Carbs and Performance: Why Low-Carb Diets Hurt Your Training

Carbohydrate is the only macronutrient with no dietary requirement, and low-carb advocates never tire of pointing that out. Your liver can manufacture glucose from amino acids and glycerol, your brain can run mostly on ketones after adaptation, and nobody has ever been diagnosed with a carbohydrate deficiency. All true. None of it answers the question a lifter actually cares about: what happens to your training when the carbs go away?

That question has a real answer, and it is more specific than either side of the diet war admits. Carbohydrate restriction does not make you weak in any general sense. It degrades a particular kind of performance -- repeated, high-effort sets in the rep ranges most people use to build muscle -- while leaving other kinds nearly untouched. Understanding which is which explains most of the confusion in the research, most of the anecdotes on both sides, and most of what you should do about it.

This article walks through the physiology of glycogen during lifting, the studies on carbohydrate restriction and strength performance, every meaningful trial of ketogenic diets in lifters, the actual carb numbers that matter, and the two topics that generate endless content but modest returns: timing and carb cycling.

What Glycogen Is and How Lifting Uses It

Glycogen is stored glucose: a branched polymer packed into your muscle and liver along with water. A trained adult carries roughly 400 to 600 grams of glycogen in skeletal muscle and another 80 to 100 grams in the liver. Muscle glycogen is selfish -- the enzyme needed to release glucose back into the bloodstream is absent in muscle tissue, so the glycogen stored in your quads can only fuel your quads. Liver glycogen maintains blood glucose for everything else, including your brain.

When you lift, the energy for each rep comes from three overlapping systems. The first few seconds of a set draw on ATP and phosphocreatine already sitting in the muscle. Past that, fast glycolysis takes over: glycogen is broken down to fuel work at rates far beyond what fat oxidation can match. Fat is a huge fuel tank with a small fuel line. It cannot deliver energy fast enough for a set of eight taken near failure, which is why hard sets run on glycogen no matter what diet you eat.

How much glycogen does lifting actually use? More than the "lifting barely touches glycogen" crowd claims, less than a marathon. The measurements are consistent across three decades:

• Tesch et al. (1986) measured a 26 percent drop in quadriceps glycogen after about 30 minutes of heavy multi-set leg work (5 sets each of squats, leg presses, and leg extensions).
• Robergs et al. (1991) found 6 sets of leg extensions at 70 percent of 1RM depleted quadriceps glycogen by roughly a third, with depletion concentrated in the fast-twitch fibers doing most of the work.
• MacDougall et al. (1999) showed even a single set of biceps curls to failure dropped biceps glycogen by about 12 percent, and three sets by about 24 percent.
• Koopman et al. (2006) confirmed substantial glycogen use in type II fibers after a full resistance session, again concentrated where the force is produced.

A realistic hypertrophy session -- 15 to 25 hard sets spread over a muscle group -- lands in the range of 25 to 40 percent depletion in the muscles trained. That will not empty the tank. It does not need to. Performance does not fall off a cliff at zero glycogen; it degrades as concentrations drop, because glycogen depletion within the specific fibers doing the work impairs their ability to sustain output. Fast-twitch fibers deplete first and disproportionately, and they are the fibers you care about.

The classic demonstration that diet moves this needle is Bergstrom et al. (1967), the study that invented carb loading. After a glycogen-depleting exercise bout, subjects ate a low-carb, mixed, or high-carb diet for several days. Time to exhaustion on a subsequent endurance test roughly tripled from the low-carb condition to the high-carb condition, tracking measured muscle glycogen almost linearly. That was cycling, not lifting, but it established the core fact everything else builds on: the glycogen you start a session with is set by the carbohydrate you ate in the days before it.

What Happens to Performance When Carbs Drop

If glycogen fuels hard sets, restricting carbs should hurt lifting performance. The research says it does -- but selectively, and the selectivity is the whole story.

