-E	−.13	.02	−.16**	−5.27	−.17	−.08
	S	−2.17	.14	−.44**	−15.08	−2.46	−1.89
	M-E x S	−.06	.02	.09**	3.38	.02	.09

Note. B = unstandardized beta coefficient. SE = standard error of B. β = standardized regression coefficient, LLCI = 95% lower limit confidence interval; ULCI = 95% upper limit confidence interval, M-E: morningness-eveningness, N: neuroticism, C: conscientiousness, S: Stability, x = interaction between variables

^⁎p < .05; ^⁎⁎p < .001

In order to provide a deeper insight into the obtained interaction effects, a floodlight technique (Spiller et al. 2013) was applied. To identify the Johnson-Neyman region(s) of the moderator where the effect of the primary variable was (non)significant, Hayes (2017) PROCESS software was applied.

Moderator: conscientiousness

The interaction between morningness-eveningness and conscientiousness was significant and indicated that with higher conscientiousness, the relationship between eveningness and depressive symptoms was weaker. Model 1 explained around 24% of the variance in depressive symptoms.

The floodlight technique revealed that the Johnson-Neyman point (i.e., the threshold for significance of the effect of focal predictor, i.e., morningness-eveningness, on the outcome variable, i.e., depressive symptoms) was located at 4.19 (0.78 after centering) in conscientiousness. This means that from low values of conscientiousness up to this point (which corresponds to the value of +1.50 SD above the mean) the association between eveningness and depressive symptoms was significant, whereas above this point eveningness was not a significant predictor of depressive symptoms. The interaction is presented in Figure 1, panel A, whereas the Johnson-Neyman regions illustrating the threshold of significance for the simple effects of morningness-eveningness on depressive symptoms for different levels of the moderator (conscientiousness) is provided in Figure 1, panel B.

Figure 1. The interaction between morningness-eveningness and conscientiousness in predicting depressive symptoms (panel A) and Johnson-Neyman regions representing the threshold for significance of the effect of the focal predictor (morningness-eveningness) on the outcome variable (depressive symptoms) for different levels of moderator (conscientiousness) (panel B).

Moderator: neuroticism

The interaction between morningness-eveningness and neuroticism was significant and indicated that with lower neuroticism, the relationship between eveningness and depressive symptoms was lower. Model 2 explained around 35% of the variance in depressive symptoms.

Further exploration revealed that the Johnson-Neyman point for the significance of the effect of the focal predictor (morningness-eveningness) on the outcome variable (depressive symptoms) was located at 2.18 (–0.97 after centering) in neuroticism. This means that from low values of neuroticism to this point (which corresponds to the value of – 1.59 SD below the mean) the association between eveningness and depressive symptoms was not significant, whereas above this point eveningness was a significant predictor of depressive symptoms. The interaction is presented in Figure 2, panel A, whereas the Johnson-Neyman regions are provided in Figure 2, panel B.

Figure 2. The interaction between morningness-eveningness and neuroticism in predicting depressive symptoms (panel A) and Johnson-Neyman regions representing the threshold for significance of the effect of the focal predictor (morningness-eveningness) on the outcome variable (depressive symptoms) for different levels of moderator (neuroticism) (panel B).

Moderators: conscientiousness and neuroticism

When both interaction terms between morningness-eveningness and conscientiousness, as well as between morningness-eveningness and neuroticism are introduced to the model, only the former remains significant in predicting depressive symptoms. However, both direct effects of conscientiousness and neuroticism continue to be significant, with the former showing the highest magnitude of the effect. Model 3 explained above 37% of the variance in depressive symptoms.

Moderator: alpha-stability

The interaction between morningness-eveningness and alpha-stability was significant and indicated that with higher alpha-stability, the relationship between eveningness and depressive symptoms was lower. Model 4 explained over 33% of the variance in depressive symptoms.

Further analysis showed that the Johnson-Neyman point for the effect of focal predictor (morningness-eveningness) on the outcome variable (depressive symptoms) was located at 10.95 (1.16 after centering) in alpha-stability, which corresponds to +1.01 SD above the mean. It indicates that from low values of alpha-stability up to this point, the association between eveningness and depressive symptoms was significant, whereas above this point eveningness was not a significant predictor of depressive symptoms. The interaction is presented in Figure 3, panel A, whereas the Johnson-Neyman regions are provided in Figure 3, panel B.

