<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Tejal Daga on Medium]]></title>
        <description><![CDATA[Stories by Tejal Daga on Medium]]></description>
        <link>https://medium.com/@dagatejal?source=rss-79076ec3d006------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*tgs1Kw93xNqXoSmb</url>
            <title>Stories by Tejal Daga on Medium</title>
            <link>https://medium.com/@dagatejal?source=rss-79076ec3d006------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 05:44:37 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@dagatejal/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Microsoft 365, 2026: When Failover Became the Failure]]></title>
            <link>https://medium.com/@dagatejal/microsoft-365-2026-when-failover-became-the-failure-a8bc7eab7c67?source=rss-79076ec3d006------2</link>
            <guid isPermaLink="false">https://medium.com/p/a8bc7eab7c67</guid>
            <category><![CDATA[technical-program-manager]]></category>
            <category><![CDATA[risk-management]]></category>
            <category><![CDATA[incident-management]]></category>
            <category><![CDATA[microsoft]]></category>
            <category><![CDATA[case-study]]></category>
            <dc:creator><![CDATA[Tejal Daga]]></dc:creator>
            <pubDate>Wed, 18 Nov 2026 22:13:25 GMT</pubDate>
            <atom:updated>2026-11-18T22:16:24.627Z</atom:updated>
            <content:encoded><![CDATA[<p><em>A program-level review of the November 12–13, 2026 Microsoft 365 outage: latent defects, capacity risk, cascading failover and dependency management.</em></p><p>On November 12, 2026, most Microsoft 365 users lost mailbox access for nearly 24 hours. Microsoft’s <a href="https://next.ink/wp-content/uploads/2026/11/PostIncidentReport.pdf">preliminary Post Incident Report (PIR)</a> attributes the outage to three conditions that converged on the networking gateway fronting Exchange Online. None was critical on its own.</p><p>I focus on the controls rather than the defect: what should have prevented, detected or contained the failure. Microsoft’s final PIR is due November 20, so some findings may change.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*M6ECznBcGyE6S0D3.jpg" /></figure><h3>Incident summary</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5XOql9U9yIdxZQRCtyudEQ.png" /></figure><h3>What happened</h3><p>Connections to Exchange Online and many other Microsoft 365 services pass through a networking gateway, a fleet of Windows-based front-door servers. Three contributing factors converged on it:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Lg5DaZeEuhUGqSFQ0YejRQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CIJfo6XvgbHnkHS-dG4FoA.png" /></figure><p>Teams, OneDrive, SharePoint and Office apps could fail over to an alternate path and recovered earlier. Exchange Online could not, “for scale reasons,” and because SharePoint notifications, Teams meeting scheduling and Copilot compliance records all depend on Exchange, its outage widened the blast radius (<a href="https://office365itpros.com/2026/11/16/exchange-online-outage-nov26/">Office 365 for IT Pros</a>). Microsoft EVP Charles Lamanna stated there was “no indication this is the result of malicious or cyber activity” and described the event as “cascading failures + thundering herds at our network perimeter” (<a href="https://office365itpros.com/2026/11/16/exchange-online-outage-nov26/">Office 365 for IT Pros</a>, <a href="https://www.itpro.com/software/microsoft/microsoft-365-outage-company-exec-lifts-lid-on-the-cascading-failure-that-caused-lengthy-disruption">ITPro</a>).</p><h3>Analysis: a compound failure amplified by its own safety net</h3><p>Each contributing factor was within tolerance on its own. The failure sat in two places: risk aggregation, since nothing assessed the factors together, and recovery automation, which turned a load-dependent fault into a fleet-wide one.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*whKv7U0mh1qFy51XsVXvbg.png" /></figure><p>Quotes from the <a href="https://next.ink/wp-content/uploads/2026/11/PostIncidentReport.pdf">preliminary PIR</a>.</p><p>The automated recovery gap is the most instructive. Failover assumes a node-level fault. When the fault is load-dependent and present on every node, redirecting traffic amplifies it instead of isolating it, the thundering-herd pattern Lamanna described. Microsoft’s completed repair items target exactly this: monitoring of redirection patterns and guardrails on traffic distribution.</p><h3>Secondary impact: incident comms as an attack surface</h3><p>On November 13, <a href="https://www.microsoft.com/en-us/security/blog/2026/11/16/high-volume-phishing-campaign-uses-microsoft-365-outage-lures-target-multiple-industry-sectors-2/">Microsoft Threat Intelligence</a> observed a high-volume phishing campaign posing as Microsoft 365 service-health notices about the outage.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3ISzW4VyotoKb6Q3WsWqsg.png" /></figure><p>Microsoft’s guidance is that legitimate service-health notices never send users to non-Microsoft domains to recover mail. The program lesson: the incident communication plan is part of the response, and an unclear status channel is a gap attackers will fill.</p><h3>Pre-incident RAID view</h3><p>Mapping the PIR onto a RAID log as it might have stood on November 11 shows where the outage was hiding: in the assumptions column.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WuqY8gugPeUcKE1Wr1euXQ.png" /></figure><p>Each assumption was reasonable on its own. None was tracked as a risk with an owner. Beyond asking what happens when a single component fails, this incident raises a harder question: <strong>what happens when several routine things happen at once?</strong></p><h3>Takeaways: the engineering and TPM lenses</h3><p>The engineering lens asks how to build the system so the failure can’t happen. The TPM lens asks how to run the program so that, if it does, it is detected, owned and contained.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*K7ZwHrACOfr6Yu7Bd9I1jQ.png" /></figure><p>Engineering reduces the probability of failure. Program management reduces its blast radius by making combined risk visible, owned and reviewed together.</p><h3>Sources</h3><p><strong>Official Microsoft sources</strong></p><ul><li><a href="https://next.ink/wp-content/uploads/2026/11/PostIncidentReport.pdf">Microsoft, Preliminary Post Incident Report for Microsoft 365, EX1550543 and MO1550821</a>, November 15, 2026. Republished copy; Microsoft 365 admins can view the original in the admin center.</li><li><a href="https://www.microsoft.com/en-us/security/blog/2026/11/16/high-volume-phishing-campaign-uses-microsoft-365-outage-lures-target-multiple-industry-sectors-2/">Microsoft Security Blog, “High-volume phishing campaign uses Microsoft 365 outage lures to target multiple industry sectors”</a>, November 16, 2026</li></ul><p><strong>Coverage and analysis of the report</strong></p><ul><li><a href="https://bleepingcomputer.com/news/microsoft/microsoft-says-months-old-bug-caused-global-microsoft-365-outage">BleepingComputer, “Microsoft says months-old bug caused global Microsoft 365 outage”</a>, November 16, 2026</li><li><a href="https://www.itpro.com/software/microsoft/microsoft-365-outage-company-exec-lifts-lid-on-the-cascading-failure-that-caused-lengthy-disruption">ITPro, “Microsoft 365 outage: Company exec lifts lid on the ‘cascading failure’ that caused lengthy disruption”</a></li><li><a href="https://office365itpros.com/2026/11/16/exchange-online-outage-nov26/">Office 365 for IT Pros, “The Great Exchange Online Meltdown of November 2026”</a>, November 16, 2026</li><li><a href="https://office-watch.com/?p=78199">Office Watch, “Microsoft 365 outage: what went wrong, in plain English”</a>, with the full preliminary PIR text</li><li><a href="https://www.theguardian.com/technology/2026/nov/16/microsoft-global-outage-cause">The Guardian, Microsoft global outage cause</a>, November 16, 2026</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a8bc7eab7c67" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Your Gut Was Listening the Whole Time…]]></title>
