Dogs can detect low blood sugar in diabetics?

[fingersmith]

I was walking through a shopping centre and noticed a sign taped to the wall(see attached blurry .jpg). It's about Pixie, a "diagnostic support dog", who will apparently become "agitated or upset" when it's diabetic owner needs sugar.
I'd never heard of this before and was sceptical, so I've tried to look up some information and haven't really come up with anything solid to confirm this.
Some things I have turned up:
BBC News article (http://news.bbc.co.uk/1/hi/northern_ireland/7107550.stm) stating Dr Deborah Wells has been given £10,000 for a study.
Her survey (http://www.psych.qub.ac.uk/surveys/diabetes/) mentioned in that article.
This general article (http://www.stanfordhospital.com/newsevents/newsletters/2008Newsletters/April08.htm) about it, which turned up a lot in some truncated incarnations.
Dogs4Diabetics (http://www.dogs4diabetics.com/) who claim to train dogs to do this.
I have to admit I don't have much experience with dogs or diabetics but it piqued my interest, anyone came across this before or have any observations?

[scepticalan]

When a person with diabetes suffers a hypoglycemic attack (low blood sugar) they often give off a strong chemical smell (acetone).

Nurses who deal with people with diabetes can usually smell when they have low blood sugar levels (if they're attentive enough) and i would logically extend this to dogs, as their sense of smell is much more acute. Perhaps a dog would be able to detect this sooner. A dog would also be around the diabetics most of the time and their detection may become refined like the nurses above.

I would assume that the dogs themselves would act as an alarm bell, prompting the care staff, or diabetics, to up the sugar levels and prevent an attack.

I haven't heard of this before, but was talking about the strong smell given off by low blood sugar attacks and the extension seems logical.

[fingersmith]

Thanks for the reply scepticalan, I'll look into that.

edit: the forum is playing games with me...

[fls]

When a person with diabetes suffers a hypoglycemic attack (low blood sugar) they often give off a strong chemical smell (acetone).

Nurses who deal with people with diabetes can usually smell when they have low blood sugar levels (if they're attentive enough) and i would logically extend this to dogs, as their sense of smell is much more acute. Perhaps a dog would be able to detect this sooner. A dog would also be around the diabetics most of the time and their detection may become refined like the nurses above.

I would assume that the dogs themselves would act as an alarm bell, prompting the care staff, or diabetics, to up the sugar levels and prevent an attack.

I haven't heard of this before, but was talking about the strong smell given off by low blood sugar attacks and the extension seems logical.

I think you are confusing ketoacidosis associated with high blood sugar with hypoglycemia (low blood sugar).

It could simply be a behavioural issue. People with hypoglycemia have specific changes in their behaviour and dogs are good at picking up non-verbal cues.

ETA: Oops, I should have read the article before responding. They said the same thing I just said. :)

Linda

[fingersmith]

When a person with diabetes suffers a hypoglycemic attack (low blood sugar) they often give off a strong chemical smell (acetone).

Nurses who deal with people with diabetes can usually smell when they have low blood sugar levels (if they're attentive enough) and i would logically extend this to dogs, as their sense of smell is much more acute. Perhaps a dog would be able to detect this sooner. A dog would also be around the diabetics most of the time and their detection may become refined like the nurses above.

I would assume that the dogs themselves would act as an alarm bell, prompting the care staff, or diabetics, to up the sugar levels and prevent an attack.

I haven't heard of this before, but was talking about the strong smell given off by low blood sugar attacks and the extension seems logical.

Can anyone clarify this for me? I've looked up symptoms of hypoglycemia and the lists don't seem to include acetone breath. Diabetes symptoms bring this up but not hypoglycemia specifically which is what is being claimed.

edit: linda clarified as I was posting, thanks.

[The Atheist]

Well, a dog can smell a bitch from some immense distance while we can't smell it from 6'.

Not quite sure how to put this, but diabetics definitely taste different to non-diabetics, so the idea that they could give off a smell noticeable by a dog isn't a great leap.

[Moochie]

When a person with diabetes suffers a hypoglycemic attack (low blood sugar) they often give off a strong chemical smell (acetone).

Nurses who deal with people with diabetes can usually smell when they have low blood sugar levels (if they're attentive enough) and i would logically extend this to dogs, as their sense of smell is much more acute. Perhaps a dog would be able to detect this sooner. A dog would also be around the diabetics most of the time and their detection may become refined like the nurses above.

I would assume that the dogs themselves would act as an alarm bell, prompting the care staff, or diabetics, to up the sugar levels and prevent an attack.

I haven't heard of this before, but was talking about the strong smell given off by low blood sugar attacks and the extension seems logical.

Now they should train the dog to go to the fridge and...


M.

[fingersmith]

It could simply be a behavioural issue. People with hypoglycemia have specific changes in their behaviour and dogs are good at picking up non-verbal cues.

