Eveningness and distinctness of the circadian rhythm in men are related to altered neural responses to gain and loss

Abstract

Circadian rhythmicity is considered to be one of the factors modulating the functions of the brain’s motivational system. To date, research has primarily focused on the unidimensional morningness-eveningness continuum (phase of the circadian rhythm), while the second dimension – distinctness (subjective amplitude) has been neglected. In this study, we explored the relationships of distinctness and eveningness with neural correlates of motivated behavior. In 37 healthy men (aged 20–30), we measured the characteristics of circadian rhythmicity using the Morningness-Eveningness-Stability-Scale improved (MESSi) questionnaire and employed additional psychometric assessments. To determine the gain- and loss-motivated behavior, we used fMRI during the Monetary Incentive Delay task. We found that higher distinctness was positively related to the: activation in the ventral tegmental area during the anticipation of punishment; activation in the left occipital pole during the negative feedback; elevated neuroticism and behavioral inhibition system scores. Eveningness was associated with increased activation in the left frontal pole, middle temporal gyrus, inferior temporal gyrus, and temporal fusiform cortex during the positive feedback. These results improve our understanding of the relationship between distinctness, eveningness, and motivated behavior. They constitute a promising direction for further studies on the multidimensionality of circadian rhythms.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-07049-8.

Introduction

The mesolimbic system, also known as the brain’s motivational system, is formed by projections of midbrain dopamine neurons of the ventral tegmental area (VTA) to the striatum, prefrontal cortex (PFC), amygdala, hippocampus, and many other structures of the limbic system. This circuitry is responsible for mediating the physiological and cognitive processing of reinforcement, during which the brain associates various events with a positive or desirable outcome. The mesolimbic system regulates an extensive range of behavioral outcomes^1,2 and plays a crucial role in the pathophysiology of mood and addictive disorders³. For this reason, it is important to discover the factors that modulate the functions of the motivational system⁴. Interestingly, increasing evidence points to circadian rhythmicity as one of such factors.

Circadian rhythmicity in humans is represented by a complex phenotype with varying levels of alertness, mood and motivation observed across a 24-hour cycle. To describe these daily rhythms, a characterization of at least two separate dimensions: chronotype (rhythm phase) and distinctness (subjective amplitude of the rhythm) are needed⁵. The chronotype represents an individual’s preferred time of activity (i.e., morningness-eveningness), while the distinctness – the strength of this preference, understood as the subjective amplitude between hyper- and hypo-activation phases (e.g., high distinctness will manifest itself in the feeling that “There are moments during the day when I feel unable to do anything”, while low distinctness – “I can focus at any time of the day”)^6–8.

Although both components can influence the mesolimbic system functions, to date, the vast majority of investigations have solely used a unidimensional assessment of morningness-eveningness⁹. Those studies demonstrated the association between chronotype and activation within this circuitry, and the circadian impact was observed in measures of both reward anticipation and reward receipt (for a review, see: Byrne et al., 2019⁴). For example, Hasler and colleagues¹⁰ examined the relationship between the chronotype and neural responses in mesolimbic structures: ventral striatum (VS) and medial prefrontal cortex (mPFC) to a rewarding task and found chronotype-related differences during both the anticipation and outcome phases of the task. The evening-type group, compared to the morning-type group, showed relatively less mPFC activation during reward anticipation and relatively greater VS activation during reward receipt. The authors suggest that this pattern could plausibly reflect reduced regulatory control and elevated reward sensitivity among the evening-types. Moreover, at the behavioral level, eveningness was shown to be associated with negative mood and affect^11,12 increased anxiety¹³ propensity to depression^13–15 more use of substances, including alcohol, nicotine, and caffeine^13,16 and vulnerability to addiction¹⁷which have been consistently related to dysregulations in reward processing^18,19. This further points to the association between evening chronotype and altered reward-related brain function. Furthermore, several studies have shown activation within the mesolimbic system during the anticipation of a loss^20,21. As the motivation to not lose money is inherent to each loss trial, such activation most probably reflects the participants’ goal of avoiding loss. Some data indicate an association between circadian rhythmicity and the functioning of the brain’s motivational system during the anticipation of a loss. LeMoult and colleagues²² investigated circadian modulations of the mesolimbic system function, but included a circadian proxy other than chronotype, namely, diurnal cortisol rhythm. The results showed that cortisol awakening response (CAR) could predict activation in the VS during the anticipation of both gain and loss²². Although that finding suggests that circadian rhythmicity could modulate the brain’s motivational system activity related to loss avoidance, to date, no study has documented the role of chronotype or distinctness in this process.