            <link>https://medium.com/@dagatejal/your-gut-was-listening-the-whole-time-e753c1f90360?source=rss-79076ec3d006------2</link>
            <guid isPermaLink="false">https://medium.com/p/e753c1f90360</guid>
            <category><![CDATA[mental-health]]></category>
            <category><![CDATA[gut-health]]></category>
            <category><![CDATA[nutrition]]></category>
            <category><![CDATA[science]]></category>
            <category><![CDATA[stress-management]]></category>
            <dc:creator><![CDATA[Tejal Daga]]></dc:creator>
            <pubDate>Mon, 16 Nov 2026 14:06:01 GMT</pubDate>
            <atom:updated>2026-11-16T14:06:01.514Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Part 2 of “I Studied Depression in 2018. In 2026, the Research Got Personal”: the gut, metabolism, movement and food</em></p><p>In <a href="https://medium.com/@dagatejal/i-studied-depression-in-2018-in-2026-the-research-got-personal-5a497af2f890">Part1</a> I wrote about how, in 2026, a long stretch of anxiety, panic attacks, broken sleep and repeated failure sent me back to the research I’d first read as a student in 2018. That part covered the brain: helplessness, overthinking, stress chemistry, and what the evidence says helps.</p><p>But the deeper I read, the more one thing kept coming up that I hadn’t expected: THE GUT.</p><p>I’d always filed “gut feeling” under metaphor. It turns out the gut and brain are wired together, a single stressful event can change the gut, gut microbes help make chemicals the nervous system uses, and blood sugar, movement and diet are all linked to mood in studies that follow people for years. Some of this has even been tested in clinical trials.</p><p>This part is also where the science is youngest, so I’ve tried to be honest about which findings come from humans, which from mice, and which are only associations.</p><p><em>Same note as before: I’m an engineer summarizing research, not a clinician. Talk to a doctor before changing medications or starting supplements. In the US, call or text 988 if you’re in crisis.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/896/1*0E6M27TrFaKQ6hsUD-Fhng.jpeg" /></figure><h3>1. The gut-brain line runs both ways</h3><p><strong>Stress reaches the gut.</strong> In a Belgian study, healthy volunteers gave a public speech, a classic lab stressor. Afterward their small intestine was measurably more permeable, but only in people whose cortisol rose. Injecting the stress hormone CRH reproduced the effect, and blocking a type of immune cell (mast cells) prevented it (<a href="https://pubmed.ncbi.nlm.nih.gov/24153250/">Vanuytsel et al., 2014</a>). So the knot in your stomach before a big moment isn’t imaginary: the gut lining itself responds.</p><p><strong>The gut reaches the brain.</strong> The main highway is the vagus nerve. In mice, feeding a <em>Lactobacillus</em> strain reduced anxiety- and depression-like behavior and stress hormones, and changed receptors for GABA, the brain’s main calming chemical. When researchers cut the vagus nerve, the effect disappeared (<a href="https://pubmed.ncbi.nlm.nih.gov/21876150/">Bravo et al., 2011</a>).</p><p>But here’s the honest part. When the same team gave that exact strain to healthy men for eight weeks, it did nothing for stress, mood or cognition compared with placebo (<a href="https://doi.org/10.1016/j.bbi.2016.11.018">Kelly et al., 2017</a>). Their paper was titled <em>Lost in Translation?</em> That gap between mice and people runs through this whole field.</p><p><strong>Gut microbes help make serotonin.</strong> Much of the body’s serotonin is made in the gut, where it helps regulate digestion, and in mice, specific gut bacteria drive a large share of that production (<a href="https://pubmed.ncbi.nlm.nih.gov/25860609/">Yano et al., 2015</a>). That’s gut serotonin, not the brain’s supply, so it’s not a shortcut to “fixing” mood chemistry. It does show how closely gut microbes and the body’s signaling chemicals are linked.</p><h3>2. What studies in people show</h3><p>Mouse studies show what <em>can</em> happen. These are the human studies that made me take the gut seriously.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9mznJ4KJ7uWAqP7t_9kutg.png" /></figure><p>Taken together: the link between gut bacteria and mood shows up in large human populations, and in animals the microbes can transfer mood-like behavior. What we don’t know yet is how much of the human link is cause, how much is consequence (mood changes diet, sleep and stress, which change bacteria), and how much both share a third cause.</p><h3>3. Metabolism: blood sugar, weight and mood</h3><p>This was the finding that surprised me most, because it’s so measurable.</p><p><strong>Insulin resistance came first.</strong> Researchers at Stanford used a Dutch study that followed adults with no history of depression for nine years. A higher triglyceride-to-HDL ratio, a routine blood marker of insulin resistance, went with an 89% higher rate of new major depression. People who developed prediabetes during the first two years had 2.66 times the risk (<a href="https://doi.org/10.1176/appi.ajp.2021.20101479">Watson et al., 2021</a>). Because insulin resistance was measured <em>before</em> depression appeared, this is stronger evidence than a one-time snapshot, though still observational.</p><p><strong>Weight and mood feed each other.</strong> Across 15 long-term studies of 58,745 people, obesity raised the later risk of depression by 55%, and depression raised the later risk of obesity by 58% (<a href="https://repub.eur.nl/pub/23061">Luppino et al., 2010</a>).</p><p><strong>The likely bridge is inflammation</strong>, plus a disrupted stress-hormone system (<a href="https://doi.org/10.1176/appi.ajp.2021.20101479">Watson et al., 2021</a>), the same systems that run through Part 1. Inflammatory markers are reliably raised on average in depression and act on the brain’s mood and motivation circuits (<a href="https://pubmed.ncbi.nlm.nih.gov/26711676/">Miller &amp; Raison, 2016</a>).</p><p>The useful part: triglycerides, HDL and fasting glucose are routine blood tests. If you’ve been through a long, stressful stretch, they’re worth asking about at your next checkup.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Utp5pbKVIfO9QBd6VLBScQ.png" /></figure><h3>4. Movement: letting the tension out</h3><p>Part 1 covered what exercise does for the brain. Here I wanted to understand the body side, because for me stress never stayed in my head. It showed up as tightness, and moving was what loosened it.</p><p><strong>Stress tightens muscles even when you’re sitting still.</strong> Mental stress on its own raises activity in the trapezius, the muscle across the top of your shoulders and neck, with no physical load at all. The proposed problem is recovery: if stress never lets those muscles fully rest, the small fibers that switch on first and off last get overworked, and over time that may turn into chronic pain (<a href="https://doi.org/10.1002/ajim.10038">Lundberg, 2002</a>). That last step is a model built mostly on snapshot data, not a proven chain, but it matched what I felt.</p><p><strong>Working the tight muscle helps.</strong> In a 10-week trial of 48 women with chronic neck muscle pain from monotonous jobs, strength training for the painful neck and shoulder muscles cut their worst pain by about 79%. General fitness on a bike helped only a little, and only briefly (<a href="https://doi.org/10.1002/art.23256">Andersen et al., 2008</a>).</p><p><strong>Steps count, whatever the number.</strong> I aim for 8,000 to 10,000 steps a day, but the research says any increase helps. Across 33 studies of 96,173 adults, people walking 7,000 or more steps a day had a 31% lower risk of later depression, and each extra 1,000 steps went with a 9% lower risk. Even 5,000 to 7,499 steps went with fewer depressive symptoms than under 5,000 (<a href="https://doi.org/10.1001/jamanetworkopen.2024.51208">Bizzozero-Peroni et al., 2024</a>).</p><p><strong>Every kind of movement counts.