I approached Peter Attia’s Outlive with trepidation. Books about how to live a longer, healthier life are a cursed genre. Eat blueberries! Don’t eat cheese! Check every food you buy for carrageenan! Follow my specific twelve-step exercise plan! Worry constantly about phthlates! Naively believe every probably-won’t-replicate study IN MICE!
But Outlive outperformed my expectations. It is a realistic, balanced, science-based book that provides useful advice for living a longer, healthier life.
Attia is careful to lay out what we do and don’t know about lifestyle choices for health and longevity. For example, he explains why we can’t rely on nutrition epidemiology: no one accurately recalls whether they eat pizza once a week or twice a week; confounding variables—such as unhealthy people being more likely to choose not to drink alcohol—abound; the studies the media clutches its pearls over have small effect sizes; often, different studies show contradictory effects. He lays out when we can generalize from animal models and when we can’t. He uses new and exciting research methods, like Mendelian randomization.
As you might expect, most of the advice in Outlive is pretty standard: you should exercise regularly, eat fiber, avoid drinking alcohol, get enough sleep, and try to be happy. By far the most interesting part of the book is Attia’s detailed mechanical explanations of how the body works. How are HDL and LDL cholesterol different, and why does one prevent heart attacks while the other causes them? Why does exercise prevent insulin resistance? Why is diabetes correlated with nonalcoholic fatty liver disease? After reading Outlive, I have a better understanding of my body as a physical system, made of chemicals that take up space and go to particular places and trigger specific processes in systems made up of other chemicals.
Attia also has a very… particular… approach to longevity, which many of my readers might find appealing. Intuitive-eating gurus and Attia both agree that every person metabolizes food differently, so advice that works well for one person might not work well for someone else. Intuitive-eating gurus suggest that you should listen to your body, pay attention to your cues and cravings, and notice what meals make you feel good and energetic. Attia, on the other hand, suggests that everyone should do three months of continuous glucose monitoring to see what spikes their glucose.
I like a man who says “you can reverse high cholesterol through diet, but if you think that sounds unpleasant you can just take a statin.”
Attia doesn’t promise miracles: he’s clear that the most important factor in longevity (especially in whether you live for over a century) is genes, not behavior. But everyone can have a longer and healthier old age than they would otherwise. Strength, dexterity, and endurance all decline with age, but they decline as a percentage of your peak abilities: if you’re strong when you’re thirty (able to benchpress your bodyweight), you’re far more likely to be strong when you’re eighty (able to pick up your grandkids).
A few points I found interesting:
Dietary cholesterol has essentially nothing to do with heart disease. If you eat too much cholesterol you just poop it out. The research that allegedly showed that cholesterol causes heart disease was done in rabbits, which (unlike humans) absorb all the cholesterol they eat into their blood.
Metabolic syndrome is (to vastly oversimplify) when your body sends calories to places in your body where they aren’t needed or are even harmful. Only about two-thirds of obese people have metabolic syndrome, while 20-40% of nonobese adults have metabolic syndrome. Metabolic syndrome increases all-cause mortality by about 50%. Conversely, metabolically healthy obese people have no increased risk of mortality. Metabolic syndrome tends to cause fatness—misallocated calories are often stored as fat—but some people are naturally fatter than others and that doesn’t pose a health risk.
Centenarians who die are generally sick and disabled for less time than non-centenarians. The average 90-year-old who will live to 100 is healthier than an average 60-year-old.
The effect of exercise on longevity dwarfs the effect of any other lifestyle choice (other than not smoking). Any amount of exercise is better than no exercise at all. An ideal program includes:
Four 45-minute sessions weekly of moderate cardio, not so hard that you can’t talk but hard enough that you don’t want to.
One hourlong session weekly of cardio as hard as you can: after a brief warmup, alternate between four minutes of exercising as hard as you physically can and four minutes of light walking.
Two hourlong sessions of weight training.
15 minutes daily of mobility and balance work, as well as an hourlong session twice a week.
Falls and other injuries are a common cause of disability and death in old age. Even if you don’t die or experience a long-term disability from the fall itself, the inactivity while you recover can cause you to permanently lose muscle mass. Therefore, it’s very important to reduce your risk of falling when you’re old by doing exercises that improve balance, stability, and grip strength.
While long-term calorie restriction does consistently make every species of animals live longer, it probably wouldn’t work in humans. We study long-term calorie restriction in lab animals, who are protected from injury and disease. If you tried to actually eat 25-30% fewer calories than you naturally would for the rest of your life, you would probably die younger, because you’d get sick or break a bone and not have the calories to heal.
“Nutrition is relatively simple, actually. It boils down to a few basic rules: don’t eat too many calories, or too few; consume sufficient protein and essential fats; obtain the vitamins and minerals you need; and avoid pathogens like E. coli and toxins like mercury or lead. Beyond that, we know relatively little with complete certainty. Read that sentence again, please.”
He later adds “Stop overthinking nutrition so much. Put the book down. Go outside and exercise.”
A caveat: Attia seems to believe that his self-help and popular science book would be more interesting if it were also a memoir. Chapters 1-3 and 17 are entirely skippable.
> Conversely, metabolically healthy obese people have no increased risk of mortality.
I am not sure about this. *Outlive* says this info comes from a meta-analysis, it doesn't cite a source but I believe it's from Hui et al. (2010), "Metabolic syndrome and all-cause mortality: a meta-analysis of prospective cohort studies", based on the fact that the book says the meta-analysis had "a mean follow-up time of 11.5 years" and that was the only meta-analysis I found with that fit that criterion. But I looked over Hui et al. and it doesn't say anything about how obesity or BMI interacts with metabolic syndrome, so I don't see what Attia is basing this claim on.
> Metabolic syndrome is (to vastly oversimplify) when your body sends calories to places in your body where they aren’t needed or are even harmful.
Worth describing how [metabolic syndrome is defined](https://en.wikipedia.org/wiki/Metabolic_syndrome) because it's pretty simple. Someone is diagnosed with metabolic syndrome if they fit at least 3 out of 5 criteria: abdominal obesity, high blood pressure, high blood sugar, high triglycerides, and low HDL.
EDIT: I looked a bit more and found a [different meta-analysis](https://www.jacc.org/doi/full/10.1016/j.jacc.2010.05.034) that reports on BMI and metabolic syndrome (MetS), but only one out of 87 studies actually looked at BMI, that study being [Song et al. 2007](https://sci-hub.hkvisa.net/https://www.ajconline.org/article/S0002-9149(07)01619-0/abstract), "Comparison of usefulness of body mass index versus metabolic risk factors in predicting 10-year risk of cardiovascular events in women". And that one study contradicts Attia's claim. Namely, it found that obese women without MetS had more CVD than normal-weight women without MetS. (It didn't look at all-cause mortality.) Although it's still true that if you control for MetS, most of the harm of obesity disappears.
Here's the table of CVD incidence by BMI for MetS and no-MetS, respectively:
- BMI <25: 7.9% and 2.2%
- BMI 25-29.9: 7.1% and 2.4%
- BMI 30+: 6.1% and 2.6%
"Chapters 1-3 and 17 are entirely skippable."
Lol this is the part I need most from any review. "When does the filler end and the useful part start?"
So many books could have been pamphlets or blog posts.