FOCUS Erly feeding prctices seem to hve no impct on the risk of developing celic disese during childhood Primry Prevention of Celic Disese: Environmentl Fctors with Focus on Erly Nutrition by Ann Chmielewsk et l. Key insights Previously, the min evidence for the influence of erly nutrition prctices on the risk of celic disese cme from observtionl studies. Recently, the results from two lrge rndomized controlled studies (the PREVENTCD nd CELIPREV trils) were published. The im of these studies ws to evlute the effect of the timing of gluten introduction on the risk of celic disese in t-risk children. Contrry to previous thought, the timing of gluten introduction in n infnt s diet does not influence the risk of developing celic disese. Current knowledge Celic disese is n utoimmune enteropthic disorder in which dietry gluten nd relted prolmins ply mjor pthogenic role. Geneticlly susceptible individuls hrbor bckground in which vrints of humn leukocyte ntigens (HLA) hplotypes DQ2/DQ8 re the min predisposing fctors. The symptoms of celic disese rnge from symptomtic to gstrointestinl nd non-gstrointestinl presenttions of vrying severity. Celic disese ffects round 1% of the generl popultion in Europe, gretly ffecting ptients qulity of life nd incurring significnt cost burden on society. There hs lwys been gret del of inconsistency in the literture regrding the impct of the timing of gluten introduction on the risk of celic disese. Prcticl implictions The only vilble tretment for ffected individuls is strict, gluten-free diet. Until recently, specil emphsis ws plced on erly nutrition, nmely, the timing nd mode of gluten in- Fctor Brestfeeding s protective fctor Timing of gluten introduction Delyed gluten introduction Mode of delivery (cesren) Genetic bckground Supporting evidence No No No No Yes Fctors influencing the risk of developing celic disese: n updte troduction s preventive mesure. However, results from the PREVENTCD nd CELIPREV trils do not support the current recommendtions. The ge of the child t gluten introduction (between 4 nd 12 months) hs no effect on the prevention of celic disese. Gluten should only be introduced in line with generl recommendtions for strting infnt complementry foods. In children with no genetic predisposition for celic disese, the timing nd mode of gluten introduction does not influence disese risk. Recommended reding Szjewsk H, Shmir R, Chmielewsk A, et l: Systemtic review with met-nlysis: erly infnt feeding nd coelic disese updte 2015. Aliment Phrmcol Ther 2015;41:1038 1054. 2015 S. Krger AG, Bsel E-Mil krger@krger.com www.krger.com/nm
Gluten: Friend or Enemy? Published online: November 26, 2015 Primry Prevention of Celic Disese: Environmentl Fctors with Focus on Erly Nutrition Ann Chmielewsk Młgorzt Pieścik-Lech Hni Szjewsk Rnn Shmir b Medicl University of Wrsw, Wrsw, Polnd; b Schneider Children s Medicl Center, Sckler Fculty of Medicine, Tel Aviv University, Tel Aviv, Isrel Key Messges Celic disese (CD) is n utoimmune disese cused by gluten, which requires lifelong gluten-free diet. Brestfeeding does not prevent CD. The timing of gluten introduction into the infnt s diet does not influence the risk of developing CD. Genetic predisposition (presence of HLA-DQ2 nd/or DQ8) is the min fctor influencing the development of CD. Key Words Celic disese Environment Gluten Brestfeeding Nutrition Abstrct Celic disese (CD) is common utoimmune disorder cused by ingestion of gluten. When dignosed, it should be treted with lifelong, strict gluten-free diet. Erly infnt feeding prctices hve been suggested s mens of preventing CD. In the lst few decdes, observtionl dt hve suggested tht brestfeeding, especilly t the time of introducing gluten into the infnt s diet, s well s the time nd mode of gluten first being given to child could prevent or dely the occurrence of CD. As result, recommendtions dvised tht it is prudent to void both erly (<4 months) nd lte (>7 months) introduction of gluten, nd to introduce gluten grdully while the infnt is still being brestfed, s this my reduce the risk of celic disese, type 1 dibetes mellitus, nd whet llergy. Recently, the results of two lrge rndomized trils hve shown tht brestfeeding in generl, brestfeeding during gluten introduction, nd erly or delyed gluten introduction do not influence the totl risk of CD in geneticlly predisposed individuls. Introducing gluten t 4 versus 6 months in very smll mounts, or t 6 versus 12 months, resulted in similr rtes of CD in these children. Thus, erly feeding prctices seem to hve no impct on the risk of developing CD during childhood. In children without the genetic predisposition, the ge nd mode of gluten introduction do not influence the risk nywy. 2015 S. Krger AG, Bsel Bckground Celic disese (CD) is n immune-medited systemic disorder elicited by gluten nd relted prolmins in geneticlly susceptible individuls, which is chrcterized by the presence of vrible combintion of gluten-dependent clinicl mnifesttions, CD-specific ntibodies, E-Mil krger@krger.com www.krger.com/nm 2015 S. Krger AG, Bsel 0250 6807/15/0676 0043$39.50/0 Ann Chmielewsk, MD Deprtment of Peditrics Medicl University of Wrsw Dzildowsk 1, PL 01-184 Wrsw (Polnd) E-Mil chmielewsk @ gmil.com
