Thursday, November 03, 2011

Alcohol and Diabetes

Risk of death for diabetics has not been something that has gotten better and better over time, but I was not really pleased when I read this study some time back. It looked at records of people in Finland who had been diagnosed with type 1 diabetes between 1970 and 1999, below age 30. It then divided them into twelve groups. First, it sorted them into a group diagnosed at ages 0-14, and a group diagnosed ages 15-29, which it called early and late onset. It then divided each of those groups into six groups based on the year of diagnosis; diagnosed in 1970-1974, 1975-1979, 1980-1984, 1985-1989, 1990-1994,1995-1999. Then it looked at risk of death within various lengths of time from diagnosis in each group, hoping to see that the risk had gone down for those diagnosed in later years. And that is what it found- for those in the "early onset" group. That is not what it found for those in the later onset group. The risk of death doubled between the earliest time period group and the later.

I was somewhat reassured to find that the extra deaths could mostly be explained by something that I am not at risk of dying from. That something? Alcohol and drugs. The total number of deaths in the total groups were reassuringly small. However, when looking only at the twenty years after diagnosis, and looking only at those diagnosed in 1970-1989 (so that twenty years have passed), 83 out of 273 deaths in the older onset cohort and 63 out of 230 deaths in the younger cohort were due to alcohol or drugs. The rate increased over time in the late diagnosed group and decreased in the early diagnosed group- I suspect the decrease in the early diagnosed group is related to the percentage in that group that were diagnosed towards the younger end within the 0-14 age range (that is, there are more people diagnosed at age 2, who are not as likely to die from alcohol by age 22, and less people diagnosed age 10, who are more likely to be dying from alcohol by age 30).
So anyways: in the 20 years after diagnosis, the percentage of deaths from alcohol in the older cohort, who were all between ages 15 and 50 when they died, was very close to 25%. That's 1/4 of all deaths in that group.
In contrast, in Finnish people who died in 1999-2007 between ages 15 and 50, 15% of all deaths were from alcohol related causes- and the incidence of alcohol related deaths in Finns is on the rise, so I think it's safe to say that risk of death from alcohol is considerably higher if you are a young adult with type 1 diabetes. Source

It is possible that diabetics are simply more likely to drink than are other Finnish people, and maybe that's the reason they die of alcohol related causes more often. I don't think that's the whole reason though. I think the reason is that being drunk and diabetic is dangerous.

And that fits well with the warning label that is on all insulin boxes sold in the United States. The warning says: Do Not Drink Alcoholic Beverages While Taking This Medicine.

But it doesn't really jive with several very large studies done here in the US. This study quotes studies that point out that moderate drinking is associated with a lower risk of developing type 2 diabetes compared to both no drinking and heavy drinking (heavy drinking directly impairs pancreatic function) but this study itself looked at a huge group of diabetic adults, mostly type 2 but also type 1, and looked at how drinking correlated with A1c. It found A1cs of that were .5% lower (like the difference between an A1c of 9.2% and 8.7%) in those who had 2-3 drinks per day v those who never drank. However, this study had so few type 1s that we almost disappear from the data, and it definitely didn't look at death rates- it was a survey, meaning that you had to be alive to participate. It didn't even look at complication rates. However, similar studies have found lower complication rates (SLIGHTLY lower) in type 2 diabetics who consume alcohol.

The reason for this is actually not hard to find. Alcohol (in all sorts of people) impairs liver function for up to twelve hours. The liver releases sugars. Impairing the liver therefore reduces blood sugar for some time after alcohol consumption. When you drink with a meal, you may be reducing or even preventing your mealtime blood sugar spike.

Despite this potential long term benefit to drinking with diabetes, and despite the fact that most people with either type of diabetes who drink never have a health problem from drinking, there are some clear risks to drinking that are even more important for diabetics. Here they are:
-Drinking and being drunk makes it hard to make good decisions. Diabetes demands good decisions.
-Short term having the liver not release sugar can cause hypoglycemia for hours after drinking. Unfortunately, drinking makes it more likely that a person will become hypoglycemic and also decreases people's ability to feel hypoglycemia.
-Short term having he liver not release sugar causes a rise in ketone levels in all sorts of people. In type 1 diabetics this rise is more problematic, especially for those who run blood sugar high enough to put them at risk for ketoacidosis anyways.
-Long term risks of alcoholism are especially dangerous when you are diabetic. Although diabetes is the most common cause of ketoacidosis in the United States, alcohol withdrawal is the second most common cause. You also do not want to develop alcoholic liver or pancreas disease (OUCH).

The American Diabetes Association recommends that diabetics not drink heavily; frankly, nobody really should be drinking heavily.
I think the following things are the most important to remember around diabetes and drinking:
1. It's safe to abstain from drinking.
2. It's very possible to drink with diabetes without problems.
3. Drinking with diabetes may make you hypoglycemic for up to twelve hours later, and glucagon will not help if your liver is impaired.
4. You are less likely to feel your hypos if you are intoxicated, so check your blood sugar more, and maybe correct a teeny bit less.
5. Don't drink alone.
6. Don't drink enough to pass out; that way people can assume that if you are passed out, it's from diabetes.

1 comment:

Nathan said...

It's an interesting study about the Finns. I have been doing some reading about diet and I think I came across some data that showed that they are one of the people groups that had a higher death rate from cardiovascular heart disease despite not fitting the high-fat/ cholesterol hypothesis. I wonder if there is another source of inflammation that can account for their CHD levels.

Anyway, I'm a homebrewer and have been making my own alchoholic beverages for over 10 years. I agree with your statements. Moderation is absolutely important. My glucose control is fine with 1-2 or less drinks a day, but more than that has obvious negative effects. If I drink more than that there can be hypos overnight and/ or insulin resistance the next day.

I also find that the beers that I like have a significant amount of carbs, so having multiple in a session always results in hyperglycemia. Lately, my favorite has been a homebrewed dry hard cider which seems to have very little carbohydrate in it.