A general postlaunch monitoring framework for functional foods tested with the phytosterol/-stanol case

The regulations and/or directives in force for functional foods primarily focus on the warrant of safety before the particular foods reach the consumer. Aspects that come into the picture after marketing are not structurally and/or regulatory dealt
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Viewpoint A general postlaunchmonitoringframework forfunctional foodstested with thephytosterol/-stanolcase Nynke de Jong*, Martine M.Ros, Marga C. Ocke´ and HansVerhagen National Institute for Public Health and theEnvironment (RIVM), Centre for Nutrition and Health,P.O. Box 1, 3720 BA Bilthoven, The Netherlands (Tel.: D 31 30 274 4135; fax:  D 31 30 274 4466; e-mail:nynke.de.jong@rivm.nl) The regulations and/or directives in force for functional foodsprimarily focus on the warrant of safety before the particularfoods reach the consumer. Aspects that come into the pictureafter marketing are not structurally and/or regulatory dealt withat this moment. This absence of clear guidelines about respon-sibility, timing and contents of a postlaunch monitoring (PLM)system hamper the establishment of an internationally stan-dardized and stakeholder-adopted framework. The current pa-per describes a proposal for PLM and is illustrated with a casestudy on phytosterols/-stanols. Introduction Consumption of functional foods is increasing and thismay exert benefits for the particular user, although potentialhealth risks cannot be excluded. The current set of EU rulesconcerning functional foods focuses among others on themandatory premarketing safety evaluation for novel foods,the minimum and maximum safe Upper Levels for vitaminand mineral fortification, positive lists of substances to beallowed for fortification, registration of herbal productsand guidance for acceptable nutrition and health claims(see for example, http://www.efsa.europa.eu/en.html andhttp://ec.europa.eu/food/intro_en.htm). The regulationsand/or directives in force primarily focus on the warrantof safety before the particular foods reach the consumer.Aspects that come into picture after marketing are notstructurally and/or regulatory dealt with. For example, wehardly know anything about (long term) exposure andsafety under free conditions of use, such as adverse effectsin potential risk groups, over consumption of specific ingre-dients, and interaction effects with nutrients and/or drugs.A postlaunch monitoring (PLM) system for functionalfoods deals with these issues. Experiences with PLM fornovel foods are limited to a few cases (Allgood, Kuter,Roll, Taylor, & Zorich, 2001; Butchko & Stargel, 2001;Lea & Hepburn, 2006; Wal, Hepburn, Lea, & Crevel,2003). A lot of this postmarket research effort has flowninto the investigation of safety aspects instead of effective-ness. Nevertheless, with an increasing supply of functionalfoods there is an increasing interest in PLM in the regula-tory agencies and the scientific community. The  objective of a PLM system is to systematically monitor (un)expectedhealth effects of functional food consumption after market-ing and under customary conditions of use. However, theabsence of clear guidelines about responsibility, timingand contents of PLM hamper the establishment of an inter-nationally standardized and stakeholder-adopted system.We have developed a theoretical framework for PLM.Also we have tried out this framework with a case studyon phytosterol/-stanol enriched margarines, but the frame-work should be able to handle other types of products aswell. A general PLM approach An overall PLM framework should consist of the follow-ing phases (Fig. 1):(a) passive signalling of consumer complaints through forexample consumer care telephone lines; * Corresponding author. 