Research Report
Levi Berge-Wells, Nate Estrella
2023-03-24
Research Question
To what degree does the intensity of food insecurity affect the relationship with health outcomes in San Diego County?
Introduction
The importance of this research cannot be overstated, as it sheds light on a critical issue that affects many urban areas and has far-reaching consequences for the health and well-being of those who live in them. Food inequality is a persistent and pressing problem in urban areas, particularly in cities. San Diego county serves as a relevant case study, as it is classified as a food desert. By examining this specific area, we can gain a deeper understanding of how food shortages affect different groups of people, and who is most impacted by limited food access. By limiting the scope of our research to San Diego county, we can ensure that our findings are more targeted and specific. This will not only make our research more manageable but will also make it easier to draw meaningful conclusions. Furthermore, because everyone in the class lives in the region, it will foster more interesting discussions within our class. This report will hopefully shed light on just how prevalent the issue of food insecurity is within San Diego County.
Literature Review
Food insecurity is a growing public health issue worldwide. In many countries, including the United States, an increasing number of individuals and families experience limited or uncertain access to sufficient food for a healthy lifestyle. Food deserts have been an issue plaguing many parts of the United States, so there are fortunately many literary and analytical works to help bolster our study.
Beginning with works that cover the issue of food deserts here in southern California, sandiegocounty.gov gives insight into just how prevalent the issue of food security is. Here in San Diego, around forty percent of all San Diegans live over a mile away from the nearest supermarket. That may not sound like an extensive distance to travel. Still, when you consider the fact that about twenty percent of people live below the poverty line in San Diego, it becomes clear to see that an alarming amount of San Diegans have no method of transportation to reach these markets (Rhone. “Low-Income and Low-Foodstore-Access Census Tracts, 2015–19). As a result, nearly one in four people are nutrition insecure according to the San Diego Hunger Coalition in their article entitled “State of Hunger in San Diego County.” In addition to this, it is also stated by the same source that this has disproportionately affected those that are a part of lower SES groups as well as minority ethnic groups (Rhone. “Low-Income and Low-Foodstore-Access Census Tracts, 2015–19). Thanks to the rise of costs in the supermarket as well as the fact that the average cost of a trip to the grocery store is thirteen percent higher than the national average, it is becoming increasingly difficult for even one person to sustain themselves on a three meal a day diet (Josephine). The methods used to obtain the information surrounding food insecurity are a series of surveys conducted by the San Diego Hunger Coalition which were then analyzed to create a figure that represents the population of people that are food insecure along with the demographics of those that are food insecure.
A recent article titled “The Connection between Food Insecurity and Health Outcomes” sheds light on the complex relationship between food insecurity and health outcomes. The authors suggest that food insecurity can lead to poor dietary choices, resulting in malnutrition, which in turn can lead to the development of chronic diseases such as obesity, diabetes, and heart disease. The study’s authors argue that implementing a food insecurity screening and referral program within the University of California San Diego (UCSD) Student-run Free Clinic Project (SRFCP) program has the potential to make a significant impact on the health outcomes of low-income individuals in San Diego. The results of the study suggest that the implementation of a food insecurity screening and referral program can be an effective way to address these health problems. The program was successful in identifying patients who were experiencing food insecurity, and it was also able to connect them with community resources that could help address their needs. It is clear that more research is needed to identify effective strategies for addressing this issue, but this study provides a promising example of how healthcare providers can play a role in addressing food insecurity and its related health problems.
“Using Surveillance Data to Understand Overweight and Obesity in a County in California” published in the Journal of Public Health Management and Practice presents an exploratory study on the use of a community-based Body Mass Index (BMI) surveillance system in San Diego to combat the obesity epidemic in the local population. The article addresses important issues surrounding the prevalence of overweight and obesity, access to healthy foods and food insecurity, and the effectiveness of public health interventions in combating these issues. The researchers draw on data collected from the San Diego BMI Surveillance System, which is a tool designed to monitor and track rates of overweight and obesity at the local level. The study has shown that the current state of the BMI Surveillance System captures a representative sample of children and older adults, and it has the potential to become a robust tool to effectively combat obesity in the community. The article explores a range of strategies used to address the obesity epidemic in San Diego, including the Healthy Foods Farmers Market Fresh Fund Incentive Project and the Healthy Schools Summer Meals Program. The Fresh Fund project is aimed at improving access to and consumption of nutritious food among low-income populations, while the Summer Meals Program provides a much-needed service to children during the summer months when access to free and reduced-price school meals is limited. The article presents a comprehensive review of the literature on overweight and obesity in the United States, highlighting the growing prevalence of these conditions in both adults and children. The study emphasizes the need for public health strategies that address the complex relationship between socioeconomic factors, access to healthy foods, and food insecurity. The research has several potential limitations, including the small sample size of participants in the study, and the lack of follow-up data on participants in the Fresh Fund project. Despite these deficiencies, the study provides valuable insights into the critical role of community-based BMI surveillance systems in monitoring and tracking rates of overweight and obesity, and the importance of public health interventions that target the complex socioeconomic factors that contribute to these conditions.