Leveritt and Abernethy (1999) ran the cleanest early demonstration. Lifters spent two days on a very low carbohydrate diet or a mixed diet, then performed 3 sets of squats at 80 percent of 1RM to failure plus isokinetic knee extension testing. After carb restriction, subjects completed significantly fewer squat reps across the 3 sets. Isokinetic performance -- brief, maximal efforts with full rest -- did not differ. Two days of low carb was enough to measurably shrink high-effort squat volume while leaving short maximal output alone.

Mitchell et al. (1997) found the opposite: after glycogen-lowering exercise plus 3.5 days of low carbohydrate intake, subjects performed 15 sets across five exercises no worse than in the high-carb condition. Honest accounting requires including this study, and it points at the variable that reconciles the literature: the demands of the test. Moderate volume, longer rests, and submaximal loads give the phosphocreatine system and remaining glycogen enough room to cover the work.

The comprehensive treatment is Henselmans et al. (2022), a systematic review in Nutrients covering 49 studies of carbohydrate intake and resistance training performance. The pattern across the whole literature:

• For sessions under about 45 minutes of moderate volume, carbohydrate manipulation rarely changes performance. Stored glycogen covers it.
• For longer, higher-volume sessions (roughly 10 or more sets per muscle group, or total sessions past the hour mark), higher carbohydrate availability supports more total reps and better maintenance of output in later sets.
• Training fasted (more than 8 hours since the last meal) is where pre-exercise carbohydrate most reliably helps.

Read that pattern next to the depletion data and the picture is coherent. A short, heavy session does not use enough glycogen for availability to bind. A long hypertrophy session does. The lifter most hurt by low carbohydrate is precisely the one training with the volumes that build muscle: multiple hard sets, moderate reps, accumulated over an hour or more. The lifter least hurt is the low-volume strength specialist pulling heavy singles and triples with five-minute rests.

Keto for Lifting: The Full Trial Record

The ketogenic diet -- carbohydrate low enough (usually under 50 g/day) to shift the body toward ketone production -- is the strongest version of the low-carb claim, so the trials in lifters are worth walking through individually. There are not many, and they mostly point the same direction.

Vargas et al. (2018): the hypertrophy test

Twenty-four trained men ran 8 weeks of the same hypertrophy program in a small caloric surplus, randomized to a ketogenic or non-ketogenic diet, measured by DXA. The non-keto group gained about 1.3 kg of lean mass. The keto group gained none -- lean mass was flat to slightly negative -- despite the surplus and identical training. Keto did reduce fat mass slightly while the control group's fat held steady. If your goal is building muscle, this is the most directly relevant trial that exists, and keto failed it.

Wilson et al. (2020): the carb reintroduction wrinkle

Twenty-five resistance-trained men, 10 weeks keto versus western diet, followed by a carbohydrate reintroduction in week 11 for the keto group. Reported lean mass gains ended up similar between groups -- but the keto group's lean mass jumped substantially during the reintroduction week, which is exactly what refilling glycogen and water looks like on a DXA scan. Strength and power gains were comparable. The study is frequently cited as pro-keto; the reintroduction spike is a reminder that "lean mass" on a scan includes glycogen and water, a theme that runs through this entire literature.

Greene et al. (2018): the weight-class result

Fourteen powerlifters and Olympic weightlifters completed a 3-month crossover: usual diet versus ad libitum keto. On keto they lost about 3.3 kg of body mass while lifting performance was maintained. For strength athletes who need to make a weight class, that is a useful result: keto dropped weight without dropping the total. Note what the training was -- low-rep, high-load, long-rest work, the exact style the depletion data predicts should survive carb restriction.

Paoli et al. (2012): the gymnasts

Eight elite artistic gymnasts spent 30 days on a very low carbohydrate ketogenic diet. Strength performance on their battery of tests held steady while body fat dropped. Same pattern: short maximal efforts, bodyweight athletes with a strong incentive to be light, performance preserved.

Kephart et al. (2018): the CrossFit signal

In CrossFit trainees, 3 months of keto reduced body mass, but the keto group showed a decline in vastus lateralis thickness relative to controls -- a direct ultrasound hint of the lean tissue cost, in a training style heavy on glycolytic work.

