Cysteine, Census Counts and a Medicare Slack · Morning Gist
Saturday, October 105 stories · 16 min
Cysteine, Census Counts and a Medicare Slack
A doctor's view of GLP-1s in clinic, two findings from aging research, and two policy moves that could reshape U.S. health apps and data.
Picked from 31 new articles across 12 sources · arrow keys for other days
Today in 30 seconds
30s left
Adam Cifu describes how GLP-1 drugs changed his primary care practice, including who he now treats and where he goes beyond trials. story 1 ↓
In mice unable to make cysteine from methionine, cutting methionine did nothing; cutting cysteine delivered the metabolic benefits linked to longer life. story 2 ↓
A 1,700-member government Slack, mostly industry, is shaping Medicare's AI app policy; officials called a planned app directory a "sales engine." story 3 ↓
Cancer patients' healthy tissue may carry more mutations than other people's, hinting that faster mutators could be flagged for earlier screening. story 4 ↓
Proposed 2030 census rules would count only citizens and green-card holders and drop race questions, distorting disease rates; comments close Nov. 2. story 5 ↓
RFK Jr. announced a federal program to identify and treat vaccine injuries; experts worry it will overstate rare harms.
1/5Sensible Medicine· Adam Cifu, MD· Oct 9
A primary care doctor on what GLP-1 weight-loss drugs changed in practice
Internist Adam Cifu reflects on nearly ten years of prescribing GLP-1 drugs. Some patients keep weight off after stopping, he now raises treatment with more people, and patients no longer say "I don't overeat."
Why it matters · It shows how an evidence-minded doctor uses these drugs day to day, including where his own judgment goes beyond what trials have shown.
Know this first
Context from the editor, not the article · 31 sec
GLP-1 receptor agonists copy a gut hormone called glucagon-like peptide-1. They slow stomach emptying and reduce appetite. They were first used for type 2 diabetes. Semaglutide (Ozempic, Wegovy) and tirzepatide (Mounjaro, Zepbound) produce much larger weight loss than older obesity drugs. Most trials report average losses of roughly 15–20% of body weight. Studies also find that most people regain much of that weight after they stop the drug. Body mass index (BMI) is weight in kilograms divided by height in meters squared. A BMI of 30 or more counts as obesity. Before these drugs, the main options were modest medications, bariatric (weight-loss) surgery, and advice about diet and exercise. That advice rarely worked over the long term.
Image: Sensible Medicine
2/5Fight Aging!· Reason· Oct 9
Lifespan benefits of low-methionine diets may come from cutting cysteine
A new open-access paper uses worms and genetically engineered mice to separate two sulfur-containing amino acids. Cutting cysteine, not methionine, produced the metabolic and lifespan benefits long credited to "methionine restriction."
Why it matters · If cysteine is the real lever, researchers testing dietary or drug ways to mimic calorie restriction may be aiming at the wrong amino acid.
Know this first
Context from the editor, not the article · 31 sec
Calorie restriction extends lifespan in many lab animals. One leading explanation is that it boosts autophagy, the cell's recycling of damaged parts. Cells detect scarcity through sensors for particular nutrients, and essential amino acids are among the most important signals. For decades, researchers have reported that diets low in methionine, an essential amino acid, reproduce many effects of calorie restriction in rodents, even when the animals eat more food. Methionine and cysteine are the only two sulfur-containing amino acids used to build proteins. Cysteine is "non-essential" because the body can make it from methionine through a series of reactions called the transsulfuration pathway. Animal foods are rich in both amino acids. Plant foods contain less of them.
Image: Fight Aging!
3/5KFF Health News· Amanda Seitz, Maia Rosenfeld and Darius Tahir· Oct 9
Tech firms shape Medicare AI policy through an unannounced government Slack
KFF Health News reviewed thousands of messages from a Medicare-run Slack workspace. Tech companies used it to lobby officials for federal payment and easy access to patient records, with little public input.
