Effects of daily low oxygen exposure on weight status, body composition, and metabolic health in adults with obesity: protocol for a randomized, double-blind, controlled-feeding study

Abstract

Background:

Diet and exercise are the cornerstone of obesity prevention and treatment. However, a substantial number of individuals are non-responsive to existing weight-loss interventions and obesity rates continue to rise. Daily intermittent exposure to low-oxygen conditions may aid in current weight-loss strategies by increasing resting metabolic rate and decreasing appetite. Whether in-home, overnight, normobaric hypoxic exposure promotes body weight loss in adults with obesity remains unknown.

Methods:

Fifty adults with obesity (BMI: 30–39.9 kg/m²) will complete this double-blind, parallel-arm, randomized, controlled-feeding clinical trial. Participants will be provided with a weight maintenance diet for 2 weeks while undergoing baseline measurements. Following the weight maintenance phase, an energy restricted diet (500 kcal/day below weight maintenance needs) will be provided in combination with either overnight exposure to normobaric hypoxia (8 h/night, 15% oxygen, elevation ~2640 m) or normoxia (8 h/night, 21% oxygen, elevation ~60 m), using a commercially available, in-home tent system, for 8 weeks. The primary outcome is the difference in body weight change between interventions. Secondary outcomes include measures of body composition, total and resting energy expenditure, energy intake from an ad libitum meal, insulin sensitivity and glycemic control, sympathetic tone, iron absorption and indicators of iron status, the gut microbiome, appetite, psychosocial factors, and sleep quantity and quality.

Discussion:

Chronic, overnight, low oxygen exposure may provide a novel intervention to supplement current weight-loss strategies, inform new strategies to accelerate weight loss, and aid long-term weight management efforts in adults with obesity.

1. Introduction

Obesity and associated metabolic conditions are a significant public health burden, costing the U.S. ~$150 billion annually [1]. Obesity is both a disease, affecting 1 in 3 Americans, and a risk factor for other chronic diseases, such as cardiovascular disease, type 2 diabetes, and 13 forms of cancers. Diet and exercise are the cornerstone of obesity prevention and treatment. However, a considerable number of individuals are non-responsive to existing lifestyle weight-loss interventions and obesity rates continue to rise.

Daily exposure to low-oxygen conditions may aid weight loss. Moderate (1500–3500 m) and higher (≥ 3500 m) altitude environments are naturally hypoxic due to the lower atmospheric pressure [for reference, Baton Rouge, LA is ~60 m (partial pressure of inspired O₂, P_iO₂ = 159 mm Hg), Denver, CO is ~1600 m (P_iO₂ = 129 mm Hg), and Pikes Peak, CO is ~4300 m (P_iO₂ = 90 mm Hg)]. A cross-sectional study conducted in U.S. adults showed an inverse relationship between elevation of residence and body mass index, even after adjusting for lifestyle factors such as dietary intake and physical activity [2]. Human interventional studies consistently show that sea-level natives exposed to moderate- or high-altitude continuously for 5 or more days lose weight [3]. These studies have mostly been conducted in populations where weight loss may be detrimental (e.g., mountaineers and military personnel).

Only four studies have tested the effects of living for an extended period at high altitude (hypobaric hypoxia) on weight status among individuals with obesity [4–7]. In one study, adult males with obesity had a 1.5 kg reduction in body weight after 7 d at moderate altitude (2650 m), despite no changes in physical activity. Weight loss was attributable, in part, to a 24% reduction in ad libitum energy intake and a 15% increase in basal metabolic rate but was greater than expected based on these measures alone [4]. The other 3 studies, conducted in adults with obesity traveling to altitudes of 1700–1900 m for 2–3 weeks, found modest or no changes in body weight compared to an active control group [5–7]. These findings may suggest the altitude stimulus was insufficient to induce clinically relevant or statistically significant findings, as a dose-response relationship has been demonstrated between elevations ≥ 2200 m and weight loss in a meta-analysis of intervention studies conducted mostly in normal weight adults [3]. Specifically, weight loss at moderate (≥ 2200 to 3500 m), high (3500–5300 m), and extreme (> 5300 m) altitude was 1.7 kg (95% CI 1.3, 2.0), 2.3 kg (1.6, 3.0), and 4.9 kg (3.7, 6.2), respectively, after prolonged exposure (days to months) [3].

Moderate- or high-altitude exposure as a means for weight loss in individuals with overweight or obesity has been limited by the logistical constraints of an extended, continuous stay at altitude or within a hypobaric altitude chamber. A potential alternative may be intermittent exposure to normobaric hypoxia, which requires decreasing the percentage of oxygen in the air but involves no changes in atmospheric pressure. A recent study in adolescents with obesity (n = 35) demonstrated a 1.6 kg greater weight reduction in those who slept 10 h/night in a low oxygen environmental chamber (14.7% oxygen, −8.7 ± 2.2 kg) compared to those who slept in normal conditions (~21% oxygen, −7.1 ± 2.2 kg) for 4 wk as part of a weight-loss camp that included 4 h/d of structured exercise [8]. This study demonstrates the efficacy of nightly exposure to normobaric hypoxia for the augmentation of weight loss in adolescents [8]. However, the intervention still required participants to travel a far distance and stay away from their homes for 4 wk, which limits the feasibility and applicability of the intervention to all individuals. A safe, less expensive, and more logistically feasible alternative for intermittent exposure to hypoxia at sea-level is the use of individual low-oxygen systems, which are commercially available (https://hypoxico.com/), portable, and can be set up at home (Figure 1). These systems were developed for and are commonly used by athletes to gain the physiological benefits of training or sleeping in hypoxic conditions.

