Depressed patients and their notion of time

Back to summary |Download this issue

Centre de Psychiatrie
et Neurosciences
Paris-Descartes University
and Sainte-Anne Hospital
(CMME), Paris

Depressed patients
and their notion of time

by P. Gorwood,France

Time perception involves different parameters that have cognitive dimensions, such as arousal, attention, memory, and mood. The neurobiological mechanisms of time processing seem to differentiate intervals of milliseconds (which concern motor tuning, and for which the cerebellum is mainly involved), hours and days (which define circadian rhythms, with the suprachiasmatic nuclei being in charge), and seconds and minutes (required for counting and estimating time; a complex fronto-striato-thalamic circuit probably being implicated). Perception of the speed of time is slower in depressed patients, because of abnormal subjective time experience and objective time judgment, and is probably explained by the additive effects of decreased arousal, attention, and memory processes. The possibility of a direct role of mood has also not been eliminated. Abnormal perception of time and decreased speed of the internal clock are both observed in depressed patients, and give a relevant and different insight into major depressive disorder. Although numerous clues are already available (regarding clinical and neurobiological findings), the role, mechanism, and etiology of abnormal time perception in depression has probably not been studied enough.

Medicographia. 2010;32:133-138 (see French abstract on page 138)
“On the wings of time, sadness is flying…”
LA FONTAINE (The young widow, Book VI)

The capacity to analyze and adapt to the temporal parameters of a specific situation is an everyday requirement, and sometimes a life threatening necessity. We usually have to be “on time” for a large number of common activities, but fine synchrony and tuning of motor activity is particularly solicited when running away from major danger…

Time perception is needed at three major levels according to different time ranges

Temporal judgments are constructions of the brain. Defining ranges of time in time perception is particularly important, as the mechanisms seem to differ depending on the time range involved: in the circadian range, the suprachiasmatic nuclei (SCN) of the hypothalamus have a clear role, whereas the millisecond range, which may be more relevant for motor activity, may especially involve the cerebellum (Figure 1, page 134). To give a simple example, eating strawberries at lunch implies that (i) you have perceived that it is around midday; (ii) you took the decision to buy them; and (iii) you are able to transfer each strawberry successfully from plate to mouth.

Figure 1
Figure 1. Time ranges involved in the organization of behavior.

It is not surprising that time perception is a complex entity, as these three events imply the involvement of different time ranges and neurological structures, which one can more or less distinguish (Figure 1). Thus clear-cut distinctions between below-second and above-second time ranges when analyzing the neurobiology of timing may be artificial, and for some authors,1 even misleading. Proposing certain heuristic models and trying to sum up the present knowledge about the neurobiology of timing could be helpful. Nevertheless, from the start, we should distinguish between the different aspects of timing and analyze the different elements that may be involved.

Subjective time perception is influenced by mood state

Our emotional state influences the way we feel time is passing. Everybody has experienced the fact that time drags when we are feeling bored or sad, whereas time flies when we are feeling excited or happy.

Major depressive disorder, which may represent a state of extreme sadness, is defined by a series of core symptoms that include a decrease in appetite, sleep, sexual desire, energy, and psychomotor activity. The latter aspect is quoted in nearly all instruments assessing depression, and can be analyzed by a specific instrument (the Psychomotor Retardation Rating Scale). This scale, devoted to psychomotor retardation, includes one item devoted to the “patient’s perception of the flow of time.”

Apart from these clinical aspects, the role of time perception in depression has also been analyzed at the phenomenological and psychopathological levels by Janet and Minkowski. In melancholia, according to Minkowski, time lived is slower and sometimes stops altogether: the present loses clarity, the future is diminished as change seems less and less possible, and the past looms into the present in the form of guilt and regret.

The feeling in depression that time passes more slowly has received limited attention in the literature,2-9 and has sometimes led to negative results,2,5,7,10 probably because accurately assessing such complex and subjective feelings brings methodological difficulties.

The discrepancies are partly explained by differences in methodology, patients, and aspects of time perception. Nevertheless, as detailed in the next paragraphs, there are six types of evidence that the internal clock of depressed patients is abnormal. Indeed; (i) the subjective feeling of slower time experienced during depression has been detected and replicated; (ii) depressed patients have been found to have more abnormal time judgment than controls for this aspect; (iii) more severely depressed patients are slower than less severe ones; (iv) a significant correlation has been reported between the severity of depression and time estimate abnormalities; (v) when depression improves, time judgment also improves; and (vi) the opposite of depression, ie, mania, is associated with an increased speed of the internal clock.

