The Wellness Brew - Issue 33

In 2001 the federal government warned pregnant women about mercury in fish. The warning worked, and they ate less fish. Two decades of research later the agencies reversed course and now recommend two to three servings a week, because the shortfall was doing more harm than the mercury. Almost nobody got the update. That’s one of several places where what gets repeated and what the evidence says have come apart.

But first, we filtered the noise — here's what's worth knowing this week.

A 2026 study in PeerJ put 21 men and 21 women through bench press testing at 75% of their one-rep max. On a single set taken to failure there was no meaningful difference at all: women averaged 9.0 reps, men 8.6. The gap showed up in the multi-set protocol, sets of five with 90 seconds of rest until failure, where women completed 58 total reps against 29 for the men. An exploratory analysis pointed to how quickly the women recovered between sets rather than how long they lasted within one. Women tend to have more fatigue-resistant type 1 muscle fibers and can recover faster. This suggests women could benefit from shorter rest periods compared to men. Forty-two people and one exercise, so treat it as a lead rather than a finding. — Nuckols et al., PeerJ, 2026.

Researchers at East China Normal University inserted a glucose-sensing switch into E. coli so the bacteria produce GLP-1 in response to rising glucose rather than on a fixed schedule. Published in Nature in August. It worked as designed in mice and in diabetic macaques, with no sign of harm. The appeal is easy to see: a dose that answers to what you actually ate instead of a set amount injected once a week. Two things to hold onto. It has not been tested in a single human being, and the researchers say those trials are some way off. Genuinely interesting, but years away from mattering to anyone. — Nature, 2026.

The list of foods a pregnant woman is told to avoid has barely changed in thirty years. Soft cheese, deli meat, sushi, limit the fish, limit the coffee. It gets discussed at the first appointment, printed on a sheet, and repeated by every app and relative for the next nine months.

Some of it is well supported. Some of it was overtaken by better evidence years ago and nobody updated the guidance. And then there’s half of the picture that matters more — what to actually eat, and when — which gets a fraction of the attention, even though that's where most of the research has been pointing for two decades.

Here's what holds up on both sides.

Alcohol

One of the most common questions is about the occasional glass of wine. Is it safe or not? Of course heavy drinking is not safe and the evidence is clear on that, but what about one glass of wine? 

A 2017 systematic review pooled 26 studies looking specifically at light drinking, which it defined as 32 grams of alcohol a week or less. That works out to roughly two standard drinks across a whole week. That’s more than the single occasional glass most people are asking about, which makes it a useful upper bound.

At that level, babies were slightly more likely to be born smaller than expected for their gestational age: an 8% increase in the odds, consistent across seven studies. For preterm birth, the estimate pointed the same direction but didn't reach statistical significance, and the authors said plainly it was compatible with no effect at all. For nearly every other outcome, there wasn't enough data to pool. [1]

The reviewers described the evidence base as sparse. Their own conclusion was that guidance could reasonably recommend abstaining as a precaution, and that it should also be honest about how thin the underlying evidence is.

US guidance takes the first half of that. The CDC's position is that there's no known safe amount and no safe point in pregnancy, including while trying to conceive. [2]

Worth being precise about what that means. "No known safe amount" is not the same statement as "any amount has been shown to cause harm." The first is true. The second isn't, at the low end. But the gap between them is an argument for abstaining rather than against it: the outcome is permanent, there's no way to identify in advance which pregnancies are susceptible, and nine months is not a long time to go without.

Glyphosate, at the point in pregnancy that seems to matter

This one is new enough that most people haven’t seen it.

In July, researchers at NYU published results from 1,450 pregnant women in New York City who gave urine samples twice, once between 4 and 18 weeks and again between 18 and 25 weeks. Women in the top quarter of glyphosate concentrations in the later sample had roughly twice the odds of spontaneous preterm birth as women in the bottom quarter. The earlier samples showed no association. Neither did medically induced early deliveries, only the spontaneous ones. [3]

A large Canadian study of 1,765 women asked the same question in 2024 and found nothing at all, but it measured first-trimester urine only, and its authors flagged that as a limitation. [4] The two results may not conflict so much as describe different moments.

