Placebo effect in blood sugar

[TomBAvoider]

Diabetic's blood-sugar levels affected by false ingredient labels

In a Harvard University test, type-2 diabetics' blood sugar levels were found to be more strongly affected by perceived sugar levels of a sweet drink than its actual sugar levels.

[Ron Put]

This is an interesting study, thanks.

I wonder if there was a relatively uniform effect among the participants, or if some were significantly more susceptible than others?

[Sibiriak]

Thanks.  I haven't read the actual study (I'm not that interested),  but  a few comments on the story caught my eye: 

 

Quote

Scientist

There is so much half-assed science in the world. No where in the report do they give the actual data or run the simple tests needed to really back up the claim.

Their main finding was supported by a nonlinear model which they chose because it gave them the answer they wanted: “we used the quadratic model, due to its AIC fit and because it matched our hypothesized pattern of blood glucose responses.” When not paired with a test of difference between the raw data, that should not be pass review.

Despite the claim that “Blood glucose levels increased in accordance with how much sugar participants believed they consumed rather than how much they actually consumed.” They tested all participants with the same amount of sugar, so there is no data at all on whether the psychological differences are trivially small compared to the actual difference of drinking the sugar or not.

 

 

Quote

RHD

I think it’s more likely that thinking the carbs were high, their bodies released stress hormones, which raises blood sugar. If you could do the test with people who were naive about blood sugar, I bet the numbers wouldn’t change as much. I notice my blood sugar going up on stressful work days, and it is consistently low all summer. (I’m a teacher with summers off work.)

 

 

Quote

ejeffrey

It sounds like they only tested that blood sugar levels were affected by perceived sugar levels so they didn’t compare the magnitude of the effect to a difference in actual sugar intake.

As a crude proxy you can look at the blood sugar level vs. time:

Participants given a sugary drink but told it was zero calorie had their blood sugar rise from 140 to 170. Those told it was high sugar rose to 185.

You can’t actually isolate this from the rest of the effects that might have caused all participants blood sugar to change throughout the experiment, but taken at face value this means that the actual sugar caused a 30 point increase, while the label caused only an additional 15 points above that.

I reiterate: this is not a particularly valid control, but it definitely does not support the conclusion that the label matters more than the contents.

 

 

 

Edited October 5, 2020 by Sibiriak

[TomBAvoider]

Well, I mean, obviously the placebo effect in general is not simply "magic", there is a mechanism of action in every case. If for example, you are stressed out because you think you are taking in excess sugar (and I would count myself in that category) and that stress causes the release of hormones which in turn spike your blood sugar levels, that's a mechanism of action. And we know that the expectation of a meal triggers physiological reactions - see the famous Pavlov's dog example. Perhaps expecting a sugar spike prepares the body in a similar way. All in all, it's seemingly far less surprising than may seem at first glance. YMMV.