“…In addition, it com-Prtses control subjects suffering exclusively from or-}hopedic disorders and admitted to a hospital different rom those which provided the cases. Although this Was necessary under the existing conditions it may, nevertheless, raise questions concerning the compara-Epidemiological evaluation of the role of nutrition in the development of gastric cancer [or its postulated precursors (Munoz and Matko, 1972;] has relied mainly on case-control investigations (Wynder et al, 1963;Acheson and Doll, 1964;Meinsma, 1964;Higginson, 1966;Graham et at., 1967;Hirayama, 1971;Haenszel et al, 1972;Bjelke, 1974;Madan et al, 1974;1975;Haenszel et al, 1976a,b;) although risk correlations (Haenszel, 1958;Hakama and Saxen, 1967;Munoz et al, 1968; Armstrong and Doll, 1975;Joosens and Geboers, 1981;Tominaga et al, 1982;Correa et al, 1983;Kono et al, 1983) and cohort studies (Hirayama, 1982) have also been undertaken. Most authors interpret the collective evidence as indicating that salted and starchy foods increase the risk of gastric cancer whereas fresh fruits and vegetables decrease this risk (Graham, 1975;Haenszel et al, 1976a;Weisburger, 1979;Hirayama, 1982;Nomura, 1982;Tominaga et al, 1982;Correa et al, 1983;Kolonel et at., 1983;Mirvish, 1983;Palmer and Bakshi, 1983).…”