Time perception is abnormal in major depressive disorder

Subjective time experience can be assessed with a visual analog scale by asking the subject to mark how slow or fast the flow of time is experienced at the moment of investigation. In at least two studies, this has been found to be significantly slower in depressed patients than controls.3,11 After 23 depressed inpatients completed a self-rating questionnaire of time awareness, it was also found that they felt that time passed more slowly than the same number of matched nonpsychiatric controls.5 In another study, when compared with controls,depressed patients indicatedon a verbal reportmeasure that they experienced time as passing more slowly.9 Direct questioning about the speed of time is even simpler: this was carried out in a study half a century ago, and a very significantly higher incidence of the slowing down of the experience of time was reported in the depressed state compared with the recovery state.7 Interestingly, at the clinical level, the depressed patients reported a slowing, and some even an apparent stopping, of the passage of time, describing their experiences in evocative terms (“Every hour seems a year to me”; “It is terribly slow–interminable”; “Time? It is standing still”).7

An easy way to assess “time estimation” is to ask a patient to estimate the time length for a task that has a fixed time length. Twenty-five endogenous depressive patients underestimated a 30-second interval by 6 seconds, whereas 12 healthy controls overestimated this interval by more than 10.12 In another trial, 30 severely depressed hospitalized patients overestimated 160-second, 240-second, 15-minute, and 30- minute time intervals compared with 30 controls.9 Depressed patients also overestimated time lengths of 12-minute spans compared with controls.11

“Time production” is amore proactive compound of time judgment. The subject is asked to produce a certain time span, using active demonstration of, for example, the “go” and “stop” signals. When asked to produce a 35- and 90-second time span, depressed patients were found to produce shorter time lengths (29 and 64 seconds on average) than controls.11 Grinker et al13 showed that patients with the most severe form of depression had the shortest estimation of standard durations of 1 and 3 seconds.

Quantitative approaches have also led to relatively homogeneous results. A feeling of being unwell was found to be accompanied by a more pronounced time estimation error.12 With the temporal bisection task, the higher the depression score, the shorter the signal duration was judged to be.14 Grinker et al13 also obtained a significant correlation between the individual depression scores and time estimates in a discrimination task.

In one study, when the depression score improved with treatment, an analog scale assessing subjective time experience tended to normalize.3 Depressive patients have not only been compared with healthy controls, but also with patients with mania. Interestingly, on a visual analog scale, controls report a balanced experience of the flow of time, manic patients an enhanced experience, and depressive patients a slowed experience of time flow.11 When assessing time production (giving a “stop” signal when the proposed duration is supposed to be finished), the intermediate position of controls between manic and depressed patients was only observed for a shorter duration (7 seconds), and did not reach statistical significance.11

Figure 2
Figure 2. The scalar expectancy theory.

Time perception also involves motor, arousal, attention, and memory processes

Depression has a strong cognitive impact, but the motor components should not be neglected when assessing time perception, especially in the production of time intervals. Indeed, most of the previously described studies were based on a wide range of time lengths, from 1 or 2 seconds, to minutes, and even hours. Furthermore, they frequently rely on temporal judgment, which also requires the production and timing of a motor response (such as tapping or counting). In the knowledge that psychomotor retardation is a core feature of depression, it is difficult to dissociate the role of the motor component from that of the timing component in temporal performance, even though these two components may be related.14 Indeed, when specifically assessing the duration of movement patterns in depressed patients with melancholia, Lemke et al15 showed that the median of repetitive movements was higher in depressed patients than in the control group, regardless of the presence or absence of medication.

In order to avoid this confusing impact of motor retardation, a temporal bisection task was tested in depressed patients. In the test, two standard durations both below 2 seconds, one short and one long, are presented to subjects who have to categorize each probe (of variable duration) as being closer to the long or to the short standard duration. With this approach, a shift toward shorter durations, therefore an underestimation of time (ie, slower internal clock), was observed in depressed patients.14

A similar approach in another study initially gave the same type of results,16 but mainly for long intervals (above 1 second), probably because longer intervals require supplementary cognitive resources.

A comprehensive model of interval timing
The “scalar expectancy theory” was initially proposed by Gibbon1 and was largely developed later.17 As simplified in Figure 2, this theory is built on the idea of a neuronal pacemaker, which provides repetitive and regular pulses. Pulses are then gated (in order to define the beginning and the end of the duration to assess), and stored as an accumulated value (the number of pulses) in stored memory. The assessment of the duration of a specific task is then compared with reference memories, leading to a decision. This model was considered to successfully predict the outcomes of a large proportion of behavioral, pharmacological, and anatomical work in the field.17

