Impact of intermittent fasting on sleep physiology: A prospective observational study using smartwatch technology

Abstract

Time-restricted eating (TRE) following a 16:8 pattern is a popular form of intermittent fasting with established metabolic benefits. However, its impact on sleep and physical activity remains unclear, particularly when assessed using wearable technologies. This study investigated the short-term effects of TRE on sleep physiology and activity levels by using a commercially available smartwatch. In this prospective observational study, 35 healthy adults (mean age, 30 years; 66% female) completed a 3-day baseline phase, followed by 7 days of a 16:8 TRE regimen. Objective data on sleep, heart rate, and activity were continuously recorded using Withings ScanWatch. Subjective sleep quality, including the Pittsburgh Sleep Quality Index, was assessed pre- and postintervention. Statistical analyses were performed using paired t tests. No significant changes were observed in total sleep duration, light or deep sleep, sleep latency, or nocturnal awakenings between baseline and fasting periods. The daily step count, energy expenditure, and heart rate parameters remained stable. A nonsignificant trend towards slightly reduced deep sleep was noted, which is consistent with the results of previous studies. Subjective sleep assessments reflected objective data, with most participants reporting no perceived changes in sleep quality during intermittent fasting. Short-term 16:8 intermittent fasting did not significantly affect the sleep physiology, physical activity, or heart rate in healthy adults. The subjective and objective measures were closely aligned. These findings support those of previous reviews suggesting that TRE does not influence sleep architecture in healthy populations. Further studies are needed in metabolically at-risk cohorts over longer durations.

1. Introduction

A healthy active lifestyle has become the focus of social debate in recent years. Nutritional interventions remain the cornerstone of weight management strategies, with fasting representing one of the oldest and most widely practiced approaches.^[1–3] Among various fasting regimens, time-restricted eating (TRE), particularly following a 16:8 pattern—16 hours of fasting followed by an 8-hour eating window—has gained popularity because of its simplicity and potential metabolic benefits.^[1,3,4]

Intermittent fasting (IF), including TRE, has been associated with favorable effects on insulin sensitivity, weight reduction, immune modulation, and cellular regeneration.^[3–5] These benefits make IF a promising strategy for disease prevention and healthy aging.^[2,3,6] However, beyond metabolic outcomes, its impact on behavioral and physiological processes, such as sleep remains inconclusive.^[7]

Sleep plays a critical role in both mental and physical health and is closely associated with autonomic function. Disruptions in sleep quality have been linked to an increased risk of depression, anxiety, hypertension, and cardiovascular disease.^[3,8,9] Although dietary habits are known to influence sleep, studies of the effects of TRE on sleep outcomes have yielded heterogeneous results.

In 2021, a review summarized 9 studies and found that TRE had no consistent influence on sleep parameters.^[10] In detail, the authors observed no effects on the Pittsburgh Sleep Quality Index (PSQI). The results for sleep latency and sleep efficiency were variable, with no clear overall data trend. Interestingly, one-third of the included studies employed devices to assess objective sleep parameters and daily activity.^[10]

A recent systematic review of 6 randomized controlled trials by Bohlman et al reported no consistent effects of TRE on sleep parameters.^[11] Interestingly, all studies assessed subjective sleep parameters, while only 3 studies additionally recorded objective markers of sleep quality using the Oura Ring (Oura Health, Finland) or scientific Actigraph devices (Actigraph, USA). However, a closer examination of the included studies revealed subtle patterns, and 1 study reported subjective improvement in sleep quality during IF. In contrast, 1 study found a decrease in sleep duration and an increase in sleep onset latency, and 2 studies reported reduced sleep efficiency. Despite these findings, pooled data analysis suggested no IF-related effects on sleep.^[11] Independent of these observations, patients often subjectively report changes in sleep quality associated with their dietary habits. Perceived sleep quality without changes in objectively measured sleep duration suggests that subjective experiences may not be fully captured by standard study designs.^[12]