The timing pattern is the interesting part, and it’s also the part the study couldn’t pin down. The formal test of whether the two windows genuinely differ fell short of statistical significance, so the pattern is suggestive rather than settled. There were 37 spontaneous preterm births across the whole cohort, which is a small number to rest a finding on. And this is observational work, which can’t show that glyphosate caused anything.

One point in its favour: in this same group of women, eating more nuts and whole grains tracked with higher urinary glyphosate. Those are foods tied to better birth outcomes, so that particular confounding should be pulling the association toward zero rather than inflating it. Worth knowing alongside it: the study’s senior author sits on the board of Beautycounter, a personal care company that markets chemical avoidance.

Still, exposure is close to universal and reducing it is cheap enough that waiting for certainty isn’t necessary.

Three practical levers. Buy organic where it’s affordable, since organic production prohibits glyphosate. Look for products carrying glyphosate residue free certification, which tests the finished product rather than inferring from how it was grown. And know where it concentrates: glyphosate-tolerant crops get sprayed during the growing season, which covers corn, soy, canola and sugar beet, while wheat, barley, oats and pulses like lentils and chickpeas may get an application shortly before harvest to dry the crop down for easier combining. [5] Those two categories account for most of where residues turn up.

Listeria foods

This isn't one to argue with. Pregnancy makes listeriosis roughly ten times more likely, and about one in six cases occur in pregnant women. For Hispanic women the figure is 24 times, which the FDA attributes specifically to queso fresco-style cheeses. [6]

The reason to take it seriously isn't the odds, which are low. It's that the mother may feel like she has mild flu while the pregnancy ends in miscarriage or stillbirth. Unpasteurized milk and cheese, queso fresco even when pasteurized, deli and luncheon meat unless heated until steaming, refrigerated pâté, and refrigerated smoked seafood unless it's cooked into something.

Low probability of occurring, but severe outcome if it does and almost no cost to avoiding it. For those reasons, this rule makes sense.

High-mercury fish

The actual high-mercury list is seven items: king mackerel, marlin, orange roughy, shark, swordfish, Gulf of Mexico tilefish, and bigeye tuna. [7]

That's it. Most other fish are fair game, which you’ll see in the next section.

Too much preformed vitamin A

Vitamin A comes in two forms and only one of them is a problem. Preformed vitamin A, the kind in liver, eggs, dairy and most supplements, gets absorbed almost completely and goes straight into circulation. Beta-carotene, the kind in carrots and sweet potatoes and leafy greens, has to be converted first, and the body slows that conversion down once it has enough. You can’t overdose on carrots. You can overdose on the other kind.

The ceiling is 3,000 micrograms a day of preformed vitamin A, and above it the vitamin can cause birth defects, affecting the eyes, skull, heart and lungs. The risk concentrates in the first trimester, when organs are forming. [8]

A single three-ounce serving of beef liver contains more than double that ceiling. And on a supplement label, only the retinol portion counts toward the limit, so a product listing 3,000 micrograms that’s mostly beta-carotene is nowhere near it. The form matters just as much as the number.

Caffeine over 200 mg

The American College of Obstetricians and Gynecologists reaffirmed its position in 2023 that moderate intake under 200 milligrams a day isn't a meaningful contributor to miscarriage or preterm birth. Above that, they say the evidence conflicts and decline to make a recommendation either way. [9]

So 200 milligrams, roughly a 12-ounce coffee, is a ceiling rather than a target.

Folate, and the window that closes first

This is the single best-evidenced thing in pregnancy nutrition, and most people get it slightly too late.

In the first month after conception, a flat ribbon of cells running down the embryo's back curls upward and seals itself into a tube. That tube becomes the brain and the spinal cord. If it fails to close completely, the result is a neural tube defect: spina bifida when the gap is toward the bottom, anencephaly when it's at the top.

