Diets · Metabolic & Cardiometabolic
time-restricted eating increases cardiovascular disease mortality risk
In plain terms: Does an eight-hour daily eating window increase the risk of death from cardiovascular disease?
Insufficient Diets
Part of: 🥗 time-restricted eating
RefutedContestedStrong support
consensus score 0.68
⚖️ Thin evidence — read the needle loosely. The score shows which way the studies lean, but there are too few independent, high-quality ones to place it firmly. Expect this to move as better evidence arrives.
Not established. The finding comes only from re-analyses of a single US survey, which can show a link but not cause, and people who are already ill tend to eat less and over fewer hours. Until another study finds the same thing, it remains unproven.
🧭 Read this score with how it was built in mindWe re-read the papers behind these claims and corrected how each one is counted
This claim came from a 2024 conference abstract. Re-reading the three analyses behind it showed they are three reports from one national survey (NHANES), counted as one group, and that an umbrella review filed as contradicting it was re-pooling trials of other outcomes. On one observational source the verdict cannot be more than 'not enough evidence', in either direction; the earlier 'refuted' overstated what the library held.
📅 Last reviewed: 2026-07-15 ⓘ
Evidence ladder
How far up the ladder this claim has climbed. A high consensus on a low rung means "consistent so far," not "proven in people."
Top evidence so far: Population patterns (Observational)
MechanismIn-vitroAnimalObservationalRCTMeta-analysis
How the studies fall
2 support 0 contradict 0 tested null 1 mixed · 3 sources, 1 independent group
What the evidence shows
A causal increase in cardiovascular mortality has not been established. The reported association comes from re-analyses of one US survey, not independent replications; that neither establishes harm nor proves safety. (The assertion under test is the 2024 AHA-abstract claim.)
The evidence (10)
| Source | Grade | Stance | Quality | Finding |
|---|---|---|---|---|
| Kibret KT, Peeters A, Tegegne TK, Mesfin YM, Nichols M 2025 · Curr Nutr Rep | meta-analysis | contradicts | moderate | [FT-verified 2026-09-25] OFF-SCOPE on the object. Kibret 2025 network meta-analysis of 56 RCTs (3,965 adults, 4-104 wk) pooled ONLY six cardiovascular RISK FACTORS - 'body weight, waist circumference, fat-free mass, blood pressure (SBP and DBP), fasting blood glucose, or LDL cholesterol' - as mean differences; no cardiovascular events or deaths were an outcome. The TRE node is also not an 8-h window: it pools 16/8, 18/6, 10-h and 15-h trials ('The most common TRE methods are the 16/8 and 14/10 method'). TRE vs usual diet: SBP -3.18 mmHg (-5.22 to -1.13), DBP -3.24 (-4.69 to -1.79), FPG -3.74 mg/dL (-6.01, -1.46); LDL null. A marker result neither supports nor contradicts a mortality claim. |
| Sebastian SA, Shah Y, Arsene C 2024 · Disease-a-Month | observational | tested-null | low | Scoping review states the reported about 91% higher CVD-mortality association with under-8h eating windows CANNOT establish a causal link and calls the evidence insufficient and mixed. |
| Molani-Gol R, Safari S, Esmaeil Zadeh Tolouei S, Rafraf M 2026 · Nutr Res | meta-analysis | contradicts | high | Umbrella of 9 meta-analyses (15 RCTs, n=6386 T2D) pooled only risk markers: TRE 'significantly reduced the concentration of fasting blood sugar (WMD = -7.514 mg/dL ...), glycated hemoglobin (WMD = -0.428 ...)' and 'systolic blood pressure (WMD = -3.960 mmHg, 95% CI [-5.495, -2.425])'; no cardiovascular events or mortality reported. |
| Mao Z, Grant H, Kritchevsky SB, Newman AB, Farsijani S 2025 · Aging Cell | observational | supports | moderate | NHANES 2003-2018, n=33,052 adults 20+, median 8.1 y, 1277 cardiovascular deaths (heart disease or cerebrovascular). Exposure is habitual eating window (first to last >0 kcal intake) averaged over two 24 h recalls, not deliberate TRE. Overall population, fully adjusted RCS Model 3: 'Shorter eating windows (≤ 8 h/day) were associated with significantly higher cardiovascular mortality (HR ≥ 1.4) and 95% confidence intervals were consistently above 1', against the spline nadir ('The lowest risk was observed at an 11–12 h eating window in Model 3'); nonlinearity p = 0.009. For CVD the long end was not significant overall ('Although longer eating windows showed an upward trend in hazard ratios, the association was not statistically significant'), significant only in White participants, so on CVD the main result is one-sided. Text gives only an HR floor with no numeric CI; the categorical <8 h vs 12.0-12.99 h CVD HR is plotted in Figure 2 only (cannot-tell). Stronger in older adults (HR ≥ 1.5), men (≥ 1.7), Whites (≥ 1.5). Moderate: the <8 h group was poorer, more food-insecure, lower energy intake and diet quality; authors concede 'residual confounding cannot be ruled out'; first-year-death exclusion reported for all-cause only (HR 1.27, 1.01-1.59). |
