Organizations such as The American College of Sports Medicine (ACSM), The International Society for Sports Nutrition (ISSN) and the International Olympic Committee (IOC) have defined guidelines for athletes [35,36,37]. As these few existing recommendations for mainly high-performance athletes were only partially applicable to this study collective, the nutrient intake was compared with intake recommendations of the D-A-CH for the general population. However, the D-A-CH does not specify any certain reference values for ambitious recreational athletes .
In general, recreational athletes can be supplied with all micronutrients through a balanced mixed diet. But, it is unknown whether a vegetarian and especially vegan diet can provide all the important nutrients for athletes.
The type, duration and intensity of sport determines the energy requirements. The ISSN recommends an energy intake from 7.510.0MJ (18002400kcal) for athletes with general physical activity levels of 3040min three to four times a week . In order to assess the energy demand, the ACSM recommends various options (e.g. based on the daily recommended intake, the basal metabolic rate and a factor of physical activity or metabolic equivalents) . The IOC refers to the fat-free mass (3045kcal/kg FFM/day) . Our subjects trained an average of three times a week for about 60min, which corresponds to an estimated physical activity level value of about 1.7 (sedentary work and recreationally active) . More than half of each group did not reach the recommended energy intake, which is not uncommon in endurance athletes . There were no differences among the groups, which agrees with the results of Lynch and colleagues, who compared 35 vegetarian athletes with 35 omnivores .
Carbohydrates are the most important sources of energy and many endurance athletes strive to consume carbohydrates to benefit from full glycogen stores . Depending on the intensity and type of training or competition, gender, and external influences, an absolute amount of 37g/kg BW is recommended for people with general physical activity of about 3060min/day 34 times a week up to about 1 hour a day [35,36,37]. Thus, participants in the present study achieved the recommendations for carbohydrate intake [35,36,37]. Similar to previous studies with non-athletes [41,42,43,44], the VEG group had the highest intake of carbohydrates (55.2, 51.459.0 EN%) compared to OMN (46.7, 43.649.8 EN%; pOMN-VEG=0.002) and LOV (49.4, 45.553.3 EN%; n.s. compared to VEG), which can be explained by the increased intake of potatoes and fruit, since the intake of whole-grain and cereal products, pastries and sweets were similar for all groups.
The protein needs of athletes have been widely discussed [45,46,47]. The three societies recommend a range of 1.22.0g/kg BW for most exercising individuals (including general fitness ) [35,36,37]. According to the IOC and ACSM, the recommended amount also applies to vegetarians. The average protein intake of all three groups was within the reference range. In addition to absolute protein intake, it is important to consider the quality of the proteins . Protein sources were mainly meat, meat products and sausages, fish and dairy products for the OMN group, milk, dairy products, and eggs for LOV, and cereal products, legumes and soybeans for VEG. In general, a high biological value can be achieved with each of these three diets. Compared to the reference values of the WHO, on average, all groups met the reference range for amino acid intake . Hence, it can be assumed that all three groups including VEG had an adequate protein and amino acid supply. Our findings are consistent with the literature, which has shown that non-athlete LOV and VEG appear to be within the range of recommendations for protein intake [44, 48].
Dietary fats are valuable energy sources and have structural and regulatory functions. Dietary recommendations for adequate fat intakes vary widely and depend on the level of training and body composition goals [35,36,37]. While the ACSM recommends a daily intake of 2035 EN% but not less than 20 EN% fat , the IOC advises an intake of 1520 EN% fat, depending on the type of sport . By contrast, both D-A-CH and ISSN recommend a fat intake of 30 EN% [33, 35]. Most subjects in the three groups reached the recommendations of the D-A-CH , ISSN and ACSM. In addition, it is important to evaluate the PUFA intake of athletes, which was below the reference value in all three groups . PUFAs play a pivotal role in health due to their precursor function as regulatory lipid mediators. The International Society for the Study of Fatty Acids and Lipids recommends a daily sum EPA+DHA intake of 0.5g, which was achieved by the OMN group (0.54, 0.230.85g), but not by LOV (0.08, 0.040.12g; pOMN-LOV=0.003) or VEG (0.09, 0.010.17g; pOMN-VEG<0.001) . PUFA intakes in LOV and VEG within this study can be classified as inadequate, which is consistent with other studies regarding non-athlete vegetarians and vegans . The EPA/DHA supplements were only consumed occasionally in the VEG and LOV groups. The resulting LA:ALA ratios in the VEG (1:5.71) and LOV groups (1:5.30) were within the reference range . The OMN group showed higher LA:ALA ratios (1:8.04), which are consistent with the results of the German Nutrition Survey .
It is generally thought that athletes consume high amounts of micronutrients via dietary supplements due to their increased health awareness . However, several studies have shown insufficient micronutrient intake in athletes [54, 55]. There are no specific recommendations for micronutrient intakes in recreationally active individuals, which differ from the general populations guidelines. However, in the view of the ACSM, ISSN and IOC, an adequate supply of micronutrients is assured with a balanced mixed diet. A possible insufficient supply to vegetarians of zinc, iron, riboflavin, cobalamin and vitamin D is described in the ACSM and IOC guidelines [36, 37], while the ACSM additionally mentions calcium, pyridoxine and folate. A specific risk of an insufficient micronutrient supply with a vegan diet is not mentioned.