Figure 1.

Normobaric hypoxic in-home tent system.

Two small studies have used these hypoxic systems to evaluate the effects of normobaric hypoxia on glucose control and insulin sensitivity in adults [9, 10]. In one study, 8 males who were overweight slept 10 consecutive nights (10 h/night) in individual hypoxic tents (15% oxygen) set up in a clinical lab [9]. Following the 10 days of overnight hypoxic exposure, participants had significant weight loss (1.2 kg), decreased fasting glucose concentrations, and increased insulin sensitivity as determined by hyperinsulinemic-euglycemic clamp. The other study included 8 males with type 2 diabetes who slept 14 consecutive nights (7–12 h/night) in individual hypoxic tents (15% oxygen) set up in their homes [10]. Post-intervention, participants had decreased area under the curve (AUC) for plasma glucose concentrations in response to a 2-h oral glucose tolerance test, no changes in insulin AUC, and a trend towards increased insulin sensitivity; however, changes in body weight were not reported. Whether in-home, overnight, normobaric hypoxic exposure, compared to normobaric normoxic exposure, promotes weight loss in adults with obesity remains unknown.

2. Trial objectives

The objective of this randomized, double-blind, parallel-arm, controlled-feeding trial is to evaluate changes in body weight and composition, assess determinants of energy balance (intake and expenditure), and measure modulators of energy intake and expenditure, following 8 wk of calorie restriction (500 kcal/d below weight maintenance needs) in combination with either overnight exposure to normobaric hypoxia (8 h/night, 15% oxygen, ~2640 m) or normoxia (8 h/night, 21% oxygen), using a commercially available, in-home tent system, in adults with obesity.

2.1. Primary study objectives

1. Determine body weight and composition responses to passive, intermittent exposure to normobaric hypoxia, compared to normobaric normoxia, for 56 days (8 h/night).

2. Evaluate whether the rate of weight loss differs acutely (1–2 wk) compared to chronically (7–8 wk) between normobaric hypoxic and normobaric normoxic exposure groups.

3. Investigate weight loss maintenance for 4 weeks post-intervention.

2.2. Secondary study objectives

1. Assess determinants of energy balance (energy intake and energy expenditure) in response to passive, intermittent exposure to normobaric hypoxia, compared to nomobaric normoxia.

2. Measure modulators of energy intake and energy expenditure in response to passive, intermittent exposure to normobaric hypoxia, compared to normobaric normoxia, for 56 days. Modulators of energy intake include subjective appetite ratings, circulating appetite hormones, psychosocial factors, acute mountain sickness (AMS) symptoms, and sleep quality. Modulators of energy expenditure include oxygen saturation, heart rate variability, urinary catecholamines, gut microbiome composition and activity, hydrogen and methane gas concentrations in breath, and circulating short-chain fatty acid concentrations.

3. Assess markers of health status in response to passive, intermittent exposure to normobaric hypoxia, compared to normobaric normoxia, for 56 d including continuous glucose concentrations, substrate oxidation, insulin sensitivity, and iron status and absorption.

2.3. Tertiary study objectives

1. Assess sex-dependent responses to low environmental oxygen during sleep.

2. Assess individual variation in body weight based on oxygen saturation levels.

3. Assess individual variation in body weight responses to low environmental oxygen during sleep using Accumulating Data to Optimally Predict Obesity Treatment (ADOPT) core measures.

3. Methods

3.1. Institutional review board approval and trial registration

The study protocol was approved by the Institutional Review Board at Pennington Biomedical Research Center (PBRC, Baton Rouge, LA; 2022-046-PBRC LOWS). The ClinicalTrials.gov identifier is NCT05289310.

3.2. Trial design

Adults with obesity will be recruited for a 3-phase randomized, double-blind, parallel-arm, controlled-feeding trial (Figure 2). The first phase (pre-intervention phase; days −14 to −1) is a 2-week controlled-feeding period designed to maintain each participant’s current body weight. Phase 2 (intervention phase; days 0 to 56) will be an 8-week controlled-feeding period designed to induce a 500 kcal/d energy deficit to produce a 1–2 lb/wk weight reduction. On the first day (day 0) of phase 2 (intervention), participants will be randomized to spend 8 h/night in a tent, set up around their bed in their home for 8 wk while exposed to either normobaric hypoxia at ~15% oxygen (achieved with nitrogen dilution, equivalent to ~2590 m elevation) or normobaric normoxia at ~21% oxygen (equivalent to sea level). After completing phase 2, the tent will be removed from the participant’s home and participants will be asked to continue weighing themselves for 4 weeks (phase 3, post-intervention phase; days 57 to 84).

Figure 2.

Experimental design.

3.3. Randomization and blinding

On the first day of the intervention period (phase 2), participants will be randomized to one of two groups, normobaric hypoxia (experimental) or normobaric normoxia (control), using randomized permutated blocks with 6 subjects in each block. The randomization plan will be stratified by sex. Both participants and trial personnel will be blinded to the intervention allocation, outcome assessment, and statistical analysis.