| Billingsley HE, St-Onge MP, Alonso WW, Kirkman DL, Kim Y, Carbone S 2024 · Nutr Metab Cardiovasc Dis | observational | mixed | low | NHANES 2003-2018 adults with SELF-REPORTED heart failure (n=991, falling to 606 in model 5; 227 CVD deaths; habitual eating window from two 24-h recalls, dichotomized at the 11.02 h mean — no <=8 h category). 'With full adjustment in model 5, an extended eating window ≥11.02 h was associated with a decreased risk of cardiovascular mortality vs. an eating window <11.02 h (HR 0.36 [95 % CI 0.16–0.81]).' So in that one model the shorter window carried higher CVD mortality — the claim's direction, but at an 11 h cut, not 8 h, in HF patients only, and model-dependent: the authors' own summary is that 'none of the time of eating variables was significantly associated with all-cause or cardiovascular mortality' once day-of-week is adjusted. Recorded as mixed (cultivar ruling 2026-09-25: significant in one model, null in the others and in the authors' final model). Fragile: 'an extended eating window ≥11.02 h did not display a favorable association with mortality in models 1–3'; with a day-of-week covariate 'model 5 HR 0.51 [95 % CI 0.22–1.22], P = 0.130'. Authors: 'we cannot be certain that participants had HF', the cutoff is 'somewhat arbitrary', and 'the causality of all included observations cannot be determined.' Clock times of first/last meal were null for CVD mortality (off-subject here). |
| Los J, Broeders W, Bahrar H, Bulut Ö, Bekkering S, den Boeft A, Sumpter N, Cetinyurek-Yavuz A, Hol IH, Netea MG, Cornel JH, El Messaoudi S, Riksen NP 2026 · Journal of the American Heart Association | RCT | contradicts | low | Randomized open-label crossover, n=19 post-MI patients: 'a 2‐week TRE intervention (eating allowed between 8 am and 2 pm) on metabolomic parameters, innate immune cell function, and systemic inflammation'. TRE 'led to reduced neutrophil counts, lower neutrophil‐to‐lymphocyte ratio, decreased neutrophil CD11b expression', but 'No significant change in CRP was observed (percentage of difference, −6.1% [95% CI, −47.1 to 66.7]; P adj=0.892)'. No cardiovascular events or mortality were measured; the authors call for 'robust clinical trials with cardiovascular end points'. 6-h window, not 8-h. |
| Peng L, Chen L, Xia J, Zhou J, Shi N, Cheng R, Shu L, Huang J, Hu M, Liao J 2026 · Med Sci Sports Exerc | RCT | contradicts | low | Randomized 4-arm parallel trial (NCT05865639), 56 sedentary adults aged >=45 (47 completers analysed, 11/11/11/14), 8 weeks: 'restrict their daily caloric intake to an 8-h window' (10:00-18:00, 16-h fast), with a TRF-alone arm separable from the exercise arms. 'The primary outcomes were changes in cardiovascular functions' (HRV, FMD, ABI, PWV); vs control 'TRF showed significant increases in both SDNN and RMSSD', FMD rose only in exercise arms, and 'No significant alteration was found in all PWV parameters between groups.' Adverse events were a symptom questionnaire (headache, constipation etc.). No cardiovascular events or deaths were measured; the claim's object (CV mortality) is not observed. |
| Peters B, Schwarz J, Schuppelius B, Ottawa A, Koppold DA, Weber D, Steckhan N, Mai K, Grune T, Pfeiffer AFH, Michalsen A, Kramer A, Pivovarova-Ramich O 2025 · Sci Transl Med | RCT | contradicts | moderate | 2-wk-per-arm eTRE vs lTRE crossover, n=31: 'Twenty-four-hour glucose, lipid, inflammatory, and oxidative stress markers showed no clinically meaningful between- or within-intervention differences.' Biomarkers only; no deaths or cardiovascular events reported. |
| Zhang Z, Zhao H, Tao Z, Jiang M, Pu J 2024 · Nutrients | observational | supports | low | NHANES 2005-2018, n=10,561 adults over 60, median 6.66 y. Exposure is habitual fasting duration, '24 − time of last calorie intake + time of first calorie intake' (the complement of the eating window), from two 24 h recalls; no <=8 h eating-window category exists. Top fasting quartile (>12.38 h fasting, eating window under ~11.6 h; mean 13.81 h fasting) vs reference Q3 (10.58-12.38 h fasting): 'Quartile 4 displayed significantly elevated CVD mortality (HR: 1.56, 95% CI: 1.08–2.25) compared to Quartile 3' (Table 2 gives 1.58, 1.10-2.28). The long-eating-window end shows no CVD excess (Q1 HR 0.95, 0.70-1.30); per hour of fasting 'HR: 1.04, 95% CI: 1.00–1.09'. Quality low: text, table and figure legend disagree on key numbers, Q4 carries more diabetes/CKD/hypertension, and authors concede 'further research is needed on whether the longer fasting length preceded, was concurrent with, or occurred after the onset of risk factors and disease diagnoses'; 'few participants in this study would have deliberately used TRE'. |
| Jamshed H, Steger FL, Bryan DR, Richman JS, Warriner AH, Hanick CJ, Martin CK, Salvy SJ, Peterson CM 2022 · JAMA Internal Medicine | RCT | contradicts | high | 14-wk RCT; 'The co–primary outcomes were weight loss and fat loss. Secondary outcomes included blood pressure, heart rate, glucose levels, insulin levels, and plasma lipid levels.' No mortality or CV-event endpoint; diastolic BP −4 mm Hg (95% CI, −8 to 0), other cardiometabolic factors similar. |
Disagree, or know a study we missed?
We grade by evidence, not opinions. The way to weigh in is to point us to a study we haven't cited (check the evidence table above first), or to flag a problem with one we have. Every submission is reviewed; if it holds up, the grade updates and shows in Science Changes Its Mind.
Opens a short form. You'll sign in with Google so submissions are tied to a real account — we don't display your identity, and we only accept a link we can verify (PubMed, DOI, ClinicalTrials.gov).
Educational only, not medical advice. Grades and scores reflect published evidence weighted by study design and quality; see the methodology.