Athlete dietary nutrition investigation method

Abstract

The invention discloses an athlete dietary nutrition investigation method. The method comprises the following steps: firstly, collecting basic information of participants; blood samples of the participants are collected for subsequent biomarker analysis; after collecting the blood sample, the participants complete a first FFQ for evaluating their dietary habits; after the first-time FFQ, the participant records weighed food for three days, including two workdays and one weekend day, for evaluating the effectiveness of the first-time FFQ; the method comprises the following steps: analyzing biomarkers in a blood sample and food intake of three-day weighing, and comparing the biomarkers and the food intake with the result of the first FFQ to evaluate the effectiveness of the first FFQ; after the first FFQ is completed for one month, the participant completes the second FFQ again so as to evaluate the repeatability of the FFQ; and evaluating the consistency of the questionnaire at different time points by comparing the results of the first FFQ and the second FFQ.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of sports technology, and relates to a method for investigating the dietary nutrition of athletes. BACKGROUND

[0002] A well-balanced and targeted nutrition plan is essential for maintaining the health and optimal performance of athletes. Given the diverse nutritional needs during different training phases, regular dietary assessments are necessary for enabling athletes to fine-tune their nutrition strategies accordingly. However, traditional methods of dietary assessment for athletes remain challenging, mainly due to the strict requirements of training programs and logistical limitations brought about by off-site training and competition venues. Therefore, it is crucial to find a nutrition assessment tool that is both convenient and reliable, capable of monitoring nutrient intake on a regular basis. At the same time, by examining the correlation between dietary patterns and athletic performance, effective nutrition intervention strategies can be identified for the athlete population. SUMMARY

[0003] One of the embodiments of the present disclosure is a method for investigating the dietary nutrition of athletes, which includes the following steps: first, collecting the basic information of the participants; collecting blood samples of the participants for subsequent biomarker analysis; after collecting the blood samples, the participants complete a first FFQ for assessing their dietary habits; after the first FFQ, the participants conduct a three-day weighed food record, including two weekdays and one weekend day, for assessing the effectiveness of the first FFQ; by analyzing the biomarkers in the blood samples and the food intake during the three-day weighed food record, the results of the first FFQ are compared to assess the effectiveness of the first FFQ; one month after completing the first FFQ, the participants complete a second FFQ for assessing the repeatability of the FFQ; by comparing the results of the first FFQ and the second FFQ, the consistency of the questionnaire at different time points is assessed.

[0004] One of the beneficial effects of the embodiments of the present disclosure is to propose a dietary nutrition questionnaire for athletes and a questionnaire investigation method, which proves the effectiveness and accuracy of the investigation method through the verification of the relative effectiveness and repeatability of the short semi-quantitative food frequency questionnaire for athletes. BRIEF DESCRIPTION OF DRAWINGS

[0005] The above and other objects, features and advantages of the exemplary embodiments of the present application will be more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which:

[0006] Figure 1 Flowchart of subject selection and determination according to one of the embodiments of the present application.

[0007] WFR - weighed food record; FFQ - food frequency questionnaire.

[0008] Figure 2 Study protocol schematic according to one of the embodiments of the application.

[0009] WFR1 and WFR2 - in training state; WFR3 - in rest state.

[0010] Figure 3 Bland-Altman plot according to one of the embodiments of the application.

[0011] Figure 3 The agreement of 3-day weighed food records (3DWFRs) with the first food frequency questionnaire (FFQ1) in (A) energy, (B) protein, (C) fat and (D) carbohydrates is shown (n = 97). The agreement of 3-day weighed food records with the first food frequency questionnaire in the assessment of energy, protein, fat and carbohydrate intake is shown by Bland-Altman plots, with a sample size of 97.

[0012] Figure 4 Intestinal-type classification and microbial diversity in fecal samples schematic according to one of the embodiments of the application.

[0013] Figure 5 Microbial community composition comparison between different groups schematic according to one of the embodiments of the application.

[0014] Figure 6 Comparison of amino acid metabolism function abundance between different groups schematic according to one of the embodiments of the application. DETAILED DESCRIPTION

[0015] There is currently no convenient and effective tool for assessing dietary intake of professional athletes. The present disclosure aims to evaluate the relative validity of a short semi-quantitative food frequency questionnaire (FFQ) by comparison with 3-day weighed food records (3DWFRs) and corresponding serum biomarkers, and to assess its reproducibility. Relative validity was assessed by Spearman correlation coefficients, cross-quintile classification, weighted kappa statistics, and Bland-Altman analysis, while reproducibility was assessed by Spearman correlation coefficients and intraclass correlation coefficients (ICCs) between two FFQs. The results showed that the median crude correlation coefficients (CCs) between the first FFQ and 3DWFRs were 0.331 (0.219 to 0.568) and 0.292 (-0.035 to 0.455) for energy and nutrients, and food groups, respectively. The FFQ-estimated intakes of omega-3 polyunsaturated fatty acids (EPA, DHA, and EPA+DHA) were significantly correlated with the corresponding serum biomarkers, with correlation coefficients ranging from 0.389 to 0.520. Weighted kappa statistics showed acceptable agreement (>0.2) for most items, ranging from -0.076 to 0.581, except for vitamin C, vegetables, and oils and fats. Misclassification into extreme quintiles was rare, with median misclassification rates of 2% (1% to 12%) and 3% (0 to 10%) for nutrients and food groups, respectively. Bland-Altman analysis showed good agreement between the FFQ and 3DWFRs, with more than 90% of data points falling within the limits of agreement (LOA) for all assessed nutrients and food groups. In reproducibility analysis, the median crude CCs and ICCs were 0.574 (0.423 to 0.643) and 0.668 (0.558 to 0.763) for energy and nutrients, and 0.681 (0.242 to 0.764) and 0.640 (0.371 to 0.787) for food groups, respectively. In summary, the short FFQ showed acceptable relative validity and reproducibility for most nutrients and food groups, indicating its potential as a valuable tool for assessing dietary intake and nutritional status of young athletes in China.

[0016] One of the objectives of the present disclosure is to validate a short semi-quantitative food frequency questionnaire (FFQ) for assessing dietary intake of professional athletes, in order to achieve an athlete dietary nutrition survey analysis method. The relative validity and reproducibility of the FFQ were evaluated by comparison with 3-day weighed food records (3DWFRs) and serum biomarkers. The results showed that the FFQ had acceptable validity and reproducibility in the assessment of most nutrients and food groups, indicating its potential for practical application.

[0017] The present disclosure is a cross-sectional study involving healthy professional athletes aged 14 to 30 years, conducted by a certain Institute of Sports Science in 2022-2023. Participants were recruited through posters distributed in a certain professional sports team between September 4 and November 20, 2022. Inclusion criteria required athletes to be at least second-level players, ranked in the top 24 in championships organized by provincial and local sports bureaus, regularly eat at designated restaurants in training bases, and have no known diseases or recent injuries. Initially, 143 athletes expressed interest, but after laboratory interviews, 41 were excluded due to irregular eating patterns, age restrictions, recent medication, or refusal to participate in a three-day weighing agreement. The final cohort included 102 professional athletes, with 49 males and 53 females, with an average age of 20 years. Sports events included cycling, track and field, handball, badminton, martial arts, gymnastics, volleyball, swimming, boxing, shooting, and archery. The study was approved by the Ethics Committee of a certain Institute of Sports Science (LLSC20220021), and all participants provided written informed consent before the study began. Figure 1 A flowchart showing subject selection and determination is presented.

[0018] The short semi-quantitative FFQ in this disclosure is adapted from the 25-item FFQ developed by Gao et al. for assessing the habitual dietary intake of the past year in the elderly population in the Changfeng Study. In this disclosure, the recall period of the FFQ was modified from 1 year to 1 month, mainly considering that athletes usually change their dietary strategies according to different training periods, which makes short-term dietary assessment more suitable for actual needs. In addition, due to the particularity of athletes' diet and the potential role of nutrients in athletes' health and performance, four items (legumes, instant food, sports drinks, and protein powder) and one open question about the use of nutritional supplements were added. Once the answer to the use of nutritional supplements is "yes", the athlete needs to report in detail the brand and specification of the supplement, the dose, the duration of use, and the frequency of use. The nutrient content provided by the supplement is calculated based on the above information. Finally, the FFQ questionnaire includes 29 items and one open question. The 29 items include flour, rice, dumplings, whole grains and their products, potatoes and their products, soybeans and their products, legumes, green leafy vegetables, dark vegetables, light vegetables, mushrooms and algae, fresh fruits, nuts, livestock meat, poultry meat, dairy products, eggs, freshwater fish, seafood, desserts, instant food, puffed food, candied fruits, sweet drinks, beer, yellow rice wine, white wine, sports drinks, and protein powder. Participants were asked how often they consumed each food and how many servings they consumed each time. The frequency includes nine options: "never", "less than 1 time per month", "1-3 times per month", "1-2 times per week", "3-4 times per week", "5-6 times per week", "once a day", "twice a day", and "three or more times a day", and the serving size is given in easy-to-evaluate measurement units such as "fist", "palm", "chunk", "handful", as well as common container units such as "spoonful", "bag", "cup", "bottle", and "spoon" to reduce the burden on participants. These units are finally converted to the intake value of each item based on experience and laboratory measurements. For fixed food items in the FFQ, the daily food intake is the product of the daily food intake frequency and the amount of food intake each time. The energy and nutrient content of each food item is calculated based on the average energy and nutrient values of 3-10 representative foods that make up the item, based on "Chinese Food Composition Table (6th Edition)". These foods are determined by investigating the most frequently purchased food ingredients in the athletes' canteen, combined with the results of previous dietary surveys specifically for athletes.

[0019] Each participant was given a portable food scale and some disposable cutlery for holding food, and detailed instructions were explained to the participants both orally and in writing. In brief, the present disclosure required participants to individually weigh and record the weight of each dish, then take a photo through the mobile phone and upload the weighing data into the designated folder, while filling in the dish name and the ingredients of the mixed food. After the meal, the participants were instructed to weigh and take a photo of the uneaten food in the same way. If there was no food left, there was no need to upload more information. The weight of the disposable cutlery had been obtained in advance, so the participants did not need to zero or re-weigh the empty bowl, in order to minimize operational errors and reduce the burden on the participants. On weekdays, the participants usually had meals at the designated canteen, and the preparation proportions of all dishes had been obtained in advance and loaded into the nutrition calculator program; on weekends, participants who might leave the training base for meals were required to carry the food scale to weigh the food in the same way as on weekdays, upload the food photo, and describe the ingredients of the mixture and the approximate proportions of the food in detail. The nutritionists checked the folder at mealtime and asked the participants to re-shoot, re-upload or re-fill in if the photo was not clear or lacked information, in order to ensure the accuracy and effectiveness of the results. Finally, the data of the 3DWFRs were calibrated by four trained nutritionists and input into the nutrition calculator program (Beijing Bowen Information Technology Co., Ltd.), which automatically calculated and summarized the daily energy and nutrient intake according to the latest Chinese Food Composition Table (6th edition).