Burke et al. (2017): the endurance warning shot

Not a lifting study, but too well-controlled to skip. In elite race walkers, 3 weeks of low-carb high-fat eating increased fat oxidation dramatically -- and impaired exercise economy, wiping out the performance improvement that the same block of intensified training produced in the high-carb groups. Fat adaptation is real; free performance from it is not. The oxygen cost of using fat as fuel is higher, and at high intensities that bill comes due.

The synthesis writes itself. Keto preserves maximal strength at low volumes and reliably drops body weight. It has never beaten a higher-carb diet for muscle growth in a controlled trial, and in the best hypertrophy-specific test it lost outright. Paoli's own review group put it diplomatically in 2019, describing keto and hypertrophy as a "frenemy" relationship: usable for cutting and weight-making, working against you when the goal is tissue gain. The evidence for lifting on keto is not good, and it is not good in a specific, predictable way.

Why the Scale and the Bar Both Lie on Low Carb

Every glycogen molecule is stored with roughly 3 grams of water. Deplete 400 grams of glycogen and about 1.2 kg of associated water goes with it -- before counting reduced gut contents from a lower-residue diet. This single fact explains the two most persuasive low-carb illusions.

The scale illusion. The first week of low carb routinely drops 1.5 to 3 kg. It feels like the fastest fat loss of your life. It is glycogen and water leaving, and it returns just as fast when carbs do -- which is also why post-diet "rebounds" look so alarming and mean so little.

The scan illusion. DXA and bioimpedance count glycogen-bound water as lean mass. A keto group therefore starts any study with an instant 1 to 2 kg lean mass handicap that has nothing to do with contractile tissue, and erases it the week carbs return, as Wilson 2020 showed on the way back up. When reading body composition studies in this space, the honest question is always: how much of the difference is muscle, and how much is what fills muscle?

The bar lies in the other direction. A flat muscle moves less weight for reasons that are partly mechanical -- muscle volume contributes to leverage and stiffness -- and partly fuel. Lifters who feel weak and deflated in week two of a low-carb cut are not imagining it, and they also have not lost the muscle they think they have. One or two days of high carbohydrate intake restores both the fullness and most of the missing reps, which tells you what was actually gone.

Minimum Carb Intake for Lifters

Endurance sport guidelines of 6 to 10 g of carbohydrate per kg of body weight per day get copy-pasted into lifting advice constantly, and they are overkill. A lifter burning 30 to 40 percent of local glycogen a few sessions per week does not have marathon-level resynthesis demands. The honest numbers:

Slater and Phillips (2011) suggested 4 to 7 g/kg/day for strength sports as a reasonable envelope. Escobar et al. (2016) reviewed the actual glycogen cost of resistance training and argued that typical bodybuilding-style training is fully supported toward the lower end of that range, since a session touching 30 to 40 percent of glycogen in one or two muscle groups is comfortably restocked within 24 hours at moderate intakes. The ISSN's 2018 review places general fitness populations at 3 to 5 g/kg and reserves higher intakes for high-volume or multiple-session training.

During a cut, carbohydrate is the flexible macro: protein gets set around 1.6 to 2.2 g/kg, fat holds a floor of roughly 0.6 to 0.8 g/kg for hormonal and general health reasons, and carbs fill whatever calories remain. That arithmetic usually lands a cutting lifter between 2 and 4 g/kg -- lower than ideal for performance, high enough to keep training productive. Dropping below about 2 g/kg while training hard is where the rep-count data says you start paying a bill every session.

Carb Timing Around Training

Timing is the most over-discussed topic in sports nutrition relative to its effect size. The mechanistic findings are real; the practical significance mostly is not.

Post-workout

Ivy et al. (1988) showed that delaying carbohydrate by 2 hours after glycogen-depleting exercise cut the glycogen resynthesis rate roughly in half during the early window. That finding launched a thousand shaker bottles. Parkin et al. (1997) then showed the part that matters for lifters: with a normal daily carbohydrate intake, muscle glycogen at 8 hours and 24 hours post-exercise is identical whether the first carbs arrive immediately or 2 hours later. The early window changes the trajectory, and the destination is the same by the next session.

