Improving Maternal Health and Birth Outcomes Through FreshRx: A Food-Is-Medicine Intervention

Abstract

Purpose

Food insecurity has far-reaching consequences for health and well-being, especially during pregnancy and postpartum periods. This study examines a food-is-medicine approach that aimed to reduce food insecurity, maternal stress, depression, anxiety, preterm labor, and low birthweight.

Design

Pre-post interventional study of FreshRx: Nourishing Healthy Starts, a pregnancy focused food-is-medicine program led by a local hunger relief organization and obstetrics department.

Setting

St. Louis, Missouri, a Midwestern U.S. city with higher-than-average infant mortality, low birthweight, and preterm birth rates.

Sample

Participants (N = 125) recruited from a local obstetrics clinic had pregnancies earlier than 24 weeks gestation; spoke English; and were enrolled in Medicaid. At baseline, 67.0% reported very low food security and none reported high food security, while 34.7% indicated depressive symptoms.

Intervention

FreshRx included weekly deliveries of fresh food meal kits, nutrition counseling and education, care coordination, and supportive services.

Measures

18-Question U.S. Household Food Security Survey, Edinburgh Postnatal Depression Scale, birthweight, gestational age.

Analysis

Single arm pre-post analysis.

Results

Average gestational age of 38.2 weeks (n = 84) and birthweight of 6.7 pounds (n = 81) were higher than rates for the general population in the area. For study participants who completed a sixty-day post-partum assessment, 13% (n = 45) indicated maternal depression (P < .01).

Conclusion

Food-is-medicine interventions may be an efficient, effective, and equitable tool for improving birth and maternal health outcomes.

Purpose

Food insecurity is “a household-level economic and social condition of limited or uncertain access to adequate food” ¹ with wide-ranging consequences such as increased adverse health outcomes and healthcare costs. ² During pregnancy, food insecurity can increase maternal stress, depression, and anxiety^3,4 as well as preterm labor and low birthweights. ⁵ Notably, the U.S. ranks last against peer countries in low-birthweight and pre-term deliveries, ⁶ while Black and Hispanic families in the U.S. experience increased risk of food insecurity,^5,7 less access to prenatal care, lower birthweight infants, and higher infant mortality. ⁸ Research also indicates that increased nutrition reduces adverse birth outcomes and chronic disease risk later in life, ⁹ making food security particularly critical during pregnancy. While households with lower income levels may qualify for Federal Nutrition Assistance Program support, food insecurity persists for pregnant individuals with recent studies suggesting rates ranging from comparable to the general population (12.8%) to as high as 69% within certain groups.^10,11 This research brief presents results from an innovative food-is-medicine intervention that integrated fresh food provision and counseling during pregnancy alongside traditional medical treatment within a local healthcare system.

Methods

Design

FreshRx: Nourishing Healthy Starts (FreshRx NHS) is a food-is-medicine delivery program developed by Operation Food Search, a hunger relief organization in St. Louis, Missouri. FreshRx NHS addresses the cyclical relationship between food insecurity and chronic disease—in which chronic disease incidence can increase food insecurity and food insecurity can increase risks of chronic disease ¹² —by providing nutrition support during pregnancy. Program staff formed and facilitated a participant advisory council that met monthly to advise on program design, guided food options, and reviewed study materials and assessment tools. The study was approved by the Washington University in St. Louis Institutional Review Board.

A baseline and sixty-day post-partum assessment collected sociodemographic and health information; household food insecurity; maternal mental health; participation in Federal Nutrition Assistance Programs; and other home cooking and nutritional information. Participants received a $25 gift card for completing each program assessment.

Sample

Staff at a local obstetrics care center that provides services regardless of an individual’s ability to pay screened for program eligibility during routine prenatal care visits using the validated Hunger Vital Sign™ two-item screening assessment. ¹³ Eligible participants needed to be pregnant; earlier than 24 weeks gestation; receiving care throughout pregnancy; English speaking; older than 14 years of age, and a Missouri resident enrolled in one of three state Medicaid Plans. Additionally, eligible participants resided in a pre-determined area with proximity to a central location to facilitate food deliveries and pickups. Overall, 125 participants enrolled in the FreshRx NHS program between September 2018 and July 2021.

Measures

While baseline food insecurity and maternal depressive symptoms were collected for all enrolled participants (N = 125), data collection including the completion of validated measures post-intervention varied. This included the 18-question USDA Household Food Security Survey Module (n = 21), ¹⁴ which measured food insecurity, and the Edinburgh Postnatal Depression Scale (EPDS) (n = 45), which assessed for maternal depressive indicators.^15,16 Birthweight (n = 81), gestational age (n = 84), and NICU days (n = 33) were collected via Electronic Health Records.

