Chemicals and Fertility: Environmental Exposure and Health

The peer-reviewed research on environmental chemicals and reproductive outcomes — sperm count, sperm quality, ovarian reserve, polycystic ovary syndrome, endometriosis, miscarriage, IVFIn Vitro Fertilization — assisted reproductive technology where eggs are fertilized outside the body and embryos transferred to the uterus. success rates — hedges every claim carefully. Associated with, not causes. Correlates with, not proves. If that feels frustrating, you're reading it correctly: the field is frustrating. What follows applies that research to specific chemicals and specific outcomes, with every claim hedged the way the underlying science hedges it. The alternative is overreach, and the audience for this article is people who will verify citations, so overreach would destroy the only thing that makes the article worth reading.

This is not medical advice — for anything resembling clinical recommendations, the next step is a reproductive endocrinologist, not a blog post. Our guide to endocrine disruptors covers the broader biology of how these chemicals interact with hormones. This article applies that science specifically to fertility: what specific chemicals the studies found associated with reproductive outcomes, what they did not find, and which modifiable exposures couples trying to conceive can reasonably reduce while they work through the clinical questions with someone qualified to answer them.

The EU-wide annual health cost of endocrine-disrupting chemicalA synthetic chemical that mimics, blocks or alters the body's own hormones — often at doses far below the threshold for classical toxicity. Shortened to EDC. exposure sits at a median of €157 billion annual population-level health burden from endocrine-disrupting chemicals across the EU per year, with reproductive and neurodevelopmental effects dominating the calculation Trasande et al. 2015. That figure is not the same as 'chemicals caused your infertility.' It's the Monte Carlo simulation of the population-level burden attributable to EDC exposure across dozens of outcomes and hundreds of assumptions — IQ loss, autism, ADHD, obesity, adult diabetes, cryptorchidism, male infertility, and testosterone-linked mortality. But it's also the closest thing the field has to a headline number, and it's orders of magnitude larger than the costs that get cited in regulatory impact assessments.

Fertility is multi-causal, and environmental chemical exposure is one factor among several. The well-established modifiable factors are smoking, heavy alcohol use, elevated BMI, heat exposure (for men), certain medications and age. These each have stronger evidence for a direct causal effect on fertility outcomes than any specific environmental chemical does. If you are a man who smokes and uses a hot tub daily, worry about the cigarettes and the hot tub first. The chemical layer is additional, not a substitute for lifestyle basics.

That said — and this is where the article gets interesting — the chemical evidence is now substantial enough that it's reasonable to treat environmental exposure as one of the modifiable factors, not a fringe concern. The question isn't whether chemicals affect fertility in absolute terms. The question is whether they affect it enough, alongside the lifestyle factors, to matter for a specific couple trying to conceive. The evidence for that is mixed but growing, and the practical interventions are low cost. The rest of this article is a tour of what we actually know.

Global mean sperm concentration has fallen by 51.6% global decline in mean sperm concentration from 1973 to 2018 since 1973, according to the largest meta-analysis of its kind — 223 studies studies pooled in the largest meta-analysis of global sperm count trends spanning 1973 to 2018 Levine et al. 2023. The rate of decline is accelerating. Before 2000, sperm counts fell by roughly 1.16%/year sperm count decline rate pre-2000. After 2000, they fell by 2.64%/year sperm count decline rate post-2000 — more than double. The analysis extended the earlier 2017 finding (Western men only) to the global population for the first time.

The data is not without caveats. Meta-analyses of sperm counts are prone to selection bias in how original studies recruited participants. Early studies and recent studies used somewhat different measurement protocols. The rate of decline appears bimodal — fertile men (proven fathers) show almost no decline, while unselected population samples show the full effect. That last finding is strange and important: it suggests that whatever is driving the decline isn't affecting the entire male population uniformly, but is disproportionately affecting the lower end of the distribution — where subfertility lives.

