Top Exercise Scientist Changes his Mind on Maximum Protein, Fasting, & Training to Failure

Brad emphasizes that the old dogma of every set to failure with forced reps is unnecessary. The evidence base, while still emerging, consistently shows that staying 2–3 reps away from failure produces comparable muscle growth. He notes that all available studies are all-or-nothing (all sets to failure or none), and none have followed elite lifters near their genetic ceiling, where the benefit of occasional failure might be larger. He also distinguishes between single-joint and multi-joint exercises: going to failure on a lateral raise is less taxing than on a squat, so it might be safer to push isolation exercises closer to failure. Crucially, Brad insists that you must learn what true failure feels like by occasionally reaching it, otherwise you’ll overestimate your RIR and undertrain. The protocol is not about training easy; you still need to work very hard, just not to absolute concentric failure.

Brad’s view has shifted from 'always full ROM' to 'the stretched position is the most critical.' He cites studies showing that lengthened partials produce comparable hypertrophy to full ROM, while shortened partials blunt growth. He notes that John Meadows speculated about static holds in the lengthened position as a post-failure strategy, and early evidence hints this may augment hypertrophy. Brad suggests that exercises like 21s, which include a shortened portion, might be unnecessary because that range adds little. The practical application is to prioritize exercises that load the muscle in its most stretched state (e.g., dumbbell curls with supination, full-depth squats for quads, Romanian deadlifts for hamstrings) and, when needed, perform only the bottom half of the movement. He cautions that for strength athletes, full ROM is still necessary for complete strength development.

Brad outlines that the old rule of 30g per meal was based on acute oxidation studies that didn't account for whole-food meals or the body’s ability to use amino acids over extended periods. The famous 100g vs 25g study showed prolonged MPS, but it used a single bolus and didn't compare to 50g. Practically, Brad finds the protein cap irrelevant because real meals with fiber, fat, and other nutrients slow absorption, making even large protein servings usable. He suggests that for most people, three meals (breakfast, lunch, dinner) with adequate protein suffice. Bodybuilders might benefit from a fourth meal or a slow-digesting protein before bed, but the evidence for that is weak. The key is to not stress over meal frequency—total daily protein is what moves the needle.

Brad’s review of the literature suggests BFR is a promising tool, particularly for type I fiber growth. Because type I fibers are more endurance-oriented, the low-load, high-rep nature of BFR combined with metabolic stress may preferentially stimulate them. Some studies show mild to moderate hypertrophy benefits beyond what is achieved with the same loads without restriction. However, BFR cannot replace heavy training for maximal strength. It is best used as an adjunct, especially for individuals who need to reduce mechanical strain on joints. He also discusses the theory that BFR-induced swelling does not stretch the fascia to allow more growth (like synthol claims); there’s no evidence for that. The real benefit likely lies in metabolic stress and increased fiber recruitment.

Brad’s earlier belief that training a muscle twice a week is always superior has evolved. He now sees that the need for higher frequency depends on volume. He cites evidence that at low to moderate volumes, once-weekly training can be adequate, but when per-session volume exceeds about 10 sets, splitting is likely beneficial. The hypothesis is that there is a ceiling on the muscle protein synthetic response to a single bout, and exceeding it adds diminishing returns. By splitting volume across days, you can stimulate MPS multiple times while maintaining session quality. He says 20 sets per muscle per week, done as 10 on Monday and 10 on Thursday, is better than 20 sets in one bro-split session.

Brad's master’s student Tommy Herman led a study that examined pre-exhaust supersets (e.g., leg extension immediately followed by squat). The theory, originally from Arthur Jones, was that pre-fatiguing the target muscle would make it fail before the synergists or lower back in the squat, thus achieving deeper inroading. However, the study showed no hypertrophic benefit; in fact, there was a slight negative trend across all muscles measured. Brad attributes this to the compromise in mechanical tension: the load on the compound movement was significantly reduced, which attenuated the primary driver of hypertrophy. He notes that while not a large detriment, the evidence suggests that pre-exhaust is not the best strategy.

Brad clarifies that the only context where immediate post-workout carb timing matters is if you’re training again within the same day. For example, a soccer player with morning practice and an evening match would benefit from rapid post-morning carbs to restore glycogen. He notes there’s a ~50% reduction in glycogen repletion if you miss the 2-hour window, but by the next day, the deficit is erased. For bodybuilders doing Arnold-style double splits, rapid carb timing is relevant. For the vast majority of lifters doing one session a day, there’s no need to rush carbs post-workout. This aligns with the overall theme that total daily intake trumps acute timing.