0924-2244/$ - see front matter    2008 Elsevier Ltd. All rights reserved.doi:10.1016/j.tifs.2008.05.001 Trends in Food Science & Technology 19 (2008) 535 e 545  (b) active signalling of hazardous and/or potential benefi-cial health effects based on active (pre- and postmarket)data searching;(c) assessment of the relevance of the data from (a) and(b);(d) quantification of the effects on a population (group)level;(e) balancing the beneficial (positive) and the hazardous(negative) effects, i.e. risk  e benefit assessment; and(f) risk  e benefit management.Below we will describe each phase separately and willadd specific PLM case study information about phytos-terol/-stanol enriched foods where possible. PLM necessary for phytosterol/-stanol enriched foods? The applicability of the proposed PLM system was ex-plored in a case study on phytosterols/-stanols. These areused as a novel food ingredient with plasma cholesterollowering activity. A wide variety of phytosterol structuresexist but the phytosterols found most frequently in natureare  b -sitosterol, campesterol and stigmasterol. Phytosta-nols are saturated phytosterols and the major phytostanolsare sitostanol and campestanol. Due to their structuralsimilarity with cholesterol, phytosterols/-stanols presum-ably displace cholesterol from mixed micelles thereby in-hibiting the absorption of cholesterol from the intestine,albeit that the exact mechanism remains unknown (Katan et al. , 2003). Many premarket controlled clinical trialshave demonstrated that intake of 2 g/day of phytos-terols/-stanols reduces serum LDL cholesterol concentra-tions by approximately 10% (Katan  et al. , 2003; Law,M., 2000; Law, M. R., 2000). A reduction in the risk of heart disease of about 25% after about 2 years wouldbe expected for this 10% reduction in LDL cholesterol(Law, M., 2000). In addition, the safety of phytos-terols/-stanols has been studied extensively (Baker et al. , 1999; Hepburn, Horner, & Smith, 1999; Lea, Hep-burn, Wolfreys, & Baldrick, 2004; Sanders, Minter,Howes, & Hepburn, 2000; Waalkens-Berendsen, Wolter-beek, Wijnands, Richold, & Hepburn, 1999; Weststrate,Ayesh, Bauer-Plank, & Drewitt, 1999; Ayesh, Weststrate,Drewitt, & Hepburn, 1999; Wolfreys & Hepburn,2002).The US Food and Drug Administration has ac-cepted that phytosterols/-stanols are Generally Recog-nized as Safe (GRAS). The former Scientific Committeeon Foods (SCF) of the EU also approved the market in-troduction of yellow fat spreads enriched with phytos-terols according to the Novel Foods Regulation 258/ 1997 (Scientific Committee on Food, 2000; ScientificCommittee on Food, 2002). Phytostanols were launchedon the market prior to Regulation 258/1997 and thereforeare not legally regarded a novel food ingredient. Never-theless, safety and efficacy studies have been done forthis ingredient as well (Slesinski, Turnbull, Frankos, Wol-terbeek, & Waalkens-Berendsen, 1999; Turnbull, Frankos,Leeman, & Jonker, 1999; Turnbull, Frankos, van Delft, &DeVogel, 1999; Turnbull, Whittaker, Frankos, & Jonker,1999). The results of the (pre)-market assessment indicatethat phytosterols/-stanols have great potential in the pre-vention of coronary heart disease. However, lack of long term experience in a free-living situation leavesa possibility of unforeseen effects. In several boxes wewill describe in what manner we may be able to fit thecurrent knowledge about safety and effectiveness of phy-tosterols/-stanols in some of the PLM phases. To stick tothe case study objectives we mainly describe figures ap-plicable to the Netherlands. Passive signallingof consumer complaints (a) Active signallingof health effects basedon investigation of data (b) Assessment of the relevanceof the data from a and b Quantification of the health effects, ona population/group levelBalancing beneficial and hazardous effects,i.e. benefit-risk analysese PLM system for health effects FooddatabaseSupplementsdatabase Assessment of dietary intake Assessment of nutritionaland health status Risk/Benefit managements Fig. 1 . Graphical presentation of the PLM paradigm. 536  N. de Jong et al. / Trends in Food Science & Technology 19 (2008) 535  e 545   Signalling phase Both the active as well as the passive signalling of healtheffects have the objective to evaluate whether there is evi-dence of any potential health effect that can be related tofunctional food exposure that might need further researchand follow-up. Passive signalling With passive signalling we mean the collection, regis-tration and analyses of consumer complaints, such as themanufacturers’ consumer care lines, the care lines of thenational food safety authorities and/or consumer organi-sations. Another option may be the (inter)national phar-macovigilance centres on side effects of drugs, as theyhave general expertise on data management, signal detec-tion and causality assessment (van Puijenbroek, Hepburn,Herd, & van Grootheest, 2007). In the Netherlandscaregivers, patients and manufacturers are able to reportdrug-related side effects and submit this to LAREB (TheNetherlands Pharmacovigilance Centre: www.lareb.nl). In-formation about the side effect, the suspected drug, per-sonal patient data, details from the prescriber andproducer have to be filled out. If necessary, additional in-formation for causality assessment (e.g. blood measure-ments and scan results) can be obtained as well.LAREB divides signal analyses into three steps. First,each individual report is evaluated by a single expertand a multidisciplinary scientific staff. Second, statisticalcalculations are carried out weekly on the existing data-base to examine if a certain combination of reported sideeffect and drug occurs more often than can be ascribedto a random statistical likelihood. Third, a comparisonwith data from the WHO database is made. The overallevaluation is based on an assessment of all available(predominantly clinical) data of the individual patientand additional clinical data retrieved through medicalspecialists. Important factors that play a role in causalityassessment are as follows: time span, plausibility, biolog-ical working mechanisms and exclusion of other explana-tions such as epidemics of viruses, cosmetics etcetera. Incase of supposed serious adverse effects, manufacturersare obliged to report on these effects to the Dutch gov-ernmental Medicines Evaluation Board, which may thendecide on additional investigations, adaptation of thedrug-accompanying instruction leaflets or drug removalfrom the market. Theoretically, it should be possible toimplement a similar registration system for consumercomplaints on functional foods. As such, attaining knowl-edge on the many confounding variables and on food e drug interactions is an effort well spent. The downsideof the dedication of drug-oriented bodies is the relativeunfamiliarity of the public to report food-related com-plaints to these types of institutes, and the apprehensionof food industries to get associated to the drug safetynetwork. A link between pharmacovigilance centres andthe national Food Safety Authorities may be anacceptable option as the latter institutes are better knownby the consumers. A presentation of the potential playersin the PLM field is presented in Fig. 2. Active signalling In addition, also active signalling should be opera-tional. As it is neither feasible nor necessary to focuson each and every functional food (ingredient), criteriashould be developed to prioritize the focus of PLM.Also, the methodology should be developed, as well asthe decision criteria that indicate which signals shouldbe followed-up. To start with the prioritisation require-ment: ranking the importance of the topics taking into ac-count that a low score does not mean full stop but onlydelayed follow-up. Ranking may be based, e.g., on thetype of functional food (e.g. ‘novel’  vs.  ‘non-novel’ ingre-dients), potential consumption levels from all foods con-taining the same ingredients, the window between therecommended intakes and the safe Upper Levels, serious-ness of side effects that may occur in humans, and timespan that has passed after market introduction. It is possi-ble to separate the prioritisation step into two sections:one with a product focus and one with an ingredient focus(Table 1).With respect to the methodology, one could start withthe listing of effects reported in the literature, sales vol-umes, consumption levels calculated through nationalfood consumption surveys, the safe Upper Levels, therecommended daily intakes, and the no or lowest adverseeffect levels, etc. in line with the FDA approach (Com-mittee on the Framework for Evaluating the Safety of Di-etary Supplements, 2003) for a framework to evaluate thesafety of dietary ingredients. Liaison with stakeholders isimportant to retrieve information. We propose to havea final PLM ‘go’ or ‘delayed go’ decision been laiddown at an  ad hoc  expert committee existing of stake-holders (e.g. scientists, government, and consumer repre-sentatives). This committee should take the socialrelevance of a potential case into account. A case mightbe relevant if the health effect will occur in a few per-cent of the population, but a case might also be relevantif the health effect is severe in only a very specific groupof the population. The decision making process of the in-dependent committee should be transparent. The latterhas to develop a systematic approach to assess the seri-ousness of the topic. Active signalling for phytosterol/-stanolenriched foods Prioritisation on product level Products