On the muscle-building side, Staples et al. (2011) tested whether adding 50 g of carbohydrate to 25 g of post-exercise protein improved net muscle protein balance. It did not -- protein alone maxed the response, and the insulin required to blunt muscle protein breakdown is reached with protein by itself. Aragon and Schoenfeld's (2013) review of the "anabolic window" reached the same conclusion: total daily protein and carbohydrate swamp any timing effect for people training once per day.

Pre-workout

The Henselmans review's clearest positive finding: carbohydrate helps when training fasted or after long gaps since eating. A meal containing 1 to 4 g/kg of carbohydrate in the 1 to 4 hours before training covers this completely. If you train in the early morning without eating and your sessions are long, a small pre-session carb dose (25 to 50 g) is one of the few timing interventions with reliable support.

Intra-workout

Intra-workout carbs earn a spot in exactly one lifting scenario: sessions past roughly 90 minutes of genuinely high volume, or two-a-days. For a normal 60 to 75 minute session preceded by meals, sipping glucose mid-session is fuel for muscles that are not asking for it.

Carb Cycling: Tool or Gimmick?

Carb cycling -- alternating higher-carb training days with lower-carb rest days -- is sold with metabolic language: keeping insulin sensitive, staying "fat adapted" while still fueling sessions, tricking the metabolism out of adaptation. The direct evidence for any of that in lifters is close to nonexistent. No randomized trial has shown better fat loss or body composition from cycling carbohydrate versus a flat intake at the same weekly calories and protein.

The adjacent endurance research gets borrowed as support. "Train low" protocols, including the sleep-low approach of Marquet et al. (2016), showed that deliberately training some sessions with low glycogen can amplify endurance-relevant cell signaling and improved time-trial performance in cyclists. Whatever that means for mitochondria, it has no demonstrated relevance to hypertrophy -- and notably, Camera et al. (2012) found that lifting with low glycogen does not impair the acute muscle protein synthesis response, so neither side of the carb war gets to claim the signaling literature for lifting.

Strip the metabolic story away and a defensible version remains: carb cycling as scheduling. Within a fixed weekly intake, shifting carbs toward heavy training days and away from rest days matches fuel to demand, makes big sessions feel better, and lets rest-day meals lean on protein and volume foods that make a deficit tolerable. Expect it to change your week's experience. Do not expect it to change what the mirror shows at a given weekly calorie and protein intake.

Who Can Actually Get Away With Low Carb

The evidence supports a short list of lifters for whom low carbohydrate costs little or serves a purpose:

• Low-volume strength specialists. Heavy singles, doubles, and triples with long rests run on phosphocreatine. Greene 2018 and Paoli 2012 both show this style surviving keto intact.
• Athletes cutting to a weight class. The glycogen-water drop is a feature when the scale is the opponent, provided there is time to refuel before performing.
• People with medical reasons. Epilepsy management and some metabolic conditions have legitimate clinical keto uses that outrank gym considerations, with a physician involved.
• People whose adherence transforms on low carb. A diet you follow beats a diet you abandon. If cutting carbs kills your appetite and nothing else does, the training cost can be a fair trade during a fat-loss phase.

Outside that list -- and specifically for anyone training with meaningful volume to build muscle -- the trade is bad. You give up reps in the sessions that drive growth, you carry a permanent glycogen-water handicap on every scan and photo, and you gain no fat-loss advantage that a calorie-matched higher-carb diet does not deliver, a result the tightly controlled metabolic ward work of Hall et al. (2016) confirmed when the ketogenic phase produced no acceleration in body fat loss despite the diet swap.