Tble 1. Durtion of brestfeeding nd risk of CD [dpted from 9, 10] First uthor [Ref.], study Durtion of BF Conclusion Interventionl studies Vriezing [7], PREVENTCD Lionetti [8], CELIPREV Observtionl studies Included in the systemtic review by Akobeng [26] Auricchio [22] BF <30 dys or bottle-fed vs. BF >30 dys Risk of CD higher if BF shorter Ascher [37] Flth-Mgnusson [35] Medin BF durtion 2.5 months (CD) vs. 4 months (controls) Risk of CD higher if BF shorter Greco [23] BF <90 dys or bottle-fed vs. BF >90 dys Risk of CD higher if BF shorter Ivrsson [36] Medin BF durtion 5 months (CD) vs. 7 months (controls) Risk of CD higher if BF shorter Peters [24] CD decresed by 63% for BF durtion >2 vs. <2 months Risk of CD higher if BF shorter Decker [25] Hummel [33] /Ziegler [59], BABYDIAB No significnt ssocition between CD utoimmunity nd BF durtion Jnsen [31], Genertion R Study No significnt ssocition between CD utoimmunity nd BF durtion Norris [27], DAISY Study No significnt ssocition between CD utoimmuniztion nd BF durtion Roberts [28] Størdl [32], MoB b Welnder [29] Ziegler [30] No significnt ssocition between CD utoimmunity nd BF durtion BF = Brestfeeding. Sme popultion. b The Norwegin Mother nd Child Cohort Study. HLA-DQ2 or HLA-DQ8 hplotypes, nd enteropthy [1]. The clinicl spectrum of the disese my rnge from gstrointestinl nd non-gstrointestinl symptoms of different intensity to n symptomtic presenttion [1]. CD ffects pproximtely 1% of the generl popultion in Europe, which mkes it reltively common disese of significnt socil impct [2]. A lifelong, strict, gluten-free diet remins the only vilble tretment for ffected individuls. The qulity of life of people suffering from CD my be decresed, nd the disese imposes significnt cost burden on society [3 5]. Due to the bove resons, primry prevention hs been dvocted nd specil emphsis ws put on erly nutrition, nmely brestfeeding, s well s the time nd mode of gluten introduction s preventive mesures [6]. Until recently, the min body of evidence for the influence of erly nutrition prctices on the risk of CD ws derived from observtionl studies, the results of which cnnot be interpreted s cuslity, but rther s ssocitions. Thus, the pursuit of rndomized trils ws strongly recommended. In October 2014, two lrge rndomized controlled trils (RCTs) performed by two different reserch tems tht were designed to exmine the effect of different timings of gluten introduction were published, delivering long-wited dt on the effect of different timings of gluten introduction [7, 8]. One of the 2 studies, the Prevent Coelic Disese (PREVENTCD; http://www.preventcd.com) project, investigted the hypothesis of possible induction of tolernce to gluten in geneticlly predisposed children through the introduction of smll quntities of gluten during the period of brestfeeding, either t 4 or 6 months of ge [7]. The other study, the Risk of CD nd Age t Gluten Introduction (CELIPREV) tril, compred the risk of CD immuniztion nd overt CD in children t incresed risk in whom dietry gluten hd been introduced t 6 months compred to 12 months of ge [8]. These 2 studies, previous systemtic review [9], nd recent systemtic review [10] tht incorported the results of the 2 RCTs served s the bsis for this current review. Brestfeeding nd CD Brestfeeding is the optiml wy of feeding n infnt for mny helth-relted resons, nd exclusive brestfeeding for 4 6 months of ge is recommended worldwide [11, 12]. There re severl mechnisms by which brestfeeding hs been linked to the prevention of CD. Given tht brest milk is bundnt in fctors involving pssive immunity, such s lysozyme, lctoferrin or IgA ntibodies, brestfeeding my prevent gstrointestinl infections tht hve been linked to CD development [13 15]. Gut permebility, which hs been shown to be de- 44 Chmielewsk/Pieścik-Lech/Szjewsk/ Shmir
