Studies of acrylamide level in coffee and coffee substitutes: influence of raw material and manufacturing conditions

* Corresponding author:+DQQD0RMVND'HSDUWPHQWRI)RRGDQG)RRG6XSSOHPHQWV1DWLRQDO)RRGDQG1XWULWLRQ,QVWLWXWH 3RZVLĔVND6WUHHW:DUVDZ3RODQGSKRQHID[ HPDLOKPRMVND#L]]ZDZSO ‹&RS\ULJKWE\WKH1DWLRQDO,QVWLWXWHRI3XEOLF+HDOWK1DWLRQDO,QVWLWXWHRI+\JLHQH

678',(62)$&5</$0,'(/(9(/,1&2))(($1'&2))((

68%67,787(6,1)/8(1&(2)5$:0$7(5,$/

$1'0$18)$&785,1*&21',7,216

Background. Many animal studies have shown that acrylamide is both neurotoxic and carcinogenic. The first reports of DFU\ODPLGHDFWXDOO\KDYLQJEHHQIRXQGLQIRRGVWXIIVZHUHSXEOLVKHGLQE\WKH6ZHGLVK1DWLRQDO)RRG$JHQF\LQFRQ-junction with scientists from the University of Stockholm. It has since been demonstrated that acrylamide arises in foodstuffs by the 0DLOODUGUHDFWLRQLHEHWZHHQIUHHDVSDUDJLQHDQGUHGXFLQJVXJDUVDWWHPSHUDWXUHV!ž&&RIIHHLQIDFWIRUPV RQHRIWKHSULQFLSDOGLHWDU\VRXUFHVRIDFU\ODPLGHZKHUHLWLVQRUPDOO\GUXQNLQODUJHTXDQWLWLHVWKURXJKRXWPDQ\FRXQWULHV ZRUOGZLGHWKDWLQFOXGHV3RODQG7KXVLWFRQVWLWXWHVDPDMRUGLHWDU\FRPSRQHQWLQDZLGHUDQJHRISRSXODWLRQJURXSVPDLQO\ ranging from late adolescents to the elderly.

Objectives. To determine the acrylamide level in commercial samples of roasted and instant coffee and in coffee substitutes by LC-MS/MS method. The influence of coffee species and colour intensity of coffee on acrylamide level was also detailed. Materials and methods. $WRWDORIVDPSOHVRIFRIIHHZHUHDQDO\VHGZKLFKLQFOXGHGWKDWZHUHJURXQGURDVWHGFRIIHH 11 instant coffees and 3 coffee substitutes (grain coffee). Analytical separation of acrylamide from coffee was performed by liquid chromatography followed by tandem mass spectrometry (LC-MS/MS). To evaluate the colour intensity of ground roasted coffee and instant coffee we used method of arranging (sequence).

Results. The highest mean acrylamide concentrations were found in coffee substitutes (818 μg/kg) followed by instant coffee —JNJDQGWKHQURDVWHGFRIIHH—JNJ2QHVLQJOHFXSRIFRIIHHPOGHOLYHUHGRQDYHUDJHIURP—JDFU\O-DPLGHLQURDVWHGFRIIHHWR—JLQFRIIHHVXEVWLWXWHV7KHUHZHUHQRVLJQLILFDQWGLIIHUHQFHVLQDFU\ODPLGHOHYHOEHWZHHQ the coffee species ie. $UDELFD vs 5REXVWD or a mixture thereof. The various methods of coffee manufacture also showed no differences in acrylamide (ie. freeze-dried coffee vs agglomerated coffee). A significant negative correlation was observed EHWZHHQDFU\ODPLGHOHYHOVDQGWKHLQWHQVLW\RIFRORXULQURDVWHGFRIIHHWKLVZDVQRWWKHFDVHKRZHYHUIRULQVWDQWFRIIHH Conclusions. It was demonstrated that roasting process had the most significant effect on acrylamide levels in natural FRIIHHKRZHYHUWKHUHZHUHQRUHODWLRQVKLSVIRXQGZLWKFRIIHHVSHFLHV'XHWRWKHKLJKDFU\ODPLGHOHYHOVGHPRQVWUDWHGLQ FRIIHHVXEVWLWXWHVUHFRPPHQGHGDPRXQWVVKRXOGEHGHILQHGDQGPDQXIDFWXUHUVVKRXOGEHREOLJHGWRUHGXFHVXFKOHYHOVLQ these products. Key wordsDFU\ODPLGHGLIIHUHQWW\SHVRIFRIIHH/&0606PHWKRGFRIIHHVSHFLHVFRORULQWHQVLW\ 675(6=&=(1,( :SURZDG]HQLH:OLF]Q\FKEDGDQLDFKQD]ZLHU]ĊWDFKZ\ND]DQRĪHDNU\ORDPLGPDG]LDáDQLHQHXURWRNV\F]QHLNDQFHUR-JHQQH:NZLHWQLXU6]ZHG]ND1DURGRZD$JHQFMDGVĩ\ZQRĞFLRUD]QDXNRZF\]8QLZHUV\WHWXZ6]WRNKROPLHSR UD]SLHUZV]\RSXEOLNRZDOLGDQHR]DZDUWRĞFLDNU\ORDPLGXZSURGXNWDFKVSRĪ\ZF]\FK$NU\ORDPLGSRZVWDMHZĪ\ZQRĞFL w reakcji 0DLOODUGDSRPLĊG]\ZROQąDVSDUDJLQąLFXNUDPLUHGXNXMąF\PLSRGZSá\ZHPWHPSHUDWXU\!ƒ&:\VRNLH VSRĪ\FLHWHJRQDSRMXZZLHOXNUDMDFKW\PWDNĪHZ3ROVFHVSUDZLDĪHSURGXNWWHQPRĪHE\ü]QDF]ąF\PĨUyGáHPDNU\OR-DPLGXZGLHFLH]ZáDV]F]DXRVyEGRURVá\FKDOHWDNĪHVWDUV]HMPáRG]LHĪ\ &HOEDGDĔ2]QDF]HQLH]DZDUWRĞFLDNU\ORDPLGXZUyĪQ\FKURG]DMDFKNDZ\PHWRGą/&0606RUD]RFHQDZSá\ZXJDWXQNX NDZ\VSRVREXSURGXNFMLRUD]EDUZ\SURGXNWXILQDOQHJRQDSR]LRPEDGDQHJR]ZLą]NX 0DWHULDáLPHWRG\0DWHULDáGREDGDĔVWDQRZLá\SUyENLNDZ\ZW\PSUyEHNPLHORQHMNDZ\SDORQHMSUyEHNNDZ\ UR]SXV]F]DOQHMLSUyENLVXEVW\WXWyZNDZ\NDZD]ERĪRZD$NU\ORDPLGR]QDF]RQRPHWRGąFKURPDWRJUDILLFLHF]RZHM VSU]ĊĪRQHM]WDQGHPRZąVSHNWURPHWULąPDV/&0606'RRFHQ\EDUZ\SUyEHNNDZ\Z\NRU]\VWDQRPHWRGĊNROHMQRĞFL (szeregowania).

