Prevalence of Nutrition and Health-Related Claims on Pre-Packaged Foods: A Five-Country Study in Europe

nu ien s A icle P e alence o Nu i ion and Heal h-Rela ed Claims on P e-Packaged Foods: A Fi e-Coun y S udy in Eu ope Sophie Hieke 1,*, Ne a Kuljanic 1, Igo P a s 2,3, K is a Mikla ec 2, Asha Kau 4, Ke y A. B own 5, Be nade e M. Egan 5, Ka ja P ei e 6, Azucena G acia 7and Mike Rayne 4 1Eu opean Food In o ma ion Council, ue Paul-Emile Janson 6, 1000 B ussels, Belgium; [email p o ec ed] 2Nu i ion Ins i u e, T žaška ces a 40, 1000 Ljubljana, Slo enia; igo [email p o ec ed] (I.P.); [email p o ec ed] (K.M.) 3Bio echnical Facul y, Uni e si y o Ljubljana, Jamnika je a 101, 1000 Ljubljana, Slo enia 4B i ish Hea Founda ion Cen e on Popula ion App oaches o Non-Communicable Disease P e en ion, Uni e si y o Ox o d, Ox o d OX3 7LF, UK; [email p o ec ed] (A.K.); [email p o ec ed] (M.R.) 5Food, Consume Beha iou and Heal h Resea ch Cen e, Uni e si y o Su ey, Guild o d GU2 7XH, UK; ke y.b own@su ey.ac.uk (K.A.B.); m.egan@su ey.ac.uk (B.M.E.) 6Ins i u e o Consume and Beha iou al Resea ch, Saa land Uni e si y, 66123 Saa b ücken, Ge many; [email p o ec ed] 7Ag i ood Resea ch and Technology Cen e o A agon, 50059 Za agoza, Spain; [email p o ec ed] *Co espondence: [email p o ec ed]; Tel.: +32-2-506-8981 Recei ed: 11 Decembe 2015; Accep ed: 25 Feb ua y 2016; Published: 3 Ma ch 2016 Abs ac : This s udy is pa o he esea ch unde aken in he EU unded p ojec CLYMBOL (“Role o heal h- ela ed CLaims and sYMBOLs in consume beha iou ”). The i s phase o his p ojec consis ed o mapping he p e alence o symbolic and non-symbolic nu i ion and heal h- ela ed claims (NHC) on oods and non-alcoholic be e ages in i e Eu opean coun ies. P e-packaged oods and d inks we e sampled based on a s anda dized sampling p o ocol, using s o e lis s o a s o e loo plan. Da a collec ion ook place ac oss i e coun ies, in h ee ypes o s o es. A o al o 2034 oods and d inks we e sampled and packaging in o ma ion was analyzed. A leas one claim was iden i ied o 26% (95% CI (24.0%–27.9%)) o all oods and d inks sampled. Six pe cen o hese claims we e symbolic. The majo i y o he claims we e nu i ion claims (64%), ollowed by heal h claims (29%) and heal h- ela ed ing edien claims (6%). The mos common heal h claims we e nu ien and o he unc ion claims (47% o all claims), ollowed by disease isk educ ion claims (5%). Eigh pe cen o he heal h claims we e child en’s de elopmen and heal h claims bu hese we e only obse ed on less han 1% (0.4%–1.1%) o he oods. The ca ego y o oods o speci ic die a y use had he highes p opo ion o NHC (70% o oods ca ied a claim). The p e alence o symbolic and non-symbolic NHC a ies ac oss Eu opean coun ies and be ween di e en ood ca ego ies. This s udy p o ides baseline da a o policy make s and he ood indus y o moni o and e alua e he use o claims on ood packaging. Keywo ds: nu i ion claims; heal h claims; heal h symbols; ood supply; moni o ing 1. In oduc ion 1.1. Backg ound The use o nu i ion and heal h- ela ed claims (NHC) on oods and non-alcoholic be e ages (hence o h oods and d inks o jus oods) is egula ed in many de eloped coun ies. In he Eu opean Union (EU), egula ions we e ha monized in 2006 by Regula ion (EC) 1924/2006 [ 1 ]. The Regula ion applies o all nu i ion and heal h claims made on ood packaging. The gene al p inciple is ha NHC Nu ien s 2016,8, 137; doi:10.3390/nu8030137 www.mdpi.com/jou nal/nu ien s Nu ien s 2016,8, 137 2 o 16 should no be misleading and should be subs an ia ed by gene ally accep ed scien i ic da a [ 1 , 2 ]. Since he EU Regis e o heal h claims made on ood en e ed in o o ce in Decembe 2012 [ 3 ], ood p oduce s in he EU a e allowed o use only au ho ized heal h claims and ela ed gene al non-speci ic heal h claims. One excep ion is claims ha a e s ill in he p ocess o scien i ic e alua ion, e.g., bo anicals [4]. NHC as well as hei symbolic ep esen a ions (examples o heal h symbols would be he Du ch Choices logo, he No dic Keyhole and he Finnish Hea Symbol) may help consume s iden i y oods ha a e heal hie op ions, bu li le is known abou how such claims a e used by consume s in eal-wo ld shopping si ua ions. The pan-Eu opean esea ch p ojec “Role o heal h- ela ed CLaims