Usage of Plant Food Supplements across Six European Countries: Findings from the PlantLIBRA Consumer Survey

Background The popularity of botanical products is on the rise in Europe, with consumers using them to complement their diets or to maintain health, and products are taken in many different forms (e.g. teas, juices, herbal medicinal products, plant food supplements (PFS)). However there is a scarcity of data on the usage of such products at European level. Objective To provide an overview of the characteristics and usage patterns of PFS consumers in six European countries. Design Data on PFS usage were collected in a cross-sectional, retrospective survey of PFS consumers using a bespoke frequency of PFS usage questionnaire. Subjects/setting A total sample of 2359 adult PFS consumers from Finland, Germany, Italy, Romania, Spain and the United Kingdom. Data analyses Descriptive analyses were conducted, with all data stratified by gender, age, and country. Absolute frequencies, percentages and 95% confidence intervals are reported. Results Overall, an estimated 18.8% of screened survey respondents used at least one PFS. Characteristics of PFS consumers included being older, well-educated, never having smoked and self-reporting health status as “good or very good”. Across countries, 491 different botanicals were identified in the PFS products used, with Ginkgo biloba (Ginkgo), Oenothera biennis (Evening primrose) and Cynara scolymus (Artichoke) being most frequently reported; the most popular dose forms were capsules and pills/tablets. Most consumers used one product and half of all users took single-botanical products. Some results varied across countries. Conclusions The PlantLIBRA consumer survey is unique in reporting on usage patterns of PFS consumers in six European countries. The survey highlights the complexity of measuring the intake of such products, particularly at pan-European level. Incorporating measures of the intake of botanicals in national dietary surveys would provide much-needed data for comprehensive risk and benefit assessments at the European level.


Introduction
Botanicals and their derivatives/preparations are used throughout Europe for health purposes, with increased usage in the general population as well as among specific subgroups encompassing children and pregnant women or those suffering from diseases such as cancer among others [1][2][3][4]. Botanicals are used in many different types of products, including foods, (teas and juices), food supplements such as plant food supplements (PFS), herbal medicinal products (HMP), homeopathic products, cosmetics, biocides etc [5]. These different product categories are regulated by specific legislation, depending on the intended use of the product.
The European Union (EU) Directive on Food Supplements (2002/46/EC) defines dietary supplements (which include PFS) as [6]: ''…foodstuffs the purpose of which is to supplement the normal diet and which are concentrated sources of nutrients or other substances with a nutritional or physiological effect, alone or in combination, marketed in dose form, namely forms such as capsules, pastilles, tablets, pills and other similar forms, sachets of powder, ampoules of liquids, drop dispensing bottles and other similar forms of liquids and powders designed to be taken in measured small quantities''.
The marketing of a product as a PFS however, depends on national legislation, which differs widely across Member States. Countries vary in the extent to which products are regulated, as well as in the process of regulatory control. Some countries have regulated the use of botanicals in detail (including negative and positive lists), some apply specific conditions of use, (including maximum usage levels or warnings for the consumer), and in others less specific requirements exist. An added complexity lies in the application of the basic European ''principle of mutual recognition'', whereby any product that is lawfully marketed in one Member State can be sold in all 27 Member States [5].
Moreover, the same botanical may be used as a food supplement and as a medicinal product, depending on the intended use of the product and both food supplements and medicinal products often share the same form of presentation (powders, pills or tablets). Hence the legal status of products differs from one country to another, resulting in a complex market environment. This so-called borderline issue between PFS and HMP is a major obstacle to the marketing of PFS in the European Union [5].
