Can Cannabis Help Diabetics?

Diabetes and Cannabis

 I have seen this relationship between cannabis and diabetes pop up quite frequently now, both in work and within social media. I tend to have not given it much credence after looking to see if I can find a good amount of credible research in the area previously, I didn’t yield too much. However, yesterday diabetes UK decided to post a link to an article written on their website, with implications that cannabis can be used to control diabetes better. I know that this is what they were implying by the comments on the article published, mostly people asking if this was true and whether cannabis can be used to help with diabetes. Sadly, these genuine questions were met with ‘I’ve been taking it for years and it has helped me’ or ‘of course it does, big pharma wont pay as it helps too much and they make money from this’ – because no one who works in the pharmaceutical industry will have diabetes.

Diabetes UK - Cannabis and Diabetes

It appears that diabetes UK jumps on every avenue to promote anything to do with diabetes, and I mean anything. Shortly after publishing the cannabis and diabetes post, they published a post showing a link between aloe vera and better blood sugars, something that I wrote about recently, that has very little scientific evidence. The cannabis post itself appears to be somewhat informative but also quite lacking in information. They say that cannabis causes ‘mild’ physical and mental symptoms in subjects – when compared to other drugs. Which is like saying that vinegar is safe to drink by the litre when compared to chloroform.  They neglect to mention that heavy and prolonged use can lead to paranoia or psychotic states which require hospitalisation. But hey, at least they stress the fact that no one has ever died from cannabis use.  It would be interesting to know how many people had died as secondary result of cannabis ingestion, e.g: careless driving. They state that there is direct evidence for cannabis reducing inflammation and direct evidence to state that type II diabetes is caused by inflammation, something of which is highly based of scientific speculation at this point in time. They go through listing different types of studies looking at diabetes related illness from neuropathy to blood sugar control whilst heavily putting emphasis on one particular…*ahem* “published study”.

Study

Well, I play it fast and loose with the word ‘study’, I’m going to call it a ‘post’ from here on. The post was published on and by ‘AAMC’, the American Alliance for Medical Cannabis (also referred to here as a 'published study'. Seriously). If red flags haven’t appeared already in your mind, you are in the right place. And it was indeed a post, not a study. It was essentially un-referenced information about what cannabis can allegedly do – I don’t have the time, nor the patience, to go through all of the information. They may as well have written it on a lamppost for all I care, it appeared overly idealistic. Any stories of cannabis regulating blood sugar is purely anecdotal at this point, which is, to their credit, disclosed. Surely, this is careless of diabetes UK to propagate, this is so far removed from a study it is ridiculous.

Any Evidence?

Is there any evidence that cannabis can help regulate blood sugars? When people strongly believe something, they tend to focus on one particular paper that confirms and conforms with their views, even if this is one study saying the opposite of thousands, commonly seen with homeopathy or accupuncture. So it comes as no surprise that I see a particular paper associated with this over and over again, which I had never read, until yesterday evening, that is. Many of the pro-cannabis groups really propel and reference to this paper to emphasise the effects of cannabis on diabetes. The American Journal of Medicine published a study in 2013 entitled: The impact of marijuana use on glucose, insulin and insulin resistance among US adults – this is an open access paper, so you can go read this for yourselves. They conclude that marijuana use was associated with lower fasting insulin and Homeostatic Model Assessment (Insulin Resistance) (HOMA-IR), which is essentially a quantification of the beta-cell production (insulin secretion) and insulin resistance - but this model itself isn't without its issues according to the American Diabetic Association.

This is really one of the only studies looking at cannabis use on blood glucose regulation, which they appear to pride themselves on. They gather 4657 people and self-assess their cannabis use by they asked three questions:

·       Have you ever, even once smoked marijuana or Hashish?

·       How long has it been since you last consumed marijuana or hashish?

·       During the previous 30 days, on how many days did you use marijuana or hashish?

When you study using subjects using self reporting data you have to ask yourself, how reliable is the information? Well, when you are dealing with people who are doing drugs, how accurate is the information? Do they remember how much they had? When they had it exactly? How many ‘drags’ of a ‘joint’? A myriad of questions pop up here. Nevertheless the study divided them into groups of never used (n=2103), past users (n=1975) and current users (smoked in the last 30 days, n=579). These groups were then tested to find the HOMA-IR and fasting insulin levels. They compared this to subjects’ information from a previous medical examination in 2010 and between the groups. They found that glucose levels were higher in those that had never (reportedly) used cannabis (103.5mg/Dl) and lower in those that were currently using (99.7mg/Dl). Astonishingly, they also use and reference to HbA1c levels being lower, by 0.1 (5.5 in those never used and 5.4 in those past and current). This is the average blood sugar over 8-12 weeks. Comparing that to information from 2010 will tell you nothing really about an overall group of people.

