Low Energy Availability and Increased Risk of Relative Energy Deficiency in Sport (RED-S) During a 3767-km Thru-Hike on the Pacific Crest Trail: A Case Study

Abstract

Long-distance “thru-hiking” has extraordinary physical demands and has become increasingly popular. This report describes a man (55 y) who thru-hiked the Pacific Crest Trail in 2021 and was at risk of developing the relative energy deficiency in sport (RED-S) syndrome. Hiking distance was 3767 km over 128 d. Eighty-eight days (69%) were full days of hiking, covering 38±8 km/d (mean±SD) in 7.9±1.6 h/d. Exercise energy expenditure above rest (heart rate vs indirect calorimetry regression method) was 2834±1518 kcal/d, total energy expenditure was 5702±1323 kcal/d, and energy intake was 4141 kcal/d. Body mass decreased by 9%, and fat mass (dual-energy X-ray absorptiometry) decreased by 46%. Energy availability (energy intake minus exercise energy expenditure) was 19.3 kcal/d/kg fat-free mass, indicating low energy availability (defined as <30 kcal/d/kg). Dual-energy X-ray absorptiometry–measured spine bone mineral density (BMD) decreased by 8.6%, with little to no decrease in total hip (−1.0%) and femoral neck (−1.5%) BMD. Total cholesterol, low-density lipoprotein cholesterol, and triglycerides increased by 24, 39, and 57%, respectively. Within 8 mo after the hike, BMD and serum lipids nearly or fully returned to baseline. No changes in high-density lipoprotein cholesterol, glycemia, or blood pressure were observed. According to guidelines, these observations are consistent with a moderate risk of RED-S, and a medical evaluation and treatment plan are advisable in order to avoid clinical manifestations (eg, stress fractures, anemia, psychological disturbances). To minimize RED-S risk, it may be prudent for thru-hikers to optimize energy availability by moderating daily hiking distances and/or increasing food intake.

Keywords

bioenergetics energy availability bone loss dyslipidemia metabolic adaptation

Introduction

Relative energy deficiency in sport (RED-S) is a medical syndrome caused by insufficient energy availability to meet the needs for both physical activity and basic biological functions.¹ It was originally identified in female athletes as a triad of low energy intake (often secondary to disordered eating), bone demineralization, and amenorrhea. Relative energy deficiency in sport is now known to occur in both men and women, with manifestations that may also involve endocrine, metabolic, immunological, cardiovascular, psychological, growth/developmental, and other conditions; additionally, it may result from low food intake that is intentional (eg, disordered eating) and/or unintentional (eg, poor food access) and/or from high energy expenditures.¹

Wilderness backpacking trips that cover the full length of a long-distance hiking trail in a single season are often referred to as “thru-hikes” and typically involve hiking 20 to 40 km/d, 6 to 7 d/wk, for 4 to 6 mo. Because of the high energy cost of these endeavors, paired with limited food access in the wilderness, the risk of RED-S may be high. The present report describes several physiologic effects of a long-distance thru-hike, as observed in a middle-aged man who completed the 4265-km Pacific Crest Trail (PCT) in the western United States in 2021. Information presented in this report may shed light on medical issues that may become increasingly common, given the many-fold increase in the popularity of thru-hiking in recent decades.^2,3

Case Presentation

The case was a 55-y-old Caucasian man who had a lifelong history of competitive and recreational endurance sports and activities. He consumed a lacto-ovo vegetarian diet. Medical history included osteopenia, with no osteoporotic fractures. He was not taking prescription medications but took oral supplements of vitamin B12 (2000 microgram/wk) and D3 (2000 IU/d, except during the thru-hike) and ibuprofen as needed. The hiker did not seek medical care for any reason during the hike, and we did not screen for clinical manifestations of RED-S or other medical conditions.

