Outcomes: children’s food consumption

Children’s food consumption was assessed with a 51-item semi-quantified food frequency questionnaire (FFQ) developed in the DAGIS study and designed to measure particularly the consumption of fruit and vegetables and sugary foods and drinks. Six food groups were assessed: fruit and vegetables; dairy products; fish, meat and eggs; cereal products; beverages; and other foods. An earlier, 47-item version of the same FFQ, excluding the questions on food amounts, showed acceptable test-retest reproducibility [41] and validity for ranking food group consumption against 3-day food records [42]. After the validation and reproducibility studies, 4 food items concerning salty bakery products (not used in this study) and the amount questions (used in this study) were added to the FFQ.

Parents filled the FFQ online. They first reported how often the child had eaten the specific foods or drinks and then answered the questions on amount consumed per day. The FFQ questions on amounts are found in Supplement 1. The questions concerned the previous 7 days. The frequency of consumption was asked about as follows: how often during the seven previous days has the child eaten the listed foods and drinks. Answer options were “not once”, “X times per week” or “X times per day”. The parent either ticked the “not once” option or reported how many times per week or per day the child had consumed the listed foods and beverages. The questions concerning the amounts of foods consumed included between 6 and 12 answer options per food/drink (Supplement 1). Answer options were based on the usual consumption amounts of these foods/drinks found in earlier studies in this age group [43, 44]. Pictures of different portion sizes and their weight were available for each food [45]. Calculating the amounts consumed per day were made as follows: first, eating occasions per week were calculated from the eating frequency questions, with seven consumption frequencies per week used as the maximum. Secondly, the midpoint (except for the second lowest and highest answer options) of each answer option on the amount consumed per day was multiplied by the consumption frequency per week. The amounts used in the formation of the variables are shown in Supplement Table 1. Lastly, the amount consumed per week was divided by seven to calculate the consumption amount per day (g/day).

The FFQ measured only those foods and drinks consumed outside ECEC hours: foods eaten at ECEC centers were left out. In Finland, all municipal ECEC centers provide the foods eaten during childcare hours, and parents cannot reliably report the consumption of these foods.

Four sum variables were formed for this study to represent children’s consumption of FV, sugary everyday foods, sugary treats, and SSB (in g/day). 1) FV consumption comprised the sum of the amount in grams (g) eaten per day of four food items: a) fresh vegetables; b) cooked and canned vegetables; c) fresh fruit; and d) fresh and frozen berries. 2) Consumption of sugary everyday foods comprised the sum of the amount in grams (g) eaten per day of five food items: a) flavored yogurt; b) puddings; c) berry and fruit soups; d) sugared cereals and muesli; and e) berry, fruit and chocolate porridge. 3) Consumption of sugary treats comprised the sum of the amount consumed of five food items: a) ice-cream; b) sweet cookies and cereal bars; c) cakes, cupcakes, and other sweet pastry; d) chocolate; and e) sweets per day. 4) Consumption of SSB comprised the sum consumed per day of three food items: a) flavored and sugar-sweetened milk and plant-based drinks; b) sugar-sweetened juice; c) and sugar-sweetened soft drinks. In the analyses, the difference in follow-up and baseline value of each food variable was used as the outcome variable to represent the change in food consumption.