This year, a report from Centers for Disease Control and Prevention revealed that America’s obesity rate has reached a record high. New Yorkers are not so safe from the obesity epidemic, as more than half of adult New Yorkers are either overweight or obese. Studies show that the rise in the obesity epidemic is partly due to disparities in food environment; it is harder for some to eat healthier because their options are limited.
Our project intends to look deeper into the relationship between food environment in NYC and obesity rate along with diabetes rate and stroke hospitalization rate.
We achieve the DOHMH New York City Restaurant Inspection Results data from NYC Open Data. The dataset contains information on restaurant name and location (zip codes), cuisine type, inspection date and inspection scores and grades.
The New York City Community Health Profiles is achieved from the NYC Department of Health and Mental Hygiene. It captures the health information, as well as the demographic description, for 59 community districts across the city. Our main outcomes of interest are Obesity (percent of adults who are obese (BMI of 30 or greater based on self-reported height and weight, age-adjusted), Diabetes (percent of adults who report ever being told by a healthcare professional that they have diabetes, age-adjusted) and Stroke (rate of hospitalizations due to stroke per 100,000 adults, age-adjusted)
We assess restaurant geographical distribution by calculating percentage of fastfood restaurants in each neighborhood. Multiple linear regession models are fitted between the three chronic disease outcomes (i.e. obesity, diabetes and stroke) and the percentage of fastfood restaurant for the 59 neighborhoods, adjusting for percentage of grade A restaurants in the neighborhood (only in the models for obesity and stroke), race, poverty, smoking status and exercise.
Model 1: Obesity = 31.8995111 + 60.8831247 * fastfood_percent + -29.3718425 * grade_percent + -0.0910361 * racewhite_rate + 0.0529318 * poverty + 0.7596112 * smoking + -0.2303902 * exercise
Model 2: Diabetes = 17.5930386 + 17.7244617 * fastfood_percent + -0.0775925 * racewhite_rate + 0.0476476 * poverty + 0.1521791 * smoking + -0.1439686 * exercise
Model 3: Stroke_hosp = 125.8598204 + 580.4988399 * fastfood_percent + -413.0988299 * grade_percent + -1.3704923 * racewhite_rate + 1.5830542 * poverty + 6.576646 * smoking + 1.922242 * exercise
Every 10% increase in the number of fastfood restaurants is associated with 6.09% increase in the obesity prevalence (p= 0.00056211), 1.77% increase in the diabetes prevalence (p= 0.0029406) and additonal 58 stroke hospitalization cases per 100,000 adults (p= 0.0014128) for a neighborhood, while adjusting for race, poverty, smoking status, exercise and percent of grade A restaurants in the neighborhood.
There are significant relationships between chronic disease outcomes (i.e. obesity, diabetes, stroke) and the geographical distribution of fast-food restaurants in New York City.
The screencast of our project can be found here.
The Github repository of our project can be found here.