Assessments were performed 3 d before the hike started, 10 d after completing the hike, and at 8 and 12 mo after hike completion. The delay between the final day of hiking and the posthike assessments was a consequence of the time required for the hiker to get to the laboratory from the remote location where he finished the hike; it also allowed for acute effects of exercise (eg, water and glycogen balance, inflammation) to dissipate before testing. Assessments included body weight, body composition and bone mineral density (BMD) via dual-energy X-ray absorptiometry (Lunar Prodigy, GE HealthCare, Inc, Chicago IL), fasting venous blood lipids and glycemia metrics (Quest Diagnostics Inc, Secaucus NJ), and blood pressure. As described in detail below, food intake was evaluated during the hike and energy expenditure was estimated retrospectively.

The thru-hike occurred from May 24, 2021 through September 28, 2021 (128 d), including 88 full days of hiking (69% of all days), 15 partial hiking days, 15 nonhiking days for resupply, and 10 d off-trail for a family emergency. Total hiking distance was 3767 km (2-s tracking intervals, InReach Explorer+, Garmin Inc, Olathe KS), which accounted for sections that were not hiked due to trail closures for wild fires and added sections of trails that were near the PCT but not the PCT itself. Full hiking days consisted of 38±8 km of hiking over 7.9±1.6 h of moving time (mean±SD for multiple days), which is equivalent to 4.8 km/h while actively hiking. Backpack weight varied depending on the amounts of consumable items (food, water, and cooking fuel) being carried and ranged from 6.5 kg (“base weight,” when depleted of consumables) to 22.7 kg (after a full food resupply and with 8 L of water for long stretches without water access). However, loads at the extremes of this range were rare, and most of the time the load was estimated to be 10 to 20 kg.

Body mass decreased by 9% during the hike, fat mass decreased by 46%, and fat-free mass (FFM) increased by 7%. These changes returned to near-baseline levels at 8 and 12 mo after the hike (Table 1). Spine BMD decreased by 8.6%. Little to no change was observed for total hip (−1.0%) and femoral neck (−1.5%) BMD, potentially because these regions are exposed to more loading and they contain more cortical bone.¹ Bone mineral density values mostly recovered by the 8-mo follow-up (Table 1).

Table 1

Changes in body composition, blood pressure, blood lipids, and glycemia

	Prehike	Posthike	8-mo follow-up	12-mo follow-up
Body mass (kg)	86.5	78.8	85.5	85.3
BMI (kg/m²)	25.3	23.0	25.0	24.9
Fat mass (%)	22.7	13.0	22.3	21.7
Fat mass (kg)	20.0	10.8	19.2	18.6
Fat-free mass (kg)	67.9	72.6	67.1	66.9
L1-L4 spine BMD (g/cm²)	0.932	0.852	0.929	0.922
L1-L4 spine BMD (T-score)	−2.5	−3.1	−2.5	−2.5
L1-L4 spine BMC (g)	56.55	52.79	57.57	56.44
Total hip BMD (g/cm²)	0.802	0.794	0.797	0.788
Total hip BMD (T-score)	−2.1	−2.1	−2.1	−2.2
Total hip BMC (g)	31.63	31.51	31.91	31.43
Femoral neck BMD (g/cm²)	0.780	0.768	0.777	0.778
Femoral neck BMD (T-score)	−2.2	−2.3	−2.3	−2.2
Femoral neck BMC (g)	4.73	4.61	4.78	4.82
Total cholesterol (mg/dL)	153	190	157	150
LDL cholesterol (mg/dL)	72	100	70	70
Non-HDL cholesterol (mg/dL)	88	122	82	85
HDL cholesterol (mg/dL)	65	68	75	65
Triglycerides (mg/dL)	81	127	50	74
Hemoglobin A1c (%)	5.3	5.2	—	5.0
Average glucose (mg/dL)	105	103	—	97
Fasting glucose (mg/dL)	85	86	—	81
Systolic BP (mm Hg)	108	110	116	107
Diastolic BP (mm Hg)	72	78	78	74

BMI, body mass index; BMD, bone mineral density; BMC, bone mineral content; LDL, low-density lipoprotein; HDL, high-density lipoprotein; BP, blood pressure.