cresed in brestfed infnts, is nother importnt plyer in CD pthogenesis [16, 17]. Additionlly, brestfeeding theoreticlly might help with the development of tolernce to gluten by exposing n infnt to some gluten frctions digested with humn milk [18, 19]. Moreover, the role of the gut microbiot is lso being rised in the context of CD prevention due to significnt differences in microbil ptterns observed in brestfed infnts compred to formul-fed infnts [20]. The bove-listed fctors my be some of the plusible explntions for the protective potentil of humn milk ginst CD. Exclusive or Any Brestfeeding nd Durtion of Brestfeeding Previous studies hve provided no cler evidence on whether exclusive brestfeeding compred with formul or mixed feeding reduces the risk of CD or only delys the onset of symptoms [21 24]. Contrdictory results hve been reported on the effect of ny brestfeeding compred to no brestfeeding [24, 25]. In the systemtic review by Akobeng et l. [26] (involving 6 observtionl studies), protection ginst CD with longer durtion of brestfeeding ws reported. However, observtionl studies tht followed did not confirm this finding [25, 27 30]. The most recent dt from lrge cohort studies still brought no cler evidence with regrds to the role of brestfeeding nd CD prevention. A prospective Genertion R cohort including 1,679 Dutch children positive for HLA-DQ2 nd/or DQ8 evluted the possible ssocition of the timing of gluten introduction nd brestfeeding durtion with CD utoimmunity t 6 yers of ge. It ws shown tht brestfeeding for longer thn 6 months ws not relted to lower risk of CD utoimmunity [31]. Interestingly, the Norwegin Mother nd Child Cohort Study (MoB) involving more thn 100,000 children showed tht brestfeeding for longer thn 12 months ws ssocited with higher risk of CD [32]. A Germn cohort study (BABYDIAB) observed more thn 1,500 children of prents suffering from type 1 dibetes nd mesured islet nd CD utoimmunity. No ssocition between the durtion of brestfeeding nd risk of CD utoimmunity ws found [33]. In the cross-sectionl ETICS study (Exploring the Iceberg of Celics in Sweden), 2 birth cohorts were compred [34]. The durtion of brestfeeding ws longer in the 1997 cohort (9 months) compred to the 1993 cohort (7 months), being similr to tht of the Swedish generl popultion. The pooled results for 5 observtionl studies [22, 23, 25, 28, 33] showed tht ny brestfeeding compred with no brestfeeding 900 800 700 600 500 400 300 200 100 0 No BF Any BF Non-CD ptients CD ptients Fig. 1. No brestfeeding versus ny brestfeeding nd the risk of CD. Results from the PREVENTCD study (no significnt difference between groups). BF = Brestfeeding. hd no effect on the risk of developing CD [odds rtio (OR) 0.69, 95% confidence intervl (CI) 0.30 1.59] [10]. The results of brestfeeding durtion on the risk of CD re summrized in tble 1. In 2014, the results of the 2 forementioned interventionl trils were published. The PREVENTCD study ws multicenter double-blind plcebo-controlled rndomized tril, crried out in 8 countries (Croti, Germny, Hungry, Isrel, Itly, The Netherlnds, Polnd, nd Spin), involving 944 children with HLA-DQ2 or HLA- DQ8 positivity who hd t lest 1 first-degree reltive with CD [7]. Children were rndomly ssigned to receive plcebo (n = 469) or 100 mg of immunologiclly ctive gluten (n = 475) dily from 16 to 24 weeks of ge. Both groups received incresing mounts of gluten from 6 months onwrd. The min outcome mesure ws the frequency of CD, which ws confirmed by intestinl biopsy t 3 yers of ge. The CELIPREV study ws multicenter rndomized interventionl tril conducted in Itly [8]. Dietry gluten ws introduced t 6 months of ge (n = 297) or t 12 months of ge (n = 256) in infnts t risk for CD, defined by the presence of first-degree reltive with CD nd positive test for HLA-DQ2 or HLA-DQ8 (tested lter during the study). The presence of specific CD utontibodies nd overt CD t 5 yers of ge were the min outcome mesures. The results of both the PREVENTCD ( fig. 1 ) nd CELIPREV studies showed tht exclusive s well s ny brestfeeding did not significntly influence the development of CD. Neither did the durtion of brestfeeding ( tble 1 ). Despite the bove, one needs to cutiously in- Prevention of Celic Disese 45