Figure 3. The interaction between morningness-eveningness and alpha-stability in predicting depressive symptoms (panel A) and Johnson-Neyman regions representing the threshold for significance of the effect of the focal predictor (morningness-eveningness) on the outcome variable (depressive symptoms) for different levels of moderator (alpha-stability) (panel B).

Moderators: other personality traits

We have also tested analogical models for the remaining personality traits (agreeableness, extraversion and openness to experience). The model with openness to experience was nonsignificant, while both direct effects of agreeableness and extraversion were significant. Next, in each model we added an interaction term to explore the potential moderating effect, however, all were nonsignificant (see supplementary materials).

Discussion

The main aim of the present study was to investigate whether personality traits have the potential to amplify or attenuate the negative emotional consequences of eveningness. Based on the conducted conceptual analysis we expected that out of the Big Five traits, conscientiousness and neuroticism would manifest the most profound effects, as they are not only systematically related with diurnal preference and depressive symptoms, but also comprise vital regulatory features with respect to emotional functioning (Watts and Norbury 2017).

All the Big Five personality traits, except for openness to experience, proved to be significantly associated with both chronotype and depressive symptoms. Among them, the strongest correlations were observed for conscientiousness and neuroticism. Our results support earlier findings regarding the higher depressive symptoms among the E-types (e.g., Levandovski et al. 2011; Merikanto et al. 2015). Moreover, both chronotype and depressive symptoms proved most strongly related to conscientiousness, neuroticism, and alpha-stability. Morningness was associated with higher conscientiousness and lower neuroticism. In contrast, higher depressive symptoms were associated with lower conscientiousness and higher neuroticism.

A more in-depth investigation of the interplay between these variables with a series of linear regression models showed that eveningness, higher neuroticism, lower conscientiousness, and lower alpha-stability were significant predictors of higher depressive symptoms. Further analyses of potential interactions between chronotype and personality traits demonstrate that the relationship between eveningness and depressive symptoms diminished with (1) higher conscientiousness, (2) lower neuroticism, and (3) higher alpha-stability. It is important to underline that a significant interaction effect does not determine which of its two components acts as a moderator – in a cross-sectional design a distinction between a focal predictor and a moderator is made arbitrarily, usually based on conceptual prerequisites (Hayes 2017). Here, the obtained effects could also be understood conversely – they may indicate that certain personality traits are particularly significant for depressive symptoms among E-types. However, in this case, the inverted interpretation of the effects seems to lead to the very same conclusion: having an adaptive personality profile is particularly important for evening chronotypes, or, in other words, being an E-type may turn out particularly costly if one manifests low conscientiousness and high neuroticism. Thus, conscientiousness, when high enough, may protect E-types from depressive symptoms. Also, lower neuroticism in E-types may be conducive to lower depressive symptoms.

A similar result also applies to the higher-order personality trait alpha-stability, which includes the features of conscientiousness, agreeableness, and emotional stability. Among E-types, having high levels of alpha-stability may be a protective factor against higher depressive symptoms. It is also worth repeating that the relationship between chronotype and alpha-stability is consistent with a neurobiological model presented by DeYoung et al. (2007) and it may be the variability in a serotonergic function that underlies the interplay between alpha-stability, chronotype, and depressive symptoms. In addition to the neurobiological basis, alpha-stability may play an adaptive role in functioning at the non-optimal time of day, as in its essence it includes adapting to social norms and the dispositions to cope with stress and negative emotions.

