Diet Quality and COVID-19: Socioeconomics Matter
What’s happening? Nutrition and COVID-19 has been a frequent topic in the media, journals and in the PEN System in the last year and a half. Recently, researchers investigated the relationship between diet quality and COVID-19 risk and severity and its interaction with socioeconomic deprivation among close to 600,000 participants of the smartphone-based COVID-19 Symptom Study (1). The PEN Team thought that an analysis of this study was important to highlight the complexity of the socioeconomic gradient in this diet-health link.
What did the authors find?
The authors’ findings suggest that a high diet quality was associated with a lower risk of COVID-19 and severe COVID-19 (1). Diets were considered high quality if they included healthy, plant-based foods. While the results show an overall effect of diet quality on COVID-19 risk, the effect of socioeconomic deprivation was greater. Specifically, the attributable COVID-19 risk to diet score was 32% (95%CI 18 to 46) while the attributable risk to socioeconomic deprivation was 38% (95%CI 27 to 50) and the additional combined effect (interaction) was 30% (95%CI 2.1 to 57%).
Digging deeper into the study’s conclusionsThe authors’ conclusions need a closer look to unpack just how much of a protective effect against COVID-19 can be explained by a diet characterized by plant-based foods. As this study revealed, the complex interweaving of diet quality and the social determinants cannot be considered independent estimates of effect. Thus, it may be less about what one wants to eat and more about what is accessible, available and possible given one’s socioeconomic situation.
- Researchers often adjust away the effects of other variables without reporting how strongly each variable is associated with the outcome. By reporting the associations between COVID-19 risk with both their diet score and the intersection of socioeconomic deprivation, the complex relationship of risk factors and COVID-19 infection is better represented.
- Worse health outcomes can be predicted by an individual’s socioeconomic position, including education, income, race, unemployment, food insecurity and housing (2). This study’s findings fit with what is known, that social determinants account for 30-55% of health outcomes. The most successful way to improve health is through addressing long-standing inequities, “which requires action by all sectors and civil society" (2). By understanding how influential social determinants are to health outcomes, dietitians' efforts to help can be more supportive and better directed to assist.
- The study did not directly compare COVID-19 risk between those who consumed a plant-based diet versus those who consumed an omnivorous diet to arrive at their conclusions. There was no differentiation in whether those with a high scoring diet characterized by plant-based foods emphasized vegetarian or vegan dietary patterns, making it less clear which perceived benefits of eating plant-based may exist. With no comparison between diet groups, it becomes harder to imply causation and harder to understand how this information might be helpful.
- The Plant-based Diet Score was calculated based on a 27-item Leeds Short Form Food Frequency Questionnaire, relying on self-reported data collection (1), which is prone to measurement error and bias.
- Each daily serving of vegetables, fruit and whole grains reported consumed pre-pandemic (February 2020) were rated positively while each daily serving of foods considered less desirable were rated negatively (fruit juices, refined grains, potatoes, sugar-sweetened beverages, sweets and desserts, dairy foods, fish, eggs and meat) (1). Some foods, including nuts, animal fat, vegetable oils, tea and coffee were not asked (1). The developer of this diet score found that animal food intake “ranged from 5–6 servings/day in the highest [diet quality] decile to 3–4 servings/day in the lowest decile” (3). Therefore, while the diet score was reduced by animal product consumption, there was no requirement for animal product avoidance to achieve a high Plant-based Diet Score (1).
- This study relied on statistical differences based on a P-value of 0.05 as an indicator of important differences between groups (1). However, when a study has a large sample size, as this one did with over half a million participants, even small differences between groups are likely to be found to be statistically significant. To accurately interpret findings from a study with such a large sample size, it is helpful to not solely rely on statistical significance tests and to look at the magnitude of the effects. Both the diet score and socioeconomic deprivation were both statistically significant and found to be of important magnitudes.
What now?This study highlights the complexity of the association between diet, socioeconomic status and health outcomes. Dietitians are aware of the importance of healthy eating to support the immune system, the need to assess individuals who are marginalized for adequate food access and to advocate for changes that impact the social determinants of health of clients and populations. In doing so, the disparities in COVID-19 risk can better be addressed to support individual and population well-being.
See Additional Content/Resources:
- Merino J, Joshi AD, Nguyen LH, Leeming ER, Mazidi M, Drew DA, et al. Diet quality and risk and severity of COVID-19: a prospective cohort study. Gut. 2021 Sep 6;gutjnl-2021-325353. doi: 10.1136/gutjnl-2021-325353. Online ahead of print. Abstract available from: https://pubmed.ncbi.nlm.nih.gov/34489306/
- World Health Organization. Social determinants of health. [cited 2021 Sep 27]. Available from: https://www.who.int/health-topics/social-determinants-of-health#tab=tab_1
- Satija A, Bhupathiraju AN, Spiegelman D, Chiuve SE, Manson JE, Willett W, et al. Healthful and unhealthful plant-based diets and the risk of coronary heart disease in US adults. J Am Coll Cardiol. 2017 July 25; 70(4): 411–22. doi:10.1016/j.jacc.2017.05.047. Abstract available from: https://pubmed.ncbi.nlm.nih.gov/28728684/