We have almost as 4 times as many subjects in ‘never used’ compared to the ‘currently using’, which is going to bias the numbers, right? Especially in something as varied as blood sugar. The 'used cannabis' group is anyone who has consumed cannabis, but not in the last 30 days. If you used it for example 4 years previously, once, you would be put into the 'used cannabis group'. Are we saying that one use of cannabis, for example a smoke of a 'joint' is enough to say that cannabis helps even 4 years after? Or is it the fact we have less subjects and that have used compared to the 'never' group, so there is less number variation, bringing the average down? My bets on the latter of the two. 

The study also provides statistics where we adjust multi-variable (they don't state which) excluding diabetics, which makes you ask the question, how many of the subjects in this study, assessing diabetic care with cannabis, were diabetic? What type or diabetes did they have? We have no clue, because the information isn’t provided or even hinted at. Then we get to a couple of sentences within their discussion section which read:

“In addition, data on marijuana use were self-reported and may be subject to under- estimation or denial of illicit drug use. However, underestimation of drug use would likely yield results biased toward observing no association.”

Why would it be only underestimated? Why are we assuming they are not overestimating? Evermore, the sentence following shows some heavy bias. They state in the beginning of the study that little work has been done in this area and this is the first research of this kind. So why would we expect to see results biased toward no association? Why wouldn’t we see results that show the opposite, that is; there is an opposite effect to what they were looking for – cannabis makes diabetes worse. This shows they were looking for a correlation of 'cannabis makes you loose weight and effectively lowers your overall blood sugar levels and insulin resistance' effectively saying;  the cannabis could only improve diabetes or it wouldn’t make a difference at all. Their hypothesis that it would improve diabetes hasn't been proven correct, when you look at the numbers, you don’t have a staggering difference between them, even if you ignore the biased amount of subjects in each group, there are still a lot of variables. The average HbA1c levels within the subjects (5.5-5.4%) are very good to say the subjects are diabetic, suggesting they have it under control with medical intervention or are type II subjects with good physical fitness and healthy eating habits. Another major questions in studies such as this is what was the concentration? They weren't all consuming the same batch of cannabis from the same plant, so there is going to be a significant difference in the levels of THC in the cannabis itself. In 2008, the Home Office released a study (PDF) stating that in 2007 the average cannabis potency of cannabis was 4.1-47%. Not only is this a large range, there is a high probability that they won't be ingesting the same form or strain and thus the same potency with each use. It is also improbable that they have remembered every consumption over their life-time, let alone the exact amount being consumed.

So to summarise this study, we have:

• No indication of the type of diabetes. Type one is insulin not being secreted, whereas type two is insulin resistance, which is what they are reviewing.
• Manipulated data and say that they exclude diabetics in a statistical analysis, but no mention of the population that is diabetic.
• Self reported data, not accurate.
• 4 times as many of the control (never smoked) as the current users, and twice as many as the past users leading to biased results with large gaps in the 'used previously' category. 
• Half of the subjects self-reported they were inactive, 41.1% of which were in the ‘never smoked’ control category. Exercise is proven to help with diabetes and weight loss. This could be a direct correlation which isn’t assessed in the study.
• Average HbA1c levels amount subjects (5.5-5.4%) are very good, showing that the diabetes is already well controlled.   
• How strong is this cannabis? What is the dose? There are a lot of unanswered questions here with a high amount of unattended variables.
• High inaccuracy in results, it is hard to see where comparisons are being drawn between numbers

I appreciate a study with something of this nature cannot really be carried out within reason, that is just drugging people and checking their HbA1c levels.  However, there are institutions in places where cannabis is legal, maybe there’s a good starting point to make for a much more accurate and reliable study.

Conclusion

If you live in the UK, the source for diabetic care should really be focused on Diabetes.org.uk, as they appear to be much better at scientific communication and don’t appear to jump on every scientific story involving the word ‘diabetes’. Any stories of cannabis helping with diabetes are purely anecdotal, which isn’t really evidence. Much like the subjects in the American Journal of Medicine use anecdotal information to recollect the use of cannabis. Anecdotal evidence isn’t clinical evidence, and with that it is irresponsible of a website promoting awareness and giving a platform for people with an illness to propagate such misinformation. The section on cannabis in the diabetes.org gives a much more accurate representation of the effects and the effects on people with diabetes. Raising the point that if you are taking cannabis you can get spaced out and forget to take medication and if you get the munchies, your blood-sugar is going to rise. To diabetes UK’s credit, they do have this information at the bottom of the page, bottom, after telling you how many ways it can help you with bogus 'studies'.