We started by asking whether places with more fast food had higher obesity. After 113 counties' worth of regressions, the answer was… not really. The pattern that did show up was about money. This is what it looks like across the fourteen counties of Massachusetts.
Last semester, in a MATH 345 statistics project, we drew a random sample of 113 U.S. counties across all five Census regions. The plan was to pair each county's fast-food restaurants per 1,000 residents against its adult obesity rate, fit a regression, and see if there was a clear upward trend.
But it actually didn't. The line we got basically showed no correlation. So we kept adding variables which were: median income, poverty rate, demographics. This drew us to a different conclusion. This essay redoes that finding inside Massachusetts and let's the users navigate the data themselves.
Small correlation and a big cloud of dots. Nantucket is the obvious outlier with high density, moderate obesity. The slope barely tilts, and slightly the wrong way. There is barely a story here.
A clear downward slope. Middlesex and Norfolk, the wealthy commuter counties are located at bottom right. Hampden and Berkshire are located at left. The relationship tells us the more obvious story.
In Massachusetts, Nantucket has the most fast-food per capita (1.58 / 1k) and obesity at 27%. Hampden has nearly the least (0.61 / 1k) and obesity at 30%. The amount of fast-foot locations are average across each county.
— p = 0.003 for poverty as a predictor in the multiple regression on 113 U.S. counties —
Switch layers below. Hover for numbers. Click any county to pin it Drag the slider to fade out counties below a poverty threshold.
Each dot is one of Massachusetts's fourteen counties. The dashed line is the fitted regression. Drag a box on the chart to highlight those counties on the map above.
Bristol, Worcester, and Hampden — three of the lowest-income counties — sit at the top of the obesity axis (32%, 31%, 30%). Norfolk, Middlesex, and Barnstable — three of the highest-income — sit at the bottom (22%, 23%, 24%).
Fast-food density, by contrast, varies almost independently of where a county lands on this chart. Nantucket has the most fast-food restaurants per capita in the state (1.58 / 1k), and roughly average obesity. Franklin has the least (0.46 / 1k), and obesity above the state mean.
Each line is one Massachusetts county's adult obesity rate, year by year, from 2011 to 2025. Click a line to isolate it. Toggle below to color by income tier the low-income counties tend to stick out.
Fast food is something that can be easily observed. Income, on the other hand, is more of a structural concept. It is simpler to pinpoint a location on a map for the former rather than the latter, which is one reason why the topic that surrounds obesity a lot of the time returns to the topic of fast food restaurants. However, the data which encompasses the 113 counties across the nation, as well as the fourteen counties in Massachusetts from 2011 to 2025 shows a contradictory trend.
None of this means that having access to healthy food doesn't matter; it does, and any truthful model would take it into account, as well as food deserts. But if you had to choose one factor to predict if a location has higher obesity rates, you'd choose its median household income over the number of fast-food locations. The map already shows where the income goes.