Ramadan and the Timing of Scheduled Sessions on an Arabic Digital Mental Health Platform: A Three-Cycle Interrupted Time-Series Analysis (2023–2026)

Background: Ramadan shifts sleep timing and daily routines, but, to our knowledge, its association with the timing of mental health care has not been measured across repeated cycles. Methods: We analysed de-identified records from an Arabic-language digital mental health platform in Saudi Arabia (30 September 2023–29 April 2026; 33,738 adults; 43,190 scheduled sessions). Ramadan 1445, 1446 and 1447 AH were compared with the 30 days before and after each month, using session-level logistic regression for session timing and negative-binomial regression for daily volume, adjusted for a flexible growth trend, day of week and holidays. Mood labels and PHQ-9 and GAD-7 scores were secondary outcomes. Results: During Ramadan, the share of sessions scheduled between 22:00 and 05:59 rose by 19.2 percentage points (95% CI 16.9–21.7; odds ratio 2.69), with increases in each cycle (26.6, 14.7 and 20.0 points) and in every sensitivity analysis; the gain came mainly from midday hours (12:00–17:59; −12.9 points) rather than the evening (−5.8). Daily volume did not change significantly (rate ratio 0.91, 0.81–1.02), although estimates depended on the trend specification and on where the series ended. Raw post-Ramadan increases in volume could not be separated from platform growth. Symptom scores did not differ, and mood shifts were inconsistent. Conclusions: On this platform, Ramadan was associated with a large, recurring shift in scheduled sessions into late-night hours, without a consistent change in volume or self-reported symptoms.