SWEET is reviewing evidence of long-term benefits and potential risks involved in switching to S&SEs in the context of public health and safety, obesity, and sustainability. Prospective associations – food behaviours in the past with health status now – between consumption of S&SEs and health are being explored using data collected from six cohort studies in four countries. But “what do we know about the content of S&SEs in different foods?” As is so often the case with research, the answer is, “not as much as we really need to know”. Mark Roe, Senior Food & Nutrition Data Scientist, EuroFIR AISBL, Brussels (BE) explains in this informative blog.
Food consumption and composition data – do we know what sweeteners are in food?
Mark Roe, Senior Food & Nutrition Data Scientist, EuroFIR AISBL, Brussels (BE)
In a previous article (https://sweetproject.eu/dietary-guidelines-and-informed-personal-choice-sugar-sweeteners-and-complexity/), my EuroFIR colleague, Siân Astley, wrote about the need for individuals to reduce free sugar intakes and challenges for the food industry associated with reducing free sugar content in foods. She also pointed out that our understanding of the impact of sweeteners and sweetness enhancers (S&SEs) on health is limited largely to reduced intakes of free sugars.
SWEET is reviewing evidence of long-term benefits and potential risks involved in switching to S&SEs in the context of public health and safety, obesity, and sustainability. Prospective associations – food behaviours in the past with health status now – between consumption of S&SEs and health are being explored using data collected from six cohort studies in four countries. Each of the studies collected information about foods consumed using either food frequency questionnaires (FFQs) or food diaries. FFQs assess habitual diet by asking how often which food items or food groups are consumed over a given period whilst food diaries record what was eaten over a defined period. Quantities of each food or food group consumed by participants is combined with information describing the nutrient composition of these foods (i.e., carbohydrates, fats, proteins, vitamins, and minerals) and information related to participants’ health of participants. Subsequent analysis should reveal any associations between dietary components and health.
All this sounds very straight forward until you ask the question, “what do we know about the content of S&SEs in different foods?” As is so often the case with research, the answer is, “not as much as we really need to know”, especially when we are trying to understand their impact on health. Ideally, we would like to know exactly what S&SEs are used, and in what quantities, in foods. Even if we were only interested in total sugar or total carbohydrate content, which is not the same thing, such information is not fully available and not in a standardised form.
EuroFIR AISBL (www.eurofir.org) provides high quality, validated national food composition data and supporting information; values have been checked for quality, standardised, and are as a result comparable. FoodEXplorer (https://www.eurofir.org/foodexplorer) allows simultaneous searching of nutrient composition data from 39 national datasets and the results can be downloaded for offline analysis. The datasets can also be licenced for use in third party products and accessed in real time using APIs (application programme interfaces). Whilst the data behind the foods and nutrient values have been harmonised, and value descriptions utilise specific vocabularies, there are limitations to national food composition datasets and some are particular to S&SEs.
Most national food composition datasets provide data for generic foods rather than branded products. Information about branded foods, including nutrient content and ingredients as shown on product labels, is available from retailers’ and manufacturers’ websites and other sources, but most are not publicly available in a single place or, if they are, access is costly. Combining food composition data with FFQ and food diaries can be challenging and time consuming and is particularly difficult when data are limited in availability and specificity, which is often the case for S&SEs. Lisa Heggie, Clare Llewellyn and colleagues from University College, London (UCL, UK) are dealing with just that problem in relation to research on the impacts of free sugar and low-calorie sweetener intakes on healthy growth, appetite, and taste preferences in childhood.
Consumption data from Gemini study (https://www.geministudy.co.uk/) food diaries are being combined with S&SE contents for foods consumed by twin children. Data for individual sugars can be drawn from the national UK food composition dataset (otherwise known as McCance & Widdowson) but, in many cases, the contents of S&SEs need to be estimated based on product ingredients. Many foods include S&SEs in the list of ingredients, but it is more common for mixtures of sweeteners to be declared (e.g., Acesulfame K & Sucralose, Aspartame & Acesulfame K) and the quantity of individual sweetener(s) is not known. Currently, Lisa is undertaking the laborious process of compiling these data and is working towards a possible method for estimating sweetener quantities in diets, with advice from Corey Scott (Cargill). If anyone has information about the amounts of S&SEs used in individual foods, it would be helpful if this could be shared with UCL ([email protected]).
Interest in S&SEs is a relatively new research field, saccharin was first used in a beverage in the US in 1952, so it is hardly surprising that information is either difficult to find or unavailable. However, similar problems also apply to other nutrients (e.g., minerals, vitamins, individual fatty acids) and new ingredients. Why is this the case, when there is considerable interest in food, nutrition, and health, is unclear. Research projects often work in isolation and there has been limited re-use of data harmonisation approaches and handling methods, despite widespread recognition of the need for these tools. Although projects have attempted to address these problems in the past, their focus has often been on linking food and health data (i.e., web-based portals or smartphone and tablet apps to collect food intake information and prompt lifestyle changes). While these solutions have been innovative, and data linking and sharing has become relatively straight forward, there has been much less interest in data quality.
Food composition, food consumption, and classical nutrition research have long been unfashionable, despite underpinning human health, and these areas are significantly under-resourced compared with others. There is a continual need to chase funding and current trends are focused on areas such as the microbiome and genome. The number of researchers with sufficient expertise to address food and nutrition data problems seems to be declining and, as with other scientific fields, increased competition for funding and short-term employment contracts are not helping! Consequently, there is a need to make the most of limited resources and increase sharing of data and expertise.
EuroFIR is a partner in a wide range of research projects related to food and nutrition, including SWEET, and collaboration and dissemination of results as well as training are core components of our work; we help ensure food data generated by any project are FAIR (findable, accessible, interoperable, and re-usable). In addition, projects like FNS-Cloud (https://www.fns-cloud.eu/) and Blue-Cloud (https://www.blue-cloud.org) are starting to address the problem of fragmented resources, which lack critical mass and are not easily accessible, by launching cloud solutions, federating existing and emerging datasets and developing new services to support re-use by researchers. That said, if the information is not available, as with S&SEs in foods, we risk missing important evidence for understanding the impact of different substances in our diet on human healthy across our lifetimes.