This could be part of it, however less likely when they're talking about the owner being asleep.
Well, a dog can smell a bitch from some immense distance while we can't smell it from 6'.

Not quite sure how to put this, but diabetics definitely taste different to non-diabetics, so the idea that they could give off a smell noticeable by a dog isn't a great leap.

Well I can't say I've tasted a diabetic so I'll have to take your word for it! I suppose the question would be do they smell different the lower the blood sugar.

It doesn't seem too implausible that there could be something to this, however there doesn't yet appear to be any evidence it is fact. Unfortunately looking at the currently ongoing study mentioned above, I don't think a positive result would cut the mustard as proof. An online survey of 100 people seems like it's bound to attract people who believe their dog is reacting to this, wide open to confirmation bias.

[fls]

This could be part of it, however less likely when they're talking about the owner being asleep.

The story says the owner thinks he had a seizure in his sleep - hardly a subtle event. :)

Linda

[Rolfe]

Hypoglycaemia is a very strong stressor. It's used that way experimentally. I wouldn't be surprised if dogs could smell someone with high levels of adrenaline and cortisol due to the effects of these hormones on the body - especially the adrenaline. First I've heard of anyone actually harnessing such a response though. Better not take the dog to watch a marathon, I suspect.

Rolfe.

[fingersmith]

The story says the owner thinks he had a seizure in his sleep - hardly a subtle event. :)

Linda

Haha, yeah I thought that too, the amazing detection of a seizure was the catalyst for the guy to set up his organisation! I was actually thinking of various other bits and pieces I'd read, which I realised I haven't posted here, including this collection of anecdotes on the British Medical Journal site (http://www.bmj.com/cgi/content/full/321/7276/1565) which don't mention seizures.

[edd]

I think you are confusing ketoacidosis associated with high blood sugar with hypoglycemia (low blood sugar).

It could simply be a behavioural issue. People with hypoglycemia have specific changes in their behaviour and dogs are good at picking up non-verbal cues.

That's my thought. My mother likes to think she can spot my hypoglycaemia. What she can spot are the behavioural clues, which aren't exclusive to hypoglycaemia. Certainly when one's mother is nagging you.

I imagine a dog would pick this up fairly well, and the false positives as we all know just get ignored.

[The Atheist]

Well I can't say I've tasted a diabetic so I'll have to take your word for it! I suppose the question would be do they smell different the lower the blood sugar.

Rolfe makes a good point on that - it could be a smell, it could be a combination of smells.

It doesn't seem too implausible that there could be something to this, however there doesn't yet appear to be any evidence it is fact. Unfortunately looking at the currently ongoing study mentioned above, I don't think a positive result would cut the mustard as proof. An online survey of 100 people seems like it's bound to attract people who believe their dog is reacting to this, wide open to confirmation bias.

Quite right, especially where pets are concerned.

Hypoglycaemia is a very strong stressor. It's used that way experimentally. I wouldn't be surprised if dogs could smell someone with high levels of adrenaline and cortisol due to the effects of these hormones on the body - especially the adrenaline. First I've heard of anyone actually harnessing such a response though. Better not take the dog to watch a marathon, I suspect.

Rolfe.

That makes sense too.

Pity we can't ask the dogs.

[fls]

Haha, yeah I thought that too, the amazing detection of a seizure was the catalyst for the guy to set up his organisation! I was actually thinking of various other bits and pieces I'd read, which I realised I haven't posted here, including this collection of anecdotes on the British Medical Journal site (http://www.bmj.com/cgi/content/full/321/7276/1565) which don't mention seizures.

I feel compelled to point out that that is published in the Christmas issue (http://www.bmj.com/cgi/content/full/321/7276/0) of the BMJ. It's possible that the stories are real, but every year several medical journals devote their Christmas (or holiday) issue to publishing strange, silly, and made-up research. There are several clues including the 'signatures' of the authors. :)

ETA: I agree that it's still plausible for dogs to detect changes in sleeping people considering some of the possibilities mentioned in the article and by Rolfe.

Linda

[Dogdoctor]

My best guess at this would be if the dogs can detect anything it would likely be the change in behavior that the owner goes through when they become hypoglycemic. I don't see the usefulness in this type of service dog since the owner will likely be able to feel when their blood sugar is low so the dog will only serve to confirm what the diabetic person already knows. Certainly a quick blood test performed by the diabetic person would be more accurate.

[The Atheist]

I don't see the usefulness in this type of service dog since the owner will likely be able to feel when their blood sugar is low so the dog will only serve to confirm what the diabetic person already knows. Certainly a quick blood test performed by the diabetic person would be more accurate.

Good point - the diabetics I know would be awake to it before a dog told them

As Linda said, it could be a joke too. Wouldn't be the first one involving dogs.