While there has been significant research pointing to the associations between chronotype and reward processing, investigations dealing with distinctness are rather sparse. However, recent research has emphasized the role of distinctness, which correlates negatively with life satisfaction and positively with negative emotionality^23–25. Moreover, higher subjective amplitude values are positively correlated with past negative time perspective and negatively correlated with focusing on the future²⁶. Relationships of distinctness with more spontaneous mind wandering and neuroticism have also been observed^27–30. Importantly, higher distinctness may be more strongly related to negative emotionality than eveningness preference²⁷. In addition, distinctness proved to be connected with seasonality and an avoidance-oriented coping style³¹. Avoidance-oriented coping is considered a non-adaptive strategy that is related to depression^32,33 negative affect, perceived stress, and experiential avoidance³⁴. Although all these observations point to the relationship between the level of distinctness and the function of the motivational system, this relationship has never been directly tested.

In summary, research on circadian rhythmicity as a modulator of motivational system function has linked eveningness to altered reward processing, whereas the relationship between evening chronotype and loss processing has received less research attention. Furthermore, the role of distinctness in reward and loss processing remains unknown. To address the open question about the relationship between eveningness, distinctness, and neural correlates of reinforcement processing, we performed an fMRI experiment. We used the Monetary Incentive Delay (MID) task (Fig. 1), which is a common method of assessing gain- and loss-motivated behavior^35,36. We analyzed brain activity during anticipation and outcome (feedback) phases of the MID task using both whole-brain and region-of-interest (ROI) analysis approaches. In the ROI analysis, we focused on the ventral tegmental area (VTA), because it is not only the core of the mesolimbic system^37,38 but also serves as a circadian oscillator that regulates incentive motivation across the day³⁹. We expected to find circadian differences in the VTA during the anticipation phase (motivation phase, characterized as ‘wanting’) rather than during the outcome phase (feedback phase, characterized as ‘liking’⁴⁰), because mesolimbic circuits are considered primary substrates of the former. Moreover, we examined the subjects’ reinforcement-related behavior. We employed a battery of psychometric questionnaires, which allowed us to assess individuals’ sensitivity to punishment and reward, personality traits, positive and negative affect, behavioral inhibition system and behavioral activation system (BIS/BAS), and anxiety levels (for details, see: Methods). Building on the previous findings, it was expected that variability in behavioral and neural correlates of reinforcement processing would be associated with both eveningness and distinctness.

Fig. 1.

The monetary incentive delay task paradigm. At the beginning of each trial, participants obtain the graphical cue with assigned monetary value – reward cues (square with 3 lines represents big win: +5 PLN (Polish Zloty, currency in Poland, 1 PLN equals 0.30 USD), square with 1 line – small win: +1 PLN), punishment cues (circle with 3 lines represents big loss: − 5 PLN, circle with 1 line – small loss: − 1 PLN). After the anticipation phase, the white-filled square on the scanner screen is displayed, signaling the moment of reaction – the participant should click on the pad as fast as possible to gain a reward or avoid punishment, depending on the monetary cue. There are also oddball trials (+/–10 PLN) in which inhibition of the automatic reaction is required. The oddball trials were not analyzed in terms of brain activity due to their low repetition number, and their main goal was to keep participants focused on correctly processing the meaning of cues. The task consisted of 60 trials. For details, see: Methods.

Results

The subjective characteristics of the participants’ circadian rhythms were assessed using the subscales for eveningness and distinctness. The higher the score in the eveningness subscale, the more evening-oriented the individual is expected to be. Similarly, the higher the score in the distinctness subscale, the stronger the within-day fluctuations in mental processes. Additionally, we collected data on participants’ predisposition to experiencing insomnia episodes. The values were distributed normally according to the Kolmogorov-Smirnov test (p > 0.05). The demographic summary of the sample is presented in Table 1.

Table 1.

Demographic summary of the cohort (37 male participants) used in the study.

Variable (n = 37)	Value range	Mean ± SD	Median
Age (years)	20–30	24.4 ± 2.1	25
Eveningness	10–24	18.6 ± 4.6	20
Distinctness	10–25	18.2 ± 3.0	18
AIS	1–16	5.6 ± 3.2	5

Eveningness and distinctness were measured using the Morningness-Eveningness-Stability-Scale improved (MESSi) questionnaire^8,41. Theoretical ranges of eveningness and distinctness subscales were 5–25. Potential insomnia episodes were controlled using Athens insomnia scale (AIS)⁴² with theoretical range 0–24. SD standard deviation.

Behavioral results

To assess the relationship between the distinctness, eveningness and psychometric parameters, an Ordinary Least Squares (OLS) regression analysis across multiple dependent variables was performed, followed by False Discovery Rate (FDR) correction for multiple comparisons. Distinctness has a statistically significant positive association with neuroticism (β = 1.36, p = 0.016) and BIS (β = 0.63, p = 0.023). We also observed trend-level associations between distinctness and sensitivity to punishment (β = 0.40, p = 0.088), openness (β = 0.89, p = 0.067), and attention deficits (β = 1.36, p = 0.067). Coefficient values and confidence intervals from the OLS regression analysis for all of the measured psychometric dependent variables are shown in Fig. 2. Detailed data for all psychometric values are in Supplementary Table S1.

