Female Endurance Runners Have a Healthier Diet than Males—Results from the NURMI Study (Step 2)

as been recognized to be an important indicator of physiological, psychological, and nutritional characteristics among endurance athletes. However, there are limited data addressing sex-based differences in dietary behaviors of distance runners. The aim of the present study is to explore the sex-speci c differences in dietary intake of female and male distance runners competing at >10-km distances. From the initial number of 317 participants, 211 endurance runners (121 fe- males and90 males) were selected as the nal sample after a multi-level data clearance. Participants were classi ed to race distance (10-km, half-marathon, marathon/ultra-marathon) and type of diet (omnivorous, vegetarian, vegan) subgroups. An

c differences in dietary intake of female and male distance runners competing at >10-km distances. From the initial number of 317 participants, 211 endurance runners (121 fe- males and90 males) were selected as the nal sample after a multi-level data clearance. Participants were classi ed to race distance (10-km, half-marathon, marathon/ultra-marathon) and type of diet (omnivorous, vegetarian, vegan) subgroups. An online survey was conducted to collect data on sociodemographic information and dietary intake (using a comprehensive food frequency question- naire with 53 food groups categorized in 14 basic and three umbrella food clusters). Compared to male runners, female runners had a signi cantly greater intake in four food clusters, including “beans and seeds”, “fruit and vegetables”, “dairy alternatives”, and “water”. Males reported higher intakes of seven food clusters, including “meat”, “ sh”, “eggs”, “oils”, “grains”, “alcohol”, and “processed foods”. Generally, it can be suggested that female runners have a tendency to consume healthier foods than males. The predominance of females with healthy dietary behavior can be potentially linked to the well-known differences between females and males in health attitudes and lifestyle patterns. Keywords: sex; gender; nutrition; dietary assessment; food frequency; protein; fruit; vegetables; distance running; half-marathon; marathon 1. Introduction The importance of sex-related comparison in sports nutrition topics has been widely discussed over the past decade [1]. It is well-established that the nutritional requirements of athletes are potentially affected by physical and physiological differences between males and females [2,3]. These sex-based differences seem to be more predominant in ultra-endurance athletes who are recommended to pay superior attention to their speci c nutritional needs due to the prolonged training/racing activities [4,5]. Sex differences in endurance performance are not limited to the menstrual cycle that causes unfavorable effects on training procedures in female athletes (mainly due to the asso- ciated challenges and anemia rather than hormonal uctuations) [6,7]. Evidence shows that Nutrients2022,14, 2590.

Nutrients2022,14, 2590 2 of 17 females have a lower oxygen-carrying capacity (due to fewer erythrocytes and hemoglobin levels) than males, which can affect their endurance performance negatively [8]. In addition, females are more susceptible to developing thyroid disorders compared to males [9] result- ing in performance-limiting outcomes, including fatigue [10]. However, males seem to be more prone to cardiovascular abnormalities as it has been shown that cardiac death and coronary heart disease are more prevalent in males than females [11,12], which increases the likelihood of unfavorable health- and performance-related consequences. Considering the fact that male athletes are characterized as being more in uenced by risky behaviors such as performance-enhancing substance abuse [13,14], their cardiovascular health is of greater concern. Research indicates that in muscle metabolism pathways during endurance activities, females have a higher capacity to utilize muscle lipids as fuel, and males rely more on muscle and liver glycogen resources [15,16]. To achieve an optimal level of en- durance performance, however, females may need further training adaptations compared to males [17,18] due to the basic sex-speci c physical differences (e.g., body mass, muscle mass, and fat mass) [6,19]. Nutritional requirements and patterns may also be affected by sex, whether dependent or independent of the mentioned physical and physiological differences between males and females. It has been shown that female athletes have a greater prevalence of uninten- tional caloric imbalance than males in order to reach and maintain the appropriate body composition required for an optimized level of endurance performance [6,18,19]. Females have also been reported to be generally more health conscious than males, which also can be associated with their attitudes towards food choice, including a greater intake of fruits, vegetables, and whole foods [20]. In contrast, it has been shown that males are more motivated to increase physical activity in their daily routines rather than modifying their nutritional habits [21]. Generally, the various health- and lifestyle-related beliefs between females and males have been predicted to be responsible for up to 50% of sex-speci c dietary choices [20]. Dietary assessment is a crucial part of sports nutrition practice, which helps