Total cholesterol increased by 24% (+37 mg/dL), low-density lipoprotein cholesterol increased by 39% (+28 mg/dL), and triglycerides increased by 57% (+46 mg/dL) (Table 1). All values remained within the recommended ranges and returned to baseline by the 8-mo follow-up. High-density lipoprotein cholesterol, glycemia measures, and blood pressure did not change.

Food intake was documented by the hiker with video and audio recordings. Most food was acquired as bulk acquisitions during resupply stops, but food and alcoholic beverages were also consumed and documented while visiting towns for resupply and when they were given to the hiker by altruistic “trail angel” individuals who voluntarily help hikers in various ways. Aggregate food consumption was recorded for 51 d, including 36 d early in the hike and 15 d near the end of the hike. Aggregate nutrient analyses of all documented foods were later performed (Food Processor software, version 11.0, ESHA Research Inc, Salem OR); the results were divided by 51 d to represent average daily intakes. Energy intake averaged 4141 kcal/d, with energy from carbohydrate and protein being within the acceptable macronutrient distribution range (AMDR). Energy from fat was above the AMDR (Table 2) at least partly because the hiker sought fat (including olive oil and powdered coconut milk added to other foods) as a calorie-dense energy source. Protein intake (1.3 g/kg/d) met the recommended dietary allowance of 0.8 g/kg/d. Much of this protein came from the small amounts present in many foods and also from protein-rich foods that were consumed frequently (eg, nuts, beans, hummus, and rice) or occasionally (cheese, eggs, and protein-enriched energy bars). Intakes of other selected nutrients that are relevant for vegetarians, bone health, and lipids are presented in Table 2. Examples of meals or foods consumed during periods of various food access conditions are presented in Table 3.

Table 2

Energy expenditure and nutrient intakes

	Value	Recommendation
Basal metabolic rate (kcal/d)	1923±11	—
Sleeping energy expenditure (kcal/d)	577±3
Hiking energy expenditure (above rest) (kcal/d)	2834±1518	—
Nonhiking physical activity (above rest) (kcal/d)	795±263	—
Thermic effect of food (kcal/d)	414	—
Total energy expenditure (kcal/d)	5702±1323	—
Energy availability (kcal/kg fat-free mass)	19.3	≥40 (IOC)
Total energy intake (kcal/d)	4141	—
Carbohydrate (g/d)	519	—
Carbohydrate (g/kg/d)	6.0	—
Carbohydrate (% of energy intake)	50	45–65 (AMDR)
Total sugars (g/d)	113	—
Added sugars (g/d)	32	<50 (DV)
Total dietary fiber (g/d)	53	≥30 (AI)
Protein (g/d)	116	—
Protein (g/kg/d)	1.3	≥0.8 (RDA)
Protein (% of energy intake)	11	10–35 (AMDR)
Fat (g/d)	173	—
Fat (% of energy intake)	37	20–35 (AMDR)
Saturated fat (g/d)	42	<20 (DV)
Saturated fat (% of energy)	9	<10 (DV)
Dietary cholesterol (mg/d)	160	<300 (DV)
Alcohol (g/d)	17	—
Alcohol (% of energy intake)	2	—
Sodium (mg/d)	6777	≤2300 (UL)
Calcium (mg/d)	1206	≥1000 (RDA)
Vitamin D (IU/d)	145	≥600
Vitamin B12	7	2.4 (RDA)
Iron (mg/d)	47	≥8 (RDA)
Zinc (mg/d)	8	≥11 (RDA)
Caffeine (mg/d)	234	—

Values are mean±SD; SD represents day-to-day variability. SD values are not available for nutrient intakes because the values are based on aggregate food intake data, not day-by-day intakes. Vitamin B12 intake does not include a 2000 microgram/wk supplement. Vitamin D intake does not include a 2000 IU/d supplement that was discontinued during the hike because of the large amounts of endogenous production (from sun exposure) that the hiker anticipated.

