**Micronutrients Dietary Supplementation Advices for Celiac Patients on Long-Term Gluten-Free Diet with Good Compliance: A Review**

**Mariangela Rondanelli 1,2, Milena A. Faliva 3, Clara Gasparri 3, Gabriella Peroni 3, Maurizio Naso 3, Giulia Picciotto 3, Antonella Riva 4, Mara Nichetti 3, Vittoria Infantino 5, Tariq A. Alalwan <sup>6</sup> and Simone Perna 6,\***


Received: 24 April 2019; Accepted: 27 June 2019; Published: 3 July 2019

**Abstract:** *Background and objective*: Often micronutrient deficiencies cannot be detected when patient is already following a long-term gluten-free diet with good compliance (LTGFDWGC). The aim of this narrative review is to evaluate the most recent literature that considers blood micronutrient deficiencies in LTGFDWGC subjects, in order to prepare dietary supplementation advice (DSA). *Materials and methods*: A research strategy was planned on PubMed by defining the following keywords: celiac disease, vitamin B12, iron, folic acid, and vitamin D. *Results*: This review included 73 studies. The few studies on micronutrient circulating levels in long-term gluten-free diet (LTGFD) patients over 2 years with good compliance demonstrated that deficiency was detected in up to: 30% of subjects for vitamin B12 (DSA: 1000 mcg/day until level is normal, then 500 mcg), 40% for iron (325 mg/day), 20% for folic acid (1 mg/day for 3 months, followed by 400–800 mcg/day), 25% for vitamin D (1000 UI/day or more-based serum level or 50,000 UI/week if level is <20 ng/mL), 40% for zinc (25–40 mg/day), 3.6% of children for calcium (1000–1500 mg/day), 20% for magnesium (200–300 mg/day); no data is available in adults for magnesium. *Conclusions*: If integration with diet is not enough, starting with supplements may be the correct way, after evaluating the initial blood level to determine the right dosage of supplementation.

**Keywords:** celiac disease; vitamin B12; iron; folic acid; vitamin D; long-term GFD therapy (LTGFD); LTGFD with good compliance (LTGFDWGC)

## **1. Introduction**

Celiac disease (CD) is an immune-mediated systemic disorder triggered by the ingestion of gluten and prolamines in genetically predisposed individuals. It is characterized by inflammation of the small bowel mucosa—the immune reaction—which occurs after ingestion of gluten that leads to intestinal villous atrophy, crypt hyperplasia, and increased number of intraepithelial lymphocytes [1]. CD is a multifactorial disease and its pathogenesis involves both genetic and environmental factors [2]. Genetic composition for the development of the disease is evident. In fact, more than 90% of celiac patients are human leukocyte antigen (HLA)-DQ2 haplotype positive and almost all of the rest carry HLA-DQ8. These genes are necessary but not sufficient for CD development [3,4]. The predisposing

DQ2 and DQ8 heterodimers are composed of the association of α and β chains. A recent meta-analysis showed that the HLA genotypes coding for DQ2 or DQ8 heterodimers, but also those including only the alleles of the respective β chains (regardless of the concomitant presence of DQ2 or DQ8 α chains) have an increased risk of developing pediatric CD [5]. Recently, another meta-analysis evaluated the predictive values of HLA-DQB1\*02 allele, suggesting the major relevance of this specific allele, rather than the expression of the full DQ2 and/or DQ8 heterodimers, in raising the risk to develop pediatric CD [6]. In addition, a risk gradient according to single or double copy of HLA-DQB1\*02 has been revealed [6]. Gluten ingestion represents the major environmental factor, contributing to the development of the pathology, but there are several other conditions involved in the etiology of CD, including viral infections, gut microbiota, breastfeeding, early life feeding practice, and smoking [3,4].

CD can occur at any stage of life and with a great variety of signs and symptoms. In fact, it is considered a multisystem immunological disorder rather than a disease restricted only to the gastrointestinal tract. Consequently, it is important to make diagnosis not only in individuals with classic gastrointestinal symptoms, but also in subjects who have a more nuanced or extra-intestinal clinical features, since the consequences can be important in both cases [2]. To date, nutritional therapy has been the only effective treatment for patients with CD that demands a strict compliance with a gluten-free diet (GFD). Non-adherence to the GFD increases the risk of morbidity and mortality, as a result of associated conditions, which include infertility, skeletal disorders and malignancy. Once diagnosed, patients should be tested for micronutrient deficiencies, including iron, folic acid, vitamin B12, and vitamin D [7].

The 2013 American College of Gastroenterology guidelines reported that micronutrient deficiencies (in particular iron, folic acid, vitamins B6 and B12, vitamin D, copper, and zinc) are frequent in celiac patients at the time of celiac diagnosis. Therefore, patients with newly diagnosed celiac disease, micronutrient deficiencies should be found and integrated. These tests should include iron, folic acid, vitamin D, vitamin B12 and more [7].

Following the United Kingdom 2015 National Institute for Health and Care Excellence guidelines, it was reported that some patients with celiac disease may need additional nutritional supplements, mainly in the early stages after diagnosis, suggesting, however, that this should be identified through an appropriate ongoing monitoring and that integration should begin after a full evaluation [8].

These two guidelines are derived from, and in agreement with, the more recent reviews demonstrating that in celiac patients, at time of diagnosis, nutritional deficiencies are often found in vitamins and minerals, such as folic acid, vitamin B12, vitamin D, calcium, magnesium and zinc.

However, at the same time, in subjects undergoing GFD for a long time with good compliance, it has been described that micronutrient deficiencies may persist due to an inadequate full reintegration of the mucous membrane [9]. Some patients with long-term treated CD may still have abnormal small bowel mucosa and persistent villous atrophy on follow-up, with or without ongoing or recurrent symptoms, despite an apparently GFD [4,10]. According to Lanzini et al., the complete recovery of duodenal mucosa with histological normalization, after a median 16 months GFD in patients diagnosed at an adult age occurs only in 8% of cases [11]. The majority of adult patients achieving remission with intraepithelial lymphocytosis (65%) and a substantial proportion showing no-change (26%) or deterioration (1%) of duodenal histology [11]. This condition seems to be more common in adults older than an age of 50 years [12], but occurs even in 19% of children who underwent follow-up biopsy at least 1 year after starting the GFD [13]. When other possible causes of villous atrophy are excluded, refractory celiac disease is diagnosed [4]. Even in the absence of symptoms this condition is not positive, because it may predispose to severe complications, such as osteoporosis and malignancy [14].

Moreover, gluten-free products are usually low in some micronutrients, such as magnesium and folic acid, and gluten-free cereals found in nature have a lower magnesium content compared with gluten-containing ones [9].

This topic is highly debated in the literature. In fact, there is a widespread agreement on the importance of supplementation at the time of diagnosis, but there is still no consensus for when and what additional nutrients are needed in subjects on long-term GFD (LTGFD).

Given this background, the aim of this narrative review is to evaluate the literature that considers blood nutritional deficiencies in celiac subjects on LTGFD therapy with good compliance (LTGFDWGC) in order to prepare dietary supplementation advice for these patients.

## **2. Materials and Methods**

The present narrative review was performed following the steps by Egger et al. [15] as follows:


## **3. Results**

This review included 73 eligible studies and the dedicated flowchart is shown in Figure 1.

Table 1 shows the reviews made on nutrient deficiencies in celiac patients at time of diagnosis and after LTGFDWGC.

Table S1 shows the studies concerning circulating levels and supplementation of micronutrients in celiac patients after LTGFDWGC.

The literature shows that nutritional deficiencies, considered by evaluating the blood values of these micronutrients, in celiac subjects on LTGFD with good compliance, relate to vitamin B12, folic acid, vitamin D, calcium, iron, magnesium, zinc, selenium, thiamine, riboflavin, niacin and vitamin K (Table 1).

**Figure 1.** Flowchart of the study.

**Table 1.** Reviews on nutrient deficiencies in celiac patients at time of diagnosis and after GFD.



**Table 1.** *Cont*.

## *3.1. Vitamin B12*

This research was carried out based on the keywords "vitamin B12" AND "supplementation" AND "long-term GFD with good compliance" AND "celiac patient" OR "celiac disease". Of the 13 studies that were taken into account, 6 were review-type papers, 3 were prospective studies, 2 were observational studies and 2 were randomized controlled trials.