[0020] Venous blood samples were collected from fasting subjects and placed in vacuum tubes without anticoagulant. Serum was separated by centrifugation at 3000 x g for 15 minutes, and multiple aliquots of each sample were stored at -80°C for subsequent analysis of serum fatty acids. Serum total fatty acids were quantified by using a fatty acid methyl esterification kit produced by Sigma-Aldrich Extraction was performed according to its instructions. Lipid determination was performed using a GC-MS / MS system (Agilent 7010B, USA) by quantifying serum fatty acids by precursor ion scanning corresponding m / z values. The calibration curve of the standard showed a very good correlation between concentration and response. All calibration curves were reproducible. The concentration of each individual fatty acid was expressed as a percentage of the total area under the peak. Here, by collecting and analyzing serum biomarkers, by collecting venous blood samples from subjects and separating serum, then using a GC-MS / MS system for quantitative analysis of fatty acids. This method can provide objective biological evidence about the dietary intake of participants, helping to verify the accuracy of the questionnaire.

[0021] The mean intake of energy, nutrients and food groups calculated from the three-day weighed food records were used as a reference for dietary intake. Since most of the variables were not normally distributed, data were presented as medians and interquartile ranges. The mean difference between 3DWFRs and FFQ was calculated by the following formula: (FFQ-3DWFR) / 3DWFR*100%; in addition, intake values between both methods were compared using the Wilcoxon rank-sum test. The relative validity of the FFQ to assess energy, nutrient and food group intake was investigated by crude Spearman correlation coefficients, energy-adjusted correlation coefficients and attenuation-adjusted correlation coefficients. The residual method proposed by Willett et al. was used to calculate energy-adjusted variables and attenuation-adjusted correlation coefficients were calculated following the method proposed by Rosner et al. The relationship between fatty acid intake estimated from the FFQ and serum biomarkers was investigated by Spearman correlation analysis. In the present disclosure, items with coefficients above 0.30 were considered to show acceptable validity, with values between 0.30-0.39 indicating moderate validity, 0.40-0.59 indicating moderate validity and >0.60 indicating high validity. The agreement between FFQ and 3DWFRs was assessed by classifying participants into quintiles according to the distribution of energy, nutrients and food groups for each method. The percentage of athletes classified in the same, adjacent and extreme quintiles was estimated. The kappa statistic was also calculated, with weighted kappa values above 0.2 considered acceptable. The agreement between the two methods for energy and macronutrient intake estimates was also assessed using the Bland-Altman method by constructing scatter plots of the mean intake and the absolute difference in intake between the two methods. The line showing the mean difference in intake and the limits of agreement (LOA), defined as the mean difference ± 1.96 SD, are shown in the plots. In the test of reproducibility, Spearman correlation coefficients, energy-adjusted correlation coefficients and intraclass correlation coefficients (based on average ranking (k=2), absolute agreement, two-way mixed effects model) between the first and second FFQ were calculated. The following classification was used: poor reproducibility: <0.50, moderate reproducibility: 0.75-0.90 and excellent reproducibility: >0.9. P values lower than 0.05 were considered statistically significant. All statistical analyses were performed using IBM SPSS Statistics ver. 27. Since multiple statistical methods were used to assess the relative validity and reproducibility of the FFQ, including Spearman correlation coefficients, energy-adjusted correlation coefficients, attenuation-adjusted correlation coefficients, kappa statistic and Bland-Altman analysis. These methods allow a comprehensive assessment of the accuracy and agreement of the questionnaire, ensuring the reliability of the results.

[0022] Characteristics by sex, in the study participants, 52% were female athletes, with an average age of 20 years and an average BMI of 22.5 kg / m2 The athletes represented 11 different sports, with 47.1% being national-level athletes, 28.4% being first-level athletes, and 24.5% being second-level athletes. The average training experience was 8.5 years, and the average weekly training time was 28.4 hours. The baseline characteristics of the study participants, including gender, age, BMI, sport, and training experience, were analyzed. This information helped understand the representativeness of the study sample and provided background information for subsequent analysis.

[0023] 3DWFRs and the first FFQ relative validity of energy, nutrients, and food groups. For energy and nutrients, the FFQ overestimated the intake of most nutrients, except for fat, vitamin A, vitamin E, sodium, and iron. Based on the percentage difference, the intake difference between the two methods was relatively small for energy, carbohydrates, dietary fiber, cholesterol, vitamin A, vitamin E, calcium, potassium, magnesium, iron, and manganese (percentage difference <10%). However, significant differences were observed for the intake of vitamin B1, vitamin B2, niacin, vitamin C, vitamin D, sodium, magnesium, zinc, selenium, and copper (p<0.05), with percentage differences ranging from 23.13% for vitamin B2 to 905.81% for vitamin D. For food groups, the FFQ overestimated the intake of fruits, poultry and meat, dairy, and legumes and nuts, while underestimating the intake of grains and potatoes, vegetables, fish and seafood, eggs, and fats. Except for fish and seafood, significant differences were observed for the intake of food groups between the two methods, with percentage differences ranging from 10.91% to 146.19%.

[0024] This section presents the results of the comparison between the FFQ and the 3DWFRs in terms of energy, nutrient, and food group intake. The FFQ overestimated or underestimated the intake of most nutrients and food groups, which may be related to the design of the questionnaire and the bias in the participants' responses. By comparing the two methods, the accuracy of the FFQ can be evaluated, and the basis for further improvement can be provided.

[0025] Spearman's crude correlation coefficients (CCs) ranged from 0.3 to 0.5 for energy and most nutrients (83% of nutrient crude CCs were above 0.3). The median crude CC was 0.331, ranging from 0.219 for vitamin C to 0.568 for cholesterol. After energy adjustment, CCs decreased for most nutrients except dietary fiber and vitamin C. The median energy-adjusted CCs and deattenuated CCs for energy and nutrients were 0.226 and 0.234, respectively. For food groups, the Spearman's crude CCs for grains and potatoes, poultry and meat, eggs, and dairy products were above 0.3, with eggs showing the highest CC of 0.455. Vegetables and legumes and nuts showed relatively weak crude CCs of 0.193 and 0.187, respectively, while fats showed the lowest crude CC of -0.035. After energy adjustment, CCs decreased for most food groups except fruits. The median energy-adjusted CCs and deattenuated CCs for food groups were 0.198 and 0.320, respectively. This section presents the correlation analysis between FFQ and 3DWFRs. The accuracy of FFQ in estimating intakes of different nutrients and food groups was assessed by Spearman's correlation coefficients. The results showed that the correlations for most nutrients and food groups were in an acceptable range, but the correlations for certain items such as vegetables and legumes and nuts were lower, which might need further improvement in the design of the questionnaire or the accuracy of the reference method.

[0026] Fatty acids from dietary intake and serum levels of participants from a subsample (n = 54). The CCs between the first FFQ-estimated intakes of EPA, DHA, and EPA+DHA and the corresponding serum levels were 0.389 (95% CI: 0.007, 0.672), 0.520 (95% CI: 0.171, 0.753), and 0.463 (95% CI: 0.097, 0.718), respectively. This section presents the correlation between FFQ-estimated intakes of omega-3 polyunsaturated fatty acids (EPA, DHA, and EPA+DHA) and serum biomarkers. The results showed that the FFQ had a moderate degree of correlation in estimating intakes of these fatty acids, which indicated that the FFQ was reliable in recalling omega-3 polyunsaturated fatty acid-rich foods, although the correlation was slightly lower than some studies that specifically assessed omega-3 intakes.

[0027] Individual rank agreement results between 3DWFRs and the first FFQ based on quintiles of energy, nutrients, and food group distribution. For energy and nutrients, the median percentage of participants correctly classified in the same or adjacent quintile was 61% with an extreme misclassification rate of 2%. Misclassification rates were less than 10% for energy and most nutrients except for vitamin E (12%). High agreement (extreme quintile proportion <3%) was observed for energy, fat, carbohydrate, cholesterol, phosphorus, potassium, sodium, magnesium, zinc, selenium, and copper. For food groups, the median percentage of participants correctly classified in the same or adjacent quintile was 61% with an extreme misclassification rate of 3%. Fruits, poultry and meat, and eggs showed high agreement, while oils showed the poorest agreement (10%). Weighted kappa statistics indicated acceptable agreement (>0.2) for energy and most nutrients and food groups except for vitamin C, vegetables, and oils. This section assessed the agreement between the FFQ and 3DWFRs through quintile classification and weighted kappa statistics. Results showed high classification agreement for most nutrients and food groups, but lower agreement for certain items such as vitamin C, vegetables, and oils. This could be related to limitations in the questionnaire design or reference methods that need to be further improved for increased accuracy.

[0028] Bland-Altman analysis was used to assess the level of agreement by plotting the relationship between the mean and the difference in daily intakes between 3DWFRs and the first FFQ. Figure 3 Bland-Altman plots for energy and macronutrients are presented. In each plot, the solid black line represents the mean difference in intakes between the FFQ and 3DWFRs, while the two solid red lines represent the limits of agreement (LOA, mean difference ± 1.96 SD). For energy and macronutrients, most data points fell within the LOA, and most data points were close to the mean line, indicating an acceptable level of agreement between the two methods. Bland-Altman analysis data for energy, nutrients, and food groups, with the percentage of data points within the LOA for all nutrients and food groups exceeding 90%.

[0029] This section demonstrated the agreement between the FFQ and 3DWFRs through Bland-Altman analysis. Bland-Altman plots showed the mean difference and the limits of agreement in intakes between the two methods, with most data points falling within the limits of agreement, indicating good agreement between the FFQ and 3DWFRs, supporting the validity of the FFQ.

[0030] Reproducibility results for energy, nutrients, and food group intakes assessed by two FFQs. For energy and nutrients, the second FFQ underestimated the intake of most nutrients except for cholesterol, vitamin A, and calcium. The median crude CC was 0.574, ranging from 0.423 for vitamin C to 0.643 for vitamins B1 and B2. After energy adjustment, the CC values varied in direction for different nutrients, with a median energy-adjusted CC of 0.530. The median intraclass correlation coefficient (ICC) was 0.668, ranging from 0.558 (95% CI: 0.379, 0.697) for vitamin A to 0.763 (95% CI: 0.645, 0.845) for vitamin B1. For food groups, the second FFQ overestimated the intake of cereals and potatoes and fish and shellfish, while it underestimated the intake of vegetables, fruits, poultry and meat, and oils and fats. The median crude CC, energy-adjusted CC, and ICC for food groups were 0.681, 0.632, and 0.640, respectively. Legumes and nuts consistently showed the lowest CCs (0.436, 0.242, and 0.371, respectively) across all analyses. This section presents the reproducibility analysis between the two FFQs. The agreement between the two FFQs in the assessment of energy, nutrient, and food group intakes was evaluated by Spearman’s correlation coefficient and intraclass correlation coefficient. The results showed that the reproducibility was good for most nutrients and food groups, but it was low for certain items such as legumes and nuts, which could be related to the questionnaire design or inconsistency in the participants’ responses.

[0031] The present disclosure created a short FFQ containing 29 items and one open-ended question to maximize the athletes’ cooperation and compliance, thus enhancing the practicality and reliability of the FFQ. The FFQ showed acceptable validity and reproducibility in most nutrients and food groups, making it an effective tool to assess habitual nutrient intake during specific training periods in young athletes.

