The generalized estimating equation regression analysis did not show increased prevalence of hypertension, diabetes or urinary stones with increasing urinary cadmium in this highly exposed population(urinary cadmium ≥5 mg/g creatinine). In our previous study of a large-sized population who lived in these contaminated areas and had various levels of urinary cadmium, cadmium exposure was found to significantly increase prevalence of hypertension and urinary stones ,but less signifi- cantly correlated with diabetes (Swaddiwudhipong etal.,2010c, 2011). Schwartz et al.(2003) used data from the NHANES among the US adults, who generally had low cadmium exposure, and found a positive association between urinary cadmium and diabetes prevalence. The present study suggests that he dose- response effects of cadmium related to the three disorders may not be noticeably detected in those with high urinary cadmium levels. Cadmium-exposed persons may express comparable toxic health effects after prolonged excessive exposure. However, the small sample size of the study might also lower the power to detect a significant association. Similar explanations might also be applicable to the findings that correlations of urinary cadmium with urinary excretion of total protein and calcium were not demonstrated in the study persons, but were found in a large- sized survey population with low-and high-level cadmium exposure in these areas (Swaddiwudhipong etal.,2010a, 2011).