3.4. Participants

Fifty adults will complete the study. Participants must be 22 – 65 years, have a BMI between 30–39.9 kg/m², be born at an altitude less than 2,100 meters, and be residing in Baton Rouge, Louisiana, or surrounding areas. All inclusion and exclusion criteria are listed in Table 1. For premenopausal females, visits will be scheduled at the start of the follicular phase of their menstrual cycle to control for hormonal fluctuations.

Table 1.

Study inclusion and exclusion criteria.

Inclusion criteria

Exclusion criteria

Men and women aged 22 – 65 years Obese (BMI 30–39.9 kg/m2) Born at altitudes less than 2,100 meters (~7,000 feet) Willing to consume only the meals provided during the weight maintenance and calorically restricted phases of the study Willing to have blood stored for future use Willing to refrain from smoking, vaping, chewing tobacco, and dietary supplement use throughout the entire study Willing to have a hypoxic tent set up in their home and spend 8 consecutive hours per night in the tent for 8 weeks Lives with a family member, partner/spouse, or roommate who can hear and respond in the unlikely event that the blood oxygen, air oxygen, or air carbon dioxide monitoring alarm would sound.

Living in areas that are more than 1,200 m (~4,000 feet), or have traveled to areas that are more than 1,200 m for five days or more within the last 2 months Prior severe cardiovascular events such as stroke or myocardial infarction Congestive heart failure Any disease/condition that interferes with oxygen delivery/transport (e.g., kidney disease) Clinically diagnosed diabetes or a fasting blood glucose >126 mg/dL Current insulin treatment Treatment with sulfonylureas or glitinides Chronic obstructive pulmonary disease (COPD) Taking any medication(s) that interfere with oxygen delivery/transport for 4 weeks prior to and throughout the entire study (e.g., includes sedatives, sleeping aids, tranquilizers and/or any medication that depresses ventilation, diuretics, alpha and beta blockers). Initiation of antidepressant medication within the last 3 months Change in medications 1 month before the study Previous bariatric surgery (or other surgeries) for obesity or weight loss Medically diagnosed sleep apnea or other sleep disorders deemed significant by the Medical Investigator Oxyhemoglobin saturation levels <95% while seated and awake or <78% overnight in normoxic conditions (assessed during screening) Adjusted oxygen desaturation index (ODI) >5/hour overnight in normoxic conditions (assessed during screening) Evidence of prior high-altitude pulmonary edema (HAPE), high-altitude cerebral edema (HACE) diagnosis, or severe high-altitude sickness Diagnosis or family history of sickle cell anemia/trait Hemoglobin <13 g/dL for males, <12 g/dL for females Blood donation within 8 weeks of beginning the study Present condition of anabolic steroid (e.g., testosterone) or other substance abuse issues Consumption of >21 drinks/week for females or >28 drinks/week for males, or unwilling to consume ≤ 2 drinks/week throughout the study Unstable weight in the last 3 months (gain or loss ≥5% of body weight) Women who are pregnant, lactating, or are planning to become pregnant Shift workers (e.g., individuals awake or working between midnight and 5AM) Adults unable to consent Prisoners

3.5. Eligibility determination

Interested individuals will undergo a multi-stage screening process. Potential participants will undergo web and/or telephone pre-screening to assess initial eligibility based on inclusion/exclusion criteria (Table 1). If eligible, individuals will undergo 2 screening visits. During screening visit 1, the study consent form will be reviewed and signed. Upon consent, study staff will determine eligibility based on participant questionnaire responses (e.g., medical history and exercise and diet habits) and measured height, weight, blood pressure, heart rate, markers of health status in blood, and oxygen saturation. If eligible to continue, participants will be scheduled for screening visit 2 with the study physician to discuss their results, review their medical history, and undergo a physical exam.

3.6. Controlled dietary intervention

Participants in both groups will be given and asked to consume all foods and beverages during phases 1 and 2 (Table S1). During phase 1 (weight maintenance), the Mifflin-St. Jeor equation [11] will be used to estimate initial energy needs for meal provisions with an activity factor determined from self-reported physical activity levels. Energy intake will be adjusted with 100 kcal unit foods as needed to maintain body weight within ±2%. For phase 2, individualized energy intake will be 500 kcal/d below the final energy level determined to maintain weight during phase 1. The macronutrient composition of the phase 1 and phase 2 diets will be 55% carbohydrate, 30% fat, and 15% protein. The proportion of total energy provided at each meal will be as follows: 24% breakfast, 10% morning snack, 30% lunch, 30% dinner, and 6% evening snack. A Registered Dietitian designed the study diets and will calculate individualized energy needs and oversee preparation and administration of the diets.

3.7. Normobaric hypoxic or normoxic intervention

At the start of the intervention (day 0), both groups will receive the energy-restricted diet and spend their first night in the tent, set up around their bed in their home. Participants will sleep 8 h/night in the tent, either in normobaric hypoxia set to maintain 21% oxygen or normobaric hypoxia set to maintain 15% oxygen achieved through nitrogen dilution. For the first three nights, the normobaric hypoxia group will be gradually exposed to overnight hypoxic conditions. The first night, hypoxic conditions will be set at 19% oxygen (~762 m), the second night at 17% oxygen (~1676 m), and the third night at 16% oxygen (~2134 m). For the remainder of the study, oxygen concentrations will be set at 15% oxygen (~2591 m). Unblinded research staff will visit the participant’s home on the first day of the intervention to set up the tent system and will return the next 3 days to adjust tent settings. Thereafter, the unblinded staff will visit the participant’s house weekly to check the tent settings. These visits will be scheduled for participants in both groups to maintain blinding.