[0032] In the present disclosure, the weighed food records (WFRs) method was used as the primary reference method to assess the relative validity of the FFQ. Although the process of weighing food is more tedious, unlike the traditional 24-hour dietary recall method (24HDR), WFRs do not rely on the accuracy of memory, provide data closest to actual intake, and can avoid the generation of homogeneity errors in the FFQ. Therefore, it provides a more robust approach to minimize the likelihood of overestimating the relative validity, resulting in a reasonable validity result. Furthermore, the fatty acid intake, particularly omega-3 polyunsaturated fatty acids, of athletes has drawn particular attention in the present disclosure, as these fatty acids have been reported to be associated with inflammation levels and performance in athletes. Since the nutritional analysis software used to analyze the 3D WFRs data cannot automatically aggregate the daily intake of fatty acids, serum biomarker measurements were introduced as an alternative reference method to assess the relative validity of fatty acid intake. Although assessing nutrient intake through serum biomarkers can be affected by individual metabolic variations, it can effectively alleviate the inherent limitations of traditional dietary intake assessment methods, such as recall bias or subjectivity. To the best of the present disclosure's knowledge, this is the first time that WFRs and biomarkers have been used simultaneously as reference methods to assess the validity of FFQs for professional athletes.

[0033] A previous study examined the correlation of a FFQ containing 138 foods, 20 beverages, and 14 condiments with Japanese university athletes, which showed a median of 0.29 and 0.36 for the crude correlation coefficients (CCs) of energy and nutrients for male and female athletes, respectively. Another similar study was conducted in Japanese athletes, which demonstrated a median of 0.407 (ranging from 0.222 for dietary fiber to 0.550 for carbohydrates) for the crude CCs of nutrients using a FFQ containing 62 food items and four supplement questions. In comparison with the present disclosure, these studies used longer FFQs. Furthermore, a meta-analysis assessing the relative validity of FFQs in healthy adults indicated that longer FFQs covering a wider range of food items tended to show stronger correlations with the reference method. However, the final analysis results of the present disclosure showed that the short FFQ of the present disclosure provided CCs for energy and most nutrients comparable to or slightly lower than those observed for longer FFQs. In the present disclosure, 83% of the nutrients showed crude CCs exceeding 0.3, with a median CC of 0.331. Among them, energy and macronutrients directly related to athlete health and performance, such as protein, carbohydrates, fats, and cholesterol, as well as micronutrients of interest, including vitamin D, calcium, and iron, achieved acceptable to moderate levels of validity (CCs = 0.315-0.568). However, the energy-adjusted CCs for most nutrients decreased, which can be attributed to the large differences in energy intake between individuals and systematic errors in energy intake in the FFQ and 3D WFRs.

[0034] Furthermore, omega-3 polyunsaturated fatty acids (EPA, DHA, and EPA+DHA) showed significant correlations with the corresponding serum biomarkers as assessed by FFQ. Previous studies using Sublette et al.’s 214-item omega-3 FFQ reported CCs of 0.34, 0.40, and 0.44 for EPA, DHA, and EPA+DHA intake, respectively, taking into account blood EPA, DHA, and omega-3 index. Another study developed a short FFQ to assess EPA, DHA, and EPA+DHA intake, which resulted in CCs of 0.48, 0.73, and 0.73 with the corresponding red blood cell biomarkers, respectively. In the present disclosure, the CCs between EPA, DHA, and EPA+DHA intake and the corresponding serum biomarkers were 0.389, 0.52, and 0.463, respectively, which were comparable or slightly lower than previous studies. Although the FFQ used in the present disclosure was not specifically designed to assess omega-3 polyunsaturated fatty acid intake, it showed acceptable to moderate validity for EPA and DHA, indicating that the FFQ was reliable in recalling omega-3 polyunsaturated fatty acid-rich foods, mainly freshwater fish and seafood. However, the validity coefficient for the “fish and shellfish” food group was lower (0.287) for 3DWFRs, possibly because the random 3-day weighing method had limitations in accurately capturing the average intake of these foods, as they are not consumed every day. In contrast, exogenous serum biomarkers can more reliably reflect recent nutrient intake.

[0035] In fact, the CCs of food groups in the present disclosure (median CC of 0.292) were lower than previous studies compared to 3DWFR, which could be attributed to factors such as questionnaire length, reference method, food group classification, and cultural dietary differences. Nonetheless, acceptable CCs were found for food groups such as cereals and potatoes, poultry and meat, eggs, and dairy products. These foods were frequently consumed by athletes, with high daily intakes and stable consumption frequencies, reducing the likelihood of recall bias, resulting in higher correlations. In contrast, vegetables and legumes and nuts had significantly lower correlations, and energy adjustment failed to improve the CCs of these two food groups. The observed lower CC values for vegetables could be attributed to athletes subjectively increasing their consumption of these foods, which are perceived as “healthy”, during the weighing period. To the best of the present disclosure’s knowledge, athletes generally underestimate their vegetable intake, with their daily intakes often below recommended levels. The lower CC values for legumes and nuts could be due to their sporadic and random intake patterns, making it difficult for the 3-day weighing method to accurately capture the consumption of this food group. Future studies could require more reliable reference methods, such as increasing the weighing frequency, to obtain accurate intake data for vegetables and legumes and nuts. Among the food groups, fats showed the lowest correlation, with both the crude CC and the energy-adjusted CC being negative values. This observation is consistent with previous findings in Japanese athletes and could be attributed to the underestimation of fat amounts in Chinese stir-fried dishes or frying by the FFQ. Therefore, adjustments are needed when using this FFQ to assess fat intake, including items related to cooking methods.

[0036] While the results of the correlation analysis suggest that the FFQ has acceptable validity, it is clear that it tends to overestimate the intake of most nutrients and shows significant bias in estimating food group intakes, consistent with many previous studies. The bias in absolute intake estimates is an inherent limitation of FFQs and needs to be considered carefully when using FFQs to assess individual-level nutrient or food group intakes. However, the main role of FFQs is to rank nutrient intakes in a population so that subjects can be reasonably classified, which is essential for subsequent studies to investigate the relationship between dietary patterns and athletic performance. Therefore, it is sensible to integrate the assessment of agreement into the framework of validity assessment. The quintile cross-classification agreement, weighted kappa coefficient, and Bland-Altman plots were used in the present disclosure to assess the agreement between the FFQ of the present disclosure and the reference method. The median percentage of nutrients and food groups classified in the same or adjacent quintiles was 61%, while the median misclassification rate was 2% for energy and nutrient components and 3% for food groups. These findings are very similar to the results of the validation study of the FFQ conducted in Japanese athletes, indicating good quintile agreement with the reference method. In addition, the weighted kappa values fell within the acceptable threshold range (0.2-0.6) for most nutrients and food groups. The Bland-Altman plots showed that the data points for most macronutrients and food groups fell within the limits of agreement (LOA), close to the average intake, with less than 10% of the data points distributed outside the LOA, further confirming the high agreement between the FFQ and the 3DWFRs.

[0037] Finally, the reproducibility of the one-month interval between the two FFQs was tested. In the present disclosure, the reproducibility CCs and ICCs for most nutrients and food groups ranged from 0.5 to 0.8, comparable to previous studies conducted in athletes and other healthy adult populations. In addition, in terms of absolute intake, the second FFQ underestimated the intake of most nutrients and food groups compared to the first FFQ. This difference can be attributed to the 3DWFRs conducted between the two FFQs, which provided participants with a more direct understanding of food portion sizes, allowing them to adjust their questionnaire answers more accurately. This is also why the present disclosure used data from the first FFQ in the relative validity assessment to more accurately reflect the results of the validity test.

[0038] The brevity of the FFQ is an advantage of the present disclosure. Shortening the questionnaire completion time can maximize the support of athletes and coaches, thereby increasing the likelihood of obtaining a sufficient sample size. In addition, the final validity and reproducibility results of the present disclosure are comparable to other similar studies, confirming the effectiveness of the short FFQ of the present disclosure in terms of project selection and food list composition.

[0039] Therefore, the beneficial effects of this disclosure include: First, it highlights the effectiveness and reproducibility of the short FFQ in assessing athletes' dietary intake, pointing out its potential for practical application. Second, it discusses the advantages and limitations of weighed food records and serum biomarkers as reference methods, and their role in validating the effectiveness of the FFQ. Next, it compares the results of this disclosure with those of other studies, highlighting the performance of the short FFQ in assessing nutrient and food group intake, and discusses the correlation of ω-3 polyunsaturated fatty acid intake assessment. Furthermore, it analyzes the reasons for the low correlation between food groups and proposes suggestions for future research.

[0040] This disclosure validates the first short FFQ developed for professional athletes. Including items related to protein powder, sports drinks, and specialty nutritional supplements in the short FFQ is crucial for athletes, as these items significantly contribute to macronutrient and micronutrient intake. The short FFQ demonstrates acceptable validity and reproducibility across most nutrient and food groups, and good consistency with reference methods. It can serve as a useful tool for assessing habitual intakes of athletes during specific training periods, including macronutrients and key nutrients such as vitamin D and omega-3 polyunsaturated fatty acids. Furthermore, it will become a key tool for further exploring the complex relationship between dietary patterns and athlete health and / or performance. Therefore, the short FFQ of this disclosure has good validity and reproducibility in assessing athletes' dietary intake and can serve as an important tool for studying the relationship between athletes' dietary patterns and health and performance.

[0041] Figure 1 The Chinese explanations of the English and Chinese terms are as follows:

[0042] Subject recruitment: Recruiting test subjects

[0043] 143 athletes registered; Initial lab visit; Failed to meet inclusion criteria.

[0044] 102 athletes included: 102 athletes were included.

[0045] 41 athletes excluded: 41 athletes were excluded.

[0046] 97 athletes completed 3DWFRs (3-day weigh-in food records).

[0047] 102 athletes completed two Food Frequency Questionnaires (FFQs).

[0048] 54 athletes provided blood samples. Biomarker validity analysis; Relative validity analysis; Reproducibility analysis.

[0049] Figure 2 The Chinese explanations of the English and Chinese terms are as follows:

[0050] Validity, 3DWFRs (Three-Day Weighing Food Records), FFQ1 (First Food Frequency Questionnaire), FFQ2 (Second Food Frequency Questionnaire), Weekday, Weekend, WFR1 (First Day Weighing Food Record), WFR2 (Second Day Weighing Food Record), WFR3 (Third Day Weighing Food Record), Baseline, Blood Sample, Reproducibility, Biomarkers.

[0051] Figure 3 The Chinese explanations of the English and Chinese terms are as follows:

[0052] Difference in energy intake, FFQ1 (First Food Frequency Questionnaire), 3DW (Three-Day Food Weighting Record), kcal / d (kcal / day), Mean (mean value), Mean+1.96SD (mean value plus 1.96 times the standard deviation), Mean-1.96SD (mean value minus 1.96 times the standard deviation).

[0053] Below is an example of a meal frequency questionnaire.

