A customer strolled up to the counter one night when I was working in a retail pharmacy:
“My doctor says I have prediabetes. I don’t want to take any drugs. Do you have something natural I can use to cut my blood sugar?”
I looked at him in the eye, and pointed at his sizeable midsection. “Sir, if you’re at risk for diabetes, and you don’t want to take medication, the single best thing you can do for yourself is lose some weight.”
He grinned and asked, “Great – what supplement can I take to help me?”
This type of discussion occurs all the time. A patient has been assessed by their physician, and informed that they have a medical problem of some sort. The patient, reluctant to accept the physician’s evaluation, heads to the pharmacy for a second opinion. In some cases, the patient may question the physician’s advice: “All my physician wants to do is prescribe drugs.” Yet there’s a disconnect when it comes to strategies for management. More often than not, non-drug approaches are rejected out-of-hand (probably because the sample I speak with have already made the decision to buy something). And in those that are leery of medical management, there’s often a willingness to consider anything that’s available without a prescription – particularly if it’s perceived as “natural.” Natural products are gentle, safe, and effective, while medicine is thought of as unnatural, harsh, and potentially dangerous. This is the naturalistic fallacy, nothing more. Purveyors of supplements leverage the naturalistic fallacy into the marketing strategy of choice for almost all supplements and “alternative” medicines. And it leads to bad health care decisions.Alternative medicine for diabetes is big business, because the public health burden of diabetes is massive, and growing. In 1985, the worldwide prevalence was 30 million people. In 2000, it was 150 million. By 2030, it could be 250 million. Why are more people being diagnosed with diabetes? Obesity, sedentary lifestyles, and an aging population. At its core, diabetes is a disease of sugar (glucose) management. Insulin, secreted by the pancreas, allows cells to use glucose. When the pancreas doesn’t produce insulin, it’s called Type 1 diabetes. This is an autoimmune disease that strikes early in life, and was a death sentence until insulin was discovered. When the pancreas can produce insulin, but the amount is insufficient, or when there’s a problem with the uptake of insulin into cells, it’s termed type 2 diabetes. 90% of all diabetes is type 2. Typically a disease of older adults, type 2 diabetes can potentially be treated without drugs of any kind, but success rates are low and medication is eventually advisable. There’s also gestational diabetes, a disease of pregnancy, and prediabetes, where blood sugars are elevated, and diabetes is an expected future diagnosis.
Given the prevalence of diabetes and the chronic nature of the disease, it’s no surprise that CAM is a popular treatment option. I don’t see a lot of CAM use in Type 1 diabetics. Insulin is the primary treatment, it works well, and patients can objectively measure their own blood sugar. Type 1 diabetics don’t seem to experiment with supplements that might alter their blood sugars. Those patients end up hospitalized or dead.
It’s the patients with type 2 diabetes that lean towards supplements. While lifestyle modifications are the foundation for managing diabetes (exercise, weight loss, and smoking cessation), drug treatment is usually also required. There are an array of prescription drugs like metformin and glyburide with a long history of use and demonstrated efficacy. Some drugs even decrease mortality – the primary outcome we’re after. But proper treatment has also been shown to the reduce the risk of an array of other consequences: Diabetes is the biggest cause of blindness, kidney failure and non-traumatic amputation. Diabetes is associated with an elevated risk of cardiovascular disease, too. Yet despite the irreversible consequences of diabetes, and the availability of effective medications, type 2 diabetes remains poorly-controlled in many, often because of poor self-management.
The diagnosis of diabetes, and the effectiveness of treatments can be objectively measured. Fasting plasma glucose (FPG) measurements and then the oral glucose tolerance test accurately measure insulin function, and guide diagnosis. While routine blood sugar monitoring (with test strips) is generally unnecessary in Type 2 diabetes, measurement gives a point estimate of blood sugar levels. Glyclated hemoglobin (A1C) levels reflect overall blood sugar trends, with higher levels associated with more complications of the disease. Interestingly, super-intensive blood glucose lowering isn’t associated with additional risk reduction, and it increases the risk of side effects due to too-low blood sugar. Treatment goals are individualized (hey, it’s “holistic”), balancing a number of factors including risks as well as a patient’s ability to manage complex treatment plans.
The Evidence for Cinnamon
Given the consequences of diabetes, self-management is something I want to encourage, not discourage. Without a commitment from the patient to take an active role in managing their diabetes, any treatment plan is doomed to fail. So is self-treatment with supplements a wise idea? There’s an array available, and patients regularly ask about the latest treatment “Big Pharma doesn’t want you to know about”. That treatment used to be chromium. Ginseng was popular for a time, too. Fenugreek and bitter melon are used as well. The treatment that seems most popular now is cinnamon. Like any other herbal remedy, most sources will tell you that it’s been used for “thousands of years” as a medicinal herb. As a treatment for diabetes, I have my doubts. While reports of diabetes go back to 1552 BCE, the ability to effectively measure any diabetes treatment only goes back a few decades. Interest in cinnamon as a treatment seems to have started with in vitro tests but gained some plausibility in 2003, when a study from Alam Khan suggested several grams of cassia cinnamon per day could lower fasting blood glucose. Khan randomized Type 2 diabetes to 1g, 3g, or 6g of cinnamon for 40 days. All three groups improved their fasting blood glucose, and blood lipid levels, but there was no effect on A1C.
