Lung Bugs, Brain Fog, and the 2am Chatbot

The Bacteria Living in Your Lungs May Affect Your Anxiety

What if the bugs living in your lungs — not just your gut — are quietly dialing your anxiety up or down?

You've probably heard that gut bacteria can affect mood. This study pushes that idea somewhere genuinely new: the lungs. Using mice bred to develop asthma-like symptoms, researchers identified a subset that also developed anxiety-like behaviour. When they transplanted lung bacteria from those anxious mice into healthy mice, the healthy mice became anxious too. That single experimental move is important: it implicates the lung microbiome — the community of microbes living in your airways — as a causal player, not just a passenger. The mechanism they traced involves a compound called formononetin, produced by a family of lung bacteria called Acidobacteria. Think of formononetin as a key that unlocks a particular door in the brain. That door is the BDNF/TrkB pathway — a signalling system that neurons use to stay healthy and communicate. When formononetin is present, it calms overactivated brain immune cells called microglia in the hippocampus, the region most tied to memory and emotion. When the researchers blocked that door with a drug called ANA-12, the anxiety-relieving effect disappeared entirely, confirming the pathway was essential, not coincidental. The catch is significant. This is entirely in mice — asthma-specific mice at that. Whether the same pathway operates in humans with asthma, let alone in people with anxiety disorders more broadly, is unknown. Formononetin appears in some medicinal plants, so human exposure isn't zero, but a clinical application is a long way off. A small but real mechanistic step — and one that expands our map of where brain-relevant signals originate.

Why Major Surgery Can Leave Older People Mentally Foggy

Between 10% and 65% of elderly patients come out of major surgery with lasting memory and thinking problems — and a new review has a specific molecular suspect.

If you've watched an older family member seem somehow different after a major operation — slower, more forgetful, less themselves — you've witnessed what doctors call perioperative neurocognitive disorder, or PND. It's distressingly common and poorly understood. A review published in Neuroscience Bulletin now argues the culprit may be a process called RNA splicing going wrong inside brain cells under surgical stress. Here is the analogy. Your DNA is like a long raw recording of everything your body knows how to do. Before a cell can use that information, molecular editors cut the tape, discard filler sections, and stitch the useful parts together — that is RNA splicing. When those editors make mistakes, wrong proteins get built: some trigger inflammation, others disrupt the junctions between neurons that make memory and thought possible. The review synthesises evidence showing that anaesthesia and surgical trauma — especially in older brains with less repair capacity — push these molecular editors into error-prone states. The cascade that follows includes inflammation in the hippocampus, disrupted neurotransmitter receptors, and impaired cognitive function. The authors point to specific splicing errors, including ones affecting a receptor called P2X7 that controls inflammatory signals, and a splicing factor called SLM2 that shapes how neurons connect. The honest catch: this is a narrative review. The authors read existing studies and built a theory — they collected no new data. Several cited findings come from animal models or small mechanistic experiments. Whether targeting RNA splicing will produce usable prevention strategies for PND in humans remains a genuinely open question. But it gives researchers a more specific molecular address to investigate, which is progress.

For Some People With Depression, Chatbots Help Manage Daily Emotions

At 2am when you're spiralling and there's no one to call, does it actually help to talk to a chatbot?

The study published in Acta Psychologica looked at Chinese adults who reported depressive symptoms and examined how they interacted with social chatbots — AI companions built for conversation and emotional support, not clinical therapy — in their everyday lives. The finding: those who formed what researchers call a parasocial relationship with a chatbot showed better emotion regulation and greater psychological resilience. Think of it like a pressure valve. When emotional pressure builds and no human outlet is available, having something to talk to — even a non-human something — may release enough tension to prevent a worse spiral. The chatbot does not fix the underlying problem. But it may stop things from deteriorating in the moment, which is not nothing when you consider how many people have depressive symptoms and no immediate access to professional support. The honest limits here are considerable, and they matter. This is an observational study, meaning we cannot tell whether chatbots caused better resilience, or whether more resilient people are simply more likely to try chatbots in the first place. Participants self-reported their symptoms rather than receiving a clinical diagnosis — a real difference. The study was conducted in China, where the cultural context around mental health and technology use is distinct; results may not transfer directly elsewhere. None of that makes the finding uninteresting. It means: chatbots might be genuinely useful as a daily coping tool for some people, and that deserves proper longitudinal investigation — not reflexive dismissal, and not uncritical enthusiasm either.

Step back and look at today's three stories together. Lung bacteria shaping anxiety through a molecular messenger. Genetic editors in brain cells going haywire after the stress of surgery. A chatbot acting as a pressure valve for people who have nowhere else to turn at 2am. What they share is a view of mental health as a whole-body, whole-life problem — one that doesn't begin and end in a therapist's office. The lung microbiome story expands our map of where brain-relevant signals come from. The RNA splicing story gives clinicians a more specific molecular address for a real and underdiagnosed post-surgical condition. The chatbot study asks us to be honest about the gap between who needs mental health support and who actually has access to it — and whether technology can meaningfully, safely bridge part of that gap. All three stories are early. None are finished answers. But they are asking the right questions.

The lung microbiome story is the one I'd follow most closely: if formononetin shows similar effects in human asthma-anxiety studies, that opens a genuinely novel therapeutic direction. On chatbots, watch for regulatory signals — the FDA has been quietly developing a framework for AI-based mental health tools, and clearer guidance could arrive within the year. The open question I'd most want answered from today's chatbot study: do any benefits persist over months of use, or do they fade as novelty wears off?

• Wikipedia: RNA splicing — what it is and how it works
• Wikipedia: Gut–brain axis (the broader framework the lung study builds on)
• Wikipedia: Parasocial interaction — the psychology behind chatbot bonding