[0054]

[0055]

[0056] According to one or more embodiments, a method for investigating the diet of athletes. The method includes formulating a FFQ questionnaire for a group of athletes, and using the results of the questionnaire to extract and name the dietary patterns of the group of athletes, to investigate the characteristics and regularities of their dietary intake, and to determine the effects of the dietary patterns on nutritional status, thereby providing guidance for the scientific and reasonable diet of athletes.

[0057] Further, the research method also includes: for a group of athletes, modifying the FFQ questionnaire reported in the literature for Chinese population, formulating a FFQ questionnaire for the group of athletes, and using rank test, Spearman correlation coefficient, Cronbach's alpha, and intra-class correlation coefficient (ICC) as indexes to complete the reliability and validity test of the questionnaire. The FFQ questionnaire that has passed the reliability and validity test is applied to the dietary investigation of professional athletes, the results of the investigation are subjected to factor analysis to extract the dietary patterns, and the dietary patterns are named according to the composition of the food in each component, while using chi-square test, logistic regression analysis, one-way ANOVA, and multiple linear regression to understand the effects of different dietary patterns on different nutritional indicators. The athletes involved are from 4 training bases, a total of 263, including 105 males and 158 females.

[0058] Results: 1. The FFQ questionnaire has acceptable reliability and validity. The intakes of tubers, vegetables, seafood and eggs are underestimated in the FFQ questionnaire, while the intakes of fruits, poultry and dairy products are overestimated. In addition, the intakes of energy, fat, cholesterol, total vitamin A, sodium and iron are also underestimated, while the intakes of protein, carbohydrate, insoluble dietary fiber, thiamin, riboflavin, niacin, total vitamin E, vitamin D, calcium, phosphorus, potassium, magnesium, zinc, selenium and copper are overestimated. 2. The results of factor analysis show that the dietary patterns of the athletes can be classified into 5 types, namely, the Mediterranean-like dietary pattern, the Western-like dietary pattern, the fast-food-like dietary pattern, the snack-like dietary pattern and the supplemental high-carbohydrate dietary pattern. The proportion of female athletes in the snack-like dietary pattern is significantly higher than that in the Western-like dietary pattern and the supplemental high-carbohydrate dietary pattern (p<0.01). 3. The factor scores of different dietary patterns are divided into Q1 group (low score group) and Q2 group (high score group) from low to high. The results of logistic regression analysis show that the supplemental high-carbohydrate dietary pattern is a protective factor for abnormal body mass index (BMI) (p<0.05), and the OR value of Q2 group is 0.41 (95% Cl: 0.21-0.81) compared with Q1 group. In addition, the Mediterranean-like dietary pattern is a protective factor for abnormal serum ferritin (SF) (p<0.05), and the OR value of Q2 group is 0.47 (95% Cl: 0.24-0.90) compared with Q1 group, without excluding gender, age and athlete level. 4. The results of one-way ANOVA show that the BMI of the Western-like dietary pattern is lower than that of the supplemental high-carbohydrate dietary pattern in the male adolescent athletes (p<0.05). 5. The results of multiple linear regression analysis show that the Western-like dietary pattern can significantly positively affect the blood urea nitrogen (BUN) in the male adolescent athletes, and the influence coefficient is 0.638 (p<0.05).In the female group, the Mediterranean-like dietary pattern could significantly positively affect FFM and negatively affect TG, with the effect coefficient of 4.634 (p<0.05) and -0.208 (p<0.05), respectively; the fast-food-like dietary pattern could significantly negatively affect BUN and positively affect HDL, with the effect coefficient of -0.891 (p<0.01) and 2.709 (p<0.001), respectively; the supplementary high-carbohydrate dietary pattern could significantly positively affect HB, with the effect coefficient of 0.997 (p<0.05); the western-like dietary pattern could significantly positively affect TC and negatively affect TG, with the effect coefficient of 0.388 (p<0.05) and -0.197 (p<0.05), respectively, but the Mediterranean-like dietary pattern had a stronger negative effect on TG than the western-like dietary pattern. In the male group, the western-like dietary pattern could significantly negatively affect T, with the effect coefficient of -49.854 (p<0.05). In the female group, the Mediterranean-like dietary pattern could significantly positively affect FFM, with the effect coefficient of 2.129 (p<0.05); the snack-like dietary pattern could significantly negatively affect BUN, with the effect coefficient of -0.652 (p<0.05); the western-like dietary pattern could significantly negatively affect HDL, with the effect coefficient of -0.87 (p<0.05).

[0059] The research method of the embodiments of the present disclosure specifically comprises:

[0060] (1) Extraction and naming of dietary patterns

[0061] The FFQ questionnaire that has passed the reliability and validity test is applied to the dietary survey of professional athletes to obtain dietary intake information, and the dietary patterns of the athlete group are extracted through exploratory factor analysis, and the characteristics of the food composition in each pattern are named respectively, and the differences in dietary patterns of different genders, age groups (adolescent group, adult group) and athlete levels (athlete level and above, first level, second level and below) are discussed.

[0062] (2) Analysis of the correlation between dietary patterns and nutritional indicators

[0063] The correlation between different dietary patterns and different nutritional indicators was analyzed by different statistical methods. In terms of nutritional indicators, two parts were included, body measurement indicators and blood test indicators. The former included body mass index (BMI), fat percentage (FAT%), and fat-free mass (FFM), and the latter included testosterone (T), blood urea nitrogen (BUN), hemoglobin (HB), ferritin (SF), and blood lipids (including high-density lipoprotein HDL, low-density lipoprotein LDL, cholesterol TC, and triglycerides TG). To exclude the influence of different training periods, these indicators were tested and collected regularly within 1 year, and the average value within 1 year was taken as the final correlation study data.

[0064] (3) Extraction of dietary patterns

[0065] After KMO (Kaiser-Meyer-Olkin) and Bartlett's sphericity test, it was determined that the data was suitable for factor analysis. The principal component extraction method in factor analysis was used to analyze the dietary patterns. At the same time, in order to make each factor have a clear professional meaning, the analysis of dietary patterns also needs to be combined with the final determination of the broken stone chart, eigenvalue extraction, food combination rationality, variance contribution rate and professional knowledge, and after calculating the factor score of individual dietary pattern, the mode with the highest factor score is taken as the dietary pattern with the highest individual compliance degree. At the same time, in order to understand the differences of different gender, age and athlete level in dietary pattern, the results of dietary pattern analysis were subjected to chi-square test.

[0066] (4) Correlation analysis between dietary patterns and nutritional indicators

[0067] The statistical methods used in the correlation analysis between dietary patterns and nutritional indicators included: ① BMI, HB, SF, HDL, LDL, TC and TG were used as categorical variables and chi-square test was performed to determine whether the abnormality of each indicator was different under different dietary patterns; ② According to the median of factor scores, the scores of each dietary pattern were divided into Q1 group (low score group) and Q2 group (high score group) from low to high. Q2 group represented higher adherence to the dietary pattern, and binary logistic regression analysis was performed with normal and abnormal conditions of the classification indicators (BMI, HB, SF, HDL, LDL, TC and TG) as dependent variables and dietary patterns as independent variables after adjusting for gender, age and athlete level; ③ All indicators can be used as continuous numerical variables and ANOVA was performed to determine whether there is a difference between indicators under different dietary patterns; ④ Linear regression was performed with factor scores of different dietary patterns as independent variables and numerical values of each indicator as dependent variables to determine the trend of different dietary patterns affecting different indicators in different groups of subjects. All statistical analyses were completed in SPSS22.0 software, and the test level a = 0.05.

[0068] All subjects (n = 263) received FFQ questionnaire survey. Exploratory factor analysis was performed on the questionnaire results, KMO = 0.793, Bartlett's sphericity test chi-square value was 2335.326, p < 0.001, indicating that the data was suitable for factor analysis. Principal component analysis was used to extract factors, and according to the scree plot and eigenvalue greater than 1.3, five factors were finally extracted, and the cumulative variance contribution rate after orthogonal rotation was 51.155%.

[0069] The principal component analysis was used to reduce the dimension of 26 indexes, and 5 common factors were finally selected. The component matrix was obtained, and the different foods contained in each common factor were classified according to the standard of factor loading absolute value ≥ 0.40. On this basis, the composition of foods contained in different patterns was used to name the dietary patterns. According to the content of each factor, the first principal component included light-colored vegetables, red-orange-purple vegetables, whole grains, fungi and algae, green vegetables, potatoes and starch products, poultry, and river fish; the second principal component included green vegetables, livestock meat, poultry, eggs, fresh fruits, milk and dairy products, desserts, rice, and stuffed rice; the third principal component included eggs, puffed food, desserts, sweet drinks, instant food, river fish, rice, and seafood; the fourth principal component included fungi and algae, seafood, nuts, candied fruits, soybeans and products, and mixed beans; and the fifth principal component included sports drinks, refined rice, protein powder, and stuffed rice. Therefore, the first principal component dietary pattern (dietary pattern 1) was named as the Mediterranean-like dietary pattern, the second principal component dietary pattern (dietary pattern 2) was named as the Western-like dietary pattern, the third principal component dietary pattern (dietary pattern 3) was named as the fast-food-like dietary pattern, the fourth principal component dietary pattern (dietary pattern 4) was named as the snack-like dietary pattern, and the fifth principal component dietary pattern (dietary pattern 5) was named as the supplemental high-carbohydrate dietary pattern.

[0070] According to the results of exploratory factor analysis, the K-means clustering analysis method was used to cluster the factor scores of 26 types of food of 263 athletes. The factor scores of the top 5 common factors of each athlete were clustered, the number of clusters was selected as 5, and finally the number of people in each group, the difference between groups, and the significance level were obtained. Group 1 had 3 athletes, accounting for 1.14%, group 2 had 91 athletes, accounting for 34.6%, group 3 had 3 athletes, accounting for 1.14%, group 4 had 145 athletes, accounting for 55.13%, and group 5 had 21 athletes, accounting for 7.99%. The analysis of variance showed that there were differences between groups, p<0.001.

[0071] The gender was divided into male and female groups, the age was divided into two groups with 18 years as the boundary, and the athletes were divided into three groups of master level and above, first level, and second level and below. The chi-square test was performed on the five dietary patterns, and the results showed that there were statistically significant differences between the dietary patterns in different genders, and the proportion of women in the snack-like dietary pattern was significantly higher than that in the Western-like dietary pattern and the supplemental high-carbohydrate dietary pattern (p<0.01) compared with men.

