This secondary analysis will use the comprehensive nuMoM2b dataset from the National Institute of Child Health and Human Development (NICHD) to explore whether food systems with traditional Indigenous foods are associated with better/improved maternal health outcomes for Indigenous women in the U.S. Indigenous women are defined as those who self-identify as American Indian, Alaska Native, Native Hawaiian, or Other Pacific Islander. On average, pregnant Indigenous women experience high rates of gestational diabetes, hypertension, and postpartum complications. One way to improve maternal health outcomes is through a dietary intake of foods that have traditionally been part of an Indigenous diet as they tend to be nutritious, low in fat, and high in protein. This analysis will use the nuMoM2b dataset in a novel way since past research using this data has never focused on Indigenous Peoples as the main topic of study nor has it statistically modeled an Indigenous health solution.
This project hypothesizes that Indigenous women in their first pregnancy who partake of traditional Indigenous foods will have lower odds of gestational diabetes, hypertension, and postpartum complications. First-time pregnancies tend to be riskier than others, therefore analyzing possible solutions such as an Indigenous diet could help reduce this risk. This project will use only one dataset, NICHD’s nuMoM2b dataset, and will be accessed through the GREI repository, Elsevier’s Mendeley Data.
The 2010-2013 dataset contains 16 self-identified American Indian/Alaska Natives and 34 Native Hawaiian/Other Pacific Islanders, according to the dataset’s webpage. In addition, 563 multiracial women may fit the Indigenous identification to be included in the analysis. The team may consider expanding the sample to the Hispanic or general population for a more robust analysis that explores if an Indigenous diet benefits all pregnant women, regardless of race/ethnicity. Data will be handled in an ethical and non-discriminatory manner as findings will not be disaggregated by Tribal affiliation to support the Indigenous data sovereignty of Tribal nations and because that information is not in the dataset.
The research method will use multivariate logistic regression models to test the relationship between an Indigenous diet and maternal health outcomes. The binary dependent variables to individually test in each model will be instances of hypertension, gestational diabetes, and postpartum complications. The independent variables will be an Indigenous diet, age, marital status, education level, low income, physical activity level, and spirituality. The Indigenous diet variable will be created by narrowing 120 food and beverage items in the nuMoM2b dataset to those considered Indigenous foods and beverages such as organ meats, salmon, halibut, corn, squash, potatoes, and beans. Data is based on a participant’s diet for the three months preceding conception.
Data quality measures will ensure the most accurate statistical results. Classification bias will be checked and corrected through undersampling, oversampling, or synthetic data generation, depending on which results in the highest AUC. Variables will be tested for multicollinearity as measured by VIFs to ensure variables are not too similar to each other. Performance metrics include checking for overdispersion through cumulative chi-square density and an ANOVA chi-square test to check the overall effect.
Six-month timeline for project deliverables:
The outcomes of this project will consist of descriptive statistics about the population of Indigenous women having their first pregnancy and the statistical regression results of the association between an Indigenous diet and maternal health outcomes. Policy suggestions will be shared as part of the data findings, which will be written for publication in a scholarly journal to help further the academic research of this topic and population.
An additional stylized fact sheet with the key findings will be created to share with the general public, making it accessible and easy to read. The fact sheet will be shared through relevant networks at the University of Arizona such as the Native Nations Institute, and distributed to Tribal maternal health professionals, Tribal leaders, Tribal food sovereignty experts, and media outlets like Indian Country Today. The team will create an email list of individuals and organizations from these key audiences to distribute the fact sheet to them.
To adhere to the FAIR Principles, the project’s cleaned dataset, R code, and databook will be uploaded to Elsevier’s Mendeley Data, a GREI repository. This will make the project dataset easily findable, accessible, interoperable, and reproducible by other researchers who may want to further the research topic using this cleaned version of the dataset. For the analysis to be replicated by other researchers, the R code will feature comments in each step of the analysis and will use seed numbers where possible.
This project incorporates the CARE Principles by focusing on a health issue impacting Indigenous pregnant women that could be addressed through an Indigenous solution. The norm for secondary data analysis usually lacks a major focus on Indigenous Peoples and often reports on their shortcomings by comparing them to other groups. By focusing on what could strengthen their health rather than reporting on what is lacking or deficient, this project’s findings will provide a collective benefit for Indigenous Peoples by supporting Indigenous women in their first pregnancies. This project could provide an Indigenous solution that may improve their maternal health outcomes.