Tble 2. Brestfeeding during the introduction of gluten nd risk of CD [dpted from 6, 7] First uthor [Ref.], study Conclusion Interventionl studies Vriezing [7], PREVENTCD Lionetti [8], CELIPREV Observtionl studies Included in met-nlysis by Szjewsk [10] Flth-Mgnusson [35] Protective Peters [24] Protective Ivrsson [36] Protective Norris [27], DAISY study Aronsson [38], TEDDY study Størdl [32], MoB study Ascher [37] Pooled Hummel [41], BABYDIAB Ivrsson [34], ETICS Brestfeeding predisposing to CD? terpret these dt, s both trils were designed to evlute the risk of CD mong children who were rndomly ssigned to the introduction of gluten t vrious ges, not the effect of brestfeeding on the risk of developing CD. Brestfeeding t the Time of Gluten Introduction Previously, met-nlysis by Akobeng et l. [26] suggested tht brestfeeding t the time of gluten introduction might be protective. However, it ws uncler whether it provided long-term prevention ginst CD or only delyed the onset of the disese [24, 35 37]. Results of the recently published cohort studies did not confirm the protective effect of brestfeeding t the time of gluten introduction on developing CD [31, 38]. Also, the pooled results for 7 observtionl studies showed tht brestfeeding t the time of gluten introduction hs no effect on the risk of developing CD compred with formul feeding (OR 0.88, 95% CI 0.52 1.51) [10]. The results of both recently published rndomized studies, PREVENTCD nd CELIPREV, did not show protective effect of introducing gluten during brestfeeding [7, 8]. Agin, cution is needed when interpreting the results, s neither of these 2 RCTs ws designed to ddress directly the effect of brestfeeding on CD. A summry of the studies showing the effect of brestfeeding t the time of gluten introduction is presented in tble 2. Timing of Gluten Introduction There hs been considerble del of inconsistency in the literture with regrds to the impct of the timing of gluten introduction on the risk of CD. In 2008, bsed on the results of observtionl studies, the best vilble evidence t tht time, the Europen scientific bodies [the Europen Society for Peditric Gstroenterology, Heptology nd Nutrition (ESPGHAN) nd the Europen Food Sfety Authority (EFSA)] recommended tht it is prudent to void both erly (<4 months) nd lte (>7 months) introduction of gluten nd to introduce gluten grdully while the infnt is still being brestfed [39, 40]. This pproch ws considered to decrese the risk of CD (lso the risk of type 1 dibetes nd whet llergy). Since then, substntil body of evidence from both cohort studies nd, most importntly, lrge RCTs hs become vilble. Dt from both interventionl nd observtionl studies re summrized in tble 3. In the previously published prospective cohort study by Norris et l. [27], both erly (<3 months of ge) nd lte (>7 months of ge) introduction of gluten to children t incresed risk of CD nd type 1 dibetes ws ssocited with n incresed risk of CD utoimmunity. Other studies did not report such effects [24, 29, 30, 35, 36]. Two new lrge cohort studies [31, 32] reported dt compring the introduction of gluten before nd fter 6 months of ge. The Genertion R Study found tht lter compred to erlier exposure ws not significntly ssocited with CD utoimmunity [31]. The Norwegin Mother nd Child Cohort Study (MoB) showed tht lter but not erlier exposure to gluten incresed the risk of CD, lthough this difference ws of borderline significnce (djusted OR 1.27, 95% CI 1.01 1.65) [32]. In The Environmentl Determinnts of Dibetes in the Young (TEDDY) study, first exposure to gluten before 17 weeks, between 17 nd 26 weeks, or fter 26 weeks of ge ws compred, nd no difference ws found in the risk of CD utoimmunity or CD between the groups [38]. Another cohort study (BABYDIAB) reported no difference in CD utoimmunity in infnts with gluten introduction before or fter 3 months of ge [33]. The ETICS study of cross-sectionl design compred 2 birth cohorts of 12-yer-olds nd found significnt difference in the totl prevlence of CD in children born during the CD epidemic (in 1993, when gluten ws introduced from 6 months of ge) nd those born fter the epidemic (in 1997, with gluten introduction involving smll mounts given from 4 to 6 months of ge) [34]. The results of interventionl trils designed to ssess the role of the timing of gluten introduction were pub- 46 Chmielewsk/Pieścik-Lech/Szjewsk/ Shmir
Tble 3. Time of gluten introduction nd risk of CD or CD utoimmunity [dpted from 6, 7] First uthor [Ref.], study Interventionl studies Vriezing [7], PREVENTCD Lionetti [8], CELIPREV Hummel [41] /Beyerlein [42] Sellitto [43] Observtionl studies Flth-Mgnusson [35] Ivrsson [36] Norris [27], DAISY Study Peters [24] Welnder [29] Ziegler [30] Jnsen [31], Genertion R Study Størdl [32], MoB b Hummel 2007 [33], BABYDIAB Aronsson [38], TEDDY Ivrsson [34], ETICS Results No significnt difference in CD risk for different time of gluten introduction (4 vs. 6 months) t 3 yers 6 vs. 12 months: RR 2.36 (1.27 4.36) t 2 yers predisposing; no significnt difference t 5 yers (primry outcome) No significnt difference in CD nd CD utoimmunity for different time of gluten introduction (6 vs. 12 months) No significnt difference in CD utoimmunity risk for different time of gluten introduction (6 vs. 12 months) No