,1752'8&7,21

)RU RYHU WKH ODVW  \HDUV DFU\ODPLGH KDV EHHQ manufactured on an industrial scale as a substrate for V\QWKHVLVLQJSRO\DFU\ODPLGHSRO\PHUVWKLVEHLQJXVHG LQLQGXVWULDODQGGULQNLQJZDWHUILOWUDWLRQDVILOOLQJV DVZHOODVLQWKHSDSHUWH[WLOHDQGFRVPHWLFVLQGXVWULHV Numerous studies have demonstrated that acrylamide is neurotoxic and may damage the central and peripheral nervous systems (respectively CNS and PNS) in both animal models and humans exposed to acrylamide at WKHZRUNSODFH>@

Cell culture studies show that acrylamide is genoto-xic and amongst other things it results in chromosomal DEHUUDWLRQVVLVWHUFKURPDWLGH[FKDQJHJHQHSRO\SORLG\ DQGRWKHUPLWRWLFGLVUXSWLRQLQPDPPDOLDQFHOOV>@ Animal studies demonstrate an increase in multi-organ WXPRXUVWKDWLQFOXGHWKHIRUHVWRPDFKWHVWHVFOLFWRULDO JODQGPDPPDU\JODQGRUDOFDYLW\VNLQDQGWK\URLG gland when acrylamide was supplied in the drinking ZDWHU>@

$OUHDG\ LQ  WKH ,QWHUQDWLRQDO$JHQF\ IRU Research on Cancer had assigned acrylamide to those substances that are ‘probably carcinogenic to humans’ JURXS$>@7KH-RLQW)$2:+2([SHUW&RPPLW-WHHRQ)RRG$GGLWLYHVLQVWDWHGWKDWWKHHVWLPDWHG exposure limits to dietary acrylamide are low for such a substance that exhibits genotoxicity and cancerogenicity DQGWKXVFRQVWLWXWHVDKXPDQKHDOWKKD]DUG>@

,Q IRRGVWXIIV DFU\ODPLGH IRUPV LQ WKH 0DLOODUG reaction between free asparagine and reducing sugars. The main dietary sources of acrylamide are potato SURGXFWV)UHQFKIULHVDQGSRWDWRFULVSVFHUHDOVHJ EUHDGEUHDNIDVWFHUHDOIODNHVDQGFDNHVWRJHWKHUZLWK FRIIHH>@

When foodstuff acrylamide levels were monitored LQ(XURSHGXULQJLWZDVIRXQGWKDWDQDYHUD- JHRI—JNJZHUHSUHVHQWLQURDVWHGFRIIHHZLWKPD-[LPXPRI—JNJRISURGXFW0XFKKLJKHUDPRXQWV ZHUHPHDVXUHGLQLQVWDQWFRIIHHPHDQRI—JNJ

DQG LQ FRIIHH VXEVWLWXWHV PHDQ RI  —JNJ In the latter case the acrylamide content reached a PD[LPXPRIDERYH—JNJLQLQGLYLGXDOVDPSOHV >@8SWLOOQRZLQ3RODQGWKHUHLVKRZHYHUDODFNRI any data on acrylamide levels in instant coffee and in coffee substitutes.

Many studies demonstrate that coffee to be a signi-ILFDQWGLHWDU\VRXUFHRIDFU\ODPLGHHVSHFLDOO\LQWKRVH countries with a high coffee consumption. The presence RIFRIIHHLQDQDGXOWGLHWPD\GHOLYHUXSRIWKHHQ-WLUHGDLO\LQWDNHRIDFU\ODPLGH>@ :KHQ FRIIHH LV URDVWHG WKH ODUJHVW DPRXQWV RI acrylamide form at the initial stage and swiftly attain PD[LPXPOHYHOV1HYHUWKHOHVVWKHORQJHUWKHURDVWLQJ FRQWLQXHVWKHQDFU\ODPLGHOHYHOVGHFUHDVH> @,WDOVRDSSHDUVWKDWWKHDPRXQWVRIDFU\ODPLGHPD\ depend on the coffee species. When the acrylamide content of $UDELFD and 5REXVWDFRIIHHVDUHFRPSDUHG WKHQWKHODWWHUVKRZUDLVHGOHYHOVLQPRVWFDVHVKRZHYHU WKHVHDUHQRWVWDWLVWLFDOO\VLJQLILFDQW>@$QRWKHU SRLQWWKDWQHHGVVWUHVVLQJLVWKDWDWSUHVHQWWKHUHDUHDV yet no effective tools available for decreasing acryla-PLGHOHYHOVLQFRIIHHXQOLNHWKRVHLQWKHRWKHUDIIHFWHG IRRGVWXIIVVXFKDVFHUHDOVDQGSRWDWRSURGXFWV>@

The aim of this study was to determine the acryla-mide level in commercial samples of roasted and instant coffee and in coffee substitutes by LC-MS/MS method. The influence of coffee species and colour intensity of coffee powder on acrylamide level was also detailed.

