*Geek Box: A Priori Indices in Nutritional Epidemiology
Most approaches to dietary assessment in nutritional epidemiology focus on foods and nutrients as the exposure of interest. However, there are a number of different methods of analysing total dietary patterns, or characteristics of dietary patterns. An ‘a priori indices’ is a fancy way of saying a scoring index to quantify the healthfulness of a dietary pattern, or to quantify specific characteristics of the diet.
Such indices are considered ‘a priori’, because they are derived from analysing the overall wider research and deciding in advance that, for example, fruits and vegetables are healthy. An example of this type of dietary assessment would be the Alternate Healthy Eating Index 2010 [AHEI-2010]. Based on the original 1995 Healthy Eating Index, the AHEI-2010 consists of 11 dietary components with a maximum of 10-points for each component, contributing to a total score of 110. The dietary components associated with lower risk of disease, including vegetables, wholegrains, whole fruit, nuts and legumes, long-chain omega-3 fatty acids, and polyunsaturated fats, have a points score that rises with increasing consumption of the dietary component from 0 up to a maximum score of 10.
In contrast, dietary components associated with negative health outcomes, including sugar-sweetened beverages and fruit juices, red/processed meats, trans fats, and sodium, are score inversely to consumption, i.e., 0 for high intakes up to 10 for low intakes. This allows for the overall healthfulness of an individual’s diet pattern to be quantified in a single number, and the overall scores in a cohort can be divided into different levels and analysed in relation to disease outcomes.
Diet indices have been developed for inflammation – the Dietary Inflammatory Index II – the financial cost of diet, a Mediterranean diet score, and low-carbohydrate diets. The low-carbohydrate diet index was first described by Halton et al. (8). Intakes of carbohydrate, protein, and fat, were divided into deciles – 10 levels – each according to intake as a percentage of total energy intake. Participants with the lowest intake of carbohydrate received a score of 10, the next level 9, the next level 8, and so on down to the highest level of carbohydrate at 0. However, protein and fat were scored in the opposite direction: the highest levels received 10 points, while the lowest received a score of 0. The total points score available was therefore 30 [10 for each macronutrient].
These scores can then be analysed for the relationships between high or low scores, and incidence of disease endpoints. A high score would therefore reflect having the lowest carbohydrate intake alongside the highest protein and fat intakes, thus capturing the exact dietary pattern – at the level of macronutrients – that proponents of the LCHF diet recommend. One of the major advantages to scoring indices like this is that they are inherently adaptable. For example, the LC diet score can be calculated separately for animal protein and fat, and for protein and fat from vegetable sources. Thus, it can be used to further refine the definition of the diet pattern beyond the broad macronutrient definition.