Pillar
Environment
Public claims about chemicals and autoimmune disease run further ahead of the evidence than almost anywhere else in medicine. People can't be assigned to a chemical, so the human research watches instead, and watching can't separate an exposure from the job, the income, and the neighborhood it comes with. What follows is what the studies measured, with their sizes, so you can judge them.
What the research found.
One national United States survey looked at PFOA, which is a PFAS chemical. It was associated with rheumatoid arthritis at an odds ratio of 1.63, with the true figure put between 1.41 and 1.89. Two related chemicals came in at 1.41 and 1.40, both well clear of one.
Lv and colleagues, International Journal of Molecular Sciences, 2025
That survey also looked at the whole PFAS mixture together, rather than one chemical at a time. The link got a great deal smaller than for any single chemical, and that is the figure closest to real life, because exposure is never to one chemical alone. The odds ratio was 1.06, put between 1.02 and 1.10.
Lv and colleagues, International Journal of Molecular Sciences, 2025
The VITAL trial gave 25,871 adults vitamin D or a dummy pill, following them a median of 5.3 years. Fewer people on vitamin D got a new autoimmune disease. The hazard ratio was 0.68, put between 0.48 and 0.94.
Two years after that trial stopped, the fish oil group no longer showed a real effect. The hazard ratio was 0.85, with a range from 0.67 to 1.08, so it crossed one and the effect had gone. Reading the trial and its follow-up together gives a different answer from reading either one alone.
Costenbader and colleagues, Arthritis and Rheumatology, 2024
Eating organic food was associated with lower C-reactive protein in a snapshot of 3,815 American adults aged over fifty. The figure was minus 0.080 once overall diet quality was allowed for. No autoimmune outcome was measured in that study at all, which is the limit most worth knowing about it.
One review pooled 10 studies and 4552 people with rheumatoid arthritis. Against never smoking, risk rose 26 percent at 1 to 10 pack-years, at a relative risk of 1.26 from 1.14 to 1.39. It roughly doubled at 21 to 30 pack-years, at 1.94 from 1.65 to 2.27.
Di Giuseppe and colleagues, Arthritis Research and Therapy, 2014
One review pooled six case-control studies on silica and ANCA-associated vasculitis. Ever having been exposed came out at an odds ratio of 2.56, from 1.51 to 4.36. It was 3.95 for microscopic polyangiitis and 3.56 for granulomatosis with polyangiitis.
One review pooled 31 studies on things breathed in at work, and seven came out linked to rheumatoid arthritis. Those seven were silica, asbestos, solvents, pesticides, fertilizers, animal dust, and engine exhaust. The relative risks ranged from 1.25 to 1.49, which is a modest size and a wide reach.
One Swedish study covered 11285 people with rheumatoid arthritis, where silica-exposed men came out at 1.4 for antibody-positive disease, from 1.2 to 1.6. That figure rose to 2.3 with more years of exposure behind it. A job is what gets measured here, and not a habit, which sets silica apart from smoking.
One Boston study covered 95 women with lupus, where silica exposure lasting more than a year came out at 4.3, from 1.7 to 11.2. Exposure there was rated by reviewers who didn't know who had lupus, which removes one of the most common ways this kind of study goes wrong. That Boston study found nothing for solvents, at 1.04 from 0.34 to 3.2.
One genetic analysis, built to test cause instead of association, found nothing at all for any of four pollutants in rheumatoid arthritis. Fine particulate matter came in at 0.71, from 0.27 to 1.91, and nitrogen dioxide at 0.482, from 0.173 to 1.343. Both of those ranges cross one, so neither of them counted as a real effect at all.
A second genetic analysis asked the same question and disagreed with the first one. Nitrogen oxides came in at 1.47 for rheumatoid arthritis, from 1.01 to 2.14, and at 7.26 for lupus, from 2.25 to 23.40. A range spanning a tenfold reach is a question rather than an answer, so the studies here neither show cause nor rule it out.
Why this subject is hard
Two things make this subject hard, and neither of them is scientists being careless or anybody hiding anything. The first is that people can't be assigned to a chemical, so the human evidence measures what people were exposed to and then sees who gets ill, and people with more exposure differ in job, income, housing, and diet. Adjusting for that never removes all of it. The second is that lab work on how chemicals affect immune cells is far more common than human work on who gets sick, so what you find to read leans heavily toward the dish.
The exposures with the most behind them
Smoking beats everything else here, because it's the only exposure on this page where risk climbs as exposure climbs, which is one of the better signs that a link is real. Silica comes next, and it's the one exposure where a job is the thing being measured. Three research groups looked at three different diseases, in rheumatoid arthritis, lupus, and ANCA-associated vasculitis, and all three pointed one direction. That agreement is why silica sits apart from everything else on this page, and the sizes with their ranges are in the list above.
