Why Americans Can’t Lose Weight
Analytics for Good — Ashley Osiason
Celebrity fad diets, expensive weight loss programs, and impossible fitness regimes by Instagram influencers all promising to drop 15 pounds in just 6 weeks: Why aren’t they working?
Let’s face it — every American, especially women, has at least once glanced in the mirror and had a negative thought about their body. Despite parading body icons like GiGi Hadid on Fashion Week runways in New York City, America has a major problem: the vast majority of people don’t have bodies that look like celebrities. Even worse, many don’t have bodies that are healthy. In fact, the U.S. has one of the highest adult obesity rates, standing at 42.4 percent.
In this project, I seek to answer what the real ‘secrets’ to weight loss are, but this time, using data. By using data to answer this question, we can better understand what lifestyle changes Americans need to make to live healthier, happier lives (and maybe feel more confident about their bodies in the process!).
Due to the challenge of de-identifying health and wellness data in the U.S., I selected an anonymized dataset that compiled eating habits, physical conditions, and lifestyles of individuals from Colombia, Peru, and Mexico. The data contains over 2000 entries of survey data from these individuals, and will be incredibly helpful for understanding key characteristics of overweight individuals.
What does it mean to be obese?
Before we begin our analysis, I wanted to provide a brief explanation of the definition of obesity. Obesity is often defined using Body Mass Index, a figure which adjusts weight by height to objectively determine if a patient is obese. According to the CDC and in our data:
If your BMI is less than 18.5, it falls within the underweight range.
If your BMI is 18.5 to <25, it falls within the normal.
If your BMI is 25.0 to 30, it falls within the overweight range.
If your BMI is 30.0 or higher, it falls within the obesity range.
Our data is further subdivided into categories, where a higher category reflects higher BMI. The graph below depicts the relationship between height and weight using our data:
Natural Factors that Influence Weight
I also wanted to preface this analysis by emphasizing our focus on attaining a healthy weight — not unrealistic body expectations. For example, age plays a large factor in weight, because metabolism slows over time. This relationship is portrayed below:
Now let’s get physical!
Physical activity is said to be a main driver of weight loss. The data strongly affirms this claim: active people are much more likely to have normal weight. As we can see from the chart below, averaging at least 1.5 days of physical activity per week is an incredibly common characteristic of non-overweight individuals.
This is not all that surprising. Plus, a lot of people don’t have the time or resources to go to the gym multiple times per week. However, what if we considered alternative ways to incorporate physical activity into our daily lives? The graph below studies the relationship between different methods of commuting to work.
Above, we can see that those who walk as their primary mode of transportation are 2 or even 3 times more likely to be normal or underweight. For those with longer commutes, bike riding is also a healthy option for incorporating physical activity into your daily routine. The proof is in the data: it is not necessary to sweat it out while bench pressing at 5am each morning at the gym. Getting fresh air on the way to work can help you lose those extra pounds.
However, during the Coronavirus pandemic, most people weren’t necessarily physically going to work due to the CDC’s social distancing guidelines. Does excessive screen time lead to unhealthy weight?
According to the data, not really! In fact, underweight and normal individuals were among some of the highest technology users, averaging about 4 hours daily. While this seems surprising, it’s important to note that younger people tend to use technology more, and younger people tend to have higher metabolisms. This trend is demonstrated in the graph below.
Regardless, we can take this trend as evidence that ditching your iPhone isn’t going to help you lose weight. Instead, keep it in your pocket so you can track your steps!
The Real Secret to Weight Loss: Be Born that Way.
Despite the tabloids that promise ‘easy’ weight loss, it is important to admit that some people’s struggles with obesity are much more challenging than others. I took a look at patients’ family histories. Unfortunately, there is a direct relationship between family history of obesity and being overweight. According to the data, every single Type II and III Obese patient had an obese family member. As well, over 90% of Level II Overweight and Type I Obese patients have obese family members. In contrast, less than half of normal and underweight patients had obese family history.
Is the cause of obesity is entirely genetic, or do some families have cultures and norms that promote healthier lifestyles?
For some people, family values may influence the likelihood of obesity, rather than genetic predisposition. For example, people with family history of obesity are four times less likely to count calories.
Relationship Between the Rest of the Variables
There are so many factors that influence weight. Below is a correlation plot between all of the variables considered in our data set. Red squares indicate an strong inverse relationship between two variables, while blue squares indicate a positive relationship between two variables.
As you can see, counting calories has an inverse relationship between weight and BMI, while frequent consumption of high calorie foods has a slight positive relationship with weight and BMI. This reflects the importance of diet in weight loss.