Marinac · JAMA oncology 2016 · prospective cohort study · n=2413

Prolonged Nightly Fasting and Breast Cancer Prognosis.

Cited 184 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective cohort study

PubMed 27032109 · doi:10.1001/jamaoncol.2016.0164 · record verified 2026-08-30

What was done

A prospective cohort analysis was conducted using data from 2,413 non-diabetic women aged 27 to 70 years with early-stage breast cancer enrolled in the Women's Healthy Eating and Living study. Nightly fasting duration was calculated from 24-hour dietary recalls at baseline, year 1, and year 4. Clinical outcomes included invasive breast cancer recurrence and new primary breast tumors (mean 7.3 years follow-up), and breast cancer or all-cause mortality (mean 11.4 years follow-up). Hemoglobin A1c, C-reactive protein, and self-reported sleep duration were also evaluated.

What was found

Participants had a mean nightly fasting duration of 12.5 (SD 1.7) hours. In repeated-measures Cox models, fasting less than 13 hours per night was associated with a statistically significant increase in breast cancer recurrence risk compared with fasting 13 or more hours per night (HR 1.36; 95% CI, 1.05-1.76). Associations with breast cancer mortality (HR 1.21; 95% CI, 0.91-1.60) and all-cause mortality (HR 1.22; 95% CI, 0.95-1.56) were not statistically significant. Each 2-hour increase in nightly fasting duration was associated with lower HbA1c levels (β = -0.37; 95% CI, -0.72 to -0.01) and longer sleep duration (β = 0.20; 95% CI, 0.14-0.26).

Why it matters

This study provides human prospective data suggesting that extending nightly fasting to 13 hours or more may reduce breast cancer recurrence, offering a potential non-pharmacologic lifestyle intervention linked to improved glucoregulation and sleep.

Limits

The observational design cannot establish causality or exclude residual confounding. Fasting time was estimated from self-reported 24-hour dietary recalls rather than objectively tracked timing. Associations with breast cancer-specific and all-cause mortality did not reach statistical significance, and women with diabetes were excluded, limiting generalizability.

Cited by