Fig. 2.

The ordinary least squares (OLS) regression results for all psychometric parameters. Distinctness (left panel) and eveningness (right panel) were used as predictor variables. Dots and numbers represent β coefficient values, horizontal bars - confidence intervals (95% CI). Statistically significant associations are marked in blue, * represents p-value ≤ 0.05, trend-level associations are marked with #, which represents p-value ≤ 0.1. Psychometric parameters were assessed using the Sensitivity to Punishment and Sensitivity to Reward Questionnaire (SPSRQ^43,44), NEO-Five Factor Inventory (NEO-FFI^45,46), Generalized Anxiety Disorder (GAD-7⁴⁷), Positive and Negative Affect Schedule (PANAS^48,49) Behavioral Inhibition System and Behavioral Activation System (BIS/BAS^50,51), and Adult ADHD Self-Report Scale (ASRS^52,53).

Neuroimaging results

Considering the associations observed between circadian characteristics and neuroticism and BIS, we expected to find neural differences in reward-punishment processing. In order to determine neural correlates of these explored processes, a Monetary Incentive Delay (MID) task was employed (for details, see: Methods). We analyzed both anticipation and feedback phases of the MID task.

To investigate our hypotheses, we used both whole-brain and region-of-interest (ROI) analysis approaches. To mitigate factors that may potentially modulate the functionality of the investigated brain regions (e.g., the synchrony effect⁵⁴ or seasonal changes in mood and cognitive performance⁵⁵), all fMRI examinations were carried out from 1 PM to 5 PM during two weeks in summer (for details, see: Methods)..

Anticipation phase

In the anticipation phase, we contrasted (1) each trial type (reward / punishment) against the implicit baseline, (2) high value trials against low value trials within each trial type, (3) and reward anticipation trials against punishment anticipation trials. An ordinary least squares (OLS) analysis performed on the region-of-interest (ROI) level for the VTA activity showed the distinctness-related difference during the anticipation of punishment vs. anticipation of reward trials, Fig. 3a. There were no significant neuroimaging findings associated with eveningness (Table 2).

Fig. 4.

Neuroimaging results for the feedback phase. Top panel (a): left Frontal Pole, left Inferior Temporal Gyrus, left Temporal Fusiform Cortex, left Middle Temporal Gyrus activity in positive feedback phase is positively associated with the eveningness (Z-stat threshold = 3.1, cluster correction p < 0.05). Plot of the association between contrast of parameter estimates and eveningness was obtained using the mask for all significant clusters. Linear regression line is shown on the chart. Bottom panel (b): left Occipital Pole activity in negative feedback phase is positively associated with the distinctness of the circadian rhythm (Z-stat threshold = 3.1, cluster correction p < 0.05). Plot of the association between contrast of parameter estimates and eveningness was obtained using the mask for the significant cluster. Linear regression line is shown on the chart. Activations are shown on standard human brain template (MNI152). Coordinates (x, y, z) were chosen to show the best representation.

Table 2.

Significant positive associations between the distinctness and neuroimaging results for the anticipation phase.

Circadian dimension	Contrast	MNI (x, y, z)	Area	Voxels	Statistics	p-value
Distinctness	Punishment anticipation > Reward anticipation	0, −17, −14	B ventral tegmental area (VTA)	32	T = 3.03	0.005 a
Distinctness	Punishment anticipation > implicit baseline	0, 24, 62	B superior frontal gyrus (SFG) B supplementary motor area (SMA)	261	Z = 3.44	0.033 b

B bilateral.

^a Region-of-interest analysis.

^b Whole-brain analysis at a threshold of Z = 2.8, corrected for multiple comparisons with cluster-level family-wise error rate approach. Due to the use of a slightly less stringent threshold, this result is considered as exploratory and should be interpreted with caution.

Fig. 3.

Neuroimaging results for the anticipation phase. Top panel (a): differences in the activity between the anticipation of punishment vs. anticipation of reward trials for the ventral tegmental area (VTA) are positively related to distinctness. Bottom panel (b): bilateral superior frontal gyrus (SFG) and supplementary motor area (SMA) activity in anticipation of punishment phase is positively associated with the distinctness of the circadian rhythm (Z-stat threshold = 2.8, cluster correction p < 0.05). Due to the use of a slightly less stringent threshold, this result is considered as exploratory and should be interpreted with caution. Activations are shown on standard human brain template (MNI152). Coordinates (x, y, z) were chosen to show the best representation. Linear regression lines are shown on the charts.