[0072] Among them, dietary pattern 1 = Mediterranean-like dietary pattern, dietary pattern 2 = Western-like dietary pattern, dietary pattern 3 = fast-food-like dietary pattern, dietary pattern 4 = snack-like dietary pattern, and dietary pattern 5 = supplemental high-carbohydrate dietary pattern

[0073] Further, the median of each dietary pattern factor score was used as a cut-off point to divide the subjects into Q1 (low score group) and Q2 (high score group) for each dietary pattern. The normal and abnormal conditions of BMI, HB, SF, HDL, LDL, TC, and TG were used as dependent variables, and the dietary pattern was used as an independent variable. After adjusting for gender, age, and athlete level, a binary logistic regression model was used to analyze the relationship between the normal / abnormal conditions of the above indicators and the dietary pattern. In the supplemental high-carbohydrate dietary pattern, the abnormal occurrence of BMI was affected by the factor score, and the OR value of model 2 in the Q2 group was 0.41 (95% Cl: 0.21-0.81) compared to the Q1 group, indicating that after excluding confounding factors, the supplemental high-carbohydrate dietary pattern was a protective factor for the abnormal occurrence of BMI. In addition, in the Mediterranean-like dietary pattern, the abnormal occurrence of SF was affected by the factor score without excluding the confounding factors, and the OR value of model 1 in the Q2 group was 0.47 (95% Cl: 0.24-0.90) compared to the Q1 group, indicating that the Mediterranean-like dietary pattern was a protective factor for the abnormal condition of SF in the crude model.

[0074] The subjects were also grouped by gender and age (adolescent males, adolescent females, adult males, and adult females), and one-way ANOVA was used to analyze the differences in dietary patterns and BMI, FAT%, FFM, T, BUN, HB, SF, HDL, LDL, TC, and TG between different groups. The results showed that only in the adolescent male subjects, the BMI of the Western-like dietary pattern was lower than that of the supplemental high-carbohydrate dietary pattern, and the difference was statistically significant.

[0075] According to the grouping rules in the above one-way ANOVA, the confounding effects of athlete level were excluded, and the factor scores of different dietary patterns were used as independent variables, and the values of each indicator in different groups were used as dependent variables for linear regression to determine the trend of the effect of dietary patterns on the size of each nutritional indicator.

[0076] In adolescent males, the Western-like dietary pattern can significantly positively affect BUN, with an influence coefficient of 0.638 (p<0.05), indicating that an increase of 1 in the factor score of this pattern will increase BUN by 0.638.

[0077] In the female adolescents, the Mediterranean-like dietary pattern could significantly positively affect FFM, and could significantly negatively affect TG, with the effect coefficient of 4.634, p<0.05, indicating that the factor score of the pattern increased by 1, FFM increased by 4.634, and the effect coefficient of TG was -0.208, p<0.05, indicating that the factor score of the pattern increased by 1, TG decreased by 0.208; the fast food-like dietary pattern could significantly negatively affect BUN, and positively affect HDL, with the effect coefficient of BUN being -0.891, p<0.01, indicating that the factor score of the pattern increased by 1, BUN decreased by 0.891, and the effect coefficient of HDL was 2.709, p<0.001, indicating that the factor score of the pattern increased by 1, HDL increased by 2.709; the supplementary high-carbohydrate dietary pattern could significantly positively affect HB, with the effect coefficient being 0.997, p<0.05, indicating that the factor score of the pattern increased by 1, HB increased by 0.997; the western-like dietary pattern could significantly positively increase TC, and could significantly negatively affect TG, with the effect coefficient of TC being 0.388, p<0.05, indicating that the factor score of the pattern increased by 1, TC increased by 0.388, and the effect coefficient of TG was -0.197, p<0.05, indicating that the factor score of the pattern increased by 1, TG decreased by 0.197. Although both the Mediterranean-like dietary pattern and the western-like dietary pattern could significantly negatively affect TG, the standardized regression coefficient of the Mediterranean-like dietary pattern was -0.354, which was greater than the standardized coefficient -0.364 of the western-like dietary pattern, so compared with the western-like dietary pattern, the Mediterranean-like dietary pattern had stronger negative effect on TG in the population.

[0078] In the male adults, the western-like dietary pattern could significantly negatively affect T, with the effect coefficient being -49.854, p<0.05, indicating that the factor score of the pattern increased by 1, T decreased by 49.854.

[0079] In the female adults, the Mediterranean-like dietary pattern could significantly positively affect FFM, with the effect coefficient being 2.129, p<0.05, indicating that the factor score of the pattern increased by 1, FFM increased by 2.129; the snack-like dietary pattern could significantly negatively affect BUN, with the effect coefficient being -0.652, p<0.05, indicating that the factor score of the pattern increased by 1, BUN decreased by 0.652; the western-like dietary pattern could significantly negatively affect HDL, with the effect coefficient being -0.87, p<0.05, indicating that the factor score of the pattern increased by 1, HDL decreased by 0.87.

[0080] Based on the literature review and the scatter plot and variance contribution rate of the statistical results, five dietary patterns were finally determined and named according to the food composition and characteristics of each dietary pattern. In the first principal component, the food composition was mainly composed of vegetables, cereals and tubers, aquatic products and poultry meat, which was similar to the food composition of the Mediterranean dietary pattern, and was named as "Mediterranean-like dietary pattern". In the second principal component, the food composition was mainly composed of poultry and livestock meat, eggs, milk, desserts and refined rice and flour, which was similar to the Western dietary pattern, and was named as "Western-like dietary pattern". In the third principal component, the food composition was mainly composed of instant food, desserts and snacks, and aquatic products, which was similar to fast food, and was named as "fast food-like dietary pattern". In the fourth principal component, the food composition was mainly composed of nuts, candied fruits, soybeans and products, and aquatic products, which was similar to snack foods, and was named as "snack-like dietary pattern". In the fifth principal component, the food composition was mainly composed of sports drinks, protein powder and refined rice and flour, which was consistent with the characteristics of high protein, high carbohydrate and low fiber, and was named as "supplemental high-carbohydrate dietary pattern". This is the first time to analyze the characteristics of athletes' dietary intake from the perspective of overall dietary pattern. The five dietary patterns extracted each have their own characteristics, and basically reflect the overall picture of athletes' dietary characteristics. In addition, from the number of people in different dietary pattern groups, the number of people in the Mediterranean-like dietary pattern and the fast food-like dietary pattern is the smallest, indicating that the typical healthy and non-meal dietary patterns are not representative in professional athletes, which may be related to the traditional nutritional concepts and closed management of athletes. The snack-like dietary pattern, the Western-like dietary pattern and the high-carbohydrate dietary pattern containing more sports nutrition supplements are the mainstream among athletes, and after comparing between genders, it was found that compared with the snack-like dietary pattern, the Western-like dietary pattern and the supplemental high-carbohydrate dietary pattern both have a higher proportion of men than women, which is more consistent with the actual situation.

[0081] In addition to the correlation analysis of dietary patterns and index abnormalities, the dietary pattern factor scores were divided into low score group (Q1) and high score group (Q2) by median, so as to explore the relationship between the two groups of each pattern and the possibility of abnormality of each index. The results showed that after excluding the confounding effects of gender, age and athlete level, the supplementary high-carbohydrate dietary pattern may be a protective factor for abnormal BMI, including underweight and overweight, indicating that the higher the adherence of the subjects in the high-carbohydrate dietary pattern group, the less likely the abnormality of BMI. This is an interesting finding, although more research evidence from the general population shows that high-protein dietary patterns can help reduce body weight, increase lean body mass ratio and maintain appropriate BMI, but there is little evidence to support high-carbohydrate intake patterns in maintaining normal BMI. This conclusion may be closely related to the special energy metabolism characteristics of the athlete group. Athletes engage in high-intensity training, and whether it is glycolysis or aerobic oxidation, the energy supply is highly dependent on carbohydrates, making high-carbohydrate intake patterns necessary for sports consumption, which may be different from the general population. Therefore, it is necessary to conduct dietary pattern research in specific populations. Unfortunately, this disclosure failed to effectively analyze body composition using more reliable instruments such as DXA, thereby revealing more relationships between dietary patterns and body composition. In addition, this disclosure found that the Mediterranean-like dietary pattern was a protective factor for abnormal SF, although the effects of gender, age and other confounding factors were not excluded, but the conclusion still has certain guiding significance. SF abnormalities are common in the athlete group, and iron deficiency can easily affect hemoglobin production, thereby affecting oxygen carrying capacity during exercise and affecting athletic performance. Although red meat is the main source of dietary iron, the Western-like dietary pattern rich in red meat does not provide a favorable protective factor for SF phenotypic outcomes, while the Mediterranean-like dietary pattern with balanced and comprehensive nutrient intake, low red meat intake and mainly vegetable and whole grain food sources shows a friendly outcome for SF, which may be related to sufficient micronutrients or the interaction between nutrients, which is more conducive to iron absorption. Mediterranean diet is a globally recognized healthy dietary pattern, and its health benefits have been gradually discovered, especially the protective effect on the cardiovascular system has been continuously confirmed. However, there are few reports on the relationship between Mediterranean pattern and iron nutrition and its application in the athlete group. This disclosure is limited by the number of samples, and the conclusion on the relationship between Mediterranean-like dietary pattern and iron nutrition is still lack of sufficient convincing power, but this result is worth further attention and in-depth discussion to explore more evidence between dietary patterns and aerobic capacity.

[0082] In order to further reveal the differences of nutritional indicators in different dietary patterns in different athlete subgroups, the subjects were divided into underage men, underage women, adult men and adult women for difference comparison by using stratified analysis method considering the influence of age and gender. The results showed that in underage male subjects, the BMI value of subjects in western-style dietary pattern was significantly lower than that in high-carbohydrate supplement pattern. It showed that in underage male athletes, it was more necessary to rely on high-protein dietary pattern for the purpose of weight loss or reducing BMI.

[0083] Meanwhile, in order to explore the trend influence of different dietary patterns on the values of each functional index in underage males, underage females, adult males and adult females, in linear regression analysis with the factor score of the dietary pattern as the independent variable, the value of each index as the dependent variable and excluding the influence of the athlete grade, it is found that the Mediterranean-like dietary pattern can positively affect FFM in underage females and adult females, that is, in these two populations, the higher the adherence to the Mediterranean-like dietary pattern, the higher the level of FFM; in addition, the pattern can also significantly negatively affect TG in underage females, that is, the higher the adherence to the Mediterranean-like dietary pattern in this population, the lower the level of TG. It can be seen that the Mediterranean-like dietary pattern has a beneficial effect on the FFM of female athletes and the TG level of underage female athletes, which is consistent with the conclusions of most previous studies in the general population, indicating that the Mediterranean-like dietary pattern also has certain application potential in the athlete population. As for the Western-like dietary pattern, the higher the adherence to the pattern in underage males, the higher the level of BUN, and the higher the adherence to the pattern in adult males, the lower the level of T; in addition, the higher the adherence to the pattern in underage females, the higher the level of TC, and the higher the adherence to the pattern in adult females, the lower the level of HDL. It can be seen that the Western-like dietary pattern has an adverse effect on the functional index of male athletes, especially the T value, and the blood lipid level of female athletes, especially the high BUN, which may cause hyperuricemia and gout. Since the Western-like dietary pattern is the mainstream dietary pattern in athletes, it is worth paying attention to and further studying. As for the fast food-like dietary pattern, the higher the adherence to the pattern in underage females, the lower the level of BUN, and the higher the level of HDL, indicating that the fast food-like dietary pattern has a certain protective effect on the BUN and HDL of underage female athletes, which may be related to the fact that the protein source in this pattern is mainly eggs and aquatic products, thereby avoiding the intake of too much saturated fat. As for the snack-like dietary pattern, the higher the adherence to the pattern in adult females, the lower the level of BUN, which may be related to the fact that this dietary pattern cannot provide sufficient high-quality protein, and athletes should be cautious when adopting this dietary pattern. As for the supplemental high-carbohydrate dietary pattern, the higher the adherence to the pattern in underage females, the higher the level of HB. It can be seen that the typical high-carbohydrate dietary pattern of athletes has a protective effect on the HB level of underage females. Previous interventions to low HB in athletes often considered high-protein and iron nutrition supplementation, and were more inclined to the Western-like pattern. The present disclosure shows different results, indicating that the high-carbohydrate pattern rich in protein seems to have more advantages. The influence of carbohydrates and their interaction with other nutrients on the HB of athletes is worth further studying.