The absorption of dietary vitamin B12 occurs mainly in the terminal ileum through an active, specific and saturable transport mechanism. Vitamin B12 is released from food proteins after exposure to gastric acid. Vitamin B12 links to a salivary and gastric R protein; then pancreatic proteases destroy the R protein in the duodenum, releasing cobalamin which creates a complex with intrinsic factor (IF) that is secreted by the parietal cells in the stomach. The complex B12-FI migrates up to the terminal ileum aided by intestinal peristalsis, and binds itself through its proteic fraction to a specific cellular receptor. The complex dissociates and cobalamin enters the enterocytes of the small intestine. When the vitamin is administered orally in high doses, a small proportion along the entire intestine is absorbed through a passive diffusion mechanism.

Absorption site remains relatively preserved in patients with CD, so deficiency of vitamin B12 should be unusual. Nevertheless, numerous studies have shown that circulating levels of this vitamin are inadequate in about 5–40% of patients with CD at diagnosis [19–22] and in about 2.9-41% of patients following a GFD [20,21].

A real link exists between CD and vitamin B12 deficiency, but it has not been established. Some studies have shown that GFD and, where required, supplementation with vitamin B12 is effective in resolving neurological complications associated with deficiency of this vitamin. It has been shown that concentration of vitamin B12 tends to normalize in patients with a LTGFD [17,23].

However, there is evidence that supplementation may also be useful in subjects undergoing GFD. Hallert et al. [24] conducted a double-blind study to evaluate the effects of supplementation with B vitamins in adult CD patients for a long time, which involved daily administration of 0.8 mg of folic acid, 0.5 mg of cyanocobalamin, and 3 mg of pyridoxine for a period of 6 months. In these patients, there was improvement of psychiatric symptoms, and a significant return to normal vitamin B12 values with reduction of homocysteine values, which is often elevated in patients with vitamin B12 deficiency. Indeed, the catabolism of homocysteine requires vitamin B12 and folate. Consequently, hyperhomocysteinemia may reflect a deficit of both nutrients [25]. Great attention to the levels of homocysteine is needed in patients with CD. Celiac patients appear to have an increased risk of venous

thromboembolism and vascular disorders [26] and high levels of homocysteine is a risk factor for these chronic diseases [27]. Supplementation of vitamin B12 and folate tends to decrease homocysteine values [24], so it could represent a prevention behavior.

In some patients, it is therefore necessary to integrate this element, even when following a strict GFD. In such cases, administration of vitamin B12 can be given via the oral or intramuscular routes.

In a study carried out by Bolaman et al. on general populations with megaloblastic anemia due to deficiency of cobalamin, oral supplementation was as effective as intramuscular treatment. Oral administration seems to be less costly and more tolerable than intramuscular delivery [28]. Furthermore, a review carried out by Vidal-Alaball et al. on general populations showed that in patients with a deficiency of vitamin B12, oral administration of 2000 mcg/day or 1000 mcg/day, followed by 1000 mcg/week and then 1000 mcg/month, can be as effective as intramuscular administration in showing improvement in hematological and neurological levels [29].

In patients with CD, supplementation is recommended for those in which there remains a blood deficiency despite GFD. Hallert et al. have shown how the oral administration of 500 mcg of cyanocobalamin in subjects undergoing GFD is effective in restoring the homocysteine value (which is an indirect measurement of vitamin B12 and folate). It suggests that the absorption after oral administration, especially in subjects undergoing GFD, is effective [24]. Furthermore, Theethira et al. suggested measuring vitamin B12 levels at diagnosis and then every 1–2 years for symptoms, and to treat with 1000 mcg orally until levels normalize, and then considering daily gluten-free multi vitamin/mineral supplementation [30].

In conclusion, considering the site of absorption (terminal ileum) of vitamin B12, which remains relatively preserved in patients with CD, deficiency of this vitamin should be infrequent; however, circulating levels of this vitamin could remain inadequate up to 41% in LTGGFDWGC patients. Given this background, in addition to its pivotal role in preventing hyperomocysteinemia, an annual routine follow-up of blood vitamin B12 level is mandatory in subjects undergoing LTGFD. Regarding dose and route of administration, the literature showed that in celiac patients with vitamin B12 deficiency, oral administration of 1000 mcg of vitamin B12 until levels normalized, followed by daily gluten-free multi-vitamin/mineral supplementation with 500 mcg of vitamin B12 is effective [30].

## *3.2. Iron*

This research was carried out based on the keywords "iron" AND "supplementation" AND "long-term GFD with good compliance" AND "celiac patient" OR "celiac disease". Of the 21 studies that were taken into consideration, 8 were prospective studies, 5 were reviews, 3 were observational case studies, 2 were case control studies, 1 was a randomized controlled trial, 1 was a report and 1 was a guidelines.

Iron deficiency often occurs in celiac patients, and it is followed in many cases by iron deficiency anemia. Studies have shown that the prevalence of this event in patients with newly diagnosed CD seems to be between 10–80%. The prevalence of deficiency after 6 months of GFD is about 70%, after 1 year, it is about 50%, and after 2 years, it is about 40% [17,19,31–37].

Iron is an essential trace element, being part of the heme structure, the non-protein component of numerous iron proteins (such as hemoglobin, myoglobin and cytochromes). Its excretion cannot be controlled, so the amount of iron in the body depends mainly on its absorption, which takes place in the duodenum and proximal jejunum.

Iron deficiency anemia in CD patients mainly arises from malabsorption, although the possibility of intestinal bleeding cannot be excluded and must be considered [38,39].

In the general population, initial treatment of iron deficiency should be continued until hemoglobin and iron stores are normalized. This goal is usually obtained with oral iron administration. Although it is occasionally recommended to take iron supplements before breakfast in order to increase the absorption, this significantly reduces the tolerance. For this reason, it seems reasonable to suggest the administration with food. All ferrous salts, including ferrous fumarate, ferrous lactate, ferrous succinate, ferrous glutamate, and ferrous sulphate, share similar bioavailability. Preparations of iron glycinate represent a valid therapeutic alternative, since they have a good bioavailability and a lower frequency of side effects, such as constipation [40–42].

Treatment with oral iron is, in the general population, slow in reaching its goal, and good compliance is required to be successful. In addition to this, anemia is often severe, and a quick response is necessary. Sometimes the tolerance is poor, and in these situations the use of parenteral iron is fully justified. The efficacy and safety of parenteral iron sucrose use have been demonstrated in several clinical studies and have been confirmed by extensive clinical practice [43]. To supply the quantities required, several doses are needed. Other intravenous drugs such as ferric carboxymaltose have been introduced [44] and would require fewer infusions to provide the required dose. Ferric carboxymaltose is a robust and stable non-dextran intravenous iron formulation with the advantage of having a very low immunogenic potential, and therefore is not predisposed to anaphylactic reactions. Its properties permit the administration of large doses (15 mg/kg; maximum of 1000 mg/infusion) in a single and rapid session (15-min infusion) [45].

Therapy on the general population should start with a low dose and the intake should be constant until the iron deposits are not restored [40]. It is fundamental that treatment of an underlying cause should prevent further iron loss, but all patients should have iron supplementation, both to correct anemia and to replenish body stores. This is achieved most simply and cheaply with oral ferrous sulphate 200 mg twice daily. Lower doses may be as effective and better tolerated, and should be considered in patients not tolerating traditional doses. Other iron compounds (e.g., ferrous fumarate, ferrous gluconate) or formulations (iron suspensions) may also be tolerated better than ferrous sulphate. Oral iron should be continued for 3 months after the iron deficiency has been corrected so that stores are replenished [46].

Regarding celiac patients, in subjects in which iron supplementation is needed, it should be started orally. In the decision on when to start, some authors suggest undertaking supplementation in the moment in which the intestinal lesions are healed [40], while other studies suggest taking the supplement immediately at the time of diagnosis, without waiting for the healing of the mucosa [47]. In most of these patients, GFD is enough to solve the framework of anemia [48], although it may take a long time. In other cases it is necessary to help the patient with supplementation [36]. A study carried out on 25 pediatric patients with CD and iron deficiency showed good efficacy of oral administration of iron, (investigated with ferrous bisglycinate chelate 0.5 mg per kg body weight, reaching a maximum of 28 mg) both in patients with GFD and in those newly diagnosed [47]. A study carried out on celiac pediatric patients showed that the therapeutic dose in pediatric patients with an iron deficiency is 3 mg of elemental iron per kg body weight per day. The prophylactic dose in pediatric patients is 2 mg of elemental iron per kg body weight per day reaching a maximum dosage of 30 mg per day [36].

