Poster: The relation between exposure to environmental selenium and cancer risk - A failure of observational research?
- Details
- Published on Sunday, 24 August 2014 12:00
Introduction: Selenium is a metalloid element with both nutritional and toxicological properties. Changes in environmental exposure to selenium might modify cancer risk, according to epidemiologic and laboratory studies. However, such relation is extremely controversial, since the first observational studies and one randomized controlled trial (RCT) originally suggested an inverse relation between selenium intake and cancer, while most recent studies including the large RCTs carried out in the US have shown no effect or adverse effects on cancer risk.
Methods: We meta-analysed the results of fifty-six observational (cohort and cohort-nested case-control) studies, including over 1,200,000 participants, and of nine randomized controlled trials (RCTs), with 46,304 total participants, using random effects models. We pooled the incidence of any cancer and of prostate cancer, one of the most commonly investigated outcomes, which was originally suggested to be prevented by increasing selenium intake.
Results: For observational studies, we found a reduced cancer incidence (summary OR 0.69, 95% CI 0.53-0.91). In contrast, RCTs showed that selenium supplementation had little effect on the risk of any cancer (RR 0.94, 95% CI 0.81-1.10), and limiting the analysis to trials with a low risk of bias, the RR further approached unity (1.01, 95% CI 0.93-1.10)(Figure1).
For prostate cancer, observational studies indicated a considerably decreased risk (OR 0.72, 95% CI 0.62-0.86), while RCTs showed little evidence of a beneficial effect of selenium supplementation (RR 0.92, 95% CI 0.75-1.12). When we limited the analysis to RCTs with low risk of bias, no effect whatsoever on prostate cancer risk emerged (RR 1.01, 95% CI 0.90-1.14)(Figure2).
Conclusions: The results of the most recent, well-designed RCTs investigating the relation between selenium and risk of any cancer or of prostate cancer have been entirely disappointing, suggesting no effect or adverse effects of the metalloid (such as an increased risk of skin cancer), in contrast with a previous RCT and with several observational studies. It is not clear why observational and experimental studies yielded such different results. The causative factors may include exposure misclassification in the observational studies, which were based on overall selenium content in peripheral biomarkers and
not on levels of single selenium compounds in target issues, or more generally confounding and other biases or effect modification by genetic factors. Overall, methodological issues concerning the relation between selenium and cancer risk have important implications for epidemiological research and for public health recommendations.
