Figure 5.

Comparison of estimated effect with “true” effect for various BMI thresholds. N = 70,261, ${\theta }^2=0.0256$, and 95% confidence intervals are generated over 1000 bootstrap resamples. “True” corresponds to the sample estimate using BMI as the exposure; “naive” corresponds to using the binary measure as the exposure ${\beta }_D$; and “latent” corresponds to the latent variable estimator ${\beta }_L$ of Section 3.2. BMI, body mass index