</strong> An umbrella review covering 1,039 trials and 128,119 people found physical activity reduced depression and anxiety by a medium-sized amount. Harder sessions helped depression more, and yoga and other mind-body exercise did best for anxiety (<a href="https://doi.org/10.1136/bjsports-2022-106195">Singh et al., 2023</a>). Many of the underlying reviews were low quality, but the direction was consistent.</p><p><strong>And it reaches the gut.</strong> When 32 previously inactive adults did six weeks of endurance exercise, their gut bacteria shifted, and in lean participants the bacteria made more short-chain fatty acids, the family butyrate belongs to. Most of the changes faded after six weeks without exercise (<a href="https://doi.org/10.1249/MSS.0000000000001495">Allen et al., 2018</a>). Movement works while you keep doing it.</p><h3>5. Food: the best-tested gut lever</h3><p>Of the gut-focused changes, diet has the best human evidence, including an actual randomized trial in people with clinical depression.</p><p><strong>The SMILES trial.</strong> 67 adults with moderate-to-severe depression, most already in treatment, were randomized to 12 weeks of dietitian support for a Mediterranean-style diet or to social support. Remission was 32% versus 8%, and the benefit tracked how much people changed their diet, not weight loss or exercise (<a href="https://doi.org/10.1186/s12916-017-0791-y">Jacka et al., 2017</a>). It’s one small trial, but it was a real one.</p><p><strong>Diet patterns over time.</strong> In a meta-analysis of observational studies, people who ate most like a Mediterranean diet had about a third lower risk of developing depression, and a less inflammatory diet went with lower risk too (relative risk 0.76) (<a href="https://doi.org/10.1038/s41380-018-0237-8">Lassale et al., 2019</a>).</p><p><strong>Doing this as a vegetarian.</strong> I’m a pure vegetarian: no meat, fish or eggs. The SMILES diet included fish and some lean meat, so I wanted to know whether this evidence applies to someone who eats like me.</p><p>The honest answer: the <em>quality</em> of a plant-based diet matters more than the label. A 2021 meta-analysis of 13 studies (49,889 people) found vegetarians had higher depression scores on average, though the studies varied widely and the authors say no firm conclusions can be drawn yet (<a href="https://doi.org/10.1016/j.jad.2021.07.098">Ocklenburg &amp; Borawski, 2021</a>). In 180,532 UK Biobank participants, a <em>healthy</em> plant-based diet (whole grains, fruit, vegetables, nuts, legumes) went with 8% lower depression risk, while an <em>unhealthy</em> one (refined grains, sugary drinks, sweets) went with 15% higher risk (<a href="https://doi.org/10.1093/ageing/afad070">Wu et al., 2023</a>).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nn8hPE4ZjayzR2z2T8bqSg.jpeg" /></figure><p>I’m not here to argue that a vegetarian diet is better or worse. What I can share is what I’ve seen work for me as a vegetarian: balancing three things on my plate. Here’s that balance, with the research I found to back it:</p><ul><li><strong>Fiber:</strong> dal, rajma, chana, vegetables, fruit and whole grains. A large review in <em>The Lancet</em> found the biggest health benefits at 25 to 29 g of fiber a day, while most people eat under 20 g (<a href="https://doi.org/10.1016/S0140-6736%2818%2931809-9">Reynolds et al., 2019</a>). Fiber is also what gut bacteria ferment into butyrate, the compound linked to better quality of life in section 2, and higher fiber intake went with lower odds of depression across observational studies (<a href="https://doi.org/10.1016/j.ctim.2020.102621">Fatahi et al., 2021</a>).</li><li><strong>Protein:</strong> dal and other legumes, paneer, curd, tofu and soy, nuts and seeds, spread across meals rather than saved for one.</li><li><strong>Carbs:</strong> whole over refined. Whole wheat, millets, brown rice or oats in place of white rice, maida and sugary snacks where you can. The same <em>Lancet</em> review found whole grains carried benefits much like fiber’s (<a href="https://doi.org/10.1016/S0140-6736%2818%2931809-9">Reynolds et al., 2019</a>).</li></ul><p><strong>One thing to check: vitamin B12.</strong> It comes almost entirely from animal foods. Across 18 studies, B12 deficiency was common in vegetarians of every age and diet type, and more common in people vegetarian since birth. The authors recommend regular B12 supplements (<a href="https://doi.org/10.1111/nure.12001">Pawlak et al., 2013</a>). It’s a simple blood test to ask for.</p><p><strong>Ultra-processed food.</strong> In 31,712 women followed from 2003 to 2017, those eating the most ultra-processed food had 49% higher risk of depression. The link was driven by artificial sweeteners and artificially sweetened drinks. Women who cut 3 or more servings a day had 16% lower risk (<a href="https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2809727">Samuthpongtorn et al., 2023</a>).</p><p><strong>Fermented foods.</strong> In a Stanford trial, adults who raised fermented foods (yogurt, kefir, kimchi, fermented vegetables, kombucha) to about six servings a day for 10 weeks had more diverse gut bacteria and lower levels of 19 inflammatory proteins, including IL-6. A high-fiber diet over the same period didn’t lower them (<a href="https://doi.org/10.1016/j.cell.2021.06.019">Wastyk et al., 2021</a>). The trial measured immune markers, not mood.</p><p><strong>Probiotics.</strong> A review of randomized trials found probiotics helped as an add-on to antidepressants but not clearly on their own (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC7915600">Nikolova et al., 2021</a>). An 8-week pilot trial in people only partly helped by antidepressants found less depression and anxiety than placebo, and the authors call for larger trials (<a href="https://doi.org/10.1001/jamapsychiatry.2023.1817">Nikolova et al., 2023</a>).</p><p><strong>A cup of tea and a breather.</strong> One more thing I’ve noticed in myself: every so often, a cup of lavender or ashwagandha tea helps calm my nerves. I didn’t find studies linking these teas to the gut, so take this as my experience rather than science. But it’s an easy thing to try when you need to slow down and catch a breather. Sitting with something warm while you slow your breathing is also close to the exhale-focused breathing in Part 1, which does have trial evidence. One caution: concentrated ashwagandha supplements have been linked to rare cases of liver injury (<a href="https://doi.org/10.1111/liv.14393">Björnsson et al., 2020</a>), so check with a doctor before taking it regularly in that form.</p><h3>6. What I’d take from all this</h3><p>First, the limits, because this field gets oversold. Most of the mechanism evidence comes from mice. Most human microbiome studies are snapshots that can’t separate cause from effect. And the one probiotic that looked great in mice did nothing in healthy men. Anyone selling a “gut reset” for anxiety is ahead of the science.</p><p>That said, the low-risk changes line up with the best evidence:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Foxa3eQMK9mn0_RXuU2s4A.png" /></figure><h3>The thread through both parts</h3><p>When I started reading, I thought of anxiety as something happening in my head. What the research showed me is that a long stretch of uncertainty, constant failures and things not working out can turn into anxiety, and that anxiety doesn’t stay in your head. It reaches the body at every level: stress hormones that won’t switch off, a reward system that goes quiet, muscles that never fully rest, a gut that shifts with every stressful week.</p><p>And much of that isn’t your body failing you. It’s your body trying to protect you. The numbness that turns the volume down, the stress chemistry that rises and then adjusts as the strain drags on, the worry that feels like it’s keeping you safe: these are defenses your body deploys in response to the spiral, mostly without you ever noticing. Some of them outlast their usefulness, and that’s where the real work starts.</p><p>These two articles are my way of compiling what I read and what I lived through (still living through to some extent). They’re also a reminder, to me as much as anyone, that your body is keeping track of everything you go through. Tough periods are hard to navigate, but that gives us a responsibility too: to work hand in hand with our bodies, find the small, consistent things that actually help (movement, sleep, food, people, and talking to someone qualified), and understand what’s happening to us in the moment.