Food Composition Analysis and its Implications for Dietary Planning
Whatever your goals as a body builder or keen trainer, be they cutting to drop body fat for a competition or your summer holiday or bulking to gain mass, the macronutrient composition of the diet is frequently discussed and calculated. Many will cut their carbohydrate intake to maybe 100g a day or less while cutting, maintaining a relatively high level of protein. We hear people talking about their protein needs: "I eat 1.5g protein per pound of body weight a day". Cutting involves creating a calorie deficit, often 500kcal a day less than is typically eaten.
These numbers have been discussed at length for years and will continue to be debated and keep the fitness magazines and diet books flying from the shelves. However, are the figures for macronutrient composition that we derive from food labels or food composition tables to be trusted? Many calculate their daily intake to a precise degree but if the figures used are not accurate, then you may be wasting your time. In this article, I will explain how food composition databases are derived and how the information on food labels is produced to put in perspective the dietary fine tuning that many body builders feel compelled to carry out.
Food composition tables
Nutritional calculators are software programmes, such as Nutricalc, Compeat and so on. They contain databases of many foods such as ingredients, recipes and manufactured items, along with their nutrient breakdowns. These databases are composed of generic data sets, such as McCance and Widdowson's Composition of Foods or the United States Department of Agriculture (USDA) nutritional data. These generic data are calculated on behalf of the British or US government, in laboratories, by analysing a sample of each food for all the macro and micronutrients of relevance.
These kind of generic databases are useful for giving an estimate of what is in an ingredient or a food. Recipes can be entered into a nutritional calculator, specifying amounts of each ingredient and an estimate of the nutritional breakdown of the recipe can be derived.
However, just because the information in these databases is both precise and extensive, it can only provide an average estimate of the composition of the food you are eating. These datasets are only updated every 10 or 20 years, due to the huge cost of analysing so many foods for so many nutrients in the lab. Therefore, the information is likely to be considerably out of date. In addition, international and regional differences in nutritional composition may have implications for using the US software programme Fitday for calculating dietary intake. Fitday uses USDA food composition data and US measurements (cups and fl oz) and products. Also, as the databases are updated only infrequently, newer or specialist foods such as quinoa, whey protein or pomegranate are not represented.
What's in a label?
Most products bought in supermarkets and food shops will have some level of nutritional information on the label. This is usually per 100g product, as it is sold. So 100g of raw Tesco chicken breast would have, say 106 calories, 1 gram of fat and 24 grams of protein in its raw state as purchased. However, any preparation of the chicken will have effects on the nutrient content and relative proportions, as discussed below. This should be taken into account when calculating nutritional intake but bear in mind that the process of accounting for all the different preparation techniques is neither quick nor simple.
The nutritional information on a product label applies only to an average sample of what may be a highly variable food. Data are derived, usually in a laboratory, using a small number of samples of a product, giving an average set of figures. For example, three chicken breasts may be blended together raw, a small sample taken and tested for nutrients. The results are then taken as being representative of all chicken breasts. The nutritional composition of mixed dishes, such as a ready made lasagne, are even more liable to analytical inaccuracies. A small sample will be taken from maybe three lasagnes and blended before a tiny sample of those three dishes analysed for nutritional content. Imagine if, on all three occasions, a particularly meaty part of the lasagne were taken with no cheese or sauce and very little pasta. The nutritional breakdown will be very different from that produced if varied samples had been taken.
Even with relatively basic foods, the amount of variation in nutritional content can be vast. The reasons for this variation can be seen below.
Sources of variation in the nutrient composition of food
- Time – as food is stored, either before or after purchase, the nutrient content will change. Fruit and vegetables may have been cold stored after harvest for many weeks or months before they reach the supermarket, to ensure an all year round supply of produce. In cold storage, nutrient loss (mainly of vitamins and minerals) is minimal but as soon as they are exposed to ambient temperatures, losses accelerate. Try to eat fresh produce as quickly as possible to minimise these effects.
- Preparation – the method of preparation and amount of processing of food may have drastic effects on its nutrient composition.
- Frying food, either shallow, stir or deep frying will cause foods to absorb fat. The amount depends on how much fat is used (deep frying will introduce more fat than stir frying) and how absorbent the food is (battered fish and aubergine will absorb more fat than egg or mange tout).
- However, frying, or indeed any kind of heating (grilling, baking etc.) of food will often cause a loss of water. This is why, however you cook meat, for example, the end product is smaller than the raw product due to loss of water. Many other products lose water when you cook them. This has the effect of concentrating nutrients, so, per 100g cooked product, the nutrient levels will be higher than for the raw product.
