Longitudinal studies of diurnal variations in depression: a sample of 643 patient days

Intraday Activity Levels May Better Reflect the Differences Between Major Depressive Disorder and Bipolar Disorder Than Average Daily Activity Levels

It is important to establish an objective index to differentiate mood disorders (i.e., bipolar disorder; BD and major depressive disorder; MDD). The present study focused on the pattern of changes of physical activity in the amount of activity intraday, and examined the relationship between activity patterns and mood disorders. One hundred and eighteen inpatients with MDD or BD in a depressive state provided the activity data by using wearable activity trackers for 3 weeks. In order to illuminate the characteristic patterns of intraday activities, Principal Component Analysis (PCA) was adopted to extract the main components of intraday activity changes. We found that some of the PCs reflected the differences between the types of mood disorder. BD participants showed high activity pattern in the morning and low activity pattern in evenings. However, MDD showed the opposite. Our results suggest that activity tracking focused on daytime activity patterns may provide objective auxiliary diagnostic information.

Diurnal mood variation in depression: a signal of disturbed circadian function?

BACKGROUND

Diurnal variation in mood is a prominent symptom of depression, and is typically experienced as positive mood variation (PMV - mood being worse upon waking and better in the evening). The present study sought to advance understanding of PMV by measuring daily mood variation in non-clinical individuals with varying levels of depressed mood. Based on research into normative variation in mood and evidence that circadian amplitudes may be decreased in depression, it was hypothesised that compared to those with low levels of depression, individuals meeting Centre for Epidemiological Studies-Depression Scale (CES-D) cut-off for probable depression would exhibit an attenuated circadian component in diurnal variation of Positive Affect (PA).

METHOD

Ninety-nine young healthy women (mean age=21.5, SD=3.0) living on a normal sleep-wake schedule provided mood reports every two hours between 0800 h and 2200 h for 7 days.

RESULTS

The high depression group (CES-D > or = 23, n=22) exhibited a pattern of diurnal variation consistent with PMV (increased PA in the evening relative to the morning). As predicted, evidence was also found that the high depression group was characterized by a decreased circadian component to diurnal variation in PA relative to the low depression group (CES-D < 23, n=77).

CONCLUSIONS

It is provisionally concluded that diurnal mood variation in depression can usefully be understood from the perspective of weakened circadian function. Findings are discussed in terms of limitations of the study's naturalistic design and future research avenues identified.

Neuropsychological function and diurnal variation in depression

Twenty-six patients suffering from DSM-III-R major depressive episode with diurnal variation of mood were examined at approximately 8 AM and 8 PM, using a neuropsychological test battery. We found that tests of executive function, especially the verbal fluency test, were sensitive measures of diurnal variation, but that personality measures were relatively robust in that they were not significantly affected by diurnal variations of mood. The relevance of detecting and assessing diurnal changes of mood with reliable, objective measures of performance is discussed.

The circadian system and the therapeutics of the affective disorders

Establishing that a circadian rhythm is abnormal tells us little about the cause, which can arise from changes in the patient's lifestyle, irregularities of the body clock or a malfunction in the process of entrainment of the clock. In a clinical context, such a range of possible explanations implies differences in the most appropriate mode of treatment. Against this background, the conventional view that the underlying abnormality in endogenous depression is due to a disorder of the body clock is challenged. The challenge is based on difficulties of interpretation of the clinical data and the results of studies on circadian rhythms in patients. It is suggested that the state of the circadian system in depression resembles its state in healthy individuals after time-zone transitions or in shift work maladaptation syndrome and that this disturbance should be seen as resulting from changes in the phasing of external zeitgebers rather than from an abnormality in the clock itself.

A longitudinal study of diurnal mood variation in depression; characteristics and significance

The course of 39 depressed in-patients' daily mood was recorded by means of frequent self-ratings during their entire stay (in total 3718 days). The frequency of diurnal variations largely varies between subjects without clear dichotomy in 'diurnal' and 'non-diurnal' subjects and the occurrence of diurnal variations is rather irregular. Mood variability measures rather than average daily mood improvement correlate with the response to sleep deprivation. These observations do not support theories of chronobiological rhythm disturbances in depression. It is argued that depressed subjects largely vary in susceptibility to stimuli. Signals generated by the biological clock or by processes related to the sleep-wake cycle are considered examples of such stimuli.