Comparable results have been reported for other fasting methods, such as alternate-day fasting (ADF), which alternates unrestricted eating with days of restricted caloric intake (~500–600 kcal). A recent systematic review found no significant influence of ADF on the sleep parameters.^[13] Nonetheless, a 2023 study comparing ADF to continuous calorie restriction found advantages in subjective measures such as perceived sleep quality and daytime fatigue, again without changes in objective sleep metrics.^[14] The less frequently studied 5:2 fasting approach shows similar neutral effects on sleep outcomes, according to current evidence.^[7]

To date, most studies have relied on scientific trackers or self-reported questionnaires such as the PSQI.^[15,16] Although these tools provide valuable insights, they may not fully capture naturalistic, real-world variations in sleep behavior. Consumer-grade wearable devices such as smartwatches offer an opportunity to bridge this gap by providing continuous, unobtrusive, and user-friendly tracking of both sleep and physical activity parameters.

This prospective observational study aimed to investigate the short-term effects of the 16:8 intermittent fasting protocol on both sleep physiology and daily activity levels in healthy adults. We used a commercially available smartwatch (ScanWatch, Withings, France) to capture objective data on sleep architecture, heart rate, and activity complemented by validated subjective assessments.^[17] By integrating multiple data sources in an ecologically valid setting, we sought to explore the complex interactions between intermittent fasting, sleep quality, autonomic function, and physical activity in everyday life.

2. Methods

2.1. Study population

This prospective observational cohort study was conducted at the Ludwig-Maximilians-University (LMU) Hospital in Munich, Germany, between July 7, 2024, and January 15, 2025. A total of 35 healthy adults were enrolled. Participants were recruited through local advertisement targeting students, staff and affiliates of LMU Hospital. Importantly, none of the participants were hospitalized patients, all were community-dwelling individuals. Inclusion criteria were: age ≥ 18 years, self-reported good general health and no current or chronic medical conditions. Exclusion criteria included any history of chronic illness, psychiatric disorders (including eating disorders), known substance use disorder or current pregnancy. All participants provided written informed consent prior to inclusion. The study was conducted in accordance with the Declaration of Helsinki and German Data Protection Laws. The study protocol was reviewed and approved by the institutional ethics committee of LMU Munich, Germany (approval number: #23-0972).

2.2. Data acquisition and study protocol

After a 3-day baseline period under habitual eating conditions, participants were instructed to follow a 16-hour fasting regimen (16:8 protocol) for 7 consecutive days. Each participant chose their own 8-hour daily eating window according to personal preferences and daily schedule. The study team provided standardized written instructions and a brief verbal explanation regarding the fasting protocol, but no specific guidance on caloric intake, meal composition or dietary content was given. Participants were explicitly informed that they could consume noncaloric beverages (e.g., water, unsweetened tea, and black coffee) during the fasting period. No meal tracking, supervision, or personalized dietary support was provided during the intervention period. This approach was designed to simulate real-life adherence to intermittent fasting and maintain the observational nature of the study.

Body weight was recorded at the beginning and end of the intervention period. Participants remained in their usual home environments throughout the entire 10-day study period (3 days baseline, 7 days fasting). No inpatient stays or study site visits were required after the initial enrollment and instruction session.

Each participant received a Withings ScanWatch (Withings, France) for continuous monitoring of physical activity, heart rate and sleep parameters, including total sleep duration, sleep stages (light and deep sleep), the number and duration of awakenings, sleep latency, and wake-up duration.^[17] The watch records these parameters via accelerometer and photoplethysmographic sensors and syncs data automatically via Bluetooth to the Withings smartphone app. Participants were instructed to install the app, pair the device with their personal smartphone and to assure daily synchronization. The study team accessed data through the secured Withings web-based research dashboard at the end of the study period. No participant interaction was needed for data collection beyond the daily synchronization. All measurements were collected passively and unobtrusively, allowing for ecologically valid real-world assessment. No recharging was required during the 10-day study period.