Why it matters · Seniors may soon get AI health apps that Medicare promotes and pays for. Those apps could pull medical records broadly, and nobody has to show they work.
Know this first
Context from the editor, not the article · 28 sec
CMS (the Centers for Medicare & Medicaid Services) runs Medicare, the federal insurer for about 68 million older and disabled Americans. Over the past year the Trump administration has pushed a "Health Technology Ecosystem" plan. It aims to let patients send their medical records to apps of their choice and to bring AI tools into routine care. Two other pieces matter here. The FDA (Food and Drug Administration) can regulate some health software as a medical device. And the Federal Advisory Committee Act requires outside groups that advise agencies to meet in public and include a fair balance of viewpoints, so that industry can't quietly capture policy.
The workspace's code of conduct, which shows CMS had quietly set up a Technology Working Group.1 / 2
4/5Fight Aging!· Reason· Oct 9
Cancer patients' healthy tissue may carry more mutations than other people's
A study in Cancer Research found cancer cells carry about four times more DNA mutations than normal cells. The normal tissue of cancer patients also carried more than that of people without cancer, suggesting some people mutate faster.
Why it matters · If some people build up mutations faster from birth, measuring that rate could help pick out who needs earlier or more intensive cancer screening.
Know this first
Context from the editor, not the article · 31 sec
Cells pick up small DNA changes, called somatic mutations, every time they divide. The longer you live, the more of them build up. In tissues that keep renewing themselves, such as skin, gut lining and blood, a mutated stem cell can pass its changes to large patches of descendant cells. The body ends up as a patchwork of genetically slightly different clones, called somatic mosaicism. Cancer usually starts when one cell collects the wrong mix of "driver" mutations, which push growth. That is one reason cancer becomes more common with age. Researchers have long known tumours carry far more mutations than the normal tissue they grow from, but it has been unclear why the gap is so large.
Image: Fight Aging!
A new paper in Cancer Research compares mutation counts in normal and cancerous cells from five organs. Cancer cells carried about four times as many mutations as non-cancerous cells. The summary available doesn't say which organs were studied, how many people were sampled, or exactly how the cells were sequenced.
Three ways to explain the gap
The authors tested three hypotheses:
5/5STAT· Nancy Krieger· Oct 9
Epidemiologist warns proposed 2030 census rules would distort U.S. health statistics
Harvard epidemiologist Nancy Krieger argues that proposed changes to the 2030 census would shrink and shift population counts. Disease and death rates would then appear to jump, and racial health gaps would stop being measured.
Why it matters · Nearly every U.S. disease rate, health disparity figure and survey weight uses census counts as its denominator, so changing the count changes what the data show.
Know this first
Context from the editor, not the article · 27 sec
Every ten years the U.S. Census Bureau tries to count every resident, as the 14th Amendment requires. Its mission is to count "everyone once, only once, and in the right place." The count decides how House seats are split among states and how districts are drawn. It also feeds many federal funding formulas. In public health, census figures are usually the denominator of a rate: deaths or cases divided by the population at risk. The census also sets the sampling frames and weights that national health surveys use to make their samples represent the whole country. Race has been asked on every census since 1790.
Adam Cifu, an 1 who writes for Sensible Medicine, is looking back on almost ten years of wide 2 use. He makes clear that this is a reflection. He doesn't review trials, and everything below comes from his own clinic, not from controlled data.
Why he ranks them highly
Cifu doesn't call GLP-1s as revolutionary as statins or HIV protease inhibitors. But because obesity is so common, he thinks their overall impact is comparable. The first drug in the class, 3 (2005), came from a peptide in Gila monster venom. It had to be injected twice a day and caused so much nausea that he almost never prescribed it. That changed with 4, a long-acting version approved in late 2017, and 5 (2022). Tirzepatide also acts on a second gut hormone, 6 (glucose-dependent insulinotropic polypeptide).
