Rationale and protocol for time-based energy intake goals within an obesity care lifestyle intervention: Daily eating patterns for total health study

Abstract

This publication outlines the rationale and protocol for the Daily Eating Patterns for Total Health (DEPTH) trial, detailing its motivation, aims, design, intervention, and measures. This randomized clinical trial examines how time-based energy goals applied within a 12-month obesity care lifestyle intervention influences percent weight loss, dietary temporal patterns, sleep regularity, and appetite regulation. One hundred seventy-four adults (ages: 25–60 y; BMI: 27–45 kg/m²) will be randomized to one of three interventions: DEPTH-Morning, DEPTH-Evening, or DEPTH. All participants will receive 33 sessions of an obesity care lifestyle intervention that includes five separate dietary goals: (1) 1200–1500 kcal/day, (2) ≤ 30% energy from fat, (3) a 12-h eating window, (4) first eating occasion within 60 min of awakening, and (5) 3 meals and 1 snack per day. They will also aim for 200 min/wk. of moderate- to vigorous-intensity physical activity. DEPTH-Morning participants will be given the goals to consume 70% of energy in two meals and one snack during the first six hours of the 12-h eating window, and 30% of energy in one meal in the last six hours of the eating window (morning-loaded distribution). DEPTH-Evening will be given goals to adopt the opposite pattern (afternoon/evening-loaded distribution). DEPTH participants will have no specific time-based energy intake goals. Measures of anthropometrics, dietary intake, sleep, and appetite regulation will be assessed at 0, 3, 6, and 12-months. This study will provide insights into the role of time-based energy intake goals, a form of chrononutrition, on obesity care outcomes.

1. Introduction

Dietary strategies used in obesity care traditionally focus on how much and what foods to eat [1,2]. More recently, there has been growing interest in meal timing strategies, such as how energy is distributed throughout the day, a form of chrononutrition, which is believed to promote circadian entrainment (coordination of the internal circadian clock to external rhythmic time-cues) [3–6]. A few investigations have examined time-based energy intake goals within reduced energy diets during brief (≤3 months) obesity care interventions and report that goals that encourage more energy intake in the morning and less energy intake in the afternoon and evening produce better weight loss outcomes than the reverse pattern [7–11]. However, these studies have not provided details on the actual temporal eating patterns achieved or whether circadian entrainment occurs.

To begin to address these gaps, we conducted an 8-week pilot study examining the influence of energy distribution timing on weight loss and circadian rhythms, using sleep onset and wake times as markers of circadian alignment [12]. Our morning-loaded energy group experienced a greater percent weight loss and improved regularity in sleep onset and wake times, suggesting stronger circadian entrainment than the evening-loaded energy group.

One hypothesized mechanism for greater weight loss when more energy is consumed earlier in the day is improved appetite regulation [11,13,14]. While hunger reaches its highest level around 8 pm, morning-loaded energy intake goals within reduced energy diets have produced reduced hunger, diminished desire to eat, and increased feelings of fullness compared to afternoon or evening-loaded energy intake goals [11,14]. However, the direct link between appetite regulation and weight loss remains unexplored in this area.

Chronotype, an individual’s preferred timing of daily activities, may moderate the effect of time-based dietary goals [15]. Evening chronotypes generally consume more energy later in the day, which may raise obesity risk and reduce success in obesity treatment [16]. This chronotype may more greatly benefit from time-based energy intake goals encouraging greater energy intake earlier in the day.

This publication describes the rationale and protocol of the Daily Eating Patterns for Total Health (DEPTH), a 12-month randomized clinical trial (RCT) designed to enhance understanding of the influence of energy distribution timing on long-term weight loss outcomes. The primary aims of the study are to learn the influence of time-based energy intake goals on longer-term weight loss and the influence of these goals on eating temporal patterns, sleep regularity, and appetite regulation. It is hypothesized that a morning-loaded energy distributed pattern will improve percent weight loss (primary aim), that this energy distribution pattern will produce greater sleep regularity and enhance appetite regulation (primary aim), and enhanced appetite regulation will mediate percent weight loss (exploratory aim). In addition, it is hypothesized that chronotype will moderate the relationship between the morning-loaded energy distribution pattern and percent weight loss (exploratory aim) (see Fig. 1). This investigation will use state-of-the-art methods to document temporal eating patterns and circadian entrainment.

