Background and ObjectivesHigh adherence to the Mediterranean diet (MeDi) has been associated with slower age-dependent cognitive decline and better cardiovascular health (CVH). We examined the association between adherence to MeDi and white matter (WM) integrity in community-dwelling Hispanic or Latino adults. In secondary analysis, we assessed whether CVH and WM integrity were pathway variables between MeDi and global cognition (GC).MethodsData from Study of Latinos-Investigation of Neurocognitive Aging-MRI Ancillary Study were analyzed. Dietary intake was collected during the baseline visit (2008-2011) using 24-hour recalls, from which a MeDi score ([MeDiS], range 0-9) was derived. Brain MRI scans with diffusion tensor imaging were obtained between 2017 and 2022. WM integrity was assessed using total WM volume (tWM), WM hyperintensity (WMH) volume, fractional anisotropy (FA), and free water (FW). GC was ascertained between 2015 and 2018 using a composite score derived from 4 standardized cognitive tests. CVH was evaluated at baseline using the Life's Essential 7 score ([LE7], range 0-100), a modified version of the existing Life's Essential 8 score, in which diet was excluded to avoid collinearity with our exposure. We used linear regression models that controlled for age, sex, and socioeconomic factors to investigate the association of MeDiS with WM integrity. We performed mediation analysis to explore whether CVH and WM integrity were pathway variables between diet and GC.ResultsA total of 2,642 participants with a mean age of 64.3 years (95% CI 63.4-65.1, 44% male) were included. The average MeDiS was 5.0 (95% CI 4.9 to 5.1), and the LE7 score was 66.6 (95% CI 65.-67.6). Higher MeDiS was associated with lower WMH volume (beta = -0.08, 95% CI -0.11 to -0.04), higher tWM volume (beta = 0.05, 95% CI 0.005-0.09), lower FW (beta = -0.04, 95% CI -0.08 to -0.002), and higher fractional anisotropy (FA) (beta = 0.09, 95% CI 0.05-0.13). WMH, tWM, and FA mediated the association between MeDiS and GC. In addition, there was serial mediation from MeDiS on GC through LE7 score, WMH, tWM, and FA.DiscussionHigher adherence to MeDi is linked to better WM structural integrity, which, together with CVH, mediates the association between MeDi and GC.
Invertebrate pollination and herbivory dynamics of rare orchids are poorly documented. This is particularly true for species of Platanthera, complicating efforts to conserve rare species within this widespread genus. We report five new pollinators and several associated insect interactions for the U.S. federally threatened white fringeless orchid (Platanthera integrilabia) based on 3–18 years of direct field observations, including remote (Raspberry Pi) cameras, across seven Kentucky wetlands. We describe sphinx moth pollination for the first time in this species: Hyles lineata, Eumorpha pandorus. Dolba hyloeus, and Hemaris diffinis at open wetlands, and documented the northern cloudywing skipper butterfly (Thorybes pylades) in semi-open forested wetlands. Habitat characteristics, including canopy cover and floristic diversity, strongly influenced pollinator presence; orchid population size and proximity to open corridors and diverse herbaceous vegetation predicted sphinx moth pollinator presence and frequency. Across sites and years, multiple diurnal and nocturnal Lepidoptera visited or pollinated P. integrilabia, with differing pollinia transfer efficiency and visitation frequency rates, indicating functional redundancy and response diversity that buffers reproduction when individual species fluctuate, a lower risk strategy than reliance on a single specialist. Numerous other visitors (butterflies, bees, hummingbirds) foraged on flowers without carrying pollinia. Invertebrate predators (spiders, praying mantids), and seven species of invertebrate herbivores (including aphids, weevils, caterpillars) were also documented. Among these, aphids are suspected contributors to arrested inflorescence development, potentially lowering fecundity. By documenting these interactions, we provide conservationists and land managers actionable insight for improving the viability of P. integrilabia and other native orchids threatened with extinction.
Clinical AI systems' lack of interpretability limits their adoption in evidence-based medicine. To address this challenge, we propose a computational framework that harnesses generative AI's medical knowledge to create interpretable structural causal models (SCMs) for clinical decision support, quality improvement evaluation, and population health management. We evaluated our approach through a case study using data from the Midwest Healthcare Conference Causal Diagram Challenge, where we compared transformer-based large language models against human performance on a complex causal reasoning task: estimating COVID- 19 treatment effects through target trial emulation. Both groups designed SCMs to evaluate glucocorticoid treatment effects on 28-day mortality using real-world data from more than 2,000 hospitalized patients, benchmarked against published RECOVERY randomized controlled trial results. The best performing SCMs achieved bootstrap coverage rates exceeding 90% for two of three severity strata. Both human and AI models demonstrated equivalent clinical plausibility (n=3 expert reviewers) and similar statistical performance, though both struggled with critical disease severity. Ablation experiments comparing SCM-based approaches against traditional potential outcomes methods revealed SCMs achieved 76-98% coverage versus 1-37% for traditional methods. These results suggest that structural causal models can effectively bridge the interpretability gap in clinical AI by providing essential scaffolding for reliable causal inference and enabling meaningful human-AI collaboration while preserving methodological rigor essential for evidence-based medicine.
Ectomycorrhizal fungi (EMF) produce mycelia with variable extension and complexity, which can be classified according to soil 'exploration types' (ETs). ETs have received attention as one of the few mycorrhizal trait frameworks, but without an empirical classification of ET functional diversity and environmental preferences, understanding and interpreting EMF biogeographic patterns has been difficult. We conducted a synthesis combining: comparative EMF genomics to describe functional divergence in decomposition and nutrient cycling genes across ETs; and EMF trait distribution modeling across continental Europe, pairing soil and root EMF surveys to establish biogeographic ET niche profiles. We demonstrate a signature of ETs encoded in EMF genomes, which is independent from phylogeny and linked to biomass production strategies. EMF ET relative abundances were separated by soil, root, and dominant tree leaf type habitats and exhibited unique correlations with forest biotic (e.g. plant productivity and plant pathogen densities) and abiotic (e.g. nitrogen deposition and soil pH) conditions. These findings support a theory that EMF niche partitioning can be partially explained by extraradical mycelial traits, with underlying variation in ET biogeography likely arising from distinct decomposition and nutrient cycling potentials. We also identify important limitations to this trait framework and provide a guided outlook for future research.
