“The darkest hour is just before dawn.”
Our mood or feeling might indeed improve as night gives way to morning, but is there scientific truth behind this? For anyone who has felt the crushing weight of midnight anxiety only to find the experience dissipates as the sun rises after a sleepless night, the answer might seem obvious. Yet until recently, we lacked robust evidence about how our mental health fluctuates throughout the day.
Diurnal mood variation—the phenomenon where mood systematically changes across the day—has been recognised as a core feature of depression for decades (Wirz-Justice, 2008). Many people with depression experience morning lows, afternoon slumps, and evening worsening (Wirz-Justice, 2008). These patterns reflect the profound influence of our biological clocks on mental health. The circadian system, controlled by a master clock in our brain, orchestrates near-24-hour rhythms in hormone secretion (like cortisol and melatonin), neurotransmitter activity, and physiological processes that directly impact our mood (Walker et al., 2020). This biological clockwork helps explain why sleep disruption and circadian misalignment are so tightly linked to mood disorders. A previous Mental Elf blog has explored how circadian rhythms influence our experience of suicidal thoughts.
We also know that seasonal changes profoundly affect wellbeing. The shortening days and reduced light exposure of winter can disrupt circadian rhythms and are associated with lower mood, sometimes triggering seasonal affective disorder (SAD) in vulnerable individuals (Amendolara et al., 2025), and lower overall life satisfaction (as discussed in this Mental Elf blog by Melanie de Lange).
Now, a new study by Bu and colleagues offers unprecedented insights into time-of-day patterns of mental health and wellbeing, analysing nearly 1 million observations from over 49,000 adults during the COVID-19 pandemic period.
Many people with depression experience morning lows, afternoon slumps, and evening worsening, reflecting the influence of our biological clocks on our mental health.
Methods
The researchers analysed data from the University College London COVID-19 Social Study, which collected repeated measurements from 49,218 adults in England between March 2020 and March 2022. Participants completed questionnaires at various times throughout the day, providing a natural experiment to examine diurnal patterns. The study didn’t require people to fill out surveys at specific times. Instead, researchers sent out survey invitations and let people respond whenever was convenient for them. This created a natural experiment where they could see how people’s mental health varied depending on when they chose to respond.
The study measured six aspects of mental health and wellbeing:
- Depression: Patient Health Questionnaire-9, a 9-item questionnaire screening for depressive symptoms.
- Anxiety: Generalised Anxiety Disorder-7, a 7-item tool measuring anxiety symptoms.
- Happiness: Single-item measure representing hedonic wellbeing (pleasure-focused).
- Life satisfaction: Single-item measure also part of hedonic wellbeing.
- Feeling life is worthwhile: Single-item measure representing eudemonic wellbeing (meaning-focused).
- Loneliness: UCLA Loneliness Scale-3, a 3-item questionnaire for social wellbeing
Data were analysed during linear mixed-effects models, testing how time of day, day of the week, season, and year affected each outcome.
Nearly 50,000 adults completed mood assessments at various times throughout the day over two years during the COVID-19 pandemic.
Results
The study initially included nearly 50,000 participants and mostly consisted of women (76.4%), highly educated people (68.1% had degrees), and lacked ethnic minorities (5.9%). To address this bias, researchers used a statistical technique called “weighting” – essentially adjusting the numbers so the results would better represent the general English adult population. This didn’t mean excluding people out of the study; instead, it meant giving more importance to responses from underrepresented groups. After this adjustment, the sample looked much more like England’s actual population: the final sample comprised 50.8% women, 34.2% with higher education, and 14.6% from ethnic minority backgrounds. The age range was well spread out, and about 1 in 5 people (20.2%) self-reported having been diagnosed with a mental health condition.
Time-of-day patterns
The findings reveal a clear diurnal pattern: people generally wake up feeling their best and feel progressively worse as the day wears on, reaching their lowest point around midnight. This pattern was remarkably consistent across multiple mental health measures:
- Depression and anxiety symptoms were lowest in the morning and steadily increased throughout the day, peaking at midnight.
- Hedonic wellbeing (happiness and life satisfaction) showed morning peaks, declined sharply until midday, and then showed a small evening recovery, before dropping to their lowest at midnight.
- Eudemonic wellbeing (feeling life is worthwhile) had the most dramatic variation, with a prominent morning peak, steep midday drop, notable evening recovery, and sharp decline at midnight.
- Social wellbeing (loneliness) remained relatively stable throughout the day, showing the least time-of-day variation.
Day-of-week patterns
Mental health also varied significantly by day of the week:
- Depression and anxiety were highest mid-week (Wednesday/Thursday), suggesting a “midweek slump” effect.
- Wellbeing measures showed greater variation on weekends compared to weekdays, with sharper morning peaks and steeper evening declines.
- Tuesday and Wednesday morning disrupted the usual pattern – people experienced higher depression and anxiety despite the general morning relief, possibly reflecting work-related stress early in the week.
Seasonal effects
The study found robust seasonal patterns in mental health:
- Summer showed the best mental health across all measures, with lower depression/anxiety and higher wellbeing scores.
- Winter had the poorest mental health outcomes, in line with the well-known “winter blues” phenomenon.
- Spring and autumn fell between these extremes, showing intermediate levels of mental health.
Year-on-year improvements
Interestingly, mental health steadily improved across the study period (2020-2022), with all measures showing better scores in later years. This likely reflects adaptation to pandemic conditions and gradual easing of restrictions.