The article “Food Insecurity Among Hispanic/Latino Youth” investigates how food insecurity is associated with a variety of health correlates. Researchers have identified a range of Despite the complexity of these factors, researchers have also found that addressing food insecurity among Hispanic/Latino youth can have significant positive impacts on health outcomes. This includes improvements in physical health, such as reduced rates of obesity and chronic disease, as well as improvements in mental health and overall well-being. factors that contribute to food insecurity among Hispanic/Latino youth. These include poverty, lack of access to healthy food options, and insufficient social support. These issues are compounded by other social determinants of health, including inadequate access to healthcare, limited education opportunities, and discrimination. Despite the complexity of these factors, researchers have also found that addressing food insecurity among Hispanic/Latino youth can have significant positive impacts on health outcomes. This includes improvements in physical health, such as reduced rates of obesity and chronic disease, as well as improvements in mental health and overall well-being.
One major obstacle came in the form of an idea rather than something physical like missing data or having to recode how the data is grouped (will be explained in more detail in Data and Methods). Instead, a theory was posed to us that explains that there is a social phenomenon that wealthy people tend to live further away from market centers as they tend to live in a more suburban area. This is a significant point as it could render our whole study obsolete by going against our definition of being food insecure. However, we believe that even in the face of this theory, we still have a strong argument and case study due to the fact that wealthy people are far less likely to suffer from serious ailments such as diabetes, hyperlipidemia, hypertension, overall heart failure, and so on. In an article entitled “Do wealth and inequality associate with health in a small-scale subsistence society,” the authors explain that wealthy people tend to live longer and have less life-threatening ailments due to the fact that they have money and access (Jaeggi et al.) By having money, it is inherently easier to have health coverage, access to nutritious foods, as well as avoiding health conditions (Jaeggi et al.). Using this information, we can infer that all data we are finding are more relevant to people that are actually food insecure and live far away. We decided that this inference is logical based on the fact that there are people that are food insecure based on accessibility as referenced in the data set we have on the intensity of food insecurity. Refraining from going further into detail so as to not shift the focus of our project, we decided this was an important point to talk about to further legitimize our research.
Data & Methods
In our project, we chose to analyze datasets surrounding health outcomes as well as look at datasets regarding food accessibility. More specifically, we gathered existing data on health conditions that are common in people with either poor diets or nutrition-lacking diets. These diagnoses include hypertension, hyperlipidemia, diabetes, and overall heart failure. This health data comes from the California Department of Health and was compilted by the San Diego County Health & Human Services Agency. We will speficially be uses the variable that represents the number of people who will die from the disseas out of one hundred thousand. The dataset provides a column with the overall rate and then another with the age-adjusted rate. We chose to use the age-adjusted data because we feel it will help to control for potential confounds. The data covering health outcomes are grouped by city, so we will use it to see where the highest rates of the aforementioned ailments are concentrated within San Diego County. In regards to the datasets surrounding food accessibility, we used data from the USDA food access research atlas that describes food insecurity. The variables we will be examining in the data set contain the proportion of people in a census tract that have low access to food. There are several variables with different definitions of low access to food, for example not having access to a grocery store within “1 mile for urban areas and 10 miles for rural areas”. The dataset also has variables that represent the proportion of low-income people who have low income to food, based on those same definitions. To investigate the relationship between the intensity of food insecurity and health outcomes we are comparing the data for food insecurity defined as living more than half a mile from a store to data for food insecurity defined as one mile or more from stores. The USDA food insecurity data is reported at the census tract level, and the San Diego County health data is labeled at the city/locality level, such as “La Jolla”. To merge these datasets for comparison we needed a way to connect the different geographic levels. To do this we used the HUD-USPS zip-census tract dataset. This dataset associates each census tract to its corresponding zip code. This tool allowed us to group our USDA food insecurity data by zip code. We then used random sampling to select twenty locations across San Diego County represented in the Health Outcomes datasets. We then used online tools to find the zip code that best covers the locality we chose. The locations are:
## Location zip
## 1 CHULA VISTA 91902
## 2 LA MESA 91941
## 3 CENTRAL SAN DIEGO 92101
## 4 DEL MAR 92014
## 5 ESCONDIDO 92025
## 6 EL CAJON 92020
## 7 SAN MARCOS 92078
## 8 POWAY 92064
## 9 KEARNY MESA 92123
## 10 VISTA 92083
## 11 CARLSBAD 92008
## 12 OCEANSIDE 92054
## 13 LAKESIDE 92040
## 14 IMPERIAL BEACH 91932
## 15 ENCINITAS 92024
## 16 SPRING VALLEY 91977
## 17 SANTEE 92071
## 18 ALPINE 91901
## 19 LEMON GROVE 91945
## 20 RAMONA 92065We have made sure that each dataset we collected pertains to different areas in San Diego county in hopes to have clear and accurate data that allows our analysis to paint a clear picture of what areas are most affected by food insecurity here in San Diego County. We chose these datasets in hopes to establish correlations between the intensity of food insecurity in an area and health outcomes. It is also important to note that all of the data we have collected have been gathered from official sources such as census reports and peer-reviewed studies.
One issue we ran into was the abundance of missing data. The missing data mostly came from the health outcome datasets. More specifically, the missing data was demographic specifics such as age range, gender, and race. Despite this, we decided use the data for. In addition to this, we originally wanted to include more measures of health within our study such as hyperlipidemia and more. However, we could not as we did not have fully complete data sets. This is what unfortunately occurred with the hyperlipidemia dataset, as the dataset was filled with NA values. With the scarce sample size provided by the hyperlipidemia dataset along with not having the necessary tools to go out and conduct research on hyperlipidemia rates in San Diego County, we just decided to drop the dataset from our study. Despite not having as many measures of health as we wanted, we are certain that we still have enough to find a relationship between the intensity of food insecurity and health outcomes.
As far as methods are concerned, to investigate the relationship between food insecurity and health outcomes we employed a linear regression line. The model we are using to fit the data is linear. We will be comparing the death rate of disease at two different level of food insecurity
Results
Before comparing the disease death rates at differing levels of food insecurity we first need to establish that there is a meaningful relationship between the independent and dependent variables. We generated these four linear regression models.
##
## Call:
## lm(formula = Rate_hypertension + Rate_heart_disease + Rate_diabetes ~
## lapophalfshare, data = locations)
##
## Residuals:
## Min 1Q Median 3Q Max
## -36.292 -12.595 1.237 15.985 25.095
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 141.3904 34.4557 4.104 0.00213 **
## lapophalfshare 0.3640 0.5266 0.691 0.50516
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 20.71 on 10 degrees of freedom
## (8 observations deleted due to missingness)
## Multiple R-squared: 0.0456, Adjusted R-squared: -0.04984
## F-statistic: 0.4778 on 1 and 10 DF, p-value: 0.5052##
## Call:
## lm(formula = Rate_hypertension + Rate_heart_disease + Rate_diabetes ~
## lapop1share, data = locations)
##
## Residuals:
## Min 1Q Median 3Q Max
## -26.532 -11.553 -5.624 13.072 31.404
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 145.5833 13.2027 11.027 6.44e-07 ***
## lapop1share 0.6165 0.3847 1.602 0.14
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.91 on 10 degrees of freedom
## (8 observations deleted due to missingness)
## Multiple R-squared: 0.2043, Adjusted R-squared: 0.1247
## F-statistic: 2.567 on 1 and 10 DF, p-value: 0.1402##
## Call:
## lm(formula = Rate_hypertension + Rate_heart_disease + Rate_diabetes ~
## lalowihalfshare, data = locations)
##
## Residuals:
## Min 1Q Median 3Q Max
## -26.221 -11.435 -1.723 12.089 36.690
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 138.651 20.332 6.819 4.63e-05 ***
## lalowihalfshare 1.667 1.243 1.341 0.21
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.52 on 10 degrees of freedom
## (8 observations deleted due to missingness)
## Multiple R-squared: 0.1524, Adjusted R-squared: 0.06762
## F-statistic: 1.798 on 1 and 10 DF, p-value: 0.2096##
## Call:
## lm(formula = Rate_hypertension + Rate_heart_disease + Rate_diabetes ~
## lalowi1share, data = locations)
##
## Residuals:
## Min 1Q Median 3Q Max
## -29.421 -14.538 -3.105 12.953 32.650
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 155.582 10.844 14.347 5.36e-08 ***
## lalowi1share 1.436 1.418 1.013 0.335
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 20.19 on 10 degrees of freedom
## (8 observations deleted due to missingness)
## Multiple R-squared: 0.09307, Adjusted R-squared: 0.002378
## F-statistic: 1.026 on 1 and 10 DF, p-value: 0.3349From the summaries it is clear that unfortunately our dataset does not produce a statistically significant relationship between any of the four response variables and disease death rates. We can see that the observed p-value for all of our cases is above even the 90% confidence interval.
