Hi everyone,
In 315 adults, packing all 75 weekly minutes of interval training into a single session cut body fat almost as much as spreading it across three days. If time is the thing keeping you from exercising, that is useful. A few other studies I thought were worth your time: adults over 75 who stopped their preventive statin were no more likely to die three years later than the ones who kept taking it, three servings a day of full-fat dairy did not move cholesterol or body fat at all, an AI that read 10,000 sleep studies found the apnea number you get handed does not predict death while the discarded signals do, and a self-guided pain program beat sitting with a therapist on every outcome measured. Plus a podcast on alpha lipoic acid.
This Week’s Podcast Spotlight
Episode 67: Alpha Lipoic Acid: Real Help for Diabetic Nerve Pain and Aging Mitochondria
Alpha lipoic acid is a compound your own mitochondria build, and it sits on pharmacy shelves everywhere labeled as an antioxidant. When I went to the literature for this episode, I found I had it almost completely backwards. An oral dose has a half-life of about 30 minutes and is essentially gone from your blood inside 90, which is nowhere near enough time to mop up free radicals across trillions of cells, and yet the effects can last days or weeks. What it actually does is pull a cellular alarm called NRF2, which switches on hundreds of your own antioxidant, detoxification, and repair genes, and flip AMPK, the same fuel gauge that a lot of the benefits of fasting and exercise run through. The strongest human evidence sits in diabetic nerve pain, where Germany has approved it as thioctic acid for decades and the randomized trials going back to ALADIN in 1995 landed on 600 mg a day as the sweet spot.
Three points from this one:
- It is a signal, not a scrubber. The molecule clears your bloodstream in 90 minutes, but the defense genes it switches on through NRF2 keep working for days after it is gone.
- Your body already makes it. An enzyme called lipoic acid synthase builds it inside your mitochondria, and that production falls with age and metabolic stress, which is why researchers call it conditionally essential rather than essential.
- The weight loss data is real but small. In a trial of 360 adults with obesity, the higher dose group lost about 2 percent more weight than placebo over 20 weeks, nowhere near what the GLP-1 medications do.
This Week in Health Science
Five studies from this week’s research, and I think a few of them are practically useful.
One 75-Minute Workout a Week Cut Body Fat Almost as Much as Three
The question most people actually want answered is not what is optimal, it is what is the least they can get away with. Researchers in Hong Kong randomized 315 adults with belly fat to the same weekly dose of high-intensity interval training, 75 minutes, and only changed how it was packaged: one long session or three shorter ones. After 16 weeks, measured by DXA scan rather than a bathroom scale, the once-weekly group carried 0.8 kg less body fat than the control group and the thrice-weekly group carried 1.0 kg less. The 0.3 kg gap between the two schedules could easily be chance. Waist size and fitness improved in both training groups, and there were no study-related adverse events in a group of previously inactive adults, though the first four weeks were a graded build-up rather than a cold start into hard intervals. Two things temper it. The absolute fat loss is modest, so nobody is seeing a transformation in the mirror. And by week 32, four months after training stopped, the fat differences had faded while the smaller waists held. This is a habit to keep, not a course to finish.
Key finding: In 315 adults with central obesity, one 75-minute interval session a week cut body fat by 0.8 kg over 16 weeks (P = 0.0107) versus 1.0 kg for the same minutes split across three sessions, with no meaningful difference between the two schedules (P = 0.36).
My book, 2mm, is almost here. It is the story of seven years inside the operating room, where every decision is measured in millimeters and every millimeter matters. If you want to be the first to know when it is available, leave your email at 2mmbook.com.
Stopping a Statin After Age 75 Was as Safe as Staying On It
French researchers recruited 1,160 adults aged 75 and older from 297 primary care offices, all of them on a statin purely to prevent a first heart attack or stroke, and randomly assigned them to keep taking it or stop. Three years later, 7.2 percent of those who stopped had died, compared with 7.9 percent of those who continued. Side effect rates were within a percentage point of each other. Median age was 80, two thirds were women, a third had diabetes, so this was not a cherry-picked group of unusually healthy seniors. Statins earn their place in a specific group, mostly people who already have established cardiovascular disease, and results like this one are a reminder of how far past that group the prescribing has drifted. Older patients accumulate medications the way a garage accumulates boxes, and nobody wants to be the person who takes one out. The population studied here is a defined one: no heart attack, no stroke, no known blockage disease, already 75 or older. If you have had an event, the evidence for staying on the statin is strong and this study says nothing about you. Three years is also short for a drug that works over decades, so a 76 year old with 18 more years ahead was not really represented. Bring it up at a visit rather than acting on a headline.
