Dr. Axe Wire · Study decoded · sleep

Irregular Sleep Patterns Linked to Depression and Anxiety in 22,894 Adults

A wearable-device study of nearly 23,000 adults found less regular sleep schedules tracked with more severe depression, anxiety and loneliness.

By Dr. Josh Axe · November 13, 2026 · 1 source · Watch

A wearable-device study of 22,894 adults, published November 12, 2026 in Scientific Reports, found that people with less regular sleep schedules reported more severe depression, anxiety and loneliness.

Researchers analyzed data from 22,894 adults enrolled in the TemPredict Study, using wearable devices to calculate each person's Sleep Regularity Index (SRI), a score reflecting how consistent a person's sleep-wake timing is from day to day. Participants also self-reported their depression, anxiety and loneliness levels. Using all available tracked days, average SRI was 74.93 among people with severe depression versus 83.74 among those with depression within normal limits. People with severe anxiety averaged an SRI of 78.94 versus 83.67 for those within normal limits, and people with elevated loneliness averaged 81.79 versus 83.75 for those without it. The pattern was graded: SRI differences from normal ranged from small to large depending on symptom severity, with standardized effect sizes spanning -0.18 to -1.32 for depression, -0.13 to -0.69 for anxiety and -0.27 for loneliness. Each 5-point drop in SRI was linked to higher odds of more severe symptoms, and results held up in a separate 7-day window measured before COVID-related restrictions began.

A lower score on this index means someone's sleep and wake times shift more from day to day rather than following a steady rhythm. In general, the body's internal clock relies on consistent daily timing cues, and disrupted circadian rhythms have been tied to mood regulation in other research. The authors describe sleep regularity as a potentially modifiable factor worth testing in future intervention studies, rather than stating that fixing sleep timing would ease symptoms.

Sleep Regularity Score by Depression Severity

Scientific Reports, 2026, Soltani et al.

This was a cross-sectional study, meaning sleep regularity and mental health were measured at the same time, so it shows a link, not proof that one causes the other. Depression, anxiety and loneliness were self-reported rather than clinically diagnosed, and the wearable-derived score does not capture sleep quality or total sleep time. Larger intervention studies that actually change sleep timing are the next step to test whether regularity itself shifts mood.

Dr. Axe’s Take

Sleep timing consistency matters as much as total hours, and this nearly 9-point gap in sleep regularity between people with severe depression and those with normal mood backs that up. The study cannot prove irregular sleep causes worse mood, but the graded pattern across depression, anxiety and loneliness is a signal worth acting on. This week, I would set a fixed bed and wake time, including weekends, and get outside for morning sunlight to anchor the circadian rhythm. Pair that with cutting off screens an hour before bed. Most people chase sleep duration and ignore timing consistency, and this data says that is a mistake.

— Dr. Axe

Take the Cortisol Quiz →

Study summary

Who: 22,894 adults in the TemPredict Study who wore sleep-tracking wearables and self-reported depression, anxiety and loneliness symptoms.
Design: Cross-sectional observational study comparing wearable-derived Sleep Regularity Index scores across self-reported mental health severity categories.
Dose: Not applicable; sleep regularity was measured via wearable device data, not an intervention.
Primary result: Mean SRI was 74.93 in severe depression vs. 83.74 within normal limits; 78.94 vs. 83.67 for severe anxiety; 81.79 vs. 83.75 for elevated loneliness.
Secondary: Standardized effect sizes ranged -0.18 to -1.32 for depression, -0.13 to -0.69 for anxiety and -0.27 for loneliness; each 5-point SRI decrease was linked to higher odds of worse symptom severity.
Funding / conflicts: Funders listed in the journal record: Medical Technology Enterprise Consortium; StartSmall Foundation; Oura Health Oy. Conflicts not reported in the abstract.
Date: Nov 12, 2026