shown that males are more motivated to increase physical activity in their daily routines rather than modifying their nutritional habits [21]. Generally, the various health- and lifestyle-related beliefs between females and males have been predicted to be responsible for up to 50% of sex-speci c dietary choices [20]. Dietary assessment is a crucial part of sports nutrition practice, which helps identify nutritional inadequacy (that commonly occurs following restrictive diets) and optimize dietary strategies for improving performance and health. Nutritional concerns, particu- larly energy de ciency, are more critical in both male and female long-distance runners compared to those who run in shorter races [18,22]. Likewise, nutritional requirements are positively associated with increasing intensity, duration, and frequency of running/training sessions [18,23]. Data show that typical daily foods may not ful ll the nutritional needs of endurance runners to support their physiological requirements [22,24]. This concern is more serious for endurance athletes who follow unbalanced and/or inappropriately- planned diets, which has been shown to occur in all diet types (e.g., omnivorous or plant- based diets) [25–27]. It has been reported that even ultra-endurance events can be com- pleted successfully without any health-related consequences by athletes who consume only plant-based foods [28,29]. This nding supports that by following the well-recognized dietary guidelines, appropriately planned plant-based diets can maintain the health of long-distance runners [28,29]. Regardless of the well-established sex differences in physical, physiological, and nutritional characteristics of general populations [30], there is limited evidence comparing dietary intake between male and female endurance athletes, particularly distance runners. Despite the advancement of knowledge in illustrating sex-based differences, the majority of sports nutrition topics have a paucity of female-speci c examinations, resulting in the misapplication of many scienti c conclusions for female athletes [31]. Available studies regarding the nutrient requirements of endurance athletes [32–34] are not consistent in covering all sex-based differences, or they did not distinguish race distance and diet type of female and male endurance runners [35,36]. Therefore, the present study was conducted to investigate and compare the dietary intake of female and male distance runners across different subgroups of diet type and race

regarding the nutrient requirements of endurance athletes [32–34] are not consistent in covering all sex-based differences, or they did not distinguish race distance and diet type of female and male endurance runners [35,36]. Therefore, the present study was conducted to investigate and compare the dietary intake of female and male distance runners across different subgroups of diet type and race distance. It was hypothesized that female runners have a dietary intake more advantageous to health.

Nutrients2022,14, 2590 3 of 17 2. Materials and Methods 2.1. Study Design and Ethical Approval The present study is a part of the Nutrition and Running High Mileage (NURMI) Study Step 2. The study protocol [37] was approved by the ethics board of St. Gallen, Switzerland (EKSG 14/145; 6 May 2015) with the trial registration number ISRCTN73074080. The methods of the “NURMI Study Step 2” have been previously described in detail [38,39]. 2.2. Participants and Experimental Approach Endurance runners were mainly recruited from Austria, Germany, and Switzerland and were contacted via social media, websites of organizers of marathon events, online running communities, email lists, and runners' magazines, as well as via additional/other multi-channel recruitments and through personal contacts. Participants were asked to com- plete an online survey within the “NURMI Study Step 2”, which was available in German and English (https://www.nurmi-study.com/en were provided with a written description of the procedures and gave their informed consent before completing the questionnaire. The following inclusion criteria were initially required for successful participation in the “NURMI Study Step 2”:(1) writteninformed consent; (2) at least 18 yearsof age; (3) questionnaire Step 2 completed;(4) successfulparticipation in a running event of at least half-marathon distance in the past two years. Female and male participants were further categorized according to race distance and kind of diet. Race distance subgroups were half-marathon and (ultra-)marathon (data were pooled since the marathon distance is included in an ultra-marathon); the shortest and longest ultra-marathon distances reported were 50 km and 160 km, respectively. However, a total number of 74 runners who completed the 10-km distance, but had not successfully participated in either a half-marathon or a marathon, also provided accurate and useable answers similar to runners competing over half-marathon or higher. In order to avoid an irreversible loss of these valuable data sets, those who met the inclusion criteria (1) to (3) were kept as additional race distance subgroup. Dietary subgroups were omnivorous (or Western diet, with no restriction on any food items), vegetarian (devoid of all esh foods, including sh and shell sh, but including eggs and/or dairy products), and vegan

