Serum selenium levels as indicators of body status in cancer patients and their relationship with other nutritional and biochemical markers

A cross-sectional study of serum selenium levels in patients (n=59) with different types of cancer from southeastern Spain was carried out using hydride generation atomic absorption spectrometry. The subjects were divided into four groups according to the cancer location (respiratory, digestive, haematological and gynaecological groups). Serum selenium levels in all patients (54.41±24.80 mg/l) were significantly lower (P<0.001) than those determined in control groups [healthy subjects from the same area (n=130) and institutionalized elderly people (n=93)]. Mean serum selenium concentrations were not significantly different among the four groups considered (P>0.05). Linear regression analyses performed on serum selenium levels and biochemical markers (total cholesterol, triglycerides, transaminases, uric acid and urea) did not establish any statistically significant correlation (P>0.05). No significant relationships between serum selenium concentrations and sex or age of patients was observed (P>0.05). Given the marked overlap between the two ranges of the populations (the means are within approx. 1/2 S.D.) the predictive values of serum selenium are low. Thus, there is indeed a statistical significance between the means, but selenium cannot be used to determine whether or not a patient has cancer disease.