Longevity & Aging · Metabolic & Cardiometabolic
circadian and peripheral-clock desynchronization irregular or late eating independently-drives obesity and metabolic disease separate from diet and calories
In plain terms: Does eating/sleeping at the wrong body-clock time cause metabolic harm on its own?
Strong support Longevity & Aging
Part of: • Circadian & Light
RefutedContestedStrong support
consensus score 0.94
Yes — independent controlled human studies and shift-work epidemiology show misalignment worsens glucose/insulin/BP even with identical food, so the timing effect is genuinely separable from what and how much you eat.
📅 Last reviewed: 2026-07-14 ⓘ
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: All trials, pooled (Meta-analysis)
MechanismIn-vitroAnimalObservationalRCTMeta-analysis
How the studies fall
5 support 0 contradict 0 tested null 1 mixed · 6 sources, 5 independent groups
The evidence (9)
| Source | Grade | Stance | Quality | Finding |
|---|---|---|---|---|
| Hemmer 2021 · Nutrients | observational | supports | moderate | Independent review: shift work is a risk factor for obesity, T2D, hypertension, metabolic syndrome, though real-world confounders (diet, sleep) remain hard to fully separate. |
| Scheer, Hilton, Mantzoros, Shea 2009 · Proc Natl Acad Sci U S A 2009;106(11):4453-4458 | RCT | supports | high | Independent Brigham forced-desynchrony experiment, 10 healthy adults on a recurring 28-h day: 'Subjects ate 4 isocaloric meals each 28-h day.' Under ~12 h misalignment the paper reports it 'systematically decreased leptin (-17%, P < 0.001), increased glucose (+6%, P < 0.001) despite increased insulin (+22%, P = 0.006), completely reversed the daily cortisol rhythm (P < 0.001), increased mean arterial pressure (+3%, P = 0.001), and reduced sleep efficiency (-20%, P < 0.002)', and 'caused 3 of 8 subjects ... to exhibit postprandial glucose responses in the range typical of a prediabetic state.' Calories held constant, so this is causal evidence for the claim's calorie-independent half in humans — but the endpoints are risk biomarkers over days, not obesity or diagnosed disease, and the 20% drop in sleep efficiency means the effect is not isolated from sleep loss. |
| Leung 2020 · Chronobiol Int | meta-analysis | supports | moderate | Systematic review and meta-analysis of acute day-vs-night postprandial studies in healthy adults, with the meal held identical within the same participant: 'Included studies met the following criteria: had a day-time test between 0700 - 1600h, a nighttime test between 2000 and 0400h, the test meals were identical and consumed by the same participant at both day and night time points, preceded by a 3-h fast (minimum).' Result: 'Meta-analysis for glucose showed a lower postprandial glucose response in the day compared to during the night, after an identical meal (SMD = -1.66; 95% CI, -1.97 to -1.36; p < .001)', and 'Meta-analysis also showed a lower postprandial insulin response in the day compared to during the night (SMD = -0.35; 95% CI, -0.63 to -0.06; p = .016).' The identical-meal within-participant design isolates a time-of-day effect from intake and from the person, which is the claim's 'separate from diet and calories' half tested directly. The endpoint, however, is acute postprandial glycemia, not obesity or incident disease, and the authors keep the disease step as inference: 'Our results suggest poor glucose tolerance at night compared to the day. This may be a contributing factor to the increased risk of metabolic diseases observed in those who habitually eat during the night, such as shift workers.' Quality moderate rather than high: 'Fifteen studies met the eligibility criteria, ten of which were included in the meta-analyses', the insulin effect is small with a CI reaching -0.06 and 'findings from included studies ineligible for meta-analysis were inconsistent', the search closed February 2018, and heterogeneity, risk-of-bias and publication-bias assessments are not in the abstract-only text available. |
