Eating 3 Hours Before Bed Improves Blood Pressure and Blood Sugar Overnight

Cardiovascular and metabolic diseases are strongly linked to modern eating patterns that span 14–15 hours per day, compressing the overnight fast and often placing the last meal close to or even after bedtime. Time-restricted eating (TRE) has emerged as a promising countermeasure, but most existing protocols use fixed daytime windows that may not align with individual sleep schedules — a critical oversight given that sleep itself is a period of active cardiometabolic regulation. This Northwestern University trial was designed to fill that gap by testing a personalized, sleep-aligned extended overnight fasting (EOF) approach.

The study enrolled 39 overweight or obese adults aged 36–75 (recruited 2018–2024) in a randomized parallel-arm controlled trial. Participants were randomized 1:1, stratified by sex, to either the EOF intervention (13–16 hours of overnight fasting, last meal ≥3 hours before habitual bedtime) or a control condition (habitual fasting of 11–13 hours). Both groups were instructed to dim lights to <100 lux during the 3 hours before bed to control for photic inputs to the circadian system. The intervention lasted a minimum of 6 weeks (mean 7.5 weeks). Comprehensive assessments were conducted during two 4-day/3-night inpatient stays — one at baseline and one post-intervention — including polysomnography, ambulatory blood pressure monitoring (ABPM), hourly overnight blood draws for cortisol, and a morning oral glucose tolerance test (OGTT).

For the co-primary outcomes, EOF significantly improved nighttime diastolic blood pressure (DBP) dipping compared to control. Specifically, the EOF group showed greater nocturnal DBP dipping, with more participants converting from non-dipper (<10% dip) to dipper status (≥10% dip). The second co-primary endpoint, the Matsuda Index of whole-body insulin sensitivity, did not reach statistical significance — though this may reflect the relatively short duration or the OGTT timing constraints of the protocol. Nighttime heart rate was significantly lower in the EOF group, and heart rate variability (HRV) in the high-frequency band — reflecting parasympathetic activity — was significantly higher, indicating improved autonomic balance during sleep. Nighttime serum cortisol was also significantly lower in the EOF group across the 7-hour sleep period, consistent with reduced overnight sympathoadrenal activation.

On morning glucose metabolism, the EOF group showed lower glucose levels during the OGTT and a significantly higher 30-minute insulinogenic index — calculated as (Insulin₃₀ − Insulin₀) / (Glucose₃₀ − Glucose₀) — indicating improved acute pancreatic beta-cell responsiveness. The 60-minute OGTT glucose, a known predictor of type 2 diabetes risk, was also lower in the EOF group. These improvements occurred without significant changes in caloric intake or macronutrient composition, supporting the conclusion that meal timing — not caloric restriction — was the active mechanism. Sleep architecture measures from polysomnography were assessed but the full PSG results are noted in the paper with supporting details in supplemental materials.

The clinical implications are notable for both patients and practitioners. This intervention required no special diet, no calorie counting, and no impractical early cutoff times — participants simply moved their last meal 3 hours earlier relative to their own bedtime. The approach is highly individualized and practically achievable, particularly for middle-aged and older adults who face age-related declines in autonomic function and glucose regulation. The significant improvements in nocturnal BP dipping and HRV are especially clinically relevant, as non-dipping BP pattern is independently associated with increased cardiovascular morbidity and mortality. Caveats include the modest sample size (n=39), laboratory-based rather than real-world setting, and the failure to demonstrate improvement in the Matsuda Index, suggesting effects on insulin sensitivity may require longer intervention periods or larger samples.