</p><p>Everyone experiences hard times differently. For me, knowing the science behind it has always been the most reliable way through periods of anxiety. It let me look at what was happening analytically instead of being overwhelmed by it. If these articles do even a little of that for you, they’ve done their job.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*x0ePSWcjhqRW_yV7Dft3kg.jpeg" /></figure><p><strong>I went looking for answers in my head and found half of them in my body. The gut doesn’t replace anything in Part 1. It’s the other end of the same wire.</strong></p><p><em>If you’re struggling right now, please talk to a doctor or therapist. In the US, call or text 988 any time.</em></p><h3>References</h3><p><strong>The gut-brain connection</strong></p><ul><li>Vanuytsel et al. (2014). Psychological stress and CRH increase intestinal permeability in humans. <em>Gut</em>. <a href="https://pubmed.ncbi.nlm.nih.gov/24153250/">PubMed</a></li><li>Bravo et al. (2011). Ingestion of Lactobacillus strain regulates emotional behavior via the vagus nerve. <em>PNAS</em>. <a href="https://pubmed.ncbi.nlm.nih.gov/21876150/">PubMed</a></li><li>Kelly et al. (2017). Lost in translation? <em>Lactobacillus rhamnosus</em> (JB-1) fails to modulate stress in healthy men. <em>Brain, Behavior, and Immunity</em>. <a href="https://doi.org/10.1016/j.bbi.2016.11.018">DOI</a></li><li>Yano et al. (2015). Indigenous bacteria from the gut microbiota regulate host serotonin biosynthesis. <em>Cell</em>. <a href="https://pubmed.ncbi.nlm.nih.gov/25860609/">PubMed</a></li></ul><p><strong>Human studies</strong></p><ul><li>Valles-Colomer et al. (2019). The neuroactive potential of the human gut microbiota in quality of life and depression. <em>Nature Microbiology</em>. <a href="https://www.nature.com/articles/s41564-018-0337-x.pdf">Link</a></li><li>Radjabzadeh et al. (2022). Gut microbiome-wide association study of depressive symptoms. <em>Nature Communications</em>. <a href="https://pubmed.ncbi.nlm.nih.gov/36473852/">PubMed</a></li><li>Kelly et al. (2016). Transferring the blues: depression-associated gut microbiota induces neurobehavioural changes in the rat. <em>Journal of Psychiatric Research</em>. <a href="https://pubmed.ncbi.nlm.nih.gov/27491067/">PubMed</a></li><li>Zamani, Alizadeh-Tabari &amp; Zamani (2019). Anxiety and depression in irritable bowel syndrome. <em>Alimentary Pharmacology &amp; Therapeutics</em>. <a href="https://doi.org/10.1111/apt.15325">DOI</a></li></ul><p><strong>Metabolism</strong></p><ul><li>Watson et al. (2021). Incident major depressive disorder predicted by three measures of insulin resistance. <em>American Journal of Psychiatry</em>. <a href="https://doi.org/10.1176/appi.ajp.2021.20101479">DOI</a></li><li>Luppino et al. (2010). Overweight, obesity and depression: meta-analysis of longitudinal studies. <em>Archives of General Psychiatry</em>. <a href="https://repub.eur.nl/pub/23061">Link</a></li><li>Miller &amp; Raison (2016). The role of inflammation in depression. <em>Nature Reviews Immunology</em>. <a href="https://pubmed.ncbi.nlm.nih.gov/26711676/">PubMed</a></li></ul><p><strong>Food</strong></p><ul><li>Jacka et al. (2017). A randomised controlled trial of dietary improvement for adults with major depression (SMILES). <em>BMC Medicine</em>. <a href="https://doi.org/10.1186/s12916-017-0791-y">DOI</a></li><li>Lassale et al. (2019). Healthy dietary indices and risk of depressive outcomes. <em>Molecular Psychiatry</em>. <a href="https://doi.org/10.1038/s41380-018-0237-8">DOI</a></li><li>Samuthpongtorn et al. (2023). Consumption of ultraprocessed food and risk of depression. <em>JAMA Network Open</em>. <a href="https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2809727">Link</a></li><li>Wastyk et al. (2021). Gut-microbiota-targeted diets modulate human immune status. <em>Cell</em>. <a href="https://doi.org/10.1016/j.cell.2021.06.019">DOI</a></li><li>Nikolova et al. (2021). Updated review and meta-analysis of probiotics for clinical depression. <em>Journal of Clinical Medicine</em>. <a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC7915600">PMC7915600</a></li><li>Nikolova et al. (2023). Probiotics as adjunctive treatment in patients with depression: a randomized clinical trial. <em>JAMA Psychiatry</em>. <a href="https://doi.org/10.1001/jamapsychiatry.2023.1817">DOI</a></li></ul><p><strong>Vegetarian eating</strong></p><ul><li>Ocklenburg &amp; Borawski (2021). Vegetarian diet and depression scores: a meta-analysis. <em>Journal of Affective Disorders</em>. <a href="https://doi.org/10.1016/j.jad.2021.07.098">DOI</a></li><li>Wu et al. (2023). Quality of plant-based diet and the risk of dementia and depression. <em>Age and Ageing</em>. <a href="https://doi.org/10.1093/ageing/afad070">DOI</a></li><li>Melina, Craig &amp; Levin (2016). Position of the Academy of Nutrition and Dietetics: vegetarian diets. <em>Journal of the Academy of Nutrition and Dietetics</em>. <a href="https://doi.org/10.1016/j.jand.2016.09.025">DOI</a></li><li>Reynolds et al. (2019). Carbohydrate quality and human health. <em>The Lancet</em>. <a href="https://doi.org/10.1016/S0140-6736%2818%2931809-9">DOI</a></li><li>Fatahi et al. (2021). Dietary fiber and depression symptoms: a meta-analysis of observational studies. <em>Complementary Therapies in Medicine</em>. <a href="https://doi.org/10.1016/j.ctim.2020.102621">DOI</a></li><li>Pawlak et al. (2013). How prevalent is vitamin B12 deficiency among vegetarians? <em>Nutrition Reviews</em>. <a href="https://doi.org/10.1111/nure.12001">DOI</a></li></ul><p><strong>Movement and muscle tension</strong></p><ul><li>Lundberg (2002). Psychophysiology of work: stress, gender, endocrine response, and work-related upper extremity disorders. <em>American Journal of Industrial Medicine</em>. <a href="https://doi.org/10.1002/ajim.10038">DOI</a></li><li>Andersen et al. (2008). Effect of two contrasting types of physical exercise on chronic neck muscle pain. <em>Arthritis &amp; Rheumatism</em>. <a href="https://doi.org/10.1002/art.23256">DOI</a></li><li>Bizzozero-Peroni et al. (2024). Daily step count and depression in adults. <em>JAMA Network Open</em>. <a href="https://doi.org/10.1001/jamanetworkopen.2024.51208">DOI</a></li><li>Singh et al. (2023). Physical activity for depression, anxiety and psychological distress: an umbrella review. <em>British Journal of Sports Medicine</em>. <a href="https://doi.org/10.1136/bjsports-2022-106195">DOI</a></li><li>Allen et al. (2018). Exercise alters gut microbiota composition and function in lean and obese humans. <em>Medicine &amp; Science in Sports &amp; Exercise</em>. <a href="https://doi.org/10.1249/MSS.0000000000001495">DOI</a></li></ul><p><strong>Safety note</strong></p><ul><li>Björnsson et al. (2020). Liver injury due to ashwagandha: a case series. <em>Liver International</em>. <a href="https://doi.org/10.1111/liv.14393">DOI</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e753c1f90360" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Firmware Upgrade That Didn’t Need a New Bench]]></title>
            <link>https://medium.com/@dagatejal/the-firmware-upgrade-that-didnt-need-a-new-bench-ee9c6ee21509?source=rss-79076ec3d006------2</link>
            <guid isPermaLink="false">https://medium.com/p/ee9c6ee21509</guid>
            <category><![CDATA[problem-solving]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[program-management]]></category>
            <category><![CDATA[project-management]]></category>
            <category><![CDATA[technical-program-manager]]></category>
            <dc:creator><![CDATA[Tejal Daga]]></dc:creator>
            <pubDate>Tue, 10 Nov 2026 21:24:48 GMT</pubDate>
            <atom:updated>2026-11-10T21:41:18.966Z</atom:updated>
