235‒Training principles for mass & strength, changing views on nutrition, & creatine supplementation

KEYPOINTS

• 01

  Strength and fitness are among the strongest predictors of mortality.

  Attia lines up the hazard ratios for all-cause mortality. Hypertension adds about 20%, type 2 diabetes about 30% and smoking about 50%. Being weak rather than strong adds about 250%. Having a VO2 max in the bottom 25% versus the top 2.5% adds about 400%. Muscle mass also matters but is tightly linked to strength. Norton adds that lifting is the best thing for bone density, and that strength and balance help prevent the fall in the first place.

• 02

  Progressive overload is the core principle, and many rep ranges build muscle.

  For a beginner, adding 5 lb a week to a squat (95, 100, 105 for the same reps) is progressive overload. Sets of 10–15 build lean mass and therefore strength. But strength is a specific skill, so low-rep, heavy practice is needed to express a true one-rep max. Ahead of his best squat of 668 lb, Norton practiced heavy singles, mostly around 90% and occasionally 95%.

• 03

  Use RPE, but learn what real failure feels like.

  On the rating-of-perceived-exertion scale, RPE 10 means no reps left and RPE 9 means one left. Norton cites research in which beginners and intermediates misjudged their RPE by about five when researchers then pushed them to true failure. He has shifted toward more sets at RPE 6–7 rather than grinding at RPE 8–9, which he feels lowers injury risk while still adding volume as overload. Using variations of the competition lifts reduces the anxiety of being 'tested' against previous numbers.

• 04

  Super-slow training works, but it isn't magic.

  Attia describes patients who will only do one weekly 20–30-minute super-slow session: about 8–10 machines, one set each to failure lasting roughly 90–105 seconds. Norton says there are 'many ways to skin a cat' for hypertrophy. He critiques a slow-rep study that used 30% of one-rep max for 10 reps, when a normal pace at that load allows 30–50 reps, so the comparison wasn't fair. Doing any training beats doing none.

• 05

  Testosterone matters even if relative gains look similar between sexes.

  Some cite a study showing men and women adding about 10% lean mass to argue testosterone doesn't matter. Norton points out that 10% of 70 kg is 7 kg, while 10% of 50 kg is 5 kg, and the absolute loads differ hugely. Ray Williams, the first drug-tested man to squat over 1,000 lb raw, carried about 308 lb of lean mass. Norton calls his own view on testosterone thresholds speculation.

• 06

  Creatine monohydrate is the lowest-hanging supplement fruit.

  Phosphocreatine donates phosphate to regenerate ATP. Creatine is also an osmolyte that pulls water into muscle, which may itself be anabolic. Norton takes 5 g daily even before meets. Loading (e.g., about 30 g a day for a week) saturates muscle faster but often causes GI bloating and nausea, and over the long run it doesn't matter. He hasn't seen clear evidence that more than about 5 g helps. The hair-loss worry traces to one 2009 study showing a DHT rise that has never been replicated.

• 07

  Norton changed his mind on LDL, BCAAs and fasting.

  He once favored HDL/LDL ratios and particle size. Mendelian randomization data persuaded him that LDL is causal and an independent risk factor, and he now takes a low-dose statin. His LDL had stayed around 125–150 even with less saturated fat and 60–70 g of fiber a day. He dropped BCAA products because whey beats BCAAs even at matched doses. Grant Tinsley's studies showed no significant lean-mass difference with 16:8 fasting when people train within the eating window, though bodybuilders may still prefer frequent protein.

• 08

  Eat fiber from real food: about 15 g per 1,000 calories.

  Norton gets about 60 g of fiber a day and says the evidence linking fiber and fruit and vegetable intake to lower heart disease and cancer risk is unusually consistent across meta-analyses. He cites a 2020 analysis showing that higher vegetable intake was linked to lower risk. Fiber bars count for something, but 40 g from three bars isn't equal to six to eight servings of produce. He notes the irony that his research was funded by egg, dairy and beef groups, yet he is defending plants against carnivore advocates.

• 09

  Tracking food intake reveals what you're really eating.

  Restaurant fat can turn a 500-calorie dish into 1,500 calories. Clients who swear they eat 1,500 calories often discover, once they track, that it's about 2,800, which Norton calls 'freedom.' Weighing even a brownie (about 40 g) acts as a governor on extra servings.