Prediction of the antidepressant response to total sleep deprivation of depressed patients: longitudinal versus single day assessment of diurnal mood variation

The relationship between diurnal variation of mood and the clinical response to total sleep deprivation (TSD) was investigated in 43 depressed patients. The question asked was whether the propensity to produce diurnal variations of mood or the actual mood course on the day before TSD determines the clinical response to TSD. Patients rated their mood three times daily during an experimental period of 56 days. The frequency as well as the amplitude of daily mood changes were assessed during this period. For each patient six TSDs were scheduled: two after days with a positive mood course, two after a negative mood course, and two after days without a diurnal change of mood. This strategy allowed comparisons of TSD responses within patients. Moreover, longitudinally and retrospectively assessed diurnal variation were compared with each other. It was found that patients vary largely in the occurrence of diurnal variations of mood. The propensity to produce diurnal variations either in terms of frequency or amplitude was positively correlated with the response to TSD. Within patients no differences were found in responses to TSDs applied after days with diurnal variations (positive or negative) or without diurnal variations. A second aim was to get more insight into the mechanism relating diurnal variations of mood and the TSD response. Therefore, the interrelatedness of various measures of diurnal variations, such as amplitudes and frequencies of positive or negative diurnal mood changes, was studied, as well as the relationships of these variables with TSD responses. On the basis of the strong interrelatedness it is suggested that they all reflect the same underlying mechanism, to be symbolized by an oscillator, producing positive daily fluctuations of mood.

Diurnal variation: reliability of measurement and relationship to typical and atypical symptoms of depression

We used three rating scales to study diurnal variation of mood in 37 patients with major depressive disorder (17 drug-free patients and 20 treatment refractory patients on stable regimens of antidepressant medication). The three rating scales included global self-ratings administered twice a day; an itemized, prospective, observer-rated scale administered twice a day; and the retrospective item on the Hamilton Depression Rating Scale. Z scores and Intraclass Correlation Coefficients demonstrated a poor level of agreement between the itemized, prospective scale and the self-ratings. In addition, stepwise multiple regression analysis and point bi-serial correlation showed no systematic relationship between atypical diurnal variation (i.e., mood worsening in the evening) and atypical depressive symptoms (weight gain, hypersomnia, etc.), or between typical diurnal variation (i.e., mood worsening in the morning) and typical depressive symptoms (weight loss, insomnia, etc.). This lack of relationship was observed in both drug-free and medicated patients using each of the three rating scales. We discuss possible explanations for these negative findings.

Behavioral engagement level, variability, and diurnal rhythm as a function of bright light in bipolar II seasonal affective disorder: an exploratory study

Six patients with bipolar II seasonal affective disorder (SAD) and seven normal control subjects rated their moods in winter at six fixed times each week-day during 1 week of dim and 2 weeks of bright light. The scales rated represent the mood dimension specifically associated with depression, a dimension here called behavioral engagement (BE). Compared with controls, depressed SAD subjects (1) showed lower BE levels across all rating times of the day, (2) were more likely to show diurnal variation in BE, (3) displayed more between-day instability in BE diurnal rhythm, and (4) exhibited greater short-term lability (change within 3 hours) in BE. Bright light reduced or eliminated all group differences in BE level and variability.

Mood variability in adolescents: a study of depressed, nondepressed and comorbid patients

In a study to examine the variability of mood in psychiatrically disturbed adolescents, 30 inpatients aged 13-17 reported on current depressive symptoms three times a day for seven consecutive days, using a set of visual analog scales (the Adolescent Mood Scale: AMS) to record DSM-IIIR and other depressive symptoms. Ten of the patients had no depressive diagnosis; 11 had both a depressive and an 'externalizing' diagnosis (mainly conduct disorders and substance abuse disorders), and nine had depressive diagnoses but no externalizing disorder. Variability was defined in terms of (1) range of AMS scores; (2) amount of change from one test point to the next; (3) rhythmicity, measured by the autocorrelation function across 21 test points. All three groups had high levels of depressive symptoms throughout the week. On all measures of severity of depression, the depressed girls were more depressed than the depressed boys, irrespective of comorbidity. Measures of variability, however, showed no effect of sex, but comorbid patients were more likely to have a wide range of mood scores, and reported a 45% greater amount of mood change. Only five subjects had a significant lag1 autocorrelation function, and there was no indication of diurnal rhythmicity. Implications for research and diagnosis are discussed.

Towards a model of mood responses to sleep deprivation in depressed patients

It is hypothesized that in depressed patients diurnal variation in mood (DV) is a daily recurring phenomenon, which fails to achieve expression on all days (showing a random distribution of DVs). From this perspective a meta-analysis was performed on the raw data of earlier presented studies. The effect of total sleep deprivation (TSD) on mood was examined in 14 so-called prototypical patients, showing on three successive days either positive DVs (feeling better in the evening) or inverse DVs. It was hypothesized that under baseline conditions mood follows a monotonous course with switching points at 7 AM and 11 PM and that during the TSD night the 7-AM switch took place earlier. The position of this switch was calculated, assuming that (1) before the switch the curve ran parallel to the nightly baseline curves, and (2) after the switch the curve showed a monotonous change parallel to the daily baseline curves. The best fit between predicted and measured depression after TSD was found for a switch at 3 AM, varying the switching point during the TSD night with hourly intervals. The characteristics based on prototypical patients contributed significantly to the prediction of the morning and the afternoon depression levels after TSD in a group of 53 patients (prototypical and nonprototypical).