and sYMBOLs in consume beha iou ” (CLYMBOL) has se ou o de e mine how nu i ion and heal h- ela ed claims and symbols, in hei con ex , can a ec consume unde s anding, pu chase and consump ion pa e ns ( o an o e iew o he p ojec see [5]). As a s a ing poin o he p ojec , he su ey epo ed he e aimed o unde s and wha nu i ion and heal h- ela ed claims and symbols consume s a e exposed o, i.e., on which ood ca ego ies such claims and symbols a e mos common, which ypes o claims and symbols a e used and o which nu ien s and heal h ela ionships hey e e . While some s udies on he p e alence o NHC ha e been conduc ed in coun ies whe e such claims a e ela i ely well egula ed, pa icula ly in he US [ 6 , 7 ], Canada [ 8 , 9 ], Aus alia [ 10 – 13 ] and New Zealand [ 14 ], s udies on he Eu opean ma ke a e sca ce and o da e no c oss-coun y p e alence s udy o Eu ope exis s. In 2009, as pa o he EU- unded p ojec FLABEL [ 15 ], a s udy in es iga ed he p e alence o nu i ion in o ma ion, including heal h claims, o i e ood ca ego ies ac oss all EU Membe S a es plus Tu key [ 16 ]. Signi ican coun y- o-coun y di e ences we e epo ed ega ding he p e alence o nu i ion and heal h claims. Howe e , he s udy did no analyze he di e en ypes o claims ound. Mo e in-dep h analyses a e a ailable o he UK [ 17 ], I eland [ 18 ] and Slo enia [ 19 ]. Howe e , all o hese su eys we e comple ed be o e he EU Regis e o heal h claims came in o e ec in 2012. Some o he s udies a e also a ailable in which he au ho s selec i ely ocused on speci ic ood ca ego ies, o example dai y oods [ 20 ] and b eak as ce eals [ 21 ]. Las ly, he p e alence o nu i ion and heal h claims was also ecen ly in es iga ed in Se bia, a candida e coun y o EU membe ship [22]. While moni o ing he ood supply is an impo an public heal h issue, he numbe o oods a ailable and he di e si y o e ail en i onmen s make his challenging. A global ini ia i e ha add esses he ha moniza ion o such da a collec ion is he INFORMAS ini ia i e (In e na ional Ne wo k o Food and Obesi y/Non-Communicable Diseases Resea ch, Moni o ing and Ac ion Suppo ). This ne wo k has p oposed a s ep-wise app oach o su eying ood labeling depending on esou ces, s anda dized me hods o he sampling o e ail ou le s and oods and p io i ies o he labeling and ela ed in o ma ion o be collec ed [ 22 ]. While he selec ion o all oods wi hin all ood ca ego ies migh be conside ed ideal [ 22 ], a huge olume o da a is gene a ed when s udies a emp o be en i ely comp ehensi e [ 19 ]. The e o e, a sampling app oach o he selec ion o oods is needed pa icula ly o s udies in es iga ing di e en coun ies. The p esen s udy desc ibes a no el app oach on how o sample a selec numbe o oods ac oss a ious coun ies, using he ood ca ego iza ion scheme [ 23 ] ha has been adop ed by he Global Food Moni o ing G oup o u u e moni o ing o he ood supply [ 24 ]. 1.2. Resea ch Ques ions The aim o his s udy was o in es iga e he p e alence o nu i ion and heal h claims as ound on oods and d inks ac oss i e di e en Eu opean coun ies. The ollowing esea ch ques ions we e iden i ied: ‚ Wha p opo ion o p e-packaged oods and d inks a ailable in-s o e in he i e coun ies ca y NHC? ‚Wha p opo ion o hese claims is symbolic? ‚Wha ypes o NHC can be ound on p e-packaged oods and d inks? Nu ien s 2016,8, 137 3 o 16 ‚To which nu ien s o o he ood componen s do NHC e e ? ‚To which heal h- ela ionships do heal h claims e e ? ‚Which ypes o oods and d inks ca y NHC? 2. Me hods and Ma e ials Labeling da a we e collec ed om p e-packaged oods and d inks sampled in i e EU coun ies: he UK, he Ne he lands, Ge many, Slo enia and Spain. The selec ion o coun ies aimed a a geog aphical sp ead ac oss Eu ope. Th ee ypes o s o es we e selec ed in each coun y (wi h a o al o 15 s o es o e all), wi h he aim o co e ing a ange o di e en e ail ou le s, in o de o map consume exposu e o ood and d ink p oduc s ac oss a a ie y o shopping places: a la ge supe ma ke /na ional e aile , a discoun e and a neighbo hood s o e. Di e ences in he pene a ion o a ious ood labeling in o ma ion in di e en e ail ou le s we e epo ed ecen ly [ 19 ], con i ming he a ionale o he app oach used in his s udy. S o es we e selec ed based on he accessibili y o hei ne wo k wi hin