- Other foods may gain water on cooking. This applies to dried foods, or those that require soaking such as pulses, dried veg (mushrooms, tomatoes) and pasta. Rice, instant potato products and other grains such as quinoa and couscous gain a lot of water on boiling, some tripling in size. This has the effect of diluting the nutrient levels compared to the unprepared product.
- Preparation of vegetables will have the effect of depleting vitamin and mineral levels. Boiling vegetables can decrease the micronutrient content by 50% or more, especially for water soluble vitamins such as vitamin C and the B vitamins. Overcooking or keeping vegetables warm for prolonged periods will lead to further losses.
- Peeling products will decrease the weight e.g. 33% of a banana is skin; around 10% of an onion is skin. Generic databases of nutritional information may or may not be based upon the edible proportion of each food. However, when analysing nutritional intake, only the part that is actually eaten should be considered.
- Analytical variation – in the laboratory, the inaccuracies associated with analysing for nutrients are multifactoral. Machinery is only accurate to a certain degree, human differences in preparation methods may cause variation in results, reagents used may differ in potency depending on how long they have been stored and so on. The smaller the amount of the nutrient being analysed, the greater the potential for variation. Labs often quote that they can only be accurate to within +/- 20% of the actual figure. Therefore, if a product contained 100mg vitamin C per 100g, the true figure may be anywhere between 80 and 120mg.
- Natural variation – Foods such as fruit, vegetables, grains / cereals, meat and fish will show a high level of natural variation. Plant foods will have a varying nutrient content depending on the mineral content of the soil in which they are grown (for example, wheat from the US is much higher in selenium than in Europe, due to the high soil content of selenium), the amount of water received, the amount of fertiliser used, the degree of pest damage encountered, the altitude at which they were grown, the amount of shade, the temperature and so on. Meat and fish will show variable levels of nutrients dependent upon how active the animals were when being reared (farmed fish and battery hens may have a higher fat to protein ratio compared to wild fish and free range chickens). Composition of meat and fish will also depend upon what the animal ate. Farmed fish and animals fed on omega 6 rich grains and feed will have a higher omega 6 content, so farmed salmon may actually have a higher omega 6: omega 3 ratio than wild salmon. Experiments have been performed on cattle reared on linseed based feed. The meat was shown to have a higher omega 3 content compared to those not receiving the feed.
- Bioavailability / biological value – while not a source of nutrient variation in the conventional sense, the bioavailability of nutrients is key to determining what we actually absorb and therefore utilise within the body. Some plant based nutrients, for example, are less well absorbed than animal based ones, most notably iron (animal-based haem iron versus plant-based non-haem iron). Certain nutrients may inhibit or enhance the absorption of others. For example, phytates in cereals inhibit iron and calcium absorption, whereas vitamin C facilitates the absorption of iron. Therefore, a bowl of bran cereal is best consumed with a glass of orange juice to ameliorate the negative effects of the phytates in the cereal on iron uptake.
It is therefore important, when analysing a recipe or food diary to take all these (and many more) sources of variation into account, using conversion factors, where available (e.g. in the annexes to McCance and Widdowson's Composition of Foods), to produce a meaningful result.
Conclusions – do I analyse my dietary intake?
Given the myriad sources of variation in nutrient content of any foods, we should treat food labels and food composition tables merely as a guide with which to compare foods. Treating nutritional information as gospel should be avoided, as should any precise calculations to the last calorie, gram of fat or milligram of vitamin C. There are simply too many sources of variation and too many ways to go wrong in calculation of dietary intake to get a precise result.
Therefore, focus on the whole diet, think about variety of foods and ignore the fine detail. If you think you need more protein, then increase the portion sizes of the protein portion of each meal until you feel you are getting enough. Likewise with carbohydrate, or any other nutrient. The more you do this, the more you will get to know how your body works, without relying upon precise calculations. Every body is different and their nutritional needs will vary for one person to the next and at different life stages. Obviously if you are clinically deficient or overdosing in a nutrient (usually a vitamin or mineral) then biochemical tests will reveal whether dietary changes or supplements are required. However, the recreational or competitive bodybuilder can happily ignore the precise nutritional profile of each individual food and focus on the total diet.
- Givens, D. I. and Gibbs, R. A. (2006) Very long chain n-3 polyunsaturated fatty acids in the food chain in the UK and the potential of animal-derived foods to increase intake. Nutrition Bulletin, 31, 104-110.
- Musgrove, C (2006) Food information databank systems – everything you ever wanted to know... Nutrition Bulletin, 31, 255-261.
- Food Standards Agency (2002) McCance and Widdowson's The composition of foods, sixth summary edition. Cambridge: Royal Society of Chemistry
- Boaden, D (2008) Step by step guide to calculating the nutritional composition of school lunch recipes. Personal communication.