The whole-brain neuroimaging analysis performed at the threshold of Z = 3.1 revealed no significant findings for both distinctness and eveningness. However, the exploratory analysis with a slightly less stringent thresholding (Z = 2.8, with cluster correction p threshold set to 0.05.) revealed associations between distinctness and negative reinforcement processing. Distinctness was positively associated with the activity of the bilateral Superior Frontal Gyrus (SFG) and supplementary motor area (SMA) during the anticipation phase of the punishment trials (across both big and small punishments), Fig. 3b. The whole-brain analysis results suggest the involvement of brain areas related to working memory and attention^56,57 which is consistent with the predictions – to effectively perform the task, participants were required to remember the meaning of monetary cues. However, it is important to note that these whole-brain findings were obtained using a less restrictive statistical threshold (Z = 2.8), and should therefore be interpreted with caution. We report them as exploratory results in the hope that they can serve as a guide for further research. The results are detailed in Table 2 and Fig. 3.

Table 3.

Significant positive associations between the eveningness, distinctness and neuroimaging results for the feedback phase.

Circadian rhythm dimension	Contrast	MNI (x, y, z)	Area	Voxels	Statistics	p-value
Eveningness	Positive feedback > implicit baseline	−22, 48, 36	L frontal pole	24	Z = 4.61	0.0089
		−58, −54, −8	L middle temporal gyrus	20	Z = 4.27	0.028
		−44, −46, −16	L inferior temporal Gyrus, L temporal fusiform Cortex	22	Z = 4.11	0.016
Distinctness	Negative feedback > implicit baseline	−16, −94, −10	L occipital pole	15	Z = 4.43	0.043

L left.

In our dataset, 3 participants obtained Athens Insomnia Scale values higher than 10, which indicated potential sleep deprivation. To check whether their brain activity was significantly impacted, we re-conducted the whole brain and region-of-interest analyses excluding these participants. The results remained unchanged (see: Supplementary Table S2).

Feedback phase

In the whole-brain analysis of the feedback phase, the general linear model (GLM) was constructed based on each subject’s behavior during the task. Contrasts were defined as: (1) positive and negative feedback relative to an implicit baseline, and (2) positive feedback relative to negative feedback. Maps (z-stat) for each contrast were thresholded at p < 0.001 (Z = 3.1) with cluster correction p threshold set to 0.05. The whole-brain analysis revealed positive association between eveningness and brain activity in the following regions for positive feedback trials: left Frontal Pole, left Inferior Temporal Gyrus, left Temporal Fusiform Cortex, left Middle Temporal Gyrus (Fig. 4a). Moreover, in the negative feedback trials, we found a positive association between the distinctness and left Occipital Pole activity (Fig. 4b). The ROI analysis of the VTA activity data revealed no statistically significant associations with distinctness and eveningness. The results are detailed in Table 3 and Fig. 4.

Monetary incentive delay (MID) task behavioral results

During task performance, participants’ reaction times (RT) were measured. Statistical analyses of reaction times were conducted by fitting a linear mixed-effects model (LMM) for each reaction time measure: median RT for (1) big win (+ 5 PLN), (2) small win (+ 1 PLN), (3) big loss (-5 PLN), and (4) small loss (-1 PLN) trials. RT was modeled as a function of two predictors: distinctness and eveningness. LMM accounts for both fixed effects (circadian characteristics) and random effects (participant-specific variability). We found that elevated distinctness was related to faster reactions on the big win trials (β = −0.695, p < 0.001), as well as slower reactions on the small win trials (β = 0.808, p < 0.001), and the big loss trials (β = 1.281, p = 0.013). Detailed information is included in the Supplementary Table S3. No statistically significant relationships were found between circadian characteristics and (1) final monetary score, (2) number of incorrect responses (omitting responses when due to provide one, and failure to suppress responses when indicated to do so).

Discussion

This study reveals that individual differences in circadian rhythm distinctness and eveningness are reflected in neural activity during the performance of the Monetary Incentive Delay (MID) task. Namely, higher levels of distinctness are related to enhanced neural activity in the left occipital pole during loss feedback trials, increased activation of the ventral tegmental area (VTA) during punishment-anticipation trials in comparison with reward-anticipation trials, and (in an exploratory analysis with slightly less stringent thresholding) increased activity in the bilateral superior frontal gyrus (SFG) and supplementary motor area (SMA) during punishment anticipation trials. The neuroimaging findings were paralleled by a positive relationship between distinctness with neuroticism and elevated scores on the behavioral inhibition system (BIS). Furthermore, eveningness was shown to be related to enhanced neural activity during win-feedback trials (in the left frontal pole, inferior/fusiform temporal gyrus and middle temporal gyrus). We did not observe any statistically significant relations between eveningness and psychometric characteristics.