[0084] Therefore, the conclusions of the embodiments of the present disclosure include:

[0085] (1) The dietary patterns of professional athletes can be classified into five categories: Mediterranean-like, Western-like, fast-food-like, snack-like, and supplemental high-carbohydrate. These categories represent the dietary characteristics of athletes in the city.

[0086] (2) The supplemental high-carbohydrate dietary pattern is a protective factor for abnormal BMI in athletes, and the Mediterranean-like dietary pattern is a protective factor for SF abnormalities.

[0087] (3) In the underage male group, the BMI value of the Western-like dietary pattern is lower than that of the supplemental high-carbohydrate dietary pattern.

[0088] (4) The Mediterranean-like dietary pattern has a beneficial effect on the FFM of female athletes and the TG level of underage female athletes. The Western-like dietary pattern has an adverse effect on the functional indicators of male athletes, especially the T value, and the blood lipid levels of female athletes. The fast-food-like dietary pattern has a certain beneficial effect on the BUN and HDL levels of underage females. The snack-like dietary pattern only affects the BUN level of adult females. The supplemental high-carbohydrate dietary pattern only has a beneficial effect on the HB level of underage females.

[0089] According to one or more embodiments, a method of analyzing the composition and metabolism of the gut microbiome of athletes. Due to the growing interest in the bidirectional interaction between exercise and the gut microbiome, however, the specific effects of different training modalities on the gut microbiome of professional athletes have not been fully understood. The present disclosure aims to explore the differences in the gut microbiome between Chinese elite athletes from the road cycling, handball, shooting, and weightlifting disciplines, and to compare them with a control group of university students. By performing high-throughput metagenomic sequencing and metabolomic analysis on fecal samples from 69 professional athletes and 22 control subjects, and by collecting dietary intake data through a food frequency questionnaire, the present disclosure found that significant differences in the composition, function, and metabolic phenotype of the gut microbiome were observed among athletes. Road cycling athletes had the highest relative abundance of Prevotella, while handball and shooting athletes were dominated by Bacteroides. Weightlifting athletes exhibited the highest microbial diversity and were rich in butyrate-producing bacteria, including Clostridium, Ruminococcus, Dorea, and Faecalibacterium prausnitzii. They also exhibited the highest potential in microbial functions related to amino acid metabolism. Notably, the characteristic flora of weightlifting athletes were positively correlated with bile acid metabolites, suggesting a potential link to the bile acid regulation pathways affected by long-term strength training.

[0090] The embodiments of the present disclosure relate to a method for establishing and analyzing an athlete's gut microbiome model. The embodiments of the present disclosure study the independent effects of exercise on the composition and function of the gut microbiome. Due to long-term high-intensity training programs, professional athletes exhibit unique physiological adaptations and metabolic characteristics compared to the general population, which makes them an ideal model for studying the effects of exercise on the gut microbiome. Understanding the specific characteristics of the gut microbiome of athletes and their potential role in athletic performance is crucial for developing interventions aimed at improving athlete performance and recovery. These knowledge also provides insights into the broad role of the gut microbiome in human health and disease, and possible directions for future therapeutic strategies.

[0091] Current research has not fully elucidated the characteristics of the gut microbiome of professional athletes and the specific effects of different training programs. In fact, strength and endurance exercise elicit different physiological adaptations. Endurance exercise can affect the composition and function of the gut microbiome by altering cardiorespiratory function and intestinal permeability, while strength training affects skeletal muscle and metabolism, which are also closely related to the regulation of the gut microbiome. At the same time, the differences between athletes of different sports or between athletes of different levels in the same sport are rarely directly compared between athletes of different sports.

[0092] The embodiments of the present disclosure recruited four different groups of athletes - road cyclists, handball players, shooting athletes, and weightlifters - and one non-athlete control group for metagenomic sequencing and fecal metabolomics analysis. By examining the composition, functional characteristics, and metabolite differences of the gut microbiome under different training backgrounds, the present disclosure attempts to reveal the factors that cause differences in the characteristics of the gut microbiome of athletes. The control group consisted of college students who had no regular exercise habits.

[0093] The inclusion criteria included: 1) aged 18 to 30 years; 2) physically healthy, without metabolic diseases or acute infections; 3) athletes with no recent sports injuries. The exclusion criteria are as follows: 1) use of antibiotics or probiotic supplements in the past 3 months; 2) suffering from gastrointestinal diseases, acute and chronic cardiovascular diseases, metabolic diseases, or immune system diseases; 3) smoking and alcohol abuse. A total of 91 eligible subjects were finally included: 18 cyclists, 18 handball players, 19 shooting athletes, 14 weightlifters, and 22 non-athletes. All subjects provided written informed consent. The fecal samples were collected during the same phase of the training cycle of the athletes, specifically during the off-season training period. The fecal samples were collected using a screw cap fecal collection tube (SARSTEDT, Germany), immediately frozen at -20°C after collection, and transported to the laboratory within one to two days after collection, stored at -80°C for subsequent analysis.

[0094] The sample collection and analysis process of the embodiments of the present disclosure includes:

[0095] (1) DNA extraction and metagenomic sequencing

[0096] Weigh 0.5 grams of fecal sample, use Fast Pure Fecal DNA Isolation Kit (magnetic bead method) (MJYH, Shanghai, China) to extract total genomic DNA according to the manufacturer's instructions. The concentration and purity of the extracted DNA were determined using SynergyHTX and NanoDrop 2000, respectively. The quality of the DNA was evaluated using a 1% agarose gel.

[0097] The DNA extract was fragmented to an average of about 400 bp using Covaris M220 (Gene Company Limited, China) for paired-end library construction. The paired-end library was constructed using NEXTFLEX Rapid DNA-Seq (Bioo Scientific, Austin, Texas, USA). Paired-end sequencing was performed using Illumina NovaSeq X Plus (Illumina, Inc., San Diego, California, USA) and NovaSeq X Series 25B Kit according to the manufacturer's instructions (www.illumina.com). The metagenomic sequencing data related to the present disclosure has been stored in the NCBI database (accession number: PRJNA1186265). TM X Plus (Illumina, Inc., San Diego, California, USA) and NovaSeq X Series 25B Kit according to the manufacturer's instructions (www.illumina.com). The metagenomic sequencing data related to the present disclosure has been stored in the NCBI database (accession number: PRJNA1186265).

[0098] (2) Metabolite extraction and UHPLC-MS / MS analysis

[0099] Weigh 50 milligrams of solid sample, add to a 2 milliliter centrifuge tube, and add one grinding bead with a diameter of 6 millimeters. Then, add 400 microliters of extraction solution (methanol: water = 4: 1 (by volume)) containing 0.02 milligrams / milliliter of internal standard (L-2-chlorophenylalanine) for metabolite extraction. The sample was ground using a Wonbio-96c frozen tissue grinder (Shanghai Wanbo Biotechnology Co., Ltd.) at -10°C and 50 Hz for 6 minutes, followed by low-temperature ultrasonic extraction for 30 minutes (5°C, 40 kHz). The sample was placed at -20°C for 30 minutes, and then centrifuged at 4°C at a speed of 13000g for 15 minutes. Then, the supernatant was transferred to an injection vial for LC-MS / MS analysis.

[0100] LC-MS / MS analysis was performed using a Thermo UHPLC-Q Exactive HF-X system equipped with an ACQUITY HSS T3 column (100 mm x 2.1 mm i.d., 1.8 μm; Waters, USA) at Majorbio Bio-Pharm Technology Co. Ltd. The mobile phase included 0.1% formic acid in water / acetonitrile (95:5, by volume) (solvent A) and 0.1% formic acid in acetonitrile / isopropanol / water (47.5:47.5:5, by volume) (solvent B). The flow rate was 0.40 mL / min, and the column temperature was 40 °C. The injection volume was 3 μL.

[0101] Mass spectrometry data were collected in positive and negative modes using a Thermo UHPLC-Q Exactive HF-X mass spectrometer equipped with an electrospray ion source. The optimal condition settings were as follows: source temperature, 425 °C; sheath gas flow rate, 50 arb; auxiliary gas flow rate, 13 arb; ion spray voltage, -3500 volts in negative mode and 3500 volts in positive mode; and normalized collision energy for MS / MS, 20-40-60 eV rolling. The full MS resolution was 60000, and the MS / MS resolution was 7500. Data acquisition was performed using a data-dependent acquisition mode. Detection was performed over a mass range of 70-1050 m / z. Thus, UHPLC-MS / MS here is the full name of Ultra-High Performance Liquid Chromatography-Tandem Mass Spectrometry, which is an advanced analytical method combining chromatographic separation and mass spectrometric detection.

[0102] (3) Bioinformatics analysis for metagenomics

[0103] Obtained data were analyzed on the Majorbio cloud platform (www.majorbio.com). fastp (https: / / github.com / OpenGene / fastp, version 0.20.0) was used to remove adapters from raw sequencing reads and to remove low-quality reads (length < 50 bp or quality value < 20 or containing N bases). BWA (http: / / bio-bwa.sourceforge.net, version 0.7.17) was used to align reads to the human genome and remove any matches related to reads and their paired reads. MEGAHIT (https: / / github.com / voutcn / megahit, version 1.1.2) was used to assemble quality-filtered data. Contigs with a length of ≥ 300 bp were selected as the final assembly results. Prodigal (https: / / github.com / hyattpd / Prodigal, version 2.6.3) was used to predict open reading frames (ORFs) of each assembled contig, and ORFs with a length of ≥ 100 bp were retrieved. CD-HIT (http: / / weizhongli-lab.org / cd-hit / , version 4.7) was used to construct a non-redundant gene catalog with a sequence identity of 90% and a coverage of 90%. SOAPaligner (https: / / github.com / ShujiaHuang / SOAPaligner, version soap2.21 release) was used to estimate gene abundance of a specific sample with an identity of 95%.

[0104] (4) Metabolomics data processing

[0105] UHPLC-MS raw data were converted into a universal format using Progenesis QI (Waters, Milford, USA) with baseline filtering, peak identification, peak integration, retention time correction, and peak alignment. At least 80% of the metabolic features detected in any one group of samples were retained. After filtering, the minimum value in the data matrix was selected to fill in missing values, and each metabolic feature was normalized to the sum. At the same time, quality control (QC) sample variables with a relative standard deviation > 30% were excluded, and log10 logarithm was performed to obtain the final data matrix for subsequent analysis.

