Do I have to eat carbohydrate straight after training?
Only when the next hard session is within about 8 h: early carbohydrate starts refuelling, and over a day the total eaten decides (leading-group review, B+).
Main caveat: Narrative mini-review. No search or appraisal; the numbers are the authors' reading of mostly small, mostly male biopsy studies (Table 1's studies have 6 to 33…
Burke, van Loon and Hawley review what speeds or slows the storage of muscle glycogen after exercise, with most attention on two awkward cases: a short gap between sessions and too little carbohydrate or energy. They describe storage as biphasic.
In the early phase (0-4 h) glycogen depletion "provides a strong drive for its own resynthesis" and about 1 g/kg body mass of carbohydrate optimises it; in the later phase (4 to 24 h) intake should meet the fuel needs of the next session, and the type, form and pattern of intake matter less than the total.
With no carbohydrate at all, glycogen is still made at 1–2 mmol/kg wet weight per hour (from gluconeogenesis, or lactate after high-intensity work); with large amounts eaten early and continued, the highest rates are typically 5–10, and at a mean of 5–6, 20–24 h are normally needed to normalise glycogen after extreme depletion.
They reread the study behind the "window of opportunity" (Ivy 1988, REF-056): its key finding, they write, was that synthesis stayed very low until carbohydrate arrived, so eating early starts refuelling rather than catching a brief period of enhanced storage; that matters when only 4-8 h separate sessions, while more than 8 h may make up for a late first meal.
A daily "glycogen storage threshold" appears at about 7-10 g/kg; protein added to carbohydrate helps below that threshold and not when carbohydrate exceeds about 1 g/kg/h. Damage slows the refill: after soccer matches glycogen was still below resting levels at 24 h, and in type II fibres at up to 48 h.
Supercompensation is probably reached within 36–48 hours of the last session with rest and enough carbohydrate.
Our reading: this is the review behind "refills within about a day" — but its 20–24 h figure is the authors' statement for extreme depletion, citing a 1991 review, not a time course measured here; for a short gap the early rate and the start time matter, for a day or more the total.
Protocol
Design: narrative mini-review; no search strategy, inclusion rules or appraisal described (n = not applicable); 117 references, three tables (other dietary constituents; guidelines; research questions) and one figure (the evolution of loading protocols).
Version read: the author accepted manuscript (43 pages, line-numbered) in the Liverpool John Moores University repository, listed by OpenAlex for this DOI. The publisher's free PDF sits behind a bot check and was not read, so small differences from the version of record cannot be excluded.
Scope: glycogen structure and synthesis (glycogenin, proglycogen and macroglycogen, three subcellular pools); amount, timing and type of carbohydrate; energy availability; protein and fat co-ingestion; alcohol; creatine, caffeine, fenugreek, hydroxycitrate and conjugated linoleic acid; dehydration; muscle damage; local cold and heat; supercompensation protocols; current guidelines and open questions.
Studies behind Table 1 (other dietary constituents): 6 to 33 participants each, mostly men (the table's own counts).
Funding and interests: the accepted manuscript has no funding, acknowledgment or disclosure section, and Crossref lists no funder; the published version's disclosure statement was not read.