During the outcome phase of the MID task, higher levels of distinctness are related to preferential processing of loss feedback in the left occipital pole containing the primary visual cortex. These results are complementary to our previous work⁵⁸ in which we showed negative correlation between subjective amplitude and both grey matter volume and cortical thickness in the left primary visual cortex, which plays a crucial role not only in visual perception but also in the down-regulation of negative emotions⁵⁹. Preferential processing of loss feedback most probably reflects the subjects’ negative emotions caused by a loss. According to loss aversion theory, individuals prefer to avoid loss rather than acquire a gain and it is even suggested that losses are psychologically twice as harmful as gains are beneficial⁶⁰. In light of this theory, the activity pattern in the visual cortex might suggest that people with higher distinctness manifest higher levels of loss aversion. In line with this view, we observed higher VTA activation in individuals with high-distinctness in the loss-anticipation vs. win-anticipation contrast. VTA is a core part of the mesolimbic system and activation of this structure has already been shown during anticipation of a loss^20,21. As such, the presented results most probably reflect assigning higher motivational value to potential negative outcomes by the salience-encoding dopaminergic subsystem, further pointing to the link between higher levels of distinctness and preference to avoid losses. Furthermore, it should be stressed that some previous findings indicate the association between circadian rhythmicity, as measured with diurnal cortisol rhythm, and the brain’s mesolimbic system activity during loss avoidance²². Our work, however, is the first one to provide evidence for the potential role of distinctness in this process.

With appropriate caution due to the use of a slightly lower statistical threshold, we need to note that individuals with higher distinctness manifest increased bilateral activation within the SFG and SMA in punishment-anticipation vs. implicit baseline contrast. The SFG and SMA are involved in working memory and executive functions⁶¹ as well as in preparation for the adequate motor response to avoid punishment⁶². In the broader context, however, these brain areas are additionally involved in emotion regulation⁶³. As the anticipation phase of the MID task contains working memory components, and working memory performance may be affected by loss aversion⁶⁴ our results might reflect the modulation of working memory functions by motivation through loss aversion in individuals with higher levels of distinctness. However, it should be acknowledged that this interpretation, based on an exploratory finding, remains preliminary and requires further investigation.

The relationship between distinctness and altered processing of losses is additionally supported by significant differences in the reaction times during performance of the MID task. Namely, we observed an increase in reaction times for big loss trials (slower reactions). Participants with greater circadian distinctness may respond slower in loss trials as part of an avoidance strategy. In other words, the longer reaction times could reflect a hesitation or increased effort to avoid loss. Furthermore, we observed a decrease in reaction times for big win trials (faster reactions), and an increase in reaction times (slower reactions) for small win trials, which could be interpreted as diminished motivational value of smaller rewards, with the observed preferential processing of loss potentially playing a part in the process.

To embed neuroimaging results within a broader, psychological context, we also collected psychometric data from participants. We found positive associations of the subjective amplitude of the rhythm with neurotic personality type and scores on the behavioral inhibition system (BIS). Neuroticism and the BIS are related to similar individual emotional experiences⁶⁵. Individuals who score high on the scale of neuroticism are more likely to experience negative emotions such as fear or anxiety, and are prone to lacking adaptive coping styles when dealing with stressors⁶⁶. Higher BIS scores are similarly linked to avoidance behavior⁶⁷ and increased experience of negative emotions such as frustration and anxiety⁵⁰. These characteristics align the previously established relationship of high subjective amplitude with negative emotionality^27–30 and unadaptive, avoidance-oriented coping style³¹. Combined with fMRI results, they additionally indicate that high distinctness is associated with avoidance behavior, manifested in particular by the tendency to avoid losses.

In addition to our exploration of the distinctness of the circadian rhythm, we found a statistically significant positive relationship between eveningness and neural activity in the left frontal pole, middle temporal gyrus, inferior temporal gyrus, and temporal fusiform cortex during the positive feedback. Frontal pole contributes to many aspects of cognition, including feedback processing, and plays a role in feedback-based learning⁶⁸. Cortical areas within the inferior, fusiform and middle temporal gyri are involved in several high-order cognitive functions including visual information processing⁶⁹. Specifically, the inferior and fusiform gyri belong to the ventral visual pathway, while the middle temporal gyrus is a part of dorsal visual pathway⁷⁰. Processing of visual information in these pathways can be influenced by the affective significance of stimuli^70,71. In light of these findings, our results might indicate that eveningness is associated with enhanced processing of positive outcomes of the MID task. Although this result is consistent with previous findings showing elevated reward sensitivity among individuals with an evening chronotype¹⁰ we did not observe statistically significant relationships between eveningness and psychometrically-measured sensitivity to reward. This may have resulted from the limited sample size. Moreover, most studies linking evening chronotype and emotionality used the Morningness–Eveningness Questionnaire (MEQ) questionnaire. While the eveningness scale and the MEQ are correlated, differences between these tools may explain some of the discrepancies in our results.

Importantly, the observed associations are correlational in nature, and thus do not imply causality. While we highlight the potential role of the eveningness and distinctness of the circadian rhythm in shaping the neural responses, the directionality of these relationships remains to be fully understood.

Limitations

Although every effort was made in the careful design of this study, several limitations should be acknowledged.