In the present study, magnesium, calcium, iron, vitamin D and cobalamin were the most frequently supplemented nutrients. Cobalamin intake was strongly dependent on supplementation, especially for both female and male VEG. Half of the VEG group supplemented cobalamin and, thus, had a significantly higher intake compared to the D-A-CH reference values of 4g per day . However, the very high cobalamin intake of the vegan supplement user can be classified as uncritical [56, 57]. In addition, the absorption rate decreases with increasing dosage. As expected, subjects of the VEG group who did not take cobalamin supplements had a marginal intake. Additionally, the dietary intake of the LOV group was insufficient, especially for males, who had cobalamin intakes below the recommendations, regardless of supplementation. However, although consuming cobalamin-rich foods such as meat, meat products and fish, its intake was still inadequate in one-third of the OMN group. Cobalamin is considered critical for VEG, but adequate intake should be ensured for every diet.
Due to high riboflavin levels in animal products, it was not surprising that the OMN group consumed the highest amounts, although, on average, VEG and female LOV reached the recommendations, which agrees with previous studies in non-athletes [58, 59]. In contrast to Eisinger and colleagues, who showed high intakes of riboflavin in LOV endurance runners , only female LOV achieved the reference values. Pyridoxine intake exceeded the recommendations in the VEG group due to the high consumption of vegetables, legumes, nuts, and seeds, which has already been shown by other studies with non-athletes [58, 61]. The VEG group showed a high folate intake due to the high amount of folate in green vegetables, yeast, and nuts, while the folate intake of most OMN subjects was insufficient. These results are consistent with the German Nutrition Survey  and studies with athletes .
Similar to cobalamin, vitamin D intake was strongly dependent on the use of supplements. This becomes clear by comparing the vitamin D intake between SU and non-SU. On average, the VEG group (19.9, 2.7537.0g) was closest to the recommendations of 20g per day compared to OMN (8.29, 2.2114.4g) and LOV (4.52, 1.1410.4g). However, the intake of vitamin D was considerably higher in SU compared to non-SU. Hence, the mean values for the vitamin D intake in the VEG group (including SU and non-SU) should be treated with caution. This also applies to the OMN and LOV group, although not quite as strongly pronounced. However, it is worth mentioning that an adequate vitamin D status can only be evaluated with the endogenous 25-hydroxyvitamin D status in the blood .
Similar to other studies with non-athletes [42, 58], the highest iron intake from food (excluding supplements) was observed in VEG subjects compared to LOV and OMN. In addition, the VEG group had the highest iron intake via supplements compared to the other two groups. A total of more than 85% of VEG subjects achieved the recommendations compared to only ~50% in OMN and LOV. Male subjects of all groups were above the recommendations with more than 10mg per day, independent of supplementation. Female OMN and LOV subjects achieved the recommendation of 15mg daily only after supplementation. Interestingly, the VEG group reached the iron intake recommendations solely via food and not via supplements. The literature on the iron supply of athletes is inconsistent. Some studies found an adequate [63, 64] and others an inadequate iron intake in athletes . High-performance athletes might have increased requirements due to biochemical adaptations (e.g. increased blood formation and increased enzymatic antioxidant defense) and increased iron losses via sweat, urine, and feces, which results in a higher risk of iron deficiency anemia . In addition to absolute amounts, the bioavailability of different iron species should be considered. Despite the exclusive consumption of plant-based iron of the VEG group, LOV and OMN also consumed predominantly plant iron sources. While plant-based foods contain non-heme iron, mainly in trivalent form (Fe3+), which has a poor bioavailability of 15%, meat and fish contain about 70% of the total iron in the form of heme iron, which can be absorbed much better at 1020% [66, 67]. Hence, the lower iron intake in OMN subjects compared to LOV and VEG does not necessarily result in a lower status. Moreover, further influences on bioavailability (promoting substances such as ascorbic acid or lactic acid and inhibiting substances such as phytic acid or oxalic acid, which occurs in vegetable foods) must be taken into account (the same applies to zinc, magnesium, and calcium). Therefore, only functional parameters, such as transferrin and ferritin, indicate an adequate supply status.
The present results show that calcium is a critical nutrient . As expected, calcium intake was highest among OMN subjects, although more than half were below the reference range. The highest number of subjects with an intake below the reference range for calcium was found in the LOV group (64.0%), although they consumed milk and dairy products. The athletes dietary intake of calcium should be improved independently of dietary habits due to the importance of bone health, and normal nerve and muscle function . The mean intake of zinc was within the reference range for all groups, although male LOV subjects were slightly below. Female participants and non-SU of all three groups reached the recommendations . Interestingly, the zinc supply was similar in OMN and VEG subjects, although animal-based foods are rich in zinc and the zinc supplement intake in the VEG group was considerably lower than OMN. These results reveal that zinc-rich plant-based foods can secure adequate zinc supply. The literature on zinc supply is inconsistent. Some studies observed a slightly lower but adequate intake of zinc in vegetarians and VEG compared to OMN [43, 48, 58], other studies found no differences between vegetarian and OMN endurance athletes .
The fact that the data of dietary intake relied on self-reported data by subjects should be considered. Both under- and over-reporting are further sources of error in dietary records. Since the use of iodized salt is voluntary in Germany and a precise indication about the dietary intake is critical, the values of iodine intake should be considered with caution. Furthermore, there are limitations regarding the nutrition software that shows data gaps, especially regarding vegan products. We did not consider the water intake of the subjects, which might also influence nutrient (e.g. mineral) supply.