Key Results
What the review states, and whose number it is (rates in mmol per kg of wet muscle per hour):
| Point | Number as printed | Basis in the review |
|---|---|---|
| Storage with no carbohydrate after exercise | 1–2 mmol/kg w.w./h | cited studies (gluconeogenesis; lactate after high-intensity exercise) |
| Highest rates, large carbohydrate intake started early | typically 5–10 mmol/kg w.w./h | the authors' summary |
| Time to normalise after extreme depletion | 20–24 h, at a mean 5–6 mmol/kg w.w./h | the authors, citing Coyle 1991 (their ref 30) |
| Daily intake at which storage plateaus ("glycogen storage threshold") | ~7-10 g/kg body mass | synthesis of 24-h studies including two dose-response studies |
| Larger early intakes (>1 g/kg/h) in repeated small feedings | 30-50% higher rates (10–11) over the first 4 h | four cited studies |
| Immediate vs delayed carbohydrate (Ivy 1988, REF-056) | 7.7 mmol/kg ww/h in the first 2 h with immediate intake vs ~4.4 when the same feeding was delayed 2 h | ref 47; Ivy's own abstract gives 4.1 for the delayed group once fed (see Limitations) |
| Gap at which the first meal's timing matters | 4-8 h; >8 h may compensate for a delay | the authors' reading, citing one study (ref 78) |
| Protein co-ingestion | helps with 0.5–0.8 g CHO/kg/h; no further effect when CHO >1 g/kg/h; 0.3-0.4 g/kg maximises it (Table 2: ~20-25 g) | cited studies and the review by Betts and Williams (ref 13) |
| Resting glycogen | ~80-85 (untrained, mixed diet), ~120 (endurance-trained), up to 200 after supercompensation | cited studies |
| Short loading | ~90 to ~180 mmol/kg ww with 24 h of rest and high carbohydrate, stable over two more days; the true loading phase from the last session was ~36 h | Bussau 2002, REF-891 (ref 25) |
| Repeated loading | 48 hours of 12 g/kg/d and rest raised glycogen above resting the first time, not the second | one study of well-trained cyclists (ref 62) |
| Guidelines (Table 2) | 0-4 h: ~1 g/kg BM/h in frequent small feedings; daily: moderate load 5-7, heavy 6-10, extreme 8-12 g/kg/24 h; loading: 8-12 g/kg/24 h for 36-48 h with a taper, for events > 90 min, in the absence of muscle damage | reproduced from guidelines (refs 23, 24, 105) |
Muscle damage: rates are impaired after muscle-damaging eccentric contractions and impacts (less GLUT4 translocation and glucose uptake), in later rather than early (0-6 h) recovery; after soccer matches glycogen stayed below resting levels at 24 h in both fibre types and at up to 48 h in type II fibres, despite relatively high carbohydrate intakes.
Meals vs snacks, liquid vs solid: over longer recovery the frequency and form of the recommended amount did not change storage, despite different blood glucose and insulin responses.
Alcohol: in well-trained cyclists, about 120 g of alcohol that displaced recovery carbohydrate reduced storage over 8 h and 24 h; a direct effect of blood alcohol was unclear (REF-1013).
Caffeine: 8 mg/kg with 1 g/kg/h of carbohydrate raised the 4-h rate (13.7 ± 4.4 vs 9.0 ± 1.8 mmol/kg ww/h, 7 men), while 1.7 mg/kg/h with 1.2 g/kg/h did not change the 6-h rate (7.1 ± 1 vs 7.1 ± 1, 14 men); the authors see no explanation and question caffeine for sleep reasons.
Dehydration: a loss of about 5% body mass did not impair glycogen storage over 15 h after cycling (one study).
Mechanisms
Measured (in the cited studies, as the review summarises them): after glycogen-depleting exercise, glycogen synthase is activated by depletion, insulin sensitivity rises and the membrane becomes more permeable to glucose; the enhanced action of insulin on muscle can persist for up to 48 h. Eccentric damage reduces GLUT4 and glucose uptake.
Authors' interpretation: the biphasic pattern partly reflects two pools of glycogen granules — proglycogen, which is made faster and plateaus after 24 h, and macroglycogen, made more slowly for up to 48 h — and the amount of glycogenin has, in their words, a direct influence on how much glycogen the cell can store, while what sets glycogenin is "largely unexplored" (one of their open questions).
Authors' interpretation: the "window" is the start of refuelling, not a gate: synthesis stays low until carbohydrate arrives, and glycogen's negative feedback on its own synthesis may equalise stores over time, so a longer gap forgives a late start.
Authors' interpretation: the "paradox" that frequent feeding raises early rates but not 24-h storage — keeping blood glucose and insulin up matters in the first hours and when total intake is short, not once the total is above the threshold.
Inference (the base's): the "about a day" refill assumes a non-damaging session; after a long downhill run or a match the review's own damage data point to a slower refill, but no time course after running is given.
Practical Application
Do I have to eat carbohydrate straight after training? Only when the next hard session is within about 8 h: early carbohydrate starts refuelling, and over a day the total eaten decides (leading-group review, B+).
What it is good for as a source: the biphasic mechanism — depletion drives early storage, total intake decides the later phase — and the reread of the "window" study, as the authors' synthesis of small biopsy studies. It does not measure a refill curve: "20–24 h" is the authors' figure for extreme depletion, citing a 1991 review.
Short gap (under about 8 h): carbohydrate as soon as practical after the first session, about 1 g/kg/h over the first 4 h in frequent small feedings, foods of moderate-high glycaemic index (Table 2, reproduced from consensus guidelines). This is the case where timing matters; the base's guide draws that band at under ~6 h, as practice (SYNTH-85).