order to avoid an irreversible loss of these valuable data sets, those who met the inclusion criteria (1) to (3) were kept as additional race distance subgroup. Dietary subgroups were omnivorous (or Western diet, with no restriction on any food items), vegetarian (devoid of all esh foods, including sh and shell sh, but including eggs and/or dairy products), and vegan diet (devoid of all foods from animal sources, including honey) [40,41] with a minimum of 6-month adherence to the self-reported diet types. 2.3. Data Clearance From the initial number of 317 endurance runners, a total of 106 participants were excluded from the data analysis. Of these, 46 participants did not meet the basic inclusion criteria. In order to control for a minimal status of health linked to a minimum level of tness and to further enhance the reliability of data sets, the Body Mass Index (BMI) approach following the World Health Organization (WHO) standards [42,43] was applied. On this basis, one participant with a BMI 30 kg/m 2 was excluded from the data analysis since rst other health-protective and/or weight loss strategies other than running are necessary to safely reduce body weight. Further, as a result of the speci c exclusion criteria for the present study, an additional number of 25 runners were identi ed and excluded for consuming 50% carbohydrates of their total dietary intake (which is lower than the minimum level recommended for maintaining a health-performance association [25,44,45]). Moreover, 34 participants with con icting statements on water intake (e.g., stated never drinking water) were excluded from the analysis to avoid con icting data on dietary intake [44]. In addition, a total of 24 runners (11%) had to be shifted to other dietary subgroups: 4 vegan runners: respectively 2 to omnivores and 2 to vegetarian samples; and 20 (9%) vegetarian runners had to be shifted to the omnivores subsample. However, 89% (n= 187) of the recreational runners correctly assessed their kind of diet. As the nal sample, 211 runners (121 women and 90 men) with complete data sets were included for statistical analysis. Figure the present study.

to omnivores and 2 to vegetarian samples; and 20 (9%) vegetarian runners had to be shifted to the omnivores subsample. However, 89% (n= 187) of the recreational runners correctly assessed their kind of diet. As the nal sample, 211 runners (121 women and 90 men) with complete data sets were included for statistical analysis. Figure the present study.

Nutrients2022,14, 2590 4 of 17Nutrients 2022, 14, x FOR PEER REVIEW 4 of 18 subgroups: 4 vegan runners: respectively 2 to omnivores and 2 to vegetarian samples; and 20 (9%) vegetarian runners had to be shifted to the omnivores subsample. However, 89% (n = 187) of the recreational runners correctly assessed their kind of diet. As the final sam- ple, 211 runners (121 women and 90 men) with complete data sets were included for sta- tistical analysis. Figure 1 shows the participants’ enrollment and classifications within the present study. Figure 1. Participants’ enrollment and classifications by sex. 2.4. Measures and Statistical Modelling Based on the food frequency questionnaire (FFQ) of the “German Health Interview and Examination Survey for Adults (DEGS)” (DEGS-FFQ; with friendly permission of the Robert Koch Institute, Berlin, Germany) [46,47], participants were asked to report their regular food intake based on the consumption frequency (single-choice out of 11 options ranging from “never” to “5 times a day”) and quantity of a broad variety of specific die- tary items (single-choice from various options depending on the food group) particularly in the past four weeks, including meals eaten while out, i.e., in restaurants, canteens, at friends’ houses, etc. Based on the 53 food groups of the DEGS-FFQ and following the Nova classification system of the Food and Agriculture Organization (FAO, Rome, Italy) [48–51], subgroups of foods were categorized with the corresponding questions pooled for a total of 17 food clusters in order to perform quantitative and qualitative data analyses (Table 1). Self-re- ported data, including sociodemographic information, motive(s) for diet type adherence, and pooled food frequency, were linked to sex-based groups. Figure 1.Participants' enrollment and classi cations by sex. 2.4. Measures and Statistical Modelling Based on the food frequency questionnaire (FFQ) of the “German Health Interview and Examination Survey for Adults (DEGS)” (DEGS-FFQ; with friendly permission of the Robert Koch Institute, Berlin, Germany) [46,47], participants were asked to report their regular food intake based on the consumption frequency (single-choice out of 11 options ranging from “never” to “5 times a day”) and quantity of a broad variety of speci c