Plant food supplement usage data at EU level are scarce with reports providing PFS market data as opposed to data reported directly by the consumer [7]. Surveys on the intake of botanicals have been conducted primarily in the context of the intake of dietary supplements in general [8] or as part of surveys of complementary and alternative medicine (CAM) therapies [9], and issues such as the legal distinction between HMP and PFS have not been taken into account. A recent systematic review evaluating the demographic characteristics and health status factors associated with CAM use reported that the majority of population based consumption studies had been conducted in the USA (64% of the 110 identified studies), and of these, 13% were in Europe, with the majority carried out in Scandinavia (7%) and the United Kingdom (5%) [4]. Studies have been limited by the heterogeneity of definitions used, study designs and objectives making it difficult to compare results and to extrapolate conclusions. The ambiguity of categories such as ''natural medicine'', ''herbal remedies'' or ''herbal medicine'' and what constitutes ''dietary supplements'' makes it nearly impossible to attain reliable estimates of the prevalence of PFS usage in Europe, with only limited data available at national levels [9][10][11] but not at the European level. A study by the European Advisory Services (EAS) on ''The use of substances with nutritional or physiological effect other than vitamins and minerals in food supplements'' [7], provided information on European market and regulation data, and highlighted the need for obtaining PFS usage data in order to plan, monitor and evaluate national and European policies, as in other regions of the world. One such example is the United States of America, where the Alternative Health/CAM supplement of the National Health Interview Survey (NHIS) has been collecting data on botanical dietary supplements for some years now [12][13][14].
The European Food Safety Authority (EFSA) has recognised the lack of data in the sector and has published a number of reports addressing related issues, namely the recommendations for reporting the use of supplements and medicines by adults in any pan-European dietary survey or project [15], and the ''Compendium of botanicals reported to contain naturally occurring substances of possible concern for human health'', aimed to help with the safety assessment of botanicals and botanical preparations intended for use as food supplements [16].
The purpose of this paper is to describe the type and frequency of PFS usage reported in a retrospective survey of consumers in six European countries; in addition we present the most frequently used botanical ingredients in these products. We also highlight the issues associated with measuring usage of PFS in European populations and make recommendations for future research.

Ethics statement
Before initiating the fieldwork, approval for the conduct of the survey was obtained from four ethics committees: the Bioethics Commission of the University of Barcelona, Spain; the Ethics Committee of the University of Milano, Italy; the Ethical Committee of the Faculty of Medicine -Transilvania University of Brasov, Romania; and the Coordinating Ethics Committee, Hospital District of Helsinki and Uusimaa, Finland. Approval of the survey by these four ethics committees required submitting all survey material to their members for evaluation. No ethical approval for the survey was needed in Germany and the United Kingdom.
To ensure harmonisation and standardisation of the fieldwork and data collection across countries, a market research organization, European Fieldwork Group (EFG) was subcontracted to implement the survey. The survey was conducted by EFG in strict accordance with the ICC/ESOMAR Code on Market and Social Research. In all countries, informed consent was obtained verbally from all respondents after reading the survey information sheet. All data were recorded manually i.e. pen-and-paper. Recruitment of survey participants occurred in the selected cities in each country. Approximately the first 1000 individuals per country were systematically selected for screening i.e. intercepting 1 in every 5 individuals passing by to ask him/her the initial screening questions; subsequent screening selection was performed on a convenience basis i.e. intercepting individuals in places where consumers were likely to be found, such as herbal shops, pharmacies etc. Eligible respondents who agreed to participate were given an appointment at their home/workplace to complete Table 2. Distribution of screened individuals, PFS consumers interviewed and prevalence sample by country and gender.       the main survey. The appointments of those willing to participate were later reconfirmed by phone. The data were made anonymous when recorded electronically i.e. the respondents' contact details were not entered into the survey database. Instead, the market research organization assigned ID numbers to each respondent and provided PlantLI-BRA partners only the database with the assigned ID numbers.

Definition of plant food supplements in the PlantLIBRA PFS consumer survey
Although there is a legal definition of Food Supplements (EU Directive (2002/46/EC) [6] under which PFS reside, for the purposes of this research it was necessary to develop a specific definition of PFS whose main characteristic is that they contain botanical preparations as ingredients for food supplementation.
Botanical preparations are obtained by subjecting botanicals (plants, algae, fungi or lichens) to treatments such as comminution, extraction, distillation, squeezing, fractionation, purification, concentration or fermentation. These include extracts, essential oils, expressed juices, powders, etc.
