Statistics Hog-Tie Pig Farming to H1N1 Cases

By Alex Roslin
August 27, 2009
The Georgia Straight
[Click here for the story at the Straight website]

As hospitals brace for the coming flu season and a possible new surge of H1N1 cases, international data on the flu pandemic shows it has hit Canada worse than almost any other country.
And a close look at the data suggests that a key factor may be something that health authorities have largely overlooked: hog farming.
Canada had the sixth-highest number of H1N1 cases per capita and the fifth-highest per capita rate of H1N1 deaths of all 134 countries and dependencies that had reported flu cases to the World Health Organization as of July 6. (That’s the last date for reliable international comparisons, because the WHO advised countries in early July to stop reporting data on individual H1N1 cases.)
Canada’s H1N1 rate was almost 15 times the global average—23.7 lab-confirmed cases per 100,000 people, compared to an international average of 1.6 cases per 100,000, according to the WHO data. Canada’s per capita rate was double that of the U.S. and 2.5 times that of Mexico, where the pandemic is thought to have started.
Canada’s H1N1 death rate was 10 times the international average: 7.4 deaths per 10 million people, versus 0.7 globally.
It’s not clear why Canadian H1N1 rates are so high. One possibility is that Canadian medical authorities have simply sent more cases to labs for testing. But the data also suggests another possible factor: Canada’s high concentration of hog farms.
It just so happens that Canada has the world’s eighth-highest number of pigs per capita—almost 15 million pigs, or about one for every two Canadians. And an analysis of international flu data shows that H1N1 rates have strong correlations with hog farming.
In Mexico, where it probably all started, there was a moderate, statistically significant 46-percent correlation between confirmed per capita H1N1 cases in all of the country’s 32 states and its federal district and the number of pigs per capita in those states. That’s according to the data as of July 2, the date the Pan American Health Organization stopped publishing the breakdown of flu cases within countries of the Americas.
(Correlation measures the strength of the relationship between two groups of data. A correlation of 30 to 50 percent is generally considered to be moderate, 50 to 70 percent is strong, while 70 percent or higher is very strong.)
Yucatán was the Mexican state with the highest rate of H1N1 cases per capita: 92 per 100,000 people. It’s also one of the country’s hog-farming hubs, with the most pigs per capita of any state, more than one for every two people.
Argentina had the world’s highest per capita death rate from H1N1, with 15 deaths per 10 million people, or 20 times the world average of 0.7 deaths. In Argentina’s 24 provinces and its capital district, there was a 70-percent correlation between the per capita death rate and the ratio of pigs to people.
The Argentinean province that had the highest death rate was Santa Fe, with 130 H1N1 deaths per 10 million people. Santa Fe also happens to have Argentina’s highest ratio of pigs to people.
And those countries aren’t the only ones where there’s apparently a relationship between the pandemic and hog farming. Among the 39 countries and dependencies in the Americas that had reported H1N1 cases as of July 6, there was a 51-percent correlation between H1N1 cases per capita and the number of pigs per capita.
Globally, the 20 countries with the most pigs per capita had a per-capita H1N1 rate of 5.5 per 100,000—more than 3.3 times the international average of 1.6 cases. As well, their per capita death rate from H1N1 was 2.5 per 10 million, or more than triple the international average of 0.7.
“This is a very serious concern,” said Bob Martin, who headed the Washington, D.C.–based Pew Commission on Industrial Farm Animal Production, when told about the Georgia Straight’s data analysis. “It’s just another step in showing what serious impacts these large-scale swine operations can have.”
Martin’s commission released a study last year that said workers in large farms, and their neighbours, have high rates of asthma and other respiratory illnesses due to manure runoff and emissions like ammonia and fine-particle pollution. Respiratory illness makes people more vulnerable to H1N1, he said.
A high portion of H1N1 hospitalizations and deaths have occurred among people with an additional medical condition like asthma or a compromised immune system.
In an initial story in July, the Straight reported that strong correlations exist between per capita H1N1 rates and the number of pigs per person within B.C.’s five health regions and in each of Canada’s provinces.
As of July 8, Manitoba, the country’s hog-farming capital, with 2.4 pigs per person, had three times as many H1N1 hospitalizations per capita as the Canadian average and 3.7 times as many deaths per capita.
The international data puts the high Manitoba numbers into even starker perspective. Manitoba’s per-capita H1N1 rate, 65 per 100,000 people, was 40 times higher the international average and far worse than that of the country with the highest rate in the world, Chile, which had 44 cases per 100,000.
Manitoba’s death rate—41 per 10 million people—was 60 times the global average and nearly three times that of Argentina, the worst-hit country in the world in terms of deaths.
So far, Canadian public-health officials have said the flu pandemic is spreading mostly randomly, though they acknowledge it has hit some vulnerable populations harder, especially those with respiratory problems, aboriginal people, and pregnant women. Most scientists believe H1N1 originated on a huge Mexican factory pig farm, then spread between people around the world.
In Canadian aboriginal communities, H1N1 is thought to be worse because of poor health care and overcrowding. Indeed, the data confirms that Native people have been hit harder and need extra resources to deal with H1N1. The per capita number of H1N1 cases in each province had a very strong 87-percent correlation with the per capita number of aboriginal people.
That’s even higher than the 77-percent correlation between per capita H1N1 cases and the per capita number of pigs in the 10 provinces.
However, when it comes to more serious H1N1 cases that involved hospitalization and death, the correlations were stronger for hog farming. There was a 44-percent correlation between per capita H1N1 hospitalization rates and the number of aboriginal people per capita in each province, compared to a 72-percent correlation between hospitalization rates and the per capita number of pigs in each province.
H1N1 deaths per capita had an 82-percent correlation with the percentage of aboriginal people in each province, but had an even stronger 89-percent correlation with the number of pigs per capita.
“I hope the World Health Organization will start looking at the same data you’re looking at,” the Pew Commission’s Martin said in a phone interview.