(FFQ) of the “German Health Interview and Examination Survey for Adults (DEGS)” (DEGS-FFQ; with friendly permission of the Robert Koch Institute, Berlin, Germany) [46,47], participants were asked to report their regular food intake based on the consumption frequency (single-choice out of 11 options ranging from “never” to “5 times a day”) and quantity of a broad variety of speci c dietary items (single-choice from various options depending on the food group) particularly in the past four weeks, including meals eaten while out, i.e., in restaurants, canteens, at friends' houses, etc. Based on the 53 food groups of the DEGS-FFQ and following the Nova classi cation system of the Food and Agriculture Organization (FAO, Rome, Italy) [48–51], subgroups of foods were categorized with the corresponding questions pooled for a total of 17 food clusters in order to perform quantitative and qualitative data analyses (Table). Self- reported data, including sociodemographic information, motive(s) for diet type adherence, and pooled food frequency, were linked to sex-based groups.

Nutrients2022,14, 2590 5 of 17 Table 1. Modeling of the Clusters for Food Frequency (Basic Nutrition and Consumption Cluster 1 to 14; Umbrella Cluster for Preparation Cluster 15 to 17). Basic Food Clusters Cluster 1 Grains a—grains b—whole grains corn akes; white bread; white pasta muesli; wholegrain; mixed bread; wholegrain pasta; wholegrain rice; other grains Cluster 2 Legumes, nuts, and pulses pulses; nuts and seeds; legumes Cluster 3 Fruit and vegetables vegetable juice; fruit; vegetables Cluster 4 Dairy products milk; cheese; yoghurt Cluster 5 Dairy alternatives milk alternatives Cluster 6 Meat a—meat b—processed meat chicken; beef; pork; deer fried nuggets; hamburger; sausage; kebab; pork; processed meat Cluster 7 Meat alternatives tofu; seitan; tempeh; etc. Cluster 8 Fish, shell sh, and seafood Cluster 9 Eggs Cluster 10 Oils and spreads butter; margarine; oils Cluster 11 Sweets and snacks sweets; snacks; salty snacks Cluster 12 Water and unsweetened tea Cluster 13 Beverages Cluster 14 Alcohol Preparation/Umbrella Clusters Cluster 15 Protein a—plant protein b—animal protein legumes and beans; vegetables; grains (couscous, quinoa); dairy alternatives (e.g., soy products); meat alternatives dairy products; eggs; meat and processed meat products; sh, seafood, and shell sh Cluster 16 (Ultra-)processed foods and free/added sugar sugary carbonated drinks; kcal reduced/arti cially sweetened drinks; fruit juice; free sugar in tea; free sugar in coffee; cereals; sweet and savory spreads; margarine; pasta; sweets, cakes, and biscuits; salty snacks, butter; processed meat; processed plant products Cluster 17 Free/added sugar Sweet spread; sugary carbonated drinks; fruit juice; free sugar in tea; free sugar in coffee; cereals; sweets, cakes, and biscuits 2.5. Statistical Analysis The statistical software R version 4.1.1 (10 August 2021) Core Team 2018 (R Foundation for Statistical Computing, Vienna, Austria) was used to perform all statistical analyses. Exploratory analysis was done by descriptive statistics: mean values and standard deviation (SD), median and interquartile range (IQR). Chi-square tests ( 2, nominal scale) were conducted to examine the association of sex with nationality, marital status, academic quali cation, diet type, race distance, and dietary motives. Kruskal–Wallis tests (ordinal and metric scale) were approximated by using the t or F distributions or using ordinary

by descriptive statistics: mean values and standard deviation (SD), median and interquartile range (IQR). Chi-square tests ( 2, nominal scale) were conducted to examine the association of sex with nationality, marital status, academic quali cation, diet type, race distance, and dietary motives. Kruskal–Wallis tests (ordinal and metric scale) were approximated by using the t or F distributions or using ordinary least squares and standard errors (SE) with R 2 to test the association of sex with age, body weight, height, and BMI. Food cluster as the latent variable was derived by 53 manifest parameters (assessing how often and how much consumption of speci c dietary items). In order to scale the food consumption displayed by measures, items, and clusters, a heuristic index (as a new composite variable) ranging from 0 to 100 was de ned (equivalence in all items; FFQ was calculated by multiplying the reports of both questions, and dividing by

The study explores dietary differences between female and male distance runners.