Abstract
Previous research has demonstrated a clear link between late chronotype and depression. The vulnerability factors underpinning this link, however, are unclear. Here the relationship between two specific emotion regulation strategies, cognitive reappraisal and expressive suppression, and chronotype was investigated using multiple regression. Two hundred and forty participants (age range 18-80, 189 females) completed validated self-report questionnaires assaying chronotype, neuroticism, depressive symptomatology, sleep quality, and emotion regulation. Eveningness was associated with increased expressive suppression, and morningness was associated with increased cognitive reappraisal after controlling for age, gender, depressive symptomatology, neuroticism, and sleep quality. Trait expressive suppression and reduced cognitive reappraisal are known to increase depression risk. Our results suggest that eveningness is associated with impaired emotion regulation, which may confer risk for future depression. These findings suggest modifiable markers that could be therapeutically targeted to prevent the onset of depression in late chronotype individuals.
Individual differences in sleep-wake behavior (morningness-eveningness) and their relation to psychiatric disorders have attracted increasing attention in recent years (Fabbian et al., 2016). In particular, several studies have reported an association between eveningness and depression. Across the general population, late chronotype is associated with increased depressive symptomatology (Levandovski et al., 2011; Merikanto et al., 2013, 2015; Togo et al., 2017), diagnosis of depression, and use of antidepressant medication (Merikanto et al., 2013, 2015). Together, these data demonstrate a clear association between eveningness and depression, but as they are observational, they are insufficient to indicate the mechanisms whereby clinical depression emerges in late chronotype individuals.
A cardinal feature of depression is sustained negative affect. Indeed, meeting the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) criteria for a diagnosis of major depressive disorder (MDD) requires the presence of either sustained negative affect or loss of pleasure for a 2-week period. Individual differences in the habitual use of emotion regulation strategies may, therefore, play an important role in the onset and maintenance of depression, and theorists argue that a failure to effectively manage emotional responses to everyday events leads to more prolonged and severe periods of distress, which in turn evolve into clinical depression (Aldao et al., 2010).
Two specific emotion regulation strategies that have received significant attention are cognitive reappraisal and expressive suppression (Gross and John, 2003). Cognitive reappraisal, which involves the generation of a benign or more positive interpretation of a stressful event in order to reduce distress, is generally considered beneficial. Cognitive models of depression (Beck et al., 1979) place maladaptive or ineffective reappraisal strategies at the core of depression, and cognitive-behavioral therapies (CBT) for depression highlight adopting effective, adaptive reappraisal strategies. By contrast, expressive suppression, which involves inhibiting the emotional expression of the emotional experience, is generally considered maladaptive. Although effective at reducing the overt expression and possibly the subjective experience of emotion in the short-term, chronic use of this approach would be less effective at reducing emotional and physiological arousal (Gross and John, 2003). More frequent use of expression suppression and inefficient reappraisal has been reported in nonclinical groups (Gotlib and Joormann, 2010), currently depressed patients (Campbell-Sills et al., 2006), and patients in remission (Aker et al., 2014; Ehring et al., 2008). More recently, emotion dysregulation (as measured by the Difficulties in Emotion Regulation Scale [DERS]) has also been demonstrated in highly neurotic individuals (Paulus et al., 2016). As high neuroticism is a widely recognized risk factor for depression (Vittengl, 2017), ineffective emotion regulation strategies may underlie, in part, the increased vulnerability of highly neurotic individuals for depression. However, to our best knowledge, no one has yet explored emotion regulation in relation to chronotype. Here, we hypothesize that eveningness, which is associated with depression, will also be associated with reduced cognitive reappraisal and increased expressive suppression.
Methods
Participants
A total of 240 participants (mean age = 25.97 years, SD = 11.73 years, 189 females) completed an online survey (Qualtrics, Provo, UT). Of these, 150 were university students who received course credit for participation. The remaining participants were drawn from the general population and recruited via social media and personal communication. The survey was available between November 1, 2016, and January 31, 2017, and participants were free to complete the study at a time of their choosing, with 75% of surveys completed between 0800 h and 1800 h. All procedures were approved by the local Research Ethics Committee (REC), and all participants provided informed consent prior to completing the survey.