IOC, International Olympic Committee; AMDR, acceptable macronutrient distribution range as stated in the dietary reference intakes; DV, daily value (food label standards from US Food and Drug Administration); AI, adequate intake as stated in the dietary reference intakes; RDA, recommended dietary allowance as stated in the dietary reference intakes; UL, upper safe limit as stated in the dietary reference intakes.

Table 3

Examples of daily food consumption based on food availability

Meal and food items	Energy (kcal)
Sections with moderate- to high-level food access
Breakfast	680
Toaster pastries (2 48-g pastries) and frosted flakes ready-to-eat cereal (74 g, consumed dry)
Lunch	1420
Ramen noodles (2 85-g packages) mixed with instant flavored mashed potatoes (1 113-g package) and olive oil (30 mL)
Dinner	1488
Instant refried beans (1 205-g package) prepared with olive oil (30 mL) and eaten with corn chips (84 g)
Snacks (consumed throughout day)	1370
Roasted, salted peanuts (2 92-g packages)
Fig bar cookies (99 g, 7 cookies)
Other (consumed throughout day)	0
Coffee, instant, black (531 mL as prepared with water)
Water (8 L, filtered from streams or other remote water sources)
Total energy intake (kcal/d)	4958
Sections with poor food access
Breakfast	336
Honey roasted oats ready-to-eat cereal (84 g, consumed dry)
Lunch	370
Ramen noodles (1 85-g package)
Dinner	660
Stuffing mix, herb flavored, instant (1 170-g package)
Snacks (consumed throughout day)	880
Granola bars, crunch oats, and honey (2 21-g bars)
Roasted, salted peanuts (1 92-g package)
Cheese, aged white cheddar (42 g)
Other (consumed throughout day)	0
Coffee, instant, black (531 mL as prepared with water)
Water, 6 L (filtered from streams or other remote water sources)
Total energy intake (kcal/d)	2246
Occasions with all-day access to retail food sources
Breakfast	1908
Egg omelet (3 eggs with vegetables and cheese) and pancakes (3 20-cm diameter), with butter (45 g) and syrup (118 mL)
Lunch	1550
Mixed salad greens (285 g) eaten with guacamole (284 g) and flavored croutons (141 g)
Fruit and vegetable smoothie (450 mL)
Dinner	3210
Pizza, cheese, and vegetables; large (40 cm diameter)
Beer (2 473-mL portions)
Snacks (consumed throughout day)	334
Blueberries, fresh (340 g)
Bananas (1 medium size)
Other (consumed throughout day)
Coffee, brewed, black, 531 mL	0
Water, 2 L	0
Total energy intake (kcal/d)	7002

The meals and food depicted in this table are based on a recall interview with the hiker, in which he was asked to describe “typical” eating patterns. Because food consumption depended heavily on food access and other factors, 3 different scenarios are presented to illustrate the wide variation in eating patterns. For sections of the hike with “moderate- to high-level food access,” food was typically accessible (via retail stores or mailed shipments to post offices) every 3 to 5 d and without having to walk or hitchhike more than 3 to 5 miles from the trail. For sections of the hike with “poor food access,” food was typically only accessible every 7 d or more, often involved more than 10-mile walks or hitchhikes from the trail, and/or required food to be stored in a bear-proof canister, which limited food carrying capacity. For “occasions with all-day access to retail food sources,” food sources included grocery stores, convenience stores, and restaurants.