Fig. 1. Study hypotheses.

A morning-loaded energy distributed pattern will improve percent weight loss, that this energy distribution pattern will produce greater sleep regularity and enhance appetite regulation, and enhanced appetite regulation will mediate percent weight loss. Additionally, chronotype will moderate the relationship between the morning-loaded energy distribution pattern and percent weight loss.

2. Methods

2.1. Study designs and aims

DEPTH is a 12-month RCT in which participants will be randomized to one of three, 12-month, group-based delivery, obesity care lifestyle interventions: DEPTH-Morning, DEPTH-Evening; or DEPTH. This trial is funded by the National Institute of Diabetes and Digestive and Kidney Disease (1R01DK137752–01) and is registered at ClinicalTrials.gov (NCT06455995). The study was approved by the University of Tennessee Knoxville Institutional Review Board (UTK IRB). All participants provide written informed consent, and assessments will occur at 0, 3, 6 and 12 months.

2.2. Participant eligibility criteria and recruitment

This investigation will include 174 adults with overweight or obesity, aged between 25 and 60 years, with 45% and 33% self-identifying as male assigned at birth and from under-represented groups, respectively. This age range was selected to limit age-related differences in sleep, appetite regulation, and circadian timing [17–19]. A body mass index (BMI) between 27 and 45 kg/m² was selected to determine eligibility because it allows for 10% weight loss prior to reaching a BMI of <25 kg/m², and a BMI > 45 kg/m² requires more medical supervision than is provided in this trial [20].

Participants will be excluded if they report: 1) <5 days/wk. consuming energy (≥ 100 kcal) prior to 12 pm and inability to consume ≥50 kcal within one hour of awakening each day; 2) using sleep medication or <5 nights/wk. getting at least 6 h of total sleep; 3) being a shift worker/alternative shift worker who works outside of 7 am and 7 pm; 4) being diagnosed with type 1, or type 2 diabetes and taking medication that requires eating to occur at certain time periods; 5) having a heart condition, chest pain during periods of activity or rest, or loss of consciousness on the Physical Activity Readiness Questionnaire (PAR-Q) [21]; 6) not able to walk for 2 blocks without stopping; 7) having major psychiatric diseases or organic brain syndromes; 8) current participation in a weight loss program and/or taking weight loss or appetite regulation medication or having lost ≥5% of body weight in the past 6 months; 9) having underwent bariatric surgery or planning to undergo bariatric surgery in the next 12 months; 10) being pregnant, lactating, <6 months post-partum or planning to become pregnant (next 12 months); 11) planning to move outside of Knoxville metropolitan area within the period of the investigation; or 12) not having daily access to a PC or smartphone with online connectivity.

Participants will be recruited primarily through social media advertisements, with efforts to increase the participation of males and under-represented groups. Interested participants will complete a phone screening, and eligible participants will be invited to an in-person orientation, where informed consent will be obtained.

Randomization.

Factors such as sex assigned at birth (male vs female), weight status (overweight vs obesity), and under-represented group status (under-represented vs not underrepresented) are important prognostic variables. Stratification by prognostic factors is commonly used to avoid imbalances during randomization but may result in small or empty cells when the total sample size is small. This is also particularly problematic in cases (such as the current trial) where the treatment trial design involves group-based intervention wherein multiple participants simultaneously attend the same intervention and/or are unable to meet at certain times due to schedule conflicts (the three intervention groups are offered at different times on the same evening of the week for each cohort (5:30–6:30 pm, 6:00–7:00 pm, 7:00–8:00 pm), with cohorts having intervention on different evenings of the week (Monday to Thursday). To address this issue, the treatment group times (earlier evening [5:30–6:30 pm], middle evening [6:00–7:00 pm] vs later evening [7:00–8:00 pm]) will be randomly assigned to DEPTH-Morning, DEPTH-Evening, or DEPTH intervention groups using an “urn design” adaptive probability algorithm, as described by Stout and colleagues [22]. The urn design is commonly employed in clinical trials with small sample sizes (〈200) and where multiple prognostic factors need to be balanced. Simulations demonstrate that this procedure produces a higher probability of balanced randomization to treatment conditions than traditional fixed allocation procedures such as stratification. Participants will be placed into a treatment time, and then for each treatment time to be randomized, the algorithm will consider the average proportion (0%–33% vs 34%–67% vs 68–100%) on the three prognostic factors (sex assigned at birth, weight, % identifying as from an under-represented group) when evaluating group balance or imbalance. Participants will be informed of their assignment at the first intervention group meeting, which will only include individuals assigned to the same group.

