At the macronutrient level the vegan diet can be considered healthy since the distribution of macronutrients corresponds well to that proposed by the NNR. Specifically, a high PUFA:SFA ratio, as that observed in vegans, is suggested to be favorable in regard to the risk of coronary heart disease . Furthermore, a low intake of added sugar, as reported in vegans, is considered beneficial for human health  and similarly a high dietary fibre intake may provide specific benefits in relation to gastrointestinal and metabolic functions [18, 19]. Intake of added sugar has previously only been examined in one American study comprising 15 vegans and 6 omnivores, in which, no difference was reported . The low intake of added sugar in the present study might be due to a lower intake of processed food and sugar-sweetened beverages. This is only speculative since we do not have data at food item level; it is, however, supported by a recent study, in which vegans reported consuming fewer servings of sweets per day than omnivores . An evaluation of the diet in the general Danish population has been done previously .
The vegan population had, however, an insufficient intake of several vitamins, which could have a negative health effect. In contrast to our findings, previous studies, reporting intake of total vitamin A (retinol equivalents; RE) found higher intake of vitamin A in vegans compared to omnivores [5, 9, 10]. However, it is difficult to compare results across studies since the amount of retinol and beta-carotene, from which the vitamin A intake is calculated, is not presented. These opposing results could be due to different methods of calculating vitamin A RE from beta-carotene and other carotenoids. The vegan intake of retinol was very low in the present study compared to previous studies reporting this [4, 11]. In the present study the 2001 Institute of Medicine Interconversion of Vitamin A and Carotenoid Units was applied . Retinol is primarily found in animal products , why the finding in the present study seems plausible. Approximately half of the vegans did not reach the recommendations for vitamin A including both intake from diet and supplements. Symptoms of vitamin A deficiency are night blindness, dry and scaly skin, increased number of infections in the respiratory tract, the gut and the urinary tract and severe vitamin A deficiency has furthermore been associated with cancer at these sites .
The vegans participating in the present study had a low intake of riboflavin and vitamin B12, which corresponds to previous findings [4, 5, 9–11]. The major food sources of riboflavin in Nordic diets are milk and meat products , which explain the low intake of this vitamin among vegans. Recommended intake of riboflavin and vitamin B12, including intake from supplements, is not reached by 29 and 31 vegans (of 70), respectively. Little attention has been paid to the effects of riboflavin deficiency on human health. An in vitro study using duodenal biopsies demonstrated that riboflavin depletion in adult humans impairs proliferation of intestinal cells and thereby may have implications for gastrointestinal function and nutrient absorption . Without supplementation the low dietary intake of vitamin B12 among vegans could enhance risk of pernicious anemia and polyneuropathy .
In 41 (of 70) vegans the total intake of vitamin D did not meet the recommendations. The consequence of insufficient intake of vitamin D is lower absorption of calcium and phosphorus, which may impact bone metabolism. Furthermore, it has been suggested that high vitamin D plasma concentration or vitamin D supplementation is associated with decreased risk of colorectal cancer, cardiovascular disease and type 2 diabetes .
The intake of vitamin D and vitamin B12 among vegans was very low in the present study and it is lower than what has been observed in previous studies [4, 5, 9–11]. This could be due to low availability of fortified foods, which are common in some countries but was prohibited in Denmark by law until 2003 and has still not been introduced on a wider scale.
In the present study, intake of dietary sodium among vegans was low compared with the general population. Three studies have previously examined the sodium content of a vegan diet, two of which also showed lower sodium intake among vegans [8, 9], whereas the third study found no difference between a vegan and an omnivore diet . The low intake of sodium in the vegan population in this study might be due to a lower intake of processed foods, which usually contain high amounts of salt . Most of the vegans (55 of 70) did not meet the recommendation for iodine intake when including intake from supplements. Only one study has previously investigated iodine intake in vegans reporting an iodine intake from diet and supplements among vegans of only 50–70 % of the dietary reference value . In Denmark, fortified table salt is a major source of iodine and it is therefore also subject to potential under- reporting. Other sources of iodine include fish and sea plants . In general, sodium and iodine intakes are difficult to quantify and the results should be interpreted with caution . Less than half of the vegans (24 of 70) reached the recommended intake of selenium. No studies have previously examined selenium intake in vegans. An insufficient intake of iodine and selenium might potentially have negative health impact such as development of goiter. Adequate iodine intake is important throughout life, but especially in childhood and during pregnancy and breastfeeding . Low serum selenium levels (≤1 μmol/l) have been associated with increased risk of myocardial infarction in men ; however, the lower critical threshold for intake of this mineral is unknown.