Folate is what the body uses to build and close it. And the tube finishes sealing about four weeks after conception, which is roughly two weeks after a missed period. [10]

Which means that by the time a pregnancy test reads positive, the window is closing or already shut. The recommendation is 400 to 800 micrograms a day for any woman who could become pregnant, not just those trying. [8] 

The landmark trial that established this cut neural tube defects by 71%, though it used a much higher dose in women who'd already had an affected pregnancy. The everyday dose is smaller and so is the effect, but the direction and the timing are not in question. [10]

Fish, where the advice reversed and nobody noticed

In 2001 the FDA and EPA warned pregnant women about mercury in fish. The warning worked, in the wrong direction: pregnant women cut their fish intake by more than a serving a month.

Two decades of cohort studies since have found no harm from eating more fish and, in several cases, better neurocognitive outcomes in the children. [11] The agencies revised the advice in October 2021, and the current FDA position states plainly that fish intake during pregnancy is recommended because it supports a baby's cognitive development.

The current guidance is two to three servings a week, four ounces each, from their best choices list. Salmon, sardines, anchovies, cod, shrimp, canned light tuna and about thirty other options all qualify. [7]

Choline

450 milligrams a day, and only about 7.7% of women aged 20 to 40 reach that from food. [8] Eggs and beef are the richest sources by a wide margin, which is also why much of the research in this area is funded by the egg and beef industries, including the one trial that compared doses, which had 26 participants. [12] One egg has roughly 150mg of choline, so if you aren’t eating 3 eggs a day or beef, consider supplementing. 

Iodine

220 micrograms during pregnancy, up from 150 otherwise. Around 63% of pregnant women fall below the urinary threshold for sufficiency. The NIH notes that while severe deficiency clearly causes serious neurodevelopmental harm, the evidence on supplementing mild to moderate deficiency is inconsistent. [8] 

Iron and DHA, which both peak late

Iron demand climbs through pregnancy. Deficiency runs around 25% in the third trimester against 10 to 16% across pregnancy overall, and 83.8% of pregnant women fall below the estimated average requirement from food alone. [8]

DHA follows the same curve more sharply. Between 42 and 67 milligrams a day crosses to the fetus during the third trimester, concentrated in the final five weeks, and it's building the brain and the retina. [13] It's the same nutrient that makes fish beneficial. If supplementing, make sure you look for Omega-3s that specifically label DHA and are IFOS (International Fish Oil Standards) certified. 

Dates, in the last month

This one sounds like folklore and has more behind it than expected. A 2024 meta-analysis pooled 48 trials of women eating roughly six to seven dates a day during the final month of pregnancy. Compared with those who didn't, they arrived at the hospital with about a centimeter more cervical dilation, were roughly half as likely to need induction, and were more likely to go into labor on their own. The latent phase of labor was shorter by around three and a half hours. Caesarean rates were not significantly different. [14]

The caveats are real. Only 15 of the 48 trials were randomised and just four were rated low risk of bias, most came from Indonesia and Iran, and the overall certainty of the evidence was rated low to moderate. This isn't settled.

But it's six dates a day, it's food, and the only real downside is the sugar load, which matters for anyone managing gestational diabetes.

And about that prenatal vitamin

Everything above assumes the bottle does what the label says. Frequently it doesn't.

Consumer Reports tested 17 prenatal vitamins in August. Fifteen of them contained nutrient amounts that didn't match their own labels, and 31% of the 124 individual nutrient measurements came back outside acceptable thresholds. [15]

Folate was the most common failure, which is the worst possible nutrient for it to be. One A Day Prenatal Multi Gummy contained 28% of its labeled folic acid. Amazon Basics Prenatal & DHA came in at 77%, CVS Health Prenatal Gummies at 82%, and Centrum's prenatal gummy also fell short. Needed, a premium brand that markets on rigor, was below both the United States Pharmacopeia (USP) threshold and its own label. Four of the eight gummies tested were short on folate.