These findings suggest our mental health follows daily, weekly, and seasonal rhythms, with morning generally being our psychological “peak” and late nights our “trough”.
From dawn to dusk, mood fluctuates; peaking in the morning and dipping at midnight.
Conclusions
The authors concluded:
Generally, things do indeed seem better in the morning. People generally reported the worst mental health and wellbeing late in the day and in winter, and there was more variation at the weekends. Looking across different aspects of mental health and wellbeing, hedonic [pleasure] and eudemonic [happiness] wellbeing had the most variation, and social well- being was most stable.
For most people, mental health was best in the morning, dipped mid-week, and was better in summer than winter. Happiness and life purpose varied the most across the day and seasons.
Strengths and limitations
This study’s greatest strength is unprecedented scale – over 49,000 participants completed nearly 1 million observations over two years. The repeated measures design tracked the same people over time to see how their own mood changed throughout the day, rather than just comparing different individuals. By letting people complete survey naturally during their daily routines, researchers captured real-world patterns instead of artificial laboratory conditions. The comprehensive approach – measuring everything from depression to life satisfaction – gives us a complete picture of mental wellbeing.
However, there are several limitations. Most importantly, the entire study took place during COVID-19 (2020-2022), when lockdowns and remote work completely disrupted normal routines. Pandemic anxiety and isolation likely influenced mood patterns in ways that wouldn’t occur in normal times. The study was also limited to England, where daylight hours differ significantly from other parts of the world –what works for London might not apply to Sydney or Stockholm.
In addition, they didn’t measure time spent in natural light exposure, even though light is crucial for regulating our body clocks and mood (Blume et al., 2019). Without knowing whether participants spent time outdoors or stayed inside all day, we can’t determine if these mood changes reflect genuine biological rhythms or just behavioural habits. The study design, as the authors acknowledged, cannot establish causality – we don’t know if morning actually makes us happier, or if happier people are simply more likely to complete morning surveys.
Finally, while depression, anxiety and loneliness were measured using validated questionnaires, happiness and life satisfaction were assessed with single questions. Single items lack the reliability and validity of established scales, potentially explaining why positive emotions showed more variability than negative symptoms.
The COVID-19 context may have skewed results, as daily routines and light exposure were highly disrupted.
Implications for practice
These findings can help us to reshape how we approach mental health care. Clinicians need to recognise that assessment timing matters – a morning appointment might miss evening depression peaks, while afternoon evaluations could underestimate morning anxiety. This suggests the need for time-stamped symptom tracking, perhaps using ecological momentary assessment (EMA) apps that capture mood fluctuations throughout the day.
The morning wellbeing peak supports existing recommendations for early-day bright light therapy in mood disorders (Dollish et al., 2025). It also suggests that we should schedule important work during these natural high points. These patterns have profound implications for service delivery. Crisis helplines should anticipate increased demand during those dark evening hours, particularly on weekends when mood swings are most extreme. Mental health services could optimize resources by boosting evening staff levels and offering morning group sessions when people are most receptive to positive interventions. In workplaces, this research suggests aligning challenging projects with morning clarity and providing extra support during the “midweek slump”.
For researchers, these findings demand a rethink of study design to account for measurement timing. Studies must either standardise measurement times or statistically control for diurnal effects to avoid confounded results. The substantial within-person variation in wellbeing measures highlights the limitations of single-timepoint assessments.
This study opens further research questions:
- Do people with major depressive disorder show blunted morning improvements?
- Could we identify early warning signs by tracking changes in someone’s typical diurnal pattern?
Future studies should explore personalised “chronotherapies” – timing medications, therapy sessions, and interventions to each person’s unique circadian profile. Someone with evening anxiety might benefit from afternoon relaxation techniques, while another battling morning depression might need dawn light exposure.
Mental health services should consider timing when scheduling assessments and interventions to optimize care delivery.
Statement of interests
No conflict of interests to declare.
Links
Primary paper
Bu F, Bone JK, Fancourt D. (2025) Will things feel better in the morning? A time-of-day analysis of mental health and wellbeing from nearly 1 million observations. BMJ Mental Health, 28:1–8. doi:10.1136/bmjment-2024-301418
Other references
Amendolara B, Miller L, Samudrala S, McCarthy MJ. (2025) Morning-evening differences in the post-illumination pupillary response are associated with seasonal mood changes. Chronobiology International. doi: 10.1080/07420528.2025.2490496
Blume C, Garbazza C, Spitschan M. (2019) Effects of light on human circadian rhythms, sleep and mood. Somnologie (Berl), 23(3):147-156. doi: 10.1007/s11818-019-00215-x
De Lange M. (2024) The impact of daylight saving time on life satisfaction. The Mental Elf, 28 October 2024.
Donnelly N. (2025) Are ‘night owls’ more at risk of suicidal thoughts and behaviours? New review on social and circadian rhythm dysregulation. The Mental Elf, 17 March 2025.
Walker WH 2nd, Walton JC, DeVries AC, Nelson RJ. (2020) Circadian rhythm disruption and mental health. Translational Psychiatry, 10(1):28. doi: 10.1038/s41398-020-0694-0
Wirz-Justice A. (2008) Diurnal variation of depressive symptoms. Dialogues in Clinical Neuroscience, 10(3):337-343. doi:10.31887/DCNS.2008.10.3/awjustice