Testosterone has been declining alongside sperm count. In NHANESNational Health and Nutrition Examination Survey — the CDC's ongoing population health study, which tests blood and urine samples from thousands of Americans for hundreds of chemicals and biomarkers. data on 4,045 American men aged 15-39, mean total testosterone dropped from 605 ng/dL mean total testosterone in US men aged 15-39, 1999-2000 in 1999-2000 to 451 ng/dL mean total testosterone in US men aged 15-39, 2015-2016 in 2015-2016 — a decline of roughly 25% in under two decades, significant even among men with normal BMI Lokeshwar et al. 2021. The Massachusetts Male Aging Study had already reported an age-independent population-level decline in mean serum testosterone — a 'substantial and as yet unrecognised' drop greater than the cross-sectional decline typically associated with ageing itself Travison et al. 2007. Neither study identifies a specific cause. Both explicitly flag environmental exposure as a candidate.

The cleanest finding in the field is about what's on the produce, not the produce itself. Among 155 men at a Harvard-area fertility clinic, those in the highest quartile of high-pesticide-residue produce intake had 49% lower total sperm count in men eating the most pesticide-residue-heavy produce lower total sperm count and 32% lower normal sperm morphology in men eating the most pesticide-residue-heavy produce lower normal morphology than men in the lowest quartile Chiu et al. 2015. Total produce intake had no effect — the signal was specific to the pesticide burden of the produce. Eat more fruit and veg is correct. Eat more fruit and veg that was sprayed with pesticides, less so.

BPA hits multiple sperm parameters at once. Among 190 men at a Boston fertility clinic, an interquartile-range increase in urinary BPA was associated with 23% lower sperm concentration, 13% lower normal morphology and 10% higher sperm DNA damage Meeker et al. 2010. A 2024 meta-analysis pooling 18 epidemiological studies confirmed the consistent negative association between urinary BPA and sperm concentration, though the authors noted that the association with clinically abnormal sperm — values below WHO reference thresholds — was not robust Castellini et al. 2024. That is an honest caveat the primary literature makes clearly, and not a finding the lay coverage usually reports.

The phthalate evidence is now the most extensive of any single chemical class for male fertility. Tian et al. 2024 pooled 50 epidemiological studies covering both general populations and fertility-clinic cohorts Tian et al. 2024. The headline findings: metabolites of DEHP — the most common plasticiser phthalate — were inversely associated with sperm concentration, progressive motility and normal morphology. The associations held across adult-exposure cohorts and were strongest for DEHP metabolites MEHP, MEHHP and MEOHP. A separate 2024 meta-analysis of 38 studies on phthalates and male reproductive hormones found consistent reductions in total and free testosterone at higher exposure levels, with effect sizes modest but directionally consistent Cai et al. 2024. What makes phthalates different from BPA is production volume: global DEHP output is still around 8 million tonnes/year global DEHP production, approximate recent year and the half-life in the body is hours, which means exposure is effectively continuous rather than accumulated.

In young Swedish men from the general population — not a fertility clinic, which matters for generalisability — DEHP phthalate metabolites were associated with lower progressive sperm motility in the higher exposure groups, and with higher high-DNA-stainability, a marker of sperm chromatin immaturity Axelsson et al. 2015. The effect magnitudes were modest but directionally consistent with the mechanistic literature on phthalates and androgen signalling. And the phthalate story doesn't start with adult sperm counts.

Prenatal phthalate exposure shows up at birth. In 85 mother-infant pairs from the Study for Future Families, boys whose mothers had higher urinary levels of four phthalate metabolites (monoethyl, mono-n-butyl, monobenzyl and monoisobutyl) had significantly shorter anogenital distanceThe measured distance between the anus and the base of the genitals — used as a marker of androgen signalling during fetal development. Shorter distance in male infants indicates insufficient masculinisation in the womb., a standard reproductive developmental marker that parallels the feminisation seen in rodent studies of phthalate exposure. The highest-versus-lowest quartile of mono-n-butyl phthalate produced an odds ratio of 10.2 for shorter-than-expected anogenital index in boys with highest prenatal phthalate exposure for a shorter-than-expected anogenital index Swan et al. 2005. The study was small, the effect size large, and the rodent parallel was already well-characterised. It's been replicated enough times since (Swan 2015 TIDES cohort, Bustamante-Montes 2013, Suzuki 2012) that 'phthalates affect male reproductive development' is now among the better-evidenced claims in the field.