In addition to objective smartwatch recordings, participants completed standardized and custom questionnaires before and after the intervention. Subjective baseline sleep quality was assessed using the PSQI.^[15] The PSQI is a validated self-report instrument comprising 19 items grouped into 7 components: sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medication, and daytime dysfunction. Each component is scored from 0 (no difficulty) to 3 (severe difficulty), with a total global score ranging from 0 to 21, where higher scores indicate poorer sleep quality.

To evaluate perceived short-term changes during the fasting phase, participants also completed a brief study-specific self-report questionnaire developed by the authors. This questionnaire consisted of 4 items addressing general sleep quality on a 3-point Likert scale as well as the 3 components time to fall asleep, nighttime awakenings and perceived restfulness on a 5-point Likert frequency scale, with categories ranging from “never” to “very often [6–7×/wk].” This allowed for a structured comparison of subjective sleep perception before and after the intervention.

2.3. Outcome measures

The primary endpoint was the change in duration of the total, light and deep-sleep phases, as recorded by the smartwatch. Total sleep duration was defined as the cumulative time spent asleep during a single nighttime sleep episode, excluding periods of wakefulness. Sleep stages were classified as “light” or “deep” based on heart rate variability and movement patterns, as per the proprietary algorithm of the Withings ScanWatch.^[17]

Secondary endpoints included changes in sleep latency (time from lying down to sleep onset), number of nocturnal awakenings (discrete periods of wakefulness ≥ 30 seconds detected after sleep onset) and cumulative time awake during the night.

Although there are no universally accepted cutoff values for all parameters in consumer sleep tracking, general reference ranges in healthy adults are: total sleep duration: 7 to 9 hours per night, sleep latency: <20 minutes, deep sleep: 1 to 2 hours (about 20%–25% of total sleep), nocturnal awakenings: 0 to 2 per night with <30 minutes total awake time. These ranges were used descriptively to interpret the cohort’s sleep quality across the baseline and fasting periods.^[18,19]

The ScanWatch continuously recorded daily step count, total energy expenditure (in kcal) and heart rate (in bpm). Energy expenditure was subdivided into passive (resting metabolic rate) and active components (activity-related expenditure), estimated by combining movement data (via accelerometer) and physiological indicators (heart rate and motion intensity). Heart rate was recorded throughout the day and night, with minimum, average and maximum daily values automatically computed by the device’s integrated algorithm. Data were synced via Bluetooth and exported using the Withings research portal for secure storage and analysis.

2.4. Statistical analysis

Categorical variables were presented as percentages. After testing for normality using the Shapiro–Wilk test, continuous variables were reported as means and standard deviations. For comparisons between the groups, we used a two-sided paired t test. All statistical analyses were performed using GraphPad Prism 10 (GraphPad Software LLC, San Diego, CA). Sample size was based on an a priori power calculation. To detect a moderate effect size (Cohen d = 0.5) with 80% power and an alpha level of 0.05 in a paired design, a minimum of 34 participants was required. To account for potential dropout, 35 participants were recruited.

3. Results

3.1. Baseline characteristics

A total of 35 healthy participants were included in this prospective observational study. The cohort was predominantly female (66%, N = 23/35), with a mean age of 30 years and a mean body mass index of 24.9 kg/m². Of all the participants, 11% (N = 4/35) were identified as vegetarians. Two participants (6%) reported practicing regular intermittent fasting prior to the study, but paused this routine to allow for an unbiased baseline assessment. Approximately 25% (N = 8/35) of the patients were current or former smokers. Regular alcohol consumption, defined as intake on at least 1 day per week, regardless of quantity, was reported by 20% of the participants (N = 6/35). Regarding physical activity, 57% (N = 20/35) of the participants engaged in regular exercise. The self-estimated average daily step count at baseline was approximately 9000.

3.2. Subjective sleep quality

Subjective sleep quality was evaluated at baseline using the PSQI, yielding a mean global score of 6.3 ± 2.6. According to standardized cutoffs, 37% (N = 13/35) of participants were categorized as having good sleep quality, 57% (N = 20/35) as moderate, and 6% (N = 2/35) met the criteria indicative of chronic sleep disturbances.