he espec i e coun y, as well as compa able s o e cha ac e is ics (e.g., s o e size). A ood o d ink (hence o h only e e ed o as “ ood”) was de ined as a single i em a ailable o sale in he selec ed s o e. This de ini ion mean ha he same ood in di e en sized packaging could be included in he su ey, on he basis ha he packaging o he same ood in di e en sized packages may ca y di e en heal h- ela ed claims. Wi hin-coun y, exac duplica es, howe e , we e emo ed om he da abase. Only p e-packaged oods we e conside ed. The EU Regula ion on he p o ision o ood in o ma ion o consume s (EC) 1169/2011 de ines a p e-packaged oods u as “any single i em o p esen a ion as such o he ul ima e consume and o mass ca e e s, consis ing o a oods u and he packaging in o which i was pu be o e being o e ed o sale, whe he such packaging encloses he oods u comple ely o only pa ially, bu in any case in such a way ha he con en s canno be al e ed wi hou opening o changing he packaging” [25]. 2.1. Da a Collec ion In each o he i e coun ies, app oxima ely 400 oods we e sampled i.e., pu chased using a andomized sampling app oach. A powe calcula ion was conduc ed o es ima e he sample size equi ed. An assumed 50% p e alence a e o NHC was selec ed in o de o ensu e an adequa e sample size, o he de ec ion o NHC p e alence wi h 5% p ecision. The sample size needed o an es ima ion o p e alence is a maximum when he measu ed p e alence is 50%. Hence, a e adjus men o a ini e popula ion and assuming a p e alence a e o 50% o NHC, 400 oods o each coun y would p oduce con idence in e als o ˘ 5%, su icien o dis inguishing a 10% di e ence in p e alence be ween coun ies. In each o he i e coun ies app oxima ely 250 oods we e sampled om he supe ma ke / na ional e aile , 75 oods we e sampled om he discoun e and 75 om he neighbo hood s o e. De ails o he whole sample a e p o ided in Table 1. Table 1. Sample o e iew (sampling me hod and numbe o oods). UK Ne he lands Ge many Slo enia Spain To al La ge supe ma ke / na ional e aile loo plan, n= 248 s o e lis , n= 252 s o e lis , n= 248 s o e lis , n= 260 s o e lis , n= 251 1259 (62%) Discoun e loo plan, n= 75 s o e lis , n= 81 s o e lis , n= 76 s o e lis , n= 78 s o e lis , n= 78 388 (19%) Neighbo hood s o e loo plan, n= 75 s o e lis , n= 83 loo plan, n= 75 s o e lis , n= 78 s o e lis , n= 76 387 (19%) To al no o oods 398 (20%) 416 (20%) 399 (20%) 416 (20%) 405 (20%) 2034 (100%) Foods we e sampled om e ail ou le s in wo ways: ei he using a s o e/s ock lis o ailing ha a loo plan. Nu ien s 2016,8, 137 4 o 16 Re aile ’s s o e/s ock lis s we e con e ed o Mic oso Excel sp eadshee s, i no al eady in ha o ma . The ollowing ypes o p oduc we e hen excluded: (a) non- ood i ems; (b) ood supplemen s; (c) alcoholic d inks; and (d) unpackaged oods. The emaining oods we e assigned an ID numbe and he app op ia ely sized sample andomly selec ed using he RAND unc ion in Excel. The sample o oods was hen pu chased. Fo ou o he s o es, a s o e/s ock lis could no be ob ained. In hese cases a loo plan was c ea ed which mapped he layou o he s o e (including he loca ion o p omo ional s ands o o he non-aisle displays). Each sec ion/aisle was hen assigned a numbe and he numbe o oods in each sec ion/aisle was es ima ed using a ally coun e . The o al numbe o oods in he s o e was hen es ima ed and each ood loca ion assigned an ID numbe . Again, an app op ia ely sized lis o ood loca ions was andomly gene a ed using he RAND unc ion in Excel. The esea che s hen e u ned o he s o e and pu chased he oods using he lis o loca ions. Exclusion o p oduc s also occu ed pos sampling. Pos -sampling exclusion occu ed o unpackaged p oduc s selec ed inad e en ly. Foods we e selec ed and pu chased in he same ime ame (July–Augus 2013) o all coun ies. Fo pe ishable oods, he packaging was emo ed. Pho og aphs we e aken o all sides o packages ca ying a leas one claim. Two pilo s udies we e ca ied ou : one in Ge many a a la ge supe ma ke and he o he in he UK a a neighbo hood s o e. Feedback om he pilo s esul ed only in mino changes o he inal p o ocol o da a collec ion. 2.2. Da a Ex ac ion Labeling da a we e aken om packaging and en e ed in o a da abase. All NHC we e iden i ied and cha ac e is ics o he claims we