Longevity research often searches for rules that apply at every age. A new human study offers a reason to resist that shortcut. Among 2,981 older adults in Stockholm, longer everyday gaps between eating occasions were associated with faster accumulation of chronic disease over as many as 15 years. The signal was concentrated among people aged 78 and older, and the design cannot show that the meal gap caused the difference.
Published August 20 in the Journal of Internal Medicine, the analysis used the Swedish National Study on Aging and Care in Kungsholmen. Participants were at least 60 at enrollment and mostly lived in the community. Researchers estimated each person’s longest usual interval between eating occasions from self-reported meal times, then tracked a standardized count of 60 chronic conditions through repeated clinical assessments and health records.
What the cohort actually found
The shortest-gap group reported a longest daily interval of 6 to 11.5 hours. Compared with that reference, participants in the 14-to-24-hour group accumulated an estimated 0.119 additional chronic conditions per year after adjustment for age, sex, social circumstances, physical activity, smoking, sleep, diet quality, energy and protein intake, reduced food consumption and other factors. Associations appeared for the overall disease count, cardiovascular conditions and neuropsychiatric conditions, but not musculoskeletal disease.
The age split changes the interpretation. Among 1,619 participants younger than 78, the highest-gap group did not show a statistically clear increase in total disease accumulation. Among 1,362 participants aged 78 or older, the corresponding estimate was 0.099 additional conditions per year. The researchers tested 21 alternative modeling assumptions, including adjustments for frailty, functional limitations, meal frequency and medication use, and the main pattern was generally stable.
Those safeguards strengthen confidence that the association is not a simple statistical accident. They do not turn it into an intervention result. People in the longest-gap group were older at baseline, ate fewer meals, more often lived alone, reported lower energy and protein intake, used more medications and already carried more chronic conditions. Statistical adjustment can reduce measured differences, but it cannot recover every unmeasured reason an older person might go longer without eating.
Analysis: two evidence questions are being confused
The key comparison is between two different questions. Randomized fasting trials usually ask whether a planned eating schedule changes weight or an intermediate cardiometabolic marker over weeks or months. This cohort asked whether a person’s habitual meal gap tracked the rate at which multiple diagnosed diseases accumulated over years. A short-term metabolic response and a long-term healthspan trajectory are not interchangeable outcomes.
A 2025 BMJ network meta-analysis of 99 randomized trials and 6,582 adults helps define that boundary. Intermittent-fasting strategies and continuous energy restriction both reduced weight compared with unrestricted eating. Alternate-day fasting produced only a small additional weight reduction compared with continuous restriction, and most trials lasted less than 24 weeks. The synthesis evaluated intermediate cardiometabolic outcomes, not 15-year multimorbidity in adults approaching their eighties.
That makes age more than a subgroup label. Aging can alter appetite, medication routines, muscle’s response to dietary protein, digestive function, social support and the practical ability to prepare food. A long meal gap in a controlled trial may represent a deliberate protocol. The same measured gap in an observational cohort may represent choice, reduced appetite, illness, isolation or functional difficulty. Calling both exposures “fasting” hides those different pathways.
What remains uncertain
The study’s fasting measure was not a verified fasting intervention. It treated the longest interval between reported eating occasions as the exposure and assumed each occasion took no time. A snack and a full meal counted equally when setting the window. Researchers did not know whether participants intended to fast, and self-reported schedules could be misclassified. The cohort was predominantly urban, highly educated and community-dwelling, which limits extrapolation to other settings.
Reverse causation also remains plausible. Disease, frailty, cognitive change or appetite loss may lengthen gaps between meals before those factors are fully captured. The analysis adjusted for many of these possibilities and excluded vulnerable groups in sensitivity tests, but observational data cannot eliminate the problem. Nor does the reported coefficient tell an individual how changing a schedule would affect health.
The most useful next study would separate deliberate time-restricted eating from involuntary meal skipping, enroll enough adults across narrower age and frailty bands, measure food quality and protein distribution directly, and follow clinical function as well as biomarkers. That design would test whether the age split reflects biology, social vulnerability, medication constraints or a mixture of all three.
The durable lesson is about evidence architecture. Longevity claims should be stratified by age, exposure and endpoint before they are generalized. This new cohort does not prove that older adults should avoid fasting, and the trial literature does not prove that short-term metabolic changes extend healthspan. Together, the sources show why a universal meal-timing rule is ahead of the evidence.
Sources: Journal of Internal Medicine; Karolinska Institutet; The BMJ.
TENS Magazine conceptual illustration
