Late dinners and skipped breakfasts tracked dementia risk

Diet research on dementia has spent two decades on the contents of the plate. Mediterranean, MIND, oily fish, leafy greens.

A cohort study in GeroScience asked a different question, using health checkup records for 283,166 older Japanese adults: not what was eaten, but when. Two habits, both ordinary, both recorded with a single yes or no at a routine appointment.

Eating dinner shortly before bed tracked with a higher rate of dementia. So did skipping breakfast. Neither association is large, and the authors are careful about what that means.

Why meal timing might matter to the brain

The reason to ask the question at all is the body clock.

Circadian rhythms include some of the physical, mental, and behavioral changes an organism experiences over a 24-hour cycle, run by a master clock in the brain that the National Institute of General Medical Sciences calls a biological clock.

Food is one of the levers that sets it. Light and dark have the biggest influence on circadian rhythms, but food intake, stress, physical activity, social environment, and temperature also affect them.

And a clock that keeps getting reset appears to cost something. Long-term sleep loss and continually shifting circadian rhythms can increase the risks of obesity, diabetes, mood disorders, heart and blood pressure problems, and cancer. Dementia is not on that list, which is roughly where this study enters.

How the LIFE Study measured meal timing

Evidence on meal timing and dementia is limited, the authors write, which is the gap they set out to fill.

They built the cohort from linked medical insurance claims, long-term care insurance claims, and government-subsidized health checkup data across 19 Japanese municipalities in the Longevity Improvement & Fair Evidence Study. People joined between April 2016 and March 2021.

The exposures are blunt by design. Late-night dinner meant dinner within 2 h of bedtime at least 3 times a week; skipping breakfast meant missing it at least 3 times a week. Both were analyzed as dichotomous exposures (yes/no), answered once, at a checkup.

Dementia was identified in claims data rather than by testing anyone, using the international diagnosis codes health systems bill against.

What happened to dementia rates in the cohort

Among 283,166 participants (mean age 72.8 years; 58.5% women), late-night dinner was reported by 36,609 (12.9%) and skipping breakfast by 17,727 (6.3%).

During follow-up, 18,056 incident dementia cases and 16,390 deaths occurred. Incident here means newly diagnosed during the study, a term that reads like an event and means the opposite.

Both late-night dinner and skipping breakfast were associated with higher hazards of incident all-cause dementia: about 17% higher for the late dinner and about 13% for the missed breakfast.

Percentages of a percentage are easy to misread, so it is worth converting. Roughly six in every hundred people in this cohort were diagnosed with dementia. A 17% relative increase moves that to about seven. Real when spread across a country, small for any one person deciding what time to eat.

The problem this dementia study cannot solve

Dementia begins changing behavior years before anyone gives it a name, and appetite and routine are among the first things it touches. Someone eating late or missing breakfast at 73 may be showing an early sign rather than acquiring a risk.

The design mitigates this without removing it. The authors used prespecified adjustment for baseline demographic, lifestyle, and comorbidity factors and treated death as censoring, so people who died were not counted as dementia-free forever. Neither step can adjust away a symptom nobody has recorded.

The exposures are also single snapshots of habits that change, and self-reported at that. And the cohort is Japanese and older, so the eating patterns and their meanings do not transfer cleanly.

The paper is paywalled, so the authors’ own limitations section was not available for this piece.

Where meal timing sits against known dementia risks

The 2024 Lancet Commission names 14 modifiable risk factors and estimates that 45% of future dementia could potentially be prevented if all 14 risk factors were eliminated.

Meal timing is not among them. The list runs to low education, hearing loss, hypertension, smoking, obesity, depression, physical inactivity, diabetes, excessive alcohol consumption, traumatic brain injury, air pollution and social isolation, plus cholesterol and vision loss.

That is the right frame for this result. Against a smoking habit or untreated hearing loss, a late dinner is a small thing, and the evidence behind it is one observational cohort rather than decades of work.

What to do about your own dinner time

The authors put their conclusion carefully: meal timing behaviors assessed during routine health checkups may be relevant for dementia prevention strategies, although causal inference requires further study.

Read as advice, that is thin. Read as a research direction, it is reasonable: checkups already ask these questions, so the data cost nothing to collect.

For a person, eating earlier and eating breakfast are cheap, safe and supported by other arguments about sleep and blood sugar. Do them for those reasons if you want to. This study is not a reason to worry at 9pm.