First, while previous investigations linking circadian rhythmicity with negative mental health outcomes have traditionally focused on the role of eveningness preference^10,17,28,72 emerging evidence suggests the effects associated with late circadian phase are to a great extent mediated by sleep quality^73,74. In our sample, neither the circadian time preference (eveningness) nor subjective amplitude (distinctness) were related to sleep complaints, which may partly reflect the inclusion of only male participants - given that sex differences can modulate sleep-circadian interactions^75,76. Nonetheless, we acknowledge the importance of thoroughly examining the role of sleep through both subjective and objective assessments of sleep quality, continuity, and architecture in future studies.

The decision to include only young adult male participants, mainly university students, was made to reduce the number of confounding variables related to sex and age, allowing for a more controlled investigation without the added complexity. However, this limits the generalizability of our findings to a specific demographic group. Future studies should include female participants and broader age ranges to examine potential sex- and age-related differences in circadian regulation and its links to motivational processes. Additionaly, all the data were acquired in the afternoon during the summer, so the potential interplay between circadian characteristics, seasons, time-of-day variations, and ambient light should also be investigated. Finally, we also acknowledge the importance of multimodal assessment in circadian-affective research. Future studies should aim to complement self-reported measures of circadian amplitude with objective, longitudinal measures of circadian rhythmicity (e.g., actigraphy, Dim Light Melatonin Onset (DLMO), and body temperature), in order to more accurately capture the direction and nature of these associations.

Conclusions

To conclude, the present study improves our understanding of the relationship of distinctness and eveningness with loss- and gain-motivated behavior. This extends the state of knowledge by showing that increased neural processing of punishment cues and negative feedback is associated with higher distinctness, whereas enhanced neural processing of positive feedback is associated with eveningness. In addition, distinctness is related to neuroticism and avoidance behavior, confirming the associations between distinctness and punishment processing observed on the neural level. However, it is important to acknowledge the potential bi-directionality of these relationships, as our findings are correlational and do not establish causal direction. As such, the current study highlights the complex nature of the relationships between circadian rhythmicity and reinforcement processing. Our results support a multidimensional model of circadian rhythmicity. By integrating distinctness as an additional factor into analyses, we can achieve a more nuanced and comprehensive understanding of human circadian behavior.

Methods

Participants

The described study involved 37 healthy men. All subjects were aged 20–30, had normal or corrected-to-normal vision, were drug-free, and self-reported no history of either neurological, nor psychiatric disorders. The described protocol was approved by Rector’s Committee for the Ethics of Research Involving Human Participants at the University of Warsaw (decision no. 230/2023) and was performed in accordance with the Declaration of Helsinki (World Medical Association, 2008). Written informed consent prior to experimentation was obtained from all participants.

Circadian characteristics

The subjective characteristics of the circadian rhythms of the participants were assessed using the Polish version of the Morningness-Eveningness-Stability-Scale improved (MESSi) questionnaire^8,41 which includes subscales for Eveningness (rhythm phase), Distinctness (rhythm amplitude) and Morning Affect. However, in this study we focused on an in-depth analysis of the role of eveningness and distinctness of the circadian rhythm, following Carciofo’s findings¹¹ according to which distinctness and morning affect can be treated as separate predictors. Distinctness measures the subjective amplitude or the range of diurnal variation. This variation results from the ability to volitionally modulate one’s own psychophysiological state and feel the difference between hyper- and hypo-activation phases (e.g., “I can focus at any time of the day”; “There are moments during the day where I feel unable to do anything”). Higher scores show higher fluctuations. Each dimension was tested with 5 objects ranked from 1 to 5, which are summed together, making the theoretical score range from 5 to 25 points. The higher the score in the eveningness subscale, the more evening-oriented the individual is. Similarly, the higher the score in the distinctness subscale, the stronger the within-day fluctuations in mental processes. In the current dataset, the eveningness scores ranged from 10 to 24, while distinctness scores varied from 10 to 25. Additionally, we assessed participants’ predisposition to insomnia using the Athens Insomnia Scale (AIS) questionnaire⁴². In our cohort, the score range was 1 to 16, while the theoretical range is 0 to 24.

Behavioral data

The individual psychometric differences were assessed using NEO-Five Factor Inventory (NEO-FFI^45,46), Behavioral Inhibition System and Behavioral Activation System (BIS/BAS^50,51), the Sensitivity to Punishment and Sensitivity to Reward Questionnaire (SPSRQ^43,44), Positive and Negative Affect Schedule (PANAS^48,49), Generalized Anxiety Disorder (GAD-7⁴⁷) and Adult ADHD Self-Report Scale (ASRS^52,53).

Monetary incentive delay (MID) task

Each trial in the MID task consisted of 5 stages as shown in Fig. 1 (in the Introduction section). At the beginning of each trial, a graphic cue indicative of its monetary value was displayed in the scanner.