Key finding: In 1,160 adults aged 75 and older taking a statin for primary prevention, three-year mortality was 7.2 percent after stopping versus 7.9 percent while continuing, a difference of 0.68 percentage points that never approached the 5 percent harm threshold set before the trial began.
Three Servings of Full-Fat Dairy a Day Did Not Raise Cholesterol or Body Fat
Whole milk, cheese, and yogurt have been on the avoid list for decades, and that advice was built on what saturated fat does to cholesterol in isolation rather than on what happens when people eat actual food. Researchers at the University of Toronto put 74 adults with overweight or obesity into three groups for 12 weeks: a 500 calorie deficit with almost no dairy, the same deficit with three daily servings of full-fat dairy, or three servings of full-fat dairy with no calorie restriction at all. Fasting cholesterol did not move. Fat mass did not move. Resting metabolic rate did not move. The group eating full-fat dairy without a calorie cut saw systolic blood pressure drop 2.72 points, which is small enough that I would not sell it as a benefit. The part worth underlining is that only the dairy groups raised their protein and calcium intake. Trading a food that reliably delivers both for one that does not is a bad trade if the fear driving it does not hold up. The honest headline here is a null result, and null results about foods people have been told to fear are worth reading carefully. Twenty-five people per group over 12 weeks tells you about cholesterol and body composition, not about heart attacks or fractures years from now, and if you have been told to limit saturated fat for a specific lipid condition, that advice still stands.
Key finding: Over 12 weeks in 74 adults with overweight or obesity, three daily servings of full-fat dairy produced no change in fasting cholesterol, fat mass, fat-free mass, waist circumference, or resting metabolic rate, while raising protein and calcium intake that the low-dairy diet did not.
An AI Read 10,000 Sleep Studies and Found What the Apnea Score Misses
If you have had an overnight sleep study, you probably walked out with one number, the apnea-hypopnea index, which counts how often your breathing stops or gets shallow. That single count decides whether you are labeled mild, moderate, or severe, and whether you get a CPAP machine. The recording itself captured brain waves, heart rhythm, muscle tone, airflow, effort, and oxygen all night, and nearly all of it gets discarded once that one number is calculated. Cleveland Clinic researchers trained a transformer model on the raw signals from 10,000 in-lab studies and let it sort patients into five groups. The highest-risk group had 171 percent higher risk of death than the lowest over a median 15 years of follow-up, along with more atrial fibrillation, more heart attacks, and more cognitive impairment. The apnea severity categories those same patients were assigned predicted death not at all. In one analysis excluding CPAP users, severe apnea carried a hazard ratio of 1.05, which is essentially nothing. A low score reassures people and a high one alarms them, and this suggests both reactions can be wrong, because the high-risk group included people from every severity band. This is observational and the model is not something any lab can run for you today. What you can do is ask for the full report rather than the summary number, and stop treating a normal apnea score as a clean bill of health for your heart and brain.
Key finding: Across 10,000 in-lab sleep studies, AI-derived risk groups separated all-cause mortality by a hazard ratio of 2.71 (95% CI 1.93 to 3.81), while conventional mild, moderate, and severe apnea categories showed no significant link to death.
A Self-Guided Pain Program Beat Sitting With a Therapist
Cognitive behavioral therapy for chronic pain teaches skills rather than treating tissue: pacing activity, calming the alarm response, interrupting the thought spirals that make pain feel bigger. It is first-line, drug-free, and most people who would benefit never get near it because the appointments do not exist. This trial randomized 764 veterans across 9 VA health systems to either 11 weeks of a self-directed program, where an automated phone system collected daily practice reports and a coach sent personalized audio feedback weekly, or standard therapist-delivered sessions. At 4 months the self-directed group scored 5.26 on pain interference against 6.23 for the therapist group, where lower is better. The lead held at 6 and 12 months, and the self-directed group beat the therapist group on every secondary outcome, including pain intensity, catastrophizing, sleep, and depressive symptoms. This result went against my instinct, which says a trained clinician in the room should win. What the data show is that the self-directed group completed more of their expected sessions. Skills only work if you practice them, and daily reporting plus weekly feedback keeps practice on the calendar in a way a biweekly appointment does not. The 0.98 point gap sits right at the threshold researchers defined as the smallest difference a patient would notice, not past it, and the trial was open-label in a veteran population. If you have a therapist you attend consistently, keep them. If you are on a waitlist or keep dropping out, this is worth asking about.
Key finding: In 764 veterans with chronic musculoskeletal pain, self-directed CBT scored 5.26 on pain interference at 4 months versus 6.23 with a clinician, a 0.98-point advantage (95% CI 0.65 to 1.31, P < .001) sustained through 12 months.
Stay curious. Stay skeptical. And stay healthy.
Dr Kumar