Mood Correlates with Circadian Alignment in Healthy Individuals

Abstract

Objective:

To determine whether there is a correlation between mood and the alignment between the timing of the circadian pacemaker (circadian phase) and the timing of sleep in healthy, euthymic individuals.

Methods:

Participants were 25 first-year medical students (25.9 ± 3.3 years, 16 females). Mood (Profile of Mood States, brief form, POMS-B) and circadian phase (salivary dim light melatonin onset, DLMO) were assessed 4 times over 7 weeks. Circadian alignment was determined using the DLMO to average midsleep interval (Phase Angle Difference or PAD).

Results:

POMS-B score and PAD were correlated: later DLMO relative to midsleep (shorter PADs) was associated with worse mood (F_1,75=10.953, p=0.001). There was no difference in POMS-B score between females and males and no interaction between gender and PAD.

Conclusions:

There is a correlation between circadian alignment and mood among healthy individuals as has been found in individuals with seasonal and non-seasonal depression. This finding has implications for the development, prevention and treatment of mood disorders.

Introduction

The hypothalamic circadian pacemaker is known to influence the timing of mood in addition to the myriad other near-24-hour rhythms under its control. Specifically, mood has been demonstrated to be at its best during the biological day and at its worst during the biological night when assessed using both simple visual analog scales¹ and the Positive and Negative Affect Schedule (PANAS).² More recently, we have demonstrated in healthy controls that both positive and negative affect have endogenous rhythms with positive affect peaking during the biological day and negative affect peaking during the biological night when measured using both the PANAS and the Profile of Mood States, brief form (POMS-B), both good measures of mood fluctuation within the normal range.³

It is also well-established that aberrant timing of the circadian pacemaker relative to the timing of physiologically important behaviors such as sleep and wakefulness can have adverse health consequences.⁴ Such circadian misalignment commonly occurs with shift work, transmeridian travel or shifts in sleep-wake timing on weekends or work-free days (termed social jet-lag).^5,6 It is therefore logical to suppose that circadian alignment might also have an adverse impact on mood. Indeed, we and others have shown in both seasonal⁷ and non-seasonal Major Depressive Disorder (MDD)^8–10 that later timing of the circadian pacemaker (measured using the dim light melatonin onset or DLMO) relative to the timing of average midsleep during the prior week [a shorter time interval or Phase Angle Difference (PAD)] is associated with worse mood (although in a significant minority of seasonal MDD patients longer PADs were associated with worse mood).⁷ We have hypothesized that shorter PADs result in the peak in negative affect and the minimum in positive affect being shifted into the waking hours with a resultant worsening of mood.³ If this is indeed the case, then it would be expected that shorter PADs in healthy individuals would also be associated with worse mood. Our aim was to test this hypothesis in a cohort of healthy people free of mood disturbance (i.e., euthymic individuals).

Methods

Participants

Participants were 25 first-year medical students (25.9 ± 3.3 years, 16 females), at Oregon Health & Science University (OHSU) who volunteered as part of their participation in a student research project. All participants provided written informed consent and the protocol was approved by the Institutional Review Board at OHSU. Participants were required to have no clinically significant health concerns (free of any cardiac, pulmonary, hepatic, renal, endocrine, neurological or psychiatric disease) as determined by a Health and Screening Questionnaire. One participant was excluded because she was taking an antidepressant medication, two because they were routinely exposed to light during the night, and two because assessments of circadian timing were not obtained. The participants were taking daytime classes and none were engaged in night shift-work.

Protocol

Participants maintained a sleep/wake schedule of their choosing for seven weeks that they documented using a written sleep/wake diary. No limitations were placed on the timing or duration of sleep. After one week and every two weeks thereafter, participants were admitted to the Oregon Clinical and Translational Research Center at OHSU on a Friday for an assessment of mood and circadian timing (a total of 4 admissions over the 7 weeks). Each admission lasted 6½ hours, beginning 6 hours prior to the average bedtime for the prior week, and light levels remained <10 lux at the angle of gaze throughout. Participants were studied from mid-January to early March in cohorts of 6-10 individuals over the course of 3 years.

Assessment of Circadian Timing and Mood

During each admission hourly saliva samples were collected for 6 hours beginning 5.5 hours prior to the average bedtime for the prior week. No restrictions were placed on posture prior to sampling. Melatonin concentrations were measured by radioimmunoassay (ALPCO Ltd., Windham, NH). The lower limit of sensitivity was 0.5 pg/ml. Circadian phase (the timing of the circadian pacemaker) was determined using the DLMO, defined as the interpolated time when salivary melatonin crossed a 3 pg/ml threshold.^11,12

At the start of each admission, mood was assessed using the POMS-B.¹³ The POMS-B Total Mood Disturbance (TMD) score was used as the measure of mood. Midsleep was calculated from the sleep diaries as the midpoint between bedtime and wake time. Circadian alignment (PAD) was determined using the time interval, in hours, between the DLMO and the average midsleep of the prior week. To account for repeated measures, a generalized linear mixed model was used to test for an association between PAD and TMD using SPSS 28.0 (SPSS, Inc.) with gender and subject as fixed effects.