2.3. Study intervention

2.3.1. Intervention framework – DEPTH

The obesity care lifestyle intervention adheres to the US Preventive Services Task Force [23] recommendations on comprehensive, intensive lifestyle interventions for adult obesity care. DEPTH includes 33 h of contact time over the course of the 12-month intervention.

2.3.2. Dietary goals

Participants will be instructed to follow a balanced diet based on the Dietary Guidelines for Americans, 2020–2025 [24]. Five dietary goals will be consistent across all three groups: energy, fat, eating window length, first eating occasion (EO), and number of EOs (see Table 1), minimizing the influence of these variables on outcomes.

Table 1.

Dietary goals in the three interventions.

Dietary Goals	DEPTH-Morning	DEPTH-Evening	DEPTH
Energy	1200–1500 kcal/d	1200–1500 kcal/d	1200–1500 kcal/d
Fat	≤ 30% energy	≤ 30% energy	≤ 30% energy
Eating window length	≤ 12 h, starting ≤60 min of awakening	≤ 12 h, starting ≤60 min of awakening	≤ 12 h, starting ≤60 min of awakening
First eating occasion	Eating within the first 60 min of awakening	Eating within the first 60 min of awakening	Eating within the first 60 min of awakening
Number of eating occasions	4: 3 meals+1 snack	4: 3 meals+1 snack	4: 3 meals+1 snack
Time-based energy intake goals	70% within the first 6 h of the eating window, 30% within the last six hours of the eating window	30% within the first 6 h of the eating window, 70% within the last six hours of the eating window	None

The energy goal will be based on study entry weight: participants weighing ≤200 lbs. will be prescribed 1200 kcal/day, while those weighing >200 lbs. will be prescribed 1500 kcal/day [25]. These goals will be adjusted as needed over the 12-month intervention to accommodate individual weight loss, with standard clinical safeguards used to monitor weight and caloric intake for healthy loss rates [25].

For the eating window, participants will be asked to consume their daily energy within 12 h, starting within 60 min of waking. EOs will be defined as consumption of ≥50 kcal [26]. According to previous research, goals that focus only on the size of EOs across the day to influence how energy is consumed in the day (e.g., EO1 = 15% energy; EO2 = 15% energy; EO3 = 20% energy; E04 = 50% energy) can be challenging for participants to consistently meet, as there are multiple daily goals with no flexibility [7,12]. Therefore, this investigation’s time-based energy intake goals are based on the 12-h eating window, allowing for more flexibility in EO size (EO size can vary across the days, as long as the time-based energy goals within the 12-h window are met) and only requires participants to achieve two time-based energy intake goals (70% and 30% goal) per day.

The energy distribution goals in this investigation were guided by three previous interventions with large effect sizes that used notably different time-based energy intake goals [7,8,12]. DEPTH-Morning will be instructed to consume 70% of their daily energy intake, in two meals and one snack, within the first six hours of the eating window (a morning-loaded energy distribution). The remaining 30% of their daily energy intake will be consumed in one meal during the later six hours of the eating window. In contrast, DEPTH-Evening will be instructed to follow the opposite pattern, resulting in an afternoon/evening energy loaded distribution. Participants assigned to the DEPTH group will not be assigned any time-based energy intake goals (see Table 2 for potential temporal eating patterns, which shows flexibility in EO size).

Table 2.

Potential temporal eating patterns of the three interventions.

	DEPTH Morning	DEPTH-Evening	DEPTH
	6:30 am awakening time	6:30 am awakening time	6:30 am awakening time
First half of the eating window (7 am – 7 pm) – morning loaded	7 am, 20–30% energy, first meal	7 am, 30% energy, first meal	7 am, 10–20% energy, first meal
	10 am, 10–20% energy, snack		12 pm, 20–30% energy, second meal
	12 pm, 20–30% energy, second meal
Second half of the eating window (7 am – 7 pm) – afternoon and evening loaded	6:30 pm, 30% energy, third meal	1 pm, 15–25% energy, second meal	3 pm, 10–20% energy, snack
		3:30 pm, 10–20% energy, snack	6:30 pm, 30–40% energy, third meal
		6:30 pm, 25–35% energy, third meal

Participants will receive sample meal plans that meet the dietary goals for their randomized condition to support the developed of personalized eating plans. After one week of recording intake and noting general eating patterns and wake-up times, participants will develop a personalized eating plan at the second group meeting to help them achieve their assigned dietary goals. Separate plans will be created for weekdays and weekend days, since wake-up times may differ. Interventionists will regularly review these individualized eating plans and adjust them as needed.

