Caffeine and Diabetes

Does caffeine make our blood glucose (BG) levels go up?

The popular news stories about caffeine and BGs can be confusing. Some say that caffeine is bad for BG control. Others say it's good. For example, heavy coffee drinkers are at lower risk of developing diabetes than light users or coffee abstainers.

What's going on here?

James D. Lane's recent review in a new science journal, the Journal of Caffeine Research, attempted to bring some order to these conflicting views by doing a meta-analysis of the various studies of caffeine and insulin resistance and BG control.

There does seem to be good evidence that high caffeine doses cause an increase in insulin resistance (IR) in healthy, nondiabetic adults. In a healthy, nondiabetic person, however, an increase in insulin resistance wouldn't necessarily mean higher BG levels. They'd just secrete more insulin to cover the increased IR.

When you have diabetes, however, and you can't just secrete more insulin, then an increase in IR usually results in an increase in BG levels. And, indeed, research has shown higher BG levels in people with type 2 diabetes after drinking coffee.

According to Lane, more than 17 studies in nondiabetic adults from 1968 through 2010 have demonstrated transient increases in IR with moderate caffeine doses (equivalent to 2 or 3 cups of brewed coffee) in both habitual caffeine consumers and abstainers. Different studies used different methods, but the results were consistent. Thirteen of 14 studies measuring the glucose response after a carbohydrate challenge found an increase in insulin resistance with caffeine. (Those who want more details can consult the references in the Lane paper, which is free full text online.)

Many studies used pure caffeine in amounts designed to replicate the amounts in coffee, and some of the caffeine doses were infused intravenously, hardly a physiological situation, although I've sometimes thought mainlining my coffee would get me going faster in the morning.

So Terry Graham and colleagues at the University of Guelph, Ontario, studied the effects of drinking coffee, to try to mimic the real-world effect of caffeine. They also studied the effect of fat. As reported here, they found that both fat and caffeine independently increased insulin resistance and glucose levels in the 10 young healthy men in the study. The caffeinated coffee had the greatest effect alone. Both together had an even greater effect than either one alone.

Both caffeinated and decaffeinated coffee increased levels of glucagon-like peptide-1 (GLP-1).

Graham explained that coffee alone won't raise BG levels. It only increases BG levels when you eat carbs. And even drinking the coffee at the same time you eat the carbs won't have much effect. But when you drink coffee, wait a bit, and then eat carbs, your BGs will go higher. The researchers had previously shown that this effect persists through a second meal and occurs with low-glycemic-index as well as high-glycemic-index carbohydrates.

Lane concluded from the meta-analysis that "Caffeine in coffee, tea, or soft drinks causes transient insulin resistance that can produce exaggerated glucose and insulin responses when carbohydrate is consumed" [italics mine].

This qualifier made me wonder what the effect would be in people on low-carb diets. The Graham group used 75 grams of carbohydrate in their test meals. Many people on low-carb diets eat only 30 to 50 carbohydrate grams a day. Would the caffeine be important in them?

To test this question in myself I adopted the following procedure.

1. Drink my usual 2 cups of strong espresso (or decaf another day) on arising.
2. Measure BG every hour until 2 hours after lunch.
3. Wait an hour after the coffee. Then eat 26 home roasted almonds.
4. Eat a poached egg with butter an hour after the almonds (so I'd have some protein to keep me until lunch) and take oral meds, including extended-release metformin.
5. Have lunch. I had 3 oz beef, 6 spears of asparagus with 1 tsp olive oil and a few slivers of red pepper, cracker-sized wedge of LC wrap with butter, half cup of plain kefir with a few nuts.

An hour after eating lunch and measuring, I took my standard walk, about 1.4 miles.

Here are the results:

Time__CAF__DECAF
0_____87____ 89
1_____85_____81___almonds
2_____104____86___ egg and ER metformin
3_____105___104
4_____*_____101
5_____88____108___lunch
6_____133____130___walk
7_____104____119

*I was absorbed in something else and forgot to test here.

At hour 2 (1 hour after eating nuts), I thought maybe the coffee was having an effect, but at hour 3 the readings were almost identical. Graham said that research has shown that caffeine doesn't affect gastric emptying, so this could be random variation.

Because my BG was higher before lunch on the decaf day than it was on the caffeine day, although the 1-hour postlunch BGs were similar, the rise was almost twice as much on the caffeine day.

Finally, perhaps the real peak might have occurred at 1.5 hours after eating the almonds and might have been higher with the caffeine (I'd planned to measure every 30 minutes, but then I got lazy).

But if so, it came down fast enough, so I concluded that for me and my diet, the effect of the caffeine was not large enough to persuade me to give up the pleasure I get from drinking black coffee.

Because we're all slightly different physiologically, with different sensitivities to chemicals and different diets, anyone who is concerned about the effects of caffeine should try a similar test themselves. If you want, you could let us know what you found.

The effect in someone on a low-fat high-carb diet might be quite different. And the effect on most Americans, who eat a lot of carbohydrate and fat as well as a lot of caffeine, especially in today's world where coffee houses are so popular, might be more significant.

The mechanism of the increased IR from caffeine is not known. There are two major hypotheses. The first is that it's because of interference with adenosine receptors on cells. Adenosine is involved in insulin-mediated glucose transport as well as playing a role in sleepiness.

One reason coffee keeps us awake is that blocking the adenosine receptors means adenosine can't bind to the receptors and make us sleepy. Blocking the receptors also seems to reduce inflammation and affect IR. Some research has shown that blocking the receptors decreases IR. Other research shows the opposite. And a study cited by Lane suggested that adenosine receptors had no effect.

The picture is further complicated by the fact that there are different subtypes of adenosine receptors on different cells, and the caffeine may have different effects in different tissues.

The other hypothesis is that caffeine increases the release of stress hormones like cortisol and epinephrine (adrenaline), counterregulatory hormones known to increase BG levels. There is some evidence for this effect, but more research needs to be done.

Finally, there's the paradoxical evidence that heavy coffee users are at much lower risk of getting type 2 diabetes. Those who drink 7 cups a day have half the risk of those who drink 2 or less.

Perhaps there's something in coffee other than caffeine that is causing this protection. In some studies, decaffeinated coffee reduced risk as much as caffeinated.

Common sense suggests that people who drink a lot of coffee probably drink fewer sodas.

But now a study at the University of California at Los Angeles, reported in January, suggests that a hormone-binding protein called SHBG (sex-hormone binding globulin) may be involved. The levels of SHBG in postmenopausal women were increased by caffeine; decaffeinated coffee and tea had no effect.

So far, this is just a hypothesis, but it's one more clue in this complex picture. When it comes to diet and type 2 diabetes control, nothing seems to be simple.