Global geographical variation elemental and arsenic species concentration in paddy rice grain identifies a close association of essential elements copper, selenium and molybdenum with cadmium

Despite the centrality of staple grains for human well-being, both as a source of nutrients and of toxic ions, there is little understanding of where and how elements vary, and if there are particular elements that correlate. Here, for shop bought polished (white) rice we comprehensively characterize trace (arsenic species, cadmium, copper, iron, manganese, molybdenum, rubidium and zinc) and macro-nutrients (calcium, chlorine, potassium, phosphorus and sulphur) for grain purchased in 18 countries, across 4 continents, a total of 1045 samples, to investigate if there were any major differences between geographic location and elemental content, and to observe if there were any patterns in elemental distribution. Greatest variation in the median was observed for the non-essential rubidium (15-fold) and arsenic species (5-fold). Rubidium was highest in the Americas, lowest in Europe, while inorganic arsenic and dimethylarsonic acid (DMA) were low for Africa and high in the South American and European continents. Cadmium had the highest concentration in Asian samples and lowest in South America. Calcium, potassium, molybdenum and phosphorus, were highest in European and lowest in African grain, though the fold-differences were relatively low, ~ 0.2, while iron was highest in African grain and low in European, Asian and South American grain, with a ~ 2-fold difference. Selenium was also higher in Africa versus other regions, and copper, manganese and zinc where highest in American grain. Factor analysis showed that copper, cadmium, molybdenum, rubidium and selenium were strongly associated together, and these element’s factor loadings were diametrically opposed to less tightly associated calcium, chlorine, manganese, potassium, phosphorus and sulphur. Stepwise additions linear region analysis was performed on log transformed concentrations to investigate cadmium associations in more detail. Selenium was the greatest predictor of cadmium concentration, followed by molybdenum, accounting for over 50% of the contribution to the adjusted R2. Arsenic species were only weakly correlated with other elements. The implications for these findings with respect to dietary nutrition are discussed.