Questionnaires
Chronotype was assessed using the 5-item Reduced Morningness-Eveningness Questionnaire (rMEQ; Adan and Almirall, 1991). The rMEQ requires participants to indicate their response on a 4-point scale (e.g., “During the first half hour after you wake up in the morning, how do you feel?” “Very tired,” “fairly tired,” “fairly refreshed,” and “very refreshed”). This scale has good internal consistency (α = 0.86), and higher scores indicate increased morningness. Sleep quality was determined using the Pittsburgh Sleep Quality Index (PSQI; Buysse et al., 1989). This scale produces a global sleep quality score, which is created by summing each of the subscales: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbance, use of sleeping medication, and daytime dysfunction. The PSQI has high internal consistency (α = 0.83). Items are answered either using a 4-point scale or by indicating time (e.g., “During the last month, how often have you had trouble sleeping because you wake up in the middle of the night or early morning?” “Not during the last month,” “less than once a week,” “once or twice a week,” “more than 3 times a week”). Higher scores on the PSQI indicate worse sleep quality, with scores greater than 5 indicating clinically poor sleep quality. Neuroticism was assessed using the 12 neuroticism-related items taken from the revised Eysenck Personality Questionnaire (EPQ-R; Eysenck et al., 1984). Participants are required to make a binary (yes/no) response to each item, and neuroticism is calculated as the sum of “yes” responses. Emotion regulation was measured using the Emotion Regulation Questionnaire (ERQ; Gross and John, 2003). The ERQ is a 10-item questionnaire designed to measure cognitive reappraisal and expression suppression (6 and 4 items each). Participants were instructed to indicate how they control their emotions by responding to each of the 10 items using a 7-point scale ranging from “strongly agree” to “strongly disagree.” The Cronbach alpha scores of the 2 subscales (reappraisal and suppression) are, respectively, 0.81 and 0.73. Depressive symptomatology was assessed using the Centre for Epidemiological Studies Depression Scale (CES-D; Radloff, 1977). Participants were required to indicate how often (rarely, some of the time, a moderate amount of time, all of the time) they have felt a certain way in the last 7 days (e.g., “During the past week I felt that everything I did was an effort”). The CES-D has high internal consistency (α = 0.85). Participants were also asked to indicate current or previous diagnosis of depression or sleep disorder.
Statistical Treatment
Simple Pearson’s correlations were used to examine associations between demographic characteristics, sleep quality, neuroticism, depressive symptomatology, current/previous depression, current/previous sleep disorder, and chronotype. If significant at the bivariate level, these predictors were included in a regression framework (Model 1). Model 2 subsequently examined the impact of emotion regulation (reappraisal and expressive suppression) on chronotype. All statistical analyses were carried out using the Statistical Package for the Social Sciences (SPSS) v21 (IBM; New York, NY).
Results
Basic demographics and sample characteristics are shown in Table 1. Cognitive reappraisal and expressive suppression correlated with rMEQ at the bivariate level (respectively, r240 = 0.237, p < 0.001; r240 = −0.199, p = 0.002). Age, gender, depressive symptomatology, neuroticism, and sleep quality were significantly correlated with rMEQ and were thus entered as control variables in Model 1 (see Table 2). This model accounted for 18.4% of the variance in chronotype (F5,234 = 10.56, p < 0.001). Increasing age was associated with morningness, whereas depressive symptomatology was associated with eveningness. Model 2 included age, gender, depressive symptomatology, neuroticism, and sleep quality as well as cognitive reappraisal and expression suppression. This model explained 24.3% of the variance in chronotype and was a significantly better fit than Model 1 (F7,232 = 10.66, p < 0.001). In this final model, morningness was associated with increased reappraisal and reduced expression suppression (Table 3).
Descriptive statistics.
The table shows basic demographics, self-reported current/previous depression and sleep disorder, neuroticism (as indexed by the EPQ-R), depressive symptomatology (CES-D), sleep quality (PSQI), chronotype (rMEQ), and cognitive reappraisal and expressive suppression (ERQ). See the Methods for questionnaire details. Values are mean (SD) or n (%). Also included are questionnaire ranges (minimum-maximum) for the study sample and original scale.
Bivariate correlations between chronotype and sample demographics, personality, depressive symptomatology, and sleep quality.
p < 0.001.
Regression table.
R2 for Model 1 = 4.29; ΔR2 for Model 2 = 0.059 (both p values < 0.001).
p < 0.011. *p = 0.001. **p < 0.001.