3.8. Compliance assessment

Participants will visit the metabolic kitchen and diet study center 2–3 times per week during phases 1 and 2 for study food pick-up, to return coolers and empty containers, and to check in with study staff regarding diet compliance. During phase 2, an unblinded research staff member will perform weekly visits to the participant’s home during the day to check and/or adjust the tent settings; collect data from the body weight scale, oxygen saturation and heart rate monitor, accelerometer, air gas (O₂ and CO₂) monitor, and room temperature and humidity monitor; and collect questionnaires that were completed throughout the week. The unblinded research staff member will use the pulse oximeter and air gas monitor data to ensure compliance with the study protocol.

3.9. Outcome measures

3.9.1. Anthropometrics

Body weight will be measured every day during phases 1–3 (84 days total) following an overnight fast and after first toilet use, using a calibrated digital scale provided to participants to use at home (wireless weight scale UC-352BLE, A&D Medical, San Jose, CA). Fasting metabolic weights will be collected on days −14, −7, 0, 14, 42, 49, and 56 during study visits whereby weight is measured while wearing only a gown and the weight of the gown is subtracted from the total weight. Height will be measured in duplicate to the nearest 0.1 cm using a stadiometer at screening.

3.9.2. Oxygen saturation

Participants will wear a pulse oximeter (PalmSAT^® 2500A, Nonin Medical, Inc., Plymouth, MN) on their index finger every night to measure oxygen saturation (Table 2).

Table 2.

Schedule of study procedures.

	Study Visits
	Screening		Phase 1		Phase 2						Phase 3
	SV1	SV2	Visit 1 (Day - 14)	Visit 2 (Day - 7)	Entire study (Days −14 to 56)	Visit 3 (Day 0)	Visit 4 (Day 14)	Visit 5 (Day 42)	Visit 6 (Day 49)	Visit 7 (Day 56)	Days 57–84
Informed Consent	X
Height	X
Physical Exam		X
Weight	X		X	X	X (at-home)	X	X	X	X	X	X
Waist circumference				X					X
Resting blood pressure and heart rate	X			X					X
Medication reporting	X	X	X	X		X	X	X	X	X
Blood draws	X					X				X
Accelerometry					X
Pulse oximetry	X				X (day −7 to 56)
Continuous glucose monitoring			X	X		X	X	X	X	X
Questionnaires					X
24hr dietary recall*			X
4-compartment body composition				X					X
3D/2D body imaging				X					X
Doubly labeled water administration				X					X
Iron isotope administration				X				X
Blood volume measurement				X					X
Urine collection				X		X	X		X	X
Stool collection			X	X					X
Indirect calorimetry						X				X
Breath testing						X				X
Electrocardiogram						X				X
Oral glucose tolerance test						X				X
Ad libitum meal						X				X
Eating controlled diet					X
Sleep in tent					X(day 0 to 56)

^*Completed 1 to 7 days prior to visit 1. Abbreviations: SV, screening visit.

3.9.3. Four-compartment model of body composition

Body composition (total body weight, fat-free mass, fat weight, and bone mass) will be determined following a 10-hour fast using dual-energy X-ray absorptiometry (DXA; Lunar iDXA; General Electric, Milwaukee, WI) on days −7 and 49. Total body water will be measured following a 10 h fast using oxygen-18 (¹⁸O) labeled water (10% at 1.5 g/kg bodyweight) in conjunction with the doubly-labeled water (DLW) method on days −7 and 49. Upon arrival to the lab, participants will provide a urine sample and completely void their bladder, this sample will be used to measure background ¹⁸O enrichment. The participant will then consume ²H₂¹⁸O (99.8% ²H₂O at 0.06 g/kg body weight; Cambridge Isotope Laboratories, Inc.). Urine will be collected again 4 hours after dosing and ¹⁸O will be measured.

3.9.4. 2D/3D body imaging

Body circumferences and composition will be estimated using an optical imaging system (Mobile Fit Booth, Size Stream LLC, Cary, NC) on days −7 and 49. The subject will stand on a stationary platform for 45 seconds while 2–3 images are taken in each of 2 positions (facing the camera and a side view) using a tablet camera. The proprietary software will render a 3D avatar of the participant and derive results.

3.9.5. Hemoglobin mass and blood volume

Hemoglobin mass (Hbmass) and blood volume will be determined on days −7 and 49 using the automated carbon monoxide (CO)-rebreathing method (Detalo Health; Hørsholm, Denmark) [12]. CO will be administered to participants in combination with 100% O₂. The gas mixture will then be rebreathed for 4 min. Before and 6 min after CO-rebreathing, arterialized capillary blood will be collected from a fingertip to assess changes in CO-Hb (%) using a blood gas analyzer (OSM3, Radiometer; Brea, California). Red blood cell volume, plasma volume, and total blood volume will be calculated from the increase in CO.

