Depression and anxiety among fasting and non-fasting university students: a prospective longitudinal quasi-experimental study in Rajshahi, Bangladesh

Many pre-post studies have reported improvements in depression and anxiety during fasting, but these associations may reflect unobserved confounding and concurrent changes in daily routines. We examined Ramadan-period changes in mental health among university students using a prospective longitudinal quasi-experimental design. Muslim students, who generally fasted during Ramadan, were compared with non-Muslim students who generally did not fast. Data were collected from 366 students across six survey waves in Rajshahi, Bangladesh, in 2024, including two waves before Ramadan, two during Ramadan, and two after Ramadan (698 person-wave observations). The primary fixed-effects difference-in-differences (FE-DiD) analysis used the first four waves (n = 594). The Muslim × Ramadan interaction was not statistically significant for continuous PHQ-9 scores (β = 0.71, SE = 1.25), continuous GAD-7 scores (β = 1.68, SE = 1.02), or GAD-7 ≥ 10 (β = 0.04, SE = 0.18). However, PHQ-9 ≥ 10 was 17.0 percentage points lower among Muslim students during Ramadan relative to non-Muslim students (SE = 0.04); this finding was sensitive to the cutoff used and was not accompanied by a clinically meaningful continuous-score change. Subgroup analyses showed lower GAD-7 scores among Muslim males (β = −3.02) but higher scores among Muslim females (β = 2.90), with significant sex heterogeneity. Attrition analyses did not identify baseline PHQ-9 or GAD-7 scores as predictors of subsequent participation, although substantial attrition and the small non-Muslim comparison group limited precision. Overall, evidence for broader improvements in depression or anxiety symptoms was limited.