Even though the intake of iron and calcium in male vegans met the Nordic recommendations, the absorption of these minerals might not be correspondingly high. Iron and calcium are known to have low bioavailability in the context of plant-based foods. This is due to the chemical form in which the minerals are present, the nature of the food matrix and the presence of bioavailability-lowering compounds in plant-based foods (e.g. phytate, dietary fiber and oxalic acid) [28, 29]. Considering the low intake of these minerals among vegan women, a reduced bioavailability may further increase the risk of outright deficiency and related disorders. However, it is well established that the bioavailability of iron increases with the presence of ascorbic acid (vitamin C) and other organic acids . Intake of vitamin C was very high among vegans, thus potentially compensating for the low bioavailability of iron in plant–based foods . It has previously been proposed that the recommended intake of iron for vegans should be 1.8 times higher than that of omnivores because of the lower bioavailability ; in the present study the iron intake among vegan men met this recommendation.
Vitamins and minerals interact and are dependent on sufficient availability of one another in order to function and contribute to ensure human health. Examples are vitamin A and zinc, every B-vitamin as well as vitamin D and calcium. This emphasizes the importance of adequate intake of every individual mineral since, for instance, an inadequate intake of vitamin A affects the function of zinc even though this mineral is consumed in sufficient amounts  as is the case for the vegans in this study.
To examine whether the low mineral intake among vegans in this study has a negative health effect, pertinent biochemistry of study participants should be evaluated. Furthermore, a healthy diet is not only a diet balanced for macro- and micronutrients. Other components such as amount of processed food, heated food, food preservatives and additives as well as the sources of the nutrients (e.g. protein from meat versus plants) could be important in evaluating a diet. However, in the present study, these data were not available.
The present study covers the macro- and micronutrient composition of vegans in general. At present very little is known about the health effect of a vegan diet under specific circumstances such as pregnancy, lactation, childhood and advanced age; circumstances which the present study does not address.
The main strength of the present study is the method applied to assess dietary intake. A weighed four-day food record is considered the gold standard for specifying dietary intake . However, a potential error may be introduced by the diet record, as it may be a burden to weigh and register all food items, thus a person may eat more simple foods than usual. Another important strength is the availability of detailed data on dietary supplements, the most comprehensive to date in a European sample, allowing us to get a more complete picture of the micronutrient intake.
An almost unavoidable limitation with diet recording is under-reporting . In the vegan population the EI: BMR ratio was acceptable in 96.0 % of cases, indicating that under-reporting EI might not be an issue in the examined vegan population. Nevertheless, it is a potential limitation that a vegan diet is non-conventional and some of the recorded meals or food items were unavailable in the database. However, this shortcoming was evaded in the present study by constructing recipes using only completely validated food items.
Whereas the number of vegans included in some other studies exceeds the number in our study by far, most have used less accurate methods [4, 5, 8–11] and the present study is the largest to date using the gold standard method to assess dietary intake. However, while sufficiently powered to detect differences between vegans and the general population it is important to exhibit caution when drawing conclusions based on a small sample, which may not be representative of the Danish vegan population in general. Another potential source of bias arises from the fact that the included vegans represent a self-selected group of participants. Particularly, healthy lifestyle bias may be an issue. Vegans may be more concerned about healthy lifestyle, consumption of healthy foods and optimal amounts of macro- and micronutrients. Furthermore, the vegan subjects had higher educational attainment compared to the general Danish population, a factor that has been associated with healthier lifestyle in general . This might partially explain the better compliance to the recommendations at macronutrient level compared to the general Danish population.