            <content:encoded><![CDATA[<p><em>A requirement arrived with a solution attached. Here’s what happened when I asked what problem it was solving.</em></p><p>Most requirements describe a problem. Some arrive with the solution already attached, and it’s easy to start building before anyone asks whether it’s the right one.</p><p>This is a short story about one of those. A customer needed the firmware, the software that runs inside a device, updated on 4,000 new devices before launch, and they had already decided how. We delivered it a different way: no new equipment, about two weeks ahead of schedule, and an estimated $20,000–22,000 that never had to be spent.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/990/1*6A3EopkbgBf76wBCDcRdjg.png" /></figure><h3>Background</h3><p>My team handles returned devices for cable and fiber providers: screening, cleaning, functional testing, refurbishment and scrap.</p><p>One customer had 4,000 new devices packaged for a mid-March launch, all on firmware version 1.0, which had a known defect. Every unit needed version 2.0, and the manufacturer had no capacity to update them in time. The devices reached us in mid-January, leaving about eight weeks.</p><h3>The requirement</h3><p>The customer’s engineering team asked us to upgrade each device over Ethernet, using a dedicated bench: a workstation where an operator connects and updates one device at a time. Their engineers rarely work with our team directly, so they had limited visibility into what our site could already do, and Ethernet became the default assumption.</p><p>No one was wrong. The requirement simply arrived with a solution attached.</p><h3>First pass: validating the requirement</h3><p>I started where an engineer would. I reviewed the spec, tested the Ethernet upgrade steps and confirmed they worked.</p><p>But working is not the same as being the right fit. Each manual upgrade took 7–8 minutes with an operator attending, so one bench would have needed 12–13 weeks. Meeting the deadline would have meant building two new benches and staffing them across both shifts, and even then, finishing on time was not certain.</p><h3>Reframing the problem</h3><p>With the full scope in view, I raised three questions with my manager:</p><p>· <strong>Is this one-time or recurring?</strong> It was a one-time batch, so new benches would have had no work afterward.</p><p>· <strong>What is the right level of investment?</strong> New equipment makes sense for an ongoing program, far less for a single batch.</p><p>· <strong>What could we reuse?</strong></p><p>The third question changed the plan. These devices also have a fiber port, and we already update firmware over fiber for the same customer’s other products. We do it on our existing testers, automated systems that handle 10 devices at once.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/990/1*UBOZlFun2QeydzTY06q6SA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*imE8xZOkSuFbKmKIqnTv7Q.png" /></figure><h3>Validating the alternative and aligning stakeholders</h3><p>I asked the customer’s engineering team whether the firmware could be upgraded over fiber and, if so, to send the file in the format our process needs. Once they confirmed, my manager and I ran a proof of concept on 10 units.</p><p>With results in hand, we met with the program manager, the general manager and the production site’s engineer. The proposal: reuse the existing tester, with no new cabling and no new bench.</p><h3>Execution and rollout</h3><p>I built and automated the upgrade software, and it was production-ready within two weeks. We started with a pilot batch, then ran 220–250 units a day on a single shift. Results were logged in the same format as every other unit, so existing reports needed no changes, and Operations fit the work around regular production without affecting it.</p><p>Quality was built into the process. After each update, the software waited for the device to restart and confirmed it was running version 2.0. Any unit that didn’t pass after three attempts was set aside rather than shipped.</p><h3>Outcome</h3><p>The final unit was completed in the last week of February, about two weeks before launch. About 3,800 units, 95%, were verified on version 2.0. The remaining 5% never passed the check, so they were returned to the customer instead of going out with uncertain firmware. Every unit that shipped had been confirmed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*C1r2FHsqySvURIQO8WPzrg.png" /></figure><p>Daily output would have been similar either way. The difference was what it took to get there.</p><h3>What I’d do differently</h3><p>I validated the Ethernet path before I questioned it. That was useful due diligence, but it meant the scoping questions came second. Next time I’d ask about volume, timeline and existing options first, and test second. The better answer was available from day one; I just hadn’t looked for it yet.</p><h3>My role</h3><p>I led the technical work end to end: validating the original requirement, identifying the fiber alternative, confirming it with the customer and building the automation. My manager and I ran the proof of concept and presented the recommendation together. The program manager, general manager and site engineer approved the approach, and Operations scheduled the work alongside regular production.</p><h3>Key takeaways</h3><p>· <strong>Separate the problem from the proposed solution.</strong> The goal was verified firmware before launch, not an Ethernet bench.</p><p>· <strong>Scope before you build, and reuse what exists.</strong> One-time work rarely justifies new equipment.</p><p>· <strong>Validate, then align.</strong> A 10-unit proof of concept made the decision quick and evidence-based.</p><p>· <strong>Make capabilities visible upstream.</strong> The Ethernet plan came from limited visibility into what our site could do. Sharing that with customer teams early lets the next request start from the better option.</p><p>As an engineer, my instinct is still to dig into the spec and make it work. But some of the most valuable work happens one step earlier, in deciding what is worth building at all.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vNXaWvphag9lfVvZRbmAPg.png" /></figure><p>One Ethernet bench alone: 4,000 units at 7–8 minutes is about 470–530 hours, or 12–13 weeks on a single 8-hour shift. With two benches ready in early to mid-February, the last unit would have landed in late February at the earliest. On the fiber path, a second shift could have roughly doubled output to about 440–500 units a day; the Ethernet plan had no spare capacity without a third bench or overtime.</p><p><em>Ethernet figures assume an 8-hour shift with two 15-minute breaks. Fiber operator hours count the full shift, so they are an upper bound.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ee9c6ee21509" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Studied Depression in 2018. In 2026, the Research Got Personal.]]></title>
            <link>https://medium.com/@dagatejal/i-studied-depression-in-2018-in-2026-the-research-got-personal-5a497af2f890?source=rss-79076ec3d006------2</link>
            <guid isPermaLink="false">https://medium.com/p/5a497af2f890</guid>
            <category><![CDATA[neuroscience]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[mental-health]]></category>
            <category><![CDATA[anxiety]]></category>
            <category><![CDATA[self-improvement]]></category>
            <dc:creator><![CDATA[Tejal Daga]]></dc:creator>
            <pubDate>Fri, 06 Nov 2026 19:22:43 GMT</pubDate>
