Alert dogs for T1D show promise in detecting owners’ blood sugar changes

Trained “glycemia alert dogs” are more reliable in signaling out-of-range (OOR) blood sugar levels to their owners with type 1 diabetes than previously reported, according to a new study published in PLOS One.

Twenty-seven dogs who were trained at a single charity in the U.K. and paired with an owner with type 1 diabetes achieved a median sensitivity of 70 percent in detecting nearly 4,000 OOR episodes. The dogs were taught to exhibit some type of alerting behavior when they suspected hypoglycemia, such as licking their owner’s face or fetching their blood testing kit.

Unsurprisingly, the dogs were most effective at detecting hypoglycemia, the condition they were trained to respond to by the charity Medical Detection Dogs. Some of them picked up the ability to reliably alert to hyperglycemic episodes as well, because they may have been rewarded for these behaviors once they spent more time with their owners.

The median sensitivities for detecting hypo- and hyperglycemia were 83 percent and 67 percent, respectively. An average of 81 percent of the alerts occurred when their owners’ blood glucose was out of target range, with four of the dogs demonstrating a 100 percent positive predictive value. However, two of the dogs were deemed incorrect in more than half of their alerts.

“The large sample shows that the individual performance of dogs is variable, but overall their sensitivity and specificity to OOR episodes are better than previous studies suggest,” lead author Nicola J. Rooney, with the Bristol Veterinary School at the University of Bristol in the U.K., and colleagues wrote. “Results show that optimal performance of glycemic alert dogs depends not only on good initial and ongoing training, but also careful selection of dogs for the conditions in which they will be working.”

Canines who had been accredited for longer—therefore spending more time with their owners—and those who were family pets before receiving training appeared to be the most accurate at detecting high blood glucose events.

This finding, the authors noted, “suggests that hyperglycemic detection ability may be something which is facilitated by a more established dog-owner relationship.”

Also, dogs living with owners with more severe diabetes or quicker drops in glucose levels tended to perform with greater sensitivity. The authors hypothesized the charity may have placed its most promising pets with the most vulnerable owners, or the dogs may have learned to function better with a more dependent owner.  

Overall, Rooney et al. believe alert dogs can form one part of a type 1 diabetic’s much larger care plan, provided they are trained and monitored properly and additional evidence supports the findings from this study. In an ideal situation, these alert dogs could accurately signal to their owners when it’s time to check their blood sugar and spur them into quick corrective action.

“The possibility of some false negatives remains and further study with professional large-scale training establishments such as Medical Detection Dogs is required, using owner-independent measures to record both blood sugar and dog behavior,” the authors wrote.

""

Daniel joined TriMed’s Chicago editorial team in 2017 as a Cardiovascular Business writer. He previously worked as a writer for daily newspapers in North Dakota and Indiana.

Around the web

Several key trends were evident at the Radiological Society of North America 2024 meeting, including new CT and MR technology and evolving adoption of artificial intelligence.

Ron Blankstein, MD, professor of radiology, Harvard Medical School, explains the use of artificial intelligence to detect heart disease in non-cardiac CT exams.

Eleven medical societies have signed on to a consensus statement aimed at standardizing imaging for suspected cardiovascular infections.