significnt ssocition between CD nd time of gluten introduction No significnt ssocition between CD nd time of gluten introduction (different time intervls from 1 to 12 months) 1 3 vs. 4 6 months: HR 2.94 (95% CI 0.83 10.4) predisposing to CD; 7 vs. 4 6 months: HR 1.78 (95% CI 0.92 3.42) predisposing No significnt ssocition between CD nd time of gluten introduction (different time intervls from 3 to >5 months) No significnt ssocition between CD nd time of gluten introduction (different time intervls from 0 to 12 months) No significnt ssocition between CD nd time of gluten introduction (different time intervls from 3 to > 6 months) No significnt ssocition between CD utoimmunity nd time of gluten introduction (<6 vs. >6 months) <6 vs. >6 months: djusted OR 1.27 (95% CI 1.01 1.65) borderline significnce for CD risk No significnt ssocition between CD utoimmunity nd time of gluten introduction (<3 or >3 months) No significnt ssocition between CD or CD utoimmunity nd time of gluten introduction (<17 vs. 17 26 vs. >26 weeks) More CD in children born during the CD epidemic (1993; gluten introduction from 6 months of ge) thn fter the epidemic (1997; gluten introduction in smll mounts from ge 4 6 months) HR = Hzrd rtio; RR = reltive risk. Sme popultion. b The Norwegin Mother nd Child Cohort Study. lished recently. In the PREVENTCD study, the risk of CD t 3 yers of ge ws similr in the group receiving 100 mg of immunologiclly ctive gluten dily from 16 to 24 weeks of ge compred to the plcebo group. The cumultive incidence of CD ws 5.2% (95% CI 3.6 6.8) [7]. Introduction of gluten t 6 months ws compred to 12 months in 3 other studies. In the CELIPREV tril, the erlier introduction of gluten incresed the risks of CD utoimmunity (16 vs. 7%, p = 0.002) nd overt CD (12 vs. 5%, p = 0.01) t 2 yers, but it hd no effect on either risk t 5 yers of ge, which ws the primry outcome of the study [8]. Thus, the lter introduction of gluten delyed the onset of the disese without influencing the overll risk. The results for PREVENTCD nd CELIPREV re presented in figure 2. Other interventionl studies lso reported no difference between groups in the risk of CD nd/or CD utoimmunity [41 43] ( tble 3 ). Amount of Gluten t Introduction A study from Sweden nlyzed the mount of gluten ingested by children. In children younger thn 2 yers of ge, the risk of developing CD ws greter when gluten ws introduced into the diet in lrge mounts compred to smll or medium mounts without quntifiction in grms per dy (djusted OR 1.5, 95% CI 1.1 2.1). The effect ws not observed in older children [36]. The bove dt come from the Swedish epidemic of CD. During the 1980s, Sweden experienced fourfold increse in the Prevention of Celic Disese 47
Fig. 2. Timing of gluten introduction nd the risk of CD: dt from the PREVENTCD (no significnt difference between groups) nd CELIPREV trils [ significnt difference between groups t 2 yers but not t 5 yers (primry outcome)]. NS = Not significnt. 450 400 350 300 250 200 150 100 50 0 PREVENTCD 4 6 months PREVENTCD >6 months CD t 3 yers (primry outcome, NS) CELIPREV 6 months CELIPREV 12 months CD t 2 yers (secondry outcome, p = 0.01) CELIPREV 6 months CELIPREV 12 months CD t 5 yers (primry outcome, NS) Non-CD ptients CD ptients number of CD cses. This phenomenon ws preceded by the chnge in infntile dietry hbits: incresing dily gluten intke nd postponing the timing of gluten introduction. The number of new CD cses decresed significntly fter the nutritionl recommendtions were modified nd the mount of gluten ingested ws reduced. Dt from the Swedish epidemic suggest tht ingestion of lrge mounts of gluten t introduction, nd possibly lrge mounts lter on, increse the risk for the erly onset of CD. However, whether other elements, such s specific products given in Sweden or other environmentl fctors, plyed role in the epidemic cnnot be ruled out. Lstly, the risk of CD development in HLA-DQ2.5 homozygous nd HLA-DQ2.2/2.5 heterozygous individuls ws reported to be significntly higher thn tht in HLA- DQ2.5/non-DQ2 heterozygous individuls. The dose of the HLA-DQ2 gene might be relted to gluten epitope diversity nd the risk of CD [44]. It is plusible tht the mount of gluten needed for CD development is different in subjects of different genetic risk. The PREVENTCD study reported tht the mount of gluten t wening ws not relted to the development of CD. This conclusion ws bsed on the men dily gluten intke fter the intervention in subset of prticipnts [7]. However, of note, it ws not preplnned study outcome. Other Environmentl Fctors Genetic predisposition nd exposure to gluten re the two undenible fctors necessry