PFASPer- and polyfluoroalkyl substances — persistent 'forever chemicals' with geometric-mean human serum half-lives of 3.5 years for PFOA and 4.8 years for PFOS (Olsen 2007). evidence for male fertility markers is real but less consistent than BPA or phthalates. In a study of young Danish military recruits, men with high combined PFOS and PFOA serum levels had lower counts of morphologically normal sperm per ejaculate than men in the low-exposure group Joensen et al. 2009. A larger follow-up study with 247 men found an inverse association between serum PFOS and total testosterone but did not replicate the sperm-count finding Joensen et al. 2013. The authors attributed the non-replication to fewer highly-exposed individuals in the later sample. This is the kind of mixed evidence that makes strong claims difficult: the direction is consistent, the magnitude is not.

The strongest dose-response evidence comes from occupational exposure. Chinese factory workers with high BPA air and handling exposure at an epoxy resin plant had a clear dose-response relationship across three sexual-function outcomes, compared to unexposed controls Li et al. 2010. The caveats are important: self-reported outcomes, high occupational doses well above consumer-level exposure, and a Chinese factory population that may not generalise. But it's the strongest dose-response evidence for a direct human BPA effect on male reproductive function, and it's hard to explain away.

Male-factor infertility is implicated in roughly half of all infertility cases, either alone or in combination with a female factor. Most consumer-facing coverage of chemicals and fertility defaults to female-pelvic framing — PCOS, endometriosis, IVF — which isn't wrong, but it underplays the other half of the equation. The male side is where the exposure-to-outcome chain is mechanistically clearest, because the process being disrupted — spermatogenesis — is continuous and happens in an external organ that can be sampled directly.

Sperm production runs on a roughly 3-month cycle. A semen analysis today reflects the exposure environment of about 90 days ago, not this week. That matters practically: if a couple trying to conceive changes household products and diet tomorrow, the first semen analysis that fully reflects those changes will be three months out. It also matters mechanistically — the same 3-month window makes male fertility uniquely sensitive to any chemical with a short half-life and daily re-exposure, which describes phthalates, BPA and parabens precisely. Short half-life is a feature, not a reassurance — it means today's exposure is today's dose, every day.

The evidence stack on the male side now looks like this. The Levine 2023 meta-analysis of 223 studies establishes the population-level trend. Tian 2024 and Cai 2024 establish the phthalate associations with semen parameters and hormones across 50+ and 38 studies respectively. Castellini 2024 does the equivalent for BPA across 18 studies. Chiu 2015 provides the pesticide-residue signal. Li 2010 provides the occupational dose-response. None of these is a randomised trial. Together, they form a directionally consistent body of evidence that is, at this point, more extensive than the corresponding evidence for most clinical fertility interventions. The male fertility workup is often an afterthought in fertility clinics. The chemical-exposure side of it, more so.

Women who ate the most pesticide-residue-heavy produce had 18% lower clinical pregnancy probability for women eating the most pesticide-residue-heavy produce during IVF lower probability of clinical pregnancy and 26% lower live birth probability for women eating the most pesticide-residue-heavy produce during IVF lower probability of live birth than women who ate the least, across 325 women tracked through 541 IVF cycles in the same EARTH cohort at the Harvard fertility centre Chiu et al. 2018. Again, total produce intake wasn't associated with outcomes — only the residue-laden variety. This is the cleanest finding in the female fertility literature. It's not proof of causation, but it's the closest thing to a real-world intervention signal that the field has produced.