Participants will track their dietary intake using the Cronometer© (Revelstoke, BC, Canada) application on their PC and/or smartphone. Dietary goals will be entered into the CronometerPro© (Revelstoke, BC, Canada) software, and interventionists will provide feedback after each intervention session on progress toward meeting these goals.

2.3.3. Physical activity goals

The moderate- to vigorous-intensity physical activity (MVPA) goal is ≥200 min/wk. [27]. Participants will be encouraged to achieve ≥40 min/day of MVPA at least five days per week to establish consistent daily habits. Initially, they will be advised to complete ≥10 min/day of MVPA, 5 days per week, starting in the first week of the intervention. Each subsequent week, they will add 5 min per day until they reach the target goal. Participants will use a wrist-worn activity device (Fitbit; Mountain View, CA) to measure their daily activity, which syncs with the Cronometer© (Revelstoke, BC, Canada) application. Interventionists will provide feedback after each intervention session on progress toward meeting this goal.

2.3.4. Cognitive behavioral strategies

Participants will receive instruction in cognitive and behavioral strategies [28]. By the third month, all strategies will be introduced, with ongoing review and reinforcement to maintain behavior change. They will learn and practice self-monitoring, stimulus control, problem solving, social support, goal setting, cognitive restructuring, and relapse prevention strategies. Self-monitoring will be introduced in the first session and will remain a focus throughout the intervention, helping participants track their progress. Participants will self-monitor diet and physical activity, as described previously, and they will also receive a scale (Fitbit; Mountain View, CA) with wireless connectivity, which syncs with the Cronometer© (Revelstoke, BC, Canada) application, to measure their weekly weight at home.

2.3.5. Treatment structure

All participants will attend face-to-face group meetings weekly during months 1 to 6, twice monthly during months 7 to 9, and monthly during months 10 to 12. It is anticipated that groups will be 6–12 participants in size. In each session, participants will be weighed privately, and feedback will be provided on the relationship between their weight change and eating and activity behaviors. Each group meeting will begin with a “check-in” to review progress on goals and the application of cognitive behavioral strategies introduced in earlier sessions. The check- in will be followed by a cognitive behavioral lesson (see Table 3). These lessons are adapted from previous successful interventions, [29,30], and are identical across the three interventions, except for information related to the different dietary goals. Meetings will be ~60 min and led by an interventionist with at least a master’s degree in nutrition, exercise physiology, or behavioral psychology.

Table 3.

Intervention session topics.

Session	Topic
1	Introduction to DEPTH
2	Energy Balance
3	Ways to Reduce Energy Intake
4	Move those Muscles
5	Menu Planning
6	Working with What’s Around You
7	Goal-Setting
8	Problem-Solving
9	Eating in Social Situations
10	Talk Back to Negative Thoughts
11	It is not Just Will Power
12	The Slippery Slope of Lifestyle Change
13	How to Handle Eating Out
14	Supermarket Smarts
15	Healthy Cooking/ Recipe Modification
16	Emotions and Thoughts about Food
17	Handling Stress
18	Handling Holidays, Vacations, and Weekends
19	Lifestyle Activity
20	Sedentary Behavior
21	How to Interpret Hunger and How Food is Used in Ways Not Related to Hunger
22	Healthy Routines
23	Keeping Track for Self-Regulation
24	Relapse Prevention
25	Looking at your Progress and Developing Next Steps
26	Social Support for Being Active
27	Keeping Track More Accurately and Reduced Calorie Eating
28	Facing Frustration
29	Physical Activity: Resistance Training and Flexibility Training
30	Ways to Stay Motivated
31	Making Reduced Energy-Density Work for You
32	Physical Activity at Home
33	Becoming a Maintenance Pro

2.4. Outcome measures

Measures will be assessed by a trained researcher blinded to treatment assignment. For each assessment (0, 3, 6, and 12 months), participants will complete two measurement sessions, at least 7 days apart. In the first session, participants will be provided with an accelerometer-based wrist device to measure physical activity and sleep, as well as a continuous glucose monitor (CGM) to measure blood glucose levels. They will also receive training on using a smartphone-based ecological momentary assessment (EMA) platform that enables them to upload images of foods and beverages consumed and to complete appetite regulation measures. After the data collection period, participants will attend a second session to return the devices and complete additional questionnaires via Research Electronic Data Capture (REDCap), an electronic data capture tool hosted at University of Tennessee Health Science Center [31].