Hiking energy expenditure was estimated by performing laboratory and field assessments as part of the 12-mo follow-up, during which the case had a similar training routine and body mass to when he started the PCT. Hiking energy expenditure was estimated by using the heart rate vs energy expenditure regression approach.⁴ In brief, heart rate and energy expenditure (indirect calorimetry, Parvo Medics Inc, Salt Lake City UT) were measured in the laboratory during treadmill walking with a 15.0 kg backpack over a range of speeds and grades and were used to generate a linear regression equation. Then, during a hike on an outdoor trail, heart rate data were collected every 10 s and used to estimate hiking energy expenditure (kcal/kg/min) by using the linear regression equation. The hike for this assessment was designed to mimic a typical day on the PCT in terms of distance, pace, elevation gain or loss, trail tread surfaces, weather, and backpack weight (backpack load started at 16.5 kg and decreased to 13.7 kg as food and water were consumed). The results were then used to determine daily hiking energy expenditure for the thru-hike based on daily hiking duration and body mass from the PCT.

Net hiking energy expenditure (ie, above basal metabolism) was 2834±1518 kcal/d, which includes full and partial hiking days and nonhiking days. This is equivalent to 5.8±0.2 metabolic equivalents (METs) and aligns with published values for backpacking (5.6 and 6.0 METs above rest).^4,5 At the 12-mo follow-up, basal metabolic rate (BMR) was measured via indirect calorimetry (Parvo Medics Inc) to approximate “prehike” BMR. The measures were made in morning fasted state and in accordance with published standards.⁶ To predict daily BMR during the hike, measured BMR was adjusted for interpolated changes in FFM.⁷ The resulting estimate of BMR during the hike was 1923±11 kcal/d, which is similar to BMR predicted from FFM alone (1988±11 kcal/d).⁷ Thermic effect of food (414 kcal/d) was estimated as 10% of energy intake. Sleeping energy expenditure was estimated as 577±3 kcal/d based on sleep energy expenditure being 90% of BMR and the case’s self-reported 8-h daily sleep duration. All remaining awake time each day (ie, when not hiking or sleeping) was assumed to be light physical activity (setting up tent, filtering water, cooking, eating, etc) at 2 METs (1 MET above resting), which translates to 795±263 kcal/d. Based on all sources of energy expenditure combined, total daily energy expenditure was 5702±1323 kcal/d.

The daily energy deficit (energy intake − total energy expenditure) was 1561 kcal/d. Based on the observed 9.2 kg decrease in fat mass, the 4.7 kg increase in FFM, and the respective energy densities of these tissues (9436.6 and 1815.6 kcal/kg⁸), total body energy stores decreased by 25% (78,283 kcal) from baseline, which accounts for 682 kcal/d of the energy deficit. The remaining 880 kcal/d of the energy deficit cannot be accounted for based on the available data. Energy availability, calculated as the difference between energy intake and exercise energy expenditure above resting values,¹ was 19.3 kcal/d/kg FFM (Table 2). This value is well below the 30 kcal/d/kg FFM threshold that defines low energy availability and is associated with an increased risk of RED-S.¹

Discussion

The physiologic demands of long-distance thru-hiking are extraordinary, especially from an energy balance perspective, because energy expenditure is high and food availability and carrying capacity are limited. In the context of large increases in the popularity of long-distance thru-hiking,^2,3 paired with bioenergetic challenges, it is important to understand the physiologic and medical implications of this type of activity. The case described in this report was unable to consume enough food to meet his energy needs during his thru-hike, and this resulted in a state of low energy availability. Manifestations of this energy-deprived state were observed, including a large reduction in total body energy stores, demineralization of the spine (comparable to 30 y of aging-related loss), and large increases in serum cholesterol and triglycerides. These findings are striking, given that weight-bearing exercise has well-known beneficial effects on bone⁹ and lipid profile.¹⁰ Taken together, these observations suggest that the hiker was at increased risk of developing clinically relevant RED-S pathologies, such as stress fractures, immunological illness, anemia, and psychological disturbances. According to published medical guidelines (Relative Energy Deficiency in Sport Clinical Assessment Tool from the International Olympic Committee¹¹), the hiker was at moderate risk of RED-S and medical clearance and a medical treatment plan would be advised for him to continue in his “sport.” One other report documented the effects of a 112-d thru-hike of the PCT (44 km/d) in a 25-y-old man.¹² Body weight and composition did not change, possibly due to a very lean baseline phenotype (5.8% body fat), but he exhibited 5.2 and 3.8% reductions in spine and pelvic BMD, respectively. These changes in BMD are in line with low energy availability, especially given that weight-bearing exercise typically increases BMD.⁹ The case also exhibited vascular dysfunction as a consequence of the hike, which may also be a consequence of low energy availability.¹³

