Eating frequency is positively associated with overweight and central obesity in US adults

Abstract

Background: Evidence of the association between eating frequency (EF) and adiposity is inconsistent. Objective: With the use of data from the NHANES 2003-2012, this cross-sectional study examined the associations between EF, meal frequency (MF), and snack frequency (SF) and overweight/obesity and central obesity. Methods: Dietary intake was assessed with the use of two 24-h dietary recalls in 18,696 US adults ≥20 y of age. All eating occasions providing ≥50 kcal of energy were divided into meals or snacks on the basis of contribution to energy intake (≥15% or <15%), self-report, and time (0600-1000, 1200-1500, 1800-2100, or other). Multivariable logistic regression was used to compute ORs and 95% CIs. Results: When analyzed without adjustment for the ratio of energy intake to estimated energy requirement (EI:EER), all measures of EF, MF, and SF showed inverse or null associations. After adjustment for EI:EER, however, EF was positively associated with overweight/obesity (body mass index ≥25 kg/m2) and central obesity (waist circumference ≥102 cm in men and ≥88 cm in women). Compared with the lowest category (≤3 times/d), the OR (95% CI) for overweight/obesity in the highest category (≥5 times/d) was 1.54 (1.23, 1.93) in men (P-trend = 0.003) and 1.45 (1.17, 1.81) in women (P-trend = 0.001). The corresponding value for central obesity was 1.42 (1.15, 1.75) in men (P-trend = 0.002) and 1.29 (1.05, 1.59) in women (P-trend = 0.03). The self-report-based MF and time-based MF were positively associated with overweight/obesity, central obesity, or both, although MF based on energy contribution showed no associations. There were positive associations for all SF measures in men and for the energy-contribution-based SF in women. Conclusions: This cross-sectional study suggests that higher EF, MF, and SF are associated with an increased likelihood of overweight/obesity and central obesity in US adults. Prospective studies are needed to confirm the associations observed in this study.