An analysis of the individual PSQI components revealed that prolonged sleep latency and daytime dysfunction were the main contributors to reduced overall sleep quality (Table 1). The participants did not perceive any notable changes in their general sleep physiology during the fasting intervention. In the before-and-after comparisons, no differences were reported in subjective sleep latency, frequency of nocturnal awakenings, or perceived restfulness (Fig. 1).

Table 1.

Results of the PSQI questionnaire at baseline.

PSQI: 6.3 ± 2.6	+ +	+	‐	‐ ‐
Quality	5 (14)	26 (74)	4 (11)	0 (0)
Latency	3 (9)	17 (49)	13 (37)	2 (6)
Duration	23 (66)	8 (23)	3 (9)	1 (3)
Efficiency	2 (6)	24 (69)	7 (20)	2 (6)
Disturbance	4 (11)	31 (89)	0 (0)	0 (0)
Medication	0 (0)	1 (3)	1 (3)	1 (3)
Day dysfunction	10 (29)	12 (34)	12 (34)	1 (3)

Data is presented as N (%); N = 35. Results of individual components are rated form “+ +” through “+,” “‐,” to “‐ ‐,” in order of decreasing sleep quality indicators.

PSQI = Pittsburgh Sleep Quality Index.

Figure 1.

Subjective parameters of sleep quality. Depicted is a longitudinal evaluation of subjective sleep quality parameters at the beginning (pre) and end (post) of our study phase. All differences are not statically significant; N = 35.

3.3. Objective sleep parameters

No significant differences were observed in total sleep duration between the baseline (7.84 ± 1.20 hours) and fasting phases (7.63 ± 1.07 hours; P = .322). Likewise, the sleep stage analysis revealed no notable changes in the amount of light or deep sleep. The number of nocturnal awakenings was about 1.6 to 1.8 in both groups (P = .347), and the cumulative duration of wake phases during sleep remained unchanged at around 22 minutes throughout the study (P = .990). Sleep latency (3.65 vs 2.91 minutes, P = .515) and wake-up duration (3.92 vs 3.09 minutes; P = .278) were also comparable at baseline and during the fasting period (Table 2).

Table 2.

Objective smartwatch measurements.

	Baseline	Fasting	P value
Sleep parameters
Total duration (h)	7.84 (1.20)	7.63 (1.07)	.322
Light sleep (h)	4.47 (1.09)	4.41 (1.03)	.743
Deep sleep (h)	3.38 (0.73)	3.21 (0.72)	.106
Time awake (min)	22.3 (13.4)	22.3 (14.0)	.990
Awakenings (N)	1.59 (1.12)	1.79 (1.15)	.347
Sleep latency (min)	3.65 (6.22)	2.91 (2.56)	.515
Time to wake-up (min)	3.92 (3.32)	3.09 (3.88)	.278
Heart rate
Average HR (bpm)	64.2 (9.4)	63.7 (8.3)	.642
Minimum HR (bpm)	55.3 (8.0)	55.1 (7.0)	.838
Maximum HR (bpm)	84.2 (13.5)	82.6 (10.3)	.342
Activity
Energy passive (kcal)	1544 (254)	1552 (255)	.467
Energy active (kcal)	309 (300)	349 (426)	.348
Steps (N)	7704 (2744)	7554 (2920)	.708

Data is presented as mean (SD); N = 35. Units are indicated as follows: h (hours), min (minutes), bpm (beats per minute), HR (heart rate), kcal (kilocalories).

Beyond sleep-specific parameters, the average heart rate—both in mean values (64.2 vs 63.7 bpm; P = .642) and extremes—remained stable over the entire study period, showing no significant influence from intermittent fasting (Table 2).

3.4. Effects on daily activity

The resting energy expenditure averaged slightly above 1500 kcal/day and showed no significant difference between the baseline and fasting periods (P = .467). Active energy expenditure remained stable at approximately 300 to 350 kcal/day (P = .348). Objective step counts, recorded by the smartwatch, averaged around 7500 steps per day during the intervention, somewhat lower than the self-reported baseline estimate but similarly unaffected by the fasting protocol (P = .708) (Table 2).