Relations between depressed mood and vocal parameters before, during and after sleep deprivation: a circadian rhythm study

The mechanism underlying improvement after total sleep deprivation (TSD) was studied in 14 major depressed patients. The suggestions that (1) circadian processes and/or (2) dimensions of arousal may play a role in the response to TSD were investigated. Diurnal variation of depressed mood and of mood- and arousal-related vocal parameters was studied in relation to the effect of TSD on depressed mood and vocal parameters. During 3 baseline days, during TSD and 2 days after TSD vocal parameters and depressed mood were assessed 6 and 3 times daily respectively. The mean fundamental frequency (frequency of vocal fold vibration, F0) (presumably reflecting aspects of arousal) as well as the range of the F0 (proposed to reflect sadness) showed a clear circadian pattern with a peak at about 4.00 p.m. TSD affected the circadian organization of the mean F0 and advanced the peak of the curve. After one night of subsequent sleep this effect disappeared. In addition, improvement after TSD coincided with an increase of the mean F0. The diurnal variation of mood before TSD predicted the mood response to TSD, whereas diurnal variation of vocal parameters did not. Moreover, circadian changes in vocal parameters were not related to changes in depressed mood. These findings suggest that the diurnal variations in mood and vocal parameters are regulated by different mechanisms. Data support the presumption that circadian as well as arousal processes are involved in the mood response to TSD. Circadian changes in vocal parameters due to TSD are not likely to reflect changes in the biological clock.

Prediction of the antidepressant response to total sleep deprivation by diurnal variation of mood

The relationship between diurnal variation of mood and response to total sleep deprivation (TSD) was investigated in 131 depressed patients. This response was related to (1) the diurnal variation on the day before TSD as assessed by self-ratings of mood, and (2) the propensity to produce diurnal variations (the "diurnality") as assessed by the Hamilton Rating Scale for Depression. Three types of diurnality are distinguished: morning type (the propensity to feel better in the morning), evening type (the propensity to feel better in the evening), and a nondiurnal type. The results show that diurnality does predict the mood response to TSD. The direction of diurnality is decisive: patients who have the propensity to feel better in the evening benefit more from TSD than other patients.

The daily course of the symptomatology and the impaired time estimation in endogenous depression (melancholia)

The importance of 'typical' diurnal variations in mood (depression worse in the morning) as a diagnostic criterion of endogenous depression has been challenged in previous investigations. Disturbance in time experience may contribute towards an understanding of diurnal variation in depressive symptomatology. To examine this hypothesis a series of time estimation experiments was conducted in the course of the day. Twenty-five endogenous depressive patients (according to ICD-9) and 12 healthy controls were asked to estimate prospectively a 30-s interval on two successive days at 7.30 a.m., 11.30 a.m., 3.30 p.m. and 7.30 p.m. Simultaneously the subjects assessed their state of well-being using a visual analogue mood scale. A circadian rhythm of time estimation errors could not be detected; even in daily courses with 'typical' diurnal variations an increasingly favourable self-assessment of well-being was not accompanied by a corresponding diurnal fluctuation of time estimation. These results cast additional doubts on the significance of 'typical' diurnal variations in depressive symptomatology.

Frequency of symptoms in melancholia (depressive illness)

Symptoms of depression were quantitatively recorded using a rating scale, for 239 men and 260 women. The frequency of the various symptoms were similar for the sexes, with a high frequency for anxiety, both psychic and somatic. The results show that the symptoms of depressive patients differ from those found in the classic descriptions.

Is diurnal variation a meaningful symptom in unipolar depression?

The present study found that defining groups of patients based on the presence or absence of diurnal variation highlighted several important differences between the two groups. Diurnal patients had longer episodes, better response to treatment, and were less frequently classified as RDC agitated. Within this diurnal population, evening sufferers had more decrease in appetite than morning sufferers. Equivalent percentages of diurnal-a.m. and diurnal-p.m. received RDC endogenous diagnoses, but more diurnal-a.m. patients were found in the DSM-III melancholia category. While this study found some evidence inconsistent with the generally accepted notion regarding the constellation of symptoms associated with endogenous and non-endogenous depressions, the authors suggest that diurnal variation may represent a predictor of unipolar depression which is biologically based.