            <atom:updated>2026-11-06T19:22:43.930Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Part 1: What research taught me about anxiety, failure and brain chemistry.</em></p><p>In 2018, as an engineering student, I co-wrote two papers on depression. One used sentiment analysis to spot signs of it in text. The other was an app that estimated how severe someone’s depression was. I read the clinical literature, learned the screening questionnaires, and built models around them. To me, depression was a classification problem, but in 2026 the research got personal.</p><p>It wasn’t one bad event. It was a long stretch where things kept not working out, one after another. Somewhere in there came anxiety that wouldn’t switch off, panic attacks, and sleep I couldn’t fix no matter what I tried. I knew the vocabulary. I’d written about the symptoms. None of that helped at 3 a.m. To be clear, I’ve never been diagnosed with depression, and I’m not putting a label on what I went through. But the research on stress, anxiety and long runs of failure described a lot of it.</p><p>I went back to the papers I’d first read in 2017 and 2018 because I was tired. Tired of the mental fatigue, and of feeling like I was going in loops. Part of me doubted that scientific methods could really help with something this personal.</p><p>But the more I read, the more shocked I was. I had no idea that brain chemistry, gut health, gut microbes, even the physical pathways in the brain could be thrown off this badly by experiences like these.</p><p>It was also a wake-up call to start putting in the work. None of this is overnight; every solution takes time. But knowing what your mind and body are going through changes what an episode of anxiety or panic feels like. Looking at it from a logical point of view helps in the moments when you’re overwhelmed, don’t know what to do, and can’t put what you’re feeling into words clearly (something I’ve always struggled with).</p><p>That’s why I did this research, and why I’m writing this.</p><p>This is what I found. It’s in two parts:</p><p>• <strong>Part 1 (this one):</strong> what long stretches of failure, anxiety and overthinking do to the brain, drawing on research that spans everyday stress, anxiety and clinical depression, and what the evidence says helps.</p><p>• <strong>Part 2:</strong> the gut, metabolism and mood, which turned out to matter far more than I expected.</p><p>Every claim links to a published study or review. Where the evidence is thin or comes mainly from animals, I say so.</p><p><em>A note before we start: I’m an engineer, not a clinician. This is what I learned from reading the research, not medical advice. If you’re struggling right now, skip to “When to get help” at the end. In the US you can call or text 988 any time.</em></p><h3>1. Failure teaches the brain to stop trying</h3><p>The first thing that surprised me: giving up isn’t something the brain <em>learns</em> after enough failure. It’s the default.</p><p>The researchers who coined “learned helplessness” in the 1960s spent the next fifty years finding out they had it backwards. Passivity under prolonged bad events that you can’t control is the brain’s automatic response, driven by stress circuits deep in the brainstem. What has to be <em>learned</em> is that you have control. A region of the prefrontal cortex detects “my action changed the outcome” and switches the passive response off (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC4920136">Maier &amp; Seligman, 2016</a>). They call it the hope circuit.</p><p>Then there’s the story you tell yourself. Hopelessness theory says setbacks turn into depression when you explain them as <em>stable</em> (“this will always happen”) and <em>global</em> (“this ruins everything”) (<a href="https://www.binghamton.edu/psychology/labs/mood/pdfs/2002_handbook-of-dep_abramson-et-al_cog-vuln-stress-models-of-dep.pdf">Abramson et al.</a>). After enough rejections in a row, those explanations stop feeling like interpretations and start feeling like facts.</p><p>And then you pull back. In a study that tracked students’ phone GPS for two weeks, days with worse mood meant more time at home, and socially anxious students who had a bad day were more likely to stay in the next day too (<a href="https://www.jmir.org/2017/3/e62">Chow et al., 2017</a>). Staying in means fewer wins and fewer chances for the hope circuit to register control. The loop tightens:</p><p>A setback → “it’s always going to be like this” → rumination and stress → pulling back → fewer wins → the next setback lands harder.</p><h3>2. Overthinking feels like work. It’s actually a habit being rewarded.</h3><p>As an engineer, my instinct when something breaks is to analyze it. When things kept going wrong, that instinct turned into overthinking. It felt productive. The research says it mostly isn’t.</p><p>Psychologists call it <em>repetitive negative thinking</em>. Rumination looks back and asks “why”; worry looks ahead and asks “what if” (<a href="https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2019.01372/pdf">Kornacka et al., 2019</a>). A major review found rumination deepens negative thinking, impairs problem-solving and reduces the action needed to fix things. It also predicts <em>starting</em> a depressive episode better than it predicts how long one lasts (<a href="https://escholarship.org/uc/item/8544f7w2">Nolen-Hoeksema et al., 2008</a>). Style matters as much as content: abstract “why does this keep happening to me?” thinking makes mood worse, while concrete “what’s the next step?” thinking doesn’t (<a href="https://doi.org/10.1037/0033-2909.134.2.163">Watkins, 2008</a>).</p><p>So why can’t you just stop? Because overthinking is <strong>negatively reinforced</strong>. That doesn’t mean punished. It means it removes something unpleasant, and the brain repeats whatever brings relief.</p><p>• <strong>Worry numbs the body.</strong> Worry runs in words, which dampens vivid images and the physical rush of fear. That brief relief rewards the worrying (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC3444754">Borkovec’s avoidance theory, reviewed here</a>).</p><p>• <strong>The superstition effect.</strong> Most feared outcomes never happen, and the brain quietly credits the worry (<a href="https://doi.org/10.1093/clipsy.9.1.76">Borkovec, 2002</a>). In a study of people with anxiety who tracked their worries, 91% never came true (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC7233480/">LaFreniere &amp; Newman, 2020</a>).</p><p>• <strong>Fighting the thought backfires.</strong> Across 28 studies, suppressed thoughts came back more often afterward (<a href="https://doi.org/10.1016/S0272-7358%2800%2900057-X">Abramowitz et al., 2001</a>).</p><p><em>What breaks the loop, according to trials:</em></p><p><strong>❤️‍🩹 A worry outcome journal.</strong> Write each worry down, then check later what actually happened. People whose worries mostly didn’t come true improved the most (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC7233480/">LaFreniere &amp; Newman, 2020</a>).</p><p><strong>❤️‍🩹</strong> <strong>A fixed worry period.</strong> Postpone worries to a 30-minute slot at the same time and place each day. Two weeks of this cut worry, anxiety and insomnia more than a control practice (<a href="https://doi.org/10.1177/0145445512455661">McGowan &amp; Behar, 2013</a>).</p><p><strong>❤️‍🩹</strong> <strong>Swap “why” for “how.”</strong> This is the core of rumination-focused CBT, which helped residual depression in a randomized trial (<a href="https://doi.org/10.1192/bjp.bp.110.090282">Watkins et al., 2011</a>).</p><p><strong>❤️‍🩹</strong> <strong>Talk to yourself in the third person.</strong> Asking “why is [your name] upset?” reduced emotional brain activity within a second, with no extra effort (<a href="https://doi.org/10.1038/s41598-017-04047-3">Moser, Kross et al., 2017</a>). It helped most in people with more depressive symptoms (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC2747112">Kross et al.</a>).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*J3PMlYckV6Y00BPcy3rOaQ.png" /></figure><h3>3. What months of stress do to brain chemistry</h3><p>Short bursts of stress sharpen you. The problem is when the alarm never turns off. This is the part I <strong>most </strong>wanted to understand, because it explained the sleep and the panic.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*S8C9MqdwpL-IYDcikDJsKQ.png" /></figure><p><strong>Worry keeps the stress response running</strong> even when nothing is happening. A meta-analysis found worry and rumination raise blood pressure, heart rate and cortisol, before a stressor arrives and after it’s gone (<a href="https://doi.org/10.1037/bul0000036">Ottaviani et al., 2016</a>). That’s how prolonged anxiety and repeated failure feed each other.</p><p><strong>Sleep is part of the machinery, not a side effect.</strong> Disrupted sleep shows up in about 80% of major depression (<a href="https://www.cs.virginia.edu/~stankovic/psfiles/robempath.pdf">Dickerson et al., 2011</a>) and feeds inflammation (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC3392963">Raison et al., 2006</a>). Across 65 randomized trials, improving sleep reduced depression (g = −0.63), anxiety (−0.51) and rumination (−0.49), and bigger sleep gains brought bigger mood gains (<a href="https://eprints.whiterose.ac.uk/id/eprint/178923/">Scott et al., 2021</a>).