enriched with phytosterols/-stanols are innovativetypes of functional foods. In the Netherlands, phytostanol-enriched margarines (Benecol  ) and phytosterol-enrichedmargarines (Becel pro.activ  ) are on the market since 1999and 2000, respectively. Since 2006 the phytosterol-enriched 537 N. de Jong et al. / Trends in Food Science & Technology 19 (2008) 535  e 545   low-fat margarine, Reducol  has been introduced. Otherphytosterol-enriched food products including dairy prod-ucts have been introduced from 2003 onwards. The brandsBecel pro.activ  and Reducol  launched also yoghurttype products, and Becel pro.activ  also includes a milk drink and a yoghurt drink in different flavours. Recently,a phytosterol-enriched cheese (ColActif   ) became avail-able in a number of Dutch supermarkets. This means thatoverall at least four to five different types of products areavailable. Prioritisation on ingredient level The recommended daily intake is equivalent to 2 e 3 g of phytosterols/-stanols and is regarded as the optimal dose toreduce the LDL cholesterol levels effectively by about10%. This is mainly achieved by 2 e 3 servings per dayfrom the range of foods containing phytosterols/-stanols.The 100 ml yoghurt drinks are an exception as this productdelivers the recommended dose in one 100 ml serving. Asno further LDL cholesterol reductions are achieved with in-takes above 3 g/day, the SCF concluded to avoid intakesabove 3 g of phytosterols/-stanols per day. Although thelevel of 3 g/day is not really an established upper level,we are left with the conclusion that there is a small rangebetween the recommended intake and unnecessarily highlevels of intake of phytosterols/-stanols. PLM is indicatedfor these types of foods. Signal detection based on effects Up till now, few signals of (un)expected health effects of phytosterol/-stanol have been reported. A concern that wasdiscovered before product launch is that phytosterols/-stanolsdecrease serum  a - and  b -carotene levels (Scientific Com-mittee on Food, 2002). Some data have been presentedthat higher plasma levels of phytosterols are associatedwith risk of coronary heart disease, suggesting that phytos-terols may be atherogenic (Sudhop, Gottwald, & von Berg-mann, 2002). There are also suggestions that theconsumption of phytosterols may promote the developmentof stroke (Ratnayake  et al. , 2000; Naito, Nagata, Takano,Nagatsu, & Ohara, 2003) or lead to the formation of un-wanted hormones from  b -sitosterol and have an impacton estrogenic activity (Malini & Vanithakumari, 1993; Mel-lanen  et al. , 1996). An increase in the fecal concentration of cholesterol has been reported to be mutagenic in some  invitro  tests (Kaul, Couch, Gingerich, Bruce, & Heddle,1987; Suzuki  et al. , 1986). Several studies have indicatedan additive effect of phytosterols on LDL cholesterol low-ering when combined with a statin. However, the use of sta-tins has also been associated with an increase of cholesterol-standardized serum phytosterol concentration,which may so induce extra atherogenicity (Miettinen,Strandberg, & Gylling, 2000). On a behavioural level,one might think of interference of phytosterol/-stanol en-riched foods in a patient’s drug therapy compliance, whichmight induce a potential downside albeit that this remainsto be investigated. On the beneficial edge, in addition totheir cholesterol lowering properties, consumption of phy-tosterols/-stanols may also possess anti-cancer, anti-inflam-matory, anti-atherogenicity, and anti-oxidant activities (seefor an overview Berger, Jones, & Abumweis, 2004) but ev-idence for this is still speculative. Signal detection based on exposure An evaluation of the anticipated intake has been carriedout by the manufacturer before the phytosterol-enrichedmargarine (Becel pro.activ  ) was launched on the market.Phytosterols are added to spreads at a level of 8% as Next PLM phase Consumer complaintsManufacturers Consumer care line Nat. Food Safety Authority Consumer care line Pharmacovigilancecentre Risk managers/ Policy makers EU – Food Safety AuthorityNat. food/nutritionresearch institutes Severe casesTo be organized ina transparant way Reports back if necessary Fig. 2 . Presentation of the potential players in the PLM field. 