The BPA and polycystic ovary syndrome signal is consistent across multiple studies but worth stating carefully. Women with PCOSPolycystic ovary syndrome — a common hormonal disorder affecting roughly 1 in 10 women of reproductive age. Characterised by elevated androgens, irregular periods and often ovulatory dysfunction. carried measurably higher BPA levels than controls in the first clinical study to test the association, with BPA correlating with testosterone, androstenedione and insulin resistance markers Kandaraki et al. 2011.

The same directional pattern appeared in adolescent girls with PCOS — BPA higher in cases than controls, correlating with testosterone and hirsutism scores Akin et al. 2015. A meta-analysis pooling nine studies reported a standardised mean difference consistent with higher BPA in women with PCOS than in controls, though with high between-study heterogeneity Hu et al. 2018. None of this proves BPA causes PCOS. It does mean that women with PCOS tend to carry more BPA than women without, which is a non-trivial finding on its own.

PFAS and ovarian reserve is another area of consistent-but-not-conclusive evidence. PFOA and PFOS exposure has been associated with delayed menarche, longer menstrual cycle length, earlier menopause and reduced estrogen levels in a systematic synthesis of the literature Ding et al. 2020. In the C8 Health Project — one of the largest PFAS-exposed populations ever studied, covering communities near a DuPont plant in West Virginia — higher PFOA and PFOS exposure was associated with earlier menopause onset Knox et al. 2011. The authors themselves flagged a reverse-causality concern in their discussion: menstruation is an excretion route for PFAS, so post-menopausal women naturally accumulate more. The signal is real but the direction of causation needs hedging.

Phthalates and endometriosis have been studied several times with mixed results. In the ENDO Study — a large multicentre investigation of endometriosis natural history — multiple phthalate metabolites were associated with higher odds of endometriosis in the population-based cohort, though the signal was weaker in the matched operative cohort Buck Louis et al. 2013. Subsequent systematic reviews have pooled the evidence and found most but not all studies show an association. The overall direction points toward a phthalate link, but the inconsistency between studies means any specific risk estimate comes with wide confidence intervals.

Two studies have looked at BPA and IVF outcomes directly. Among 137 women undergoing 180 IVF cycles, a dose-response relationship emerged between urinary BPA and implantation failure, though the trend was marginal rather than strong Ehrlich et al. 2012. At a Stanford fertility centre, early-pregnancy serum BPA was measured, and women in the highest BPA quartile had 83% greater miscarriage risk for women with the highest early-pregnancy BPA levels greater miscarriage risk than the lowest quartile Lathi et al. 2014. Separately, higher PBDE flame retardant levels were associated with longer time-to-pregnancy in the CHAMACOS cohort of Latina women in California's Salinas Valley Harley et al. 2010. None of these are enormous sample sizes. The pattern across them is what the field takes seriously — not any individual finding.

Here is the honest answer: no RCTRandomised Controlled Trial — the gold standard study design where participants are randomly assigned to treatment or control groups. has directly tested whether reducing environmental chemical exposure improves sperm quality, ovulation or pregnancy rates in humans. That trial hasn't been done, partly because it's hard to run and partly because the commercial incentive to fund it is small. What does exist is strong intervention evidence that exposure itself is highly modifiable — and moderate observational evidence that modifiable exposures correlate with reproductive outcomes. The inference chain is reasonable but not closed.

Three days on a fresh-food diet — nothing from cans or plastic packaging — was enough to drop urinary BPA by 66% drop in urinary BPA after switching to fresh food for three days and DEHPDi(2-ethylhexyl) phthalate — the most common plasticizer phthalate, found in food packaging, vinyl flooring, and medical devices. metabolites by 53-56% in five families who switched from their normal diet Rudel et al. 2011. When the families returned to their normal diet, levels rebounded within days. This isn't a study of fertility outcomes. It's a demonstration that the body burden of the most-studied fertility-relevant chemicals responds quickly and measurably to dietary changes. Chemical exposure is not a permanent property of living in the modern world. It's a daily input that responds to daily choices.