The findings will be made available and accessible to Tribal leaders, which supports their authority to control the use of data about their communities. The data will be relevant to their worldviews and can be used for effective self-governance in maternal health. By considering the role of Indigenous diets in maternal health, the project aims to fulfill the responsibility of producing findings that support Indigenous cultures and worldviews. The framing of the research provides an ethical approach to bringing visibility to Indigenous Peoples and using the data to benefit their collective rights and well-being.
As this study focuses on pregnant Indigenous women and their maternal health outcomes, the resultant findings will contribute to the scientific discipline of health and well-being of Indigenous Peoples. Few studies offer insights into American Indian, Alaska Native, Native Hawaiian, and Other Pacific Islander women’s experiences with pregnancies as it relates to maternal health outcomes. To our knowledge, there is not a study published specifically focused on the first pregnancy experiences of Indigenous women in the U.S. accompanied by a statistical analysis of the relationship between an Indigenous diet and maternal health outcomes.
Understanding factors that could improve an Indigenous mother’s health is important because it is a major predictor of a child’s health and well-being into adolescence and adulthood. Therefore, Indigenous communities will benefit from having more Indigenous mothers who are healthy as they influence the health of future generations. This project will contribute valuable knowledge about improving the health of Indigenous soon-to-be mothers that will benefit Indigenous communities in this way.
This project’s findings can help drive suggestions for prevention measures that promote Indigenous foods if better maternal health outcomes are found in Indigenous pregnant women as it relates to their experiences with gestational diabetes, hypertension, and postpartum complications. This information can demonstrate the importance for Tribal nations to exercise their Indigenous food sovereignty, which is an initiative that gives Indigenous communities the right to define their own food systems. If Indigenous foods are found to have a positive impact on maternal health outcomes, then it could motivate Tribal nations to build or continue their work on developing their own food systems. The findings may also provide more support for directing funding to Indigenous food sovereignty efforts. In turn, increased Indigenous dietary strategies could lower the burden of healthcare costs. This will help not only Indigenous pregnant women but their whole community.
Another benefit this project will provide is a focus on best practices for data reuse and secondary analysis by using a valuable dataset that has been around for several years but has yet to engage in an Indigenous-only focused study. As a leading example, this project will show how to be creative in data reuse and consider populations that may not have had their data analyzed even though it is available. The study itself will follow best practices for data reuse and report the analysis and cleaned data on Elsevier’s Mendeley Data so other researchers can use it to help further research on this population and health issue.
The team comprises Britnee Johnston and Breanna Lameman who are both Indigenous and have collaborated at the University of Arizona. They bring a combination of skills in statistical analysis, health research, food security, and Indigenous data sovereignty.
Britnee Johnston is Vietnamese and Blackfeet (Native American) and a former research analyst at the Native Nations Institute. She was a winner of NICHD’s 2021 data challenge where her statistical analysis demonstrated a relationship between discrimination and postpartum complications. She has worked for the Utah Data Research Center specializing in state administrative data using R, and used statistical analysis for her Master’s project to analyze the association of educational attainment and food assistance participation. She is a second-year PhD student in American Indian Studies at the University of Arizona.
Breanna Lameman is Diné from Shiprock, Navajo Nation. She is a second-year PhD student at the University of Arizona in the health behaviors health promotion program with a focus on Indigenous food, water, & energy systems (FEWS). Her life’s work and passion are grounded in the land, Diné cultural teachings, the Diné language, and lived experience. She is a farmer, community member, relative, and aunty who understands the interconnectedness and importance of her ancestral foods. Her research focus is on Indigenous food sovereignty, food security, hydroponics, environmental justice, and the nexus of Indigenous FEWS.
Internally as a team, Johnston and Lameman will meet weekly on updated progress and assign next steps. The team is willing to submit a brief monthly report to the data challenge administrators to share progress along with direct links to the analysis and written drafts at their present stage. This monthly report will hold the team accountable to ensure timely progress is made on the project to complete it in six months.
Responsibilities will be assigned using the Asana project management software to ensure project deliverables are completed according to the timeline. Johnston will lead the cleaning of the data and coding of the statistical analysis, while both Johnston and Lameman will work on the data interpretation and writing the findings into a scholarly article and public-facing fact sheet. Johnston and Lameman will lead the review and feedback from relevant scholars and experts and manage disseminating the fact sheet. Johnston will post the data into Elsevier’s Mendeley Data.