e eco ded, i.e., whe he he claim was wo ded, pic o ial o a combina ion (bo h wo ded and pic o ial) and whe he i quali ied as a symbolic claim. Wo ded claims we e eco ded e ba im and ansla ed in o English. Pic o ial claims (o claims which we e a combina ion o wo ds and pic u e(s)) we e b ie ly desc ibed. The ype o heal h symbol was no ed. O he labeling and ood composi ion da a we e also collec ed (e.g., whe he he package had on -o -pack nu i ion labeling and in o ma ion abou nu ien con en om he nu ien decla a ion). Addi ional cha ac e is ics o he NHC eco ded included hei posi ion on pack and he numbe o imes hey appea ed on one packe . Fo heal h claims, i was also no ed whe he i was a speci ic heal h claim, meaning whe he a speci ied nu ien , o he subs ance o heal h- ela ed ing edien was s a ed o ha e a speci ic heal h e ec , o a non-speci ic heal h claim. 2.3. Analysis Once iden i ied, claims we e classi ied in o nu i ion claims (including nu ien con en claims and nu ien compa a i e claim), heal h- ela ed ing edien claims and heal h claims (including educ ion o disease isk claims, nu ien and o he unc ion claims, gene al heal h claims and child en’s de elopmen and heal h claims). All claims we e ca ego ized as being ei he symbolic o non-symbolic. A symbolic claim was de ined as a heal h- ela ed claim (whe he a nu i ion, heal h- ela ed ing edien o heal h claim) ha was pic o ial o combined wo ds and pic u es and o which he c i e ia o use o he symbol ha e been published. An example o a symbolic gene al heal h claim is he Du ch “Choices” logo [ 26 ]. An example o a symbolic nu ien and o he unc ion claims is he “Too h iendly” logo [ 27 ]. See Figu e 1. Nu ien s 2016,8, 137 5 o 16 Nu ien s 2016,8,1375o 17  Allclaimswe eca ego izedasbeingei he symbolico non‐symbolic.Asymbolicclaimwas de inedasaheal h‐ ela edclaim(whe he anu i ion,heal h‐ ela eding edien o heal hclaim) ha  waspic o ialo combinedwo dsandpic u esand o which hec i e ia o useo  hesymbolha e beenpublished.Anexampleo asymbolicgene alheal hclaimis heDu ch“Choices”logo[26].An exampleo asymbolicnu ien ando he  unc ionclaimsis he“Too h iendly”logo[27]. Symbolicgene al heal hclaim Symbolicnu ien and o he  unc ionclaim Du chchoiceslogoToo h iendlylogo Figu e1.Exampleso symbolicclaims. Colo ‐codednu i ionlabels(e.g., a icligh o %GDAlabeling)didno quali yasasymbolic claimbecause heyp esen nu i ionalin o ma ionbu dono sugges aheal h ela ionship[28].All NHCde ini ionsa econsis en wi hEURegula ion: 1. Anu i ionclaimwasde inedas“anyclaimwhichs a es,sugges so implies ha a oodhas pa icula bene icialnu i ionalp ope iesdue o(a) heene gy(calo i ic alue)i (i)p o ides; (ii)p o idesa a educedo inc eased a e;o (iii)doesno p o ide;and/o (b) henu ien so  o he subs ancesi (i)con ains;(ii)con ainsin educedo inc easedp opo ions;o (iii)doesno  con ain”[1]. 2. Nu i ionclaimswe eclassi iedasei he anu ien con en claimwheni “desc ibes hele elo  anu ien con ainedina ood(o i sene gy alue)”(e.g.,“highin ib e”and“low a ”)o a nu ien compa a i eclaim“wheni compa es hecomposi iono  he oodinques ionwi h he composi iono o he  oods”(e.g.,“highe in ib e”and“ educedsuga ”)[1]. 3. Aheal h‐ ela eding edien claimwasde inedasaclaimcommunica ing hep esenceo an ing edien (s)whichisno anu ien o o he subs anceasde inedin heEURegula ion[1]bu  whichimpliesheal hbene i s.Inmos cases, heseclaims ela ed o hecon en o ing edien s ha a econside edasaheal hy(e.g.,“Con ainsoneo you  i eaday”)o a leas aheal hie  al e na i e(e.g.,“Swee enedonlywi hb ownsuga ”). 4. Aheal hclaimwasde inedas“anyclaim ha s a es,sugges so implies ha a ela ionshipexis s be weena oodca ego y,a oodo oneo i scons i uen sandheal h”[1]. 5. Agene alheal hclaim(co e edbyA icle10(3)o  heEURegula ion[1])wasde inedasaclaim e e ing obene i s o gene alheal ho well‐being.A ypicalexampleo suchaclaimis“Good o you heal h”o “Heal hie choice(wi hin hisp oduc g oup)”. 6. Anu ien ando he  unc ionclaim(asco e edbyA icle13o  heEURegula ion[1])was de inedasaheal hclaim ha desc ibeso  e e s ooneo  he ollowing:(a) he oleo anu ien  o o he subs anceing ow h,de elopmen and he unc ionso  hebody;(b)psychologicaland beha io al unc ions;o (c)slimmingo weigh ‐con olo a educ ionin hesenseo hunge o  aninc easein hesenseo sa ie yo a educ iono  hea ailableene gy om hedie .Typical exampleso suchaclaima e“Calciumbuildss ong ee h”and“Fib ehelpsmain ainaheal hy diges i esys em”. 