This model of an internal clock has the advantage not only of being simple, but also of involving different cognitive features. Indeed, “gating” means being able to focus attention (with enough arousal ) on when to open and close the inputs, “comparing” (the number of pulses that were stored in the accumulator with reference memory) needs encoding and access tomemory contents, and “concluding,” ie, giving the stop signal as a consequence of the perception that the correct time interval has been reached, solicits motor skills. On the other hand, the scalar expectancy theory might be more relevant from a mathematical, rather than a neurobiological perspective, mainly because it is difficult to instantiate neural mechanisms that accumulate pulses over the order of minutes. This useful model was thus considered to be a “wellstructured metaphor [rather] than a diagram of the working brain.”17

The neurobiology of timing

Initially, patients with a cerebellar pathology were used to pinpoint the role of the cerebellum in timing perception, but evidence has now been enriched with neuroimagery, stimulation, or inhibition using repetitive transcranial magnetic stimulation (rTMS), and electrophysiology data (for a review, see reference 18).

Other neural regions might serve as a dedicated timing system, including the basal ganglia, the supplementary motor area, and the prefrontal cortex.18 A fronto-striato-thalamic circuit, modulated by the dopamine system, would appear to be crucial for temporal processing within the range of seconds.

The basal ganglia receive the majority of their stimuli from the cortex, the thalamus, and the midbrain, and although initially reported as being involved in motor functioning, are now considered as playing an important role in motivational and cognitive aspects of brain functioning. The cortex projects glutamatergic (excitatory) afferents to the basal ganglia, mainly through the striatum (Figure 3). Striatal γ-aminobutyric acid (GABAergic; inhibitory) outputs project to the internal globus pallidus and the substantia nigra pars reticulata. These two structures provide an inhibitory influence on the thalamus, which, in turn, provides an excitatory output back to the cortex (and partly to the striatum).

Figure 3 was built on a schema usually proposed to describe the loops involved in motor activity (especially regarding the basal ganglia). The main proposed change today is that the output signal from the cortex does not represent motor activity, but the sum of individual waves of neurons that have different frequency oscillating signals. Adding together individual neurons as a model for producing an internal clock has the considerable advantage of allowing production (and recognition) of intervals of seconds or even minutes, which are far above the usual 200-millisecond intervals of neuronal activity. Indeed, mixing three neurons with 5 Hz (a peak every 200 msecs), 6 Hz, or 7 Hz oscillating signals will produce a curve at which a peak is observed every second.17 Such peaks, from milliseconds to minutes depending on the number of neurons concerned and their individual frequency, may then serve as an output signal that will reach the striatum.

Figure 3
Figure 3. Neuroanatomy of the timing circuit.

Abbreviations: DA, dopamine; GABA, γ-aminobutyric acid; GLU, glutamate; STN, subthalamic nucleus; VA, ventral anterior, VL, ventral lateral.

In terms of the role of the cortex, according to single cell activity studies, the prefrontal cortex may be more specifically involved. Increasing activity of dorsolateral prefrontal neurons was detected during the delay period of a task soliciting postponed activity.19 Furthermore, prefrontal delay neurons can have several different patterns of activity, including increasing, decreasing, and peak firing rates during the delay.20

Interestingly, this model places the striatum as a core structure in the timing process, but also gives importance to the cortex and probably more specifically to the right dorsolateral prefrontal cortex. This model also favors a specific role for dopamine, which is in accordance with the fact that metamphetamine has an enhancing effect on timing function, and patients with Parkinson’s disease have a slower internal clock speed.18

Are the internal clocks identical for circadian and short-term interval timing?

When assessing time judgment, below-second and abovesecond intervals seem to be important, but what about intervals of hours and days? In order to learn associations with events that occur at particular times of the day, organisms use a specific circadian clock: the SCN. The SCN are located bilaterally in the anterior hypothalamus and are entrained to the external light-dark cycle by a neural pathway that transmits light information from the eyes through the retinohypothalamic tract.

Thus the neurobiological circuits for circadian and ultradian intervals seem to be different. Indeed, three types of distinction have been proposed between circadian and shorter interval timing.21 The circadian clock would be phase-based (automatically generated by the SCN), with low flexibility (the 24-hour basis can be shifted only by 1 hour a day, for example during jetlag) and constant variability, whereas the interval clock would be counter-based (adding pulses), with high flexibility (as relying on internal memory) and scalar variability (implying it has decreased precision for longer intervals).


The way we assess the speed of time forms part of our basic cognitive skill set, and is an important part of everyday analyses, decision making, and action. Such a core activity can be (artificially) distinguished according to time intervals. For intervals of below a second, most of the task concerns motor activity and may be more specifically orchestrated by the cerebellum. For intervals of over an hour, a specific structure is shared by many animals, ie, the SCN, allowing species to function both synchronically and in accordance with the night/day rhythm, and therefore to survive. For intervals of seconds to minutes, the timing circuit seems to be more complex, also involving cognitive skills. Whichever way neurons are organized as pacemakers (different models have been proposed, and their validity is difficult to prove), it is interesting that arousal, attention, and memory are involved, because of their ability to influence the inputs or outputs of the pacemaker. It is therefore not surprising that in major depressive disorder, which is known to be associated with poor attention and memory impairment,22 time seems to pass slower.