Botanical preparations can be considered as nutrients or other substances. Thus, the definition of PFS for the survey was as follows: PFS are "foodstuffs the purpose of which is to supplement the normal diet and which are concentrated sources of botanical preparations that have nutritional or physiological effect, alone or in combination with vitamins, minerals and other substances which are not plant-based. PFS are marketed in dose form, such as capsules, pastilles, tablets, pills and other similar forms, sachets of powder, ampoules of liquids, drop dispensing bottles, and other similar forms of liquids and powders designed to be taken in measured small unit quantities''.
Products that did not meet this definition, such as herbal remedies and other medicinal products based on botanicals, and those that did not meet the PFS definition in terms of dosage, such as herbal teas or juices, were excluded.

Sample population and PFS consumer definition
A cross-sectional, 12-month retrospective survey was conducted in 24 cities in six European countries -Finland, Germany, Italy, Romania, Spain and the United Kingdom. An estimated sample size of 2000 screened individuals per country was calculated in order to obtain a final sample of approximately 400 consumers per country (total N = 2400 approximately). Per country, gender and age group quotas were set as follows: 300 adults (18 to 59 years) and 100 older adults (60-and-over years), with 30-50% male and 50-70% female. All individuals were screened by means of a brief questionnaire which recorded PFS usage in the preceding 12 months. Individuals were considered eligible for inclusion if they were over 18 years old and met either of the following specified criteria, intended to capture the different usage patterns of PFS consumers: 1) They had taken at least 1 PFS in the last 12 months, in an appropriate dose form at a minimum frequency of either: a) 1 daily dose for at least 2 consecutive or non-consecutive weeks, or b) 1 or more doses per week for at least 3 consecutive weeks or c) 1 or more doses per week for at least 4 consecutive or non-consecutive weeks 2) They had taken 2 or more different PFS, in an appropriate dose form, at a minimum frequency of 1 or more doses per Table 5. PlantLIBRA's PFS consumer survey -PFS usage patterns, per product used by a respondent, overall and by gender and age group. week, with the sum of the usage period of the 2 or more products being equal to at least 4 weeks.

Instruments and variables
A short screening questionnaire was used to identify consumers who met the survey inclusion criteria; it consisted of six questions which allowed interviewers to identify eligible consumers, based on the product(s) used, the frequency and duration of use and the dose form. Eligible consumers subsequently completed a more detailed questionnaire on their PFS usage in the preceding 12 months, providing details of product/plant names, dosage forms, frequency of use, reasons for use, adverse effects, places and patterns of purchase and information sources on products. These questions were asked for each of up to a maximum of 5 different PFS used. In addition, respondents were asked to provide sociodemographic data including age, gender, level of education and employment status, as well as self-reported height and weight and further health-related lifestyle information.

Survey administration and data collection
Fieldwork and data collection for the cross-sectional survey were conducted by the international market research company EFG, from May 2011 to September 2012. The duration of the fieldwork ensured that any seasonal variability in usage of products was captured. The survey protocols and instruments -training material, information sheet, informed consent, screening and usage questionnaires-, were initially developed in English by consensus amongst the research team, and subsequently translated into the respective languages in each of the survey countries. Pilot interviews were conducted in each participating country to assess the comprehension of the questions and to determine the time required to complete the survey.
In each participating country, trained interviewers systematically screened approximately 1000 individuals during the first three months of the survey, which allowed the estimation of the prevalence rate. Subsequently, screening and recruitment were conducted on a convenience basis. The recruited eligible consumers were interviewed face-to-face and the more detailed PFS usage questionnaire completed.

Data preparation and statistical analysis
All data from the completed surveys were entered into the statistical package SPSS for Windows v. 18 (IBM Corporation, Somers, NY, USA), which was also used for data analysis.
Following review of the completed interviews by the research team in each country, a database with botanical composition data for all PFS products reported was compiled for each country and then merged into a single database. Potential product duplicates between countries were not removed. Each product was coded for its botanical ingredients in scientific, English and local names and botanicals were coded after removing duplicates between countries. Additionally, each product was categorised as a single-or multi-botanical product. To indicate the certainty of the matching of products, a series of numerical codes were used, based on those used in the National Health and Nutrition Examination Survey 2005-2006 [17]. Values ranged from 1-5, where ''1'' indicated an exact match, ''2'' a probable match, ''3'' a reasonable match, ''4'' a default match and ''5'' no match. Only products with certainty values 1 to 4 have been included in the analyses.
