What is the hygiene hypothesis, and how could it cause autoimmune disease?
The hygiene hypothesis proposes that our modern, sanitized environments deprive the immune system of the microbial 'training' it evolved to need. Without enough exposure to harmless bacteria, parasites, and other microbes—especially in early childhood—the immune system can become poorly regulated. This can lead it to overreact to harmless substances (allergies) or mistakenly attack the body's own cells (autoimmunity). One review explains that reduced pathogen exposure in high-income countries leads to 'immune dysregulation and increase in chronic inflammation, allergy, autoimmunity' [2]. The idea is not that hygiene is bad, but that a too-sterile environment may leave the immune system without the necessary practice to distinguish friend from foe.
A key mechanism involves the gut microbiome. Research shows that people with autoimmune diseases like type 1 diabetes, multiple sclerosis, and rheumatoid arthritis have distinctly different gut bacteria compared to healthy people [3]. Intestinal helminths (parasitic worms) are one factor that can alter the gut microbiome, and their disappearance from modern societies may be a missing immune regulator [3]. This aligns with the hygiene hypothesis: losing our ancient microbial companions may disrupt the gut-immune axis and increase autoimmunity risk.
What does the evidence actually show? Does it support the hypothesis?
The strongest human evidence here comes from a large U.S. study tracking antinuclear antibodies (ANA)—a marker of autoimmunity—in adolescents over 25 years. ANA prevalence jumped from 5% in 1988–1991 to 12.8% in 2011–2012, a more than doubling [1]. Crucially, teens with early-childhood asthma (a hygiene-hypothesis-related condition) were twice as likely to have ANA (odds ratio 2.07), and those with antibodies to the bacterium Helicobacter pylori (a marker of past infection) were half as likely to have ANA (odds ratio 0.49) [1]. This pattern—more allergies and fewer infections linked to more autoimmunity—directly supports the hygiene hypothesis.
Animal experiments provide even more direct causal evidence. In one study, mice raised in a 'dirty' environment with natural mouse pathogens showed decreased self-reactive T cell activity and less organ damage compared to clean-lab mice [4]. Most strikingly, when diabetes-prone NOD mice were co-housed with pet-shop mice (exposing them to diverse microbes), zero of them developed autoimmune diabetes [4]. This is a dramatic demonstration that microbial exposure can prevent autoimmunity in a controlled setting. Another review focusing on type 1 diabetes notes that early-life microbial exposure, gut microbiota diversity, and geographical factors all support the hypothesis, but also acknowledges that genetics and other variables complicate the picture [5].
What are the caveats? Is the hypothesis overblown?
The hygiene hypothesis is not a complete explanation. The same review that supports it for type 1 diabetes also points out that 'current research produces contradictory results and often overlooks factors such as genetics, which also influence the onset of T1D' [5]. For example, while the U.S. adolescent study found a link between asthma and ANA, the association with hay fever was not statistically significant, meaning it could have been due to chance [1]. This suggests the relationship is not simple or universal.
Furthermore, the hypothesis cannot be overgeneralized to all autoimmune diseases or populations. A letter on COVID-19 mortality notes that while the hygiene hypothesis might explain lower death rates in some low-income countries (due to 'trained immunity' from prior infections), factors like climate, age, gender, and reporting differences also matter [2]. The same letter warns that 'the hypothesis cannot be over generalized' and that social vulnerability can actually increase COVID-19 risk in some low-income settings [2]. So while the hygiene hypothesis offers a valuable perspective, it is one piece of a much larger puzzle involving genetics, diet, pollution, and other environmental triggers.
About These Sources
This answer is built on 5 peer-reviewed studies — published from 2022 to 2026, 3 from 2024 or later, 1 in Q1 journals — selected as the most relevant from 5 studies that passed quality screening, drawn from 44 papers retrieved from a database of over 500 million.
Sources used in this answer
Hygiene Hypothesis Indicators and Prevalence of Antinuclear Antibodies in US Adolescents
In a large U.S. study of adolescents (N=2,709) over 25 years, antinuclear antibody (ANA) prevalence rose from 5% to 12.8%. ANA were positively associated with early-childhood asthma (OR 2.07) and negatively associated with H. pylori antibodies (OR 0.49), supporting the hygiene hypothesis.
Is the hygiene hypothesis a plausible explanation for the asymmetry in COVID-19 mortality?
This letter argues the hygiene hypothesis plausibly explains lower COVID-19 mortality in some low-income regions due to 'trained immunity' from prior microbial exposure, but cautions against overgeneralization due to confounding factors like climate and reporting differences.
The gut microbiome–helminth–immune axis in autoimmune diseases
This review proposes that intestinal helminths alter the gut microbiome, and their loss in modern societies (consistent with the hygiene hypothesis) may contribute to rising autoimmune diseases, focusing on type 1 diabetes as a model.
Immune experiences alter tolerance in T cell-mediated autoimmunity.
In mouse experiments, immune-experienced ('dirty') mice showed decreased self-reactive T cell activity and organ pathology. Cohousing diabetes-prone NOD mice with pet-shop mice resulted in zero cases of autoimmune diabetes, directly demonstrating microbial protection.
The Role of the Hygiene Hypothesis in the Pathogenesis of Type 1 Diabetes
This review of type 1 diabetes finds that early-life microbial exposure, gut microbiota diversity, and geographical factors support the hygiene hypothesis, but notes contradictory results and the need to integrate genetics and longitudinal data.