2.4.1. Anthropometrics

Participants’ weight will be measured, wearing no shoes and only light clothing and generally in the later afternoon/evening, similar to when the intervention sessions occur, using an electronic scale, while height will be measured with a stadiometer following standard procedures [32]. BMI will then be calculated, kg/m², based on these measures. Height will be collected at baseline only. Weight loss and percent weight loss ([baseline weight – assessment time point weight/baseline weight] × 100) will be calculated at 3, 6, and 12 months [33]. Waist circumference will be measured using standard procedures (above the iliac crest) with a retractable measuring tape [34,35].

2.4.2. Dietary intake

Dietary intake will be assessed by three random 24-h dietary phone recalls (two on weekdays and one on a weekend day) using the five-step, multiple-pass method [36]. This method questions participants about food intake five times in the following order: quick list, forgotten foods, time & occasion, detail cycle, and final probe [36]. To enhance the validity of these recalls, two additional strategies will be employed. First, participants will use a smartphone-based, study-designed EMA platform to capture digital images of their food at the beginning and end of each EO (showing any remaining food) during the week the recalls occur. Interviewers will review these images during the 24-h recalls. Studies have shown that using images alongside 24-h recalls can improve the accuracy of reported intake without introducing systematic bias [37]. Images also are time-stamped, providing an objective record of consumption times.

Second, the Dexcom G7 CGM system (San Diego, CA) will track post-prandial glucose responses throughout the day during the week that the recalls occur. This wearable sensor and transmitter records glucose levels every 5 min with no required calibration, and will be placed on the upper posterior portion of the participant’s non-dominant arm [38]. The CGM sensor probe, a platinum/silver wire about the width of two human hairs, is placed just beneath the skin (0.5 in./1.27 cm). The system includes a research app for real-time data collection and provides time-stamped CGM readings. Research staff will note any blood glucose spike of ≥40 mg/dL, signaling a potential EO to investigate.

Before the 24-h recall, research personnel will review the timing of the photos and the blood glucose spikes. If the self-reported intake time does not align with the photos or if no EO is reported during a blood glucose spike, participants will be asked for clarification. EMA photos will be compared with participants’ reports of what and how much was consumed. It is not assumed that the EMA photos or CGM data are always correct while self-reports are incorrect, as several factors can influence EMA and CGM information (e.g., not photographing a consumed food or physical activity). All details will be reconciled through query.

Dietary data will be analyzed with the Nutrition Data Systems Software for Research (NDSR; Nutrition Coordinating Center, University of Minnesota, Minneapolis, MN). Key variables of interest will include the three-day average of energy and fat intake and the temporal eating pattern. The temporal eating pattern will cover the midpoint of energy intake (the primary dependent variable for temporal analysis), the start time of the eating window, the length of that window, the number of EOs, the percentage of energy consumed in the first half of the window, the percentage of energy consumed in the last half, and the size of EOs [39].

