Medicine & health · the experiment
Read health in the baskets
51 million receipts, 95 municipalities, food pairings that change from the cities to the countryside. Four questions, from beginner to expert, on what shopping baskets reveal about heart and diabetes medicines.
1 · Find what varies
Beginner
According to the receipts, what mainly differs between Danish municipalities?
2 · Follow vegetables and cheese
Vegetables bought with cheese make up 1.34% of the co-purchasing backbone in the urban core, and 0.38% in the most outlying municipalities.
Intermediate
What fraction of its urban share does this pairing keep far from the cities?
3 · Measure what the network adds
For cardiovascular prescriptions, age, income and population size alone explain 84% of the differences between municipalities. Adding the network's cohesion raises this to 94%. The authors report what share of the differences left unexplained by age, income and population the network accounts for.
Expert
What is that share?
4 · Place ultra-processed foods
Baskets holding both a minimally processed and an ultra-processed item carry 76.6% of the food volume analysed. Yet in all 95 municipalities, the chance that an ultra-processed product has more partners than a minimally processed one is below 0.5 (mean 0.43).
Expert
What do these two facts together show?
Baskets decoded
—