Intervention

Cognizant of common food prescription program barriers around transportation and nutrition literacy, ¹⁷ FreshRx NHS provided fresh food, cooking and nutrition supports, counseling, and increased case management during pregnancy through 6 weeks post-partum. Following review of informed consent, baseline assessment, and program orientation, participants received weekly meal kits to prepare three meals for their household focused on fresh, seasonal, and locally available foods. Participants who missed a delivery could request an additional drop off or opt to receive the following week as scheduled. Participants received meal preparation essentials such as cooking oil and other pantry staples, and program staff provided mental health services, cooking and nutrition education, and connected participants with resources to navigate cooking barriers such as malfunctioning appliances, disconnected utilities, and pest mitigation. Appendix A includes additional detail about the intervention and program design.

Analysis

For describing sample characteristics, we include all 125 participants regardless of attrition or missing data. Outcome variable sample size varies as our single arm pre-post results only include those for whom we have data for each outcome measure. We additionally obtained local birthweight and gestational age rates through the Missouri Department of Health and the Missouri Information for Community Assessment (MICA) system.

Results

We found decreased maternal depression and lower rates of pre-term and low birthweight infants. Participant demographics reflect the geographic area in which the program focused. 43.9% of participants were 16-24 years old, 46.3% were 25-34, and 9.8% were 35-44. Most participants identified as Black (77.4%) and 16.9% identified as White. At baseline, participants had high rates of very low (67.0%) and low (27.8%) household food security, with less observed marginal food security (5.2%) and no participants indicating high food security (n = 115).

To examine participant birth outcomes, we used publicly available data for St. Louis Medicaid recipients. We defined low birthweight as having a baby below 2500 grams (5.51 pounds) and pre-term births as earlier than 37 weeks as described by the World Health Organization and U.S. Centers for Disease Control and Prevention. Study participant gestational age at birth (mean = 38.2, median = 38.6, n = 84) and average birth weight (mean = 6.7, median = 6.7, n = 81) are considered normal or comparable to national rates in the U.S. (Figure 1). Compared to St. Louis citywide rates obtained from the Missouri Information for Community Assessment (MICA) system, program participant pre-term birth rate (12.0%) and low birthweight rate (9.9%) are lower than St. Louis city overall rates of 12.3% preterm and 13.0% low birthweight. Even of the participant pre-term births, the median gestational age was 36.3 weeks, and only five births required intensive care or time spent in the NICU. Notably, while overall rates are lower, Black mothers in the program still experienced a higher percent of low birthweight and pre-term births.

Figure 1.

Birthweight and gestational age at birth.

At baseline, 34.7% of participants (n = 121) demonstrated clinically significant depressive symptoms. Comparing participants for whom both pre and post-test data are available (n = 45), we observe a statistically significant (P < .01) decrease in depressive symptoms as 13.3% of participants 60 days post-partum had clinically significant depressive symptoms (Figure 2). Tables B1 and B2 in the Appendix summarize the study outcomes and provide additional detail on study participant characteristics.

Figure 2.

Maternal depressive symptoms for participants completing pre- and post-program assessments.

Discussion

Summary

Food insecurity persists despite an array of public nutrition assistance programs and community-based interventions. FreshRx NHS was the first program during pregnancy and post-partum periods to incorporate fresh food deliveries, counseling and mental health supports, and cooking and nutrition education. Participants experienced less depressive symptoms and fewer low birthweight and pre-term births compared to local rates, while previous food-is-medicine research has shown mixed results. One cluster randomized control trial of a food box delivery program found decreased household expenditures on food, however no sustained reduction of food insecurity. ¹⁸ A WIC-based fruit and vegetable voucher program during pregnancy was found to increase food security and lower the odds of preterm birth, ¹⁹ and a medically tailored meal program for adults with serious medical conditions lowered participant food insecurity and improved mental health. ²⁰ Early results from other programs, such as the Geisinger Fresh Food Farmacy focused on individuals with type 2 diabetes, estimates per-year-per-participant medical savings of $16,000 - $24,000. ²¹

Limitations

The FreshRx NHS study design and sample size did not allow for measuring isolated program components. Loss to follow-up also limited our analysis and findings for the program’s effect on food insecurity, and the lack of a control group limits our ability to make any causal claims. Furthermore, participants who opted into the program during pregnancy may not reflect those who opted out. Undoubtedly, the COVID-19 pandemic affected FreshRx NHS participant experience, study attrition, and data collection. However, among those who completed both baseline and post-partum assessments, observed study outcomes showed considerable improvement.

