The estimated associations of BPA with reported diabetes in 2005/06 was modest but
with wide confidence intervals; however, the diabetes association remained significant in pooled data (OR = 1.27, CI: 1.12 to 1.43, p = 0.0004).
However, associations with diabetes seen in the 2003/04 NHANES survey was not evident in the current study, possibly because of the reduction in statistical power due to the lower levels of BPA in 2005/06, resulting
in wide confidence intervals around estimates.
But McCormack notes that the new results actually are consistent with the data from the earlier trial, which had
a wide confidence interval and included the possibility that the product did not work.
If you wish to test that observed trends are within model distribution, you would use the standard deviation of the trends of models and you would have
the wide confidence intervals you expected.
Also, this might have reduced the power to detect relations with the chosen variables and contributed to
the wider confidence intervals of this group at the youngest and oldest age groups.