0$7(5,$/$1'0(7+2'6

6DPSOLQJ

$QDO\VHVZHUHSHUIRUPHGRQFRIIHHVDPSOHVFRQ-VLVWLQJRIVDPSOHVRIJURXQGURDVWHGFRIIHHLQVWDQW coffee and 3 coffee substitutes made by various manu-facturers. The roasted coffee samples were randomly selected as part of a monitoring programme in accordan-ce with European Commission (EC) recommendations

:\QLNL :ĞUyG SU]HEDGDQ\FK SUyEHN UyĪQ\FK URG]DMyZ NDZ\ QDMZ\ĪV]ą ĞUHGQLą ]DZDUWRĞü DNU\ORDPLGX VWZLHUG]RQR ZNDZLH]ERĪRZHMȝJNJDQDVWĊSQLHZNDZLHUR]SXV]F]DOQHMȝJNJLNDZLHSDORQHMȝJNJ-HGQDILOLĪDQND POQDSDUXNDZ\GRVWDUF]DáDZ]DOHĪQRĞFLRGURG]DMXNDZ\SU]HFLĊWQLHRGȝJNDZDSDORQDGRȝJNDZD ]ERĪRZDDNU\ORDPLGX1LHVWZLHUG]RQRLVWRWQ\FKUyĪQLFZ]DZDUWRĞFLDNU\ORDPLGXZ]DOHĪQRĞFLRGJDWXQNX]LDUQD$UDELFD vs 5REXVWD vs mieszanka $UDELFDi 5REXVWD). Sposób produkcji kawy rozpuszczalnej (liofilizowana vs. aglomeryzowana) QLHZSá\ZDáQD]DZDUWRĞüDNU\ORDPLGXZSURGXNFLHILQDOQ\P6WZLHUG]RQRLVWRWQąSXMHPQąNRUHODFMĊSRPLĊG]\ ]DZDUWRĞFLąDNU\ORDPLGXDLQWHQV\ZQRĞFLąEDUZ\NDZ\SDORQHM=ZLą]NXWDNLHJRQLHVWZLHUG]RQRGODNDZ\UR]SXV]F]DOQHM :QLRVNL8]\VNDQHZ\QLNLSRWZLHUG]DMąĪHSURFHVZ\SDODQLDNDZ\PDQDMLVWRWQLHMV]\ZSá\ZQD]DZDUWRĞüDNU\ORDPLGX ZNDZLHQDWXUDOQHM1LHSRWZLHUG]RQRQDWRPLDVW]ZLą]NXSRPLĊG]\SR]LRPHPDNU\ORDPLGXDJDWXQNLHPNDZ\=XZDJLQD Z\VRNą]DZDUWRĞüDNU\ORDPLGXZVXEVW\WXWDFKNDZ\NDZD]ERĪRZD]DVDGQHZ\GDMHVLĊRSUDFRZDQLHZDUWRĞFLZVNDĨQL-NRZ\FKGODWHMJUXS\SURGXNWyZVSRĪ\ZF]\FKRUD]SRGMĊFLHSU]H]SURGXFHQWyZG]LDáDĔQDU]HF]REQLĪHQLDDNU\ORDPLGX szczególnie w tym rodzaju kawy.

1R(&>@7KLVKDGEHHQSHUIRUPHGE\WKH 6WDWH6DQLWDU\,QVSHFWRUDWHGXULQJE\VDP-pling from grocery shops and supermarkets throughout WKHZKROHRI3RODQG7KHVDPSOHGFRIIHHZHUHIURP brands that originated from 8 manufacturers of which 3 were Polish coffee roasting companies. One sample comprised at least two commercially available packages RIWKHVDPHQDPHPDQXIDFWXUHUDQGEDWFKQXPEHULQWKH minimum quantity of 1 kg. The methods of sampling and amounts taken were in accordance with EC Regulation 3DUW%1R>@

7KHURDVWHGFRIIHHVDPSOHVFRQVLVWHGRI$UDELFD species (&RIIHD$UDELFD5REXVWD (&RIIHDFDQHSKRUD) DQGZHUHDPL[WXUHRIERWK7KHUHZHUHVDPSOHVRI instant coffee of which 9 were a mixture of $UDELFD and

5REXVWD. The coffee substitutes samples were principally

composed of roasted cereals (ie. rye and/or barley) and chicory root. One sample contained also roasted beetroot. In accordance with specifications provided by manufactu-UHUVWKHGHJUHHRIFRIIHHURDVWLQJZDVPRGHUDWHLQWZHQW\ IRXUVDPSOHVOLJKWWRPRGHUDWHLQWZRFDVHVDQGOLJKWWR GDUNDOVRLQWZR2IWKHHOHYHQLQVWDQWFRIIHHVDPSOHV seven were freeze-drying whilst four were agglomerated. All samples were transported and stored under con-GLWLRQVVWDWHGRQWKHSURGXFWODEHOV3ULRUWRDQDO\VLV sample replicates were mixed together from which a representative portion was taken which then constituted the laboratory sample.

6DPSOHSUHSDUDWLRQ

The laboratory samples consisted of 1.8 to 3 g FRIIHH SODFHG LQ FHQWULIXJH WXEHV WR ZKLFK  WR  —/ RI LQWHUQDO VWDQGDUG ZHUH DGGHG $$G₃ at a FRQFHQWUDWLRQ RI  —J/ 1H[W GHLRQLVHG ZDWHU DQG KH[DQH ZHUH DGGHG IROORZHG E\ FHQWULIXJDWLRQ PLQVDWUSP$IWHUWKHSKDVHVKDGVHSDUDWHG the organic layer was removed. The aqueous layer was

incubated in an ultrasonic water bath. After cooling the VDPSOHVWRURRPWHPSHUDWXUH&DUUH],DQG,,VROXWLRQV were added. The sample was then further purified on DFDUERQFROXPQ%DNHUERQG&DUERQPJPO DV SUHYLRXVO\ GHVFULEHG >@ 7KH PHWKDQRO IUDFWLRQ so obtained was evaporated to dryness under nitrogen and then reconstituted in 1 ml of mixture of water and PHWKDQROYYRIZKLFK—/ZDVXVHGDVWKH injection volume for chromatographic analysis.