[0106] (5) Dietary intake data collection

[0107] A validated food frequency questionnaire (FFQ) was used to assess the dietary intake of each subject. Subjects self-reported their dietary habits through the FFQ, which was then independently calculated and verified by two registered dietitians to determine the energy and nutrient intake of each subject.

[0108] (6) Statistical analysis of the obtained sample data

[0109] Data analysis was performed using SPSS version 27, and R statistical package was used for visualization of diet and gut microbiome relative abundance data. The subjects’ basic information and gut microbiome data were presented as mean and standard deviation, while diet intake data were presented as median and interquartile range. Kruskal-Wallis test and Tukey-Kramer post-hoc test were used to assess the significance of differences in alpha diversity, Firmicutes / Bacteroidetes (F / B) ratio, gut microbiome composition, function, metabolites, and diet intake between groups. Gut-type clustering analysis was performed at the genus level using JSD distance metric. Bray-Curtis distance was used to assess sample similarity, and Principal Coordinate Analysis (PCoA) was applied to visualize the beta diversity of samples between groups. PCoA based on binary-Jaccard distance was used to assess the similarity of diet patterns between groups. Partial Least Squares Discriminant Analysis (PLS-DA) was performed to model the relationship between metabolite expression levels and sample categories, and PLS-DA score plot illustrated the classification effect. Spearman rank correlation analysis was performed to examine the correlation between gut microbiome features, diet intake, and metabolites. The significance level (a) for all statistical analyses was set to a minimum of 0.05.

[0110] The following results were obtained after analysis

[0111] (1) Analysis of the dominant microbial community structure showed that all samples could be classified into two main gut types: Prevotella-dominant gut type and Bacteroides-dominant gut type. Further analysis of the sample composition within these two gut types found that the Prevotella gut type was mainly associated with cyclists. The gut type composition chart also showed that the proportion of the Prevotella gut type in the cycling group was 55.6%, which was significantly higher than that in other groups. In addition, the Bacteroides-dominant gut type was dominant in all groups except the cycling group.

[0112] (2) The Shannon and Simpson indices were used to measure microbial richness and species dominance, respectively. Although no significant difference in alpha diversity was observed between athletes and non-athlete control groups, significant differences were observed between different sports groups. The weightlifting group exhibited the highest Shannon index, which was significantly higher than that of the cycling group (p = 0.013), indicating that this group had the highest species richness and evenness. In contrast, the cycling group exhibited the highest Simpson index, which was significantly higher than that of the weightlifting and shooting groups (p = 0.035 and 0.038, respectively). This indicates that although the cycling group had lower microbial richness, the dominance of the main species was more significant.

[0113] (3) Principal coordinate analysis (PCoA) revealed significant differences in microbial composition among different groups (p = 0.001). Samples from the weightlifting group clustered more tightly in the PCoA plot and were significantly separated from other groups, indicating the high consistency of their microbial composition. Samples from the handball and shooting groups showed higher similarity in the PCoA plot. In addition, samples from the cycling group partially overlapped with those from the control group and were generally significantly separated from other sports groups.

[0114] (4) PCoA analysis of microbial functions at KEGG pathway level 3 showed similar trends as the analysis based on taxonomic composition, with significant differences in microbial functions among groups (p = 0.003). Notably, the weightlifting group was significantly different in function from the control group and other sports groups, while the cycling group had some degree of separation in function from other sports groups.

[0115] (5) In the analysis of microbial composition at the phylum level, samples from all groups were mainly composed of Firmicutes and Bacteroidetes. However, there were significant differences in the Firmicutes / Bacteroidetes (F / B) ratio among groups, with the F / B ratio of the weightlifting group being significantly higher than that of other groups.

[0116] At the genus and species levels, significant differences were observed among groups, and the top 10 different taxonomic units with the highest average relative abundance were listed. In the cycling group, the relative abundance of Prevotella and Prevotella copri was significantly higher than that of other groups, accounting for 18.82% and 15.37% of the relative abundance, respectively. The relative abundance of Bacteroides and Parabacteroides in the handball and shooting groups was significantly higher than that of the control group and / or other sports groups. The weightlifting group exhibited the highest relative abundance of Clostridium, Ruminococcus, Dorea, and Dorea. At the species level, the relative abundance of Eubacterium rectale in the weightlifting group was significantly higher than that of other groups, and it was the most dominant species in this group, with a relative abundance of 5.10%.

[0117] (6) Analysis of microbial functions at KEGG pathway level 3 identified 117 pathways that were significantly different among groups. Among them, amino acid biosynthesis, ABC transporters, carbon metabolism, ribosomes, and purine metabolism were the main different functions with relatively high average abundance. Notably, among the top 15 different functions ranked by abundance, three pathways related to amino acid metabolism—amino acid biosynthesis; cysteine and methionine metabolism; and phenylalanine, tyrosine, and tryptophan biosynthesis—exhibited significantly higher functional potential in the weightlifting group, as shown in Table 6. Figure 3

[0118] ​(7) No significant differences in energy and major nutrient intake were observed among the exercise groups; however, significant differences were observed between the exercise and control groups. Specifically, energy intake was significantly higher in the handball group than in the control group. Protein intake was significantly higher in the cycling, handball, and weightlifting groups than in the control group. Fat and dietary fiber intake was significantly higher in the weightlifting group than in the control group. Figure 4 A). Differences in intake were observed between groups regarding food categories. However, principal coordinate analysis (PCoA) of dietary patterns using the binary-Jaccard distance metric showed no significant overall differences between groups. Notably, the bias observed in the cycling group was attributed to the presence of an outlier sample. Figure 4 B).

[0119] (8) Correlation analysis was performed on the main differences among the groups with dietary intake factors, BMI, training time, and other basic information. This analysis revealed 38 significant correlations at the p<0.05 level. Figure 4 (C) Among these, BMI was significantly associated with eight microbial taxa, while protein powder and sugary beverage intake were significantly associated with five taxa, respectively. However, in terms of correlation coefficients, the correlations between taxa and BMI, as well as dietary factors, were generally weak, ranging from 0.2 to 0.4, indicating a moderate association.

[0120] (9) Supervised discriminant analysis of sample metabolites was performed using partial least squares discriminant analysis (PLS-DA), such as... Figure 5 As shown in the figure, the analysis revealed a certain degree of separation in the metabolite characteristics between different groups, indicating differences in metabolite composition. Furthermore, using a combination of univariate and multivariate statistical analysis methods, differentially expressed metabolites between specific samples from two groups were screened, identifying a total of 4207 significantly different metabolites.

[0121] (10) A correlation heatmap was presented, revealing the relationship between major differential taxa and differential metabolites. The results showed that taxa enriched in the weightlifting group, including *Dorhamnum*, *Dorhamnum*, *Dorhamnum*, *Dorhamnum*, and *Eubacterium*, were significantly positively correlated with bile acid metabolites such as cholic acid, isochoric acid, 12-ketodeoxycholic acid, 7-ketodeoxycholic acid, and deoxycholic acid. Notably, the weightlifting group showed elevated levels of primary bile acid compounds (cholic acid) and secondary bile acid compounds (isocholic acid, deoxycholic acid, and isodeoxycholic acid), significantly different from the cycling group, shooting group, or control group. These metabolites were mainly enriched in three metabolic pathways: bile secretion, primary bile acid biosynthesis, and secondary bile acid biosynthesis. Compared with other groups, the weightlifting group showed enhanced expression activity in these three metabolic pathways.

[0122] The significance of this disclosure includes:

[0123] (1) The present disclosure fills a gap by conducting a comprehensive metagenomic sequencing and metabolomics analysis of elite athletes from four different sports in China and a control group of non-athletes. The present disclosure observes significant differences in the structure and function of the gut microbiome of athletes from different sports, with little correlation to dietary intake, suggesting that training modalities can be the dominant factor in shaping the gut microbiome. Notably, endurance athletes and strength athletes exhibit unique microbiome features, suggesting different microbial functions. The potential benefits of long-term strength training in enhancing gut microbiome composition and function warrant further investigation, as they can have profound implications for the health and performance of athletes.

[0124] (2) Contrary to the commonly observed trend that athletes exhibit higher alpha diversity, the present disclosure finds no significant difference in alpha diversity of the gut microbiome between athletes and the control group of non-athletes. This unexpected finding is consistent with the notion that high-intensity training of athletes can not necessarily have better health outcomes compared to the general population. Notably, significant variations in alpha diversity are observed among athletes from different sports, with weightlifting athletes exhibiting higher richness and evenness, but lower dominance of the main species compared to road cycling athletes. These results suggest that the type of training regimen significantly influences the composition of the gut microbiome, a hypothesis that will be further explored in subsequent sections.

[0125] (3) The present disclosure's gut profiling identifies two main enterotypes: a Prevotella-dominant enterotype and a Bacteroides-dominant enterotype. The Prevotella enterotype is predominantly observed in cycling athletes, while the Bacteroides enterotype is more prevalent among other athletes and the control group. These enterotypes are consistent with the main types reported in human populations, associated with long-term dietary patterns. The Prevotella enterotype is typically associated with diets rich in complex carbohydrates and high dietary fiber, while the Bacteroides enterotype is associated with diets rich in animal-derived fats and proteins. However, in the present disclosure, using Principal Coordinate Analysis (PCoA) did not reveal significant differences in dietary patterns among the groups after correcting for one outlier sample in the cycling group. Despite the higher nutrient intake of the athlete groups, none of the groups reached the recommended dietary fiber intake, while exceeding the recommended intake of animal-derived foods, indicating a general preference for animal-derived diets. Moreover, correlation analysis failed to establish a clear relationship between enterotypes and nutrient intake, suggesting that the prevalence of the Prevotella enterotype among cycling athletes can be more related to the type of sport than to the diet. Beta diversity analysis further reveals significant differences in taxonomic composition and microbial functions, particularly in distinguishing between the cycling and weightlifting groups. These findings support the hypothesis that long-term training modalities can significantly influence the development of gut microbial community features.

[0126] (4) To elucidate the structure of the gut microbial community, the present disclosure first analyzed the microbial composition at the phylum level. Notably, the Firmicutes / Bacteroidetes (F / B) ratio of weightlifters was significantly higher than that of other groups, which is contrary to previous findings that the F / B ratio is positively correlated with maximal oxygen uptake (VO2max) level. Although the present disclosure did not directly measure VO2max, it is reasonable to speculate that the VO2max level of weightlifters can be lower than that of road cyclists, given the known association between aerobic capacity and VO2max. In addition, previous studies have shown that VO2max level is positively correlated with microbial alpha diversity. However, the results of the present disclosure showed that both the alpha diversity and the F / B ratio of weightlifters were higher than those of cyclists, which is inconsistent with the expected relationship with VO2max level. This difference suggests that the diversity and composition of the gut microbiome of athletes can be influenced by multiple training factors. In addition to cardiovascular adaptability, the unique exercise patterns, intensity, and duration associated with weightlifting can have significantly contributed to the observed characteristics of the gut microbiome of athletes.