3.9.6. Total daily energy expenditure (TDEE)

TDEE will be assessed by the DLW method [13] during phase 1 (days −7 to 0) and phase 2 (days 49 to 56). Volunteers will provide a urine sample and completely void their bladder, this sample will be used to measure background isotope enrichment. The participant will then consume 1.5 g of 10% ¹⁸O/kg body weight and 0.06 g ²H₂O/kg body weight (Cambridge Isotope Laboratories, Inc. or equivalent). Urine will be collected at 4 h, 6.5 h, 24 h, 3 d, 5 d, and 7 d after dosing. Isotopic elimination rates will be calculated using regression analysis of urinary samples and the equation developed by Speakman et al. [14] for rate of CO₂ production. Energy expenditure will be calculated by multiplying the rate of CO₂ production by an assumed food quotient of the diet 0.882.

3.9.7. Resting metabolic rate

Resting metabolic rate will be measured using indirect calorimetry (Omnical indirect calorimeter, Maastricht Instruments, Maastricht, Netherlands) on days 0 and 56 following a 10 h fast. Participants will rest in the supine position for 30 min in a quiet, dim, and temperature regulated room before measurements begin. The test will be discontinued when 20-minutes of steady state oxygen consumption ($\dot{\text{V}}{\text{O}}_2$) and carbon dioxide ($\dot{\text{V}}{\text{CO}}_2$) production are recorded.

3.9.8. Physical activity

Participants will be asked to maintain their normal daily level of physical activity throughout the study. Participants will be provided a wrist-worn device with Bluetooth capabilities (vivosmart 4, Garmin, Olathe, KS) to be worn continuously during phases 1 and 2. This device will measure steps, physical activity level, heart rate, and sleep patterns. Habitual physical activity (steps, intensity) and sleep/wake (sleep quality and duration) data will also be measured via accelerometry (wGT3X-BT, ActiGraph, Pensacola, FL) worn on the wrist during all of phase 1 and days 0–14 and days 42–56 of phase 2.

3.9.9. Oral glucose tolerance test (OGTT)

Oral glucose tolerance testing will occur at visits 3 (day 0) and 7 (day 56) (Table 2). A 75-g glucose solution will be consumed and blood samples collected by venous catheter prior to (fasting) and at 30, 60, 90, and 120 min for measurements of plasma glucose, insulin, and appetite hormones (Table 3). Insulin sensitivity will be determined by using the Matsuda Index [15]. Insulin secretion will be estimated by the Insulinogenic Index [16]. Beta-cell function will also be estimated by the Disposition Index [17]. Additional OGTT measures of 2-h glucose AUC and 2-h insulin AUC will be calculated.

Table 3.

Blood parameters.

Day:	SV	Day 0					Day 56
Minute:		0	30	60	90	120	0	30	60	90	120
Sodium	X	X					X
Potassium	X	X					X
Chloride	X	X					X
Calcium	X	X					X
Total CO2	X	X					X
Urea nitrogen	X	X					X
Creatinine	X	X					X
Total Bilirubin	X	X					X
Total Protein	X	X					X
AST	X	X					X
ALT	X	X					X
Albumin	X	X					X
Alkaline Phosphatase	X	X					X
Complete blood count	X	X					X
C-reactive protein	X	X					X
Glucose	X	X	X	X	X	X	X	X	X	X	X
HDL-cholesterol	X	X	X	X	X	X	X	X	X	X	X
Triglycerides	X	X	X	X	X	X	X	X	X	X	X
Total cholesterol	X	X	X	X	X	X	X	X	X	X	X
LDL-cholesterol*	X	X	X	X	X	X	X	X	X	X	X
Non-HDL-cholesterol*	X	X	X	X	X	X	X	X	X	X	X
Total/HDL-cholesterol*	X	X	X	X	X	X	X	X	X	X	X
VLDL-cholesterol*	X	X	X	X	X	X	X	X	X	X	X
Free fatty acids, glycerol		X	X	X	X	X	X	X	X	X	X
Insulin		X	X	X	X	X	X	X	X	X	X
Lactate		X	X	X	X	X	X	X	X	X	X
Leptin		X					X
Acylated ghrelin		X	X	X	X	X	X	X	X	X	X
Cholecystokinin		X	X	X	X	X	X	X	X	X	X
Glucagon-like peptide-1		X	X	X	X	X	X	X	X	X	X
Peptide YY		X	X	X	X	X	X	X	X	X	X
Erythropoietin		X					X
Erythroferrone		X					X
Hepcidin		X					X
Interleukin-6		X					X
sTfR		X					X
Ferritin		X					X
Serum iron		X					X
TIBC		X					X
57Fe and 58Fe							X
Short-chain fatty acids		X					X
Serum archive		X	X	X	X	X	X	X	X	X	X
Plasma archive		X	X	X	X	X	X	X	X	X	X
Whole blood archive		X					X

Each X denotes a timepoint at which the blood parameter is measured. Zero minutes is in the fasted state. Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; CO₂, carbon dioxide; HDL, high density lipoprotein; LDL, low density lipoprotein; sTfR, soluble transferrin receptor; SV, screening visit; TIBC, total iron binding capacity; VLDL, very low-density lipoprotein.

^*Calculated

3.9.10. Subjective appetite testing

Immediately prior to (fasting) and 30, 60, 90, 120 min after consumption of the glucose drink for the OGTT, subjectively rated appetite (fullness, hunger, desire to eat, and prospective consumption) will be measured on 100 mm visual analog scales. A composite satiety score (CSS) will be calculated from the individual appetite scores [18].