Respondent data were recorded in a separate database. A number of variables were created and/or recoded to facilitate reporting and analysis, including: 1) ''education level'', defined as low, medium, and high; 2) ''BMI'', which was calculated from self-reported weight Table 6. PlantLIBRA's PFS consumer survey -PFS usage patterns, per product used by a respondent, overall and by country. and height, and for which WHO criteria [18] were used to categorise individuals as underweight (BMI,18.5 kg/m 2 ), normal weight (BMI 18.5-,25 kg/m 2 ), overweight (BMI 25-,30 kg/m 2) and obese (BMI $30 kg/m 2 ); 3) ''physical activity'', calculated using the short version of the IPAQ [19] and defined as low, moderate or high. Absolute frequencies and percentages for each of the variable categories were used to describe the qualitative nominal/ordinal and discrete quantitative survey data. In turn, all data have been stratified by gender, age range and country -also using absolute frequencies and percentages and 95% confidence intervals. When describing the association between two qualitative variables (nominal or ordinal), contingency tables were used. The continuous quantitative variables (e.g. BMI, alcohol) were recoded into categorical variables.
It is important to note that when reporting the main results of the survey, the unit of analysis varies depending on the variables used, i.e. for certain variables the unit is an individual respondent, however, given the potential intake of multiple supplements by one respondent, the unit of analysis may change to the supplement level. Furthermore, all results presented in the tables represent the analysis of raw data as opposed to data weighted by the population size. Data were not weighted because of the study methodology selected, whereby all country samples were very similar in size and included only PFS consumers.

Validation study
In order to validate the PFS usage questionnaire, a validation study was conducted in which the data collected using the survey instrument were compared with a 30 to 180-day diary (used as the gold standard). The study was conducted in two of the PlantLIBRA consumer survey cities: Las Palmas de Gran Canaria (Spain) and Milan (Italy), where 48 and 49 consumers respectively were recruited using convenience sampling. The PFS usage questionnaire was completed by the respondents at the beginning and at the end of the 6-month period of the validation; during this time the consumers also completed the usage diary. Data from the last questionnaire and the diary were compared for concordance, and results are shown in Table 1, indicating a good agreement for product consumed, dose form and doses per day.

Characteristics of the PFS consumer sample
A final sample of 2359 consumers (those eligible and willing to participate) was recruited from 11783 screened individuals (Table 2). Due to different legal frameworks (different distribution of botanicals in food supplements and medicinal products), more individuals had to be screened in Finland in order to recruit the required 400 consumers. Survey respondents were recruited to fixed quotas for age and gender, which were achieved, with some differences within countries (Table 3). In Finland the proportion of adults aged 50-59 years was significantly higher (26.2%), whilst the opposite was true in Italy, where consumers in that age group constituted only 13.0% of adults. Romania had a significantly higher number of consumers in the youngest age group (30.5%), in contrast to Spain and the United Kingdom, where this age group represented only 9.5% and 9.0% of adult consumers, respectively. A significantly higher proportion of female consumers were recruited in Spain (56.7%) and in the United Kingdom marginally more males were recruited (50.3%). Across all countries, more than half of the participants (57.5%) were employed (Table 3), with the percentages slightly lower in Finland (50.9%) and in the United Kingdom (52.4%). The majority of participating consumers were educated to medium level (Table 3).
Respondents were asked a number of questions regarding health-related lifestyle factors (Table 4). Less than half of the consumers had never smoked (46.6%), less than one quarter were ex-smokers (23.1%) and less than one third were current smokers (30.3%).
More than half of the total respondents (59.3%) had not consumed alcohol or had consumed it less than once daily; more than a tenth (12.6%) reported daily alcohol consumption.
The proportion of overweight and obese people in the survey was 49.8% (Table 4). Some significant differences in levels of physical activity were noted between countries. High levels of activity were reported by 85.5% of Romanian respondents compared to a value of 42.9% across all countries.