PFAS, and what a mixture figure means
PFAS stands for per- and polyfluoroalkyl substances, and they're the exposure with the most usable evidence in joint disease. Coverage of the main analysis quotes the single-chemical figures and stops, and the next figure in the same paper changes what those mean, because the whole mixture together came out a great deal smaller. The survey behind it is also a snapshot, so exposure and disease were measured at one time point and neither can be shown to have come first. That's a real limit instead of a quibble, and it's the reason this page says concern and not proof.
What this is, and what it isn't
Not one of the studies here assigned an exposure to anybody, which makes a link a reason to look further and not proof that changing the exposure changes the disease. No study has tested whether cutting your exposure changes a disease you already have, so nothing on this page is a treatment. Anyone offering exposure reduction instead of your medicine is going far past the evidence, and acting on that risks organ damage no later treatment undoes. Cutting avoidable exposure is still sensible on its own terms, it needs no part of the disease evidence above to hold up, and there are other good reasons to want less of these chemicals in you.
What to do this week.
Filter your drinking water
This is the change with the most return for the least effort and the least expense, and it goes at the exposure with the most usable evidence behind it. After the filter it costs you nothing.
Keep food off hot plastic
Heat is what moves these chemicals out of packaging and into what you eat. Use glass or steel for hot food and for reheating, and let leftovers cool before they go into a plastic container.
Deal with household dust
A good deal of what ends up in people arrives through dust. Damp-dusting, a vacuum with a decent filter, and taking shoes off at the door handle most of it, and not one of those needs anybody selling you anything.
Treat the disease with what treats the disease
Cut exposure because it's sensible, and keep taking what your rheumatologist prescribed. No study has tested exposure reduction as a treatment for a disease you already have, so it isn't one, however it gets offered to you.
Common misconceptions.
Myth. This is just wellness paranoia.
Reality. PFAS-RA association is from NHANES, the largest population health dataset the US has. BPA, phthalates, and organochlorines are studied in peer-reviewed toxicology and epidemiology journals. The science is robust; the policy lag is the problem.
In plain words. These chemicals were studied in real science journals. One study used the biggest health survey the country has. What lags behind is the law, not the research.
Questions patients ask.
Do environmental chemicals cause autoimmune disease?
Some are associated with it, and proving cause is another thing entirely. PFAS chemicals show a stable link with rheumatoid arthritis in survey data, though that data is a snapshot, so exposure and disease were measured at one time point. People with more exposure also differ in job, income, and where they live, and those differences are hard to strip out. A link this size earns concern rather than certainty.
Should I be worried about PFAS specifically?
PFAS have the best data on this page, and the numbers still need care. Single chemicals gave odds ratios between 1.40 and 1.63 for rheumatoid arthritis, while the whole mixture together gave only 1.06. That drop is the part to hold onto, because exposure is never to one chemical alone, so the mixture figure is the one closest to real life.
Can reducing my exposure reverse autoimmune disease I already have?
No study has tested that, and it counts to be clear about it rather than vague. All the human evidence here is about who gets a disease in the first place, rather than about what happens to a disease you already have. Cutting exposure is sensible for other reasons. Calling it a treatment goes past the research, and it can delay the treatment that does work.
Is organic food worth the cost?
One large snapshot study found organic food was associated with lower C-reactive protein, and the effect shrank once diet quality was allowed for. Its own authors say healthy eating and income could explain the rest. No autoimmune outcome was measured in it at all, so what you have is a small link with one blood marker rather than a claim about disease.
What about the number of chemicals approved for use in food?
Numbers like that get passed around with no source, which is why this page gives no figure at all. Counting approved food substances depends entirely on what you decide to count, and the figures people quote trace back to other websites rather than to a government list. A number you can't check is no use to you, so ask where it came from every time.
What does the one randomized prevention trial tell me?
VITAL gave 25,871 adults vitamin D, fish oil, both, or neither, and followed them a median of 5.3 years. Vitamin D cut new autoimmune disease at a hazard ratio of 0.68. Two years after the pills stopped, the fish oil effect was gone. That's the only trial evidence that anything prevents autoimmune disease, and it's about taking a pill rather than about avoiding a chemical.
References.