Since gluten-free products are characterized by a low iron content [49], the intake of foods rich in this mineral, such as meat, should be recommended to patients initiating a GFD.

Theethira et al. suggested measuring serum iron and ferritin at diagnosis, repeating every 3–6 months until ferritin was normal, and then every 1–2 years for symptoms. Moreover, they suggested iron supplements (325mg), 1–3 tablets based on initial ferritin level until iron stores are restored, and consideration of intravenous (IV) iron for severe symptomatic iron deficiency anemia or intolerance of oral iron.

In conclusion, it seems that when long GFD, including food tips to consume adequate dietary iron, is not enough to restore iron levels (40% of LTGFDWGC subjects are iron-deficient), the right approach could be to start with oral administration of iron.

Based on this background, a semi-annual steady and routine follow-up of blood iron and ferritin levels is mandatory in subjects undergoing LTGFD [30].

## *3.3. Folic Acid*

This research was carried out based on the keywords "folic acid" AND "supplementation" AND "long-term GFD with good compliance" AND "celiac patient" OR "celiac disease". A total of 17 studies were taken into consideration. Among these studies, 7 were prospective studies, 5 were review studies, 2 were observational case studies, 2 were randomized controlled trials, and 1 was a report.

The term "folate" describes the vitamer group B, based on the main structure of folic acid, which shares the same vitamin activity. This group of vitamins is essential for the synthesis and repairing of DNA, and they also act as cofactors for the enzymes involved in several biological reactions. Folate occurs naturally in some foods, and its synthetic form, folic acid, is added to many food products to increase the dietary intake. An example of supplementation is the addition of folic acid to wheat flour, which has been introduced in 52 countries worldwide since 2007 [50].

Folate deficiency was detected in about 10–85% of adult patients with CD at diagnosis, and in about 0-20% of patients following a GFD [17,20,22,51–54]. Usually the folate deficiency in CD occurs in patients with lesions in the ileum [48,55,56].

Tighe et al. compared in the general population the effectiveness of 0.2 mg folic acid per day with that of 0.4 and 0.8 mg/day in lowering homocysteine concentrations over a 6-month period. It has been seen that folic acid significantly reduces the risk of stroke overall by 18%, but to a greater extent by up to 25% in those trials that showed greater homocysteine lowering or in persons with no history of stroke. The lowest dose of folic acid required to achieve effective reductions in homocysteine is controversial but important for food fortification policy given recent concerns about the potential adverse effects of overexposure to this vitamin. This study supports the potential benefit of enhancing folate status and/or lowering homocysteine in the primary prevention of stroke. The authors suggest that a folic acid dose as low as 0.2 mg per day can, if administered for 6 months, effectively lower homocysteine concentrations [57].

Numerous studies have shown that a GFD would be sufficient to normalize folate status [21,48,58], but one other [23] show that in a certain percentage of patients, folate levels remain low despite GFD maintained for over 10 years. A possible explanation for this phenomenon is the reduced content of folate in gluten-free foods as previously described by Thompson [49]. Another hypothesis is that in celiac patients genetic alterations of the proteins involved in absorption and metabolism of folate may be present [17].

Hallert et al. suggested providing patients with good information about folate-rich foods. They also recommended starting supplementation in patients who show blood deficiencies after GFD [17,23].

Dosage should be decided in relation to the initial value of the subject. In a study conducted by Hallert et al. on celiac patients, they administered 0.8 mg of folic acid leading to normalization of homocysteine values [24].

In conclusion, folate deficiency was detected in up to 20% of patients on LTGFDWGC. Given this background, a semi-annual routine follow-up of blood folic acid level is mandatory in these subjects. Dosage of supplementation should be decided in relation to the detected value in the subject. The literature suggests supplementation with 1 mg/day of folic acid for 3 months, followed by a reduction to 400–800 mcg/day [30] or with 0.8 mg of folic acid [24] in order to improve the poor folate status.

#### *3.4. Vitamin D and Calcium*

This research was carried out based on the keywords "vitamin D" OR "calcium" AND "long-term GFD with good compliance" AND "supplementation" AND "celiac patient" OR "celiac disease". A total of 18 studies were taken into account. Of these studies, 5 were review studies, 4 were prospective studies, 3 were case reports, 2 were observational studies, 2 were guidelines, 1 was a case-control study and 1 was a meta-analysis study.

## 3.4.1. Vitamin D

The cholecalciferol, or vitamin D3, can be synthesized in the basal layers of the epidermis starting from cholesterol, by the action of ultraviolet rays of sunlight, and this should be the main source of vitamin D for the body. Another source of vitamin D is food, from which the absorption occurs mainly in the terminal ileum.

Numerous studies have shown low levels of vitamin D in many untreated celiac patients. Vitamin D deficiency, investigated through blood value, was detected in about 8–88% of adult CD patients at diagnosis, and in about 0–25% of patients following a GFD [17].

Nevertheless, certain patients, mainly post-menopausal women, continue to present bone density levels below the normal range [59,60]. This seems to be partly due to lack of vitamin D1.

If GFD is not sufficient to bring the values in the normal range, supplementation of vitamin D and calcium is required [17,61–63]. The Endocrine Society guidelines recommend, for the general population, that serum levels of vitamin D are at least equal to 30 ng/mL, and that it is necessary to decide the dosage of supplementation in relation to the initial value of the subject [64].

A meta-analysis conducted on general populations by Shab-Bidar et al. shows that a significant increase in serum levels of vitamin D in adults is achieved with a dose of ≥800 UI/day, at least after 6–12 months of supplementation [65].

Consider that for the celiac patient, a commonly applied strategy in cases of serious vitamin D deficiency is to prescribe a "loading dose" (e.g., 50,000 UI/week for 8 weeks) followed by reduced doses, as shown by Duerksen in a case report study of a woman with CD [66].

In a study aimed at detecting the effects of calcium and vitamin D supplementation in celiac children by Muzzo et al., daily supplementation with 1000 mg of calcium and 400 UI of vitamin D for 24 months was shown to have beneficial effects on the bone mass of celiac patients in whole body and femoral neck measurements; however, these values did not reach the controls [67].

A recent study carried out by Zanchetta et al. suggests an intake of 1000–1500 mg/day of calcium in two or more divided intakes of dairy products and a dose of vitamin D necessary to maintain a blood level of 30 ng/mL [68].

Moreover, Theethira et al. suggested measuring vitamin D levels at diagnosis, repeating every 3 months until levels are normalized, and then every 1–2 years or for symptoms. If necessary, integrate with 1000 (or more-based serum level) UI/day or 50.000 UI weekly if level is <20 ng/mL.

In conclusion, blood vitamin D deficiency was detected in about 0–25% of patients following a LTGFDWGC [69].

## 3.4.2. Calcium

Calcium deficiency was detected in about 41% of adult patients with CD at the diagnosis [32] and 3.6% of treated children [70]. This seems to be due to malabsorption related to intestinal epithelial damage, but it could also be linked to a reduced expression of a protein regulated by vitamin D that controls the absorption of calcium [17,71]. Calcium absorption is impaired due to mucosal atrophy. Therefore, to avoid hypocalcemia, parathyroid hormone increases substantially (secondary hyperparathyroidism) and stimulates osteoclast-mediated bone degradation. Calcium is then obtained from the skeleton reservoir, but this high remodeling state can lead to osteopenia and osteoporosis, altering bone microstructure and increasing fracture risk [68].