During the trial, participants had to rely on previously learned associations between the cues and monetary outcomes. Following the cue, the anticipation phase occurred, which consisted of a fixation point (represented as a “+” in Fig. 1) displayed in the center of the screen. The duration of the anticipation was pseudorandomized to avoid subjects learning the clicking rhythm. In the next stage (“reaction (click)” in Fig. 1), a screen with a white, filled square appeared in the place of the fixation cross, and the subjects were instructed to press the response button as quickly as possible to obtain a reward or avoid punishment (depending on the previously displayed cue). After their reaction, participants received the feedback, displaying the result of a given trial and the total amount won so far. After the feedback phase, a fixation point (represented as a “+” in the panel entitled “waiting for the next cue” in Fig. 1). appeared to mark the end of the trial. The complete duration of each trial was 13 s.

The task consisted of 60 trials: 28 trials with reward cues (large win: +5 PLN, small win: +1 PLN), 28 trials with punishment cues (large loss: − 5 PLN, small loss: − 1 PLN) and 4 oddball trials (+/–10 PLN). The oddball trials were not analyzed in terms of the brain activity due to their low repetition number, and their main goal was to keep participants focused on correctly processing the meaning of cues. The sequence in which the cues were presented in the task was pseudo-randomized, but consistent across all subjects (the task code utilised a predefined list dictating the order of cue presentation). A summary of these cues is illustrated in Fig. 5. Participants were shown an identical screen (translated into Polish) at the commencement of the MID task session.

Fig. 5.

Description of the cues from the monetary incentive delay (MID) task. Participants were taught the meaning of these graphical cues during a training session. At the beginning of the neuroimaging session, participants were shown this exact screen to give them the opportunity to recall the monetary value of the cue. However, during the MID task, participants had to rely on their memory.

Prior to entering the scanner, the participants became thoroughly familiar with the task by memorizing the values assigned to specific graphical symbols (cues), and becoming accustomed to providing answers in the same manner as during the subsequent scanning session (the index finger of the right hand). The training session additionally enabled the measurements of the individual reaction times, which were used to tailor the threshold in the MID task to achieve a 60% success rate in the scanner for each participant. This approach was meant to ensure that the participants would be able to win in a sufficient number of trials for the reinforcement processes to take place, simultaneously demanding their continuous focus on the task. This goal was additionally ensured by informing participants that their financial remuneration for participation in the study would be equal to the amount of total gain in the MID task. Participants were told that the guaranteed reward for participating in the study was 50 PLN (13 USD), and that the total final payout they would receive would depend on their performance in the experiment (up to 110 PLN, 30 USD).

Data acquisition

The MRI data was acquired using the 3T MAGNETOM Prisma Siemens scanner at the Laboratory of Brain Imaging (LOBI), Neurobiology Center at the Nencki Institute of Experimental Biology Polish Academy of Sciences.

Scans were performed using a 32-channel head coil. The subjects held a response pad in their right hand. The right button of the response pad was used to complete the task. To ensure optimal visual acuity, participants were offered MRI-compatible glasses if necessary. For each subject T1-weighted anatomical scan, fieldmap, and T2*-weighted scans during tasks were performed. T1-weighted anatomical scans were performed using Echo Planar Sequence (GR\\IR), Slice Thickness = 1 mm, TE = 0.00296 s, TR = 2.3 s, 1 mm³ isotropic voxels. To perform the correction of magnetic field, the fieldmap scans (opposite phase encoded EPIs) were obtained. Task T2*-weighted scan was performed using Simultaneous multislice (SMS) imaging technique, characterized by 397 volumes, TE = 0.03 s, TR = 2 s, Slice Thickness = 2.5 mm, 60 slices acquired in interleaved order, Multiband acceleration factor = 2.

The MID task was programmed in the Presentation software.

Behavioral data analysis

All analyses of behavioral data were performed using the Python programming language (version 3.9.6) and the following libraries: pandas (version 1.5.1), numpy (version 1.22.1), scipy (version 1.7.3), statsmodels (version 0.13.5), matplotlib (version 3.6.2) and seaborn (version 0.13.2). To assess the effect of distinctness and eveningness on psychometric parameters, the Ordinary Least Squares (OLS) regression analysis across multiple dependent variables was performed. The Variance Inflation Factor (VIF) for both predictors was ~ 1, so well below the threshold of 5, dispelling multicollinearity concerns. The values were distributed normally according to the Kolmogorov-Smirnov test (p > 0.05). For each model homoscedasticity, the uniform variance of residuals across the range of predicted values, was confirmed via the Breusch-Pagan test and the Durbin-Watson statistic was used to check the autocorrelation among residuals and attest to the independence of errors. Together, these diagnostic tests confirmed the critical assumptions underlying the linear regression model, establishing a robust foundation for our analysis. The Ordinary Least Squares (OLS) regression model, with distinctness and eveningness as predictors, was fitted using the statsmodels library and followed by False Discovery Rate (FDR) correction for multiple comparisons. Reaction times (RT) outliers during the MID task were calculated using the Z-score for each data point. A Z-score greater than 3 or less than − 3 was considered indicative of an outlier. The linear mixed-effects model (LMM) was performed to account for both fixed effects (circadian characteristics) and random effects (participant-specific variability). Median reaction times for (1) big win (+ 5 PLN), (2) small win (+ 1 PLN), (3) big loss (-5 PLN), and (4) small loss (-1 PLN) were modeled as a function of two predictors: distinctness and eveningness.