2.4.3. Actigraphy: sleep and physical activity

The wGT3X-BT (3.5 cm × 3.5 cm × 1.0 cm) (Pensacola, FL) is an accelerometer-based device that will estimate daily sleep and time spent in MVPA at the sample rate of 30 Hz. This device features a 3-axis accelerometer, gyroscope, magnetometer, and secondary accelerometer. It will be worn on the nondominant wrist, a placement shown to be valid for both sleep and physical activity monitoring [40,41]. Participants will wear the wGT3X-BT (Pensacola, FL) continuously for a 7-day period, removing it only for swimming (bathing is permitted with the device). Five continuous days, with at least one weekend day, with 22 h of wear time will be required for inclusion in analysis [42]. Self-reported sleep onset and awakening times will be collected daily via an EMA prompt at 8 a.m., to assist with determining sleep from actigraphy. Sleep-dependent variables will include the mean for sleep onset and wake times, and total amount of sleep. Circadian alignment will be examined using the Sleep Regularity Index (SRI), a recently developed sleep metric [43], that measures the consistency of an individual’s sleep–wake patterns across days that will also be determined from actigraphy. The SRI evaluates the percentage probability of being in the same state (sleep or wake) at any two 30-s epochs (even when epoch length will be set at 5 s for other analyses) separated by 24 h, with values ranging from 0 (random) to 100 (perfectly regular). Notably, the SRI is not associated with sleep length [43], however, a higher SRI is associated with circadian alignment [44,45]. MVPA-dependent variables will include the mean for daily minutes and percent time spent engaging in MVPA.

Raw accelerometer data will be processed using the R-package GGIR [46], (https://github.com/wadpac/GGIR/wiki/Publication-list), software to process wrist-worn accelerometer data to daily estimates of sleep and activity behaviors. The GGIR pipeline will include: 1) autocalibration of the raw acceleration data on x, y, and z axes using the local gravity as the reference [47]; 2) aggregation of the acceleration signal over 5-s epochs after removing the gravitational acceleration (i.e., Euclidean Norm Minus One [ENMO]) [48]; 3) non-wear time detection using the default algorithm based on examining 15-min blocks with a 60-min sliding window; 4) classification of the sleep and awake periods using the combination of the EMA sleep data to identify potential time windows of sleep, and the van Hees 2015 algorithm to detect sustained inactivity bouts [49]; and 5) classification of the awake time into sedentary behavior, and physical activity of light, moderate and vigorous intensities using the Hildebrand 2014/2016 ENMO cutoff values of 44.8, 100.6, and 428.8 for adults following wrist-worn protocols [50,51].

2.5. Appetite regulation

The study-designed EMA platform will employ a 7-day protocol using time, semi-random, and event-based sampling strategies. Time-based sampling will take place at 8 am and 8 pm, capturing times of typically low and high hunger levels, respectively [52]. Semi-random sampling will occur during the afternoon between 12 pm and 6 pm. Both time-based and semi-random prompts will be signaled via smartphone tones/vibrations, prompting participants to rate hunger, fullness, temptation to eat, desire to eat, and control of overeating using 100 mm visual analogue scales, a validated approach to measuring appetite regulation [53,54]. During these prompts, participants will also be reminded to complete event-based sampling after each EO, using the same appetite regulation ratings. The minimum acceptable adherence to the prompts will be 60% over 7 days [55]. The mean ratings for 8 am, 8 pm, am period, pm period, and overall will be calculated.

2.5.1. Chronotype

Chronotype will be measured via the shortened version of the Munich ChronoType Questionnaire, μMCTQ [56]. Chronotype is determined by the midpoint of sleep on work-free days, with later times in the morning indicating a later chronotype preference [56].

2.5.2. Process data

Adherence with treatment will be assessed by tracking the number of days participants self-monitor diet and physical activity, as well as intervention session attendance [57,58]. For intervention delivery, interventionists will follow a structured session protocol guided by detailed treatment manuals. Intervention sessions will be audiotaped, with 33% of the recordings randomly selected to assess fidelity via a checklist of intervention components. Medication usage, particularly anti-obesity medications, will be queried at each assessment, and if differences in usage occur between intervention groups, this will be controlled in analyses.

2.5.3. Statistical and power analysis

A fixed and random effects approach to growth modeling will analyze both between- and within-subject sources of variation in longitudinal outcomes. All models will be estimated in MPlus version 8.8 (https://www.statmodel.com/index.shtml) using robust maximum likelihood estimation under missing data theory, which leverages all available data for the 174 participants [59]. Treatment differences by sex assigned at birth, race, and ethnicity will be evaluated by including these variables as covariates.

Given the anticipated non-linear relationship between time and percent weight loss, eating temporal patterns, sleep regularity, and appetite regulation (e.g. change from 0 to 3, 3 to 6 months, and 6 to 12 months, followed either stability or a slight return toward baseline values), piecewise growth modeling will be employed. Four random parameters will be modeled: the intercept (reflecting individual baseline differences) and three slopes, representing varying rates of change during 0–3, 3–6, and 6–12 months. Randomization should equalize intervention group characteristics; however, analyses will compare conditions on baseline characteristics. If differences occur, these variables will be covariates in later analyses if related to the dependent variable being examined in the model. Along with the previously described analyses, one additional model, in which the linear change from baseline to 12 months, will be examined for percent weight loss, eating temporal patterns, sleep regularity, and appetite regulation.