</p><p><strong>What about “low serotonin”?</strong> I grew up hearing depression is a chemical imbalance. A 2022 umbrella review found no consistent evidence that low serotonin causes it (<a href="https://www.nature.com/articles/s41380-022-01661-0">Moncrieff et al., 2022</a>). Critics countered that the serotonin system is still involved (<a href="https://doi.org/10.1038/s41380-023-02095-y">Jauhar et al., 2023</a>), and the original authors replied that the specific “low serotonin” claim remains unsupported (<a href="https://doi.org/10.1038/s41380-023-02094-z">Moncrieff et al., 2023</a>). The practical takeaway for me: there’s no single chemical you’re “low” on. The field has moved toward circuits and how adaptable the brain is.</p><h3>4. When nothing feels good, or bad</h3><p>There’s another pattern I wanted to understand: a flatness where good news barely lands and bad news doesn’t sting the way it used to. It can feel almost protective, as if the brain has turned the volume down so it won’t be overwhelmed.</p><p>That turns out to be a documented pattern. A meta-analysis of 19 lab studies found people with major depression reacted less to <em>both</em> pleasant and unpleasant material, with a bigger drop for the pleasant kind (<a href="https://pubmed.ncbi.nlm.nih.gov/18006196/">Bylsma et al., 2008</a>). Researchers call it <em>emotion context insensitivity</em>. In a later study, loss of pleasure, more than low mood, predicted blunted reactions in both directions (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC7252228">anhedonia and reactivity study</a>). The same flattening shows up in PTSD (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC4153581">Felmingham et al., 2014</a>).</p><p>Is it a defense mechanism? Partly, the evidence suggests. Researchers describe it as disengaging from the environment (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC8097691">Bylsma, 2021</a>). Stress hormones follow a similar shape: cortisol is high when a stressor starts and falls as it drags on (<a href="https://pubmed.ncbi.nlm.nih.gov/17201569/">Miller et al., 2007</a>). Whether this actually protects you is less clear, since blunted stress responses are also linked to worse health (<a href="https://research.birmingham.ac.uk/en/publications/the-other-side-of-the-coin-blunted-cardiovascular-and-cortisol-re/">Phillips et al., 2013</a>).</p><p>Several systems seem to turn the volume down:</p><p>• <strong>The reward circuit responds less</strong>, so wins feel flat (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC4153581">Felmingham et al., 2014</a>; <a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC3701611">Huys et al., 2013</a>).</p><p>• <strong>A stress chemical called dynorphin</strong> drives dysphoria and loss of pleasure, though this comes mostly from animal research (<a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC2819621/">Bruchas et al., 2010</a>).</p><p>• <strong>Hiding feelings</strong> dampens positive emotion, doesn’t reduce negative emotion, and raises body arousal (<a href="https://www.psychiatry.wisc.edu/courses/Nitschke/seminar/gross_2002_psychophys.pdf">Gross, 2002</a>). That explained something for me: you can feel numb and wired at the same time.</p><p>• <strong>Some antidepressants</strong> cause emotional blunting in an estimated 40–60% of users, likely by dampening sensitivity to reward (<a href="https://doi.org/10.1038/s41386-022-01523-x">Langley et al., 2023</a>). If that’s you, it’s worth raising with your prescriber.</p><p>The encouraging part: a therapy built to rebuild positive emotion, Positive Affect Treatment, beat standard threat-focused CBT on positive emotion, depression and anxiety (<a href="https://doi.org/10.1037/ccp0000396">Craske et al., 2019</a>), and has been replicated twice (<a href="https://pmc.ncbi.nlm.nih.gov/articles/10213148">Craske et al., 2023</a>; <a href="https://doi.org/10.1001/jamanetworkopen.2026.7403">Meuret et al., 2026</a>). Its core moves are things you can practice daily; I list them below.</p><h3>5. The brain physically changes, and changes back</h3><p>This was the scariest thing I read, and then the most hopeful.😥</p><p>In animal studies, weeks of chronic stress shrink neurons in the hippocampus (memory, and a brake on stress hormones) and prefrontal cortex (flexible thinking), while neurons in the amygdala (threat detection) grow (<a href="https://www.uned.es/universidad/inicio/dam/jcr:60327b17-ddd7-4902-bf9c-2d6421ee6a70/McEwen%20(2008.EJPhar).Stress%20hormones%20in%20health%20and%20disease.pdf">McEwen, 2008</a>). Weaker brakes plus a jumpier alarm. In humans, the largest brain-imaging analysis of depression (1,728 patients, 7,199 controls) found a slightly smaller hippocampus, but only in people with <em>recurrent</em> episodes, not a first one (<a href="https://repub.eur.nl/pub/86136">Schmaal et al., 2016</a>). A 2024 study also found a network that flags rewards and threats is about twice as large in most people with depression, even before it starts (<a href="https://doi.org/10.1038/s41586-024-07805-2">Lynch et al., 2024</a>). Some people may simply be wired to feel setbacks more sharply.</p><p>Now the hopeful part. McEwen describes the stress changes as largely reversible over weeks (<a href="https://www.uned.es/universidad/inicio/dam/jcr:60327b17-ddd7-4902-bf9c-2d6421ee6a70/McEwen%20(2008.EJPhar).Stress%20hormones%20in%20health%20and%20disease.pdf">McEwen, 2008</a>). And the brain grows back with the right inputs:</p><p>• A year of moderate aerobic exercise grew the hippocampus by about 2% in older adults, while a stretching group lost about 1.4% (<a href="https://pmc.ncbi.nlm.nih.gov/articles/3041121/">Erickson et al., 2011</a>).</p><p>• Twenty sessions of cognitive behavioral therapy increased gray matter in the amygdala and hippocampus, and 19 of 30 patients ended with few symptoms (<a href="https://doi.org/10.1038/s41398-025-03545-7">Zwiky et al., 2025</a>).</p><p>• Fast-acting treatments like ketamine appear to work by restoring the same synapses stress wears down (<a href="https://pubmed.ncbi.nlm.nih.gov/38220554/">Trends in Neurosciences, 2024</a>).</p><p>The first-episode finding is the practical one: acting early matters.</p><h3>6. What the evidence says actually helps</h3><p>This is the part I wish I’d had in one place. None of these is a cure, but each has trials behind it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*E93Zb9Qs8uB-VcAAq0ZH0g.png" /></figure><p>One more that sounds obvious but isn’t: <strong>therapy works, and it changes the brain</strong> (section 5). Only about 4 in 10 people with a diagnosable condition ever seek help (<a href="https://www.wseas.us/e-library/conferences/2006tenerife/papers/541-399.pdf">Culjak, 2006</a>).</p><p>For long stretches you can’t control, the research nudged me toward <strong>process goals over outcome goals</strong>. You can’t control whether you get the offer. You can control the hours practiced, the applications sent, the walks taken. Tracking those gives the hope circuit something to register.</p><h3>When to get help</h3><p>Reading research helped me understand what was going on. It’s not a substitute for a professional, and the same research is clear about when to reach out.</p><p>• <strong>Two free, validated self-checks.</strong> The PHQ-9 screens for depression and the GAD-7 for anxiety. A PHQ-9 score of 10 or more correctly flagged major depression 88% of the time (<a href="https://doi.org/10.1046/j.1525-1497.2001.016009606.x">Kroenke et al., 2001</a>); GAD-7 scores of 5, 10 and 15 mark mild, moderate and severe anxiety (<a href="https://doi.org/10.1001/archinte.166.10.1092">Spitzer et al., 2006</a>). They’re screening tools, not diagnoses.</p><p>• <strong>The two-week rule.</strong> Low mood or loss of interest most days for two weeks or more is the point to talk to a doctor or therapist; it’s the window the PHQ-9 itself asks about (<a href="https://doi.org/10.1046/j.1525-1497.2001.016009606.x">Kroenke et al., 2001</a>).