Most of the respondents (65.1%) reported not being regular consumers of food supplements other than PFS in the preceding 12 months, except for Finland ( Table 4). The proportion of nonconsumers varied from 20.7% in Finland to more than 80% in the United Kingdom and Italy. By contrast, in Finland 76.3% of the individuals were regular consumers of food supplements.
Over half of all respondents (59.5%) reported not having used CAM therapies/treatments in the past year. This is particularly the case in Italy (74.6%), Romania (80.8%) and the United Kingdom (92.6%).
Three quarters of consumers reported their health status as very good or good (75.5%), while 3.6% reported it as bad or very bad and 21.0% as neither bad nor good (Table 4).
Between countries, more consumers reported their health status as very good or good in Romania (81.3%) and in the United Kingdom (81.1%) than in other countries; though conversely the highest proportion reporting their health status as bad or very bad was also in the United Kingdom (7.6%).

PFS usage patterns
Overall, products are most often taken ''periodically'' (37.3%) with respondents also reporting using PFS when experiencing a ''flare up or worsening of a condition'' (22.2%) ( Table 5). Products are also used on a more ''sporadic basis'' (19.8%) and on ''other non-specified occasions'' (17.8%). Both men and women reported taking products on a periodic basis (39.3%, 35.6%) and this was also true for both age groups (Table 5). Periodic use was reported significantly more often in Finland (46.2%), Germany (50.7%), Italy (41.3%) and Romania (41.8%), but in Spain, ''another reason'' was most reported (46.0%) and in the United Kingdom, sporadic use (34.8%) was significantly higher than any other reason as to when products were used (Table 6).

PFS products used
Respondents reported a total of 1288 products across the six countries. At individual country level, the highest numbers of different PFS were used in Italy (289) and Spain (284); in the United Kingdom, the number of different PFS was approximately half that of the other countries ( Table 7). The number of different botanical ingredients was 491, with the maximum number of different botanicals contained in a single product being 46 and present in a German product. The United Kingdom differed from the other countries as the products reported contained a lower number of botanical ingredients (maximum 8).
In terms of the number of products used, 83.7% of all consumers reported taking one product in the preceding 12 months, with 12.3% taking two products and 4.0% using more than two products (Table 8). Generally this pattern was similar for both men and women and across the age groups, although those over 60 did report a significantly higher use of two or more products than those under 60 (19.5% vs. 15.2%) ( Table 8). At country level (Table 9), some significant differences were noted: in Finland, the percentage of consumers using two or more products was significantly higher than in all other countries (40.2%).
Overall 51.5% of consumers used a single-botanical product and 32.3% used one multi-botanical product (Table 8). There were no significant differences between males and females in this usage pattern, but consumers aged over 60 used less multibotanical products than those aged 18-59 (27.7% and 33.8% respectively) (Table 8). Overall, fewer consumers reported using two or more single-botanical products (4.4%) and two or more single-and multi-botanical products (11.9%) ( Table 8).
There were some significant differences across countries in the type of products consumed (Table 9). In the six countries, the values for single-botanical products range from 84.5% (the United Kingdom) to 20.5% (Finland). Usage of multi-botanical products was reported in all countries, with the lowest proportion (7.1%) reported in the United Kingdom (Table 9). The use of two or more single-botanical products was low in all countries as was the usage of two or more single-and multi-botanical products. Finland was an exception to the latter, with 38.2% of respondents taking multiple products ( Table 9).
The most common dose forms used (Table 10) are capsules (38.3%) and pills/tablets/lozenges (36.8%). No significant difference was observed in relation to gender or age (Table 10). Across the six countries (Table 11), solid forms are generally most popular, although capsules were used less frequently in Romania (17.7%). Liquid forms were less common in the United Kingdom (8.2%) and Germany (9.9%), but more common in Finland (26.2%) and Italy (26.4%) ( Table 11). Table 8. PlantLIBRA's PFS consumer survey -number and type of products taken, overall distribution and by gender and age group.