3.9.11. Thermic effect of glucose and substrate oxidation

The thermic effect of glucose will be measured during the OGTT with indirect calorimetry (Omnical indirect calorimeter, Maastricht Instruments, Maastricht, Netherlands) on days 0 and 56. The thermic effect of glucose will be measured in 15 min increments for 2 hours (10–25, 40–55, 70–85, and 100–115 min). Additionally, indirect calorimetry will be used to estimate whole-body substrate oxidation [19]:

$$\text{Glucose}\text{oxidation}(\text{g}/\text{min})=4.585\dot{\text{V}}{\text{CO}}_2(\text{L}/\text{min})-3.226\dot{\text{V}}{\text{O}}_2(\text{L}/\text{min})$$

$$\text{Fat}\text{oxidation}(\text{g}/\text{min})=1.695\dot{\text{V}}{\text{O}}_2(\text{L}/\text{min})-1.701\dot{\text{V}}{\text{CO}}_2(\text{L}/\text{min})$$

3.9.12. Hydrogen and methane breath testing

Hydrogen and methane breath testing will be performed to determine small intestinal bacterial fermentation on days 0 and 56 during the OGTT. Breath samples will be collected after participants exhale through a mouthpiece connected to a dual-bag system by a three-way valve. Breath samples will be analyzed for hydrogen and methane by gas chromatography (MicroLyzer Model SC, Quintron Instrument Co., Inc., Milwaukee, WI). One breath sample will be collected prior to OGTT (fasting), between 5–10 mins after glucose administration, and every 30 minutes up to 120 mins (6 samples). A single breath sample will be taken following the ad libitum meal.

3.9.13. Ad libitum energy intake

Low oxygen exposure may reduce appetite and, subsequently, energy intake [4, 20, 21]. However, since this is a controlled-feeding study with daily energy intake held constant, effects of low oxygen on energy intake will need to be assessed by an ad libitum meal. Participants will be provided an ad libitum mixed-macronutrient meal (frozen lasagna) that will occur immediately following the OGTT on days 0 and 56. Food will be provided in excess of expected consumption for one individual. Staff will provide participants with their own empty plate, instructing them to take as much as they would like, and to eat until they are comfortably full. Water in the amount of 240 g will be provided during the meal. Participants will be instructed to consume all water before completing the meal and will not be permitted additional water during the meal to control the effect of fluid intake on satiety. Upon cessation of the ad libitum meal, subjectively rated appetite will be measured on a 100 mm visual analog scale and CSS calculated [18]. Leftover food will be collected and weighed. Energy and nutrient content of the amount consumed will be determined.

3.9.14. Continuous glucose monitoring

A continuous glucose monitor (CGM; Dexcom G6, Dexcom Inc., San Diego CA) will be inserted on the abdomen and used to assess interstitial glucose concentrations. The CGM will be worn for the entirety of phase 1 and days 0–14 and days 42–56 of phase 2.

3.9.15. Electrocardiogram

A 12-lead electrocardiogram will be used to assess heart rate variability to indicate sympathetic tone, a modulator of energy expenditure. Participants will have electrode patches placed on their chest to measure the electrical activity of the heart.

3.9.16. Urinary catecholamines

Low oxygen exposure induces sympathetic activation, which may increase energy expenditure through greater catecholamine release and clearance [22]. Participants will collect their urine overnight on days −1 to 0, 13 to 14, and 55 to 56 (Table 2). They will void their bladder before bed and record the time. Any output overnight will be collected in a pre-labeled container. Participants will collect one final urine upon waking in the morning. Total urine will be weighed and catecholamine concentrations will be measured.

3.9.17. Iron status and absorption

Individuals with obesity often exhibit reduced iron status compared to those with normal weight, potentially due to obesity-related chronic inflammation [23, 24]. However, the impact of reducing chronic inflammation through diet-induced weight loss on iron metabolism remains poorly understood. Exposure to low environmental oxygen increases dietary iron absorption and availability to supply iron for erythropoiesis [25], but this approach, particularly intermittent hypoxia, has not been investigated in the context of weight loss. The following measures are designed to investigate diet-induced weight loss and overnight low oxygen exposure as potential strategies to improve dietary iron absorption and indicators of iron status in obesity. Measures of iron status (Table 3) will be assessed in blood on days 0 and 56. Stable iron isotopes will be administered on days −7 (⁵⁷Fe) and 42 (⁵⁸Fe) to assess iron absorption. Participants will consume the isotope beverage, rinse the container with water, and drink the rinse to ensure complete intake of the isotopes. Blood collected on day 56 will be analyzed for iron isotope abundance to inform estimates of fractional iron absorption. Whole blood will be mineralized by microwave-assisted digestion in nitric acid and ⁵⁷Fe and ⁵⁸Fe incorporation into red blood cells will be determined by inductively coupled mass spectrometry using standardized techniques [26]. The abundance of ⁵⁶Fe, ⁵⁷Fe, and ⁵⁸Fe in whole blood will be calculated based on the isotopic shift between the natural isotopic distribution and the final distribution after consuming the isotopes. Circulating iron will be calculated by multiplying measured hemoglobin concentration with blood volume from the CO-rebreathing method (see section 3.9.5.). Fractional iron absorption will be calculated according to guidelines and assuming 80% incorporation of absorbed iron into red blood cells [26].

