Contaminants in food supplements and associated health risks

Summary Through the increasing use and availability of food supplements on the market, safety Aconcerns relating to the safety of these food supplements are growing as well. The aim of the present PhD thesis was to investigate the presence and actual levels of contaminants of concern in selected food supplements on the Dutch market and to estimate the associated health risks. First, in chapter 1, an overview is provided on the food supplements selected for the studies, which are clay products for oral use, herbal food supplements used to enhance sexual potency and herbal food supplements used for weight loss. Furthermore, an overview is given on the Dutch en European legal provisions for food supplements. In the first study described in chapter 2 of this theses data are presented on the occurrence of metals such as lead, mercury and cadmium and the metalloid arsenic in clay products which are used via the oral route by pregnant and lactating women. For lead, the use of 34 of the 36 traditional clays and two of the 27 health clays would result in intake levels exceeding the selected health based guidance values, by up to 20-fold. In the case of inorganic arsenic, the use of 15 of the 35 traditional clays and 11 of the 27 health clays would result in intake levels exceeding the health based guidance values by up to 19-fold. The second study, described in chapter 3 of the thesis, reports data on the presence of dioxins in 33 clay products, which were collected on the Dutch market and in some African countries. Polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans (PCDD/F) were detected in clay products from the Dutch market, in concentrations ranging from 66 to 103 pg TEQ g-1, whereas PCDD/F concentrations in the suspected clay products from African countries varied from 24 to 75 pg TEQ g-1. Furthermore, in this study congener patterns in African clay products were compared with those of pooled human milk samples collected by WHO in eight African countries, to investigate a possible relation between PCDD/Fs in human milk with contaminated clay used for consumption. From the similarity between the patterns in clays and the human milk samples from the Democratic Republic of The Congo and Côte d’Ivoire, it can be concluded that there is probably a relationship with the consumption of contaminated clay. The aim of the third study, described in chapter 4 of this thesis, was to determine whether herbal food supplements on the Dutch market contain active pharmacological ingredients (APIs) known to inhibit phosphodiesterase type 5 (PDE-5), such as sildenafil and other known analogous PDE-5 inhibitors. Therefore, herbal food supplements intended to enhance sexual potency (n=71), and two soft drinks, were analysed. In 23 herbal supplements, nine different PDE-5 inhibitors were identified, in a few cases (n=3) more than one. The presence of these APIs was however not stated on the label. Subsequently, it was estimated whether intake of the supplements with the detected PDE-5 inhibitors could result in pharmacological effects. It was concluded that 18 of the 23 herbal food supplements with PDE-5 inhibitors, when used as recommended, would have pharmacological effects due to the added APIs. In the fourth study, described chapter 5 of in this thesis, another group of herbal food supplements, claiming to reduce weight, was investigated for the presence of APIs that can be used for the treatment of overweight and obesity. To this end, 30 herbal food supplements for weight loss on the Dutch market were collected and analysed for the presence of APIs with weight loss properties. In 24 samples the APIs sibutramine, desmethylsibutramine (DMS), didesmethylsibutramine (DDMS), rimonabant, sildenafil, and/or the laxative phenolphthalein were identified 41 times. The potential pharmacological effects of the detected APIs were estimated, and use of 20 of the 24 herbal food supplements, may result in potential pharmacological effects. Furthermore, a risk assessment of phenolphthalein regarding its carcinogenic effects, resulted in Margin of Exposure (MOE) values of 96-30,000. MOE values lower than the minimum required 10,000 (96-220) were calculated for the daily intake levels of four out of the ten supplements in which phenolphthalein was found. However, taking into account that weight loss preparations may be used for only a few weeks or months rather than during lifetime, MOE values may be two to three orders of magnitude higher. This study shows that the use of food supplements with sibutramine, DMS, DDMS, and/or phenolphthalein could result in both pharmacological but also other health effects. From the studies described in this thesis it can be concluded (chapter 6) that, in addition to concerns over naturally occurring endogenous toxins present in herbal supplements, the presence of exogenous contaminants in herbal supplements can pose a health concern. Furthermore, the results of the present thesis also lead to the conclusion that in order to refine the risk assessment on the presence of contaminants such as metals, metalloids, dioxins and APIs in (herbal) food supplements more precise data are required on bioaccessibility of contaminants of concern from the food matrix. Also information on groups at increased risk may need to be increased. Additionally, the presence of APIs in herbal supplements, which may originate from drop outs in the drug development process, is of concern. In order to screen for unknown APIs the use of effect-based bioassays should be considered more often, as they have been shown to be successful in detecting unexpected and as yet unknown active ingredients. From the results obtained for the food supplements included in our studies it can be concluded that consumers should be aware that food supplements may not be without risks. The overall conclusion from the work described in this thesis is that for food supplements ‘natural’ does not equal ‘safe’.