In a study relating to the persistence of calcium deficiency despite a GFD, Kavak et al. undertook the analysis of reduced intake and absorption, rather than the percentage of shortage investigated by blood values in children patients after GFD [70]. The authors reported a reduction in calcium intake in about 76–88% of patients adhering to a GFD [69]. Pazianas et al. described a reduced fractional calcium absorption in adult celiac patients adhering to a GFD despite adequate calcium intake. Taken into account their reduced fractional calcium absorption, the authors concluded that their daily dose should be at least 1200 mg per day [72]. Larussa et al. showed that asymptomatic patients following a GFD for at least 2 years showed normal circulating serum calcium and parathyroid hormone (PTH) levels [73,74]. Zanchetta et al. showed a significant reduction of bone resorption parameters and PTH values, with a significant increase in serum calcium and vitamin D after GFD. Sategna-Guidetti et al. described significant improvement of bone mineral density values in newly diagnosed CD patients after 1 year of following a GFD [53,68].

Regarding supplementation, Zanchetta et al. suggested 1000–1500 mg/day in two or more divided intakes of dairy products. The authors concluded that calcium supplementation may be an option if the patient is not able or willing to fulfill the required intake through dietary means [68]. Theethira et al. found that more than 50% of patients consume less than the recommended daily intake of calcium. They recommended that CD patients undergo regular assessments with a dietitian and that the recommended intake of calcium, including supplementation should be 1200–1500 mg/day [30].

## *3.5. Other Micronutrients*

This research was carried out based on the keywords "zinc" OR "magnesium" OR "selenium" OR "vitamin K" OR "thiamine" OR "riboflavin" OR "niacin" AND "supplementation" AND "long-term GFD with good compliance" AND "celiac patient" OR "celiac disease". Of the 15 studies that were taken into account, 9 were review studies, 2 were report studies, 2 were prospective studies, 1 was an observational study, and 1 was a case-control study.

## 3.5.1. Zinc

Zinc deficiency was detected in more than 50% of adult patients with CD at diagnosis, and between 0–40% of patients following a GFD [17]. The lack of this mineral seems to be linked in part to the its reduced absorption, due also to the degree of inflammation of the mucosa.

In the review by Theethira et al., they proposed to measure the serum zinc levels of CD patients at diagnosis and repeat after 3 months until the level is normal, followed by every 1–2 years for symptoms. They also suggested zinc supplementation between 25–40 mg/day until zinc levels were normal [30].

#### 3.5.2. Magnesium

Magnesium deficiency was detected in about 21.4% of adult patients with CD at the diagnosis, and a similar percentage (19.6%) in patients following a GFD [75].

This deficiency can be explained by malabsorption, but GFD may also lead to possible nutrient deficiencies because gluten-free products are usually lower in magnesium, and gluten-free cereals found in nature have a lower magnesium content compared with gluten-containing ones [9].

Furthermore, it has been seen that the resolution of mucosal inflammation may not be sufficient to resolve the shortage of magnesium in celiac patient. The deficiency may also be linked to a reduced intake of this mineral.

Breedon reported that some CD patients need additional magnesium supplement of 200–300 mg/day in the form of magnesium oxide or magnesium chloride, while others can improve magnesium levels through dietary means [76].

## 3.5.3. Selenium

Selenium deficiency is particularly remarkable because a GFD leads to its absence in cereal foods such as wheat and its derivatives, which are a source of selenium [77]. There are no sufficient literature data to describe a percentage of deficiency investigated through the blood level.

Reduced concentrations of selenium in whole blood, plasma, and leucocytes might develop in several ways. Firstly, GFD might contain a reduced amount of selenium compared with a normal diet, and, secondly, there might be malabsorption of selenium even when the patient is clinically well. Between the extraintestinal symptoms associated with CD, autoimmune thyroid diseases are more evident, underlining that CD-related autoimmune alterations can be modulated not only by gluten but also by various concurrent endogenous (genetic affinity, over-expression of cytokines) and exogenous (environment, nutritional deficiency) factors. The thyroid is particularly sensitive to selenium deficiencies because selenoproteins are significant in biosynthesis and activity of thyroid hormones, while other selenoproteins, including glutathione peroxidase are involved in inhibiting

apoptosis. Thus, selenium malabsorption in CD patients can be considered a key factor directly leading to thyroid and intestinal damage [78].

Studies have shown that in celiac patients, selenium supplementation between 120–200 mcg/day is within a safe range. It is important, however, not to exceed the tolerable upper limit of 400 mcg/day for selenium, as this can lead to gastrointestinal upset, hair loss and nerve damage [79].

## 3.5.4. Vitamin K

Vitamin K deficiency, investigated by markers like PIVKA-II or by prothrombin times, was detected in about 25% of adult patients with CD at the diagnosis, and it seems to return to acceptable levels in almost all patients following a GFD [72,80].

There are limited available data that relate the role of vitamin K and bone health in children and adults with CD. Pazianas et al. examined vitamin K status in children newly diagnosed with CD using prothrombin times as a marker of vitamin K status and found that approximately 35% of children were lacking this marker [72]. However, this may have been an underestimate of the prevalence of vitamin K deficiency as prothrombin time is a very insensitive marker of overall vitamin K status [72]. More sensitive markers of vitamin K status include serum levels of PIVKA-II, which is a vitamin K-dependent protein [80].

In the study carried out by Mager et al., over 25% of children were vitamin K deficient at diagnosis, investigated by PIVKA-II (which is a protein increasing in vitamin K absence), but it resolved in all children after 1 year. This seems to be due in part to improvements in vitamin K intake on the GFD. However, the remaining one-third of children and adolescents continued to have vitamin K intakes considerably lower than the adequate intake on the GFD [80].

Suboptimal dietary intake of vitamin K is common in this population, including when on a GFD. There are no sufficient literature data to recommend a specific dose of supplementation in celiac patient. Therefore, careful consideration should be given to routine supplementation of this nutrient at time of diagnosis of CD.

#### 3.5.5. Niacin, Riboflavin and Thiamin

Deficiency of other micronutrients, like niacin, riboflavin and thiamin have been described in several reviews at the time of diagnosis, although in the literature there is no accurate data on the percentage of celiac patients with such deficiencies [2,16,18]. Deficiencies of niacin and riboflavin may persist after a GFD [2,18]. Regarding thiamin, a study carried out by Shepherd et al. found that the inadequacy of thiamin was more common after GFD implementation than at time of diagnosis [81]. This can be explained by the fact that many gluten-free cereal products do not provide the same levels of thiamin, riboflavin, and/or niacin as enriched wheat flour products. As a result, a GFD that routinely includes gluten-free cereal products could be deficient in one or more of these nutrients, especially if these foods are, in large part, refined and unenriched [82].

Although there is insufficient data in the literature to recommend a dose for supplementation, it is considered useful to carry out control of blood values after diagnosis and after a period of GFD.

#### **4. Discussion**

It is evident from the analyzed reviews that there emerges an attention towards nutritional deficiencies that occur in celiac patients after an even longer period of GFD with good compliance. The suggested course of action is a half-yearly routine search in patients on LTGFD nutritional deficiencies, such as low levels of folic acid, vitamin B12, vitamin D, calcium, iron, zinc, selenium and magnesium and the need to establish a personalized supplementation plan, following patients over time, to avoid stopping of integration once the values are returned to their normal range, as shown in Table 2.


**Table 2.** Supplementation of nutrients in generic state deficiency and in celiac patient.

\* Therapy in literature in generic state deficiency \*\* Therapy in literature in celiac patient.

To help patients reduce deficiencies of minerals (calcium, phosphorus, sodium, potassium, chloride and magnesium) and trace elements (iron, zinc and selenium) it is important to advise them to introduce into their eating habits pseudo-cereals, in which the content of these elements can be twice as high as in other cereals. For example, in teff, iron and calcium contents (11–33 mg/100 g and 100–150 mg/100 g, respectively) are higher than those of wheat, barley, sorghum and rice [18].

It seems important to explain to patients that nutritional education and dietary supplementation should become part of the therapeutic process, which must last a lifetime.

## **5. Conclusions**

In conclusion, if correct GFD is not enough, and the blood levels of micronutrients remain low, it is mandatory to start with personalized supplements. In this case, it would be helpful to evaluate the

initial blood level to determine the right dosage of supplementation and repeat the examinations to keep under control values.

In any case, there are a lot of unresolved questions regarding the causes and the mechanisms that lead to these nutritional deficiencies. Further studies are absolutely required for the detailed understanding of this topic.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/1010-660X/55/7/337/s1, Table S1: Original articles concerning circulating levels and supplementation of micronutrients in celiac patients.