Botanicals used
A total of 491 botanicals -used in at least one PFS-were reported across the six participating countries. An overview of all the reported botanicals -clustered by intervals of frequency of intake (number of consumers ranging from 194 to 5)-is shown in Table 12. Based on the survey results, the eleven most frequently used botanicals (numbers of consumers ranging from 194 to 100) in descending order are Ginkgo biloba (ginkgo), Oenothera biennis (evening primrose), Cynara scolymus (artichoke), Panax ginseng (ginseng), Aloe vera (aloe), Foeniculum vulgare (fennel), Valeriana officinalis (valerian), Glycine max (soybean), Melissa officinalis (lemon balm), Echinacea purpurea (echinacea) and Vaccinium myrtillus (blueberry) ( Table 12). Table 13 shows the overall unweighted ranking of botanicals, 1-40, according to the number of consumers, in decreasing order. Table 13 also shows that when unweighted overall data are stratified by gender, only slight differences between men and women become evident and only Glycine max (soybean) was used significantly more by women than by men (Table 13).
When the overall top-40 botanical data are stratified by age groups, slight differences become evident. In the group of 18-59 year-olds, the most frequently used botanicals comply with the overall data just differing in the ranking, with Oenothera biennis (evening primrose) being the most frequently used botanical (Table 13). In the group of 60+ year-old a stronger shift can be observed (Table 13). Although Ginkgo biloba (ginkgo) is still the most reported botanical -as in the overall ranking-other botanicals are frequently used by that age group. Harpagophytum procumbens (devil's claw), Vaccinium myrtillus (blueberry) and Allium sativum (garlic) are within the most frequently reported botanicals, whereas Glycine max (soybean), Melissa officinalis (lemon balm) and Echinacea purpurea (echinacea) do not appear in the top 10 ranking.
Cross-country differences emerge when considering the overall top-40 botanicals more frequently present in PFS products in each of the individual six countries (Table 14). In the Finnish sample, products containing Glycine max (soybean) are the most frequently used, followed by those containing Echinacea angustifolia and purpurea (echinacea). German consumers reported Ginkgo biloba (ginkgo), Cynara scolymus (artichoke) and Olea europea (olive) as the most frequently used botanicals; whilst in Romania, Ginkgo biloba (ginkgo) was also the ingredient most frequently indicated, followed by Aloe vera (aloe) and Panax ginseng (ginseng). Amongst Italian consumers, Aloe vera (aloe) was the most frequently used botanical, followed by Foeniculum vulgare (fennel) and Valeriana officinalis (valerian). In Spain, PFS containing Cynara scolymus (artichoke) were the most frequently used products, followed by those containing Valeriana officinalis (valerian) and Equisetum arvense (horsetail). In the United Kingdom, Oenothera biennis (evening primrose) was by far the most frequently reported botanical ingredient, followed by Panax ginseng (ginseng) and Hypericum perforatum (St. John's wort). In addition, there is a great variation in the ranking of consumed botanicals among countries.

Discussion
The present paper reports the findings from a European multicountry survey of PFS consumers: the PlantLIBRA PFS consumer survey. Data on the usage of PFS at the European level are limited, confined in the main to commercial market data [7] as opposed to consumer survey data, as evidenced in the recent review by Bishop and Lewith (2010) [4], where only 13% of population based consumption studies were in Europe. The European Food Safety Authority (EFSA) has recognised the lack of Table 9. PlantLIBRA's PFS consumer survey -number and type of products taken, by country.  Usage of Plant Food Supplements by European Adults PLOS ONE | www.plosone.org data in the sector and has published a number of reports addressing related issues [15][16].
To our knowledge this is the first survey of consumers of PFS undertaken in Europe. In total 2359 consumers of PFS were recruited in this cross-sectional retrospective survey. Across all countries prevalence of usage is estimated at 18.8%. Vargas-Murga and colleagues (2011) [9] highlighted that comparable data at European level is difficult to identify when reviewing prevalence data from a selected number of European studies, evaluating PFS or CAM usage, with values ranging from 0.8% to 70%. All studies were based on nationally representative samples but the definition of use of supplements varied widely, in some cases being selfdefined by the participant and not distinguishing between PFS and HMP. The use of dietary supplements in a European population was measured in the European Prospective Investigation into Cancer and Nutrition (EPIC) study [8]. Usage was measured by completion of a standardised 24-hour dietary recall and included all dietary supplements that met the EU Directive 2002/46/EC. Results indicated significant differences in overall dietary supplement use between countries with herbs/plant-based supplements representing 8-17% of the products used across the ten countries.