7. A educ iono disease iskclaim(co e edbyA icle14.1(a)o  heEURegula ion[1])was de inedasaclaimcommunica ing ha  heconsump iono a oodca ego y,a oodo oneo i s cons i uen ssigni ican ly educesa isk ac o in hede elopmen o ahumandisease.Typical exampleso suchaclaima e“Plan s e ols educebloodcholes e ol.Highcholes e olisa isk Figu e 1. Examples o symbolic claims. Colo -coded nu i ion labels (e.g., a ic ligh o % GDA labeling) did no quali y as a symbolic claim because hey p esen nu i ional in o ma ion bu do no sugges a heal h ela ionship [ 28 ]. All NHC de ini ions a e consis en wi h EU Regula ion: 1. A nu i ion claim was de ined as “any claim which s a es, sugges s o implies ha a ood has pa icula bene icial nu i ional p ope ies due o (a) he ene gy (calo i ic alue) i (i) p o ides; (ii) p o ides a a educed o inc eased a e; o (iii) does no p o ide; and/o (b) he nu ien s o o he subs ances i (i) con ains; (ii) con ains in educed o inc eased p opo ions; o (iii) does no con ain” [1]. 2. Nu i ion claims we e classi ied as ei he a nu ien con en claim when i “desc ibes he le el o a nu ien con ained in a ood (o i s ene gy alue)” (e.g., “high in ib e” and “low a ”) o a nu ien compa a i e claim “when i compa es he composi ion o he ood in ques ion wi h he composi ion o o he oods” (e.g., “highe in ib e” and “ educed suga ”) [1]. 3. A heal h- ela ed ing edien claim was de ined as a claim communica ing he p esence o an ing edien (s) which is no a nu ien o o he subs ance as de ined in he EU Regula ion [ 1 ] bu which implies heal h bene i s. In mos cases, hese claims ela ed o he con en o ing edien s ha a e conside ed as a heal hy (e.g., “Con ains one o you i e a day”) o a leas a heal hie al e na i e (e.g., “Swee ened only wi h b own suga ”). 4. A heal h claim was de ined as “any claim ha s a es, sugges s o implies ha a ela ionship exis s be ween a ood ca ego y, a ood o one o i s cons i uen s and heal h” [1]. 5. A gene al heal h claim (co e ed by A icle 10(3) o he EU Regula ion [ 1 ]) was de ined as a claim e e ing o bene i s o gene al heal h o well-being. A ypical example o such a claim is “Good o you heal h” o “Heal hie choice (wi hin his p oduc g oup)”. 6. A nu ien and o he unc ion claim (as co e ed by A icle 13 o he EU Regula ion [ 1 ]) was de ined as a heal h claim ha desc ibes o e e s o one o he ollowing: (a) he ole o a nu ien o o he subs ance in g ow h, de elopmen and he unc ions o he body; (b) psychological and beha io al unc ions; o (c) slimming o weigh -con ol o a educ ion in he sense o hunge o an inc ease in he sense o sa ie y o a educ ion o he a ailable ene gy om he die . Typical examples o such a claim a e “Calcium builds s ong ee h” and “Fib e helps main ain a heal hy diges i e sys em”. 7. A educ ion o disease isk claim (co e ed by A icle 14.1(a) o he EU Regula ion [ 1 ]) was de ined as a claim communica ing ha he consump ion o a ood ca ego y, a ood o one o i s cons i uen s signi ican ly educes a isk ac o in he de elopmen o a human disease. Typical examples o such a claim a e “Plan s e ols educe blood choles e ol. High choles e ol is a isk ac o in he de elopmen o co ona y hea disease” and “Reduces he isk ac o o de elopmen o den al ca ies”. Nu ien s 2016,8, 137 6 o 16 8. A child en’s de elopmen and heal h claim (co e ed by A icle 14.1(b) o he EU Regula ion [ 1 ]) was de ined as a heal h claim whe e child en’s de elopmen and/o heal h was speci ically men ioned. Typical examples o such a claim a e “Calcium is needed o no mal g ow h and de elopmen o bones in child en” and “Fo you baby’s sa e and balanced die ”. Claims whe e child en we e no speci ically men ioned, e en i hese claims we e ound on oods in ended solely o use by child en, we e no conside ed child en’s de elopmen and heal h claims. Claims such as “na u al”, “o ganic” and “halal” we e no conside ed NHCs on he basis ha hey e e o he me hod o he p oduc ion a he han he con en . Simila ly, in o ma ion on he p esence o addi i es, p ese a i es, la o ings, e c. was no conside ed a heal h- ela ed claim. Alle gy ad ice (e.g., “con ains nu s”) was no conside ed o be an NHC no we e e e ences o he p esence o a ood o ood g oup in he p oduc whe e he e was no clea ela ionship o i s heal h bene i s, e.g., e e ences o milk con en . Claims ela ing o he endo semen o a heal h- ela ed o ganiza ion we e also no included, e.g., “[P oduc ] suppo ed by he [heal h associa ion]”. 