Like trials with any other supplement or herbal product, the primary question we must answer is “What exactly was studied?”. The cinnamon you have in your kitchen may be a single species of plant or a mix of different cultivars. Ceylon cinnamon (Cinnamommum verum) is more commonly found in the West. Cassia cinnamon (Cinnamomum aromaticum) is the version of cinnamon that’s been studied in trials. The chemical hydroxychalcone has been identified as a potential active ingredient, which is believed to modify the sensitivity of cells to insulin, enhancing their uptake. If that’s the true mechanism of action, then it would work in a manner similar to that of the drugs Avandia, Actos, and metformin (Glucophage). Given the active ingredient (or ingredients) have not yet been definitively isolated, the issue of studying cinnamon is problematic. There’s no way to assess the potency of any batch, which complicates any evaluation. And that may be a reason why the research with cinnamon is inconsistent and largely disappointing.
While the Khan study looked promising, supplementary studies have failed to consistently show beneficial effects. Vanschoonbeek gave 1.5g of cinnamon or placebo to postmenopausal women over 6 weeks. There was no effect reported on blood sugar or blood lipid levels. Baker’s 2008 meta-analysis identified 5 trials including the Khan and Vanschoonbeek studies and concluded the following:
In this meta-analysis of five randomized placebo-controlled trials, patients with type 1 or type 2 diabetes receiving cinnamon did not demonstrate statistically or clinically significant changes in A1C, FBG, or lipid parameters in comparison with subjects receiving placebo.
In that analysis, the Khan study looks like an outlier. More studies have emerged since then: Crawford in 2009 found cinnamon 1g per day reduced A1C levels compared to placebo. Suppapitiporn found no effect on any measure with 1.5g per day. Akilen, in 2010, found an effect with 2g per day. Another meta-analysis, published in 2012 and included 6 studies, concluded the opposite of Baker, and made positive conclusions:
A total of 6 clinical trials met the strict inclusion criteria and considered a total of 435 patients; follow up between 40 days–4 months, doses ranging from 1 g to 6 g per day. Meta-analysis of RCTs showed a significant decrease in mean HbA1c [0.09%; 95% CI was 0.04–0.14] and mean FPG [0.84 mmol/l; 95% CI was 0.66–1.02].
But look closer. The results may be statistically significant, but they’re not that impressive compared to medication. Cinnamon lowered A1C by 0.09%, versus the usual 1% with medication. Give A1c reflects overall glucose trends, cinnamon doesn’t look that impressive. Even at the extreme of the confidence interval, cinnamon has, at best, 10% of the efficacy of drug treatments. At worst, it’s completely ineffective.
Until the findings are reproduced consistently, and cinnamon has been show to provide a meaningful improvement in relevant measures, there is no persuasive evidence to suggest that cinnamon has potential as a useful treatment option. Drugs that work, work consistently and provide meaningful improvements in measures of the disease. Why doesn’t cinnamon work? There may be an active ingredient, but it’s present in low concentrations, and varies in content between the different batches of cinnamon used in the different trials. In that case, the active ingredient needs to be standardized and possibly isolated, which would make it a drug treatment. Or this could be yet another example of a supplement that looks promising in early studies, only to see the effect disappear as the trials get larger and control for bias more effectively.
If a drug treatment’s efficacy is questionable, the adverse event and safety profile is even more important. As a popular food additive, cinnamon seems safe when consumed at doses of a few grams per day. (1 teaspoon of the powder is about 4.75 grams). While the trials have been small and short in duration, no significant adverse events have been reported. It is Generally Recognised as Safe (GRAS), as a seasoning and flavoring. However, reversible liver damage has been reported with therapeutic use, due to coumarin, a chemical also present in Cassia cinnamon. Those with liver impairment or dysfunction may be at greater risk of harm. There are no published long-term studies with cinnamon that inform us whether chronic consumption of high doses is safe.
Cinnamon’s effectiveness as a treatment for diabetes has not been established. A prescription drug as ineffective as cinnamon likely wouldn’t pass FDA muster. Existing drug treatments for diabetes, on the other hand, are cheap, effective, and generally well tolerated. Compared to drug therapy, we don’t know if cinnamon can reduce the risk of mortality due to diabetes, or the progression to any of the other serious outcomes of diabetes. For my patients that insist on trying cinnamon, I’d caution them of the risks, and reinforce that cinnamon is no alternative for lifestyle changes and medication if necessary. It may be natural, sure, but that doesn’t mean it’s either safe or effective.