The prevalence rate reported here can be compared to rates from surveys conducted in the United States, where data on usage of dietary supplements, including herbal supplements, is collected more routinely. It is similar to the rate reported in the 2002 and 2007 National Health Interview Surveys (NHIS), 18.9% and 17.9% respectively [20]; higher than the rates of both the Eisenberg's survey [21] and the Slone survey [22], with 14% and 12.1% respectively; and lower than the 2002 Health and Diet Survey (42%) [23] or the 1999 Kaiser Permanent Medical Care Program of Northern California (KPMCP), with a prevalence of 28.3% [24]. These differences in prevalence across studies may in part be due to the distinct selected population samples, survey methodologies (i.e. sampling methods, data collection techniques) or definitions of usage, as well as possible variations in health beliefs and health behaviour of the different populations of study [9], [24].
Survey respondents were recruited to set quotas for both age and gender to reflect characteristics previously reported for dietary supplement users. Age and gender are significant determinants of the consumption of dietary supplements in general and in botanical products in particular. Previous studies on the use of dietary supplements or other herbal-related use show a higher consumption among women as compared to men [1], [17], [24][25][26][27][28] and a higher consumption among older adults as compared to younger adults [24], [29][30][31][32].  Table 11. PlantLIBRA's PFS consumer survey -PFS dose forms, per product used by a respondent, by country.  Other characteristics of dietary supplements users that have been reported previously in the literature include having higher educational attainment and socioeconomic status [24], [33][34], being less likely to smoke [10], [32], [35], being more physically active [10], [29], [32]. Bailey et al. also reported a moderate alcohol consumption (1 drink per day) among dietary supplement users as compared to nonusers. In contrast, a study by Rovira et al. in a southern European population found no differences in lifestyle factors such as physical activity, smoking, and alcohol consumption between dietary supplement users and non-users [36]. Our survey population consists exclusively of PFS consumers, but their responses to a series of questions on health-related lifestyle factors reflect some of the characteristics mentioned above. The majority of PFS consumers perceived their health status to be ''very good or good'', reflecting results reported in a number of studies on dietary supplement users [32] and CAM and dietary supplement users [24], where the answer ''very good or excellent'' has been reported for self-reported health status.
The survey results indicate that most consumers reported using one PFS product in the preceding 12 months, with 12% using two products and 4% using more than two. Individual country data show that Finnish consumers use more than one product and PFS with more than one botanical component, and the opposite is observed in the United Kingdom, where about 90% of the consumers use only one PFS and the products contain mostly only one botanical. In the United States, recent studies have reported that about half of the adults report using one or more dietary supplements [32], [37]. One of these studies also found that over half of dietary supplement consumers used a single-botanical product and one third used one multi-botanical product [32]. Similar results were found in our survey across all countries i.e. smaller numbers of consumers reported using two or more singlebotanical products (4.4%) and two or more single-and multibotanical products (11.9%).