RED-S is a medical syndrome resulting from insufficient energy available for basal physiologic functions and exercise. This insufficiency is sensed by the central nervous system, which responds through neuroendocrine adaptations, including suppression of the hypothalamic-pituitary-gonadal axis and reductions in hepatic insulin like growth factor-1 secretion.¹⁴ The net effect is to conserve energy by slowing or stopping physiologic functions that are not necessary for short-term survival (such as reproductive functions and bone remodeling); this ensures that energy remains available for vital functions and exercise. For the case described in this report, there was an energy deficit of 1561 kcal/d. While much of this deficit was covered by mobilizing 682 kcal/d from total body energy stores, 880 kcal/d of the deficit remained unaccounted for. Given that accurate assessments of energy intake and expenditure are challenging,^15,16 measurement error may explain some or all of this difference. However, a physiologic explanation is also possible. Although we did not evaluate hormones, it is conceivable that neuroendocrine changes, as described above, reduced energy expenditure by ∼880 kcal/d, thereby accounting for this otherwise inexplicable portion of the energy deficit. Indeed, the large reductions in BMD are suggestive of attenuated anabolic activity, and the increases in serum cholesterol are consistent with reduced cholesterol use for gonadal hormone production. Within 8 mo of completing the thru-hike and resuming an energy-sufficient state, BMD returned to near-baseline levels and serum lipids fully recovered; this is consistent with the normalization of neuroendocrine status upon restoration of energy availability.^13,14

The unique observations presented in this report have important wilderness medicine implications but more research is needed. Studies are needed to understand the frequency and severity of low energy availability in thru-hikers and to document the prevalence of RED-S signs and symptoms. Measures of hormonal changes and metabolic adaptations are needed to better characterize physiologic responses to thru-hiking. Finally, although nutrition guidelines for athletes exist,¹⁷ it is not clear how relevant these are for thru-hiking; thus, research on the dietary requirements for thru-hiking are needed.

Summary and Conclusion

In summary, the person described in the present report completed a 4-mo, 3767-km thru-hike, during which energy intake was not sufficient to meet all physiologic needs and a state of low energy availability ensued. Despite the beneficial effect that exercise has on bone and serum lipids, BMD decreased substantially and large increases in serum cholesterol and triglyceride were observed. While the hiker did not present with clinically relevant sequelae of RED-S, such as stress fractures, immunological illness, anemia, or psychological disturbances, our findings suggest that he was at risk of such conditions. It is possible that occasional, short periods with high energy availability (eg, resupply days with high food access and little or no hiking) may have protected against overt clinical manifestations of RED-S; however, this is speculative. Within 8 mo after the hike, the changes that were observed reverted to baseline, reflecting the transient nature of his state. In a broader scope, this report should serve to raise awareness of the energetic challenges that occur during long-distance thru-hiking. Furthermore, although many thru-hikers successfully complete their hikes without major medical problems, it may be prudent to consider low energy availability as a potential causative factor in hikers who do develop health problems. From a practical perspective, a prudent strategy for reducing RED-S risk would be for thru-hikers to optimize energy availability by moderating daily hiking distances to reduce daily exercise energy expenditure and by increasing food access and intake by resupplying more often and/or carrying more food.

Footnotes

Acknowledgments

The authors thank the person who is described in this report for the time and effort he contributed toward this project.

Author Contributions: Conceived the idea, designed the project, analyzed and interpreted the data, drafted the article and revised it critically for important intellectual content, and gave final approval of the version to be published (EPW); contributed to the conception and design of the project and analyzed and interpretation the data, critically revised the manuscript for important intellectual content, and gave final approval of the version to be published (AMF, VRP).