Inspection of standard residuals suggested that the data were not confounded by outliers (standard residual minimum = −2.42, standard residual maximum = 2.50). Assumptions of collinearity were met, as tolerance values for all variables were greater than 0.1 (0.452-0.947) and variance inflation factor (VIF) scores were less than 10 (1.06-2.21). The data met the assumption of independent errors (Durbin-Watson value = 1.99), and a histogram of standardized residuals indicated that the data contained approximately normally distributed errors as did the normal P-P plot of standardized residuals. A scatterplot of standardized predicted values showed that the data met the assumptions of homogeneity of variance and linearity. Finally, the data also met the assumption of non-zero variances (13.18-137.23).
Discussion
A burgeoning corpus suggests that late chronotype is associated with depression, antidepressant use, and depressive symptoms. Disordered affect (characterized by sustained negative affect and reduced experience of positive affect) is a cardinal feature of depression, and individual differences in emotion regulation have been linked to increased risk for depression. Here we show that eveningness was associated with maladaptive emotion regulation strategies (reduced cognitive reappraisal and increased expression suppression) after controlling for age, gender, depressive symptomatology, neuroticism, and sleep quality. Inefficient emotion regulation may, therefore, underlie in part the increased vulnerability of late chronotypes for psychiatric disorder.
Recent important population studies provide good evidence for a link between late chronotype and depression (Merikanto et al., 2013, 2015). However, the observational nature of those studies precludes determining the underlying psychological mechanisms that underpin this link. The vulnerability-stress hypothesis of depression posits that depression emerges through an interaction between a psychological vulnerability (e.g., preferential processing of negative material) and an environmental stressor (e.g., death of a loved one, job loss), and one approach to explore this hypothesis is to investigate cognitive biases in “at-risk” groups. Using this method, Chan and colleagues (2007) reported that high neuroticism (a recognized risk factor for depression) was associated with reduced latency to classify negative versus positive personality descriptors, reduced positive memory intrusion at subsequent recall, and reduced capacity to recognize happy facial expressions compared with individuals with low levels of neuroticism. Using the same “at-risk” model with late chronotypes, Berdynaj and colleagues (2016) reported increased recognition of sad facial expressions, greater recall and reduced latency to correctly recognize previously presented negative personality trait words, and reduced allocation of attentional resources to happy faces in late chronotypes. Similarly, Horne et al. (2017) reported increased recognition accuracy in late chronotypes for sad compared with happy faces independent of sleep quality, mood, age, and gender. Emerging evidence (Berdynaj et al., 2016; Horne et al., 2017), therefore, suggests that in healthy, never-depressed individuals, late chronotype is associated with negative biases in emotional processing which, if combined with adversity, may lead to depression.
In the current study, we also found that morningness was associated with adaptive antecedent-focused emotion regulation strategies (cognitive reappraisal) where action is taken before the emotion response tendency is fully activated. By contrast, eveningness was associated with expressive suppression (generally considered maladaptive), a response-focused strategy that is instigated once an emotion is underway. The consequences related to this pattern of emotion regulation are likely far reaching. Expressive suppression is considered cognitively more demanding, as this approach requires constant updating to manage and maintain emotional responses, thereby monopolizing valuable cognitive resources that would be better spent on other tasks (John and Gross, 2004). In addition, the conflict generated between the overt “controlled” behavior and inner emotion and the sense of not being true to oneself may itself lead to increased negativity, thereby contributing to continuing depressed mood (Sheldon et al., 1997).
To test the affective consequences of reappraisal and suppression, Gross (1998) invited participants to view movie clips designed to elicit negative emotions. Participants in the control group simply watched the film (no instruction), in the reappraisal condition participants were instructed to think about the film in such a way that they would feel nothing, and in the suppression condition participants were instructed to behave in such a way that an independent viewer would not know they were feeling anything. Compared with the control condition, both suppression and reappraisal were effective in reducing emotion-expressive behavior. Suppression, however, had no effect on the subjective experience of negative emotion. That is, although outwardly participants in the suppression group appeared unmoved, they felt as much emotion as the no-instruction, control group. In addition, physiologically, suppression was associated with increased sympathetic activation (increased heart rate and galvanic skin response). By contrast, reappraisal reduced both the experience and expression of negative emotion and had no impact on physiological arousal. These data and results from related studies (Gross and Levenson, 1993, 1997) suggest that expressive suppression, here associated with eveningness, does not lead to a reduction in the experience of emotion and may feed into a spiral of despair and negativity. Cognitive reappraisal, however, here associated with morningness, has a positive impact and is effective at reducing the subjective experience of negative emotion.
