The relevance of food composition data for nutrition surveys in rural Tibet: pilot study in the context of Kashin-Beck Disease

Kashin-Beck disease (KBD) is an endemic and chronic osteochondropathy. This disease principally occurs in the Tibet Autonomous Region and in several provinces of the People's Republic of China. The etiology of the disease remains obscure although environmental factors are assumed to be involved. Diet, in particular, differentiates the rural community, affected by KBD, from the other communities (nomads and city-dwellers), who remain unaffected. In anticipation of a nutrition survey, this study aimed to measure the mineral content (Ca, P, Mg, Fe, Zn, Mn, Cu, Ni, Se, Al, Sr, Mo, Cd, As, Pb, Hg, Cr, and Co) of eight Tibetan staple foods and to compare the results against two food composition tables (FCTs). Foods were sampled in twenty households selected from both an endemic and a non-endemic area of rural Tibet. Ten minerals involved in bone metabolism were measured using atomic and molecular spectrometric methods. Results revealed that a very limited number of food/constituent pairs showed a variation in mineral composition during a single year of testing for a given region. In addition, results showed significant differences in mineral content between the endemic and the non-endemic area, especially for wheat flour. Following our analysis of the mineral content of the Tibetan food samples, results were statistically compared with similar foods listed in two food composition tables: the USDA National Nutrient Database (USDA Food Search for Windows, Version 1.0, database version SR23), and the China Food Composition Table (book 1, 2nd edition). More than 50 to 60% of p-values < 0.05 were highlighted, suggesting the inappropriateness of using FCTs as a reference for nutrition surveys in rural Tibet, and emphasizing the need for analysis of traditional foods. Differences were found to be more or less marked depending on the element considered, and calcium content seemed to show the greatest difference. Although it is obviously too early for definite conclusions to be reached (insufficient number of samples by food and insufficient number of foods analyzed), it seems that the present pilot-study indicates significant discrepancies between measured and tabulated values of the mineral content of certain foods. A more complete survey would therefore seem mandatory.