Figure 1.
Deciles of putative predictors for ASD (A) and ID (B) per-male rates.
Plots for females look essentially identical, but the absolute rate of incidence is lower (data not shown). The predictor variables shown here are male congenital malformations of reproductive system (ConGenM), viral infections in males of any age (Viral_M), congenital malformations excluding malformations of the genitals in males, CongMrepM, and spontaneous abortion (spont_abort).
Figure 2.
Estimated Poisson rates of incidence of ASD (A) and ID (B) per male individual of any age.
Figure 3.
Comparison of fixed effects (geographically varying factors) governing rate variation in ASD (A) and ID (B).
The asterisks indicate the level of significance of individual regression coefficients; see the figure key and Table 1.
Figure 4.
Total state-level random effects of ASD and ID incidence in the USA: (A) ASD and (B) ID.
In the figures we color-coded the Empirical Bayes estimates of the state-level random effects, separately for ASD and ID. County- and state-level random effects model the unknown factors that vary geographically and govern differences in county-specific disease rates after accounting for all fixed effects (see Methods).
Figure 5.
Total county-level random effects of ASD and ID incidence in the USA: (A) ASD and (B) ID.
In the figures we color-coded the Empirical Bayes estimates of the state-level random effects, separately for ASD and ID. While county-specific random effects are directly comparable within the same state, comparison of these effects across different states is not meaningful, because each state-specific random effect determines the baseline disease rate for each county in the corresponding state, and these baseline rates vary across states.
Table 1.
Markov chain Monte Carlo estimates of regression weights, corresponding event rate ratios (exponential of regression weights) and 95% event estimate credible intervals.
Table 2.
Markov chain Monte Carlo estimates of covariances and correlations of random effects across two phenotypes.
Table 3.
Summary of the prior state-level studies regarding geographic clustering of ASD.