RESUMEN
OBJECTIVE: To analyse the Na content of bread by comparing the amount of salt and Na among the label, laboratory analysis and international guidelines. DESIGN: Ten selected bakeries provided 3239 randomly selected samples of bread, which were weighed on-site. Triplicate samples were retrieved from each bakery (thirty samples) for analysis. Bread production was observed, and ingredient labels were queried to determine salt weights, which were used for comparison with the laboratory analysis. Flame photometry and the method for chlorides were utilised for analysing Na. Laboratory findings were compared to nine different international nutritional guidelines for Na consumption. SETTING: Florianopolis, south of Brazil. PARTICIPANTS: Ninety independent bakeries locally producing Portuguese rolls were queried; rolls from ten conveniently selected bakeries were retrieved for further analysis. RESULTS: The average weight of the rolls was 50·2 ± 5·3 g. The average amount of salt (g) per roll, by laboratory and label analyses, was 0·69 ± 0·0 and 0·62 ± 0·1 g, respectively. The mean level of Na (mg) reported on nutrient labels (478·2 ± 93·4/100 g) was significantly lower than by laboratory analysis (618·2 ± 73·8/100 g), P < 0·001. There was a difference for Na in rolls produced in the bakeries considering the unit weight of rolls (P ≤ 0·001) per 100 g (P = 0·026) and the mode of production. The consumption of two averaged units of rolls was equivalent to 51·7 % of the Brazilian guideline daily amount for Na for children and 31 % for adults. CONCLUSIONS: The nutrient labels underreported Na values. This study strengthens the importance of monitoring Na of breads in Brazil.