**Author Contributions:** Conceptualization, M.R.; methodology, S.P.; data curation, G.P., M.A.F., M.N., A.R., M.N. and C.G.; writing—original draft preparation, M.R.; writing—review and editing, T.A.A., S.P. and G.P.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

## **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Review* **Non-Celiac Gluten Sensitivity: A Review**

## **Anna Roszkowska 1,\*, Marta Pawlicka 1, Anna Mroczek 1, Kamil Bałabuszek <sup>1</sup> and Barbara Nieradko-Iwanicka <sup>2</sup>**


Received: 2 February 2019; Accepted: 22 May 2019; Published: 28 May 2019

**Abstract:** *Background and objectives:* Grain food consumption is a trigger of gluten related disorders: celiac disease, non-celiac gluten sensitivity (NCGS) and wheat allergy. They demonstrate with non-specific symptoms: bloating, abdominal discomfort, diarrhea and flatulence. Aim: The aim of the review is to summarize data about pathogenesis, symptoms and criteria of NCGS, which can be helpful for physicians. *Materials and Methods:* The PubMed and Google Scholar databases were searched in January 2019 with phrases: 'non-celiac gluten sensitivity', non-celiac gluten sensitivity', non-celiac wheat gluten sensitivity', non-celiac wheat gluten sensitivity', and gluten sensitivity'. More than 1000 results were found. A total of 67 clinical trials published between 1989 and 2019 was scanned. After skimming abstracts, 66 articles were chosen for this review; including 26 clinical trials. *Results:* In 2015, Salerno Experts' Criteria of NCGS were published. The Salerno first step is assessing the clinical response to gluten free diet (GFD) and second is measuring the effect of reintroducing gluten after a period of treatment with GFD. Several clinical trials were based on the criteria. *Conclusions:* Symptoms of NCGS are similar to other gluten-related diseases, irritable bowel syndrome and Crohn's disease. With Salerno Experts' Criteria of NCGS, it is possible to diagnose patients properly and give them advice about nutritional treatment.

**Keywords:** non-celiac gluten sensitivity; irritable bowel disease; gluten; FODMAP; wheat allergy

## **1. Introduction**

Wheat, rice and maize are the most commonly consumed grains worldwide. These products are rich sources of starch—the basic dietary component for the growing human population [1]. Wheat contains gluten. In 1953 Dickie, van de Kamer and Weyers published a study confirming malabsorption after wheat consumption in patients with celiac disease (CD) [2]. Nowadays, gluten intake is considered to be the trigger of gluten related disorders (GRDs). In GRD, the gluten-free diet (GFD) is principal, effective and yet the only treatment method. The gluten-free market is still rising, not only because of growing interest and public awareness of GRDs, but also due to celebrities touting this diet by for weight loss and athletes for improved performance [3], which is debatable as grains should be the main source of energy in the human diet.

## **2. Materials and Methods**

Standard up-to-date criteria were followed for review of the literature data. A search for English-language articles in the PubMed database was performed. The PubMed and Google Scholar databases were searched in January 2019 with phrases: 'non-celiac gluten sensitivity', non-celiac gluten sensitivity', non-celiac wheat gluten sensitivity', non-celiac wheat gluten sensitivity', and gluten sensitivity'. More than 1000 results were found. A total of 67 clinical trials published between 1989 and 2019 was scanned. After skimming abstracts, 66 articles were chosen for this review including 26 clinical trials.

## *2.1. Gluten Related Disorders (GRDs)*

The term "gluten intolerance" includes three different conditions: CD, allergy to wheat (WA) and non-celiac gluten sensitivity (NCGS) [4]. To date, CD and WA comprise for the best known and studied entities, which are mediated by immune system [1]. WA—classified as a classic food allergy is induced by wheat (not only gluten) intake that leads to type I and type IV hypersensitivity. The crucial role in WA disorder play IgE immunoglobulins [1,5]. CD is an autoimmune disease occurring in genetically susceptible individuals with HLA-DQ2 and/or HLA-DQ8 genotypes. CD is characterized by the presence of specific serological antibodies such as: anti-tissue transglutaminase (tTG) IgA, anti-endomysium IgA (EMA) and anti-deamidated gliadin peptides IgG (DPG) [1]. There were reported cases of patients with gluten sensitivity in which allergic and autoimmune mechanisms could not be identified. They were collectively described as NCGS [1]. The NCGS or "non-celiac wheat sensitivity" (NCWS) has been a topic of interest in recent years. This trend is associated with a large number of studies concerning the syndrome [6,7]. The term NCWS is more adequate because of components other than gluten, that may contribute to intestinal and extra-intestinal symptoms [6]. In 1980, Cooper et al. described intestinal gluten-sensitive symptoms in 8 patients in whom CD was ruled out [8]. Further studies led to the definition of NCGS. NCGS is a condition characterized by clinical and pathological manifestations, related to gluten ingestion in individuals in whom CD and WA have been excluded [1,6,9,10]. Leccioli et al. described NCGS as a multi-factor-onset disorder, perhaps temporary and preventable, associated with an unbalanced diet [11].

Interestingly, II MHC haplotype HLA-DQ2 and HLA-DQ8 typical for CD is present only in about 50% of NCGS patients [1]. The main features of GRDs are summarized in Table 1.


**Table 1.** Comparison of prevalence, pathogenic, and diagnostic features of gluten related disorders (GRDs); non-celiac gluten sensitivity (NCGS), IgA anti-EMA (IgA antibodies against endomysium), IgA anti-tTG (IgA antibodies against transglutaminase), IgG anti-DGP (IgG antibodies against deamidated gliadin peptides).

\* Marsh classification.

## *2.2. Epidemiology of Gluten Related Disorders (GRDs)*

CD morbidity, based on serological results, is estimated to be 1.1% to 1.7% worldwide [12,13]. WA among children occurs with a frequency of 0.4–9% [5,14]. Due to an absence of diagnostic markers and population studies, the prevalence of NCGS is not well established [5,6]. Although studies have been conducted by several authors, this problem is still insufficiently explored. Previous data were based primarily on questionnaires for self-reported gluten sensitivity SR-GS/self-reported NCGS. According to several authors, the NCGS prevalence is from 0.6% up to 13% of the general population [15–19]. NCGS was reported more often among women [16–18], adults in the fourth decade of life [19,20] and individuals coming from urban area [18]. Among intestinal symptoms the most frequent in NCGS are: bloating, abdominal discomfort and pain, diarrhea and flatulence. The most common extra-intestinal symptoms were: tiredness, headache and anxiety [15,16,18,20]. Differentiation between NCGS and functional gastrointestinal (GI) disease—mainly irritable bowel syndrome (IBS)—may be difficult as some of the above-mentioned symptoms overlap with IBS manifestations. Van Gils et al. pointed that 37% of self-reported gluten sensitivity individuals (SR-GS) fulfilled the Rome III criteria for IBS, in contrast to 9% prevalence in the control group [18]. Similar findings were reported by Carroccio et al. IBS symptoms were reported in 44% self-reported NCWS [15]. According to research conducted by Cabrera et al., IBS, eating disorders and lactose intolerance were present more often in SR-GS individuals than in non-SR-GS group (14.3% vs. 4.7%) [16]. Herein, discussed studies indicate that SR-GS/SR-NCGS may correlate with more frequent occurrence of IBS, comparing to the general population. However, the German Society of Allergology and Clinical Immunology emphasized that the publications about NCGS suffer from certain weaknesses: absence of validated diagnostic criteria, suitable biomarkers, frequent self-diagnosis and unconfirmed etiology of reported symptoms. Thus, the prevalence of NCGS cannot be clearly established [21].

## *2.3. Gluten*

Gluten is defined as a family of proteins found in grains (wheat, rye, barley, oats). It includes two main proteins: gliadin and glutenin. Also, similar proteins such as secalin in rye, harden in barley and avenues in oats contribute to the definition of 'gluten' [22]. Gluten proteins are characterized by high proline and glutamine content, moreover, they are resistant to proteolytic enzymes in the gastrointestinal tract. In some individuals these peptides can cross the epithelial barrier and activate immune system: trigger an allergic (WA) or autoimmune response (CD) [5]. Incomplete digestion leads to significant changes in human gut and causes intestinal or extra-intestinal symptoms. Gliadin and other gluten proteins stimulate T-cells. Some authors suggested that amylase-tripsin inhibitors (ATIs) and fermentable oligo-, di-, and mono-saccharides and polyols (FODMAPs) may be associated with NCGS [11]. Another wheat constituent, known as agglutinin-carbohydrate binding protein and exorphins seem to influence immune system and induce damage of intestinal epithelium [11,22].

