Hi emily and gcoe!
I came across the same information about a week ago, during reading Hoppenfeld's Orthopaedic Neurology, chapter 2, page 66. It's about the same explanation as gcoe gave but I'll add my comments.
Remember that in the lumbar spine, most of the nerve roots travel a lot down across the canal and do not exit it at the same level of origin from the cauda equina (the last reaching the L1-L2 level and then it's ONLY the roots themselves in the canal, not like at the rest of the spine where the roots travel down just a little ALONG WITH THE CORD and then exit it through the intervertebral foraminae). As they travel down, the roots pass through different spinal levels, with the lower roots passing more than the higher ones (e.g. L5 root passes through L1-L2, L2-L3, L3-L4 and L4-L5 while L3 only through the first 2). So, in the case of a herniated disc at the lumbar spine, it is obvious that many roots should be irritated and the higher the level of hernia the more roots would be. Anatomically, according to Hoppenfeld's book, the site of greater pressure is at the root just one level down. For example, in an L4-L5 herniated disc, the L4, L5, S1 etc roots pass posterior to the disc and one could guess that all of them should be irritated. In fact, due to anatomical structure, it's the second root (L5 in this example) that is most irritated because that's where the disc will find less pressure and move towards, impinging the root that is located there. I would add that because of limited space in the canal and that the roots are close to each other (we’re not talking about cm but about mm), the FIRST root to be irritated would be L5 in our example. If the problem gets worst, though, then other structures in the area, that is the other roots too, will be irritated. That’s why symptoms do not always depict malfunction at one spine level, in spite the fact that anMRI would demonstrate a hernia only at, let’s say, L2-L3. This must be because not only L3 root is irritated but possibly L2 or L4 or even L5, in different degrees of course.
To conclude, try reading this page from Hoppenfeld’s book together with gcoe’s article and it will become obvious.
Looking forward to other comments too!
ilias