ofonorow wrote:As just pointed out, what if they lost their teeth/amalgams?
It's a valid concern, and one the researchers were aware of as they adjusted for it in the final analysis. However, one limitation of the study seems to be that they only measured number of amalgams at the baseline, although I'm only basing this on the abstract.
Abstract wrote:Altogether 1462 women aged 38, 46, 50, 54 and 60 yr were initially examined in 1968-69 in a combined medical and dental population study in Gothenburg, Sweden. Number of tooth surfaces restored with amalgam fillings was assessed. The incidences of myocardial infarction, stroke, diabetes, cancer and overall mortality were determined during a 20-yr follow-up period. Women with few amalgam tooth fillings had increased incidence of myocardial infarction, stroke, diabetes and early death compared with women with a large number of fillings. However, the significant inverse correlations between number of amalgam tooth fillings and the endpoints studied disappeared when number of teeth and socioeconomic group were included in a multivariate analysis. The study thus did not provide any evidence for a correlation between amalgam fillings and cardiovascular disease, diabetes, cancer or early death.
http://www3.interscience.wiley.com/jour ... 2/abstractHow was cardiovascular disease measured? Incidence of cardiac events?
It would appear so.
(Cholesterol levels would be more interesting).
Why? It's not a "hard" endpoint - surely, mortality and morbidity are the more important endpoints.
If they can easily determine whether someone has CVD, I'd like to know how they do it because we could then (finally) run objective studies.
They measured events - you'd need to run a fairly large study otherwise it would be underpowered to detect any statistically significant differences. Since a large trial or observational study seems unfeasible, I still think the best option would be a consecutive case-series or a placebo-controlled crossover trial, depending on the condition being treated.
It is known that the older Amalgams are "safer", I forget the year, but the alloy after some time in the 1970s became an order of magnitude more dangerous, and how do they account for root canals?
That's probably a valid question (assuming that amalgams were only measured at baseline), although even if the older amalgams were safer, one would still expect to see some effect. I simply assumed that you were refering to amalgams when you spoke of dental toxicity, because of the mercury issue. The root canal hypothesis was news to me when I read that dr. Levy suggested it as the main cause of cancer and heart disease.
So this inverse result is prima facie nonsense. Contrived. Again, the first thing I would look at before bothering to read the paper is the source of funding.[/color]
As I said before, the inverse association was there before potential confounders like socioeconomic status were taken into account. The source of funding is not given in the abstract, so I will have to look it up in the full paper whenever I get a chance to visit the library - but given the multidisciplinary nature of the study, my guess is that is some form of public funding not necessarily tied to any dental associations. I agree that the inverse result is not probable, but this is reported in the abstract as being attenuated once certain confounders were taken into account.
The null result for cardiovascular disease is confirmed by a New Zealand study, which appears to have a better design because it had access to yearly dental records (and thus a more accurate measure of amalgam exposure):
http://ije.oxfordjournals.org/cgi/conte ... 4/894#TBL4