Patterns in Food Insecurity During Pregnancy, 2004 to 2020

This survey study assesses patterns in food insecurity during pregnancy among individuals in 14 US states participating in the Pregnancy Risk Assessment Monitoring System from 2004 to 2020.


Introduction
Approximately 10% of US households are food insecure, meaning they experience limited or uncertain access to enough food for an active, healthy life. 1 Moreover, nearly 4% experience very low food security, meaning that household members have reduced or disrupted eating due to lack of money or other resources. Adults with food insecurity (FI) are most likely to be female and of reproductive age. 2 Food insecurity in pregnancy is understudied and may be associated with adverse pregnancy outcomes. 3 The US Department of Agriculture (USDA) FI surveillance report does not distinguish by pregnancy status. The last population-based FI estimates were limited to California from 2002 to 2006. 4 More recent population-based estimates on the burden of FI in pregnancy are lacking. We estimated population-based patterns of FI in pregnancy across 14 states from 2004 to 2020.

Methods
This survey study was approved by the University of Pennsylvania Institutional Review Board with a waiver of informed consent due to the use of deidentified data. We analyzed data from individuals 18 years or older who delivered infants between January 2004 and December 2020 within 14 states participating in the Pregnancy Risk Assessment Monitoring System (PRAMS) 5 (eMethods in Supplement 1). Participants were asked via questionnaire, "During the 12 months before your new baby was born, did you ever eat less than you felt you should because there wasn't enough money to buy food?" We estimated age-and state-adjusted yearly FI prevalence using the direct method, with data weighted to the 2020 distributions. We adjusted for age to account for any shifts in the age of pregnant people across the study period. We also adjusted for state because not all states consistently had data on FI across the study period or met the Centers for Disease Control and Prevention response rate threshold to be included in the data set. We tested for modification by age, race and ethnicity, and parity by including an interaction term with year. Data were weighted to account for the complex sampling design and nonresponse and to reflect population estimates.
Analysis was performed using SAS software, version 9.4 (SAS Institute). The significance threshold was 2-sided P = .05.

Results
Among 129 540 participants, most were ages 25 to 29 years (30.4%) and of non-Hispanic White race and ethnicity (68.3%). Prevalence of FI varied by individual characteristics and delivery year (Table).
The adjusted FI prevalence was 8.9% (95% CI, 7.6%-10.1%) in 2004 and 6.7% (95% CI, 6.0%-7.5%) in 2020; FI fluctuated across years, with lower prevalence in 2006 (7.9%; 95% CI, 6.8%-9.1%) and higher prevalence in 2013 (10.4%; 95% CI, 9.1%-11.6%) ( Figure). There were no interactions by age, race and ethnicity, or parity.    likely an underrepresentation of true FI prevalence in pregnancy given the question's focus on disrupted eating, which typically reflects more severe FI. 6 In 2020, national FI prevalence was 10.5%,  Estimates were adjusted for maternal age to account for any shifts in the age of pregnant people across the study period. Estimates were also adjusted for state because although the 14 participating states provided data on food insecurity, not all states consistently had data on food insecurity across the study period. All adjustments were completed using the direct method, with data weighted to the 2020 distributions. Shading represents 95% CIs.