Is biodiversity exposure associated with the risk of asthma?

Two biodiversity metrics give two different answers

The paper asks whether living alongside more environmental biodiversity, or carrying a more diverse human microbiota, is linked to developing asthma. Its central result is that the answer depends on which biodiversity measure is used, and on whether the exposure is measured in the environment or in the human body. The authors describe this as variable associations whose direction and magnitude depend on the specific metrics and analytical approaches.

Two indices carry most of the argument. Bacterial richness counts how many distinct operational taxonomic units or species are recorded in a sample. The Shannon diversity index combines that count with how evenly individuals are distributed across those species, so a community can be rich but score lower if a few types dominate. Because these indices answer different questions about a microbial community, results based on one cannot be substituted for the other.

Eighteen studies behind the pooled estimates

The authors searched Scopus, PubMed MEDLINE and Web of Science for studies published up to 5 March 2024, following PRISMA reporting standards with a protocol registered in PROSPERO. From 1,986 screened records, 18 studies published between 2011 and 2021 entered the meta-analysis, covering 31,048 subjects: 28,119 in the environmental biodiversity assessment (study sizes from 86 to 26,367 participants) and 2,929 in the human microbiota assessment (40 to 923 participants).

Included studies had to use a recognized biodiversity measure, either the Shannon index or bacterial richness, and analyze biodiversity as a continuous variable. Asthma could be defined by self-reported medical diagnosis, medical diagnosis, use of asthma medication, or a GINA-based definition. Environmental biodiversity covered microbiota in dust or soil as well as scores such as species richness indices, land-use gradients and plant diversity; human microbiota biodiversity came from skin, stool, urine and airway samples. Effect estimates were pooled as standardized mean differences or relative risks with 95% confidence intervals, using the Hedges method and a restricted maximum likelihood estimator, and only studies scoring above 3 on the Newcastle-Ottawa Scale were meta-analyzed.

Environmental richness looks protective but is heterogeneous

For environmental biodiversity measured as bacterial richness, the pooled odds ratio was 0.74 (95% CI 0.57–0.96), consistent with the authors' hypothesis that higher exposure has a protective association with asthma development. That estimate comes with two qualifications stated in the same result: heterogeneity across studies was considerable (I² = 71.8%, p = 0.003), and the funnel plot was asymmetric with Egger's test indicating publication bias (p = 0.007).

The Shannon index version of environmental biodiversity did not reach significance: summary OR 0.77 (95% CI 0.55–1.06), with considerable heterogeneity (I² = 72.4%, p = 0.027). Here the funnel plot also looked asymmetric, but Egger's test was not significant (p = 0.195). The two environmental estimates point in the same direction, yet only the richness estimate excludes 1.0, so the protective reading rests on that narrower result rather than on environmental biodiversity in general.

Human microbiota diversity tracks with higher asthma odds

The human microbiota results run opposite to the authors' original hypothesis. Across 6 studies reporting the Shannon index, people with asthma had higher diversity, with a pooled standardized mean difference of 0.31 (95% CI 0.14–0.48) and essentially no heterogeneity (I² = 0%, p = 0.88). Across 4 studies reporting bacterial richness or abundance, the pooled standardized mean difference was 0.29 (95% CI 0.10–0.49), again homogeneous (I² = 0%, p = 0.97). Neither analysis showed significant publication bias by Egger's test.

A fourth analysis asked a different question with a different metric: the odds ratio for bacterial richness and asthma across 4 study-specific estimates was 1.14 (95% CI 0.83–1.56), not statistically significant, with significant heterogeneity (I² = 62.0%, p = 0.048). The standardized mean difference compares diversity levels between people with and without asthma, whereas the odds ratio estimates association with asthma risk; the paper reports both, and the significant and non-significant results should not be collapsed into one claim.

Why environment and microbiota may diverge

The authors interpret the environmental richness finding through the biodiversity hypothesis: contact with diverse environmental microbes may modulate the immune system and reduce susceptibility to asthma. They note that several included studies examined early-life exposure, prenatal and early childhood, as a period that may strongly shape asthma and allergic disease pathogenesis, while other cited work argues that immune tolerance needs constant, lifelong microbial exposure.

For the human microbiota result, the authors offer a different explanation rather than treating diversity as automatically beneficial. High bacterial diversity in the body could reflect pathogenic bacteria disturbing local bacterial ecology and immune balance, and an imbalance in microbial taxa, with loss of beneficial organisms and expansion of potentially harmful ones, has been linked to asthma development. They also flag a gap: the human microbiota studies included in the meta-analysis did not assess the joint association between environmental and human microbial diversity and asthma development.

What the pooled numbers cannot settle

The authors caution that the summary results rest on a small number of studies and on heterogeneity in how exposure and outcome were defined, in sampling methods, in sample types for human microbiota, and in temporal variation in biodiversity. Asthma definitions varied enough that prevalence, and therefore effect estimates, could be under- or overstated, affecting consistency and comparability. Heterogeneity may also stem from cross-sectional versus longitudinal design, study duration, timing and duration of exposure, and adjustment for confounders such as socioeconomic status, medication use, diet and exposure to natural or built-up areas.

Because too few studies were available, the authors could not run sensitivity analyses by likely sources of heterogeneity, such as asthma definition or exposure timing. Measures such as Chao 1 and the inverse Simpson index appeared in some studies but were too rare to meta-analyze or to support a sensitivity analysis by biodiversity index. These constraints qualify the pooled estimates themselves, not just their interpretation.

Where the metric choice changes the message

The practical priority this paper establishes is measurement discipline. If the question is whether nearby environmental microbial richness is associated with asthma, the evidence points toward a protective association (OR 0.74) that is weakened by heterogeneity and publication bias. If the question is whether environmental biodiversity measured by the Shannon index matters, the pooled estimate does not exclude no association. If the question is about human microbiota, the standardized mean difference analyses point toward higher diversity among people with asthma, while the odds-ratio analysis for richness does not reach significance.

The authors connect these findings to policies that expand natural areas, especially in cities, and to biodiversity interventions such as encouraging time in green spaces, gardening, enriching playgrounds with microbially diverse vegetation and organic soil, and integrating green walls into built environments. Those suggestions follow from the environmental richness result, not from the human microbiota findings, and they remain associational: the meta-analysis summarizes observational studies and does not establish that changing biodiversity changes asthma risk.

Sources used in this answer