Abstract
In order to determine whether simplified indicators for usual consumption of selected food groups, specifically those derived from either the percentage of the number of daily portions, the percentage of total daily weight consumed, or both, could serve as proxies for the conventional expression of daily energy intake, these indicators were computed and compared from food-frequency data in a data set. Food consumption was reported in frequency categories and portion sizes per month, per week, or per day, and the cumulative sum was divided by 365 to provide a daily average. The survey was done in the township of Santa Cruz and three hamlets of a rural county seat in Guatemala. Data from food-frequency questionnaires from 269 individuals (55 men and 214 women aged 16 to 86 years) were analyzed. For foods of plant origin, the percentage of total energy, percentage of total food weight, and percentage of total number of portions consumed showed low correlations (r < .45). When subdivided into specific foods and food groups (fruits and vegetables, red meat, etc.), marked differences were revealed across sites and among different indicators of consumption. Despite the simplicity of calculation, neither the percentage of weight of food in a group nor the percentage of portions consumed could serve reliably as proxies for their contributions to the percentage of total energy in this rural population in Guatemala.