Brad explains he was 'bought Hookline and Sinker' into the anabolic window of opportunity, believing that missing the post-workout period would compromise gains. He details how his colleague Alan Aragon challenged him in 2013 to revisit the literature, leading to their collaboration on review papers and meta-analyses. The research showed that muscle remains sensitized to protein for up to 24 hours after a workout, and that pre-workout nutrient consumption strongly influences the urgency of post-workout intake. Therefore, the total daily intake of protein and carbs is what matters most. He does concede that for competitive bodybuilders, consuming protein shortly after training might confer a very modest benefit, but for the average lifter, it won't make a noticeable difference. He also points out that the original theory was based on acute studies that didn't account for pre-exercise feeding, which most people do. The 'catabolic' scare was overblown. Now he finds it liberating not to obsess over the clock after training.

Brad states he has done a '180' on this. The overwhelming body of evidence now shows muscle growth can be achieved with loads from about 5 reps up to 30 reps, as long as sets are taken close to failure. He notes that staying primarily in the classic 8–12 rep range is still efficient, but adding periods of higher reps (lighter loads, less joint stress) and lower reps (heavier loads, greater mechanical tension if volume is matched) may provide a more complete stimulus. He cautions that the jury is still out on whether there are unique benefits to each range, but there is no evidence that mixing rep ranges is detrimental. This shift also has practical implications for those with joint issues, who can use higher reps without sacrificing gains.

Brad recounts the hormone hypothesis was popular when he was a student. The idea was that post-exercise surges in testosterone, GH, and IGF-1, while transient, drove the hypertrophic response. However, emerging research indicates correlation does not imply causation. The acute hormonal response has not been shown to have a meaningful impact on long-term muscle growth. He now believes that even if there is a small effect, it should not dictate program design. He also notes the irony that very short rest periods (a hormone spike tactic) actually compromise hypertrophy by reducing volume load, because you can't complete as many reps in subsequent sets. This underscores that mechanical tension, not hormonal floods, is the key stimulus.

Brad explains that the emerging literature shows training to failure is not necessary for optimal gains. Studies comparing all sets to failure versus no sets to failure consistently find similar hypertrophy when the proximity to failure is close (2–3 RIR). However, he stresses that as one becomes more trained, going too far from failure (e.g., 5+ reps shy) will blunt results. He also notes limitations: studies are often all-or-nothing (all sets to failure or none), and they haven't examined elite lifters near their genetic ceiling. Brad suggests that single-joint exercises can be taken to failure more safely than multi-joint exercises, because the fatigue cost is lower. He emphasizes that lifters must learn to gauge failure by occasionally training to failure, especially early in their careers, to accurately estimate RIR. The key is to train hard, close to failure, but not necessarily beyond it every session.

Brad describes being a product of that mentality, eating six meals regimented every two to three hours. He now points to evidence that protein above 30g is not oxidized; rather, it can still be utilized for muscle protein synthesis over a longer period. He references the recent study where 100g of protein led to prolonged MPS elevation vs 25g, though he notes the limitation that 100g wasn't compared to 50g. More importantly, the ecological validity of such findings is questionable because real meals (chicken, rice, etc.) slow digestion, further extending amino acid delivery. He also discusses that as long as one consumes 3–4 meals with relatively evenly spaced protein, there is no meaningful difference. For bodybuilders, a pre-bed protein shake might offer a small edge, but for most people, not stressing over meal frequency is liberating.

Brad mentions that his group published a study showing greater growth with a total body routine compared to a split routine. He notes there are limitations to drawing causal conclusions, but the evidence overall does not support any inherent superiority of splits. Instead, the advantage of splits emerges when you need to fit in higher weekly volumes—if you're doing over 10 sets per muscle in a session, splitting those sets across days may be beneficial. However, for moderate volumes, total body training works just as well. He also adds a caution about novelty: while changing routines can be beneficial, too much variation (e.g., changing exercises every session) can hinder performance tracking and skill development, so a balance is needed.

Brad clarifies that research now clearly shows full ROM beats partials in the shortened range. However, when partials are performed in the initial (lengthened) range—the first half of the range of motion where the muscle is stretched—hypertrophy is similar to that from full ROM, assuming comparable effort. He cites evidence in arms, legs, and calves, but notes that data are limited to certain muscles and exercises. His colleague Brett Contreras still believes the glutes may benefit from shortened-position work. Importantly, for strength, angular specificity applies: you get stronger in the range you train. So for pure hypertrophy, emphasizing the lengthened portion may suffice, but strength across the full range requires training through it. He also mentions that the classic 21s (top, bottom, full) might be unnecessary since the shortened portion likely doesn't add much, while post-failure static holds in the stretched position could augment growth. This reflects a nuanced view: not all partials are equal, but the stretch is king.