Practical Recommendations

The whole article in one list:

1. Set carbs at 3 to 5 g/kg/day if you lift 4 to 6 hours per week. Go higher only if volume, conditioning, or sport demands it.
2. Never build muscle on keto. Use a surplus with adequate carbohydrate; the one direct trial shows keto forfeits the lean gain entirely.
3. While cutting, protect the floor. Protein first, minimum fat second, carbs get the rest -- and keep them at or above roughly 2 g/kg while training is hard.
4. Eat before training. A mixed meal 1 to 4 hours out. Fasted training plus long sessions is the one context where pre-workout carbs are consistently worth it.
5. Stop optimizing the post-workout clock unless you train the same muscles twice within 8 hours.
6. Use carb cycling for logistics, not magic. More carbs on heavy days, fewer on rest days, same weekly total.
7. Judge low-carb experiments honestly. Discount the first week's weight loss as water, and test your rep counts in week three before concluding anything.

Frequently Asked Questions

Do you need carbs to build muscle?

Strictly, no -- protein synthesis runs on protein and training, and Camera et al. (2012) showed a low-glycogen session still triggers a normal acute response. The cost is indirect: low carbohydrate trims your rep counts on high-volume work week after week, and volume drives growth. In the one direct trial, lifters on keto gained no lean mass while the carb-fed group gained 1.3 kg on the same program. You can build muscle without many carbs the way you can drive across the country in second gear.

How many carbs should lifters eat per day?

Three to five grams per kilogram per day covers most people lifting 4 to 6 hours per week, per the ranges in Slater and Phillips (2011), Escobar et al. (2016), and the ISSN 2018 review. High-volume training or added conditioning pushes toward 5 to 7 g/kg. Cutting usually lands between 2 and 4 g/kg once protein and fat are set.

Is keto bad for lifting?

Bad for building muscle, workable for specific jobs. Maximal strength at low volume survives it (Greene 2018, Paoli 2012), and the rapid weight drop helps athletes making a weight class. For hypertrophy the record is one clear loss (Vargas 2018), one wash with a telling carb-reintroduction spike (Wilson 2020), and a muscle-thickness decline in CrossFit trainees (Kephart 2018).

Should I eat carbs before or after a workout?

Before matters more than after. A mixed meal 1 to 4 hours pre-session covers fuel; fasted lifters benefit most from adding one. Post-workout, glycogen equalizes by 24 hours regardless of timing (Parkin 1997), and carbs add nothing to protein for muscle protein balance (Staples 2011). Rush post-session carbs only when the same muscles work again within about 8 hours.

Does carb cycling work?

It reorganizes your week; it does not change your physiology. No trial shows a body composition advantage over a flat intake at matched weekly calories and protein. Loading carbs onto heavy days is sensible logistics and can make big sessions feel better -- take it as that and nothing more.

Why did I lose weight so fast when I cut carbs?

Glycogen binds about 3 grams of water per gram. Depleting 400 to 600 grams of stored glycogen sheds 1.5 to 3 kg of water in days, which is why week one of low carb looks miraculous and why the weight reappears the moment carbs do. Fat loss still moves at the speed of your calorie deficit.