## *2.4. Amylase-Tripsin Inhibitors (ATIs)*

ATIs are albumin proteins found in wheat representing up to 4% of total proteins in grains [1]. They are highly resistant to intestinal proteases [1] and may induce release of pro-inflammatory cytokines from monocytes, macrophages and dendritic cells through activation of a toll-like receptor-4 in CD and NCGS patients [1,22]. ATIs may provoke activation of innate immune cells and intestinal inflammation [21]. ATIs activate immunological system through effect on toll-like receptor-4 in CD, that was confirmed in the research conducted by Junker et al. on mice deficient in TLR4 or TLR signaling [23]. Authors observed, that their mice models were protected from intestinal and systemic immune responses during oral ATIs intake [23]. Scientists also confirmed, that ATIs stimulate monocytes, macrophages and dendritic cells *in vitro* to produce IL-8, IL-12, TNF, MCP-1 and Regulated on Activation, Normal T-cell Expressed and Secreted (RANTES) [23].

## *2.5. Fermentable Oligo-, Di- and Mono-Saccharides and Polyols (FODMAPs)*

FODMAPs are short-chain sugars with less than 10 carbon atoms in the molecule [24]. The attention of scientists in recent years was drawn to the potential contribution of FODMAPS to pathogenesis of gastrointestinal disorders [25]. The scientists from Monash University in Australia conducted thorough analysis of a group of carbohydrates, which, despite their different structures, produced similar postprandial effects. The most prevalent forms of FODMAP include: fructooligosaccharides (FOS), galactooligosaccharides (GOS), lactose, fructose, polyols, sorbitol and mannitol. Barrett et al. created a list of food products that are good sources of FODMAP (Figure 1) and poor in short chain sugars (Figure 2) [24].

#### *Medicina* **2019**, *55*, 222

**Figure 1.** List of products being the source of fermentable oligo-, di-, and mono-saccharides and polyols (FODMAPs).

**Figure 2.** List of products low in FODMAPs.

Compounds belonging to the FODMAP group are not digested nor absorbed in the gastrointestinal tract. They have a strong osmotic effect and undergo rapid fermentation in the intestines, resulting in intestinal liquefaction, excessive gas production, bloating and pain. They may cause or exacerbate symptoms in susceptible patients with inflammatory bowel disease and irritable bowel syndrome (IBS) [24,25]. Numerous studies have confirmed the improvement in patients suffering from ulcerative colitis, Crohn's Disease and IBS following the elimination of short-chain sugars from the diet [26].

Wheat is a rich source of gluten and also contains large amounts of FODMAPs, which play a key role in NCGS development [27]. Some researchers suggest that diet low in FODMAP is beneficial for NCGS patients [25].

Considering the above research results, scientists are leaning towards renaming NCGS to a more recent NCWS [27]. It should be emphasized that a diet poor in FODMAPs should not be used without medical indications, as healthy people do not benefit from such diet [24]. Moreover, it was proven that FOS and GOS, compounds belonging to FODMAPS, alike prebiotic, favor proper colonization of intestines with *Bifidobacteria* and *Lactobacilli* bacteria and limit the proliferation of *Bacteroides* spp., *Clostridium* spp. and *Escherichia coli*. There is evidence that short-chain fatty acids (SCFA)—the product of FODMAP fermentation—have protective properties against colorectal cancer [24,27]. FODMAPs are believed to have a positive effect on lipid metabolism by lowering serum cholesterol, triglycerides and phospholipids [27]. In addition, this diet leads to calcium absorption disorders, lowering its serum levels. People resigning from products that are the source of FODMAP are at risk of vitamin and antioxidants deficiency [27,28]. Therefore, it is suggested to supplement vitamins, pro- and prebiotics when switching to the low FODMAPs diet [24,27].

## *2.6. The Salerno Experts' Criteria of NCGS*

As long as the NCGS biomarker is not available, certain limitations are included in two-step diagnostic protocol introduced in 2015. However, up to date The Salerno Experts' Criteria constitute the only accessible recommendations for diagnosis of NCGS. It should be emphasized that according to currently used criteria, NCGS should not be based only on exclusion diagnosis, which is new in comparison to the former practice [29]. Thus, the guidelines indicate the need of a standardized procedure: 6-week course of gluten-free diet—with the simultaneous, continuous assessment of symptoms and their intensity, followed by measuring the effect of reintroducing gluten after a period of treatment with GFD. A modified version of the Gastrointestinal Symptom Rating Scale (GSRS) was found to be applicable in terms of symptoms evaluation. Although limited, double-blind-placebo-controlled (DBPC) procedure remains to be the golden standard in NCGS investigation, yet, single-blinded procedure is allowed for the purposes of clinical practice [29–31]. The guidelines stress the importance of patient compliance, especially when it comes to shift to GFD, which should be discussed with a dietitian before implementation [29].

Back to the limitations—it is recommended to use gluten in the form of commonly consumed food products, during gluten challenge, rather than in the form of gluten capsules. Nevertheless, there is presumption that ATIs and FODMAPS—as the constituents of grains—interfere with the DPBC results [6,30,32]. Moreover, since the study on patients complaining about IBS-like symptoms, it was revealed that almost two-thirds of questioned patients presented nocebo effect after elimination diet, which seems to have same significant influence on performing DBPC during gluten challenge [29,33].

The fact that numerous symptoms manifested by active NCGS can be either vague or simply mimic other medical conditions, makes the diagnostic process long lasting and complex. For instance, bloating, abdominal pain, and irregular bowel movements are typical symptoms seen in IBS [20]. The overlapping symptoms of IBS, Crohn's disease and GRD are shown in Figure 3.

**Figure 3.** Overlapping symptoms in Crohn's disease, IBS and gluten-related disease.

The similarity between symptomatology of IBS and NCGS may lead to a wrong diagnosis and ineffective treatment [6]. The clinical case described by Vojdani and Perlmutter, presented a 49-year-old woman formerly diagnosed with IBS. The patient complained about abdominal pain, constipation, acid reflux and headache. Following conditions were contemplated and finally excluded: autoimmunological disorders, abnormal level of thyroid hormones, *H. pylori* infection [34]. Consequently, in the course of inappropriate therapy, the patient developed symptoms imitating systemic lupus erythematosus. Furthermore, the patient showed some improvement after corticosteroids administration, which appeared to be confusing for making right diagnosis as well [34]. Ultimately, after years of inappropriate treatment, the NCGS turned out to be the reason for patient's affliction. In addition, a few studies indicate that NCGS can be primary trigger for developing IBS. Virtually, as NCGS and IBS-like symptoms tend to overlap, the diagnostic process is particularly challenging [20,35,36].

Even though the diagnosis within the wide spectrum of bowel diseases was made, in the case of continuous therapy failure, it is crucial to reconsider NCGS as the possible cause. A clinical case of a patient with NCGS overlapping Crohn's disease has been reported. The onset of Crohn's disease is characterized mainly by unspecific symptoms, including diarrhea, weight loss, right lower quadrant abdominal pain, which proceed in a gradual way, with very harmful effects [37]. In the above-mentioned case report, the patient suffered from refractory Crohn's disease for 14 years and elevated IgG class antibodies directed against native gliadin (AGA) were detected, which shed a light on gluten related disorder. Introduction of GFD ceased diarrhea and enabled the patient to gain weight [34]. It is worth highlighting that NCGS patients are twice as likely to have AGA positivity [38].