He is blunt about lifestyle counseling. He guesses that about 2,000 patients have to be counseled on diet and exercise for one to lose weight. He says he wouldn't prescribe a drug that ineffective. That figure is his estimate, not a measured number.
Who keeps the weight off
Cifu agrees that the drugs treat obesity rather than cure it, and that most people regain weight after stopping. In his experience, two kinds of patients are the exceptions:
Meticulous trackers. They log food, calories, exercise and fiber, and keep every habit the same after coming off the drug.
People who used the drug as a full reset. One married couple in their 60s started tirzepatide with BMIs of 37–40. After 18 months they stopped at around 26. By then they had new meal plans, exercised together every day and drank much less alcohol. Their weight has stayed stable for another 18 months.
These are anecdotes. They don't tell us how common such success is.
Changes in the exam room
Before GLP-1s, Cifu rarely brought up weight with "healthy overweight" patients. These are people with a BMI of 30–40 whose cholesterol, blood sugar (7), liver tests and blood pressure are normal. He had little to offer them. Now he raises treatment with anyone whose BMI is above 35. Reactions vary from relief to flat refusal ("no interest in looking like some gaunt influencer"). Fewer patients start the drugs than he expected.
Side effects
Cifu sets aside the usual worries, such as thyroid cancer, hair loss and muscle loss. He thinks the cancer effect seems to come out about even and the rest mostly reflect eating in a calorie deficit. He also says an important harm could still emerge. The new complaint in his clinic is benign: weight loss uncovers bony landmarks, small hernias and fatty lumps (8) that worry patients into booking visits.
His last point is that many patients doubted the drugs would work because "I don't overeat." He says he has never heard that line once someone has lost weight on the medication.
Remember
This is one doctor's clinical experience, not trial evidence. His success stories don't show how often weight loss lasts.
Cifu thinks keeping weight off after stopping takes either strict tracking or lasting lifestyle changes made while on the drug.
He now offers treatment to otherwise healthy patients with a BMI above 35, but fewer accept than he expected.
The diet researchers call "methionine restriction" has a naming problem, according to the authors of a new open-access paper, covered by Reason at the longevity blog Fight Aging!. The standard lab diet cuts methionine and also contains no 1, the paired form of cysteine found in food. So animals on it are short of both sulfur amino acids. The authors argue that cysteine is doing the work.
Why the two are hard to separate
When dietary cysteine runs out, the body makes more by breaking down methionine. The key enzyme in that pathway is 2. Because cutting one amino acid drains the other, earlier experiments could not tell which shortage triggered the benefits. Those benefits are substantial. In one well-known rat study, cutting dietary methionine by 80%, with no cystine in the diet, raised both median and maximum lifespan by about 40%. Other work in normal mice, where the two amino acids were adjusted separately, had already tied the fat-reducing effects of the diet to low cysteine.
What the new study did
Worms: In 3, a small roundworm widely used in aging research, restricting only cysteine extended lifespan and improved survival under stress, whether methionine was present or not.
Mice lacking CTH: These mice cannot turn methionine into cysteine, so the two variables can finally be separated. Cysteine-free diets produced favorable metabolic effects whether methionine was normal or restricted. Restricting methionine while keeping cysteine normal had no effect.
Old mice: Keeping cysteine about 70% lower over a long period in aged CTH-knockout mice "reprogrammed the 4", meaning the links between immune function and metabolism, and improved 5 (years lived in good health).
Earlier work from the same field suggests one mechanism. Severe cysteine depletion activates brain regions that drive the sympathetic nervous system, the body's "fight or flight" wiring. That releases norepinephrine, which acts on 6 in fat cells and turns ordinary white fat into heat-burning "brown" fat. Forcing cysteine down that hard caused a rapid 30% weight loss.