Significance

Improvements in maternal mental health and birth outcomes present immediate considerations for policymakers, practitioners, and researchers. While prevention accounts for a small share of U.S. healthcare spending, health insurance plans have incentive to increase access to effective services not traditionally considered Medicaid billable. Government investment has grown as the National Institutes of Health recently included food-is-medicine as part of its first strategic plan for nutrition research and USDA has directed $69 million for food prescription programs. Wider improvement of maternal and birth outcomes would generate significant immediate and long-term cost savings alongside meaningful improvements in perinatal and birth outcomes. Cross-sector partnerships in food-is-medicine interventions, especially during pregnancy, are ripe for replication and expansion while being cognizant of the FreshRx NHS experience with study attrition and limitations. The loss to follow-up for participants for food insecurity assessment suggests future consideration for using the recently validated eight-question Abbreviated Child and Adult Food Security Scale. ²²

“So What?” Implications for Health Promotion Practitioners and Researchers

What is Already Known on This Topic?

Food insecurity is associated with numerous adverse outcomes during the critical developmental period of pregnancy. Interventions such as medically tailored meals, fruit and vegetable vouchers, and food delivery programs have demonstrated mixed results across a number of target populations and outcomes of interest.

What Does This Article Add?

FreshRx NHS is distinct in its target population, comprehensive program model, and participant improvements in maternal mental health and birth outcomes. The program design also directly addresses research-identified barriers to successful food prescription programs such as transportation and nutrition literacy.

What Are the Implications for Health Promotion Practice or Research?

Programs such as FreshRx NHS have the potential to improve outcomes for mother and child and drastically reduce healthcare expenditures through the supported provision of a holistic, comprehensive food-is-medicine program. Scaling up FreshRx NHS and similar programs for greater impact will require cross-sector collaboration across levels of direct practice, policy, and research.

Appendix A. Detailed Program Overview

A1. Program Design

This study examines a food-is-medicine delivery program called FreshRx: Nourishing Healthy Starts (FreshRx NHS). The program and research design were developed through a community-academic partnership in St. Louis, Missouri—a Midwestern city with higher infant mortality, low birthweight, and preterm birth rates than regional and national averages. The program aims to address the cyclical relationship between food insecurity and chronic disease: chronic disease incidence can lead to increased food insecurity and vice-versa, increased food insecurity has been found to complicate if not exacerbate chronic disease conditions. In addition, the program aims to provide additional nutrition support during pregnancy for households who may face multiple policy, social, and cultural barriers to accessing nutrition assistance.

FreshRx NHS offers a holistic and comprehensive intervention to address household food insecurity during pregnancy to provide enhanced access to whole, nutrient-dense foods. While distinct in its study population, setting, and comprehensive program approach, the program builds on previous and ongoing food prescription or delivery program models. In partnership with an obstetrics department at a local hospital, the program provided fresh food, cooking and nutrition supports, counseling, and increased case management during pregnancy. The program utilized participant feedback to understand and accommodate a variety of barriers through a Participant Advisory Council. This group ensured that program participant agency was respected, food and nutrition options were culturally appropriate and appealing, and that the program design was appropriate to the circumstances of participants. In collaboration with the Participatory Advisory Council, researchers reviewed and assessed the informed consent and all study assessment tools for participant comprehension and contextual appropriateness using cognitive interviews with ten participants. Feedback informed the collection of the most pertinent information while minimizing unnecessary or burdensome participant assessment.

From enrollment through 60 days post-partum, participants received weekly food shares of locally sourced protein, dairy, fruits, and vegetables from a Combined Community Supported Agriculture provider. Weekly food share size varied to reflect larger or smaller households. Food shares were delivered to participants with a pickup option also available based on their preference. Initially, food shares were comprised of raw ingredients that participants could prepare following recipes provided by the program’s dietitian. However, following input from the Participant Advisory Council, the program started offering food through “meal kits” that made preparation much simpler.

Research has demonstrated a resource gap for families participating in SNAP of $45.69 per week to achieve food security. Based on these findings, FreshRx NHS aimed to provide approximately $50 of food per participant per week. Meal kits included ingredients to prepare three distinct meals with a focus on fresh, seasonal, and locally available produce, dairy products, and protein sources. Over 200 recipes were created by a trained chef and registered dietitian to address the nutritional needs of a pregnant person with input and feedback from the Participant Advisory Council. Approximately 150 recipes were revised and selected for use. In addition to providing basic cooking tools and pantry staples, the program linked participants with Federal Nutrition Assistance Programs (eg, SNAP and WIC) and locally available community resources.

