Olfactory Contributions to Sleep-dependent Food Craving (SDFC)

This within-subject experiment uses one night of acute sleep restriction (4h) vs normal sleep (8h) to study state-dependent changes in olfactory processing. Odor-evoked blood oxygen level dependent (BOLD) responses will be measured in olfactory brain regions using functional magnetic resonance imaging (fMRI). Food intake will be measured at a buffet.

This randomized within-subject sleep-deprivation protocol is designed to examine the effects of reduced sleep on neural processing of food odors in olfactory brain areas. Subjects will be healthy, normal-weight subjects (N=30) with comparable regular sleep patterns. They will be pseudorandomly assigned to first participate in either the sleep deprived (SD) or the non-deprived (ND) session, and then undergo a 7-day sleep stabilization phase during which subjects will maintain a regular sleep schedule of 8 h (within pre-determined range of 10:30 pm to 7:30 am). After sleep stabilization, subjects will either sleep normally (ND, 8h between 11pm and 7am), or sleep for only for 4h (SD, between 1am and 5am), at home. During the sleep stabilization phase and the night of sleep deprivation, sleep duration, as well as sleep and wakeup time will be recorded using actigraphy (ActiGraph, LLC, Pensacola, Florida). All subjects will participate in both sessions (SD and ND) with the two sessions being separated by 4 weeks. In the evening of the next day, olfactory fMRI will be conducted after dinner. Isocaloric meals will be provided during the 24h before fMRI and no beverages other than water will be permitted. After fMRI, subjects will have ad libitum access to high-caloric snacks in the form of an all-you-can-eat buffet. During initial screening, subjects will rate the pleasantness of six food odors, including three high-caloric sweet odors (caramel, yellow cake, ginger bread) and three high-caloric savory odors (potato chips, pot roast, garlic butter). Based on each participant's ratings, two sweet and two savory odors that are matched in pleasantness will be selected. In addition, two non-food control odors (fir needle and celery seed) will be used. On the evening after the sleep manipulation, subjects will arrive at the imaging center at 5:15pm, and will consume an isocaloric dinner at 6pm before entering the scanner at 7pm. Neuroimaging will be performed on a Siemens PRISMA system with a 64 channel head/neck coil, using imaging sequences optimized for signal recovery in olfactory and orbitofrontal cortices. Subjects will participate in 4 runs of olfactory stimulation inside the scanner and BOLD responses will be acquired with high spatial (2 mm isotopic) and temporal resolution (2000 ms). On each trial, subjects will be visually cued to sniff, and an odor (food or non-food) or non-odorized air will be delivered. Subsequently, subjects will rate the pleasantness or the intensity of the odor (pseudo-randomized). Sniffing will be measured using an fMRI-compatible breathing belt and spirometer. A high-resolution anatomical image will be acquired for the purpose of spatial normalization and anatomical localization of the fMRI responses. Upon arrival at the imaging center, a study study nurse will insert an intravenous sterile heparin-lock catheter in the left forearm vein and initiate blood sampling. Blood will be drawn every 30 min after arrival, resulting in 5 samples: before dinner, after dinner, before fMRI, during fMRI, and after fMRI. At the completion of fMRI, participants will be provided with a buffet where they have ad libitum access to high-caloric food items (sweet [mini muffins, cinnamon buns, chocolate chip cookies, and doughnut holes] and savory high-calorie snacks [pizza bites, potato chips, hash browns, garlic bread]). Subjects will be instructed to consume as much food as they like. All food items will be weighed pre- and post-consumption, and the consumed calories will be calculated.

Participants will first be tested after 1 night with partial sleep deprivation (sleep-deprived, 4 h sleep). After a washout period of 4 weeks, they will be tested again after 1 night with normal sleep (non-sleep deprived, 8 h sleep).

Participants will first be tested after 1 night with normal sleep (non-sleep deprived, 8 h sleep). After a washout period of 4 weeks, they will be tested again after 1 night with partial sleep deprivation (sleep-deprived, 4 h sleep).