Caveats
All of this evidence comes from worms and mice, and the key mouse experiments used genetically engineered animals that cannot make cysteine. The excerpt gives no lifespan numbers for the new experiments. It reports "healthspan benefits" in old mice, not a measured extension of life. There are no human data. The 30% weight loss from forced depletion is also a reminder that cutting cysteine hard is not obviously safe. Cysteine is needed to make 7, a major antioxidant, and coenzyme A, a central molecule in metabolism.
Remember
Standard "methionine restriction" diets also contain no cysteine, so past results can't tell the two apart.
In mice unable to make cysteine, restricting methionine alone did nothing, while cutting cysteine did help.
All the evidence is from worms and mice, and there are no lifespan numbers for the new mouse work.
KFF Health News reviewed thousands of Slack messages, meeting transcripts and recordings. They show that since August 2025, 1 has run a private Slack workspace where tech companies talk directly with senior health officials. The workspace has grown to about 1,700 members. Only a handful are patient advocates, doctors or hospital representatives.
The workspace was opened by Amy Gleason, former acting head of 2 (the Department of Government Efficiency) and now a senior CMS product official. Members include Oura, Palantir and 8VC, a venture capital firm linked to Health Secretary Robert F. Kennedy Jr.'s son.
Policy made outside public view
In February, the FDA held a listening session on patient-facing chatbots. At least 35 industry groups attended, including Microsoft, OpenAI, Anthropic, Apple and Google. The invitation went out on Slack, and the meeting never appeared on the FDA's public calendar.
The workspace's code of conduct says the group is not an 3. Joseph Daval, a former FDA lawyer now at Harvard Medical School, said it works like one anyway: members discuss with each other and build consensus on what policy should be. A formal committee would have to meet independence and fair-balance rules. CMS declined to say whether the workspace is legal.
Promises to promote apps
On a February call, a CMS Innovation Center official said he hoped the agency would be a "sales engine" for members' apps. The recording was made private the day after reporters asked about it.
The 4 now promotes about two dozen apps for conditions such as obesity, diabetes and cancer. CMS has not published its full criteria for listing apps. Officials told companies that listed apps would get priority in a new program that pays them for AI chatbot advice or wearable tracking.
One listed app is Slothwise, launched this year. It charges $9.99 a month, has only a handful of reviews, and describes itself as "informational." One reviewer said they trusted it "because it was endorsed by medicare."
Access to records
Vendors have lobbied to let apps act "in the shoes" of the patient. In practice, one consent would give an app ongoing access to records on 5, the national network providers use to exchange records. Supporters say third-party vetting and an industry code of conduct are enough protection. One lobbyist pushing this sits on the board of a vetting body. Kristen Valdes, CEO of b.well, wrote that "we need to stop protecting patients from themselves."
Not everyone agreed. Fasten Health founder Jason Kulatunga warned that the system must be designed for malicious actors if it grows to thousands of apps. 6 chief John Whyte said many of these tools lack data on safety and accuracy, and no one is liable when they get things wrong.
Gleason described the record-access push as done with "no regulation, no rulemaking."
Remember
Most of the workspace's 1,700 members come from industry; patients and clinicians are barely represented.
Medicare listing could become a path to federal payment for AI apps with little evidence they work.
Vendors want one-time consent to give apps ongoing access to records on the national exchange network.
Classical:1 themselves raise the mutation rate. For example, a driver might break a 2, so the cell then piles up extra mutations as it divides.
Catastrophic: single extreme events, like a chromosome shattering, cause sudden large-scale changes to the genome.
Tail: people differ in their baseline mutation rate. Those at the high end of the distribution (the "tail") build up more mutations in every tissue, so they are more likely to hit a cancer-causing combination, and their tumours start out with more mutations.
They tested each idea against several independent lines of evidence. Each hypothesis explained the spread of mutation counts in tumours to some degree. But only the tail hypothesis easily explained one finding: normal tissue from cancer patients carried more mutations at the 3 than normal tissue from people without cancer. The classical and catastrophic explanations work only inside a tumour, so they can't explain why a patient's healthy cells are also more mutated.