Results

Participants had average (± SD) bedtimes and waketimes of 23:31 ± 0:32 and 07:18 ± 0:28, respectively. The mean TMD was 18.3 ± 12.6 which is consistent with scores in healthy control populations and below that found in disordered cohorts.¹³ The average time of the DLMO was 21:13 ± 01:17 and the average PAD was 6:12 ± 01:00 hours (consistent with averages in normal historical control populations^14,15). Generalized mixed model analysis showed a correlation between TMD and PAD: shorter (more phase-delayed) PADs were associated with worse mood (F_1,75=10.953, p = 0.001, Figure 1, relevant model coefficients provided in Supplemental Table). There was no difference in TMD between females and males (F_1,75=1.630, p=0.206), nor was there a significant interaction between gender and PAD (F_1,75=0.888, p=0.349).

Figure 1.

Relationship between mood as measured by total mood disturbance (TMD) score on the Profile of Mood States, Brief Form (POMS-B) and the Dim Light Melatonin Onset (DLMO)-to-Midsleep Interval in hours (Phase Angle Difference, or PAD), a measure of circadian alignment. Higher TMD scores represent worse mood while smaller PAD values represent a shorter DLMO-to-Midsleep Interval (indicating phase-delayed internal circadian alignment). Each data point represents the average of 4 mood and DLMO assessments. Standard errors are displayed for the TMD means.

Discussion

We have found a correlation between mood and circadian alignment among healthy, euthymic individuals. Specifically, we found that a shorter interval between the timing of the circadian pacemaker and that of sleep was associated with worse mood. An important feature of this finding is that circadian timing and circadian alignment, like mood, were all within the normal range in this cohort. Our findings are consistent with those in non-seasonal MDD^8,9 but contrast somewhat with our previous findings in seasonal-MDD, and we cannot rule out the possibility that a larger data set would reveal a subgroup of healthy individuals in which worse mood is associated with phase-advanced circadian alignment (i.e., longer PADs).⁷

The findings are limited by the small sample size (which limits the number of effects that can be estimated in the mixed model analysis), by the lack of menstrual cycle data among the female subjects, and by the particular population studied. The findings might not generalize to other populations (e.g., older individuals). Nor have we demonstrated a causal relationship between circadian alignment and mood. Nonetheless, if these findings are causal and generalizable, then they have a number of important implications. It is possible that phase-delayed changes in circadian alignment could contribute to the development of mood disorders. This is in keeping with the large body of work suggesting that the circadian system, and later circadian timing in particular, plays a role in the development of mood disorders including the associations between evening chronotype (preference for later sleep timing) and greater social jet-lag with depression symptoms.^4,16 Some evening chronotypes have been shown to have later circadian timing relative to sleep (i.e., shorter PADs)^15,17,18 and this may help explain the association of chronotype with depression symptoms.

We did not find a gender-specific association between circadian alignment and mood. Such a finding might be expected given the higher prevalence of depression among females, the demonstrated differences in circadian physiology between males and females^19,20 and previous findings showing gender-specific associations between PAD and depression severity in individuals with non-seasonal MDD who were treated with fluoxetine and time-in-bed restriction.¹⁰ Despite the lack of a statistically significant gender-specific association, Figure 1 suggests a greater association between PAD and TMD score among females, and a larger data set might be better able to assess this question. We speculate that certain populations (e.g., females) may be more vulnerable to a worsening of mood with phase-delayed changes in circadian alignment and that this may contribute to the development and/or perpetuation of mood disorders in these groups.

These findings may have implications for the prevention of mood disorders as well. Light exposure in the biological evening and early night induces phase delays,²¹ and even dim, everyday levels of evening light can contribute to phase delays.^6,22 The shift in sleep to a later time and the associated circadian phase delay that occur on weekends is specific to conditions of artificial light,⁶ and therefore it is possible that self-selected patterns of light exposure contribute to mood disorder risk in vulnerable individuals. These individuals may benefit, in terms of decreased risk of mood disorders or simply a more positive mood, by curating their light exposure to obtain more light exposure in the morning and less light exposure in the evening. Additional studies examining this question are warranted.^23,24 These data are also relevant to the use of phase-advancing oral melatonin^7,25 and morning bright light^14,26 for treating most patients with existing seasonal and/or non-seasonal MDD.

We have found a correlation between circadian alignment and mood among healthy, euthymic individuals. Our findings contribute to the growing scientific literature demonstrating a relationship between internal circadian alignment and mood in healthy individuals, as well as in those with schedule- or light exposure-induced changes in circadian timing and in patients with mood disorders.