| Xie F, Hu KS, Fu RR, Zhang YM, Xu KQ, Tan JN 2024 · BMC Endocr Disord | meta-analysis | supports | moderate | Cohort-only meta-analysis (9 articles, 10 prospective/retrospective cohorts, 'over 235,800 participants', all NOS 7-9, PROSPERO CRD42024520037): night shift workers 'exhibit a higher incidence rate of T2DM compared to non-night shift workers (HR = 1.30, 95%CI: [1.18, 1.43], P < 0.001, I 2 = 57.6%)', with a duration gradient ('>10 years: HR = 1.17, 95%CI: [1.10, 1.24]' vs '≤ 10 years: HR = 1.06, 95%CI: [1.03, 1.10]') and no publication bias on Begg's test (P = 0.371). Entirely observational, exposure is an occupational schedule rather than measured circadian phase, and definitions differ by cohort; the authors concede 'Since there are differences in the adjustment factors of each original study and not all factors can be controlled, these can lead to biases in the results' and that missing data on 'intensity of night shift work, start times, and dietary habits' limited the analysis. The paper runs no test separating circadian disruption from diet, calories or sleep loss — it supports the association half of the claim, not the independence half. Female-dominated ('nearly 200,000 women' vs 'more than 10,000 men'); the null male subgroup (HR = 1.53, 95%CI: [0.89, 2.63]) is underpowered by the authors' own account, not evidence of no effect. |
| Xi 2025 · Front Public Health | meta-analysis | supports | moderate | OFF-SCOPE for a circadian-misalignment-drives-obesity-and-metabolic-disease claim: the exposure half fits, the outcome half is a different disease family. The paper's declared question is cardiovascular only — 'This study aimed to assess the relationship between night shift work and the incidence and mortality of CVD.' — and its PICO fixes the outcome set as 'The outcomes focused on cardiovascular events (including both incidence and mortality) and encompassed major CVD subtypes such as coronary heart disease (CHD), ischemic heart disease (IHD), and stroke,' with marker-only studies explicitly excluded ('studies with self-reported cardiovascular outcomes or those reporting only cardiovascular markers'). The supplementary search strategy (Table S1) contains no obesity, BMI, diabetes or metabolic-syndrome term in any of the six databases, so a metabolic-disease result could not have been retrieved even in principle; every one of the 23 pooled cohorts in Table 1 reports a CVD, CHD, IHD or stroke endpoint and not one reports weight, glycemia or incident T2DM. Its actual finding, banked as a candidate claim, is 'Overall, this meta-analysis revealed that night shift work significantly increased the risk of total CVD events (RR = 1.13, 95% CI = 1.10–1.16) and total CVD mortality (RR = 1.27, 95%CI = 1.18–1.36).' over 23 cohorts 'with a total population of 3,340,377 participants', with a linear dose-response ('For each 5-year increase in shift work duration, the risk of CVD incidence increased by 7%'). The only bridge to obesity or metabolic disease is a secondhand mechanism sentence in the Discussion citing others' work — 'This disruption impairs autonomic regulation and hormonal homeostasis, rendering individuals more susceptible to hypertension, insulin resistance, and metabolic dysfunctions' — premise, not result. The claim's 'independently / separate from diet and calories' half is doubly unmet: the authors concede 'Second, residual confounding from unmeasured factors (dietary habits, physical activity) cannot be fully excluded.', adjustment is heterogeneous across the pool (UK Biobank and the Nurses' Health Study adjust for BMI, diet and physical activity; Bøggild 1999 and Hublin 2010 adjust for age alone), and no analysis separates schedule from intake. Certainty is low by the authors' own grading: 'GRADE evaluation indicated predominantly low to very low certainty evidence', and 'Given the observational cohort designs, initial certainty was low. Evidence for CVD incidence was downgraded to very low due to publication bias, while CVD mortality maintained low certainty absent downgrading factors.' — Egger's test on the headline incidence pool 'indicated potential publication bias.' (trim-and-fill left the estimate intact). Read at full text (JATS + supplement); the prior extract was abstract-derived and treated a cardiovascular result as a metabolic one. |
