Those who started eating after 12:00 had a higher risk of death — a large study

The link became progressively stronger: the later the first meal of the day was eaten, the higher the average risk turned out to be. However, the study was observational. It does not prove that eating a late first meal in itself shortens life: mealtimes may be linked to sleep, work, health and a host of other factors.

Details

The authors used data from the US National Health and Nutrition Examination Survey (NHANES) for the period 1999–2018.

The analysis included 31,044 people aged 40 and over. The median follow-up period was 8.6 years.

During this period, the researchers recorded 7,129 deaths and compared these with the times at which participants typically began and finished eating throughout the day.

The main question was simple: is there a link between the timing of the first and last meals and overall and cardiovascular mortality?

The results painted a fairly consistent picture.

The researchers used people whose first meal took place between 7.00 and 8.00 as the reference group.

Compared with this group, the risk of death from any cause was:

• 10 per cent higher for those eating their first meal between 8:00 and 10:00;
• between 10:00 and 12:00 — 19% higher;
• after 12:00 — 29 per cent higher.

For the latter group, the hazard ratio was 1.29, and the 95 per cent confidence interval was 1.10–1.51. In other words, the observed association remained significant even after statistical adjustments.

The authors also found associations between a later start to eating and mortality from cardiovascular disease.

However, ‘29 per cent higher’ refers to relative risk, not a claim that a person literally has an additional 29 per cent chance of dying.

The researchers looked separately at when people last ate during the day.

For comparison, they used a group whose last meal was between 7.00 pm and 8.00 pm.

For those who continued to eat after midnight, the relative risk of death from all causes was 27 per cent higher.

Interestingly, an increased risk was also observed at the other end of the scale. For those who finished eating before 7.00 pm, the risk was approximately 13 per cent higher than for the 7.00 pm–8.00 pm group.

This means that the results cannot be reduced to a simple rule such as ‘the earlier you have dinner, the better’.

The relationship regarding the timing of the last meal proved to be more complex and resembled a U-shape: a higher risk was observed both when the meal was finished very late and when it was finished earlier.

And this certainly does not mean that the researchers have established a universally ‘correct time for dinner’ — 19:00 or 20:00.

The authors attempted to translate the statistical correlations they found into a more understandable measure — life expectancy after the age of 50.

According to their calculations, for people who ate their first meal after midday, life expectancy was on average 2.46 years shorter than for the group who began eating between 7.00 and 8.00.

For those whose last meal was after midnight, the estimated difference was 2.35 years compared with the 19:00–20:00 group.

However, these figures are quite easy to misinterpret.

The researchers did not observe how the two groups of people lived their entire lives and then conclude that those who had a late breakfast died exactly 2.46 years earlier.

This is a statistical estimate of life expectancy, calculated on the basis of observed mortality rates. It would therefore be incorrect to state that ‘eating after midnight takes 2.5 years off your life’.

Another important detail is that the researchers analysed the first calorie-containing meal of the day, rather than a traditional breakfast.

If a person woke up late and ate for the first time at 13:00, for the purposes of the analysis this was the ‘first meal after 12:00’, regardless of whether the participant themselves referred to this meal as breakfast, lunch or something else.

Therefore, the results do not allow for a simple conclusion:

‘Skipping breakfast increases the risk of death by 29 per cent’.

The study specifically measured the time at which the daily period of calorie intake began.

This is particularly important for people with different sleep patterns. Someone who wakes up at 11 am and eats at midday is in a completely different situation to someone who got up at 6 am but deliberately fasted until 12:00.

The time on the clock alone does not fully reflect a person’s internal circadian rhythm.

Why might the timing of meals matter at all?

Our metabolism depends on circadian rhythms — approximately 24-hour biological cycles that regulate sleep, hormones, body temperature and a host of metabolic processes.

Consequently, the body may process the same food differently at different times of the day.

This is precisely what chrononutrition is all about — a field of research that examines the interaction between mealtimes, sleep and the body’s internal biological clock.

Previous studies had already linked skipping meals and irregular eating patterns to increased mortality. For example, an analysis of US data revealed a link between eating just one meal a day and higher overall and cardiovascular mortality rates.

However, the existence of a biologically plausible mechanism does not necessarily mean that it is the one that explains the new findings.

Eating late may be a sign of entirely different problems. This is the main limitation of the study.

The study was observational, and participants were not randomly assigned to groups with their first meal at 7.30 am, 10.30 am or 1.00 pm.

People simply went about their daily lives as usual.

And someone who starts eating after midday may differ from someone who has breakfast at seven in the morning in dozens of ways.

For example, among those who start eating late, there may be a higher proportion of people who:

work night shifts, go to bed and wake up late, suffer from sleep disorders, have chronic illnesses or a poor appetite, or practise intermittent fasting.

The authors statistically accounted for many social, behavioural, medical and dietary characteristics of the participants, but such adjustments never allow for the complete elimination of all hidden differences.

Therefore, the correct conclusion is as follows:

eating the first meal after 12:00 was associated with increased mortality — but the study did not prove that moving the first meal to the morning would, in itself, prolong life.

There is also another significant limitation.

The times of the first and last meals were determined using a 24-hour dietary recall — a questionnaire in which a person recalls what they ate and when over the previous 24 hours.

This is neither a long-term food diary nor a device that constantly monitored meals.

It turns out that a single specific day was used as an indicator of a person’s routine. But that day might have been unusual: a participant might have woken up late, been travelling, skipped a meal or come home much later than usual.

This increases the likelihood of error when assessing habitual mealtimes.

This is precisely why the results from a large sample are interesting at a population level, but are currently unsuitable for making strict recommendations to an individual.

Is it now essential to eat at 7 am?

No.

The study does not prove that there is a universal ideal time for breakfast or dinner.

Nor can it be concluded from it that a person who is used to sleeping from 2 am to 10 am should force themselves to eat at 7.30 am.

Furthermore, the authors did not compare the effect of mealtime with the influence of dietary quality; therefore, the claim that ‘when you eat is just as important as what you eat’ is not supported by this study.

The findings point to a more specific point: mealtimes may be yet another indicator linked to health and longevity, alongside diet composition, food intake, physical activity, sleep and other factors.

And the most striking pattern in the sample was simple: the later the first meal of the day, the higher the observed risk of death.

Whether this is due to the body’s internal clock, lifestyle, health status or a combination of these factors remains to be determined.

Source

Study:“Meal timing and eating window with all-cause and cardiovascular mortality and life expectancy: a population-based cohort study”.

Authors: Wen Hu, Mengyao Wang, Sibo Liu and co-authors.

Journal: European Journal of Clinical Nutrition, 2026.