Food insufficiency, SNAP status, and variability in weight and body composition: Longitudinal analysis of the National Health and Aging Trends Study Cohort 2012 - 2021

Muzi Na; Martin J Sliwinski; Penny M Kris-Etherton

doi:10.1016/j.ajcnut.2024.11.025

Food insufficiency, SNAP status, and variability in weight and body composition: Longitudinal analysis of the National Health and Aging Trends Study Cohort 2012 - 2021

Am J Clin Nutr. 2024 Nov 27:S0002-9165(24)01420-5. doi: 10.1016/j.ajcnut.2024.11.025. Online ahead of print.

Authors

Muzi Na¹, Martin J Sliwinski², Penny M Kris-Etherton³

Affiliations

¹ Department of Nutritional Sciences, Penn State College of Health and Human Development. 110 Chandlee Lab, University Park, PA 16802. Electronic address: muzi.na@psu.edu.
² Department of Human Development and Family Studies, Penn State College of Health and Human Development. 402 Biobehavioral Health Building, University Park, PA 16802; Center for Healthy Aging, Penn State College of Health and Human Development. 4th Floor Biobehavioral Health Building, University Park, PA 16802. Electronic address: mjs56@psu.edu.
³ Department of Nutritional Sciences, Penn State College of Health and Human Development. 110 Chandlee Lab, University Park, PA 16802. Electronic address: pmk3@psu.edu.

PMID: 39613031
DOI: 10.1016/j.ajcnut.2024.11.025

Abstract

Background: The impact of food insecurity and food assistance program on long-term body weight and composition is unclear.

Objective: To investigate the associations between baseline and duration of food insufficiency, the Supplemental Nutrition Assistance Program (SNAP) status, and variability in weight, BMI, and waist circumference (WC).

Methods: Data from 3,897 eligible Medicare beneficiaries aged ≥ 65 years recruited in the National Health and Aging Trends Study (2012-2021) were analyzed. At baseline, SNAP status was assessed. With repeated measurements over follow-up, baseline food insufficiency (FI) status, years with FI experience (nFI), three variability metrics for weight, BMI, and WC (i.e. standard deviation (SD), coefficient of variation (CV), root mean squared error (RMSE)) and four patterns (stable, loss, gain and cycling) were defined. Partial proportional-odds generalized ordered logit models and multinomial logistic regression models were fit to investigate the association between FI status, nFI, and SNAP status and quartiles of variability metrics and patterns, respectively.

Results: The median [IQR] follow-up years was 8 [4, 9]. Per one-unit increase in the nFI, older adults had significantly higher variability in BMI (OR: 1.08 - 1.10), weight (OR = 1.14-1.15), and waist circumference (OR = 1.11 - 1.27) by SD, CV, and RMSE. SNAP participants did not differ from eligible non-participants in any variability metrics. Older adults with FI at baseline were 2.72 times (95%CI: 1.32, 5.58) more likely to gain weight. Relative risk of weight loss (RRR = 1.29, 95%CI: 1.06, 1.56), gain (RRR = 1.34, 95%CI: 1.09, 1.65), or cycling (RRR = 1.27, 95%CI: 1.04, 1.21) increased per one-unit increase in nFI. SNAP participants did not differ in BMI, weight, or WC patterns from eligible non-participants.

Conclusions: Recurring food insecurity was associated with variability in older adults' weight and body composition. Additional interventions, beyond SNAP, are needed to prevent food insecurity and instability in body weight and composition.

Keywords: BMI; Food insecurity; Waist Circumference; food assistance; intrapersonal variability; older adults; weight change; weight pattern.