for CD to develop. However, most of the individuls positive for HLA-DQ2 nd/or HLA-DQ8 who re exposed to gluten do not develop the disese. Predisposing genes re crried by 30% of generl popultion [45], while CD occurs only in 1% [1]. Aprt from erly nutrition, other environmentl fctors must ply role. Repeted gstrointestinl infections hve been reported to increse the risk of CD [13, 14]. Incresed intestinl permebility might be custive, llowing proteins, such s glidin, to enter the lmin propri nd ctivte the utoimmune rection. Evidence, however, is not cler. Some studies hve found infections t the time of gluten introduction not to be ssocited with incresed risk of CD [46, 47]. The intestinl microbiot hs lso been investigted s potentil pthogenetic fctor in CD. Duodenl nd fecl microbiot hve been reported to be imblnced in children with CD, with incresed proportions of Bcteroides species nd reduction of Bifidobcterium species. A more fvorble pttern ws only prtilly restored fter introducing gluten-free diet [48, 49]. Administering Bifidobcterium together with gluten-free diet in children with CD ws investigted in rndomized plcebo-controlled tril [50]. A decrese in both the numbers of the Bcteroides frgilis group nd the content of secretory IgA in stools ws found, which might further confirm the role of microbiot in the pthogenesis of CD. However, it still remins uncler whether the differences in microbil ptterns between celic nd non-celic individuls re custive fctor or rther consequence of the disese. Interestingly, possible interction between genotype nd gut microbiot predisposing to CD development hs been rised in some studies nd is being further explored [20]. In infnts t high genetic risk of developing CD, 48 Chmielewsk/Pieścik-Lech/Szjewsk/ Shmir
the bcteril pttern is significntly different from tht in infnts t moderte or low risk [51 53]. Further studies on the potentil role of ltered microbiot in infnts geneticlly predisposed to CD might provide more insight [20]. Delivery by cesren section nd ssocited ltertions in the development of the enteric homeostsis during the neontl period hve been suggested to influence the incidence of CD [54]. However, more recent lrge cohort studies hve found no ssocition between the mode of delivery nd the incidence of CD [55, 56]. Vccintions, due to modultion of immunity, hve been proposed to be involved in the process of developing utoimmune diseses [57]. Scrce dt vilble with regrds to CD fil to support this hypothesis [58]. Conclusion The results of recently published interventionl studies did not confirm previous findings from observtionl studies nd did not support current recommendtions. Neither the specific timing of gluten introduction nor brestfeeding protect ginst CD in geneticlly predisposed individuls. Therefore, there is need to updte Neither the specific timing of gluten introduction nor brestfeeding protect ginst CD in geneticlly predisposed individuls current recommendtions to suggest tht since the ge of gluten introduction (nywhere between 4 nd 12 months) hs no effect on the prevention of CD, gluten should be introduced only s prt of recommendtions for introducing complementry foods in generl. It is worth emphsizing tht in children with no genetic predisposition for CD, the timing nd mode of gluten introduction does not influence the risk nyhow. Furthermore, brestfeeding for 6 months nd beyond should be promoted due to its mny positive effects on helth, but CD prevention should not be used s reson to do so; there is no need to introduce gluten while the infnt is still being brestfed. In regrd to the mode of delivery, even though it ws once suggested to be relted to CD, there seems to be no support for such n ssocition. Finlly, the gut microbiot my ply greter role in the pthogenesis of CD thn previously pprecited. The role of other environmentl fctors such s infections or vccintions remins uncler. Disclosure Sttement The uthors declre tht no finncil or other conflicts of interest exist in reltion to the content of this pper. The writing of this rticle ws supported by Nestlé Nutrition Institute. References holder Pltform on CD (CDEUSSA): Review rticle: future reserch on coelic disese position multistkeholder pltform on celic disese (CDEUSSA). Aliment Phrmcol Ther 2008; 27: 1030 1043. 7 Vriezing SL, Auricchio R, Brvi E, et l: Rndomized feeding intervention in infnts t high risk for celic disese. N Engl J Med 2014; 371: 1304 1315. 8 Lionetti E, Cstellnet S, Frncvill R, et l: Introduction of gluten, HLA sttus, nd the risk of celic disese in children. N Engl J Med 2014; 371: 1295 1303. 