Subjects will follow an isocaloric diet, designed by a dietician. The eating window for the late-TRE group is between 10 AM - 8 PM

To quantify the intrahepatic fat content, a single voxel 1H-MRS (magnetic resonance spectroscopy) will be used. Relative fat content will be expressed as the ratio of the fat peak over the cumulative fat and water peak. This will also be corrected for T2 relaxation.

Fasted and stimulated C-peptide (nmol/L) will be measured during MMTT

Biopsies will be taken from skeletal muscle and subcutaneous fat to measure key proteins in the insulin signalling pathway (Western blots)

In our study, we will deploy Continuous Glucose Monitors (CGMs) to acquire an in-depth understanding of glucose variability in participants throughout the course of the intervention. These monitors gauge glucose concentrations in the interstitial fluid (mmol/L), serving as a reliable proxy for blood glucose levels. This approach will enable us to assess key metrics such as mean, minimum and maximum glucose levels.

Fasted and stimulated de novo lipogenesis (DNL) will be measured during the MMT as 2H incorporation into fatty acids following deuterated water (2H2O) administration

We will conduct immune cell phenotyping on whole blood samples to identify and categorize various immune cell types. Additionally, we will assess immune cell function and metabolism in isolated peripheral blood mononuclear cells (PBMCs). Inflammatory markers will also be assessed in serum samples to provide a comprehensive overview of immune and inflammatory status.

Physical activity will be assessed via accelerometry. Accelerometry represents the magnitude of acceleration in any direction, over a predefined epoch.

Brain activation maps and functional connectivity will be assessed by blood oxygen dependent signals in the resting state and after visual food cues using functional magnetic resonance imaging (fMRI).

We will use the General Food Cravings Questionnaire (G-FCQ) to evaluate the frequency and intensity of food cravings among participants. The scoring for the G-FCQ ranges between 21 and 105, with higher scores indicative of more pronounced food craving tendencies.

We will utilize the Dutch Eating Behavior Questionnaire (NVE) to assess and categorize the eating behaviors and tendencies of our participants. The NVE discerns three distinct eating styles: emotional eating, external eating, and restrained eating. A higher score within a specific style suggests a predominant inclination towards that particular eating behavior.

We will employ a nine-question visual analogue scale (VAS) to assess hunger. Each question will be scored 0-100.

We will use the Yale Food Addiction Scale 2.0 (YFAS 2.0) to assess addictive behaviours. The YFAS 2.0 is designed in accordance with the DSM-5 criteria, and its scoring system mirrors the number of DSM-5 criteria fulfilled indicative of addiction.

We will use the Barratt Impulsiveness Scale (BIS) to scale impulsiveness towards food. This will help us understand how impulsivity might influence dietary behaviour. The score ranges from 30-120 and a higher score leans towards greater impulsivity.

To determine natural sleep-wake patterns, we will use the Munich Chronotype Questionnaire (MCTQ). This tool offers a comprehensive understanding of sleep behaviors, revealing a chronotype.

We will employ a semi-structured oral interview as part of our qualitative approach to understand our participants' experience and adherence to the intervention.

The delay discounting task aids in understanding the decision-making processes that might contribute to overeating and poor food choices. In this task, we will be able to measure the ability of each individual to delay immediate gratification for a greater future reward.

We use the Iowa gambling task to assess decision-making and risk-reward sensitivity. Participants choose cards from four decks, each with different reward and punishment rates, aiming to maximize their winnings. We use this task to study risk taking and impulsive behaviour.