During the program, participants received access to ongoing nutrition counseling and health coaching from a registered dietitian, care coordination and supportive services from a licensed social worker, and access to culinary nutrition education support from a trained chef. All services provided by program staff incorporated a participant-centered, trauma-informed model. Nutrition education and culinary skills were based on the needs of the participants and focused on adapting to changes in dietary requirements throughout the perinatal period. Support services provided by case managers were available to address participant crises, connect participants to relevant community resources, and provide mental health services. Supportive services utilized a combination of validated screening tools like the Edinburgh Postnatal Depression Scale and informal, conversational methods to accurately assess participant needs. Together, program components address a wide range of elements related to the experience of food insecurity through individual and household level supports that reflect and address social and environmental factors.

A2. Participant Eligibility & Enrollment

To meet program eligibility criteria, participants needed to be pregnant (earlier than 24 weeks gestation); receiving and planning to continue care throughout pregnancy; English speaking; older than 14 years of age, and a Missouri resident enrolled in one of three state Medicaid Plans. Additionally, eligible participants needed to have a primary address in one of fourteen pre-determined zip codes. The program’s geographic requirements were based on the prevalence of food insecurity in these zip codes, the presence of food deserts (determined using USDA Economic Research Service data), and proximity to a central facility that would allow food deliveries and pickups to take place within an approximately twenty-minute driving radius.

Potential participants who met the criteria were screened for food insecurity in an initial assessment conducted by their healthcare provider. During a first prenatal visit, staff at a local hospital’s obstetrics care center screened potential program participants for food insecurity using the validated Hunger Vital Sign™ two-item screening assessment. Eligible participants responded “often” or “sometimes” to either of the following two questions: [1] “Within the past 12 months, we worried whether our food would run out before we got money to buy more” or [2] “Within the past 12 months, the food we bought just didn’t last and we didn’t have money to get more.” The staff conducting the preliminary screener was trained to explain that all patients are asked about food access because of its importance to being healthy during pregnancy and referred individuals screening positive to the program through a phone call, text message, and/or email connection.

Following their referral, participants completed a program orientation and informed consent process. Next, a baseline assessment and program onboarding collected participant information including sociodemographic and health information (race and ethnicity, education level, household members, and information about current and previous pregnancies); household food insecurity; maternal mental health; participation in additional Federal Nutrition Assistance Programs (eg, SNAP and WIC); common childcare and household cooking items needed; and level of comfort and familiarity with basic cooking skills. Data were collected again at 60 days post-partum through participant assessments. Participants received a $25 gift card for completing each program assessment. Birth outcomes were collected when possible through the obstetric clinic’s Electronic Health Records. The Washington University in St. Louis Institutional Review Board (IRB) approved the study (ID #202011018).

Appendix B. Supplemental Results

In this section of the Appendix, we present two tables to provide additional detail on the study results. Table B1 outlines an array of key participant demographics at baseline, as well as their participation rates in WIC and SNAP, their baseline food insecurity and EPDS indicators, and their birth outcomes, including birthweight and gestational age at birth. Table B2 summarizes the changes in food insecurity and EPDS scores that we discuss in the main analysis.

Table B1.

Selected Study Indicators.

	Observations	%/Mean (SD)
Baseline Indicators
Race
Black/African American	96	77.42
White	21	16.94
Other	7	5.64
Age
16-24	54	43.9
25-34	57	46.34
35-44	12	9.76
Adults in household	84	2.19 (1.02)
Children in household	84	1.23 (1.24)
WIC Participation
Yes	68	56.67
No	52	43.33
SNAP participation
Yes	40	34.19
No	77	65.81
Food security
Marginal	6	5.22
Low	32	27.83
Very low	77	66.96
Clinically significant depressive symptoms (EPDS)
Yes	42	34.71
No	79	65.29
Birth Outcomes
Birthweight	81	6.67 (1.30)
Gestational age at birth	84	38.33 (1.91)

Note: Observations differ by indicator due to missing data for some participants.

Table B2.

Pre- and Post-program Outcome Summary.

	Observations	Pre-program (%)	Post-program (%)
Has clinically significant depressive symptoms (EPDS)	45	28.89	13.33
Food security	21
High		0	19.05
Marginal		4.76	14.29
Low		28.57	42.86
Very low		66.67	23.81

Note: These results only include study participants for whom there were both pre- and post-program data available. Observations differ by indicator due to missing data for some participants.

ORCID iDs

Dan Ferris https://orcid.org/0000-0003-2109-4483

Tyler Frank https://orcid.org/0000-0003-1878-423X

Ethical Statement

Ethical Approval

This project was reviewed and approved by the Washington University in St. Louis School of Medicine (IRB ID #202011018).