Using cellphones and accelerometers to track food intake and energy expenditure in rural areas of developing countries
Measuring whether people in developing countries have enough nutritious food to eat is vital work. It warns policymakers when people are running out of food, and is critical to staving off malnutrition.
It’s also time consuming and expensive. Teams fan out, often to rural, isolated areas, to do in-person interviews at homes, a months-long task.
But there might be an easier, less expensive way to do it, said Patrick Webb, Alexander McFarlane Professor of Nutrition at the Friedman School, who focuses on food insecurity and programs that try to alleviate it.
He and his colleagues are testing whether cellphone use in remote areas of Nepal is a proxy for food insecurity. Cellphones are ubiquitous even in very poor countries: Because communications towers can be built even in outlying areas, many people who never had access to landlines have adopted cellphones..
The idea was a simple one. “What if cellphone usage were some reflection of the poverty level, which is often linked to the food security level?” Webb said. “If we were able to assess the ownership of cellphones, volume of calls, or money spent on recharging cellphone plans, could that be correlated with food security?”
Webb already had teams at twenty-one locations in rural Nepal doing the old-fashioned food security surveys with more than 4,000 households for the past three years. Now he is seeing how cellphone data matches up with that on-the-ground information, controlling for other parameters, such as distance to market and road density, which can affect poverty rates.
“The preliminary results are looking quite promising,” he said. “There does seem to be quite a strong correlation between less cellphone use and more food insecurity.”
On the flip side of nutritional intake is energy expenditure: the calories people burn as they till fields, gather firewood, and go about their daily lives. If people are getting more nutrition, but are expending even more energy than before, there’s still a problem. Measuring energy expenditure of rural workers has never been easy—and it’s rarely done even now. Researchers would need to follow people around, make note of all their daily tasks, and try to tally their energy use.
But Webb and a colleague had a better idea: How about modifying fitness trackers—often used in developed countries by people trying to lose weight—and converting them to track activity of rural workers?
A startup in Cambridge, England, devised such an accelerometer for them. Worn around the waist or hung around the neck, it measures a range of motions typically employed by rural workers: hoeing, pounding, carrying buckets of water. Thanks to a Tufts innovation grant, a study is currently under way in India, Ghana, and Nepal, following men and women at different times of year; researchers are also charting their food consumption.
So far, initial results suggest the device works, Webb said. “We’re also getting some interesting pictures of the differences between men and women in some households,” he said. As expected, women are more active, Webb said, “but that doesn’t mean that men aren’t doing work. There are unexpectedly high peaks of intense activity for the men, which seem to be when they are breaking soil and doing a lot of the hoeing and tilling. But they do less of that.”
If the accelerometers are refined as more data comes in, “it could become a new measurement tool for assessing if people are doing more work than they are able to cover for in their diets,” Webb said.
Taylor McNeil can be reached at taylor.mcneil@tufts.edu.
