Is Training Safe in Wildfire Smoke?

Somewhere between five and 150 micrograms per cubic meter, the sky turns orange and a familiar question arrives: should the workout happen anyway? Barbell Medicine's Jordan Feigenbaum and Austin Baraki spend the first half of this episode answering that question with actual numbers instead of vibes, and the second half doing the same thing for a completely different piece of gym folklore: the idea that only sets taken within five reps of failure count for anything.

Start with the smoke. The relevant particle is called PM2.5, meaning anything smaller than 2.5 microns across. That is small enough to slip past the nose and throat and travel deep into the small airways and, they believe, into the bloodstream (03:52). On a clean day that reading sits around five micrograms per cubic meter. Wildfire smoke can push it past 100, sometimes past 150 (03:52). The number on a weather app, the Air Quality Index or AQI, converts that raw concentration into a 0 to 500 scale, with anything under 50 read as good air and 50 to 100 as moderate (04:11).

A 2016 paper, often cited in these conversations and referred to here as the Teneo paper, modeled years of city commuting and tried to find a tipping point where pollution exposure would cancel out the benefit of exercise. It landed on 90 minutes of cycling at 100 micrograms per cubic meter, though Feigenbaum notes the paper actually states that number three different ways across its own abstract, discussion, and a follow-up review, ranging from just over an hour to an hour and a half (05:18). It is a simulation built on city pollution, not a wildfire day, and its own authors admit it never modeled short spikes in exposure.

A more recent study, published last year in BMC Medicine, is more useful. It pooled seven cohorts, about 1.5 million adults followed for a median of 12 years, and asked what happens to the health benefit of exercise as pollution rises. Below 25 micrograms per cubic meter, roughly an AQI of 80, hitting standard exercise guidelines cut all-cause mortality by about 30% (10:38). Above that, in air worse than Los Angeles on an average day, the same amount of exercise still helped, but the benefit dropped to roughly 12 to 15% (10:38). As Feigenbaum puts it:

"The AQI, Air Quality Index of 80, is where the benefit roughly halves. We just don't have any evidence that exercise ever becomes harmful in a particular environment." — Jordan Feigenbaum [10:18]

That halving point matters, but so does what happened when researchers tested the opposite intervention. A separate model looked at six cities, including Beijing and New Delhi, and asked what would happen if active commuters switched to transit or stayed home on high-pollution days. Restricting activity did not lower mortality in a single city. In Beijing and Delhi, the model suggested mortality actually rose because people moved less (11:40).

Why lifting is a different animal than cardio

The reason lifting gets an easier pass than running or cycling comes down to minute ventilation, the total volume of air moved in and out of the lungs each minute. At rest, that's about 10 liters per minute. Walking brings it to roughly 23, and cycling to about 43 (12:26). Lifting, even at high effort, stays much closer to resting or walking levels. One study Feigenbaum cites had both healthy people and people with COPD (chronic obstructive pulmonary disease) do leg extensions at 40% and then 80% of their max; perceived effort rose sharply, but breathing rate barely moved (12:26). As Feigenbaum summarizes:

"Minute ventilation, the total volume of air that you move in and out in a minute, generally is lower with lifting." — Jordan Feigenbaum [12:20]

Indoors isn't a perfect shield, though. During recent California wildfires, researchers monitoring more than 1,000 buildings found that indoor particulate matter nearly tripled even with people staying inside, simply because outdoor air got bad enough to seep in (13:50). And a Scripps-affiliated analysis of 1.6 million respiratory hospital admissions found that city pollution raised admissions by under 1.5% per 10-microgram increase, while wildfire smoke pushed that figure anywhere from 1.3% up to 10%, a range wide enough that Feigenbaum cautions it depends heavily on the size of the fire and proximity to it, a point Baraki extends by adding wind direction (14:48).

Baraki's clinical read is that the sky's color isn't the deciding factor: a person's own vulnerability is. Someone with a history of asthma flares or COPD exacerbations during past smoke events should lean toward indoor exercise rather than skipping it, since staying inside and doing nothing is likely worse than staying inside and moving (15:42). For a healthy person facing a short, temporary smoke event, Feigenbaum argues the stakes are low either way, but Baraki pushes back, warning that missing a week can lead to months of lost momentum. What actually worries them is a week off turning into a lost habit, a bigger risk to long-term health than a few smoky lifting sessions ever was (17:54).

The other hard line: training to failure

The second question drags a different piece of gym dogma into the light: the claim that only sets taken to within five reps of failure, called reps in reserve or RIR, actually build muscle. A listener points out the contradiction. If that rule is strict, the triceps working as a secondary mover during a heavy bench press shouldn't count for anything, since they never reach that failure threshold. But lifters routinely count that triceps work as a fractional set anyway, which implies warm-up sets should count too.

A 2024 meta-regression, a study that pools data from many trials to map the shape of a relationship rather than just testing whether one exists, pulled together 140 results from 26 studies on muscle growth. The best-fit line was close to straight, meaning the growth stimulus runs continuously from failure all the way out past 20 reps in reserve, with no sharp cliff at any particular number (27:32). A separate analysis of 67 studies and about 2,000 people tested whether indirect, secondary-muscle sets should count as zero, half, or a full set toward volume, and half fit the data best (28:10).

Feigenbaum's takeaway is that any rigid cutoff is wrong in both directions:

"Any hard line sort of interpretation of the effective reps model is wrong, meaning, like, you gotta take every set to five repetitions in reserve or closer, otherwise it doesn't count." — Jordan Feigenbaum [28:35]

His practical split: for hypertrophy work on machines and isolation exercises, aim for an RPE (rate of perceived exertion, a 1 to 10 effort scale) of 7 to 9 (35:13). For strength work, once the load is heavy enough, roughly 60 to 70% of a one-rep max, staying farther from failure at RPE 4 to 7 tends to preserve bar speed and force quality (35:38).

Baraki, for his part, is less interested in the exact number. His closing point ties both halves of the episode together:

"The differences in inter-individual response to training dwarf the differences in impact of any one or combination of these decisions." — Austin Baraki [36:17]

That line could apply just as easily to the smoke question as to the failure question. Both conversations end in the same place: broad thresholds exist, and they matter more than any single day's decision. Whether the air is bad or the last rep is hard, the bigger risk isn't the exposure itself. It's letting one uncertain day talk someone out of the habit altogether.