Construct validity of the DSM-III and RDC classification of melancholia (endogenous depression)

Diagnostic schedules for any psychiatric disorder should represent a coherent concept and demonstrate adequate construct (internal) validity and transferability. Testing fit to latent class models (e.g. by Rasch model fitting test), correspondence to these requirements can be assessed. Based on data from a sample of 173 depressed in-patients, diagnostic schedules for endogenous depression (melancholia) according to DSM-III and RDC are evaluated for their ability to fit the requirements listed above. According to the data presented, the set of DSM-III-criteria for melancholia represents a coherent concept; on the contrary, this is not the case for RDC-criteria. Furthermore, application of the diagnostic algorithm applied to the defining criteria according to DSM-III has been justified; again, this was not the case for the diagnostic algorithm according to RDC. However, a simple linear diagnostic algorithm emerges from our data.

Ward environment and disturbed behaviour

Disturbed behaviour and dissatisfaction were studied in relation to two contrasting ward environments, one representing the 'medical model' of psychiatry and the other being a modified form of 'therapeutic community'. Specially designed forms were used for 18 months to obtain daily and weekly measures of behavioural disturbance from the nurses and of dissatisfaction with various aspects of the ward environment fom the patients. The results suggested that the 'therapeutic community' ward was significantly more disturbed, and that disturbance on that ward varied according to the day of the week, something which did not happen on the 'medical model' ward. Differences in disturbance between diagnoses were also found. Disturbance was significantly correlated with, and tended to precede, dissatisfaction only in the 'therapeutic community' ward.

Diurnal mood fluctuation and age

Normal adolescents and young adults tended to report better mood toward the evening; whereas middle-aged and elderly persons tended to report better mood in the morning. Limited evidence suggests the opposite trends for depressed psychiatric patients. The findings were related to theoretical considerations.

[Diurnal variations in hospitalised depressive patients during depression and when healthy (author's transl)]

Diurnal variations during the depressive phase and healthy periods were investigated in 84 hospitalised depressive patients of different nosological diagnosis. The occurrence of different rhythm-types in this population led to the conclusion that depression induces rhythmicity: those belonging to the arhythmic group when healthy showed significant increase in rhythmicity when depressed, predominantly the classical form of diurnal variation (morning with improvement toward evening). Age and sex were found to be important factors determining diurnal variation. In the course of hospitalisation, the type of diurnal rhythm remained individually constant.

Phase advance of the circadian sleep-wake cycle as an antidepressant

Sleep in depressed patients resembles sleep in normal subjects whose circadian rhythms of temperature and rapid-eye-movement sleep are phase-advanced (shifted earlier) relative to their sleep schedules. If this analogy is relevant to the pathophysiology of depressive illness, advancing the time of sleep and awakening should temporarily compensate for the abnormal timing of depressed patients' circadian rhythms. Four of seven manic-depressive patients studied longitudinally spontaneously advanced their times of awakening (activity onset) as they emerged from the depressive phase of their illness. In a phase-shift experiment, a depressed manic-depressive woman was twice brought out of depression for 2 weeks by advancing her sleep period so that she went to sleep and arose 6 hours earlier than usual. The antidepressant effect of the procedure was temporary and similar in duration to circadian desynchronization induced by jet lag in healthy subjects. This result supports the hypothesis that abnormalities of sleep patterns in some types of depression are due to abnormal internal phase relationships of circadian rhythms.

Diurnal variations in intensity of symptoms in patients of different diagnostic groups

Consecutive patients, regardless of diagnostic group, upon admission to the Department of Psychiatry, University of Umea, were asked to fill in a special form about diurnal variation in their symptomatology and sleep disturbances. Three trained psychiatrists separately reviewed the records and completed a second form comprising information about diagnosis and certain background factors. Among the background factors only age and ongoing antidepressive medication were found to be of importance. In the study no strong consistency in the symptom complex of early awakening, experience of symptomatology as worst in the morning, and a feeling of relief in the afternoon was found. However, the experience of symptomatology as a whole, anxiety and inhibition as being worst in the morning is significantly more common in patients with unipolar and bipolar affective disorders but is also common in all diagnostic subgroups and seems in no way to be specific for depressive disorders. Furthermore, the phenomenon of diurnal variation of symptomatology was not found to be of diagnostic help and should be employed very cautiously as a basis in pathogenic consideration.

The use of autocorrelation analysis in the longitudinal study of mood patterns in depressed patients

The statistical method of autocorrelation, commonly used in econometrics and engineering, was applied to the daily mood scores of ten depressive hospital in-patients. The analyses made possible the quantification of two aspects of the longitudinal course of individual patients' psychopathology, the degree of day-to-day stability and the degree of periodicity in mood. Quantification of the degree of day-to-day mood stability yielded wide variations between patients and suggested that patients might be usefully categorized in terms of this characteristic. Mood stability during periods of severe depression was found to be less pronounced than during periods of relatively moderate depression. Furthermore, the existence of 'mini-cycles', cyclical fluctuations in mood of one to two weeks' duration occurring during the course of depressive episodes, was demonstrated in three cases.