Feeling Fine in Pregnancy Was No Shield Against Dirty Air's Link to Postpartum Depression

I’d always assumed that if a smoggy city were going to wear down a pregnant woman’s mood, it would hit hardest on the one already buried in stress. A new Mexico City analysis only half agrees with me. Researchers tracked fine-particle pollution through pregnancy and then screened mothers for depression six months after delivery. The pollution link was strongest in two groups: women who had piled up bad life events, and women who said they felt little stress at all.

The work, published this week in Environmental Health, comes from researchers at Mount Sinai’s Icahn School of Medicine, Brown University and Mexico’s National Institute of Public Health and National Institute of Perinatology. They drew on PROGRESS, a cohort of pregnant women recruited through Mexico’s social security health system. Each woman’s home was matched to daily estimates of fine particulate matter (PM2.5), the mix of combustion particles, dust and other specks small enough to slip past the nose and settle deep in the lungs. The estimates came from a model built on satellite aerosol readings at one-kilometer resolution, ground monitors, weather and road density. Mothers filled out the Edinburgh Postnatal Depression Scale, which is a screening questionnaire and not a diagnosis, during pregnancy and again six months after birth. The NIH’s environmental health institute and Mexican public-health agencies funded the work, and the authors declare no competing interests.

By American standards the air was dirty. Average pregnancy exposure came to 22.9 micrograms per cubic meter, against a U.S. annual standard of 9.0 micrograms per cubic meter, so these women breathed more than double the American limit. The middle half of exposures ran from 20.3 to 24.5 micrograms, which means the 5-microgram step the researchers used for their comparisons is a little wider than that middle spread.

The detailed results come from the version of the paper the team posted in February, which covered 475 mothers. Each 5-microgram rise in average pregnancy PM2.5 went with 1.49 times the risk of screening positive for depression at six months. Splitting the cases made the signal sharper. Women who screened positive both in pregnancy and afterward showed no significant link to the air. Women who screened negative in pregnancy and positive at six months showed a relative risk of 2.67 per 5 micrograms (95% CI 1.37 to 5.21). That new-onset group held only 41 women, which explains the wide interval, and the paper doesn’t formally test the two groups against each other. This is an observational cohort, so it can show a link but can’t establish that the air caused the depression. The pattern still points at women who looked fine during pregnancy and then weren’t.

The stress data is where I had to slow down. The team measured stress two ways, and the two measures capture different things. One counted how many of eleven areas of a woman’s life (money, legal trouble, work, relationships, safety at home and in the neighborhood, health, prejudice and a few others) had produced at least one negative event in the previous six months, which makes it a tally of what happened to her. The other was a four-item Perceived Stress Scale, given in the third trimester, which asked how overwhelmed she felt. For new-onset depression, the pollution link was strong among women with many negative life events, at a relative risk of 4.71 (CI 1.72 to 12.92), and I’d have predicted that. It was just as strong among women who reported low perceived stress, at a relative risk of 4.58 (CI 1.83 to 11.49). Both estimates come from subsets of those 41 cases, so the intervals are huge, but the interaction test for perceived stress still reached significance (p = 0.035).

Why would saying you feel fine leave you more exposed?

Wait, how could your answer on a questionnaire change what a particle does once it’s lodged in your lungs? The authors offer several possibilities and call them possibilities. Women who report little stress may have built fewer coping strategies, so an invisible physical stressor lands harder. They may live in more hazardous neighborhoods with thinner support networks. And perceived stress is subjective, so a low score can sit on top of hardship a woman simply doesn’t call stress.

The biology they propose is physical. Fine particles settle in the deep lung and set off inflammation, and the authors suggest PM2.5 may lean harder on the hypothalamic-pituitary-adrenal axis, the brain-to-adrenal relay that pumps out cortisol, while pushing up inflammatory cytokines. This study didn’t measure any of that. Earlier work has looked at the pairing directly: a 2021 Heliyon paper examined particulate exposure alongside prenatal depressive symptoms and stress-hormone dysregulation.