| Gao 2020 · Chronobiol Int | meta-analysis | supports | moderate | Meta-analysis of 21 observational studies (12 cohort, 9 cross-sectional): 'shift work was associated with an increased risk of type 2 diabetes (relative risk = 1.10, 95% confidence interval = 1.05-1.14)', with 'night shift and rotating shift' both positive but 'female shift workers have increased risk of type 2 diabetes while male not observed'. Dose-response was limited to 'three cohorts among female workers' and hedged by the authors as 'there might be a positive association between duration of shift work and the risk of type 2 diabetes'. Observational only — the paper tests no separation of circadian misalignment from diet, calories or sleep loss. |
| Bonham 2016 · Chronobiol Int | meta-analysis | supports | moderate | OFF-SCOPE for a circadian-misalignment-causes-metabolic-disease claim. The paper's exposure is a work schedule and its only pooled outcome is dietary energy: 'This systematic review investigated whether the 24 h energy intake of shift workers differs to that of fixed day workers.' Its result is a null on intake - 'The standardised mean difference (95% CI) in energy intake between shift and day workers was -0.04 (-0.11, 0.03); I(2) = 54%' across 'There were 10 367 day workers and 4726 shift workers from 12 studies included in the qualitative analysis and meta-analyses.' Neither circadian misalignment nor obesity or any metabolic endpoint was measured; elevated obesity in shift workers is cited as background, not analysed. The only bridge to this claim is a hedged conjecture naming four rival explanations: 'Reported energy intakes were not different between day workers and shift workers, suggesting that other factors such as circadian misalignment, meal timing, food choice and diurnal variation of energy metabolism at night may be responsible for the increased rates of obesity observed in shift workers.' Argument-by-elimination, and weakly so - every input is self-reported intake, under-reporting is plausibly differential by shift status, and 'Qualitative results on macronutrient intakes were conflicting.' Read at abstract grade (no PMCID, no open full text); the paper's real finding is preserved as a candidate claim on shift work and energy intake. |
| Madjd, Taylor, Delavari, Malekzadeh, Macdonald, Farshchi 2021 · Br J Nutr | RCT | supports | moderate | 82 women (BMI 27-35) randomised for 12 weeks to an evening meal at 19.00-19.30 (EEM) v. 22.30-23.00 (LEM) during the same weight-loss programme: 'the EEM group had a greater mean reduction in weight (EEM: -6.74 (sd 1.92) kg; LEM: -4.81 (sd 2.22) kg; P < 0.001)', waist circumference (-8 v. -6 cm, P = 0.007), total cholesterol (-0.51 v. -0.43 mmol/l, P = 0.038), TAG (-0.28 v. -0.19 mmol/l, P < 0.001) and HOMA-IR (-0.83 (sd 0.37) v. -0.55 (sd 0.28), P < 0.001). Randomising meal timing alone moved weight and metabolic markers, which supports late eating driving worse outcomes. Caveat (abstract-only): achieved energy intake per arm is not reported, and because the early arm also lost ~1.9 kg more, the lipid and HOMA-IR gaps may be downstream of the weight difference rather than calorie-independent. |
| Pizinger 2018 · Sleep Health | RCT | mixed | low | n=6 4-phase randomized inpatient crossover holding sleep duration and food intake identical: NO effect of sleep or meal timing on insulin sensitivity — 'There were no effects of sleep and meal times or sleep × meal time interaction on Si (all P>.35), acute insulin response to intravenous glucose (all P>.20), and disposition index (all P>.60) after adjusting for sex and body mass index.' The meal-timing axis did move secondary endpoints — 'Meal tolerance test glucose and insulin areas under the curve were lower during Nm (glucose P=.11; insulin P=.0088)' and 'an effect of meal times on overnight glucose (P=.0040 and .012, respectively) and insulin (P=.0075 and .067, respectively).' Authors' own conclusion: 'Sleep timing, without concomitant sleep restriction, does not adversely affect Si and glucose tolerance, but meal times may be relevant for health.' Pilot, n=6, authors ask for replication in a larger sample; abstract-only. |
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.