Functional Magnetic Resonance Imaging (fMRI)-Based Decoding Accuracy of Food vs Non-food Odors in Piriform Cortex

In the evening after the night of the "Sleep-Deprived" and the "Non-Sleep Deprived" intervention, odor-evoked brain responses were collected using fMRI. fMRI data was analyzed using searchlight-based support vector machines and leave-one-run-out cross-validation to decode information about food vs. non-food odors. For each participant, this resulted in two maps of decoding accuracy (% data points correctly classified as food or non-food odors), one map per intervention. The measure of interest was decoding accuracy extracted from the piriform cortex.

A single time point, in the evening (~19 h) after both the night of the "Sleep-Deprived" and the "Non-Deprived" intervention.

Energy-density (kcal/g) of the food items consumed at the buffet provided after fMRI scanning in the evening after the night of the "Sleep-Deprived" and the "Non-Sleep Deprived" intervention.

A single time point, in the evening (~20 h) after both the night of the "Sleep-Deprived" and the "Non-Deprived" intervention.

Ghrelin levels in blood samples (ng/mL) collected in the evening after the night of the "Sleep-Deprived" and the "Non-Sleep Deprived" intervention.

A single time point, in the evening (~19.5 h) after both the night of the "Sleep-Deprived" and the "Non-Deprived" intervention.

Leptin levels in blood samples (ng/mL) collected in the evening after the night of the "Sleep-Deprived" and the "Non-Sleep Deprived" intervention.

A single time point, in the evening (~19.5 h) after both the night of the "Sleep-Deprived" and the "Non-Deprived" intervention.

Total calories (kcal) of the food items consumed at the buffet provided after fMRI scanning in the evening after the night of the "Sleep-Deprived" and the "Non-Sleep Deprived" intervention.

A single time point, in the evening (~20 h) after both the night of the "Sleep-Deprived" and the "Non-Deprived" intervention.

Insulin levels in blood samples (uU/mL) collected in the evening after the night of the "Sleep-Deprived" and the "Non-Sleep Deprived" intervention.

A single time point, in the evening (~19.5 h) after both the night of the "Sleep-Deprived" and the "Non-Deprived" intervention.

Cortisol levels in blood samples (ug/dl) collected in the evening after the night of the "Sleep-Deprived" and the "Non-Sleep Deprived" intervention.

A single time point, in the evening (~19.5 h) after both the night of the "Sleep-Deprived" and the "Non-Deprived" intervention.

Inclusion Criteria: age between 18 and 40 years old regular sleep schedule (average sleep duration of ≥7 h/night, habitual mid-sleep time 3-5 am [deviation ≤2 h in daily mid-sleep time], time in bed ≤9 h) right-handed fluent English speakers Exclusion Criteria: history of significant cerebrovascular disease including (but not limited to) epilepsy, stroke, multiple sclerosis, meningitis body mass index (BMI) below 18.5 (underweight) or above 24.9 (overweight or obese) history of sleep disorder job with night shifts history of major head trauma with sustained loss of consciousness history of neurosurgery, ear-nose-throat (ENT) surgery, or cardiac surgery history of cardiac pacemaker or neurostimulator implantation history of significant medical illness including cardiovascular disease, diabetes, cancer, chronic obstructive pulmonary disease, severe asthma requiring hospitalization for treatment, renal insufficiency requiring dialysis, etc. history of major psychiatric disorder including major affective disorder, obsessive-compulsive disorder, schizophrenia claustrophobia history of significant food or non-food allergy, including latex, detergents, soaps presence of known smell, taste or ENT disorder history of sinusitis or allergic rhinitis history of alcoholism or consistent drug use current smoking current use of medications that affect alertness (e.g. barbiturates, benzodiazepines, cannabinoids, chloral hydrate, haloperidol, lithium, carbamazepine, phenytoin, etc.) current pregnancy (or possible pregnancy) history of metal working, or injury with shrapnel or metal slivers

Bhutani S, Howard JD, Reynolds R, Zee PC, Gottfried J, Kahnt T. Olfactory connectivity mediates sleep-dependent food choices in humans. Elife. 2019 Oct 8;8:e49053. doi: 10.7554/eLife.49053.