Here’s how I read it, and it’s an interpretation, not something the study found. The life-events count tracks what is happening to a woman, while the perceived-stress score tracks how she rates it. In these data, a heavy load of events went with a stronger pollution link, and a woman’s sense of being fine did nothing to weaken it. The body may be keeping a tally that the questionnaire misses.

Other studies point the same way, though they aren’t all independent. An earlier analysis of the same PROGRESS cohort covered 509 mothers, and the 475 in the new paper are a subset of them. It found that each 5-microgram rise in pregnancy PM2.5 went with 1.59 times the risk of depression at six months (CI 1.11 to 2.28), and 2.58 times the risk for depression that began after the first month. Pollution in the year after birth didn’t predict depression, which pointed to pregnancy as the sensitive window.

The independent evidence comes from Los Angeles. The MADRES cohort followed 361 mostly low-income Hispanic mothers for three years after birth. Exposure to coarser PM10 in gestational weeks 12 to 28, and to nitrogen dioxide in weeks 13 to 29, tracked with nearly four times the cumulative risk of depression over those years. That study found no clear sensitive window for PM2.5 itself. A Kaiser Permanente Southern California analysis of 340,679 births found smaller effects. The odds of postpartum depression were 2 percent higher per interquartile step in PM2.5 across pregnancy and the postpartum months, and 9 percent higher for ozone, with black carbon showing the strongest link among the particle components. The effect on any one mother is small, but nobody living in that basin can opt out of breathing its air. Because those numbers come from American exposure levels, they’re the better guide for readers here. The Mexico City women breathed air far above both the old and new U.S. limits, and their data can’t tell you what happens between the two.

The EPA asked a court to throw out its own rule

About 1 in 8 women with a recent live birth report symptoms of postpartum depression, according to the CDC. The drug market’s answer comes after the fact. Zuranolone (Zurzuvae), from Sage Therapeutics and Biogen, was the first pill approved for postpartum depression, and it launched in 2023 at a wholesale price of $15,900 for a 14-day course. It treats depression once it has arrived. The exposure in this study comes earlier, in the air a pregnant woman breathes, and the federal agency that sets the limit on that air has spent the past year and a half trying to undo a tighter one.

In February 2024 the EPA lowered the annual PM2.5 standard from 12.0 to 9.0 micrograms per cubic meter. In March 2025, Administrator Lee Zeldin’s EPA announced it would reconsider the standard. In November 2025 the agency asked the D.C. Circuit to vacate its own rule, arguing that it had not done a thorough enough review of the science and had wrongly refused to weigh costs. On June 26, 2026, the court upheld the standard anyway and rejected challenges from industry groups, several states and the EPA itself. An administration that campaigned on making America healthy again sent its environmental agency to federal court to get rid of a tighter limit on particles pregnant women breathe, and cost was one of its arguments. That put the agency on the polluters’ side, and the health-first wing of the movement should say so out loud.

The authors’ practical suggestion is that careful stress profiling during pregnancy could help flag which women are at risk. Today’s screening mostly asks a woman how she feels. This work makes the case for also asking what has happened to her and what she breathes, because in these data “I’m fine” didn’t protect anyone.

If I were pregnant in a smoggy city, I’d run a HEPA purifier in the bedroom and take my walks away from the truck routes. No trial has shown that either one prevents postpartum depression. I’d do both anyway, because a filter is cheap compared with waiting for that proof, and I wouldn’t let a low score on a stress questionnaire talk me out of it.

Sources

1. Joseph et al., Environmental Health (2026)
2. Joseph et al., Research Square preprint via PMC (2026)
3. Niedzwiecki et al., Environment International (2020)
4. MADRES cohort, Science of the Total Environment (2024)
5. Sun et al., JAMA Network Open (2023)
6. Heliyon (2021), particulate matter, prenatal depressive symptoms and HPA axis dysregulation
7. Harvard Environmental and Energy Law Program, PM2.5 NAAQS tracker
8. Drug Discovery Trends, Zurzuvae treatment cost (2023)
9. CDC, Depression during and after pregnancy