</p><p><strong>Right away</strong> if you have any thoughts of harming yourself. In the US, call or text <strong>988</strong> any time, or chat at <a href="https://988lifeline.org/">988lifeline.org</a></p><p><em>Everything in this piece, from the numbness to the overthinking loops to the stress chemistry that won’t switch off, started as your brain trying to protect you. It bends, and sometimes breaks, to keep you going. Acknowledge that. Then use it. The same brain that learned to stop trying can learn to try again, and the first step is simply seeing the blocks for what they are. We got this!💛</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PV5boL-T-fKm9_cD_huTlw.png" /></figure><h3>References</h3><p><strong>Failure, helplessness and control</strong></p><p>• Maier &amp; Seligman (2016). Learned helplessness at fifty. <em>Psychological Review</em>. <a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC4920136">PMC4920136</a></p><p>• Baratta, Seligman &amp; Maier (2023). From helplessness to controllability. <em>Frontiers in Psychiatry</em>. <a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC10205144">PMC10205144</a></p><p>• Abramson et al. Cognitive vulnerability-stress models of depression. <a href="https://www.binghamton.edu/psychology/labs/mood/pdfs/2002_handbook-of-dep_abramson-et-al_cog-vuln-stress-models-of-dep.pdf">PDF</a></p><p>• Chow et al. (2017). Using mobile sensing to test clinical models of depression and social anxiety. <em>JMIR</em>. <a href="https://www.jmir.org/2017/3/e62">Link</a></p><p><strong>Overthinking</strong></p><p>• Nolen-Hoeksema, Wisco &amp; Lyubomirsky (2008). Rethinking rumination. <em>Perspectives on Psychological Science</em>. <a href="https://escholarship.org/uc/item/8544f7w2">Link</a></p><p>• Watkins (2008). Constructive and unconstructive repetitive thought. <em>Psychological Bulletin</em>. <a href="https://doi.org/10.1037/0033-2909.134.2.163">DOI</a></p><p>• Watkins et al. (2011). Rumination-focused CBT for residual depression. <em>British Journal of Psychiatry</em>. <a href="https://doi.org/10.1192/bjp.bp.110.090282">DOI</a></p><p>• Borkovec (2002). Life in the future versus life in the present. <em>Clinical Psychology: Science and Practice</em>. <a href="https://doi.org/10.1093/clipsy.9.1.76">DOI</a></p><p>• LaFreniere &amp; Newman (2020). Exposing worry’s deceit. <em>Behavior Therapy</em>. <a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC7233480/">PMC7233480</a></p><p>• McGowan &amp; Behar (2013). Stimulus control training for worry. <em>Behavior Modification</em>. <a href="https://doi.org/10.1177/0145445512455661">DOI</a></p><p>• Abramowitz, Tolin &amp; Street (2001). Paradoxical effects of thought suppression. <em>Clinical Psychology Review</em>. <a href="https://doi.org/10.1016/S0272-7358%2800%2900057-X">DOI</a></p><p>• Moser, Kross et al. (2017). Third-person self-talk and emotion regulation. <em>Scientific Reports</em>. <a href="https://doi.org/10.1038/s41598-017-04047-3">DOI</a></p><p><strong>Brain chemistry</strong></p><p>• Miller &amp; Raison (2016). The role of inflammation in depression. <em>Nature Reviews Immunology</em>. <a href="https://pubmed.ncbi.nlm.nih.gov/26711676/">PubMed</a></p><p>• Pizzagalli (2014). Depression, stress and anhedonia. <em>Annual Review of Clinical Psychology</em>. <a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC3972338">PMC3972338</a></p><p>• Ottaviani et al. (2016). Physiological concomitants of perseverative cognition. <em>Psychological Bulletin</em>. <a href="https://doi.org/10.1037/bul0000036">DOI</a></p><p>• Scott et al. (2021). Improving sleep quality leads to better mental health. <em>Sleep Medicine Reviews</em>. <a href="https://eprints.whiterose.ac.uk/id/eprint/178923/">Link</a></p><p>• Moncrieff et al. (2022). The serotonin theory of depression: an umbrella review. <em>Molecular Psychiatry</em>. <a href="https://www.nature.com/articles/s41380-022-01661-0">Link</a>; critique by <a href="https://doi.org/10.1038/s41380-023-02095-y">Jauhar et al. (2023)</a> and reply by <a href="https://doi.org/10.1038/s41380-023-02094-z">Moncrieff et al. (2023)</a></p><p><strong>Emotional numbing</strong></p><p>• Bylsma, Morris &amp; Rottenberg (2008). Emotional reactivity in major depressive disorder. <em>Clinical Psychology Review</em>. <a href="https://pubmed.ncbi.nlm.nih.gov/18006196/">PubMed</a></p><p>• Gross (2002). Emotion regulation: affective, cognitive and social consequences. <em>Psychophysiology</em>. <a href="https://www.psychiatry.wisc.edu/courses/Nitschke/seminar/gross_2002_psychophys.pdf">PDF</a></p><p>• Langley et al. (2023). Escitalopram and reinforcement sensitivity. <em>Neuropsychopharmacology</em>. <a href="https://doi.org/10.1038/s41386-022-01523-x">DOI</a></p><p>• Craske et al. (2019). Positive Affect Treatment. <em>Journal of Consulting and Clinical Psychology</em>. <a href="https://doi.org/10.1037/ccp0000396">DOI</a>; replications <a href="https://pmc.ncbi.nlm.nih.gov/articles/10213148">2023</a> and <a href="https://doi.org/10.1001/jamanetworkopen.2026.7403">2026</a></p><p><strong>The changing brain</strong></p><p>• McEwen (2008). Stress hormones in health and disease. <em>European Journal of Pharmacology</em>. <a href="https://www.uned.es/universidad/inicio/dam/jcr:60327b17-ddd7-4902-bf9c-2d6421ee6a70/McEwen%20(2008.EJPhar).Stress%20hormones%20in%20health%20and%20disease.pdf">PDF</a></p><p>• Schmaal et al. (2016). Subcortical brain alterations in major depressive disorder (ENIGMA). <em>Molecular Psychiatry</em>. <a href="https://repub.eur.nl/pub/86136">Link</a></p><p>• Lynch et al. (2024). Frontostriatal salience network expansion in depression. <em>Nature</em>. <a href="https://doi.org/10.1038/s41586-024-07805-2">DOI</a></p><p>• Erickson et al. (2011). Exercise increases the size of the hippocampus. <em>PNAS</em>. <a href="https://pmc.ncbi.nlm.nih.gov/articles/3041121/">PMC3041121</a></p><p>• Zwiky et al. (2025). Limbic gray matter increases after CBT. <em>Translational Psychiatry</em>. <a href="https://doi.org/10.1038/s41398-025-03545-7">DOI</a></p><p><strong>What helps</strong></p><p>• Ekers et al. (2014). Behavioural activation for depression. <em>PLoS ONE</em>. <a href="https://ore.exeter.ac.uk/repository/handle/10871/17506">Link</a></p><p>• Noetel et al. (2024). Exercise for depression: network meta-analysis. <em>BMJ</em>. <a href="https://rodin.uca.es/xmlui/handle/10498/32732">Link</a></p><p>• Bizzozero-Peroni et al. (2024). Daily step count and depression. <em>JAMA Network Open</em>. <a href="https://doi.org/10.1001/jamanetworkopen.2024.51208">DOI</a></p><p>• Burns et al. (2023). Day and night light exposure and psychiatric disorders. <em>Nature Mental Health</em>. <a href="https://doi.org/10.1038/s44220-023-00135-8">DOI</a></p><p>• Balban et al. (2023). Brief structured respiration practices. <em>Cell Reports Medicine</em>. <a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC9873947">PMC9873947</a></p><p>• Lieberman et al. (2007). Putting feelings into words. <em>Psychological Science</em>. <a href="https://pubmed.ncbi.nlm.nih.gov/17576282/">PubMed</a></p><p>• Gariépy, Honkaniemi &amp; Quesnel-Vallée (2016). Social support and depression. <em>British Journal of Psychiatry</em>. <a href="https://pubmed.ncbi.nlm.nih.gov/27445355/">PubMed</a></p><p>• Hunt et al. (2018). No more FOMO. <em>Journal of Social and Clinical Psychology</em>. <a href="https://doi.org/10.1521/jscp.2018.37.10.751">DOI</a></p><p>• Calvert et al. (2025). Social media detox and youth mental health. <em>JAMA Network Open</em>. <a href="https://doi.org/10.1001/jamanetworkopen.2025.45245">DOI</a></p><p>• Wilson et al. (2018). Self-compassion interventions. <em>Mindfulness</em>. <a href="https://link.springer.com/10.1007/s12671-018-1037-6">Link</a></p><p>• Kroenke, Spitzer &amp; Williams (2001). The PHQ-9. <em>Journal of General Internal Medicine</em>. <a href="https://doi.org/10.1046/j.1525-1497.2001.016009606.x">DOI</a></p><p>Spitzer et al. (2006). The GAD-7. <em>Archives of Internal Medicine</em>. <a href="https://doi.org/10.1001/archinte.166.10.1092">DOI</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5a497af2f890" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>