At present, the linkage between gluten sensitivity, such as CD, and neurological disorders seems to be obvious. So far, numerous studies have unveiled extra intestinal symptoms affecting the peripheral and central nervous system due to celiac sprue. Although not fully understood yet, a wide range of NCGS neurological complications has been reported too. The state-of-the-art knowledge on NCGS revealed its association with transient and subtle cognitive impairment, being called "brain fog" [39]. Some scientists suggest NCGS to worsen symptoms in the context of depression but further examination must be performed to comprehend and determine NCGS relation with depressive disorders [32]. Busby et al. in their meta-analysis pointed out that standardization of methods measuring dietary adherence and mood symptoms is vital in terms of future research. Nevertheless, they admit that the gluten elimination diet may be an applicable treatment for mood disorders in patients suffering from gluten-related diseases [40].

It has not been until recently, when researchers explored that NCGS may be associated with gluten ataxia (GA), as the patients with typical GA symptoms did not meet criteria for CD diagnosis [41]. NCGS symptoms are believed to originate from an innate immune response. Interestingly, autoimmune diseases are reported to be more frequent in this group of patients, comparing to sheer IBS patients [42].

## **3. Results**

Comparison of selected clinical trials concerning NCGS is shown in Table 2. In the study by Capannolo et al. patients with CD and WA were excluded while in the study by Elli et al. patients without CD, WA, IBS were enrolled. The prevalence of NCGS, CD and WA among patients with functional GI symptoms in the study of Capannolo et al. was estimated to be 6.88%, 6.63% and 0.51%. Capannolo et al. indicate that high frequency of visits due to gluten-related symptoms is not associated with high prevalence of GRDs. Ellie et al. established that 14% of patients, suspected to have NCGS because of responding to gluten withdrawal showed a symptomatic relapse during the gluten challenge. It was highlighted that GFD can have a beneficial effect even in the absence of CD or WA. However, there are certain limitations seen in both of compared above research papers. The research of Capannolo et al. was lacking blindness in GFD challenge and missing evaluation of possible influence of other food components. Besides it was conducted before Salerno Criteria were introduced (2015). A choice of timing and gluten dosage shown in the research of Elli et al. was not in line with the timing suggested by Salerno criteria. In addition, the protocol did not make use of a scheduled diet besides GFD. Moreover, a nocebo effect may be presumed, in consistence with symptomatic deterioration observed in the placebo group. Other diet variables in both studies cannot be excluded (ATIs) [43,44].


**Table2.**Comparisonof selected researchesonNCGS.


Giessen Subjective Complaint List and VAS

therapy

**Table 2.** *Cont.*


In 2015, Zanini et al. published a prospective, randomized, double-blind, placebo-controlled study on patients without CD or wheat allergy as seen in Table 2. Scientists observed 35 patients (31 females and 4 males) being on a GFD due to their own initiative because of gastrointestinal symptoms they had had on a diet containing gluten. They were switched to a diet containing gluten. Participants' ability to distinguish between flours containing gluten and gluten-free was assessed, as well as their score in the Gastrointestinal Symptoms Rating Scale (GSRS). In order to participate in the study, patients had to be over 6 months a self-prescribed GFD and have a Gastrointestinal Symptoms Rating Scale (GSRS) below 4. The CD had had to be excluded before the start of the GFD. Before the beginning of the study t-TG antibody levels were measured and patients were instructed how to keep a diet diary. After 3 months, t-TG antibody level was checked again and GSRS questionnaire was performed. The participants received 10-g sachets containing gluten-free or gluten- containing flour labeled A or B. Patients were ordered to add contents to the pasta or soup for 10 days. Then for 2 weeks there was a washout period. Then, the patients received a second sachet with the other label, which they were to consume for 10 days. The primary outcome was the ability of the participants to correctly identify flour containing gluten. The study showed that only 34% (12 participants) correctly identified gluten- containing flour. Two thirds of the participants were not able to properly identify flour containing gluten. Almost half of the participants 17 (49%) misidentified gluten-free flour as gluten-containing flour, but those patients recorded symptoms and their GSRS scores increased on the flour not containing gluten. The gluten-free flour used in this test contained FODMAP [45].

Hollon et al. in their study (Table 2) disclosure ex-vivo gliadin effect on gut permeability in patients with active celiac disease (ACD), remission celiac disease (RCD) and gluten sensitivity (GS). The results of the research indicated that in all four groups, including control group (NG), there is certain response to gluten administration [46]. Researchers reported increased permeability particularly comparing ACD and GS groups to RCD, which is due to gluten induced alteration of intestinal barrier. Furthermore, researchers by means of quantification method investigated changes in following cytokines IL-6, IL-8, IFN-γ, TNF-α, which showed no significant difference, however, in this case a short period of incubation could implicate results. It should be emphasized that lack of blindness in GFD challenge while recruiting GS group along with lack of GFD challenge in the control group are important limitation in discussed study and could impact final results [46].

Shahbazkhani et al. investigated the relationship between dietary habits in IBS patients and consequent symptom fluctuations (Table 2). In particular researchers were interested in gluten impact on wellbeing of IBS patients and weather it may induce IBS-like symptoms. After rigid inclusion and exclusion criteria, strict six-week GFD 72 patients were recruited and divided into two groups: gluten containing group (study group), gluten free group (placebo group). Symptoms were analyzed by means of visual analogue scale (VAS). The results of the research revealed significant worsening of symptoms in a study group after gluten powder challenge. Scientists reported increase of overall symptoms such as satisfaction with stool consistency, tiredness, nausea, bloating in study group comparing to the control one. The results occurred to be statistically significant [47]. Nevertheless, there was limitation such as gluten form—a packet of 100 g powder, which is not recommended anymore by Salerno criteria [29].

According to the study published in Gastroenterology, scientists discovered that FODMAPs are another wheat antigen along with gluten triggering symptoms in patients with NCGS. Biesiekierski et al. conducted a double-blind crossover trial in which participated 37 patients suffering from NCGS and IBD. The following exclusion criteria were applied: age less than 16 years, CD confirmed by genetic tests and duodenal biopsy, alcohol abuse, chronic non-steroidal anti-inflammatory drugs (NSAIDs) and immunosuppressant treatment, uncontrolled psychiatric illness. Patients who had confirmed symptoms of IBS by accomplished the Rome III criteria and symptoms well controlled on a GFD were qualified for the study. Another requirement was to follow the GFD 6 weeks before this clinical trial. The first stage of the study was identical for all participants and the task was consuming for a one week a gluten-free and low FODMAPs diet. After a 2-week washout period, patients were

randomly assigned to the three groups: high-gluten, low-gluten and placebo, without introducing FODMAP into the diet. The symptoms of the patients were measured by using 100-mm VAS scoring and Daily-Fatigue Impact Scale (D-FIS). All participants were asked to return to the second stage of this study—trial in which all patients received each diets for 3-days [48]. Gluten-specific responses were found only in 8% of patients. Scientists found a high nocebo effect and reproducibility of induction of symptoms in each arm was low [48].

Biesiekierski et al. noticed that patients with NCGS do not present a statistically significant occurrence of symptoms after introducing gluten into the diet, if at the same time they limit products rich in FODMAP (Table 2). These results may suggest that the symptoms in patients suffering from NCGS may in many cases be associated with intolerance to the contained sugars, but not hypersensitivity to gluten. Surprisingly, the patients involved into study evinced eminently high VAS ratings for their symptoms, despite being on GFD. Furthermore, an anticipatory nocebo response could influence the final results of this DBPC research. It is interesting that all participants eventually returned to GFD at the end of the trail as they 'subjectively describe feeling better' [48].

Scientists from Oslo, Skodje et al., conducted a study in which took part 59 patients on a GFD, in whom CD was excluded (Table 2). Participants were divided into three groups: receiving diet including gluten (5.7 g), fructans (2.1 g) and placebo. The clinical trial lasted 7 days and was preceded by a 1-week washout period. The following symptoms were recorded: pain, bloating, diarrhea, constipation, nausea, dizziness, weakness, sleepiness and tiredness. Participants filled a questionnaire containing 13 questions about their gastrointestinal symptoms and filled VAS. The results were measured by GSRS, Irritable Bowel Syndrome scale (GSRS-IBS), VAS, Short Form-36 (SF-36) and Giessen Subjective Complaint List [49]. Scientists observed that daily symptoms calculated using VAS score were significantly higher in fructans diet. Furthermore, they noticed that overall GSRS-IBS was higher in the FODMAPs group (38.6 g) than in the gluten group (33.1 g) and placebo (34.3 g). More ailments were recorded in the group receiving fructans, compered to two another groups. In addition, it was demonstrated that a diet rich in FODMAPSs caused greater weakness and decreased vitality compared to the placebo and gluten groups. The results of the study indicate that FODMAPs are a trigger factor of gastrointestinal complaints in patients suffering from NCGS [49]. Thus, scientists are leaning towards renaming NCGS to a more recent NCWS [27].