The same principle holds for personal care products. When Latina adolescent girls in the HERMOSA intervention study switched to low-chemical cosmetics for three days, urinary methyl paraben, propyl paraben, triclosan and a monoethyl phthalate metabolite all dropped measurably Harley et al. 2016. Again, three days was enough. Neither study measured fertility outcomes directly — but both confirmed that the modifiable exposures identified in the observational fertility literature are, in fact, modifiable on timescales faster than a single menstrual cycle.

The logical chain: chemicals X, Y and Z correlate with fertility outcomes in observational studies. Exposure to X, Y and Z drops measurably when you change your diet and personal care products. Therefore changing those products is a reasonable intervention for couples who want to reduce the modifiable part of their fertility risk. The last step is an inference, not a proof. But for couples TTC, the cost of acting on the inference is low — the alternatives are often cheaper than the things they replace — and the cost of waiting for a definitive randomised trial that may never be funded is potentially years of lost reproductive window.

The advice 'reduce your chemical exposure' is useless without a process. Here is the one we'd run in our own home if we were trying to conceive, ordered by the ratio of exposure reduction to effort. It's deliberately boring — the point is that none of this requires a lifestyle overhaul. Most of it is a weekend afternoon.

The reason the bathroom comes first is absorption time. A moisturiser sits on skin for 12+ hours. A shampoo sits on the scalp for a couple of minutes but at high concentrations. A receipt sits on fingertips for seconds at high concentration — which is why the 'dry hands' detail matters: hand sanitiser and sunscreen on the fingers dramatically increase BPA absorption from thermal paper Hormann et al. 2014. The kitchen ranks second because food is daily dose multiplied by daily re-exposure. Water ranks last because the per-litre chemical load is usually low relative to food packaging and personal care — unless your local area has known PFAS contamination, in which case it moves up the list.

The evidence-based hierarchy, ordered from strongest evidence to weakest, looks like this. Treat it as a priority list, not a checklist — if you do the first tier and nothing else, you've captured most of the available benefit.

The trade-off worth naming honestly: none of these interventions are proven to improve fertility outcomes in a randomised trial. They are plausible based on observational evidence, they are modifiable, and the switching costs are low to zero. For a couple already committed to trying to conceive, and already facing the possibility of expensive clinical intervention, the cost-benefit of lowering modifiable exposures is clearly positive. For couples not actively trying to conceive, the same logic applies but with less urgency.

If you are already working with a fertility specialist, bring the chemical exposure conversation into that clinical relationship. Most reproductive endocrinologists will have opinions on this that range from 'worth doing' to 'not strictly necessary but can't hurt.' Both are reasonable clinical positions given the current state of the evidence. The specific studies cited in this article are real, the effect sizes are real, and the hedging is honest. None of this is a substitute for a clinical evaluation, and this article is not medical advice.

Chemicals and fertility is an uncomfortable topic because the research answers the questions you most want answered — will this specific exposure affect my chances? — in the least satisfying way: with associations, with modest effect sizes, with hedged language, and with a note that no randomised controlled trial has closed the loop. That is the honest state of the field. It is not a reason to ignore the evidence. The associations are consistent enough that a couple trying to conceive has reasonable grounds to reduce the modifiable chemical inputs while they work through the clinical questions with someone qualified.

Sperm counts have fallen by more than half since the 1970s. Testosterone in young American men has dropped by a quarter in under twenty years. Women with PCOS tend to carry more BPA than women without. Maternal phthalate exposure during pregnancy is associated with shortened anogenital distance in male infants — the reproductive development signal that rodent studies predicted decades earlier. Pesticide-residue-heavy produce consumption is associated with lower IVF success rates. The Tian 2024 meta-analysis of 50 studies on phthalates and semen parameters now sits alongside Castellini 2024 on BPA and Chiu 2015 on pesticides as the three anchor findings in the male-side literature. The literature doesn't prove causation for any of these findings individually. Together, they point in the same direction. The appropriate response is neither panic nor dismissal — it's calm, informed action on the parts that are modifiable, combined with the kind of clinical support that no article on the internet can replace.