$QDO\VLVRIDFU\ODPLGHLQFRIIHHE\/&0606

Acrylamide in coffee was determined a previously RSWLPLVHGDQGYDOLGDWHG/&0606PHWKRG>@&KUR-matographic separation was achieved by the Dionex 8OWL0DWHV\VWHPRQD+\SHUFDUEFROXPQPP [PP—P7KHUPR6FLHQWLILFDIWHUDJXDUGFROXPQ with the same absorbent. Analytical conditions: flow rate RI  —/PLQ FROXPQ WHPSHUDWXUH RI ž& VDPSOH LQMHFWLRQYROXPH—/DQGDQLVRFUDWLFPRELOHSKDVH RIZDWHUPHWKDQROYYZLWKIRUPLFDFLG5XQ time was 5 minutes. Detection and quantification were SHUIRUPHGRQDQ$%6FLH[47UDSPDVVVSHFWURPHWHU XVLQJPXOWLSOHUHDFWLRQPRQLWRULQJ050LQWKHSRVLWL-ve ion mode with an electrospray capillary voltage (IS) of 97KHFDUULHUJDVZDVQLWURJHQ&85 DQGWKH LRQVRXUFHWHPSHUDWXUHZDVž&$VSUHYLRXVO\GHVFUL-EHG>@DFU\ODPLGHFRQFHQWUDWLRQVZHUHFDOFXODWHG by the peak areas under the curve originating from the IROORZLQJLRQWUDQVLWLRQVP]ĺ$$P] ĺ$$G₃ZKHUHDVWKHLRQVĺ$$ DQGP]ĺ$$G₃) were used for verification.

$VVHVVLQJURDVWHGDQGLQVWDQWFRIIHHFRORXULQJ

This was performed according to an established PHWKRGXVLQJDQPHWKRGRIDUUDQJLQJVHTXHQFH>@ +HUHDWHQPHPEHUWHDPUDWHGWKHJUDGHVRIFRORXULQJV LQRUGHURILQWHQVLW\VKRZQLQSKRWRV 6DPSOHV

 Photo 1. Samples of ground roasted coffee

 6DPSOHVVLJQLILFDQWO\OLJKWHVW

 6DPSOHVVLJQLILFDQWO\PRGHUDWHO\GDUN  6DPSOHVVLJQLILFDQWO\GDUNHVW

were arranged by members in order of increasing inten-VLW\DFFRUGLQJWRDWRVFDOHIRUURDVWHGFRIIHHDQG a 1 to 11 one for instant coffee. Results were recorded in custom-made reporting cards and then analysed by .UDPHU¶V tables where the upper division served to determine if there were any significantly lighter or darker colourings in samples amongst all those tested. The lower division was used to establish which of the samples differed.

6WDWLVWLFDODQDO\VLV

Results were calculated on a Microsoft Excel spre-DGVKHHWDQGVWDWLVWLFDOO\DQDO\VHGE\WKH6WDWLVWLFD YHUVLRQ6WDWVRIW,QF$FU\ODPLGHFRQFHQWUDWLRQVDUH given as the mean ± SD. ANOVA analysis was perfor-med to first determine if any significant differences existed between samples. When this proved to be the FDVHDSRVWKRF1,5WHVWZDVXVHGWRILQGZKHUHWKH\ lay. Linear relationships were assessed by the 3HDUVRQ correlation coefficient.

5(68/76

Concentrations of acrylamide in all the sample types of commercial coffee studied are shown in Table 1.

The highest mean level values was in coffee substitu-WHV—JNJIROORZHGE\LQVWDQWFRIIHH—JNJ whilst the lowest were for roasted coffee at 179 μg/kg (range: 61 – 397 μg/kg). The ANOVA analyses indicated that there were significant differences present between VDPSOHVS

,QPRVWRIWKHURDVWHGFRIIHHVDPSOHVQ DFU\-ODPLGHOHYHOVZHUHEHORZ—JNJRIZKLFKZHUH ORZHUWKDQ—JNJ$FU\ODPLGHFRQFHQWUDWLRQVLQD IXUWKHUVL[VDPSOHVUDQJHGEHWZHHQWR—JNJ whilst only two were higher at 386 and 397 μg/kg (Figu-UH)RUWKHLQVWDQWFRIIHHWKHUHZHUHVDPSOHVZLWK DFU\ODPLGHOHYHOVEHORZ—JNJDQGRQHVWKDWUDQ-JHGEHWZHHQ±—JNJ,QDIXUWKHUWZRVDPSOHV DFU\ODPLGHOHYHOVZHUHDVKLJKDVDQG—JNJ )LJXUH$VDIRUHPHQWLRQHGFRIIHHVXEVWLWXWHVKDGWKH KLJKHVWDFU\ODPLGHOHYHOVZKHUHWZRVDPSOHVIURPWKH VDPHPDQXIDFWXUHUEXWIURPGLIIHUHQWEDWFKHVVKRZHG DQG—JNJZKHQFRPSDUHGWRDVDPSOHIURP DQRWKHUPDQXIDFWXUHUDW—JNJ %HFDXVHDFU\ODPLGHLVKLJKO\ZDWHUVROXEOHLWFDQ be assumed that it the entire acrylamide content of com-mercial products will be present in any coffee infusion. Taking into account the variations of acrylamide levels REVHUYHGDQVLQJOHFXSRIGULQNLQJFRIIHHLVH[SHFWHG 3KRWR 6DPSOHVRILQVWDQWFRIIHH  6DPSOHVVLJQLILFDQWO\OLJKWHVW  6DPSOHVVLJQLILFDQWO\PRGHUDWHO\GDUN  6DPSOHVVLJQLILFDQWO\GDUNHVW 

Table 1. Comparison of acrylamide level in different types of coffee and coffee substitutes Type of coffee

Acrylamide content

>ȝJNJRIFRIIHH@ >ȝJFXSRIFRIIHH@

x ± SD min ÷ max x ± SD min ÷ max

ground roasted coffee1 _{179 ± 85}a* _{61 ÷ 397 “}a* _·

instant coffee _{358 ± 188}b _{· “}b _· coffee substitutes3 _“c _{· “}c _· *_{VWDWLVWLFDOO\VLJQLILFDQWGLIIHUHQFHS} 1 _{RQHWHDVSRRQJRIJURXQGURDVWHGFRIIHHZDVXVHGWRSUHSDUHDFXSRIFRIIHHPO} _{RQHWHDVSRRQJRILQVWDQWFRIIHHZDVXVHGWRSUHSDUHDFXSRIFRIIHHPO} 3 _{RQHWHDVSRRQJRIFRIIHHVXEVWLWXWHVZDVXVHGWRSUHSDUHDFXSRIFRIIHHPO}

0 50 100 150 200 250 300 350 400 450 acrylamide content [μg/kg] 0 1 2 3 4 5 6 7 8 9 10 num be r of s am pl es of ro as te d c offe e ‘check value’ for roasted coffee*

)LJXUH'LVWULEXWLRQRIVDPSOHVRIURDVWHGFRIIHHDFFRUGLQJWRWKHFRQWHQWRIDFU\ODPLGHQ  * _VRXUFH>@ 0 100 200 300 400 500 600 700 800 900 acrylamide content [ȝg/kg] 0 1 2 3 4 5 n u m b er o f s am p le s o f in st an t c offe e ‘check value’ for instant coffee*

)LJXUH'LVWULEXWLRQRIVDPSOHVRILQVWDQWFRIIHHDFFRUGLQJWRWKHFRQWHQWRIDFU\ODPLGHQ  * _VRXUFH>@ 162 188 ₁₈₀ 0 50 100 150 200 250 300 350 400

Arabica (n = 6) Robusta (n = 10) mixture of Arabica and Robusta (n = 12) ȝg/ kg of p rod uc t

WRFRQWDLQEHWZHHQ—JLQURDVWHGFRIIHHWR μg in coffee substitutes (Table 1).

The mean amount of acrylamide present in the

$UDELFDURDVWHGFRIIHHZDV—JNJZKLFKZDVORZHU EXWQRWVLJQLILFDQWO\WKDQPHDQOHYHOVPHDVXUHGLQWKH 5REXVWDVSHFLHVDW—JNJ)LJXUH &RIIHHRULJLQDWLQJIURPURDVWLQJIDFLOLWLHVLQ3RODQG KDGVOLJKWO\KLJKHUPHDQDFU\ODPLGHOHYHOVDW—JNJ FRPSDUHGWRWKRVHRIRWKHUPDQXIDFWXUHUVDW—JNJ but as before this difference was not significant.

/LNHZLVHWKHKLJKHUOHYHOVRIDFU\ODPLGHIRXQGLQ freeze-dried coffee compared to agglomerated coffee  —JNJ YV  —JNJ ZHUH DOVR QRW VLJQLILFDQW 7DEOH

7DEOH &RPSDULVRQRIDFU\ODPLGHOHYHOLQLQVWDQWFRIIHH depending on the way of production

Type of coffee Number of _samples

Acrylamide content >ȝJNJRIFRIIHH@ x ± SD min ÷ max

freeze-dried 7 “ ·

agglomerated  “ ·

The intensity of coffee colouring in the studied VDPSOHV DUH VKRZQ LQ SKRWRV  DQG  7KH UHVXOWV demonstrate significant differences of intensity in roasted coffees irrespective of the coffee species. In IDFWDQHJDWLYHFRUUHODWLRQZDVREVHUYHGU  SEHWZHHQDFU\ODPLGHOHYHOVDQGWKHLQWHQVLW\ RIFRIIHHFRORXULQJV)LJXUH7KHKLJKHVWVWDWLVWLFDOO\ VLJQLILFDQWSDFU\ODPLGHOHYHOVZHUHVHHQZKHQ WKHFRORXUZDVOLJKWHVWDW—JNJFRPSDUHGWRWKRVH 100 150 200 250 300 350 400 450 acrylamide content [μg/kg] 0 2 4 6 8 10 12 14 16 th e i nt ens ity of t he c ol or of ground roa st ed c offe e _{r = -0.5569; p = 0.0386} )LJXUH7KHFRUUHODWLRQEHWZHHQDFU\ODPLGHFRQWHQWLQFRIIHHDQGFRORULQWHQVLW\RIJURXQGURDVWHGFRIIHH 100 200 300 400 500 600 700 800 900 acrylamide content [ȝg/kg] 0 2 4 6 8 10 12 th e i n te n si ty of t h e c o lor o f i n st an t coffe e r = 0.4616; p = 0.1530

where the colour were moderately dark (197 μg/kg) or dark (168 μg/kg) (Table 3).

,Q FRQWUDVW WR URDVWHG FRIIHH WKHUH ZDV QR UHOD-tionship observed between acrylamide content and the intensity of colour for instant coffee (Figure 5). Less acrylamide was in instant coffee with the lightest colo-XULQJV—JNJWKDQFRIIHHZLWKPRGHUDWHO\GDUN or dark colourings at 381 μg/kg and 399 μg/kg respec-WLYHO\DOWKRXJKWKHVHGLIIHUHQFHVZHUHQRWVLJQLILFDQW (Table 3).

',6&866,21

The mean acrylamide content in the roasted coffee samples was 179 μg/kg ranging from 61 to 397 μg/kg depending on the product. Similar findings were obse-rved in a study by $QGU]HMHZVNLHWDO>@ZKLFKWHVWHG GLIIHUHQWFRIIHHW\SHVDYDLODEOHRQWKH86$PDUNHW ZKHUHDFU\ODPLGHOHYHOVUDQJHGIURPWR—JNJ Higher amounts of acrylamide were found in various coffee types in Denmark by *UDQE\ and )DJK >@ $FFRUGLQJWRWKLVVWXG\>@RQHOLWUHRIH[SUHVVRFRI-fee contains on average 9.5 μg of acrylamide (range 7 WR—JZKLFKLQWHUPVRIDNJRIFRIIHHJLYHVDQ DYHUDJHHTXLYDOHQWRI—JLWZDVDVVXPHGWKDW JRIURDVWHGFRIIHHZHUHUHTXLUHGWRPDNHXSOLWUH of drinking coffee). A smaller amount was demonstra-WHGE\DQRWKHUVWXG\FRQGXFWHGLQ3RUWXJDO6RDUHVet DO>@ZKHUHH[SUHVVRFRIIHHFRQWDLQHGDFU\ODPLGH from 57 to 181 μg/kg coffee (the authors claiming that DQH[SUHVVRFRIIHHLVPDGHXSRIJH[WUDFWZLWKPO ZDWHU +RZHYHU PXFK ORZHU DFU\ODPLGHFRQFHQWUD-tions were demonstrated by 6HQ\XYDand *RNPHQ>@ LQ7XUNH\ZLWKOHYHOVLQJURXQGURDVWHGFRIIHHUDQJLQJ IURPWR—JNJ

The mean levels of acrylamide in instant coffee were found to be 358 μg/kg which was twice that in roasted coffee (179 μg/kg). Similar results were obtained by RWKHUVWXGLHV$QGU]HMHZVNLHWDO>@DQG6RDUHV et al >@ ,Q WKH 86$ WKH DFU\ODPLGH FRQWHQW RI LQVWDQW FRIIHHUDQJHGIURPWR—JNJZKHUHDVWKHPHDQ acrylamide levels in commercial products in Portugal ZHUH—JNJ>@/RZHUDFU\ODPLGHFRQFHQWUDWLRQV than those presented were seen in instant coffee studies

IURP7XUNH\>@LHRIWR—JNJ6RPHZKDWKL-gher acrylamide levels (8 μg per litre of drinking coffee) ZHUHVHHQLQWKHDIRUHPHQWLRQHG'DQLVKVWXG\>@RI

*UDQG\ and )DJK7DNLQJLQWRDFFRXQWWKDWJRI

FRIIHHDUHXVHGSHUOLWUHLQGULQNLQJFRIIHHWKHQLQVWDQW coffee in the Danish study contained 597 μg acryla-mide/kg of commercial product. A monitoring study FRQGXFWHGWKURXJKRXW(XURSH>@GXULQJ GHPRQVWUDWHGVOLJKWO\KLJKHUDFU\ODPLGHOHYHOVZKHUH RQDYHUDJHLQVWDQWFRIIHHFRQWDLQHG—JNJUDQJLQJ EHWZHHQWR—JNJ

The presented study showed that coffee substitutes FRQWDLQ RQ DYHUDJH  WLPHV WKH DPRXQW RI DFU\OD-mide found in roasted coffee. The results are similar to that demonstrated in the aforementioned European VWXG\>@ZKHUHWKHPHDQDFU\ODPLGHOHYHOVIRXQGLQ FRIIHHVXEVWLWXWHVZHUH—JNJZLWKDPD[LPXP RI  —JNJ$FU\ODPLGH FRQFHQWUDWLRQV LQ FRIIHH substitutes depend on the composition of products. These products contain chicory root and roasted ce-UHDOVHJU\HEDUOH\DQGVRPHDOVRURDVWHGEHHWURRW All these ingredients contain acrylamide precursors ie. reducing sugars such as glucose and fructose as well as asparagine. The differences in acrylamide levels found may be accounted for by the variety of coffee substi-tutes sampled and of having different manufacturers and differing raw materials that thereby influence the ILQDO FRPSRVLWLRQ RI SURGXFW )RU H[DPSOH FKLFRU\ URRWVFRQWDLQWZLFHDVPXFKIUXFWRVHDVJOXFRVH>@ where the former is more reactive than the latter during DFU\ODPLGHIRUPDWLRQ)XUWKHUPRUHLWKDVEHHQQRWHG by :RURELHMand 5HOLG]\ĔVND>@WKDWWKHUHDUHODUJH differences in reducing sugars present in coffee substi-WXWHVZKLFKWKXVHIIHFWWKHOHYHOVRIDFU\ODPLGHIRXQG in a given final product.

It should be stated that none of the roasted coffee or instant coffee samples showed acrylamide levels WKDWH[FHHGHGWKHFKHFNYDOXHRIDQG—JNJ UHVSHFWLYHO\ DV VWDWHG LQ (& UHFRPPHQGDWLRQV IURP >@$WSUHVHQWWKHUHDUHKRZHYHUQRVXFKFKHFN values existing for coffee substitutes.

Although the study showed that acrylamide levels were lower in $UDELFD coffee than in 5REXVWD or mixtu- UHVWKHUHRIWKHVHGLIIHUHQFHVZHUHQRWVLJQLILFDQW+R-ZHYHUOLNHILQGLQJVKDYHEHHQVKRZQLQDQRWKHUVWXG\ by /DQW]HWDO>@ZLWKDFU\ODPLGHOHYHOVRIDQG —JNJLQUHVSHFWLYHO\5REXVWD and $UDELFDFRIIHH

Table 3. Comparison of acrylamide level in roasted and instant coffee depending on the intensity of colour Intensity of coffee color1 5RDVWHGFRIIHHQ  ,QVWDQWFRIIHHQ 

Number of samples x ± SD Number of samples x ± SD

VLJQL¿FDQWO\OLJKWHVW  “* _{ “}

VLJQL¿FDQWO\PRGHUDWHO\GDUN 5 168 ± 55 3 381 ± 119

VLJQL¿FDQWO\GDUNHVW 5 197 ± 69  “

1 _{evaluated on the basis of .UDPHU¶Vtables} *_{VWDWLVWLFDOO\VLJQLILFDQWGLIIHUHQFHS}

WKHGLIIHUHQFHVDOVREHLQJQRWVLJQLILFDQW,QDGGLWLRQ a study by $OYHVHWDO>@GHPRQVWUDWHGWKDW5REXVWD coffee had significantly higher acrylamide than $UDELFD FRIIHHYV—JNJWKHODWWHUGHSHQGLQJRQ the country from which the beans originated. It appe-DUVWKDWGXULQJURDVWLQJPRUHDFU\ODPLGHLVIRUPHGLQ

5REXVWD coffee than $UDELFDFRIIHHEXWXSWLOOQRZWKH

determining factors remain unexplained. Nevertheless it has been suggested that fresh 5REXVWD coffee beans FRQWDLQPRUHDVSDUDJLQH>@WKDQ$UDELFD ones.

The instant coffee products that appear on the Polish PDUNHWYDU\LQWKHPHWKRGVRIPDQXIDFWXUHHVSHFLDOO\ in how the coffee is dehydrated. When coffee is freeze-GU\LQJIURPIUR]HQH[WUDFWVWKHLFHLVUHPRYHGWKRXJK ZDWHUHYDSRUDWLRQKRZHYHULQDJJORPHUDWHGFRIIHHWKH extract is dried under streams of hot air up to the moment that the dissolvable granules are formed. Despite such GLIIHUHQFHVWKHFXUUHQWVWXG\GLGQRWVKRZDQ\VLJQLIL-cant association between these methods of production ZLWKDFU\ODPLGHOHYHOVWKHDSSDUHQWLQFUHDVHRI acrylamide in freeze-dried coffee was not significant.

$QHJDWLYHDQGVLJQLILFDQWFRUUHODWLRQSZDV demonstrated between acrylamide concentrations and the intensity of color which was consistent with other ZRUN7DH\PDQVHWDO>@DQG*RNPHQ& 6HQ\XYD>@ Lower acrylamide levels in dark coloured coffee com-pared to the light coloured ones reflect the sensitivity of DFU\ODPLGHWRGHFRPSRVHGXULQJURDVWLQJDFU\ODPLGH is principally formed at the start of the roasting process but its content decreases the longer the roasting conti-QXHV>@(YHQWKRXJKWKHUHZHUHGLIIHUHQFHVLQ FRORXULQJZLWKLQWKHLQVWDQWFRIIHHVDPSOHVWKLVZDV found not to be linked to acrylamide concentrations.

The presented study confirms that the roasting process is a key factor in determining the acrylamide levels in the finished product. The species bean (ie. raw PDWHULDO KDV D VPDOOHU HIIHFW DOWKRXJK LW ZRXOG EH worth checking the acrylamide content in $UDELFD and

5REXVWD coffee according to country of origin.

It can be summarised that coffee forms a significant source of dietary acrylamide within the population of persons consuming high amounts of coffee. Of special concern is the high acrylamide levels in coffee sub-VWLWXWHVZKLFKRIWHQWDNHVWKHSODFHRIQDWXUDOFRIIHH especially in the more vulnerable population groups such as children and pregnant or breast feeding women. It is therefore vital that check values are established for FRIIHHVXEVWLWXWHVWKDWQXWULWLRQDOUHFRPPHQGDWLRQVDUH modified accordingly in affected groups and that targe-WHGHGXFDWLRQLVXQGHUWDNHQ)XUWKHUPRUHEHFDXVHRIWKH wide variations of acrylamide concentrations within the VDPHJURXSRIFRIIHHSURGXFWVPDQXIDFWXUHUVVKRXOG take whatever steps are necessary to reduce such levels.

&21&/86,216

1. The study findings confirm that coffee can be an important dietary source of acrylamide. Amongst WKHYDULRXVFRIIHHW\SHVFRIIHHVXEVWLWXWHVKDGWKH highest mean acrylamide concentrations (818 μg/kg) followed by instant coffee (358 μg/kg) and then by roasted coffee (179 μg/kg).

 :HGLGQRWILQGVLJQLILFDQWGLIIHUHQFHLQDFU\ODPL-de content between coffee types or mixtures (ie.

$UDELFD vs 5REXVWD vs mixtures thereof) nor in the

methods of manufacturing instant coffee (ie. freeze--drying vs agglomeration).

3. We found a negative correlation between acrylamide levels in roasted coffee and the intensity of color of coffee. This was not observed with instant coffee.  7KH SUHVHQWHG VWXG\ FRQILUPV WKDW WKH SURFHVV RI

roasting coffee beans is a significant factor in deter-mining the acrylamide content in the final product. ,QFRIIHHVXEVWLWXWHVWKLVFULWLFDOIDFWRULVWKHFRP-position of the raw material.

Acknowledgements

7KHVWXG\IRUPHGSDUWRIDVFLHQWLILFSURMHFWHQWLWOHG µ6WXGLHV WR GHWHUPLQH VXEVWDQFHV IRUPLQJ GXULQJ IRRGVWXIISURFHVVLQJDQGWKHLULPSDFWRQKXPDQKHDOWK¶ XQGHUWDNHQE\WKH1DWLRQDO)RRGDQG1XWULWLRQ,QVWLWXWH :DUVDZ3RODQG1R'. Conflict of interest 7KHDXWKRUVGHFODUHQRFRQIOLFWRILQWHUHVW 5()(5(1&(6

1. $OYHV 5& 6RDUHV & &DVDO 6 )HUQDQGHV -2 %H

DWUL]02OLYHLUD33.: Acrylamide in espresso coffee:

,QIOXHQFHRIVSHFLHVURDVWGHJUHHDQGEUHZOHQJWK)RRG &KHPLVWU\

 $QGU]HMHZVNL'5RDFK-$**D\0/0XVVHU60.: Analysis of coffee for the presence of acrylamide by LC-0606-$JULF)RRG&KHP 3. %DJGRQDLWH.'HUOHU.0XUNRYLF0.: Determination

RIDFU\ODPLGHGXULQJURDVWLQJRIFRIIHH-$JULF)RRG &KHP

 %HODQG)$0HOOLFN3:2OVRQ*50HQGR]D0&%

0DUTXHV00'RHUJH'5: Carcinogenicity of

acry-ODPLGHLQ%&)₁PLFHDQG)1UDWVIURPD\HDU GULQNLQJZDWHUH[SRVXUH)RRG&KHP7R[LFRO 

5. %HUJPDUN (: Hemoglobin adducts of acrylamide and DFU\ORQLWULOH LQ ODERUDWRU\ ZRUNHUV VPRNHUV DQG QRQ-VPRNHUV&KHP5HV7R[LFRO

6. %RRQ3(GH0XO$YDQGHU9RHW+YDQ'RQNHUVJRHG

H[SRVXUHWRDFU\ODPLGH0XWDWLRQ5HVHDUFK 

7. %UXQQHU . 'XGOHU 9 5HLQKDUG + 5K\Q 3 5XSS

+ 6DJHU ) 6WUHXOH 0 =LPPHUPDQQ + =ROOHU 2:

Assessment of acrylamide intake by duplikate diet study. KWWSZZZEIUEXQGGHFPDVVHVVPHQWBRIBDFU\OD-mide_intake_by_duplicate_diet_study.pdf

8. &KXVWHF]NL3ĝZLGHUVNL).: Zastosowanie metod analizy VHQVRU\F]QHMZRFHQLHMDNRĞFLSURGXNWyZVSRĪ\ZF]\FK :7HRULDLüZLF]HQLD]WRZDUR]QDZVWZDSURGXNWyZVSR-Ī\ZF]\FK5HGĝZLGHUVNL):\GDZQLFWZR6**:$5 Warszawa 1987.

