4 ms·
weird methodology if you ask me but there it is: The proportion of mothers who were both older and had presumed TCE exposure was more than six-fold greater amo
by carlob 14y ago
weird methodology if you ask me but there it is:
The proportion of mothers who were both older and had presumed TCE exposure was more than six-fold greater among case infants than among control infants (3.3% [8/245] versus 0.5% [19/3780]). When adjusted for other variables, CHD risk was over three-fold greater among infants of older, exposed mothers compared to infants of older, nonexposed mothers (adjusted OR, 3.2; 95% CI, 1.2-8.7). Older maternal age, alcohol use, chronic hypertension, and preexisting diabetes were each associated with CHDs (adjusted ORs, 1.9, 2.1, 2.8, 4.1; 95% CIs, 1.1-3.5, 1.1-4.2, 1.2-6.7, 1.5-11.2, respectively), but residence close to TCE sites alone was not.
http://www.ncbi.nlm.nih.gov/pubmed/15390315 http://www.ncbi.nlm.nih.gov/pubmed/15390315
- uvdiv 14y ago"[...] were each associated with CHDs [...] but residence close to TCE sites alone was not." So, they actually found no correlation with TCE exposure. There found a correlation with one of the subgroups [TCE + other variable], one out of many, that they selected after examining the data. There's an XKCD about this: http://xkcd.com/882/ http://xkcd.com/882/ Or did I misunderstand something?
- carlob 14y agoAs far as I understand only women that were older (≥ 38) had a significant effect from being exposed (on top of the increased risk due to being older).
- Evbn 14y agoYou are misunderstanding the statistical theory. The xkcd comic was making a joke about misused priors. You have no reason to believe that this paper used a wrong prior and peer reviewers overlooked it, nor have you examined the choice of p-value.