3. Resul s Whe e con idence in e als a e p o ided in ables, hey will no be epo ed in he ex . Resul s a e discussed by coun y, claim ype and p oduc ca ego y. 3.1. Wha P opo ion o P e-Packaged Foods and D inks A ailable In-S o e in he Fi e Coun ies Ca y NHC? Twen y-six pe cen o oods ca ied a leas one nu i ion, heal h- ela ed ing edien o heal h claim (including symbolic e sions o such claims) (Table 2). The e we e almos wice as many oods ca ying nu i ion claims as he e we e oods ca ying heal h claims (21% and 11% espec i ely). Jus 4% o oods ca ied a heal h- ela ed ing edien claim. Gene al heal h claims appea ed on 7% o he oods. Only 5% o oods ca ied a nu ien and o he unc ion claim. Jus 0.6% o oods ca ied a educ ion o disease isk claim and 0.7% o oods ca ied child en’s de elopmen and heal h claims. P oduc s could ca y mo e han one claim. Thi y pe cen o oods sampled in he UK ca ied a nu i ion claim, ollowed by Spain (23%), Slo enia (19%), he Ne he lands (17%) and Ge many (16%). The e was less a ia ion ac oss coun ies in he numbe o oods ca ying heal h claims; he highes p e alence was obse ed in he Ne he lands (14%) and he lowes in Spain (7%). The oods sampled in he Ne he lands had mo e ins ances o gene al heal h claims (12%), compa ed o 4%–6% o he emaining coun ies. The highes p opo ion o oods ca ying a nu ien and o he unc ion claim was ound in Slo enia (9%), ollowed by he UK (7%), Ge many (5%), Spain and he Ne he lands (bo h 3%). O e all, he p e alence o educ ion o disease isk claims was e y low ac oss all coun ies; i was highes in he UK (1%) and lowes in he Ne he lands (0.2%). Simila ly, he p e alence o child en’s de elopmen and heal h claims anged om 0% (in he Ne he lands) o 1.3% (in Ge many). Foods ca ying claims end o ca y mul iple claims; he a e age numbe o NHC pe p oduc ca ying any claim ac oss he i e coun ies was 2.6 (Table 3). The highes numbe o claims seen on a p oduc was 17, obse ed on wo baby ood p oduc s in Spain and Ge many. The mean numbe o nu i ion claims pe p oduc ca ying any claim was sligh ly highe han he mean numbe o heal h claims (2.0 and 1.9, espec i ely). A K uskal-Wallis es was used o de e mine i he e we e any signi ican coun y di e ences in he numbe o nu i ion claims, heal h claims o any claims (NHC) o oods ha ca y such claims. The e was a signi ican coun y di e ence in he numbe o heal h claims, bu he coun y di e ences in he numbe o nu i ion claims o NHC claims o e all we e no s a is ically signi ican . The highes numbe o heal h claims obse ed on a single p oduc was 15, ound on a baby ood, and o nu i ion claims, 13 claims ound on a con ec ione y p oduc , bo h in Ge many. Nu ien s 2016,8, 137 7 o 16 Table 2. P e alence o nu i ion claims, heal h claims, and symbols in i e Eu opean coun ies. Coun y Claim Type No. o Claims . . . o Which A e Symbolic No. o Foods wi h a Claim % o Foods wi h Claim (95% CIs) All Coun ies N = 2034 Foods Nu i ion claim 865 1 423 20.8% (19.0%–22.5%) Nu ien con en claim 797 1399 19.6% (17.8%–21.3%) Nu ien compa a i e claims 68 049 2.4% (1.7%–3.1%) Heal h- ela ed ing edien claim 105 6 72 3.5% (2.7%–4.3%) Heal h claim 392 74 222 10.9% (9.6%–12.3%) Gene al heal h claim 153 64 137 6.7% (5.6%–7.8%) Nu ien and o he unc ion claim 185 9 106 5.2% (4.2%–6.2%) Reduc ion o disease isk claim 21 1 12 0.6% (0.2%–0.09%) Child en’s de elopmen & heal h claims 33 0 15 0.7% (0.4%–1.1%) Any ype o claim (NHC) 1362 81 528 26.0% (24.0%–27.9%) UK N = 398 oods Nu i ion claim 247 0 118 29.6 (25.1%–34.1%) Heal h- ela ed ing edien claim 65 3 40 10.1% (7.1%–13.0%) Heal h claim 85 2 44 11.1% (8.0%–14.1%) Gene al heal h claim 30 0 23 5.8% (3.5%–8.1%) Nu ien and o he unc ion claim 38 2 26 6.5% (4.1%–9.0%) Reduc ion o disease isk claim 10 0 4 1.0% (0.0%–0.2%) Child en’s de elopmen & heal h claims 7 0 4 1.0% (0.0%–2.0%) Any ype o claim (NHC) 397 5 140 35.2% (30.4%–40.0%) Ne he -lands N = 416 oods Nu i ion claim 154 0 70 