Table 1.

Estimated effect of 10-fold increase in prenatal exposure to mercury. The results are shown separately for Cohort 1 and Cohort 2. A p-value is provided from the test that the two mercury effects are the same, and then a joint estimate is shown. The results are based on a latent mercury exposure variable and they have been adjusted for confounders other than PCB exposure. Number of children with outcome information was about 900 in Cohort 1 and 160 in Cohort 2.

Outcome	Cohort 1		Cohort 2		Test for equality	Cohort 1 & 2
	β	95%-conf.	β	95%-conf.		β	95%-conf.	p
NES2 Finger tapping
Preferred hand	−1.966	−3.228; −0.704	3.449	−0.226; 7.123	0.0069	−1.424	−2.629; −0.219	0.021
Non preferred hand	−1.290	−2.510; −0.069	0.838	−2.108; 3.783	0.19	−0.985	−2.118; 0.148	0.088
Both hands	−2.975	−5.546; −0.404	−1.569	−8.188; 5.051	0.70	−2.795	−5.201; −0.389	0.022
NES2 CPT
Reaction time	44.027	20.491; 67.563	7.680	−30.388; 45.748	0.11	34.080	13.832; 54.327	0.0011
Ln total missed	0.139	0.027; 0.251	0.111	−0.066; 0.287	0.79	0.131	0.036; 0.225	0.0069
Boston Naming Test
No cues	−2.120	−3.239; −1.001	−2.020	−4.679; 0.639	0.95	−2.105	−3.137; −1.073	0.0001
Cues	−2.223	−3.328; −1.119	−2.0780	−4.683; 0.527	0.92	−2.201	−3.219; −1.183	<0.0001
WISC(R)
Similarities	−0.016	−0.908; 0.876	−0.969	−2.192; 0.253	0.22	−0.347	−1.071; 0.377	0.35
Sqrt. Block Designs	−0.234	−0.469; 0.002	0.097	−0.512; 0.706	0.32	−0.191	−0.411; 0.029	0.089
Digit Spans	−0.228	−0.539; 0.083	0.237	−0.348; 0.821	0.17	−0.126	−0.402; 0.150	0.37
CVLT-Children
Learning	−1.637	−3.526; 0.253	0.255	−4.311; 4.820	0.45	−1.361	−3.110; 0.387	0.13
Short delay	−0.538	−1.097; 0.021	−0.510	−2.012; 0.992	0.97	−0.535	−1.059; −0.011	0.046
Long delay	−0.723	−1.351; −0.094	−0.244	−1.669; 1.181	0.55	−0.645	−1.221; −0.069	0.028
Recognition	−0.261	−0.725; 0.203	0.760	−0.358; 1.877	0.10	−0.114	−0.546; 0.318	0.61