Recently, neurologists publishing in the journal Neurology have cast doubt on seizure alert dogs' (SADs) abilities to warn their owners of impending seizures. One report found that some epileptic owners of SADs were found to have normal brain wave activity during convulsive events. This meant that they had "psychogenic non-epileptic seizures" (PNES) as opposed to true epilepsy. The same author referenced another seizure-alert dog that supposedly didn't so much predict his owner's seizures as cause them by licking the owners face. Another report describes a Blue Heeler who warned his owner about PNES events by lying across her chest. Doubting Thomases, the authors suggest that information obtained from PNES patients about their dogs' behavior may be inaccurate. They justify there opinion noting that other data reported by PNES patients does not correlate with more objective measures. The authors question whether the seizure alert dogs may promote seizures by their very presence, as teddy bears present during EEG recording can trigger PNES. Strange, but apparently true. The question raised is do seizure alert dogs increase or decrease the frequency of PNES? A good question and one that could be fairly easily answered, I think. On a slightly more credulous note, the authors muse about what SADs might possibly be detecting, suggesting possibly a sound or odor.
Both reports cast doubt on trained (and untrained) dogs' abilities to anticipate their owners' seizures. "The current literature fails to support that canines can warn of impending seizures," one author writes. The benefit of owing a seizure alert dog may be a psychological one rather than a neurological benefit, the editorial suggests. One of the reports acknowledges that seizure-response dogs (SRDs) may be usefully trained to alert family members to an ongoing seizure and attract possibly lifesaving help. But the ability of SRDs is not in question. Neither is the fact that patients might benefit from the emotional support provided by a dog. Only the predictive ability of SADs is doubted by the scientists. Somewhere in the depths of one report, the authors refer to a study in which two dogs, on at least one occasion each, warned of a psychogenic event while under medical scrutiny. This observation is not acknowledged as a positive finding. The other report footnotes the author's later confirmation of a dog pacing prior to its owner having an epileptic seizure. Again, no comment was made. Using the black swan/white swan line of reasoning, this means that some dogs can sense seizures ahead of time. (If you see one black swan then you know black swans exist: It doesn't matter how many white swans you see, you can never prove black swans don't exist). The only things this leaves in question are how do dogs do it and how reliable are they at preemption?
I don't think it's not too much of a stretch of the imagination to believe that dogs, with their extraordinary sensory abilities, might be able to sense various impending events more quickly and more accurately than we (relatively) sensorily obtunded humans. In veterinary medicine, we often hear of thunderstorm phobic dogs who begin to pace and whine well before owners have the faintest notion that a storm is brewing. Dogs are also reported to be able to sense impending earthquakes, their owners' imminent return, and their owner's emotional state (just ask any owner). Just think of the weapons they have. Their olfactory abilities are second to none and are beyond our comprehension; they can hear the ultrasound range; and they are masters of observation and interpretation when it comes to body language. Dogs can practically see you sweat. An ABC news report a few years ago featured dogs trained to warn diabetic owners when they might were going to have a diabetic seizure. Low blood sugar is the culprit here. I have heard of this myself - first hand. A diabetic client of mine had a German shepherd who only ever approached her husband in one situation: when she was asleep and about to have a low blood sugar crisis. Also, I know a formerly seizure-prone dog trainer who's Keeshond warns her of impending epileptic events with such reliably that she no longer needs medication.
How might dogs accomplish this early warning? Using their superior sensory abilities, that's how. In the case of diabetics, affected individuals' metabolism changes prior to low blood sugar-induced seizures. The sweet smell of ketosis, which is like nail polish-remover, subtly accompanies this state. It stands to reason that wouldn't be too difficult for a dog to detect. The dog learns that when he smells that smell all hell breaks lose not long afterwards. Accordingly, he barks in alarm or licks his owner submissively when the odor occurs. Once owners appreciate that their dog displays this behavior before they have a seizure, it's case closed. The owner's behavioral change, mood change, sounds, or altered body scent could tip SADs off to impending PNES or epileptic seizures. No one is exactly sure what the trigger is in such cases. Probably not a sound cues, more likely scent, in my opinion.
One problem for the scientists is that bout-like behavior is tricky to study. Add the requirement for a cooperative canine, simultaneous video and EEG recording, a trigger you cannot identify, and you have situation that's hard to organize and replicate. Another difficulty is that learning in dogs tends to be location specific and may not transfer well from the home to the laboratory. Probably for these reasons, the reports in this issue of Neurology detail, somewhat negatively, another five cases of SADS in action while reviewing the limited literature on the subject. While the pundits sit on the fence pondering the next step, certain seizure-prone individuals can and do utilize SADs and they swear by them. Try telling them that they or their dogs are telling lies.