16.8% (13.2%–20.4%) Heal h- ela ed ing edien claim 12 0 12 2.9% (1.3%–4.5%) Heal h claim 73 50 60 14.4% (8.9%–15.2%) Gene al heal h claim 52 49 50 12.0% (8.9%–15.2%) Nu ien and o he unc ion claim 19 1 12 2.9% (1.3%–4.5%) Reduc ion o disease isk claim 2 0 1 0.2% (0.0%–0.1%) Child en’s de elopmen & heal h claims 0 0 0 0.0% Any ype o claim (NHC) 239 50 103 24.8% (20.6%–29.0%) Ge many N = 399 oods Nu i ion claim 123 0 64 16.0% (12.4%–19.7%) Heal h- ela ed ing edien claim 19 1 13 3.3% (1.5%–5.0%) Heal h claim 82 0 37 9.3% (6.4%–12.1%) Gene al heal h claim 29 0 23 5.8% (3.5%–8.1%) Nu ien and o he unc ion claim 45 0 20 5.0% (2.9%–3.2%) Reduc ion o disease isk claim 1 0 1 0.3% (–0.2%–0.7%) Child en’s de elopmen & heal h claims 7 0 5 1.3% (0.1%–0.2%) Nu ien s 2016,8, 137 8 o 16 Table 2. Con . Coun y Claim Type No. o Claims . . . o Which A e Symbolic No. o Foods wi h a Claim % o Foods wi h Claim (95% CIs) Slo enia N = 416 oods Nu i ion claim 144 0 78 18.8% (15.0%–22.5%) Heal h- ela ed ing edien claim 3 1 2 0.5% (–0.2%–1.1%) Heal h claim 88 7 52 12.5% (0.9%–15.7%) Gene al heal h claim 24 1 23 5.5% (3.3%–7.7%) Nu ien and o he unc ion claim 58 6 36 8.7% (5.9%–11.4%) Reduc ion o disease isk claim 4 0 3 0.7% (–0.1%–1.5%) Child en’s de elopmen & heal h claims 2 0 1 0.2% (–0.2%–0.7%) Any ype o claim (NHC) 235 8 103 24.8% (20.6%–28.9%) Spain N = 405 Nu i ion claim 196 1 93 23.0% (18.8%–27.1%) Heal h- ela ed ing edien claim 6 1 5 1.2% (0.2%–2.3%) Heal h claim 64 15 29 7.2% (4.6%–9.7%) Gene al heal h claim 18 14 18 4.4% (2.4%–6.5%) Nu ien and o he unc ion claim 25 0 12 3.0% (1.3%–4.6%) Reduc ion o disease isk claim 4 1 3 0.7% (–0.1%–1.6%) Child en’s de elopmen & heal h claims 17 0 5 1.2% (0.2%–2.3%) Any ype o claim (NHC) 266 17 100 24.7% (20.5%–28.9%) Bold is o main claims we looked a and he o al claims. I alic is o he sub-ca ego ies o all hese claims. Regula on size was only used o he desc ip ion o he coun ies o he le . Nu ien s 2016,8, 137 9 o 16 Table 3. Numbe o claims pe p oduc . Mean Numbe o Nu i ion Claims 1 Highes Numbe o Nu i ion Claims on a Single P oduc Mean Numbe o Heal h Claims 1 Highes Numbe o Heal h Claims on a Single P oduc Mean Numbe o Any Claim 1 Highes Numbe o Any Claims on a Single P oduc All coun ies 2.0 13 1.9 15 2.6 17 UK 2.1 11 1.9 5 2.8 15 Ne he lands 2.2 8 1.2 6 2.3 9 Ge many 1.9 13 2.2 15 2.7 17 Slo enia 1.8 8 1.7 5 2.3 11 Spain 2.1 12 2.2 11 2.7 17 p alue 0.94 <0.01 0.52 NOTE: pe p oduc ca ying a claim. 3.2. Wha P opo ion o These Claims is Symbolic? Subs an ial di e ences we e obse ed in he p e alence o symbolic claims be ween he coun ies (Table 2). The highes p e alence was ound in he Ne he lands (12%, (9%–14%)), ollowed by Spain (4%, (2%–6%)), Slo enia (2%, (1.0%–3.0%)), he UK (1%, (0%–2%)), and Ge many (0.3%, (0%–1%)). Almos all o he gene al heal h claims in he Ne he lands we e symbolic— he majo i y o which comp ised o he Choices logo. Spain had a simila ly high le el o symbolic gene al heal h claims, including logos o heal h o ganiza ions such as he Spanish Associa ion o Pedia icians. Howe e , i should be no ed ha in gene al he p e alence o symbolic claims was qui e low; al oge he only 3.88% (3%–5%) o oods we e labeled wi h a symbolic claim. 3.3. Wha Types o NHC Can Be Found on P e-Packaged Food and D inks? 64% o he claims we e nu i ion claims o which 92% (90%–94%) we e classi ied as nu ien con en claims and jus 8% (6%–10%) we e nu ien compa a i e claims (Table 2). Heal h claims accoun ed o 29% (26%–31%) o he claims eco ded. O hese, almos hal (47%, (42%–52%)) we e nu ien and o he unc ion claims, 39% (34%–44%) we e gene al heal h claims, 8% (6%–11%) we e child en’s de elopmen and heal h claims and he emaining 5% (3%–8%) we e educ ion o isk claims. The emaining claims ell in o he ca ego y o heal h- ela ed ing edien claims (8%, (6%–9%)). The p opo ion o nu i ion claims was highes in Spain (74% o all claims we e nu i ion claims), ollowed by he Ne he lands (64%), he UK (62%), Slo enia (61%), and Ge many (55%). The p opo ion o heal h claims showed sligh ly less a ia ion ac oss he i e coun ies: i was highes in Slo enia and Ge many (bo h 37%), ollowed by he Ne he lands (31%), Spain (24%) and he UK (21%). Only 14% (12%–15%) o he NHC we e nu ien and o he unc ion heal h claims. The highes p opo ion was ound in Slo enia (25%, (19%–30%)) and lowes in he Ne he lands (8%, (5%–11%)). Spain had he highes p opo ion o child en’s de elopmen and heal h claims (6%, (3%–9%)) while none we e ound on oods sampled in he Ne he lands. The UK had he highes p opo ion o disease isk claims (3%, (1%–4%)) whe eas Ge many had he lowes (only one such claim was ound, ou o a o al o 224 claims iden i ied) (Table 2). 