The specific impact of emotion has been tested, with analysis (using event-related potentials) of the way subjects react in front of neutral versus sad faces. Under emotional conditions, the P160 and P240 amplitudes have been found to be enhanced, suggesting that intentional bias for emotional stimuli attenuates the cognitive resources for time perception.23 The important role of dopamine at the neurotransmitter level, and the specific place of the prefrontal cortex in time production, both argue in favor of abnormal time perception belonging to the list of symptoms of depression; or at the least, sharing some identical neurobiological patterns. _


1. Gibbon J, Malapani C, Dale CL, Gallistel C. Toward a neurobiology of temporal cognition: advances and challenges. Curr Opin Neurobiol. 1997;7:170-184.
2. Bech P. Depression: influence on time estimation and time experiments. Acta Psychiatrica Scandinavia. 1975;51:42-50.
3. Blewett AE. Abnormal subjective time experience in depression. Br J Psychiatry. 1992;161:195-200.
4. Hoffer A, Osmond H. The relationship between mood and time perception. Psychiatr Q Suppl. 1962;36:87-92.
5. Kitamura T, Kumar R. Time passes slowly for patients with depressive state. Acta Psychologica Scandinavia. 1982;4:127-140.
6. Lehmann HE. Time and psychopathology. Ann N Y Acad Sci. 1967;138:798-821.
7. Mezey AG, Cohen SI. The effect of depressive illness on time judgment and time experience. J Neurol Neurosurg Psychiatry. 1961;24:269-270.
8. Straus E. Disorders of personal time in depressive states. South Med J. 1947; 25:254-259.
9. Wyrick RA, Wyrick LC. Time experience during depression. Arch Gen Psychiatry. 1977;34:1441-1443.
10. Hawkins WL, French LC, Crawford BD, Enzle ME. Depressed affect and time perception. J Abnormal Psychol. 1988;97:275-280.
11. Bschor T, Ising M, Bauer M, et al. Time experience and time judgment in major depression, mania and healthy subjects. A controlled study of 93 subjects. Acta Psychiatrica Scandinavia. 2004;109:222-229.
12. Kuhs HW, Kammer HK, Tolle R. Time estimation and the experience of time in endogenous depression (melancholia): an experimental investigation. Psychopathology. 1991;24:7-11.
13. Grinker J, Glucksman ML, Hirsch J, Viseltear G. Time perception as a function of weight reduction: a differentiation based on age at onset of obesity. Psychosom Med. 1973;35:104-111.
14. Gil S, Droit-Volet S. Time perception, depression and sadness. Behav Processes. 2009;80:169-176.
15. Lemke MR, Koethe NH, Schleidt M. Timing of movements in depressed patients and healthy controls. J Affect Disord. 1999;56:209-214.
16. Sévigny MC, Everett J, Grondin S. Depression, attention, and time estimation. Brain Cogn. 2003;53:351-353.
17. Matell MS, Meck WH. Cortico-striatal circuits and interval timing: coincidence detection of oscillatory processes. Brain Res Cogn Brain Res. 2004;21:139-170.
18. Keele SW, Ivry R. Does the cerebellum provide a common computation for diverse tasks? A timing hypothesis. Ann N Y Acad Sci. 1990;608:179-207.
19. Fuster JM, Bauer RH, Jervey JP. Cellular discharge in the dorsolateral prefrontal cortex of the monkey in cognitive tasks. xp Neurol. 1982;77:679-694.
20. Kojima S, Goldman-Rakic PS. Delay-related activity of prefrontal neurons in rhesus monkeys performing delayed response. Brain Res. 1982;248:43-49.
21. Hinton SC, Meck WH. The ‘internal clocks’ of circadian and interval timing. Endeavour. 1997;21:82-87.
22. Gorwood P, Corruble E, Falissard B, Goodwin GM. Toxic effects of depression on brain function: impairment of delayed recall and the cumulative length of depressive disorder in a large sample of depressed outpatients. Am J Psychiatry. 2008;165:731-739.
23. Gan T, Wang N, Zhang Z, Li H, Luo YJ. Emotional influences on time perception: evidence from event-related potentials. Neuroreport. 2009;20:839-843.

Keywords: time perception; depression; cognition; subjective; objective; neurobiology; circadian rhythm; cerebellum; cortex; suprachiasmatic nuclei