Pregnancy is complicated. What contributes to success?

The rise of data-driven analysis has given us insights into many aspects of life that were once more nuanced, and less data-driven. Whether that be sports(see: Moneyball), advertising (see: any advertisement on Instagram), or groups of friends. One of the areas of life that has quickly become the most data-driven is personal health. While doctors and hospitals and other large institutions have had access to data about health for a much longer time, the ability to quantify and identify trends in one’s own health has become much more common and accessible. Trends about metabolic rates, Vo2 Maxes, heart rates, and others are all now common knowledge to most people. This increase in data collection around health has also occurred at the state and county level, allowing the common citizen to identify trends using datasets that were previously out of reach. One of the most complex and important health possibilities has also profited from this increase in data: the process of pregnancy. Probably the most complex process that a person can go through, it also benefits the most from a complex analysis of data points, and how they affect the process. Some of the most important factors are nutrition and weight during pregnancy. These two factors, while vitally important, miss one other factor of physical health, that being exercise. Using nationwide health data from the CDC for the year of 2021, I will be conducting an analysis to see if better access to physical activity correlates to a lower incidence of newborns being underweight at birth.

Using the “Low birthweight raw value” and “Access to exercise opportunities raw value” variables, I plotted a scatter plot of their relationship:

As well as calculating their correlation, which came out to about -0.22, which is very much on the weaker side, yet still sufficient for showing a slight negative relationship. This means that increased opportunities for physical activity correlates to lower rates of underweight births. This relationship is also represented in different variables however. Variables related to physical fitness, and therefore related to access to physical activity, have a much stronger correlation than just access to physical activity. For example, there is another variable in the data, called “Poor physical health days”, which measures the amount of days that the population of this dataset would describe their physical health as poor. I created another scatter plot, and calculated the correlation of low birth weight with this variable as well:

The correlation between these two variables came out to .47, which I would call a moderate correlation, and much more strong than the correlation between access to physical activities and birth weight. A person’s physical health is obviously very important when pregnant, so this is a fairly obvious correlation. There are other obvious correlations, such as a correlation between adult obesity and low birth weight with a correlation of .26, so still very weak.

As I was conducting this analysis, I felt that I was retreading already known correlations. So I decided to dig a little deeper and look at another variable that, while still connected to my original premise, was a little tangential. In the dataset, there is a variable called “Food Environment Index” which is a scale from 1–10 that governments use to quantify proximity and income for healthy food in an area. Proximity could be how close someone is to a grocery store, while income is how much someone needs to make in an area to afford healthy food. This variable also has a moderate negative correlation to the rates of low birth weight:

The correlation value was about -.44. This was very interesting to me because it offers a couple of insights: The accessibility of healthy food is way more heavily correlated with healthier births, more so than rates of adult obesity, something that would feasibly be a direct result of having a lower food environment index. It also suggests that the kind of food a person eats while pregnant is much more important than physical activity, and about on par with the importance of the general physical health of the mother. A higher food environment index is also positively correlated with a higher access to physical exercise opportunities, which could be a muddying factor here:

Overall, I think that there is definitely a causal relationship between access to exercise opportunities and the incidence rate of babies being born underweight. Healthier mothers generally mean healthier babies. There are definitely other factors at play however, which seem to be even more important, such as strings of bad physical health, and the mother’s access to healthy foods. I feel all of these factors also hint at some systemic issues for mothers in our country, such as mothers from low income areas having access to neither of the variables shown to decrease the chance of low weight babies, and that’s something I want to explore further for my final project.

Thanks for reading!