Gathered together, the results we acquired paint a consistent picture of how certain personality traits interplay with chronotype in predicting depressive symptoms. Besides the well-established role of neuroticism and diurnal preference in predicting depressive symptoms (Klein et al. 2011), the moderating effect of conscientiousness seems to be the most important finding in the present study (notably, this effect was significant, even if investigated together with neuroticism). As we mentioned earlier these moderations can be interpreted conversely. Thus, as conscientiousness is known to be the strongest predictor of chronotype (Tonetti et al. 2009), its role in protecting against higher depressive symptoms may possibly be viewed through its ability to shift one’s chronotype to less evening-oriented. Even though such shifts seem to be possible (Antypa et al. 2017), e.g, through cognitive/behavioral interventions or chronotherapy (Boland et al. 2019; Hong and Zhang 2019), it is vital to take into account that biological and genetic factors play an important role in shaping diurnal preference (Etain et al. 2011). Thus, such changes may merely move an evening chronotype to a “shifted” category or even conflate the improved quality of sleep with a true shift in one’s circadian rhythm (Hasler et al. 2016). Lastly, it is important to note that eveningness itself should not be viewed as a maladaptive feature, as its undesirable consequences may often stem from its interactions with other factors (such as the mismatch with the social clock and resulting necessity to operate at a non-optimal time of day). Thus, the following interpretation may prove more justified and useful.

When we interpret moderation inversely, we can argue that conscientiousness, encompassing dutifulness and self-discipline, may reduce E-types’ depressive symptoms in two main, intertwined ways: through limiting the frequency of maladaptive behaviors that are associated with eveningness and by affecting sleep hygiene (Duggan et al. 2014). People with higher self-discipline and self-control may regulate their impulses more effectively and thus avoid consuming caffeine, alcohol, and tobacco as well as other sleep-impairing behaviors in the hours just before going to sleep, thus having better sleep hygiene, which in turn improves their quality of sleep. The role of conscientiousness seems even more vital when interpreted together with research showing that this trait can be fostered and developed through various interventions (Robert et al. 2017). Even a two-week intervention may affect the level of self-discipline, a facet of conscientiousness if the accompanying behavioral change becomes habitual and one sees such change as desirable and feasible (Stieger et al. 2020). These behavioral changes at the narrow level may lead to personality change at the trait level (Robert et al. 2017). In light of these findings, diurnal preference and personality traits may be viewed as important factors to consider in both prevention and diagnostics of depression. It seems that the development of higher consciousness and emotional stability may prove to be beneficial in supplementing the main therapeutic process in evening-oriented people.

Future studies could follow the present line of research to seek personality factors that alleviate other eveningness-associated affective disorders (e.g., anxiety disorders), as well as with the use of other conceptualizations of personality (e.g., more biological concepts of temperament, see e.g., Jankowski 2012) or among specific cognitive abilities. For instance, given that certain features of emotional intelligence (Stolarski and Jankowski 2015) proved significantly higher among E-types, it would be interesting to test whether these differences reflect a more general way of dealing with negative emotional consequences of eveningness. Following the positive psychology approach (Seligman and Csikszentmihalyi 2014) it would also be essential to seek the factors that would allow E-types to compensate for their lower well-being (Randler 2008). One such dimension could be extraversion, which has been shown to moderate the link between morningness and life satisfaction (Drezno et al. 2019).

Limitations

The study has several limitations that may affect the generalization of the findings. The assessment of all variables in the study was based solely on self-report information. Even though physiological measurements of circadian rhythms, e.g., using an actigraph, would be a more accurate method to assess chronotype, conducting a study in such a large sample would not be feasible, due to its high cost. Furthermore, this study could be extended to other age groups in the future, due to the regular changes in the circadian rhythm that occur throughout the lifespan. We collected the data during a one-week period in the winter months. As it is known that low daylight intensity and less exposure to it increases the risk of depression and may lead to desynchronization of the sleep-wake rhythm, repeating the study with the same group of participants in the summer months would allow for inference about the constancy of the identified relationships and the magnitude of their seasonal variability.

Conclusions

Our results underline the moderating role of personality traits in the link between eveningness preference and depressive symptoms. High neuroticism, low conscientiousness, and low alpha-stability may increase the risk of depressive symptoms, particularly among evening chronotypes. Diurnal preference and personality traits should be taken into account in both prevention and diagnostics of depression. It seems that the focus on the development of certain personality traits may prove to be beneficial in supplementing the main therapeutic process in evening-oriented individuals.

Author contributors

Joanna Gorgol designed the study, acquired both data and funding, and drafted the initial manuscript. Joanna Gorgol, Wojciech Waleriańczyk and Maciej Stolarski analyzed the data and revised the manuscript. All authors have read and accepted the final version of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Excellence initiative – research university (2020-2026) program organized by the Ministry of Science and Higher Education [project number: PSP 501-D125-20-0004316].