3.9.18. Gut microbiome composition and function

Participants will be provided with materials on days −14, −7, and 49 to self-collect a stool sample to bring to their next visit. Stool samples will be analyzed to determine microbiome composition and function with metagenomic sequencing and measurement of short-chain fatty acid concentrations.

3.9.19. Questionnaires

Questionnaires will be administered throughout the 3 study phases to assess environmental symptoms, appetite, sleep quantity and quality, psychosocial factors, and real and perceived barriers to sleeping in the tent system (Table 4). Participants will be asked to complete a single 24 h diet recall using the automated self-assessment questionnaire (ASA24) from the National Cancer Institute to assess habitual dietary intake prior to study start [27].

Table 4.

Questionnaires used to assess outcomes related to hypoxic exposure.

Outcome	Test	Description
Environmental symptoms (administered in the morning upon waking and evening before bed on days 0–7, 14, 21, 28, 35, 42, and 56)	Environmental Symptoms Questionnaire (ESQ) and the Lake Louise Acute Mountain Sickness (LL-AMS) Scoring System [30]	The prevalence and severity of Acute Mountain Sickness (AMS) will be determined from information gathered using the shortened version of the ESQ and the LL-AMS Scoring System (26 questions total).
Psychosocial Factors (administered on days 0, 14, 56)	The Positive and Negative Affect Schedule (PANAS) [31]	The PANAS is one of the most widely used scales to measure mood and emotion. This scale is comprised of 20 items, with 10 items measuring positive affect (e.g., excited, inspired) and 10 items measuring negative affect (e.g., upset, afraid). Each item is rated on a fivepoint Likert Scale to measure the extent the affect has been experienced in the past week.
	Cohen Perceived Stress Scale (PSS) [32]	PSS is the most widely used psychological instrument for measuring the perception of stress. It is a measure of the degree to which situationsinone’s life are viewed as stressful. Items ask how unpredict able, uncontroll able, and overloaded respondents find their lives. The scale also includes questions about current levels of experienced stress. The PSS is composed of 10 questions with each scored on a 5-point Likert scale.
	Big Five Inventory-2 (BFI-2) [33]	BFI-2 is a reliable and valid personality measure. The BFI-2 is a 60-item questionnaire that hierarchic ally assesses the Big Five personality domains (i.e., openness, conscienti ousness, extraversion, agreeable ness, and neuroticism) and 15 morespecific facet traits. Each question in the BFI-2 is scored with a 5-point Likert scale.
	Delay Discounting questionnaire (also known as the Monetary-Choice Questionnaire) [34]	This is a 27-item questionn aire to assess whether the participant prefers smaller immediate rewards over delayed larger rewards. The questionn aire is scored by calculating where the respondent’s answers place amid reference discounting curves, where placement amid steeper curves indicates higher levels of impulsivity.
	Palatable Eating Motives Scale (PEMS) [35]	PEMS assesses how frequently one consumed tasty foods and drinks in the past year for various reasons. Examples of tasty foods/drinks are provided (i.e., fast foods, homemade fried foods, sweets, salty snacks, and nonalcoholic sugary drinks) and motives for consumption include coping (e.g., “to forget about your problems”), reward enhance ment (e.g., “because it gives you a pleasant feeling”), social (e.g., “to celebrate a special occasion with friends”), and conformity (e.g., “because your friends or family want you to eat or drink these foods or drinks”). Each item uses a 5-point Likert-like scale for scoring.
	Generalized Self-Efficacy scale (GSE) [36]	GSE is a 10-item psychometric scale that is designed to assess optimistic selfbeliefs to cope with a variety of difficult demands in life. Items are rated on a 4-point scale.
	Patient Health Questionnaire (PHQ-9) [37]	The PHQ-9 questionn aire is a 9-item, self-reported depression questionn aire used by health professionals to identify potential cases of major depressive disorders. This questionn aire identifies feelings over the past two weeks based on a fourpoint scale. Higher scores represent greater depression severity.
Sleep Quality (administered on days 0, 14, and 56)	Pittsburgh Sleep Quality Index (PSQI) [38]	The PSQI is an instrument used to measure the quality and patterns of sleep in adults over the past month. It differentia tes “poor” from “good” sleep quality by measuring seven areas (scored on a 4-point scale): subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medications, and daytime dysfunction over the last month. Higher scores represent greater sleep difficulties.
Daily sleep tracking (before and after sleep on days −14 to −1, 0–14, and 42–56)	Sleep diary [39]	This is a questionn aire administered in the morning and evening to assess daily sleep characteristics, sleep quality, and factors that affect sleep such as mood, caffeine intake, and substance use (e.g. alcohol, caffeine, and prescriptions).
Gastrointestinal health (assessed on days 0, 14, and 56)	Gastrointestinal health log [40]	This log (29 questions) asks participants how often gastrointe stinal symptoms and related issues occurred during the past week.
Appetite testing (assessed on days 0 and 56)	Visual analog scale [18]	Immediately prior to and following (30, 60, 90, 120 min) consumption of a 75 g glucose beverage, subjectively rated appetite (fullness, hunger, desire to eat, and prospective consumption) will be measured on 100 mm visual analog scales. Upon cessation of the adlibitum meal, subjectively rated appetite will be measured again. Composite satiety scores (CSS) will be calculated from the individual appetite scoresusing the equation: CSS = (fullness + (100 – desire to eat) + (100 – unger) + (100 – prospective food consumption) / 4. A higher CSS score indicates greatersatiety.
	Meal acceptability [41]	Acceptability of the ad libitummeal (on a 9-point scale) will be assessed before the meal (appearance and odor), immediately after starting the meal (taste, texture, and overallacceptability of the meal), and at the end of the meal (appearance, odor, taste, texture, and overall acceptability of the meal).
Real and perceived barriers to sleeping in a hypoxic tent (administered on day 56)	Barriers questionnaire	This questionn aire was developed by the study team to understand real and perceived barriers to participation in research studies that use a hypoxic tent system. The 17-item questionn aire uses a scale to quantify perceived barriers(on a 4-point scale) and openended questions.

