It's kind of misleading to call them nonbenzodiazepines, since they basically try to mimic benzodiazepines in every way possible, working on the same receptors and all.
Although they are an agonist of the same type of receptors as benzodiazepines ( the GABA-A receptors), the exact subunits of the GABA-A receptors which zolpidem and benzos bind to vary quite a bit between the two. Here's some info on the subject:
Zaleplon and Zolpidem both are agonists at the GABA A ɣ 1 subunit. Due to its selective binding, Zolpidem has very weak anxiolytic, myorelaxant, and anticonvulsant properties but very strong hypnotic properties. Zolpidem binds with high affinity and acts as a full agonist at the α1-containing GABAA receptors, about 10-fold lower affinity for those containing the α2- and α3- GABAA receptor subunits, and with no appreciable affinity for α5 subunit-containing receptors
Zolpidem has a tenfold higher affinity for α1 than α2 and α3; this is very different from most benzodiazepines, which don't have such a large difference between their α1 affinity and α2/α3 affinity. Its also worth noting that the "Big Three" benzos (alprazolam/Xanax, diazepam/Valium, and clonazepam/Klonopin) have a higher affinity for α2/α3 than α1; binding to α2/α3 produces much more anxiolytic effects (and hence, higher potential for addiction), where as α1 produces greater hypnotic effects (hence, zolpidem is better for sleep, as well as being less addictive).
Although zolpidem is still physically (and slightly psychologically) addictive nonetheless, it is
far,
far less addictive than benzodiazepines.
When I was going through my benzo and zolpidem phase about 5 years ago, I remember zolpidem still producing the same effects at the same doses for weeks (perhaps months) -- I also did not keep want to taking more of it after the effects had worn off (the physical side effects at higher recreational doses deterred me from taking it more than once every 1-2 days, but it was still entertaining to use). However, I remember my dose of benzos doubling after about 10 days, and then quadrupling after a month -- eventually, the effects of benzos became just a "background feeling". I believe this is due to the half-lives of benzos being so massive, that tolerance is simply unavoidable.
With zolpidem's short half-life of 2-3 hours, tolerance has a much slower onset, and withdrawals/tolerance would mostly be limited to difficulty sleeping - where as benzos produce much more harsh withdrawal symptoms such as anxiety, malaise, and in some cases seizures.
To become physically dependent to zolpidem, one would literally have to
work at it. Taking at least 4-6 doses a day. Otherwise, the risk of taking zolpidem every day is pretty much limited to difficulty sleeping.
I agree however that it shouldn't be marketed as "non habit forming"; but when compared to benzos,
it is less habit forming (albeit still "habit forming"). Zolpidem is also not meant to be taken every night; it is meant for short term use (less than a month) to merely get a person's sleep schedule back on track, or to be taken only occasionally during severe episodes of insomnia. If a doctor is giving a patient a long-term once a day prescription to zolpidem, they aren't prescribing it correctly and its almost malpractice.