3.4. To Which Nu ien s o O he Food Componen s Do NHC Re e ? Mo e han one hi d o he nu i ion claims e e ed o i amins and/o mine als (e.g., “En iched wi h impo an i amins and mine als”) (Table 4). Mo e speci ically, 22% e e ed o i amins, o which i amin C was he mos common i amin e e enced (e.g., “High in i amin C”) and 13% o mine als, o which calcium was he mos common mine al (e.g., “A sou ce o calcium”). Almos a qua e o nu i ion claims (24%) e e ed o he a con en o a ood in some way, 12% o he suga con en and 9% o ibe . Nu ien s 2016,8, 137 16 o 16 18. Lalo , F.; Kennedy, J.; Flynn, M.A.T.; Wall, P.G. A s udy o nu i ion and heal h claims—A snapsho o wha ’s on he I ish ma ke . Public Heal h Nu . 2010,13, 704–711. [C ossRe ] [PubMed] 19. P a s , I.; Kuša , A. Consume s’ exposu e o nu i ion and heal h claims on p e-packed oods: Use o sales weigh ing o assessing he ood supply in Slo enia. Nu ien s 2015,7, 9353–9368. [C ossRe ] [PubMed] 20. Mena d, C.; Dumas, C.; Gillo , N.; Lau en , L.; Laba be, B.; I eland, J.; Vola ie , J.-L. The F ench OQALI su ey on dai y p oduc s: Compa ison o nu ien con en s and o he nu i ion in o ma ion on labels among ypes o b ands. J. Hum. Nu . Die . 2012,25, 323–333. [C ossRe ] [PubMed] 21. Debeljak, K.; P a s , I; Košmelj, K.; Kaˇc, M. “Heal hie ” and “less heal hy” classi ica ions acco ding o h ee nu ien p o iling sys ems ela i e o nu i ion and heal h claims on ood labels. Ac a Alimen . 2015 ,44, 561–565. 22. Da ido ic, D.B.; Tomic, D.V.; Pauno ic, K.Ž.; Vasilje ic, N.D.; Jo ga, J.B. Nu i ion labelling o p e-packaged oods in Belg ade, Se bia: Cu en si ua ion. Public Heal h Nu . 2015,18, 1969–1978. [C ossRe ] [PubMed] 23. Dun o d, E.; Webs e , J.; Me zle , A.B.; Cze nichow, S.; Mhu chu, C.N.; Snowdon, W.; L’Abbe, M.; Li, N.; Maulik, P.K.; Ba que a, S.; e al. In e na ional collabo a i e p ojec o compa e and moni o he nu i ional composi ion o p ocessed oods. Eu . J. P e . Ca diol. 2012,19, 1326–1332. [C ossRe ] [PubMed] 24. Dun o d, E.; Webs e , J.; Me zle , A.; Blanco, S.; Ni Mhu chu, C.; Wolma ans, P.; Snowdon, W.; L’Abbe, M. In e na ional collabo a i e p ojec o compa e and ack he nu i ional composi ion o as oods. BMC Public Heal h 2012,12, 1–7. 25. Regula ion (EU) No 1169/2011 on he P o ision o Food In o ma ion o Consume s. A ailable online: h p://eu -lex.eu opa.eu/legal-con en /EN/TXT/PDF/?u i=CELEX:02011R1169-20140219 (accessed on 29 Feb ua y 2016). 26. Roodenbu g, A.J.C.; Popkin, B.M.; Seidell, J.C. De elopmen o in e na ional c i e ia o a on o package ood labelling sys em: The In e na ional Choices P og amme. Eu . J. Clin. Nu . 2011 ,65, 1190–1200. [C ossRe ] [PubMed] 27. Cina , A.B.; Mu omaa, H. A holis ic ood labelling s a egy o p e en ing obesi y and den al ca ies. Obes. Re . 2009,10, 357–361. [C ossRe ] [PubMed] 28. Rayne , M.; Wood, A.; Law ence, M.; Mhu chu, C.N.; Albe , J.; Ba que a, S.; F iel, S.; Hawkes, C.; Kelly, B.; Kumanyika, S.; e al. Moni o ing he heal h- ela ed labelling o oods and non-alcoholic be e ages in e ail se ings. Obes. Re . 2013,14, 70–81. [C ossRe ] [PubMed] 29. Wo ld Heal h O ganisa ion. In e na ional Classi ica ion o Func ioning, Disabili y and Heal h (ICF). A ailable online: h p://www.who.in /classi ica ions/ic /en/ (accessed on 29 Feb ua y 2016). 30. Raa s, M.M.; Malcolm, R.N.; Läh eenmaki, L.; P a s , I.; Gage, H.; Clea y, A.; Klopˇciˇc, M. Unde s anding he impac o Eu opean Regula ion on he subs an ia ion and use o claims on ood and d inks: Design o he REDICLAIM p ojec and ini ial esul s. Nu . Bull. 2015,40, 340–348. [C ossRe ] © 2016 by he au ho s; licensee MDPI, Basel, Swi ze land. This a icle is an open access a icle dis ibu ed unde he e ms and condi ions o he C ea i e Commons by A ibu ion (CC-BY) license (h p://c ea i ecommons.o g/licenses/by/4.0/).