References

• Bergstrom J, Hermansen L, Hultman E, Saltin B. Diet, muscle glycogen and physical performance. Acta Physiol Scand. 1967;71(2):140-150.
• Tesch PA, Colliander EB, Kaiser P. Muscle metabolism during intense, heavy-resistance exercise. Eur J Appl Physiol Occup Physiol. 1986;55(4):362-366.
• Robergs RA, Pearson DR, Costill DL, et al. Muscle glycogenolysis during differing intensities of weight-resistance exercise. J Appl Physiol. 1991;70(4):1700-1706.
• MacDougall JD, Ray S, Sale DG, McCartney N, Lee P, Garner S. Muscle substrate utilization and lactate production during weightlifting. Can J Appl Physiol. 1999;24(3):209-215.
• Koopman R, Manders RJ, Jonkers RA, Hul GB, Kuipers H, van Loon LJ. Intramyocellular lipid and glycogen content are reduced following resistance exercise in untrained healthy males. Eur J Appl Physiol. 2006;96(5):525-534.
• Leveritt M, Abernethy PJ. Effects of carbohydrate restriction on strength performance. J Strength Cond Res. 1999;13(1):52-57.
• Mitchell JB, DiLauro PC, Pizza FX, Cavender DL. The effect of preexercise carbohydrate status on resistance exercise performance. Int J Sport Nutr. 1997;7(3):185-196.
• Henselmans M, Bjornsen T, Hedderman R, Varvik FT. The effect of carbohydrate intake on strength and resistance training performance: a systematic review. Nutrients. 2022;14(4):856.
• Escobar KA, VanDusseldorp TA, Kerksick CM. Carbohydrate intake and resistance-based exercise: are current recommendations reflective of actual need? Br J Nutr. 2016;116(12):2053-2065.
• Vargas S, Romance R, Petro JL, et al. Efficacy of ketogenic diet on body composition during resistance training in trained men: a randomized controlled trial. J Int Soc Sports Nutr. 2018;15(1):31.
• Wilson JM, Lowery RP, Roberts MD, et al. Effects of ketogenic dieting on body composition, strength, power, and hormonal profiles in resistance training men. J Strength Cond Res. 2020;34(12):3463-3474.
• Greene DA, Varley BJ, Hartwig TB, Chapman P, Rigney M. A low-carbohydrate ketogenic diet reduces body mass without compromising performance in powerlifting and Olympic weightlifting athletes. J Strength Cond Res. 2018;32(12):3373-3382.
• Paoli A, Grimaldi K, D'Agostino D, et al. Ketogenic diet does not affect strength performance in elite artistic gymnasts. J Int Soc Sports Nutr. 2012;9(1):34.
• Kephart WC, Pledge CD, Roberson PA, et al. The three-month effects of a ketogenic diet on body composition, blood parameters, and performance metrics in CrossFit trainees: a pilot study. Sports (Basel). 2018;6(1):1.
• Burke LM, Ross ML, Garvican-Lewis LA, et al. Low carbohydrate, high fat diet impairs exercise economy and negates the performance benefit from intensified training in elite race walkers. J Physiol. 2017;595(9):2785-2807.
• Paoli A, Cancellara P, Pompei P, Moro T. Ketogenic diet and skeletal muscle hypertrophy: a frenemy relationship? J Hum Kinet. 2019;68:233-247.
• Ivy JL, Katz AL, Cutler CL, Sherman WM, Coyle EF. Muscle glycogen synthesis after exercise: effect of time of carbohydrate ingestion. J Appl Physiol. 1988;64(4):1480-1485.
• Parkin JA, Carey MF, Martin IK, Stojanovska L, Febbraio MA. Muscle glycogen storage following prolonged exercise: effect of timing of ingestion of high glycemic index food. Med Sci Sports Exerc. 1997;29(2):220-224.
• Staples AW, Burd NA, West DW, et al. Carbohydrate does not augment exercise-induced protein accretion versus protein alone. Med Sci Sports Exerc. 2011;43(7):1154-1161.
• Aragon AA, Schoenfeld BJ. Nutrient timing revisited: is there a post-exercise anabolic window? J Int Soc Sports Nutr. 2013;10(1):5.
• Kerksick CM, Wilborn CD, Roberts MD, et al. ISSN exercise and sports nutrition review update: research and recommendations. J Int Soc Sports Nutr. 2018;15(1):38.
• Slater G, Phillips SM. Nutrition guidelines for strength sports: sprinting, weightlifting, throwing events, and bodybuilding. J Sports Sci. 2011;29(Suppl 1):S67-S77.
• Marquet LA, Brisswalter J, Louis J, et al. Enhanced endurance performance by periodization of carbohydrate intake: "sleep low" strategy. Med Sci Sports Exerc. 2016;48(4):663-672.
• Camera DM, West DW, Burd NA, et al. Low muscle glycogen concentration does not suppress the anabolic response to resistance exercise. J Appl Physiol. 2012;113(2):206-214.
• Hall KD, Chen KY, Guo J, et al. Energy expenditure and body composition changes after an isocaloric ketogenic diet in overweight and obese men. Am J Clin Nutr. 2016;104(2):324-333.