A wide variety of botanicals (491) is used in PFS consumed by the respondents in this survey. Overall raw data show that the most frequently (n.100) used botanicals in descending order are Ginkgo biloba (ginkgo), Oenothera biennis (evening primrose), Cynara scolymus (artichoke), Panax ginseng (ginseng), Aloe vera, Foeniculum vulgare (fennel), Valeriana officinalis (valeriana), Glycine max (soybean), Melissa officinalis (lemon balm), Echinacea purpurea (echinacea) and Vaccinium myrtillus (blueberry). These results reflect some commercial data which reported that ginkgo followed by echinacea, garlic and ginseng were the four most commercially important botanicals in the combined markets of seventeen EC Member States. In this data, echinacea and ginkgo were part of the composition of products registered as medicines [7], [9], which were excluded from our survey. Similarly, the US Food and Drug Administration 2002 Health and Diet Survey, also a 12-month retrospective study, reported the same four herbs/botanicals/or other nonvitaminnonmineral dietary supplements being the most used by its adult population -although in the following order: echinacea, garlic, ginkgo and ginseng (the latter including tea) [23]. Schaffer et al. also reported echinacea as the most consumed botanical in the Californian 1999 KPMCP survey, followed by ginkgo [24]. Differences between countries are more evident; the top list of botanicals contained in PFS for each single country complies little with the ranking of the overall data. As mentioned earlier, data were not weighted by country population size because of the study methodology which included very similar country-sample sizes of PFS consumers only, therefore caution is needed when drawing conclusions from these results at the overall 6-country level. Overall data merely describes the collected pooled data from all 6 countries. However, if the overall ranking data were to be weighted by the population size -for example the 1-5 ranking data-, the positions of the botanicals would have been only slightly altered, with Oenothera biennis (evening primrose) being the most consumed one, followed by Cynara scolymus (artichoke) Ginkgo biloba (ginkgo), Panax ginseng (ginseng) and Aloe vera (aloe). The results of the survey highlight clear differences between countries in terms of the botanicals used by consumers as PFS.   This may reflect the fact that the current legal and regulatory framework for botanicals has a major influence on the nature of the local PFS markets. The EU Directive 2002/46/EC does not provide a clear definition of what is encompassed by the term 'other substance with a nutritional or physiological effect', although it is generally accepted that botanicals and their extracts fall into this category. Current legislation varies across Europe, with significant differences in the botanical species permitted in PFS. These issues were highlighted in a recent review of the regulations applicable to PFS in the European Union by Silano et al. [38]. They provide examples of the different national approaches for the use of selected botanicals in food supplements in the EU Member States. To illustrate the above complexity, in Germany, food supplements are regulated by the German Regulation on Food Supplements [39] and the German Law on Food and Feed [40]. Positive lists are available for minerals and vitamins. Food supplements have to be registered with the Federal Office of Consumer Protection and Food Safety [41]. The BVL maintains a list of plants which are either classified as a food or a medicinal product, and which is neither considered complete nor legally binding [41]. Data on the intake of PFS in Germany is limited and, despite food supplement intake being recorded in recent health and nutrition surveys [42][43][44], no specific data was published on PFS intake. The results from the PlantLIBRA consumer survey do not include Valeriana officinalis in the German top list of botanicals used in PFS, whereas 1852 medicinal products containing Valerian exist on the market [40]. The absence of Valeriana officinalis in the German list of botanicals can be explained by its dominant presence as a HMP in the German market.
The results of this survey represent some of the first data on the usage of PFS at European level, thus addressing the existing deficit of such data by collecting retrospective data directly from consumers in six European countries. The benefits of the data collection instrument used in this study included that it was relatively straightforward to administer, did not alter habitual usage patterns and allowed the classification of individuals into categories of usage. However, the results must be considered in the light of their limitations. The sample population comprises exclusively of PFS consumers, recruited to meet very specific inclusion criteria and hence no comparisons can be made with the general population. Future studies should seek to compare users and non-users of PFS.
Further limitations relate to the retrospective nature of the data being collected. In many cases respondents needed to rely on memory to report usage of products in the preceding 12 months. Where products are available for inspection at data collection, there is a need for careful recording of product details to ensure accurate coding. The lack of a comprehensive product database containing reliable ingredient information meant a bespoke database needed to be created. Future studies should seek to collect prospective data. Prospective dietary intake surveys offer an ideal opportunity to collect data on supplement use in conjunction with data on food and beverages. Care needs to be taken to collect sufficiently detailed information about ingredients and amounts consumed. For example, in the US, the Alternative Health/CAM supplement of the National Health Interview Survey (NHIS) is part of an annual, nationally representative survey of US adults. It contains data on adults' use of 10 herbs most commonly taken to treat a specific health condition in the preceding 12 months [13]; the survey has a separate section on dietary supplements and distinguishes ''natural herbs'' from vitamins and minerals. The authors would like to encourage researchers to implement future Table 14. Cont.
surveys/studies which are necessary to overcome the bottlenecks in PFS risk and benefit assessments at the European level.