538  N. de Jong et al. / Trends in Food Science & Technology 19 (2008) 535  e 545   approved by the SCF. Based on an estimated consumptionof 27 (Eurostat data) or 22 (Dutch data) g/day margarine,phytosterol intakes of 2.2 or 1.8 g/day, respectively, wouldresult. As part of Unilever’s PLM system, the estimateddaily intake of Becel Pro.activ  was obtained using marketresearch data. These data showed that the amount of marga-rine purchased per Dutch households varied between 12and 18 g/day (0.96 e 1.44 g/day of phytosterols). The upper(95th percentile) intakes varied between 30 and 45 g/day(2.4 e 3.6 g/day of phytosterols) (Lea & Hepburn, 2006).The intake of cholesterol lowering margarines in the free-living population has been investigated in a Dutch cohortstudy (Wolfs, de Jong, Ocke´, Verhagen, & Verschuren,2006). Between 1999 and 2003  ca.  2% of the populationused cholesterol lowering margarines. The mean use of phytosterol-enriched margarines was 15  8 g/day(1.2  0.64 g/day of phytosterols) and of phytostanol-en-riched margarines this was 9  6 g/day (0.72  0.48 g/dayof phytostanols). The percentage of users that ate lessthan the recommendations was 81%. In another Dutch co-hort  ca.  5% of the population used cholesterol loweringmargarines (de Jong  et al. , 2007). Mean habitual intakeof phytosterol margarine and phytostanol margarine was14  9 g/day (1.12  0.72). Of the users, 79% ate lessthan the recommendations while 3% exceeded the intakeof 3 g/day. As the observed mean intake of phytosterols/-stanols in the free-living population appears to be lowerthan the premarket assumptions, there would be no reasonto worry about potential adverse effects. The data in thesestudies were collected before any other enriched foodsapart from margarines had been launched on the Dutchmarket. No actual data on the whole range of phytos-terol/-stanol enriched products are available. The data of the Dutch MORGEN project (1993 e 1997) had been usedto calculate the possible daily phytosterols/-stanols intakeby adults in model simulations (De Jong, Pijpers, Bleeker,& Ocke´, 2004). In the case of the replacement of allmargarines, yoghurt and cheese by products containingphytosterol esters the simulations resulted in median phy-tosterol/-stanol intake amounts of 4 e 6 g/day and the intakeabove the 90th percentile would be between the 9 and 10 g/ day. The daily intake of phytosterols/-stanols might go farbeyond the recommendations when different productswere consumed. In conclusion, intake is not widespreadamong the Dutch population, and the intake levels as mea-sured in the population do not indicate a structural subpop-ulation overdosing. Nevertheless, there are a small numberof people exceeding the recommendations. This combinedwith the increasing number of foods enriched with phytos-terols/-stanols should keep the attention to the potential of an excessive phytosterol/-stanol intake. The lack of intakedata after market launch and market penetration of multiplephytosterol/-stanol enriched foods in addition to the marga-rines leaves room for further PLM studies. Passive signalling for phytosterol-enriched margarines Up till now, the described link between the Dutch Na-tional Food Safety Authority (VWA) and the pharmocovi-gilance centre (LAREB) to monitor consumer complaints Table 1. A proposal for PLM prioritising criteriaGo Delayed go Prioritisation on product level  Type of product Innovative Non-innovativeRecently introduced  < 1 Year  > 1 YearTime span on the market 10 Years or a multiple of 10 years Less than 10 yearsNumber of products with same ingredient  > 4  < 4 Prioritisation on ingredient level  Range between RDA and UL Small range: 2 e 3 times Larger range:  > 3 times Signal detection based on effects  Adverse effects reported in humans In case of serious acute effects In case of mild effectsIn case of serious long term effectsPotential interaction (food e food, food e drug,food e supplement) a Yes, any interaction No evidence of interactionClaim level In case of health claims In case of nutrition claims Signal detection based on exposure  In case of known safety limits: anticipatedconsumption level > UL  < ULIn case of unknown safety limits High intake in small high risk group Low intake in small high risk groupWidespread consumption in total population Low to moderate consumption in total population a Physiological interaction: will be covered by adverse effects decision, behavioural interaction might for example be the replacement of drugswith foods. 539 N. de Jong et al. / Trends in Food Science & Technology 19 (2008) 535  e 545 
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks