An article in this month's (May 2013) American Journal of Medicine shows that current users of marijuana have better glucose control than do former users, who in turn have better glucose control than do abstainers. This study was a retrospective one, and needs to be validated by a forward study. This will probably never happen, for obvious reasons, much as retrospective studies showing that cigarette smokers have a lower incidence of Parkinson's disease were never followed up.
The use of marijuana was based on interviews with the 4657 members of the study; of this group 579 were current users and 1975 were prior users. Fasting glucose and insulin levels were measured, and the degree of insulin resistance was calculated. Diabetics and pre-diabetics both have insulin resistance.
This study was triggered by prior studies showing that marijuana users have a lower incidence of both obesity and diabetes, but there have been no similar studies of THC users. Current marijuana users in this study were seen to have lower levels of fasting insulin, and were less likely to be insulin resistant. They also had higher levels of HDL cholesterol and smaller waist sizes. Past users (more than 30 days before the study questionnaire) also had a lower incidence of insulin resistance than did non-users, but not as low as did current users.
The results of this study need to be followed up, but given the current federal law classifying marijuana as a Class I drug, meaning "no significant medical benefit", it is difficult to envision a prospective study being done. One possibility might be in the state of Washington or Colorado, where marijuana was recently legalized. OTOH, perhaps the federal law could be changed to make it legal to treat diabetes and the metabolic syndrome with THC and to do a study of its effect.