Daily Food Insecurity Predicts Lower Positive and Higher Negative Affect: An Ecological Momentary Assessment Study

Muzi Na, Nan Dou, Yujie Liao, Sara Jimenez Rincon, Lori A. Francis, Jennifer E. Graham‐Engeland, Laura E. Murray‐Kolb, Runze LiFrontiers in Nutrition 2022 · DOI

TL;DR

This study moves beyond viewing food insecurity as a static, long-term condition by using real-time smartphone tracking to show how it fluctuates daily. The researchers found that daily food insecurity significantly predicts lower positive mood and higher distress, with these emotional impacts intensifying during the latter half of the month as financial resources and SNAP benefits typically dwindle.

What Problem Does This Paper Address?

Most research on food insecurity (FI) treats it as a chronic state, often measured through retrospective surveys covering the past year. However, for low-income households, food access is often cyclical and volatile, tied to the timing of paychecks or the disbursement of SNAP benefits (food stamps).

The authors identify a critical gap: we don't fully understand the "micro-temporal" relationship between food stress and emotional well-being. Does experiencing food insecurity on a Tuesday immediately dampen your mood? Does this relationship change as the month progresses and the fridge gets emptier? Furthermore, the study examines how the unprecedented stress of the COVID-19 pandemic modified these daily emotional dynamics.

Methods

The researchers employed Ecological Momentary Assessment (EMA), a method that collects data in a participant's natural environment in real-time.

  • Participants: 29 low-income adults in central Pennsylvania were tracked over two 21-day "waves" (one in fall, one in winter).
  • Data Collection: Using smartphones, participants answered one morning survey about food insecurity (e.g., worrying about food running out, skipping meals) and were "pinged" five times throughout the day to rate their Positive Affect (PA) (e.g., happy, hopeful) and Negative Affect (NA) (e.g., lonely, annoyed).
  • Statistical Innovation: Instead of standard linear regressions, the team used Time-Varying-Effect-Models (TVEMs). This allowed them to see how the strength of the association between food insecurity and mood changed across the 21-day study period, rather than assuming the impact is the same on day 1 as it is on day 21.

Datasets

The study utilized data from the Food ‘N Mood pilot study, conducted between September 2019 and March 2021. This resulted in a high-density dataset of 902 person-days. The sample consisted of adults living below 185% of the Federal Poverty Line, providing a window into both pre-pandemic and mid-pandemic realities.

Key Insights and Ideas

  • The "End-of-Month" Vulnerability: The most striking insight is that the emotional toll of food insecurity isn't constant. The negative impact on mood was significantly stronger in the 3rd and 4th weeks of the month. This aligns with the "SNAP cycle," where benefits are often exhausted within the first 10 days, leading to increased psychological strain as the month ends.
  • Seasonal "Heat-or-Eat" Dynamics: In winter, the link between food insecurity and negative emotions (distress, anxiety) was more pronounced than in the fall. This suggests that competing costs—like high heating bills—exacerbate the mental health burden of food stress.
  • COVID-19 as a Sensitivity Booster: During the pandemic, the "buffer" seemed to wear thin. Unique patterns emerged where food insecurity began dragging down positive mood as early as the second week of the month, suggesting that the general instability of the pandemic made individuals more "affectively reactive" to food stressors.
  • Beyond Nutrition: The study reinforces that food insecurity is a psychological stressor, not just a nutritional one. The "worry" and "uncertainty" of food access act as daily "micro-stressors" that accumulate over time.

Key Results

  • Positive Affect (PA): Daily FI was associated with significantly lower PA in the 3rd and 4th weeks of both seasons. In fall, beta-coefficients for the 4th week ranged from -2.81 to -4.44, indicating a substantial drop in positive mood.
  • Negative Affect (NA): In winter, FI was significantly linked to higher NA (distress) during the second half of the month, with beta-coefficients ranging from 1.53 to 2.39.
  • Pandemic Impact: During COVID-19 days, the number of days where participants reported food insecurity increased dramatically (from 2 days/month pre-COVID to 10 days/month during COVID).

Limitations and Future Work

The authors acknowledge several limitations:

  • Sample Size: With only 29 participants, the study is a pilot. While the data is "dense" (many points per person), the results might not generalize to all low-income populations.
  • Demographics: The sample was predominantly female and relatively highly educated for their income bracket, which may influence coping mechanisms.
  • Directionality: While FI likely causes poor mood, the study cannot definitively rule out "reverse causality"—that people in a negative headspace might perceive or report food insecurity more acutely.

Future Work: The authors call for larger studies that include biological markers of stress (like cortisol) and explore whether interventions—such as shifting the timing of benefit disbursements—could flatten these emotional "dips" and improve long-term mental health outcomes for vulnerable populations.

Summary updated

Abstract

Food insecurity (FI) is a dynamic phenomenon, and its association with daily affect is unknown. We explored the association between daily FI and affect among low-income adults during a 2-seasonal-month period that covered days both pre- and during the COVID-19 pandemic. A total of 29 healthy low-income adults were recruited during fall in 2019 or 2020, 25 of whom were followed in winter in 2020 or 2021. Daily FI (measured once daily) and affect (measured 5 times daily) were collected over the 2nd-4th week in each month. Time-Varying-Effect-Models were used to estimate the association between daily FI and positive/negative affect (PA/NA). Overall, 902 person-days of daily-level data were collected. Daily FI was associated with lower PA in the 3rd and 4th week of fall and winter and with higher NA in the second half of winter months. Similar patterns of FI-affect relations were found pre- and during COVID-19 in the second half of a given month, while unique patterns of positive affect scores in the 2nd week and negative scores in the 1st week were only observed during COVID days. Our study supports a time-varying association between FI and affect in low-income adults. Future large studies are needed to verify the findings; ultimately, better understanding such associations may help identify, target, and intervene in food insecure adults to prevent adverse mental health outcomes.
BibTeX
@inproceedings{na2022daily,
  author = {Muzi Na and Nan Dou and Yujie Liao and Sara Jimenez Rincon and Lori A. Francis and Jennifer E. Graham‐Engeland and Laura E. Murray‐Kolb and Runze Li},
  title = {{Daily Food Insecurity Predicts Lower Positive and Higher Negative Affect: An Ecological Momentary Assessment Study}},
  booktitle = {Frontiers in Nutrition},
  year = {2022},
  doi = {10.3389/fnut.2022.790519},
}
Figure