Cancer Risks from Insulin Revisited
Not long ago the news media were proclaiming that insulin was linked to cancer, as discussed here. Some headlines even suggested that insulin caused cancer. Needless to say, some patients were worried.
In fact, the original reports never said that insulin caused cancer, just that once cancer cells were formed, the insulin (especially glargine, or Lantus) might make them grow faster.
The cancer rates seem to be higher the more insulin a person uses. It is known that cancer rates are higher in people with type 2 diabetes, who have insulin resistance and hence need more insulin than normal, either more injected insulin or higher levels of insulin they produce themselves, or both. High blood glucose (BG) levels also seem to increase cancer rates.
It is known that cancer cell growth is stimulated by a hormone called insulin-like growth factor-1 or IGF-1. And insulin cross-reacts with the IGF-1 receptor. Some insulins cross-react with the IGF-1receptor more than others, and Lantus is one of those. Whether or not using Lantus is dangerous is controversial, as described in the Diabetes Health article cited above, because different studies have had different results.
Interestingly, the synthesis of a new synthetic insulin called a "smart" insulin was recently announced. This analog insulin uses zinc to hold the insulin in a configuration that makes it work like Lantus. But unlike Lantus, it binds much less strongly to the IGF-1 recepter. In fact, it binds to the IGF-1 receptor less than human insulin. It is not yet available on the market.
Now comes an even more relevant report saying that insulin use doesn't increase cancer rates and in fact reduces them. It's high blood glucose (BG) levels that increase the cancer rates, the researchers say.
Yang and colleagues studied 4623 Chinese type 2 diabetes patients in Hong Kong, which has a highly subsidized health care system. Public hospitals provide 80% of outpatient visits, especially for chronic diseases, and drugs are dispensed on site, which means the researchers could determine when each patient began insulin use by examining hospital databases.
The researchers used an ingenious "new insulin user" study design that excluded patients who were already using insulin at the beginning of the study. This meant that insulin effects that might have occurred before the study began wouldn't affect the results.
With the new-insulin-user design, when a patient started using insulin, the analysis began, and each patient was eventually matched with 2 patients who did not use insulin, with their analysis beginning at the same time as that of the insulin user.
Yang and colleagues found that cancer incidence was lower in the insulin users than in the patients who did not use insulin. The hemoglobin A1c levels, however, were associated with higher cancer rates. Death rates were similar in both groups.
What does all this mean for us? Well, it's not simple. First, this is only one study. The results need to be replicated.
Second, the statistical analysis used to come to these conclusions is complex and not easily understood by a nonstatistician. When you have many factors that may affect the result you're looking at (in this case, cancer), you try to control for other possible factors that could affect the results. In this case, there was prior evidence suggesting that insulin and high BG levels both increased and metformin decreased cancer rates. Other factors such as lipid levels, blood pressure, sex, alcohol use, length of diabetes, and use of other drugs could also affect the rates of cancer. The researchers used complex statistical analysis to try to isolate the factors of interest: insulin use, BG levels, and cancer rates.
Other caveats are that the types of insulin used were not specified, so the study shows only the effects of insulin in general, not any specific type of insulin. The study was limited to Chinese patients, and it's possible, albeit unlikely, that other groups might react differently.
Nevertheless, I think this study should calm the fears of people who are afraid to use insulin because it might increase cancer rates. Perhaps it does, but high BG levels may be even more dangerous. We know they can contribute to heart disease. So when it comes to a choice between higher BG levels without insulin and normal BG levels with insulin, I'd vote for the insulin.
If you have a lot of insulin resistance and require a lot of injected insulin, it might make sense to see if you can control with some type of insulin that doesn't bind strongly to the IGF-1 receptors, especially if you have a history or family history of cancer. But if that doesn't give you good control, then I'd vote for the best control, regardless of the insulin type.
Nothing in life is certain, and that includes whether or not we will get a disease. Having diabetes puts us at higher risk for a lot of things, but it's not an immediate death sentence. More and more evidence is accumulating that the better controlled our BG levels are, the less likely we are to get a myriad of complications.
If we can lose a lot of weight and get a lot of exercise and control our BG levels with no drugs at all, as David Mendosa has done, that's obviously the best solution. But not everyone can accomplish this goal. Then we need to use whatever drugs it takes, including insulin, to get those BG levels under control.