9 Szjewsk H, Chmielewsk A, Pieścik-Lech M, et l: Systemtic review: erly infnt feeding nd coelic disese prevention. Aliment Phrmcol Ther 2012; 36: 607 618. 10 Szjewsk H, Shmir R, Chmielewsk A, et l: Systemtic review with met-nlysis: erly infnt feeding nd coelic disese updte 2015. Aliment Phrmcol Ther 2015; 41: 1038 1054. 11 World Helth Orgniztion: The optiml durtion of exclusive brestfeeding: report of 1 Husby S, Koletzko S, Korpony-Szbo IR, et l: Guidelines for the dignosis of coelic disese. J Peditr Gstroenterol Nutr 2012; 54: 136 160. 2 Mustlhti, K, Ctssi C, Reunnen A, et l: The prevlence of celic disese in Europe: results of centrlized, interntionl mss screening project. Ann Med 2010; 42: 587 595. 3 Kurpp K, Collin P, Mäki M, et l: Celic disese nd helth-relted qulity of life. Expert Rev Gstroenterol Heptol 2011; 5: 83 90. 4 Shmir R, Hernell O, Leshno M: Cost-effectiveness nlysis of screening for celic disese in the dult popultion. Med Decis Mking 2006; 26: 1 12. 5 Hershcovici T, Leshno M, Goldin E, et l: Cost effectiveness of mss screening for celic disese is determined by time-dely to dignosis nd qulity of life on gluten free diet. Aliment Phrmcol Ther 2010; 31: 901 910. 6 Troncone R, Ivrsson A, Szjewsk H, Merin ML; Members of Europen Multistken expert consulttion. Genev, Word Helth Orgniztion, 2001. http://www.who. int/nutrition/publictions/optiml_durtion_of_exc_bfeeding_report_eng.pdf (ccessed September 4, 2012). 12 ESPGHAN Committee on Nutrition; Agostoni C, Bregger C, Decsi T, Kolcek S, Koletzko B, Michelsen KF, Mihtsch W, Moreno LA, Puntis J, Shmir R, Szjewsk H, Turck D, vn Goudoever J: Brest-feeding: commentry by the ESPGHAN Committee on Nutrition. J Peditr Gstroenterol Nutr 2009; 49: 112 125. 13 Stene LC, Honeymn MC, Hoffenberg EJ, et l: Rotvirus infection frequency nd risk of celic disese utoimmunity in erly childhood: longitudinl study. Am J Gstroenterol 2006; 101: 2333 2340. 14 Troncone R, Auricchio S: Rotvirus nd celic disese: clues to the pthogenesis nd perspectives on prevention. J Peditr Gstroenterol Nutr 2007; 44: 527 528. 15 Solid LM: Brest milk ginst celic disese. Gut 2002; 51: 767 768. Prevention of Celic Disese 49
16 Heymn M, Abed J, Lebreton C, et l: Intestinl permebility in coelic disese: insight into mechnisms nd relevnce to pthogenesis. Gut 2012; 61: 1355 1364. 17 Shulmn RJ, Schnler RJ, Lu C, et l: Erly feeding, ntentl glucocorticoids, nd humn milk decrese intestinl permebility in preterm infnts. Peditr Res 1998; 44: 519 523. 18 Chirdo FG, Rumbo M, Anon MC, et l: Presence of high levels of non-degrded glidin in brest milk from helthy mothers. Scnd J Gstroenterol 1998; 33: 1186 1192. 19 Verhsselt V: Neontl tolernce under brestfeeding influence. Curr Opin Immunol 2010; 22: 623 630. 20 Verdu EF, Glipeu HJ, Jbri B: Novel plyers in coelic disese pthogenesis: role of the gut microbiot. Nt Rev Gstroenterol Heptol 2015; 12: 497 506. 21 Nsh S: Does exclusive brest-feeding reduce the risk of coelic disese in children? Br J Community Nurs 2003; 8: 127 132. 22 Auricchio S, Follo D, de Ritis G, et l: Does brest feeding protect ginst the development of clinicl symptoms of celic disese in children? J Peditr Gstroenterol Nutr 1983; 2: 428 433. 23 Greco L, Auricchio S, Myer M, et l: Cse control study on nutritionl risk fctors in celic disese. J Peditr Gstroenterol Nutr 1988; 7: 395 399. 24 Peters U, Schneeweiss S, Trutwein EA, et l: A cse-control study of the effect of infnt feeding on celic disese. Ann Nutr Metb 2001; 45: 135 142. 25 Decker E, Engelmnn G, Findeisen A, et l: Cesren delivery is ssocited with celic disese but not inflmmtory bowel disese in children. Peditrics 2010; 125:e1433 e1440. 26 Akobeng AK, Rmnn AV, Buchn I, et l: Effect of brest feeding on risk of coelic disese: systemtic review nd met-nlysis of observtionl studies. Arch Dis Child 2006; 91: 39 43. 27 Norris JM, Brrig K, Hoffenberg EJ, et l: Risk of celic disese utoimmunity nd timing of gluten introduction in the diet of infnts t incresed risk of disese. JAMA 2005; 293: 2343 2351. 28 Roberts SE, Willims JG, Meddings D, et l: Perintl risk fctors nd coelic disese in children nd young dults: record linkge study. Aliment Phrmcol Ther 2009; 29: 222 231. 29 Welnder A, Tjernberg AR, Montgomery SM, et l: Infectious disese nd risk of lter celic disese in childhood. Peditrics 2010; 125:e530 e536. 30 Ziegler AG, Schmid S, Huber D, et l: Erly infnt feeding nd risk of developing type 1 dibetes-ssocited utontibodies. JAMA 2003; 290: 1721 1728. 31 Jnsen MA, Tromp II, Kiefte-de Jong JC, et l: Infnt feeding nd nti-tissue trnsglutminse ntibody concentrtions in the Genertion R Study. Am J Clin Nutr 2014; 100: 1095 1101. 