- Lv Y; Zhao C; Xiang Y et al. Perfluoroalkyl Substance (PFAS) Mixtures Drive Rheumatoid Arthritis Risk Through Immunosuppression: Integrating Epidemiology and Mechanistic Evidence. International journal of molecular sciences. 2025;26. 10.3390/ijms26157518Cross-sectional
- Hahn J; Cook N; Alexander E et al. Vitamin D and marine omega 3 fatty acid supplementation and incident autoimmune disease: VITAL randomized controlled trial. BMJ. 2022;376:e066452. 10.1136/bmj-2021-066452VITAL: nationwide
- Costenbader K; Cook N; Lee I et al. Vitamin D and Marine n‐3 Fatty Acids for Autoimmune Disease Prevention: Outcomes Two Years After Completion of a <scp>Double‐Blind</scp>, <scp>Placebo‐Controlled</scp> Trial. Arthritis & Rheumatology. 2024;76:973-983. 10.1002/art.42811Two years of observational follow-up after the end of the VITAL randomised trial
- Ludwig-Borycz E; Guyer HM; Aljahdali AA et al. Organic food consumption is associated with inflammatory biomarkers among older adults. Public health nutrition. 2021;24:4603-4613. 10.1017/S1368980020005236Observational
- Di Giuseppe D; Discacciati A; Orsini N et al. Cigarette smoking and risk of rheumatoid arthritis: a dose-response meta-analysis. Arthritis Res Ther. 2014;16:R61. 10.1186/ar4498Dose-response random-effects meta-regression of 3 prospective cohorts and 7 case-control studies
- Gómez-Puerta JA; Gedmintas L; Costenbader KH. The association between silica exposure and development of ANCA-associated vasculitis: systematic review and meta-analysis. Autoimmun Rev. 2013;12:1129-35. 10.1016/j.autrev.2013.06.016Systematic review and meta-analysis of silica exposure and ANCA-associated vasculitis
- Ilar A; Klareskog L; Saevarsdottir S et al. Occupational exposure to asbestos and silica and risk of developing rheumatoid arthritis: findings from a Swedish population-based case-control study. RMD Open. 2019;5:e000978. 10.1136/rmdopen-2019-000978Swedish population-based case-control study
- Liu Q; Song X; Mauro E et al. Exposure to Occupational Inhalants and the Risk of Developing Rheumatoid Arthritis: A Systematic Review and Meta-Analysis. Arthritis Rheumatol. 2026;78:830-847. 10.1002/art.43446Systematic review and meta-analysis following PRISMA
- Finckh A; Cooper GS; Chibnik LB et al. Occupational silica and solvent exposures and risk of systemic lupus erythematosus in urban women. Arthritis Rheum. 2006;54:3648-54. 10.1002/art.22210Case-control study in four predominantly African American neighborhoods in Boston
- Cooper GS; Wither J; Bernatsky S et al. Occupational and environmental exposures and risk of systemic lupus erythematosus: silica, sunlight, solvents. Rheumatology (Oxford). 2010;49:2172-80. 10.1093/rheumatology/keq214Case-control study through the Canadian Network for Improved Outcomes in SLE
- Jiang F; Li S; Jia C. Smoking and the risk of systemic lupus erythematosus: an updated systematic review and cumulative meta-analysis. Clin Rheumatol. 2015;34:1885-92. 10.1007/s10067-015-3008-9Systematic review and cumulative meta-analysis of 12 published articles reporting 13 case-control or cohort studies from 1990 to 2015 on smoking and systemic lupus erythematosus
- Chua MHY; Ng IAT; W L-Cheung M et al. Association Between Cigarette Smoking and Systemic Lupus Erythematosus: An Updated Multivariate Bayesian Metaanalysis. J Rheumatol. 2020;47:1514-1521. 10.3899/jrheum.190733Updated multivariate Bayesian meta-analysis of 12 eligible studies comprising 3
- Parisis D; Bernier C; Chasset F et al. Impact of tobacco smoking upon disease risk, activity and therapeutic response in systemic lupus erythematosus: A systematic review and meta-analysis. Autoimmun Rev. 2019;18:102393. 10.1016/j.autrev.2019.102393Systematic review and meta-analysis of smoking and systemic lupus erythematosus across disease risk
- Yang C; Du Z; Ma J et al. Causal relationship between air pollution and rheumatoid arthritis: A two-sample Mendelian randomization study. Medicine (Baltimore). 2025;104:e42901. 10.1097/MD.0000000000042901Two-sample Mendelian randomisation using genome-wide association data for four air pollution metrics as exposures and European rheumatoid arthritis cohorts as outcomes
- Zhang M; Wang Y; Hu S et al. Causal relationships between air pollution and common autoimmune diseases: a two-sample Mendelian randomization study. Sci Rep. 2025;15:135. 10.1038/s41598-024-83880-9Two-sample Mendelian randomisation study using published genome-wide association data
This page gathers the published research on this subject into one place. The studies behind it were published between 1989 and 2026, and every figure links to the paper it came from. Those studies were peer reviewed. This summary of them was not. Dr. Sarah Luebker is reviewing these pages one at a time and has not reached this one yet, so it carries no medical review date and nothing here is her opinion or her advice to you. Each page gets updated as she reaches it. It is here in the meantime because the science is worth having in one organized place that is easy to find and easy to read. Talk to your own clinician before acting on any of it.