Di Sabatino et al. observed increased severity of intestinal symptoms (abdominal bloating, abdominal pain) and extra intestinal symptoms (foggy mind, depression, and aphthous stomatitis) among subjects with suspected NCGS (excluded CD and WA). Although, this study did not make a significant contribution in development of knowledge about NCGS and had some weaknesses such as lack of a control group, it indicates possible symptoms experienced by NCGS patients (Table 2) [50].

In order to prove that gluten is a trigger factor in patients with NCGS, Rosinach et al. conducted a study in which 18 participants were assigned to gluten or placebo groups. In 10 out of 11 patients, symptoms worsened in response to a gluten-containing diet, 7 of which were withdrawn from the study due to the severity of the symptoms [51]. There was no early termination in the placebo group although in 2 participants symptoms were observed (Table 2) [51].

Carroccio et al. collected and analyzed data from 200 patients examined in previous study with diagnosed NCWS. Their findings are interesting because about 90% of patients who maintained wheat-free diet (WFD) were characterized by significant improvement of IBS symptoms [52]. The authors came to the conclusion that NCWS is a persistent condition and patients with NCWS should therefore be correctly identified and treated with WFD (Table 2) [52].

Roncoroni et al. conducted a study on dietary exposure to different amounts of gluten in patients meeting the criteria of the NCGS [53]. Researchers observed different reactions of patients after the introduction of gluten. Some of them had a worsening of well-being and increased symptoms after a small dose of gluten, others observed this effect after the medium dose and others only after a high dose of gluten (Table 2) [53].

Carrocio et al. in their study in 2011 emphasize the link between particular food ingestion and deteriorating symptoms in a subgroup of IBS patients [54]. It clearly shows alleviation of the symptoms in 22% of IBS patients—whose previous treatment was ineffective—after eliminating gluten from the diet. Moreover, researchers excluded association of DQ2 and DQ8 haplotypes with frequent gluten sensitivity, however, patients presenting food hypersensitivity (FH) to both wheat- and cow's milk-protein were reported to be often DQ2/DQ8 positive. Fecal eosinophil cationic protein (ECP) may be useful while identifying FH in IBS-patients (Table 3) [54].

Carroccio et al. in their study published in 2012 examined individuals with non-celiac WS, diagnosed by DBPC challenge with IBS-like symptoms, compared to CD patients and IBS patients [55]. Authors described presence of two types of WS subjects: WS similar to CD and WS associated with multiple food hypersensitivity. Besides, symptoms such as anemia, weight loss, self-reported wheat intolerance, coexistent atopy, and food allergy in infancy were noticed more often in WS compared to IBS controls. Furthermore, WS individuals were characterized by higher frequency of presence IgG/IgA anti-gliadin in serum, basophil activation (assessed by flow cytometric method) and histology specific eosinophil infiltration of the duodenal and colon mucosa. This study shows the differences between non-celiac WS and other gluten-related disorders (Table 3) [55].

Volta et al. in their study, assessed the level of immunoglobulin distinctive for CD in patients with GS comparing to CD [56]. They revealed that 50% of GS patients presented IgG AGA, whereas IgA AGA was seen only in a few patients in study group. Besides, researchers observed absence of IgA EmA, IgA tTGA, IgG DGP-AGA, which are typical for CD, within GS group (Table 3) [56].

Basing on a study group conducted by Volta et al., Caio et al. continued research on AGA IgG [38]. Scientists aimed to explore GFD impact on AGA IgG titer in AGA IgG positive patients (44 individuals) with NCGS. After six months of GFD AGA IgG disappeared in all the patients (Table 3).

Carrocio et al., in another research conducted in 2015, evaluated and described frequent ANA positivity within NCWS patients group [57]. The study demonstrated ANA positivity occurring along with DQ2/DQ8 haplotypes. As it was previously discussed, DQ2/DQ8 positivity is a distinctive feature of CD rather than NCWS. Thus, researchers highlight the need of intraepithelial intestinal flow cytometric pattern, which is an accurate method identifying seronegative CD patients, in the initial diagnostic biopsy. However, scientists found autoimmune diseases (AD) particularly frequent in study group. Autoimmune thyroiditis was reported to be the most frequent AD and amounted for 22% and 24% in retrospective and prospective groups respectively Table 3) [57].

Infantino et al. similarly to Volta observed frequent IgG AGA occurrence in NCGS patients, however, the author highlights that it is still lacking diagnostic accuracy. Nevertheless, in some cases, it can be helpful in the diagnostic process of NCGS patients [58].

Papers included in Table 3. Indicate IgG AGA and ECP to be helpful diagnostic tool while diagnosing NCGS. Still they have limited application in a large group of NCGS patients and cannot be widely used in NCGS diagnostic protocol [54,58].


**Table3.**Researcheson potentialNCGSbiomarkers.



## **4. Discussion**

Nowadays, a gluten-free diet is fashionable and is promoted by many celebrities. Many people undergo this fashion and despite lack of symptoms, try to reject gluten because they believe it may harm their health. In 2016, as much as USD 15.5 billion was spent on gluten-free food sales. This value is more than twice as high as in 2011. Lack of gluten in food consumed by people who tolerate it well may not bring favorable results.

In a study conducted by Norsa et al., children with CD were tested for at least one year on a GFD diet. As many as 34.8% of children on GFD diet had high concentrations of triglycerides on fasting, 24.1% high concentration of LDL cholesterol and 29.4% increased blood pressure. In 52 out of 114 participants there were available cards with information on blood lipids concentration before GFD introduction. 24% of children on GFD had had LDL cholesterol borderline values. That was much more than before the introduction of the diet (10%). However, these data did not meet the value of statistical significance (*p* = 0.09) [59].

Studies show that gluten may have a positive effect on triglyceride levels. In a clinical trial in which 20 adults with hyperlipidemia took part, a group with a balanced diet and a group with a high gluten content (78 g per day with an average human intake of 10–15 g) were studied. The high gluten diet group had a decreased triglyceride concentration of 19.2% (*p* = 0.0003) compared to the control group after one month of the study [60]. In another study, a group of patients consuming 60 g of gluten per day had a 13% (*p* = 0.05) lower triglyceride concentration compared to the control group [61]. In a study published in 2017, the estimated gluten consumption lead to the protective effect against cardiovascular disease (HR 0.85, 95% CI 0.77-0.93, *p* = 0.002) [62].

Gluten-free products can also be more than twice as expensive as regular products [63]. There are other disadvantages of GFD. The GFD turned out to be poor in trace elements and vitamins, such as zinc, iron, magnesium, calcium, vitamin D, vitamin B12, folate, and fiber [64,65]. Furthermore, Tovoli et. al. compared scores obtained by NCWS and CD individuals using quality of life questionnaire (CDQ) before GFD introduction and after at least one year. NCWS patients still reported intestinal and parenteral symptoms, although symptoms were significantly reduced in comparison to period before GFD. Therefore, other factors influencing NCWS should be investigated [66].

Finally, based on revised research results, it is clear that NCGS still remains to be the subject of uncertainty, especially in terms of other wheat components contribution to its symptoms. There are only a few published forms of research in the last six years. It should be stressed that it is hard to compare the results of each study as obtained methods and criteria significantly vary. Moreover, the timing of onset of each research was of a great importance as some of them were conducted before Salerno criteria were introduced, which led to many interpretations and qualification protocols of patients with NCGS-like symptoms. Further investigations and seeking for biomarkers would play key role in improving of the diagnostic process and patients' follow up.

## **5. Conclusions**


**Author Contributions:** Conceptualization, A.R., M.P., B.N.-I.; Methodology and resources A.M., K.B., B.N.-I.; Visualization A.M., M.P., A.R., K.B.; Writing—Review & Editing A.R., M.P., A.M., K.B., B.N.-I.; Software K.B.; Supervision—B.N.-I.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

## **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Review*