[0127] (5) Further analysis at the genus and species level confirmed the uniqueness of the gut microbiome composition of weightlifters, with increased relative abundance of Clostridium, Ruminococcus, Dorea, and Dorea, indicating a higher evenness of taxonomic units at the genus level. These genera enriched in weightlifters are known for their butyrate-producing capacity, which is crucial for intestinal epithelial cell health, intestinal barrier function, immune modulation, and anti-inflammatory properties. Notably, the relative abundance of Eubacterium rectale was significantly enriched in weightlifters, which is a new finding.

[0128] (6) The significant increase in Prevotella and Prevotella copri in cyclists is of interest, which is consistent with the aforementioned analysis of gut types and the findings of Aya et al., who observed higher relative abundance of Prevotellaceae in cyclists compared to weightlifters. Petersen et al. linked the high abundance of Prevotella to the increased weekly training time of professional cyclists. This association between Prevotella and endurance performance is further supported by observations in marathon runners and soccer players. Contrary to Petersen’s speculation about diet as an influencing factor, the present disclosure found no significant differences in energy, carbohydrate, or dietary fiber intake between cyclists and other groups, nor was there a clear correlation with the levels of Prevotella or Prevotella copri. Prevotella bacteria, such as Prevotella copri, are equipped with polysaccharide utilization sites that facilitate the energy metabolism of complex carbohydrates. In addition, Prevotella copri is capable of detoxifying superoxide radicals and tolerating reactive oxygen species. The present disclosure believes that the high energy demand of endurance training can be met by the efficient polysaccharide metabolism of Prevotella, providing potential energy supplementation for endurance athletes. The upregulation of energy metabolism and oxidative stress pathways during endurance exercise can also favor the selective development of Prevotella. The bidirectional enhancement between Prevotella and endurance performance has considerable potential as a biomarker for identifying elite endurance athletes.

[0129] (7) The functional analysis of the present disclosure identified 117 functional pathways that were significantly different between groups, with weightlifters exhibiting unique functional potential, particularly in amino acid biosynthesis, ABC transporters, purine metabolism, and quorum sensing. Notably, weightlifters exhibited higher levels in pathways related to amino acid metabolism, including the biosynthesis and metabolism of cysteine, methionine, and aromatic amino acids Figure 3 ). These microbial functions are crucial for the health and performance of both the microbiota and the host. Amino acids are essential for protein synthesis and muscle repair, especially for athletes with high muscle turnover rates. Microbially derived amino acids can enhance muscle mass and recovery. In addition, short-chain fatty acids (SCFAs) produced during microbial amino acid metabolism can improve athletic performance. Certain amino acids have important physiological roles: cysteine is a key component of glutathione, an important antioxidant; methionine metabolites are involved in cellular methylation and regulation of gene expression; tryptophan is the precursor of serotonin, affecting mood, sleep, and cognitive function. Microbiota-derived tryptophan can also counteract the decline in blood tryptophan levels caused by exercise, suppress the production of kynurenine, and delay the onset of exercise fatigue. To some extent, weightlifters have achieved the best synergy between host needs and gut microbiome functions.

[0130] (8) To uncover these potential mechanisms, the present disclosure conducted a non-targeted fecal metabolomics analysis. This analysis confirmed the differences in gut microbiome composition, with distinct metabolic features observed between cyclists and weightlifters. Through inter-group differential analysis, weightlifters exhibited elevated levels of primary (e.g., cholic acid) and secondary (e.g., isocholic acid and deoxycholic acid) bile acids. Significant positive correlations were observed between feature taxa (including Dorea, Dorea, Dorea, Dorea, and Eubacterium rectale) and bile acid metabolites in the weightlifter group. These correlations have not been reported in previous studies focused on professional athletes, providing important insights into the development of gut microbiome features in weightlifters.

[0131] Primary bile acids are synthesized in the liver and are crucial for promoting microbial diversity. Their microbial metabolic transformation into secondary bile acids significantly influences host health and can affect exercise-mediated skeletal muscle function through the bile acid-FXR-FGF19 signaling axis. Although the cross-sectional study of the present disclosure cannot establish causality, it provides insights into the potential link between bile acid metabolism and exercise and the gut microbiome. Previous studies have shown that, unlike endurance exercise, resistance exercise decreases circulating primary bile acid levels and increases secondary bile acids, such as cholic acid. Gehlert et al. further found that primary bile acids in muscle tissue significantly increased after resistance exercise, suggesting that the observed decrease in circulating primary bile acids can be due to increased uptake by muscle tissue. Although these studies did not directly examine the intestinal transport of bile acids after exercise, it can be inferred that resistance exercise can also increase the uptake of primary bile acids by the intestine, providing the necessary substrate for the gut microbiota and potentially promoting alpha diversity and the proliferation of related taxa. Therefore, the present disclosure speculates that long-term resistance training uniquely modulates the liver-gut-muscle axis in weightlifters, contributing to the observed increase in gut microbiome alpha diversity and functional potential. However, this hypothesis requires further in-depth study and verification to provide new evidence for exercise-based strategies to modulate the gut microbiome.

[0132] The analysis method of the embodiments of the present disclosure reveals that there are significant differences in the structure and function of the gut microbiome between athletes with different training modes. Weightlifters exhibit enhanced microbial diversity and function, and their gut microbiome features are significantly correlated with bile acid metabolites. Training mode plays an important role in shaping the gut microbiome features of athletes. These findings provide new solutions for sports training practice and public health management.

[0133] In the embodiments of the present disclosure, Figure 4English terms included in the text are: Gut Microbiota - gut microbiome, Bacteroidetes - Bacteroidetes phylum, Prevotella - Prevotella genus, Enterotypes - enterotypes, Shannon Index - Shannon index (measure of species diversity), Simpson Index - Simpson index (measure of species diversity), PCoA (Principal Coordinate Analysis) - principal coordinate analysis, PC (Principal Component) - principal component, PLS-DA (Partial Least Squares Discriminant Analysis) - partial least squares discriminant analysis, Bray-Curtis Distance - Bray-Curtis distance, Weighted-UniFrac - Weighted-UniFrac method (dimensionality reduction method for principal coordinate analysis), Clustering - clustering, Boxplot - boxplot, Barplot - barplot.

[0134] Figure 4 Enterotype classification and microbial diversity in fecal samples are shown. (A) Principal coordinate analysis (PCoA) using Jensen-Shannon divergence at genus level visualizes the two enterotypes. Red represents the Prevotella-dominant enterotype, while blue represents the Bacteroides-dominant enterotype. (B) Distribution of the two enterotypes in all five groups. In the cycling, control, handball, and weightlifting groups, approximately 55.6%, 22.7%, 5.6%, and 14.3% of the samples, respectively, were found to be the Prevotella-dominant enterotype. (C) Shannon index and (D) Simpson index of the microbial communities in all five groups. *p < 0.05. PCoA using Bray-Curtis distance was compressed on the third level of the KEG hierarchy based on (E) taxonomic and (F) functional composition.

[0135] Figure 5The English and Chinese terminology includes: Phylum level, Firmicutes, Proteobacteria, Actinobacteria, Bacteroidetes, Clostridium, Dorea, Ruminococcus, Prevotella, Verrucomicrobia, Streptophylococcus, Eubacterium, Eubacterium rectale, and Parabacteroides.

[0136] Figure 5 The microbial community composition was compared among the different groups. (A) phylum-level distribution of microorganisms among the five groups. (B) Comparison of Firmicutes / Bacteroidetes (F / B) ratio among the five groups. *p<0.05, **p<0.01. (C) Comparison of relative abundance of gut microbiota at the genus and species levels among all five groups. *p<0.05, **p<0.01, ***p<0.001.

[0137] Figure 6 The English and Chinese terms include: Abundance, KEGG Pathways.

[0138] Figure 6 The abundance of amino acid metabolic functions among different groups is shown. *p<0.05, **p<0.01, ****p<0.001. ko01230: Biosynthesis of amino acids; ko00270: Cysteine ​​and methionine metabolism; ko00400: Biosynthesis of phenylalanine, tyrosine, and tryptophan.

[0139] It is worth noting that although the spirit and principles of this invention have been described with reference to several specific embodiments, it should be understood that this invention is not limited to the disclosed specific embodiments, and the division of aspects does not imply that the features in these aspects cannot be combined; such division is merely for the convenience of description. This invention is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

Claims

1. A method for investigating the dietary nutrition of athletes, comprising the following steps: Firstly, collecting the basic information of the participants; Collecting blood samples of the participants for subsequent biomarker analysis; After collecting the blood samples, the participants complete the first FFQ to assess their dietary habits; After the first FFQ, the participants conduct a three-day weighed food record, including two weekdays and one weekend day, to assess the effectiveness of the first FFQ; By analyzing the biomarkers in the blood samples and the three-day weighed food intake, the results of the first FFQ are compared to assess the effectiveness of the first FFQ; One month after completing the first FFQ, the participants complete the second FFQ to assess the reproducibility of the FFQ; By comparing the results of the first and second FFQs, the consistency of the questionnaire at different time points is evaluated.

2. The method of claim 1, wherein, After collecting the blood samples, the serum is separated by centrifugation, and the serum fatty acids are analyzed. The total serum fatty acids are extracted using a fatty acid methyl ester kit, and the serum fatty acids are quantified by scanning the corresponding m / z values from the precursor ion.

3. The method of claim 2, wherein, By Spearman correlation analysis, the correlation between the estimated fatty acid intake from the FFQ and the serum biomarkers is determined.

4. The method of claim 1, wherein, Various statistical methods are used to evaluate the relative effectiveness and reproducibility of the FFQ, including Spearman correlation coefficient, energy-adjusted correlation coefficient, attenuation-free correlation coefficient, kappa statistic, and / or Bland-Altman analysis.

5. The method of claim 4, wherein, Bland-Altman analysis is used to assess the relationship between the average and difference values of daily intake between 3DWFR and the first FFQ to evaluate consistency.

6. The method of claim 1, wherein, The consistency of the two FFQs in energy, nutrient, and food group intake assessment is evaluated by Spearman correlation coefficient and intraclass correlation coefficient.

7. The method of claim 1, wherein, Factor analysis is used to extract dietary patterns, and the patterns are named based on the composition of foods in each component. Chi-square test, logistic regression analysis, one-way ANOVA, and multiple linear regression statistical methods are used to determine the effects of different dietary patterns on different nutritional indicators.

8. The method of claim 7, wherein, The factor scores of different dietary patterns are divided into Q1 low-score group and Q2 high-score group from low to high. Logistic regression analysis is used to determine which dietary pattern is a protective factor for abnormal nutritional indicators.

9. The method of claim 8, wherein, The nutritional indicators include body measurement indicators and blood test indicators. Body measurement indicators include body mass index (BMI), body fat percentage (FAT%), and fat-free mass (FFM). Blood test indicators include testosterone (T), blood urea nitrogen (BUN), hemoglobin (HB), ferritin (SF), and blood lipids, including high-density lipoprotein (HDL), low-density lipoprotein (LDL), cholesterol (TC), and triglycerides (TG).

10. An athlete gut microbiome analysis method, characterized by, The method includes high-throughput metagenomic sequencing and metabolomics analysis of athlete fecal samples, and collecting dietary intake data through food frequency questionnaires to analyze the differences in composition, function, and metabolic phenotype of intestinal microbiomes between athletes of different sports.

Images