4. Discussion

This prospective observational study aimed to investigate the short-term effects of a 16:8 IF regimen on sleep behavior and daily physical activity in healthy adults using a commercially available smartwatch for continuous monitoring. To our knowledge, this is one of the first studies to employ consumer-grade wearable technology to objectively assess the impact of IF on sleep architecture and activity in real-life settings.

Our findings demonstrated no significant differences in objectively measured sleep parameters between baseline and fasting periods. The total sleep time, sleep latency, number and duration of nocturnal awakenings, and time spent in light and deep sleep remained unchanged. A nonsignificant trend toward slightly reduced total sleep time and deep sleep duration was observed, which is consistent with previous isolated studies that reported similar patterns under intermittent fasting conditions.^[20] However, the clinical relevance of such subtle variations remains unclear.

In parallel, no significant changes were detected in daily step counts, active or resting energy expenditure, or average heart rate, indicating that intermittent fasting did not measurably alter daytime physical activity or cardiovascular parameters. These findings align with the results of 2 recent systematic reviews that concluded that IF protocols, including time-restricted eating, have no consistent or significant effects on sleep physiology or physical activity when assessed objectively.^[10,11]

Subjective assessments via validated questionnaires closely reflected objective measurements. Most participants reported no perceived changes in sleep quality, latency, or restfulness. These results suggest a high degree of concordance between the subjective and objective measures in our cohort and support the hypothesis that short-term intermittent fasting may not influence sleep behavior in healthy individuals.

From a conceptual perspective, this study adds to the growing body of literature suggesting that the metabolic and circadian benefits of intermittent fasting do not necessarily translate into measurable changes in sleep architecture, at least in the short term and among healthy young adults. Furthermore, the use of commercial wearable devices such as Withings ScanWatch enabled unobtrusive, continuous, and ecologically valid data collection, demonstrating their potential utility in lifestyle-based research.

This study had several limitations must be acknowledged. First, the study was designed as an exploratory trial with a relatively small sample size and powered to detect moderate effects (Cohen d = 0.5). Second, the intervention period was short, and it remains unclear whether a longer exposure to intermittent fasting would yield different outcomes. Third, our cohort consisted exclusively of healthy, relatively young participants, which limits the generalizability of our results. The effects in individuals with obesity, metabolic syndrome, or other comorbidities may differ and should be addressed in future studies.

5. Conclusions

In this exploratory study, we found no significant short-term effects of the 16:8 intermittent fasting regimen on sleep physiology, physical activity, or cardiovascular parameters in healthy adults, as measured by a commercially available smartwatch. Subjective reports of sleep quality were consistent with objective data, suggesting that intermittent fasting did not notably affect perceived or measured sleep behavior in this cohort. Our findings support the growing consensus that time-restricted eating does not significantly affect sleep architecture or daytime activity in healthy populations. Future studies should explore whether longer interventions or studies in metabolically compromised populations may reveal different effects.

Author contributions

Conceptualization: Christopher Stremmel.

Data curation: Jenny Schlichtiger, Antonia Kellnar, Christopher Stremmel.

Formal analysis: Jenny Schlichtiger, John Michael Hoppe, Antonia Kellnar, Christopher Stremmel.

Investigation: Anna Strüven, Jenny Schlichtiger, Isabel Thiessen.

Methodology: Jenny Schlichtiger, Antonia Kellnar, Christopher Stremmel.

Project administration: Christopher Stremmel.

Resources: Christopher Stremmel.

Software: Jenny Schlichtiger, Christopher Stremmel.

Supervision: Anna Strüven, Christopher Stremmel.

Validation: Anna Strüven, John Michael Hoppe, Isabel Thiessen, Antonia Kellnar, Christopher Stremmel.

Visualization: Christopher Stremmel.

Writing – original draft: Anna Strüven, Christopher Stremmel.

Writing – review & editing: Jenny Schlichtiger, John Michael Hoppe, Isabel Thiessen, Antonia Kellnar, Christopher Stremmel.