 &RPPLVVLRQ 5HFRPPHQGDWLRQ RI  RQ LQYH-stigations into the levels of acrylamide in food. http:// ec.europa.eu/food/food/chemicalsafety/contaminants/ UHFRPPHQGDWLRQBBDFU\ODPLGHBIRRGBSOSGI  &RPPLVVLRQ 5HFRPPHQGDWLRQ RI  0D\  RQ WKH PRQLWRULQJRIDFU\ODPLGHOHYHOVLQIRRG(& 2II-(XU8QLRQ/  &RPPLVVLRQ5HJXODWLRQ(&1RRI0DUFK OD\LQJGRZQWKHPHWKRGVRIVDPSOLQJDQGDQDO\VLV IRUWKHRIILFLDOFRQWURORIWKHOHYHOVRIOHDGFDGPLXP PHUFXU\LQRUJDQLFWLQ0&3'DQGEHQ]RDS\UHQHLQ IRRGVWXIIV2II-(XU8QLRQ/  '\ELQJ(6DQQHU7: Forum. Risk Assessment of

Acryl-DPLGHLQ)RRGV7R[LFRORJLFDO6FLHQFHV 13. ()6$: Scientific Report. Results acrylamide levels in

IRRGIURPPRQLWRULQJ\HDUV±DQGH[SRVXUH DVVHVVPHQW()6$-RXUQDOwww.efsa. HXURSDHXHQHIVDMRXUQDOSXEKWP

 )ULHGPDQ0$'XDN/+6WHGKDP0$.: A lifetime oncogenicity study in rats with acrylamide. Fundam. $SSO7R[LFRO

15. *LHOHFLĔVND,0RMVND+ Optimisation and validation of the analytical procedure for the determination of acrylamide in coffee by LC-MS/MS using SPE clean-up. 5RF]3DQVWZ=DNO+LJ

16. *|NPHQ9ùHQ\XYD+=.: Study of colour and acrylamide IRUPDWLRQLQFRIIHHZKHDWIORXUDQGSRWDWRFKLSVGXULQJ KHDWLQJ)RRG&KHPLVWU\

17. *UDQE\.)DJW6.: Analysis of acrylamide in coffee and dietary exposure to acrylamide from coffee. Analytica &KLPLFD$FWD

18. +H)6+DQJ6/:DQJ+//L*+DQJ=0/L)/

'RQJ;0+X): Neurological and

electroneuromyo-graphic assessment of the adverse effects of acrylamide RQRFFXSDWLRQDOO\H[SRVHZRUNHUV6FDQG-:RUN(QYL-URQ+HDWK

19. Health implications of acrylamide in food. Report of D -RLQW )$2:+2 &RQVXOWDWLRQ :+2 +HDGTXDUWHUV *HQHYD6ZLW]HUODQG-XQH  ,QWHUQDWLRQDO$JHQF\IRU5HVHDUFKRQ&DQFHU,$5&: $FU\ODPLGH,$5&PRQRJUDSKVRQWKHHYDOXDWLRQRIFDU-FLQRJHQLFULVNVWRKXPDQV6RPHLQGXVWULDOVFKHPLFDOV YRO,QWHUQDWLRQDO$JHQF\IRU5HVHDUFKRQ&DQFHU /\RQ)UDQFH±

 -(&)$ -RLQW )$2:+2 ([SHUW &RPPLWWHH RQ IRRG additives: evaluation of certain food additives and con-WDPLQDQWV nd _{report of the joint FAO/WHO expert}

committee on food additive. WHO Technical Report 6HULHV

 -RKQVRQ.*RU]LQVNL6%RGQDU.&DPSEHOO5:ROI

&)ULHGPDQ00DVW5: Chronic toxicity and

oncoge-nicity study on acrylamide incorporated in the drinking ZDWHURI)LVKHUUDWV7R[LFRO$SSO3KDUPDFRO 

 .RQLQJV(-0%DDUV$-YDQ.ODYHUHQ-'6SDQMHU

0&5HQVHQ30+LHPVWUD0YDQ.RRLM-$3HWHUV 3:-: Acrylamide exposure from foods of the Dutch

population and an assessment of the consequent risk. )RRG&KHP7R[LFRO

 /DQW],7HUQLWH5:RONHQV-+RHQLFNH.*XHQWKHU

+YDQGHU6WHJHQ*+'.: Studies on acrylamide levels

LQURDVWLQJVWRUDJHDQGEUHZLQJRIFRIIHH0RO1XWU )RRG5HV

 0LODOD-*U]HODN..UyO%-XĞNLHZLF]-=GXĔF]\N

=: Composition and properties of chicory extracts rich

LQIUXFWDQVDQGSRO\SKHQROV3RO-)RRG1XWU6FL 

 0RMVND + *LHOHFLĔVND , 6WRĞ . Determination of acrylamide level in commercial baby foods and an asses-sment of infant dietary exposure. Food Chem. Toxicol. 

 0RMVND+*LHOHFLĔVND,6]SRQDU/2áWDU]HZVNL0.: Estimation of the dietary acrylamide exposure of the 3ROLVKSRSXODWLRQ)RRG&KHP7R[LFRO 

 ùHQ\XYD+=*|NPHQ9.: Study of acrylamide in coffee using an improved liquid chromatography mass spectro-metry method: Investigation of colour changes and acrylamide formation in coffee during roasting. Food $GGLW&RQWDP

 6RDUHV & &XQKD 6 )HUQDQGHV -.: Determination of acrylamide in coffee and coffee products by GC-MS using an improved SPE clean-up. Food Addit. Contam. 

 6YHQVVRQ.$EUDPVVRQ/%HFNHU:*O\QQ$+HO

OHQlV.(/LQG<5RVpQ-.: Dietary intake of

acryl-DPLGHLQ6ZHGHQ)RRG&KHP7R[LFRO 1586.

31. 7DH\PDQV ' :RRG -$VKE\ 3 %ODQN , 6WXGHU$

6WDGOHU5+*RQGp3YDQ(LMFN3/DOOMLH6/LQJQHUW +/LQGEORP00DWLVVHN50OOHU'7DOOPDGJH' 2¶%ULHQ - 7KRPSVRQ 6 6LOYDQL ' :KLWPRUH 7.: A

review of acrylamide: An industry perspective on rese-DUFKDQDO\VLVIRUPDWLRQDQGFRQWURO&ULW5HY)RRG 6FL1XWU

 7KH &,$$: Food Drink Europe. Acrylamide Toolbox  http://ec.europa.eu/food/food/ chemicalsafety/ FRQWDPLQDQWVFLDDBDFU\ODPLGHBWRROER[SGI

33. :RURELHM(5HOLG]\ĔVND.: Cereal coffees – characteris-WLFDQGDQWLR[LGDQWSURSHUWLHV%URPDW&KHP7RNV\NRO LQ3ROLVK

5HFHLYHG $FFHSWHG