To examine the non-linear association between breakfast consumption frequency and psychosomatic complaints among adolescents. Data were obtained from five waves (2002, 2006, 2010, 2014, and 2018) of the multi-country Health Behaviour in School-aged Children (HBSC) study. Psychosomatic complaints were assessed using eight items on psychosomatic complaints, combined into a composite score ranging from 0 to 32. Breakfast consumption frequency was measured by the number of days per week. Multilevel generalized additive models were applied to evaluate potential nonlinearity between breakfast frequency and psychosomatic complaints. A total of 918,564 adolescents were included, with a mean (SD) age of 13.59 (1.64) years, of whom 473,633 (51.6%) were girls. Breakfast frequency showed a significant non-linear association to psychosomatic complaints (P for nonlinearity < 0.001). Compared with daily consumption (adjusted mean score: 7.41; 95% CI: 7.39–7.43), breakfast skipping was associated with a 2.84-point (β: 2.84, 95% CI: 2.75–2.93) higher symptom score (adjusted mean score: 10.17; 95% CI: 10.08–10.26). This difference corresponded to 8.9% of the scale range (POMP: 8.9%; 95% CI: 8.6–9.2) and a SMD of 0.49 (95% CI: 0.47–0.50). All association remained significant after False Discovery Rate correction (P-FDR < 0.001) and remained consistent across all survey years. Stratified analyses indicated that the association was stronger among females (β: 3.39, 95% CI: 3.26–3.52) than among males (β: 2.02, 95% CI: 1.90–2.15) and was more pronounced in higher grades (β: 2.07, 95% CI: 1.86–2.28 in grade 5; β: 2.88, 95% CI: 2.75–3.01 in grade 7; β: 2.98, 95% CI: 2.82–3.14 in grade 9). Breakfast frequency showed an inverse, non-linear association with psychosomatic complaints among adolescents, with consistent findings across survey waves. In addition, this association was more pronounced among females and adolescents in higher school grades.
Duan et al. (Fri,) studied this question.