A large real-world study suggests that what people eat during the day may influence not only how long they sleep, but also the biological quality of that sleep. Researchers analyzing 4,793 nights of dietary records and wearable sleep measurements found that plant-forward eating patterns were associated with more deep and REM sleep, less light sleep and a lower average heart rate during the night. The findings point to daily diet as a potentially immediate regulator of sleep physiology, adding to evidence that nutrition and sleep are linked through metabolic, hormonal and autonomic pathways.
The study, published in Nature Health, examined people living their normal lives rather than participants confined to a sleep laboratory. Volunteers recorded their food intake in real time, while wearable devices tracked multiple stages of sleep and overnight cardiovascular signals. This allowed the researchers to compare day-to-day changes in eating habits with the sleep that followed on the same night. Because each person could serve as a partial comparison for themselves, the design helped reduce the influence of relatively stable factors such as age, long-term health status and habitual sleep patterns.
The investigators used a lagged-variable framework, meaning that the previous day’s diet and sleep were included as baseline information when estimating the relationship between that day’s food choices and the subsequent night’s sleep. This approach is important because sleep and diet are closely entangled. A poor night of sleep can alter appetite, meal timing and food selection the next day, while stress, exercise, illness and work schedules can affect both eating and sleeping. By accounting for the preceding day’s measurements, the researchers attempted to isolate the short-term association between dietary variation and later sleep physiology.
To estimate effects, the team applied inverse probability weighting, a statistical method commonly used in observational causal-inference research. The technique reweights observations so that people or nights with different characteristics become more comparable with respect to the dietary exposure being studied. The researchers also used bootstrap procedures to quantify uncertainty, repeatedly resampling the data to produce confidence intervals around the estimates. Although this type of analysis cannot establish causation with the certainty of a randomized clinical trial, it is designed to move beyond a simple correlation by addressing several major sources of confounding in free-living data.
Fibre density produced some of the clearest associations. Nights following days with a greater amount of fibre relative to total energy intake showed an average increase of 0.59 percentage points in deep sleep, with a 95 percent confidence interval ranging from 0.25 to 0.94 percentage points. REM sleep rose by 0.76 percentage points, while light sleep declined by 1.35 percentage points. The same dietary pattern was linked to a reduction of approximately 1.14 beats per minute in mean nocturnal heart rate. Deep sleep is generally associated with physical restoration and metabolic regulation, while REM sleep contributes to memory processing, emotional regulation and neural plasticity. Even modest shifts in sleep-stage composition could therefore be meaningful if they occur repeatedly over months or years.
The results also highlighted the potential importance of dietary diversity rather than a single nutrient alone. Participants who consumed a larger number of unique plant-based food items in a day tended to have lower nocturnal heart rates. Greater intake of whole-plant foods showed a similar relationship. In addition, the Dietary Phytochemical Index, a measure intended to reflect the proportion of energy derived from foods rich in plant compounds, was associated with a more restorative sleep profile. Higher phytochemical intake corresponded to a 0.62-percentage-point increase in REM sleep, a 1.00-percentage-point reduction in light sleep and a decrease of 0.93 beats per minute in average nocturnal heart rate.
These findings may reflect several biological mechanisms, although the study was not designed to prove which mechanisms are responsible. Fibre is fermented by gut microorganisms, producing short-chain fatty acids that can influence immune activity, glucose regulation and signaling between the gut and brain. Plant foods also supply potassium, magnesium, folate, polyphenols and other compounds involved in vascular function and nervous-system regulation. A diet rich in whole plants may reduce post-meal glucose fluctuations and systemic inflammation, processes that could otherwise interfere with stable sleep or maintain higher nighttime sympathetic nervous-system activity. The lower heart rate observed with plant diversity and phytochemical intake may indicate reduced autonomic arousal during sleep.
Meal timing showed a different pattern from food composition. A heavier evening meal, defined as a larger percentage of total daily energy consumed during the six hours before bedtime, was associated with 7.7 additional minutes of total sleep time. However, it was also linked to an increase of 0.73 beats per minute in mean nocturnal heart rate. This combination suggests that eating more energy late in the day may extend sleep duration without necessarily improving physiological recovery. Digestion, thermogenesis, glucose handling and hormonal responses remain active after a substantial meal, potentially increasing cardiovascular workload during the night even when total sleep time rises.
Not every dietary feature produced a reliable result. Short-term changes in the proportion of calories supplied by carbohydrates, fats or protein, as well as variations in micronutrient consumption, did not show robust associations after correction for multiple statistical comparisons. The researchers also tested more extreme contrasts in dietary exposure. Under those conditions, the estimated effects became larger but remained directionally consistent with the main analysis. That pattern strengthens the possibility that the observed relationships are not random, while also suggesting that ordinary day-to-day changes may have relatively modest effects that accumulate through repetition.
The study offers a fast-moving scientific message suited to an era of continuous health tracking: sleep may respond measurably to dietary choices made only hours earlier. Yet the results should not be interpreted as proof that adding fibre or moving calories to a particular part of the day will automatically transform sleep for everyone. Wearable devices estimate sleep stages rather than measuring brain activity directly with laboratory-grade polysomnography, and observational dietary logs can contain errors or omissions. Unmeasured influences, including exercise, alcohol, stress, medication use and illness, may also affect the results. Even so, the large number of person-nights, real-time dietary records, lagged analysis and objective physiological measurements provide a detailed picture of how everyday eating patterns may shape the night that follows. The emerging implication is not that one “sleep food” exists, but that a diverse, plant-rich diet and thoughtful meal timing could become practical components of strategies aimed at improving restorative sleep and overnight cardiovascular health.
Subject of Research: The relationship between daily dietary patterns, meal timing and objective sleep architecture and nocturnal heart rate.
Article Title: Impact of daily diet on sleep quality
Article References: Shkolnik, M., Sapir, G., Shilo, S. et al. Impact of daily diet on sleep quality. Nat. Health (2026). https://doi.org/10.1038/s44360-026-00182-2
Image Credits: AI Generated
DOI: https://doi.org/10.1038/s44360-026-00182-2
Keywords: sleep architecture, diet, dietary fibre, plant-based foods, REM sleep, deep sleep, meal timing, nocturnal heart rate, phytochemicals, wearable health monitoring