32 Størdl K, White RA, Eggesbø M: Erly feeding nd risk of celic disese in prospective birth cohort. Peditrics 2013; 132: 1202 1209. 33 Hummel S, Hummel M, Bnholzer J, et l: Development of utoimmunity to trnsglutminse C in children of ptients with type 1 dibetes: reltionship to islet utontibodies nd infnt feeding. Dibetologi 2007; 50: 390 394. 34 Ivrsson A, Myléus A, Norström F, et l: Prevlence of childhood celic disese nd chnges in infnt feeding. Peditrics 2013; 131:e687 e694. 35 Flth-Mgnusson K, Frnzen L, Jnsson G, et l: Infnt feeding history shows distinct differences between Swedish celic nd reference children. Peditr Allergy Immunol 1996; 7: 1 5. 36 Ivrsson A, Hernell O, Stenlund H, et l: Brest-feeding protects ginst celic disese. Am J Clin Nutr 2002; 75: 914 921. 37 Ascher H, Krntz I, Rydberg L, et l: Influence of infnt feeding nd gluten intke on coelic disese. Arch Dis Child 1997; 76: 113 117. 38 Aronsson CA, Lee H-S, Liu E, et l: Age t gluten introduction nd risk of celic disese. Peditrics 2015; 135: 239 245. 39 ESPGHAN Committee on Nutrition; Agostoni C, Decsi T, Fewtrell M, et l: Complementry feeding: commentry by the ESPGHAN Committee on Nutrition. J Peditr Gstroenterol Nutr 2008; 46: 99 110. 40 EFSA Pnel on Dietetic Products, Nutrition nd Allergies (NDA): Scientific opinion on the pproprite ge for introduction of complementry feeding of infnts. http://www. efs.europ.eu/en/efsjournl/doc/1423.pdf (ccessed June 9, 2015). 41 Hummel S, Pfluger M, Hummel M, et l: Primry dietry intervention study to reduce the risk of islet utoimmunity in children t incresed risk for type 1 dibetes: the BABYDIET study. Dibetes Cre 2011; 34: 1301 1305. 42 Beyerlein A, Chmiel R, Hummel S, et l: Timing of gluten introduction nd islet utoimmunity in young children: updted results from the BABYDIET study. Dibetes Cre 2014; 37:e194 e195. 43 Sellitto M, Bi G, Seren G, et l: Proof of concept of microbiome-metbolome nlysis nd delyed gluten exposure on celic disese utoimmunity in geneticlly t-risk infnts. PLoS One 2012; 7:e33387. 44 Vder W, Stepnik D, Kooy Y, et l: The HLA-DQ2 gene dose effect in celic disese is directly relted to the mgnitude nd bredth of gluten-specific T cell responses. Proc Ntl Acd Sci USA 2003; 100: 12390 12395. 45 Kgnoff MF: Celic disese: pthogenesis of model immunogenetic disese. J Clin Invest 2007; 117: 41 49. 46 Welnder A, Tjernberg AR, Montgomery SM, et l: Infectious disese nd risk of lter celic disese in childhood. Peditrics 2010; 125:e530 e536. 47 Lebwohl B, Green PHR, Murry JA: Seson of birth in ntionwide cohort of coelic disese ptients. Arch Dis Child 2013; 98: 48 51. 48 Colldo MC, Dont E, Ribes-Koninckx C, et l: Imblnces in fecl nd duodenl Bifidobcterium species composition in ctive nd nonctive coelic disese. BMC Microbiol 2008; 8: 232. 49 Colldo MC, Dont E, Ribes-Koninckx C, et l: Specific duodenl nd fecl bcteril groups ssocited with peditric coelic disese. J Clin Pthol 2009; 62: 264 269. 50 Olivres M, Cstillejo G, Vre V, et l: Double-blind, rndomised, plcebo-controlled intervention tril to evlute the effects of Bifidobcterium longum CECT 7347 in children with newly dignosed coelic disese. Br J Nutr 2014; 112: 30 40. 51 De Plm G, Cpill A, Ndl I, et l: Interply between humn leukocyte ntigen genes nd the microbil coloniztion process of the newborn intestine. Curr Issues Mol Biol 2010; 12: 1 10. 52 Sánchez E, De Plm G, Cpill A, et l: Influence of environmentl nd genetic fctors linked to celic disese risk on infnt gut coloniztion by Bcteroides species. Appl Environ Microbiol 2011; 77: 5316 5323. 53 De Plm G, Cpill A, Nov E, et l: Influence of milk-feeding type nd genetic risk of developing coelic disese on intestinl microbiot of infnts: the PROFICEL study. PLoS One 2012; 7:e30791. 54 Decker E, Engelmnn G, Findeisen A, et l: Cesren delivery is ssocited with celic disese but not inflmmtory bowel disese in children. Peditrics 2010; 125:e1433 e1440. 55 Emilsson L, Mgnus MC, Størdl K: Perintl risk fctors for development of celic disese in children, bsed on the prospective Norwegin Mother nd Child Cohort Study. Clin Gstroenterol Heptol 2015; 13: 921 927. 56 Sevelsted A, Stokholm J, Bønnelykke K, et l: Cesren section nd chronic immune disorders. Peditrics 2015; 135:e92 e98. 57 Writh DC, Goldmn M, Lmbert PH: Vccintion nd utoimmune disese: wht is the evidence? Lncet 2003; 362: 1659 1666. 58 Myléus A, Stenlund H, Hernell O, et l: Erly vccintions re not risk fctors for celic disese. Peditrics 2012; 130:e63 e70. 59 Ziegler AG, Schmid S, Huber D, et l: Erly infnt feeding nd risk of developing type 1 dibetes-ssocited utontibodies. JAMA 2003; 290: 1721 1728. 50 Chmielewsk/Pieścik-Lech/Szjewsk/ Shmir