When carbohydrate is short (poor appetite, an energy-restricted phase): protein added to the carbohydrate can help storage (Table 2: about 20-25 g; text: 0.3-0.4 g/kg); with carbohydrate above about 1 g/kg/h it adds nothing to glycogen (REF-151).
After damaging sessions (long downhill runs, matches, unaccustomed eccentric work), expect a slower refill; the suggestion that more carbohydrate in the first 24 h may partly offset it rests on one eccentric-exercise study (ref 29).
Limitations
Narrative mini-review. No search or appraisal; the numbers are the authors' reading of mostly small, mostly male biopsy studies (Table 1's studies have 6 to 33 participants), and the guideline table is reproduced from guidelines Burke co-authored, so it is not an independent check of them.
"20–24 h" is not a measured curve. It is stated for extreme depletion at a mean rate, with one 1991 review as the source; how long a typical training session takes to refill is not given.
Small inconsistencies in the manuscript: the delayed rate in Ivy 1988 is given as ~4.4 mmol/kg ww/h, while Ivy's abstract reports 4.1 (delayed, once fed) and 4.3 (immediate, second 2 h) (REF-056); the protein dose reads 0.3-0.4 g/kg in the text and about 20-25 g in Table 2 (about the same for a 60-70 kg athlete; our arithmetic); the Figure 1 legend says supercompensation occurs in 24-48 h, the text 36–48 hours; the text calls the 1-day loading study (ref 25, Bussau) a midpoint to "Sherman and colleagues (25)", a citation slip.
Accepted manuscript, not the version of record. Wording and table layout may differ slightly in print.
Funding and interests not printed in the manuscript read; Crossref lists no funder.
How this entry was checked and changed
Why this level
authoritative narrative review from the leading sports-nutrition and muscle-glycogen groups (Burke, Australian Institute of Sport / Australian Catholic University; van Loon, Maastricht; Hawley, ACU / Liverpool John Moores), the rubric's line for a leading-group narrative review. Its numbers are the authors' summary of mostly small biopsy studies, not pooled estimates, and its guideline table is reproduced from consensus guidelines two of which Burke led
Full text (checked 2026-10-28)
In the early post-exercise period (0-4 h), glycogen depletion provides a strong drive for its own resynthesis, with carbohydrate (~ 1 g/kg body mass) optimizing this process; in the later phase, 4 to 24 h, type, form and pattern of intake matter less than total intake (full text)
Even without post-exercise carbohydrate, glycogen synthesis occurs at 1–2 mmol/kg wet weight of muscle/h; the highest rates, typically within 5–10 mmol/kg w.w./h, come when large amounts are consumed soon after exercise and continued (full text)
With a mean glycogen storage rate of 5–6 mmol/kg w.w./h, 20–24 h of recovery are normally required for normalization of muscle glycogen levels following extreme exercise depletion (full text)
A 'glycogen storage threshold' appears to occur at a daily carbohydrate intake of ~7-10 g/kg body mass (full text)
Larger intakes (e.g. >1 g/kg/h) consumed as repeated small feedings gave 30-50% higher rates of glycogen synthesis (10–11 mmol) over the first 4 h of recovery (full text)
Ivy's study: 7.7 mmol/kg ww/h during the first 2 h with immediate intake vs ~4.4 mmol/kg ww/h when the feeding was delayed after 2 h; the key finding was that synthesis remained very low until carbohydrate feeding was initiated, which matters with only 4-8 h between sessions, while a longer (>8 h) recovery may compensate (full text)
Protein co-ingestion increases synthesis with carbohydrate below the threshold (e.g. 0.5–0.8 g CHO/kg/h) and has no further effect when carbohydrate is adequate (e.g. >1 g/kg/h); 0.3-0.4 g/kg appears to maximize the effect (full text)
Glycogen rose from ~90 mmol to ~180 mmol/kg ww with 24 h of rest and high carbohydrate intake, the true loading phase being ~36 h; supercompensation is probably achieved within 36–48 hours of the last exercise session (full text)
Table 2: in the absence of muscle damage, a carbohydrate intake of 8-12 g/kg/ 24 h for 36-48 h with an exercise taper can supercompensate muscle glycogen before events of > 90 min (full text)
After soccer matches glycogen remained below resting levels after 24 h of recovery in both fibre types and after as much as 48 h in Type II fibres (full text)