Financial/Material Support: None.

Disclosures: None.

References

Mountjoy

Sundgot-Borgen

J.K.

Burke

L.M.

Ackerman

K.E.

Blauwet

Constantini

et al. IOC consensus statement on relative energy deficiency in sport (RED-S): 2018 update. Br J Sports Med 2018; 52, 687–697.

Pacific Crest Trail Association . 2600 miler list. https://www.pcta.org/discover-the-trail/thru-hiking-long-distance-hiking/2600-miler-list/. Accessed July 3, 2023.

Appalachian Trail Conservancy . 2000-Milers. https://appalachiantrail.org/explore/hike-the-a-t/thru-hiking/2000-milers/. Accessed July 3, 2023.

Hill

L.C.

Swain

D.P.

Hill

E.L.

Energy balance during backpacking. Int J Sports Med 2008; 29, 883–887.

Ainsworth

B.E.

Haskell

W.L.

Herrmann

S.D.

Meckes

Bassett

D.R.

Jr. Tudor-Locke

et al. 2011 Compendium of Physical Activities: a second update of codes and MET values. Med Sci Sports Exerc 2011; 43, 1575–1581.

Fullmer

Benson-Davies

Earthman

C.P.

Frankenfield

D.C.

Gradwell

Lee

P.S.

et al. Evidence analysis library review of best practices for performing indirect calorimetry in healthy and non-critically ill individuals. J Acad Nutr Diet 2015; 115, 1417–1446.e2.

Cunningham

J.J.

A reanalysis of the factors influencing basal metabolic rate in normal adults. Am J Clin Nutr 1980; 33, 2372–2374.

Hall

K.D.

What is the required energy deficit per unit weight loss?. Int J Obes (Lond) 2008; 32, 573–576.

Guadalupe-Grau

Fuentes

Guerra

Calbet

J.A.

Exercise and bone mass in adults. Sports Med 2009; 39, 439–468.

10.

Martínez-Vizcaíno

Amaro-Gahete

F.J.

Fernández-Rodríguez

Garrido-Miguel

Cavero-Redondo

Pozuelo-Carrascosa

D.P.

Effectiveness of fixed-dose combination therapy (Polypill) versus exercise to improve the blood-lipid profile: a network meta-analysis. Sports Med 2022; 52, 1161–1173.

11.

Mountjoy

Sundgot-Borgen

Burke

Carter

Constantini

Lebrun

et al. The IOC relative energy deficiency in sport clinical assessment tool (RED-S CAT). Br J Sports Med 2015; 49, 1354.

12.

Heinbockel

T.C.

Craighead

D.H.

Case studies in physiology: impact of a long-distance hike on the Pacific Crest Trail on arterial function and body composition in a highly fit young male. Physiol Rep 2021; 9 e14767.

13.

Dipla

Kraemer

R.R.

Constantini

N.W.

Hackney

A.C.

Relative energy deficiency in sports (RED-S): elucidation of endocrine changes affecting the health of males and females. Hormones (Athens) 2021; 20, 35–47.

14.

Wong

H.K.

Hoermann

Grossmann

Reversible male hypogonadotropic hypogonadism due to energy deficit. Clin Endocrinol (Oxf) 2019; 91, 3–9.

15.

Rousset

Médard

Fleury

Fardet

Goutet

Lacomme

Energy intake evaluation by a learning approach using the number of food portions and body weight. Foods 2021; 10, 2273.

16.

Leenders

N.Y.

Sherman

W.M.

Nagaraja

H.N.

Kien

C.L.

Evaluation of methods to assess physical activity in free-living conditions. Med Sci Sports Exerc 2001; 33, 1233–1240.

17.

Thomas

D.T.

Erdman

K.A.

Burke

L.M.

Position of the Academy of Nutrition and Dietetics, Dietitians of Canada, and the American College of Sports Medicine: nutrition and athletic performance. J Acad Nutr Diet 2016; 116, 501–528.