In a study of 26,820 UK Biobank participants aged 53 to 86, people who described themselves as definite evening types had higher scores on composite tests of reasoning, memory and reaction time than definite morning types. The association appeared in two separate groups drawn from the database.
The finding, published in BMJ Public Health in 2024, does challenge the familiar idea that early rising is a reliable sign of mental sharpness. It does not show that staying up late makes anyone smarter, or that changing a sleep schedule would improve cognition.
This is not medical advice. I am reading one observational study and asking what it does, and does not, show. This is one study, not settled consensus.
What the researchers actually measured
Lead author Raha West and colleagues analyzed sleep and cognitive data from the UK Biobank, a large research resource that follows volunteers in Britain. The participants were divided into two cohorts according to which computerized cognitive assessments they had completed.
Cohort 1 included 10,067 people who completed four tasks: fluid intelligence or reasoning, pairs matching, reaction time and prospective memory, which involves remembering to carry out an intended action. Cohort 2 included 16,753 people who completed only the pairs-matching and reaction-time tasks.
Chronotype was determined with one self-report question. “Definitely a morning person” became the morning group, while “definitely an evening person” became the evening group. People who leaned morning or evening without choosing either extreme were combined into an intermediate category.
The researchers converted the cognitive results into standardized scores and combined them into a global score for each cohort. After adjusting for age, sex, sleep duration and several health and lifestyle variables, evening types scored 0.135 standard deviations above morning types in Cohort 1 and 0.075 standard deviations above them in Cohort 2. Intermediate types also scored above morning types in both groups.
Those are modest differences. Some coverage translated the coefficients into a 13.5 percent and 7.5 percent advantage, but a standardized-score difference is not quite the same as saying that one group was 13.5 percent more intelligent. The paper measured performance on a defined set of tasks, not intelligence as a whole.
The study found an association, not an effect of staying up late
The cross-sectional design gives the researchers a snapshot. Chronotype, reported sleep and cognitive scores were compared within the same broad period, rather than followed through a planned change over time.
That means the direction of the relationship is unresolved. Evening preference could be related to cognition through some unmeasured factor. Cognitive ability, education, work history or daily routine could influence when a person prefers to be active. Both could arise partly from other differences between participants.
Changing bedtime was not part of the study.
The authors adjusted for several plausible confounders, including age, sex, cardiovascular conditions, diabetes, smoking and body mass index. They also acknowledged important omissions. Educational attainment was not included because of incomplete data, and neither were physical activity, social activity or depression. UK Biobank participants are volunteers and are not a perfect cross-section of the general population.
The evidence therefore supports a narrower sentence: among these older UK Biobank participants, self-described evening types tended to score somewhat higher on these particular cognitive assessments.
The time on the testing-room clock was missing
The paper did not control for when participants completed their cognitive tests. This is a consequential gap because performance can change across the day, and the pattern may depend on a person’s chronotype.
A 2023 review of time of day, chronotype and cognition described a common “synchrony” pattern: people often perform better when testing occurs closer to their preferred time. The size and consistency of that effect vary by task and age, but it is strong enough that the review’s authors recommended recording both chronotype and test time.
The missing clock time matters.
If more evening types happened to take the UK Biobank tests later in the day, synchrony could explain some of their advantage. If everyone was tested at similar daytime hours, the interpretation might be different. The 2024 paper cannot separate those possibilities.
Age adds another boundary. The participants averaged 71 or 72 years old in the two cohorts, and chronotype generally shifts earlier across adulthood. A definite evening preference at 70 may identify a different mix of people and circumstances from the same preference at 20. The findings should not be carried straight into claims about students or young workers.
This was not the first UK Biobank result to favor evening types
The broad pattern did not appear from nowhere. In 2017, Simon Kyle and colleagues published a much larger cross-sectional UK Biobank analysis in Sleep Medicine. Across hundreds of thousands of participants, evening chronotype was associated with better performance on reasoning, reaction time, visual memory and prospective memory tasks, while morning chronotype was associated with poorer performance.
That earlier analysis adjusted for a wider set of variables, including education and socioeconomic status. Finding a similar pattern twice is useful, but both analyses came from the same research resource. They are not two independent populations.
Evidence from elsewhere complicates the causal story. A 2022 study in eClinicalMedicine examined 14,582 older adults in the Guangzhou Biobank Cohort. Later sleep timing was again associated with better scores in conventional observational analysis.
The researchers then used Mendelian randomization, a method that uses genetic variants as instruments to probe possible causal directions. Genetic predisposition toward morningness was not associated with cognitive performance. Better genetically predicted cognitive function, however, was associated with lower odds of being a morning type. The method has assumptions of its own, but its result does not support the simple claim that later sleep timing creates sharper cognition.
Sleep duration was a separate result
The 2024 study also compared reported sleep duration. Seven to nine hours was associated with a slightly higher composite score than fewer than seven hours in Cohort 1, but not in Cohort 2. More than nine hours was associated with lower scores in both cohorts.
Those findings are observational too. Longer sleep can accompany other differences in health or daily life, and a single self-reported estimate does not measure sleep with the precision of a wearable device or laboratory recording. The paper is not an argument for cutting sleep to become more alert, nor for moving bedtime later.
Chronotype describes a preference for timing. Sleep duration describes how long someone sleeps. Social schedules determine whether a preference can be followed. Treating all three as the same thing creates advice that the data cannot justify.
What a stronger test would need to do
A more decisive study would measure sleep timing objectively over multiple days, record the time of every cognitive assessment and test people at both preferred and non-preferred hours. It would include younger and more diverse populations, adjust carefully for education and work schedules, and follow participants long enough to establish which changes came first.
An intervention could then ask a harder question: does aligning a person’s schedule with their existing chronotype improve performance, without assuming that one chronotype is inherently better?
No reader should use a cross-sectional result to force a later schedule or ignore persistent sleep disruption. If sleep timing or sleep loss is causing ongoing difficulty in daily life, a qualified clinician can assess the individual situation more usefully than a population study can.
For now, the study weakens a cultural stereotype. It does not replace the stereotype of the virtuous early riser with a new hierarchy led by night owls.
