Each year, up to 30,000 confirmed cases of Lyme disease are reported to the U.S. government, and researchers estimate the true number of people diagnosed with the debilitating illness is at least 10 times that many. Still others go undiagnosed, as physicians have struggled to come up with a foolproof blood test for Lyme, which is transmitted by tick bites.
In search of better clues, scientists have started looking at an animal that gets bitten – and tested – far more often than people: the family dog.
During their annual visits to the veterinarian, millions of canines are tested for antibodies to the bacteria that cause Lyme disease, and a new study adds to evidence that these results may help predict where the disease is likely to crop up next in humans.
"The dog shows it early," said Shila K. Nordone, a North Carolina State University immunologist who was among the study authors.
Why? Because they go where the ticks are, said University of Georgia parasitologist Michael Yabsley, another author of the study, in the journal PLOS ONE.
"They're the ones that are outside, running around through the bushes," he said.
Over the last five years, for example, increasing numbers of dogs have tested positive in North Dakota and West Virginia, according to the study, and now human cases are starting to show up in those states as well.
In Pennsylvania, New Jersey, and other northeastern states where Lyme disease has been well established in humans for years, the numbers of positive tests in dogs were far higher.
In the Philadelphia area, Chester County canines topped the list in 2016, with 2,484 out of 14,290 – more than 17 percent of dogs – testing positive.
Nationally, among the 4,172,861 dogs for which results were available, 268,413 tested positive in 2016, a rate of 6.4 percent.
The data, compiled by the nonprofit Companion Animal Parasite Council, come from a blood test that signals exposure to Lyme and three other disease-causing agents, including heartworm. Most dogs still are tested only for heartworm, but the four-disease panel with Lyme is becoming more common, said the study authors, who also included mathematical modeler Christopher S. McMahan of Clemson University.
Among dogs that test positive for Lyme, most do not show symptoms, such as limping, and in such cases vets do not recommend treatment, said Katie McGonigle, a clinical assistant professor at the University of Pennsylvania's Ryan Veterinary Hospital.
She agreed that it makes sense to study dogs as a kind of early-warning system for human Lyme disease.
"The dogs are lower to the ground," McGonigle said. "The dogs aren't wearing clothes."
The idea of dogs as Lyme-disease sentinel has been around for a few years.
In 2011, researchers at the U.S. Centers for Disease Control and Prevention found that in counties with canine Lyme rates above 5 percent, the rate in people also tended to be higher than average (though at less than 1 percent, not nearly as high as that in dogs).
What's more, in some counties where this connection did not at first appear to hold true, it became that way. That is, these counties at first showed high rates of canine Lyme coupled with low rates of human disease, but after three years the human rates started to catch up – supporting the idea that the dogs could have predictive value.
Nordone said so far the link appears even "tighter" than what the CDC study found in 2011.
The new study, by Nordone, Yabsley, McMahan, and colleagues, described a model that can be used to forecast rates of positive Lyme tests in dogs. Among the factors that helped predict canine Lyme rates were surface water and forests, which are important habitat for deer – another animal that carries Lyme bacteria.
(Neither deer nor dogs transmit the disease to humans. The disease is transmitted when a tick bites one animal with the bacteria, then bites another.)
A key reason that it's useful to get an early warning of Lyme: Public health officials can start telling people to take preventive steps.
"It gives them a chance to educate early," Nordone said.