Two products matched their labels across every nutrient tested: Carlyle For Her and Bird&Be. Nature Made's prenatal gummy was USP Verified and still had a nutrient outside range.

Errors ran both directions. Thorne Basic Prenatal averaged 54.5 milligrams of iron against 45 on the label, above the tolerable upper limit. Seeking Health's vitamin D also exceeded its upper limit. Perelel disputes the results for its product; Target and Centrum didn't respond to Consumer Reports.

Separately, a 2025 analysis in the American Journal of Clinical Nutrition examined 47 prenatals for the two nutrients most often missing. Only 26% listed choline at all, and of those, 42% contained the amount claimed. 53% listed iodine, and only 16% of those were accurate. [16]

Both studies also detected lead and cadmium in most products tested. Neither found any product exceeding USP or Consumer Reports safety limits. That's worth stating plainly, because it's the detail most coverage leaves out.

The avoid list is shorter and more specific than what you hear from family. Seven fish, not all fish. A ceiling on caffeine, not a ban. Listeria rules that are worth following exactly because they cost nothing.

The other half is where the work is. Folate before you need it. More fish than the old advice suggested. Choline and iodine that most women don't get and most prenatals don't supply. Iron and DHA that matter most in the stretch when people have stopped paying close attention.

And a prenatal vitamin that is worth checking rather than trusting.

Four things you can do this week:

1. Start folate before you start trying.
If there’s any chance of a pregnancy in the next year, begin taking 400 to 800 micrograms of folic acid a day now. Not when you start trying, and not when a test comes back positive. The neural tube finishes closing about four weeks after conception, which is roughly two weeks after a missed period, so for many women the window has already shut by the time they know. This is the one item on the entire list where being early is the whole point, and a standard prenatal or even a basic multivitamin covers it for a few dollars a month.

2. Take the bottle off the shelf and check three things.
Is it a gummy? Four of the eight gummies Consumer Reports tested came up short on folate, one of them at 28% of what its label claimed, and gummies were the worst performers in the group. Does it list choline, and how much? Most prenatals contain little or none against a target of 450 milligrams a day. Does it list iodine? About half don’t. If it fails on choline, three eggs a day closes the gap, or a separate choline supplement will. Only two products matched their labels on every nutrient tested: Carlyle For Her and Bird&Be.

3. Eat more fish, not less.
Two to three servings a week, four ounces each. That’s current federal guidance, revised in 2021, and it’s close to the opposite of what most people took away from the 2001 mercury warning. Salmon, sardines, anchovies, cod, shrimp and canned light tuna are all on the recommended list along with about thirty others. Seven fish are genuinely off the table: king mackerel, marlin, orange roughy, shark, swordfish, Gulf of Mexico tilefish and bigeye tuna. A middle group including albacore tuna and halibut is capped at one a week. If you’ve been skipping fish to be careful, that’s the thing to change.

4. Spend your organic budget where glyphosate actually concentrates.
The new research found the association in mid-pregnancy samples, around weeks 18 to 25, which is also the stretch when most people have stopped thinking carefully about it. You don’t need to buy everything organic. Prioritize the categories where residue turns up: corn, soy and canola products, and the grains and pulses that may be sprayed before harvest, meaning wheat, oats, barley, lentils and chickpeas. On the products you buy every week, glyphosate residue free certification tests the finished product rather than inferring from how it was grown.

If you found this useful, please forward it to your friends and family. If you had this forwarded to you — sign up here.

Wellness, filtered.

The Wellness Brew

Sources:

Disclaimer: The Wellness Brew is for informational purposes only and does not constitute medical advice. The content published here is not intended to diagnose, treat, cure, or prevent any disease or health condition. Always consult a qualified healthcare professional before making any changes to your diet, supplement routine, or lifestyle.