MRI data analysis

Image processing and statistical analyses were done in the FMRIB Software Library (FSL⁷⁷).

Anatomical scans were reoriented to the standard (MNI152) orientation, bias-field corrected (RF/B1-inhomogeneity-correction), and non-linearly registered to the standard space, which was followed by brain-extraction, and segmentation⁷⁸. In turn, the fieldmap files were prepared using FSLMERGE and TOPUP. The fMRI data were analyzed using the standard pipeline in the FEAT module (FMRI Expert Analysis Tool^79,80). The first three volumes of the fMRI time series were discarded to allow for signal equilibration. The subsequent procedures included motion, slice timing, and distortion correction, removal of non-brain tissue, spatial smoothing with a Gaussian kernel of 8 mm FWHM and high-pass temporal filtering (90 s). Registration to the MNI152 standard space was obtained with a non-linear approach. The time-series were analyzed with the general linear model (GLM), controlling for the temporal autocorrelation and motion during the scan. All explanatory variables (EV, distinctness) were defined with a 3-column format and double-gamma HRF convolution. For the anticipation phase, we defined the following first-level contrasts: (1) reward anticipation vs. implicit baseline, (2) punishment anticipation vs. implicit baseline, (3) reward anticipation vs. punishment anticipation, (4) big reward vs. small reward, (5) big punishment vs. small punishment. For the feedback phase, GLM tested contrasts for: (1) win feedback (positive outcome) vs. implicit baseline, (2) loss feedback (negative outcome) vs. implicit baseline, (3) win feedback relative to loss feedback.

Two types of analyses were performed. Primarily, we ran group-level whole-brain models using FMRIB’s Local Analysis of Mixed Effects (FLAME 1). Maps (z-stat) for each contrast were thresholded at p < 0.001 (Z = 3.1) for the feedback phase and p < 0.005 (Z = 2.8) for the anftiohase with cluster correction p threshold set to 0.05. The biological clock data, i.e. eveningness and distinctness, were used as factors of interest.

In addition to the whole-brain models, we also performed region-of-interest (ROI) analyses for VTA. We focused exclusively on the VTA, as it is the core part of the mesolimbic system^37,38 and manifests the circadian oscillations across the day³⁹. The decision to use such an approach for the midbrain area was guided by its relatively small size, which deemed the cluster-extent significance thresholding a highly underpowered method. Therefore, VTA mask was created through thresholding a probabilistic ROI derived from an earlier work⁸¹ using the 0.999 cut-off, which was followed by its binarization and resampling to the resolution of fMRI data. The mask was used in the Featquery module of FSL to extract the mean parameter estimates for all the contrasts of interest from the first-level GLM whole-brain analyses. The OLS linear model was run in Python statsmodel (version 0.13.5), using the VTA parameter estimates as the dependent variables, with eveningness and distinctness serving as their predictors.

In the anticipation phase analysis, data from 37 participants were included, while the feedback phase analysis comprised 36 participants due to the exclusion of one individual with an insufficient number of negative feedback trials.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

All authors conceived the project. All authors interpreted the results and revised the manuscript. P.S.: Conceptualization, Study design, Methodology, Participant recruitment, Data collection, Formal analysis, Writing - Original Draft, Visualization, Figures preparation, Funding acquisition; M.R.Z.: Conceptualization, Methodology, Formal analysis, Writing - Original Draft; J.L.: Participant recruitment, Data collection, Writing - review & editing; A.E.S.: Writing - review & editing, Supervision; P.B.: Writing - review & editing, Supervision; I.S.: Conceptualization, Study design, Methodology, Writing - review & editing, Supervision.

Funding

This research was funded by the Ministry of Science and Higher Education (Poland) as a project under the program Excellence Initiative – Research University (2020–2026), decision no.: BOB-IDUB-622-28/2023 (IV.4.1.).

Data availability

We uploaded all data used in described analyses to the Openneuro database (https://openneuro.org/) and made it available in open-access – dataset entitled “Monetary Incentive Delay task - structural and functional images of 37 men; study of associations between circadian characteristics (eveningness, distinctness) and affective processing”⁸². Detailed description and quality assessment of this dataset are provided in our data descriptor manuscript entitled “Circadian rhythmicity and reinforcement processing: a dataset of MRI, fMRI, and behavioral measurements”⁸³.

Declarations

Competing interests

The authors declare no competing interests.