## `geom_smooth()` using formula = 'y ~ x'
## `geom_smooth()` using formula = 'y ~ x'
## `geom_smooth()` using formula = 'y ~ x'
## `geom_smooth()` using formula = 'y ~ x'
We initially chose to consider multiple health conditions together because we hypothesised that it would strengthen the relationship and make the results clearer. But because we are not getting significant results with this method. We considered that potentially combining the three health conditions was confounding the results. But graphing the relationship between a single condition, Heart Disease, and each of the four response variables shows that there still is not meaningful relationship.
Conclusion
From these results we must conclude that we cannot establish a correlation between intensity of food-insecurity and health outcomes in San Diego County. We are still confident, based on the existing literature, that there is a relationship does exits, and that it does apply to San Diego. We hypothesis our inaccuracy came from a few main sources. Firstly, the health data suffered from the fact that some localities included have relatively small populations. Therefore they will only have a few deaths per year, due to each dissease. Secondly our process of collecting and compiling data introduced several steps for error to be introduced. This primarily comes from the fact that we manually matched zip codes to localities. We did our best to choose the zip code that best covers the geographical represented by the location in the health dataset. We belive a majority of the confounding is coming from the difference between the geography covered by the zip code and the city. We feel more extensive research based on our question is needed. Having access to health data reported by census tract will allow it to be directly matched to the food insecurity data to help eliminate the room for error in this process. But with census tract level health data there is still the problem of low instances of dissease in small localities. We see this as a fundamental challenge with preforming data analysis on such small populations with so many variables involved.
Work Cited
Jaeggi, Adrian V, et al. “Do Wealth and Inequality Associate with Health
in a Small-Scale Subsistence Society?” ELife, ELife Sciences
Publications, Ltd, 14 May 2021.
Josephine, Amelia. The Cost of Living in San Diego - Smartasset. https://smartasset.com/mortgage/the-cost-of-living-in-san-diego.
Mower, C. (2021, December 10). Food accessibility: San Diego County. ArcGIS StoryMaps. Retrieved from https://storymaps.arcgis.com/stories/5a8553a6ad304ca895596d67167b4074
Potochnick, Stephanie, et al. “Food Insecurity Among Hispanic/Latino Youth: Who Is at Risk and What Are the Health Correlates?” Journal of Adolescent Health, vol. 64, no. 5, 2019, pp. 631–39
Purushothaman, V. (n.d.). Health Equity Data. Community Health Statistics Unit. Retrieved March 24, 2023, from https://www.sandiegocounty.gov/hhsa/programs/phs/community_health_statistics/
Rhone, Alana. Low-Income and Low-Foodstore-Access Census Tracts, 2015–19. https://www.ers.usda.gov/webdocs/publications/104158/eib-236_summary.pdf?v=5236.2.
Ratigan, Amanda. Countywide BMI Surveillance and Community-Level Approaches to Improve Access to Nutritious Food Among Low-Income Residents in San Diego, California. University of California, San Diego, 2015.
Smith, Sunny, et al. “Implementation of a Food Insecurity Screening and Referral Program in Student-Run Free Clinics in San Diego, California.” Preventive Medicine Reports, vol. 5, no. C, 2016, pp. 134–39, https://doi.org/10.1016/j.pmedr.2016.12.007.