The current findings and previous work suggest that eveningness is associated with impaired emotional processing and regulation (Berdynaj et al., 2016; Horne et al., 2017) and suggest possible interventions for the prevention of depression in late chronotype individuals. However, the underlying causes that lead to depression are likely multifactorial, and there is a need for longitudinal studies to explore and determine effective strategies that promote psychological well-being in these individuals. For example, experimental manipulations that allow evening-types to follow, or better match, their circadian rhythm may be effective in reducing depressive symptomatology. Alternatively, correcting phase disturbance with bright morning light, melatonin, or melatonin agonists may be useful to improve mood. Indeed, agomelatine, an agonist at melatonergic (MT1/MT2) receptors, has been demonstrated to be an effective antidepressant (Kasper et al., 2010).
In addition to our novel findings of reduced cognitive reappraisal and increased expressive suppression in evening type individuals, our data are in good agreement with previous research exploring chronotype. For example, we observed significant correlations between age, gender, depressive symptomatology, neuroticism, and sleep quality. Morning types tended to be older, to be female, to report lower depressive symptomatology and neurotic traits, and to have better sleep quality, data that are consistent with previous work in this field (Duggan et al., 2014; Merikanto et al., 2013, 2015; Randler et al., 2016).
A number of limitations associated with this study should be taken into account when considering the findings. First, chronotype status was determined using a single metric, the Reduced Morningness-Eveningness Questionnaire (Adan and Almirall, 1991); future studies may benefit from objective tools such as acrophase estimates of cortisol and core body temperature, polysomnography, and actigraphy to corroborate self-report questionnaire data. Also, the ability of individuals to accurately self-report emotion regulation strategies has been scrutinized (Robinson and Clore, 2002). Respondents may be influenced by negative mood or other biases; they may also confound the experience of emotion with its regulation, and the degree of insight and meta-cognition required to complete the questionnaires may be challenging for some participants. Future work that includes experimental measures to assess the affective consequences of different emotion regulation strategies (e.g., viewing video clips designed to elicit negative emotion; John and Gross, 2004) supported by self-report questionnaires is warranted. In addition, our anonymous online survey required that current and/or previous depression and diagnosed sleep disorder were determined by self-report. Future studies that include reference to medical history, structured clinical interview, and other known risk factors for depression such as family history of depression and stressful life events are required. Also, our study sample was predominantly female (78%), and it has been reported that emotion regulation strategies vary as a function of the menstrual cycle, with the mid-late luteal phase and follicular phase being associated with reduced cognitive reappraisal compared with the menstrual phase (Wu et al., 2014). Here, we did not record menstrual cycle phase, nor did we conduct repeat testing (e.g., to compare emotion regulation during the follicular and luteal phase). We cannot, therefore, exclude a potential impact of cycle phase on our findings. Finally, a limited number of studies have explored the potential link between eveningness, depression, and body mass index (BMI). In adolescent females, Pabst and colleagues (2009) reported that eveningness was associated with greater depressive symptomatology, as measured by the Children’s Depression Inventory, in both normal and overweight girls, although the effect was more marked in overweight individuals. More recently, data from a large adult population study by Merikanto et al. (2015) suggest that eveningness was associated with reduced BMI. Moreover, stratification of morning, intermediate, and evening types by number of depressive symptoms revealed that, compared with morning types, evening type individuals with none or one depressive symptom had lower BMI scores whereas BMI did not distinguish between chronotype groups when greater levels of depression (two symptoms present) were noted (Merikanto et al., 2015). Taken together (Merikanto et al., 2015; Pabst et al., 2009), these data suggest that the relationship between circadian preference, depressive symptomatology, and BMI is complex and varies as a function of age. In the current study, we did not measure BMI. However, the observed association between eveningness and reduced cognitive reappraisal/increased expressive suppression remained significant after controlling for depressive symptoms. Future studies may benefit from including measures of BMI, especially if the target population includes adolescents or young adults.
In conclusion, the current findings demonstrate impaired emotion regulation in late chronotypes that may confer risk for future depression. Specifically, we found that morningness was associated with cognitive reappraisal, widely theorized to be protective against psychopathology, and eveningness was associated with expressive suppression, which has been argued to be a risk factor for psychopathology. These findings suggest modifiable markers that could be therapeutically targeted to prevent the onset of depression in late chronotype individuals.
Footnotes
Conflict of Interest Statement
The author(s) have no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.