Food Insecurity, Burnout, and Social Isolation Among Resident and Fellow Physicians.

Importance: Little is known about food insecurity (FI) among graduate medical education (GME) trainees or how FI might relate to well-being outcomes in this population. Objectives: To assess the prevalence of and factors associated with FI among GME trainees and to investigate the associations of FI with well-being outcomes such as burnout and social isolation in this population. Design, Setting, and Participants: In this cross-sectional study, a survey was distributed among 3408 resident and fellow physicians at 4 geographically distinct sites within 2 US academic institutions from May 2 to June 21, 2023. All resident and fellow physicians appointed at both institutions were eligible and were invited by email to participate. Main Outcomes and Measures: The primary outcome was prevalence of FI. Measurement instruments included a 2-item FI screening tool, a 2-item measure of burnout, items addressing intent to stay at one's institution, social isolation, and demographic characteristics. Univariable analysis and multivariable Poisson and linear regression were used to characterize prevalence and factors associated with FI and associations with well-being outcomes and intent to stay at the institution. Results: Of those eligible, 1656 residents and fellows participated in the survey (a response rate of 48.6%). Of these respondents, 735 of 1458 (50.4%) who indicated their gender identified as men and 365 of 1402 (26.0%) who indicated their parental status reported having children. Of the 1551 respondents who reported their age, 519 (33.5%) were younger than 30 years, 770 (49.6%) were aged 31 to 35 years, 169 (10.9%) were aged 36 to 40 years, and 53 (3.4%) were older than 40 years. Among 1457 respondents indicating their race and ethnicity, the majority were Asian (310 [21.3%]) or White (654 [44.9%]). The overall prevalence of FI was 13.7%. FI proportions differed across training sites, with higher levels in large metropolitan locations (15.6%, 17.0%, and 21.3% for the 3 sites in large metropolitan areas vs 4.5% for the site in a small metropolitan area; P < .001). There were also differences by postgraduate year (16.9% for postgraduate year 1 vs 10.1% for postgraduate year ≥5; P = .003) and race and ethnicity (22.4% for Black or African American trainees compared with 8.4% for White trainees; P = .04). Those with FI were more likely to experience burnout (adjusted relative risk, 1.37 [95% CI, 1.18-1.60]; P < .001), were less likely to consider remaining at their institution after training (adjusted relative risk, 0.81 [95% CI, 0.68-0.98]; P = .02), and had higher social isolation scores (T-score parameter estimate 2.37 [95% CI, 0.89-3.86]; P = .002). Conclusions and Relevance: In this cross-sectional study, nearly 1 in 7 GME trainees screened positive for FI. FI was associated with important well-being outcomes including burnout. These findings suggest that academic medical institutions should pursue systemic solutions to address FI among resident and fellow physicians as a means of supporting their well-being.