A hypothetical example of the tail idea: suppose two people's stem cells divide the same number of times, but one person's cells make twice as many errors per division. That person ends up with roughly twice the mutations in every tissue. Cancer needs several specific hits in the same cell, so their chance of getting that combination rises by more than double.
Commentary and caveats
Reason, who writes the 4 blog, reads the result as support for the idea that random mutational damage spread across tissues by stem cells raises cancer odds. He places it alongside other age-related factors, such as a weakening immune system and the build-up of 5.
The evidence here is observational and the excerpt reports no effect size for the patient-versus-healthy difference, so treat it as suggestive. One alternative the summary doesn't address (this is the digest's note, not the authors'): "normal" tissue sampled near a tumour may have been exposed to the same carcinogens or local changes, sometimes called 6. That could raise its mutation count without any difference in a person's inherent mutation rate. The authors say the tail hypothesis is "tenable" and could shape prevention strategies, but they offer no test for spotting high-rate individuals.
Remember
Across five organs, cancer cells carried about four times as many mutations as normal cells.
The healthy tissue of cancer patients was also more mutated, which only the 'tail' hypothesis readily explains.
The idea that some people mutate faster is unproven; tissue sampled near a tumour may be a confounder.
On Sept. 10, the 1 published a notice proposing three changes to the 2030 census. Nancy Krieger, a social epidemiologist at the Harvard T.H. Chan School of Public Health, argues in STAT that the changes would make U.S. death and disease rates jump without any real change in health. This is an opinion piece. Her predictions are reasoned arguments, not modeled estimates.
What's proposed
Who counts. The census would count only citizens and 2 (green-card holders), not the "whole number of persons in each state." International students, work-visa holders, refugees, 3 and undocumented immigrants would be left out.
Where they count. People would no longer be counted at their usual residence on 4, April 1. They would be counted where they lived most between Jan. 3 and April 1, tied to their tax-record address. Krieger expects college students to be counted at their parents' homes. "Snowbirds," people who spend winter in warmer states, and seasonal workers would be counted in their winter states.
What's asked. Questions on race and ethnicity would be banned. So would the household-relationship item that, from 2020, captured same-sex couples.
Why rates would move
A rate is cases divided by population. Suppose a county records 1,000 deaths among 100,000 counted residents, a rate of 1%. If the new rules drop 10,000 residents from the count but the deaths stay the same, the rate becomes 1.11%. Krieger expects the biggest artificial jumps among college-age and retirement-age people, people of color and immigrants, and in some states more than others. Rates could also fall in places that gain counted residents, such as college towns' home counties or Sun Belt states. The article doesn't discuss that side.
She says the damage would spread further. Health surveys weighted to census totals would become skewed. So would funding formulas and the Census Bureau's own listed uses: planning hospitals and clinics, estimating disaster displacement, and assessing how infectious disease could spread.
The race data argument
Krieger's main example is 5. U.S. rates are higher than in any other high-income country. College-educated Black women die at higher rates than white women who never finished high school. Data on income or education alone would miss this, she argues. She also points out that racial and socioeconomic gaps in deaths before age 65 narrowed between 1966 and 1980, then widened again. She links that pattern to 6 programs and the backlash after 1980. That link is her interpretation, not something the data prove.
Process
Krieger says the changes haven't gone through the census's usual testing. She calls them unconstitutional and says their purpose is partisan advantage in House seats and districts. The original 33-day comment period was extended after public pushback. Comments are now due Nov. 2.
Remember
Shrinking the census denominator raises rates on paper even if the number of deaths and cases stays the same.
Survey weights and federal funding formulas both depend on census counts, so errors would spread well beyond the census itself.
Black women with a college degree have higher maternal mortality than white women who didn't finish high school.