CCS, Composite satiety scores; ESQ, Environmental Symptoms Questionnaire; GSE, Generalized Self-Efficacy scale; LL-AMS, Lake Louise Acute Mountain Sickness; PANAS, The Positive and Negative Affect Schedule; PEMS, Palatable Eating Motives Scale; PHQ, Patient Health Questionnaire; PSQI, Pittsburgh Sleep Quality Index; PSS, Perceived Stress Scale

3.10. Sample size calculation

A study sample size of 50 participants (n = 25 per group) was estimated to detect a 1.9 kg difference in body weight change between groups, the primary outcome variable, with an estimated standard deviation of 2.2 kg (effect size, Cohen’s d = 0.86), power set to 0.85, and α set to 0.05. Estimates were based on a previous study in adolescents with obesity that demonstrated a 1.6 kg greater weight reduction in those who slept 10 h/night in a low oxygen environmental chamber (14.7% O₂, −8.7 ± 2.2 kg) compared to those who slept in normal conditions (~21% O₂, −7.1 ± 2.2 kg) for 4 week as part of a weight-loss camp that included 4 h/d of structured exercise [8]. In another study, males who were overweight lost 1.2 ± 0.3 kg body weight after sleeping 10 consecutive nights (10 h/night) in individual hypoxic tents (15% oxygen), although there was no control group [9]. Our estimated mean difference in body weight assumes a slightly greater weight loss due to the longer study duration (i.e., 8 wk vs. 4 wk or 10 d), but remains conservative due to the shorter overnight exposure (i.e., 8 h vs. 10 h). We will enroll 30 participants per group (n = 60 total), assuming a 17% attrition rate, which is a conservative estimate based on a previous study investigating normobaric hypoxia as a weight loss intervention [8].

3.11. Statistical analysis

Statistical analysis will be performed using SAS (SAS Institute, Cary, NC). Means and standard deviations will be calculated for continuous variables and proportions will be calculated for binary/discrete variables. Continuous variables will be assessed for normality and transformed if necessary. All primary analyses will be intention-to-treat. Between group comparisons for phase 1 variables and change in outcome variables with only 2 time points will be analyzed using two sample Student’s t-tests which is robust even for non-normal and heavy tailed distributions [28]. Each outcome variable with more than 2 repeated measurements will be analyzed as change from baseline (e.g., change from baseline on day 1 through 56 for body weight) using a mixed effect linear model [29]. Models will include treatment (hypoxia or sham), day, and treatment-by-day interaction as fixed effects covariates in the model. The random effect will use a covariance matrix to account for the correlation within-subjects over time. Least squares means from the model will estimate interaction effects. For outcomes with significant treatment-by-day interactions, within group comparisons across days and between group comparisons within each day will be made using two sample Student’s t-tests based on the least squares means and adjusted using the Bonferroni correction. Outcome variables with repeated time points within a day (e.g., blood parameters) will be analyzed as absolute values using a mixed effect linear model. Models will include treatment (hypoxia or sham), day, time, and all possible interactions as fixed effects covariates in the model. The random effect will use a covariance matrix to account for the correlation within-subjects over time. Least squares means from the model will estimate interaction effects. For outcomes with significant interactions, comparisons will be made using two sample Student’s t-tests based on the least squares means and adjusted using the Bonferroni correction.

4. Discussion

Body weight gain, specifically accumulation of adipose tissue, is multifactorial and, as such, requires a range of prevention and treatment options that are equally as diverse. This investigation will be the first to manipulate overnight environmental oxygen concentrations to promote body weight loss in adults with obesity. Findings from this study will examine the determinants and modulators of energy balance in response to hypoxic exposure, which may inform new or modified strategies to accelerate weight loss, aid long-term weight management efforts, and benefit metabolic health among those with obesity.

ABBREVIATIONS

ADOPT

Accumulating Data to Optimally Predict Obesity Treatment

AMS

Acute Mountain Sickness

AUC

Area under the curve

BMI

Body mass index

CGM

Continuous glucose monitor

DLW

doubly labeled water

DXA

Dual X-ray absorptiometry

F_iO₂

inspired oxygen fraction

Hbmass

Hemoglobin mass

HH

Hypobaric hypoxia

Kcal

kilocalories

NH

normobaric hypoxia

OGTT

oral glucose tolerance test

P_iO₂

Partial pressure of inspired oxygen

PBRC

Pennington Biomedical Research Center

RMR

resting metabolic rate

TDEE

total daily energy expenditure