Similarly, exploratory analyses will investigate whether appetite regulation mediates the relationship between time-based energy intake and weight loss and whether chronotype moderates the effect of the intervention on weight loss. To examine mediation, multilevel (mixed) mediation models will be employed, utilizing the product-of-coefficients approach [60]. To assess moderation, the dependent variables (percent weight loss, eating temporal patterns, sleep regularity, and appetite regulation) will be regressed on the main effects for intervention groups, baseline values of the dependent variables, baseline moderator variable (chronotype), and the interaction between the intervention groups and the moderator. This strategy can be combined with the product of coefficients approach and model constraints in MPlus (https://www.statmodel.com/index.shtml) to explore moderated mediation and produce a plot of the indirect effect across the range of the moderator(s) [61,62]. Significant interactions will be followed up with simple slopes analyses for each treatment condition, and in the case of continuous variables, high (i.e., one SD above the mean or clinical cut-off) and low (i.e., one SD below the mean or clinical cut-off) moderator values will determine the nature of the moderated effect. All continuous variables will be centered prior to the tests for moderation. Variables that directly predict the outcome, but do not interact with treatment will be retained as covariates.

Sample size calculations presume 2-sided hypothesis testing at 12-month follow-up, with type one error rate (alpha) equal to 0.016 (three comparisons). There will be 58 participants randomized to each group, and all analyses will be conducted with the intent-to-treat principle. To reject with 80% power the null hypothesis of no treatment difference vs. the alternative that the treatment difference is f = 0.4 [equivalent d effect size is 0.8] or greater [7,8,12], 46 participants per group are required. This provides 80% power to detect a difference of 2.24% weight loss between groups (our preliminary work found a difference of 4.1% weight loss between groups), given a standard deviation of 8.0% at 12 months [29]. This sample size will allow for adequate statistical power with 20% attrition at 12 months.

3. Discussion

Recent interest has emerged regarding the temporal pattern of eating as a dietary intervention for obesity care. While early studies have focused on intermittent fasting or time restricted eating, less attention has been given to how energy consumption is distributed throughout the day [63]. Research on energy distribution timing within obesity care yields promising results [7–12]. However most investigations have examined only short-term outcomes, and few have reported on the temporal eating patterns achieved by participants or how these chrononutrition prescriptions influence circadian entrainment, a factor believed to be important for enhanced outcomes [64].

Time-based dietary goals are hypothesized to influence health by enhancing synchronization of biological and behavioral circadian rhythms, thereby promoting circadian entrainment [65]. Eating can cause circadian misalignment when it does not align with the suprachiasmatic nuclei (SCN) in the hypothalamus [66]. The SCN functions as the master clock for the mammalian circadian timing system, coordinating vital processes such as sleep-wake cycles, metabolic and hormonal rhythms, and the general temporal organization of physiological processes within the body [67–71]. Timing of eating can serve as a nonphotic zeitgeber (external cue for biological rhythms) for peripheral oscillators in cells and organs. When eating is misaligned with the SCN, the peripheral oscillators may become desynchronized, increasing the risk of negative physiological outcomes [66,72]. Conversely, aligning eating to SCN output can support circadian entrainment, reducing that risk [66,67,72]. It is hypothesized that a morning-loaded energy distribution supports entrainment [12], which may enhance appetite regulation.

The current study will provide valuable insights into the role of time-based energy intake goals on obesity care outcomes. The methodologies, including the use of CGM with dietary assessment, are state-of-the-art and comprehensive. This investigation will be the first in this area to provide comprehensive data on temporal eating patterns. Consecutive 24-h periods of actigraphy data will allow an examination of circadian entrainment, particularly through sleep regularity. These data, along with EMA measures of appetite, will help uncover how time-based energy intake goals relate to weight loss outcomes. Finally, the examination of how chronotype influences outcomes will identify which chronotype aligns best with a morning-loaded energy distribution. The concept of individualizing obesity care treatment (“precision nutrition”) will be tested in future trials.

Data availability

No data were used for the research described in the article.