There are no common disorders - only extremes of quantitative traits.
That's the argument made by Plomin, Haworth, and Davis in a great review paper just published online in Nature Reviews Genetics. One of the main themes presented here is that as a disorder becomes more common, a case-control design becomes less and less powerful to identify associated genetic variants because cases become less extreme and the control group becomes increasingly contaminated by "near cases".
Some qualitative disorders have obvious relevant quantitative traits - BMI for obesity, blood pressure for hypertension, mood for depression. I recently authored a review with Dana and Marylyn making a similar argument in the context of pharmacogenomics research. The authors admit, however, that quantitative measurements are not at all apparent for some disorders, such as alcoholism, arthritis, autism, dementia, diabetes, or heart disease.
The review also has a glossary for the uninitiated, encouraging the use of quantitative vocabulary, like linear regression or ANOVA instead of logistic regression, variance and mean differences rather than risk, and covariance rather than comorbidity.
The conclude with statement that the extremes of a distribution are important medically, but there is no scientific or statistical advantage in analyzing artificially constructed disease labels that evolved historically on the basis of symptoms rather than etiology.
Common disorders are quantitative traits (NRG AOP).