Chapters Transcript Anxiety, Stress, and Insomnia – Clinical Updates 2025 Course: NYU Langone Health's Annual Focus on Men's Health 2025 Good afternoon, everyone. I'm delighted to, as the chair of psychiatry, to bring a uh psychiatry perspective informed by primary care. I started my career. I did not go directly into psychiatry. I uh did advanced training in internal medicine and I was a family doctor for a year before I entered my psychiatry residency program, and um I've always had a deep love for primary care and great pleasure in collaborating with primary care, and a lot of my research is directed at trying to provide tools that are Uh, accessible, scalable, and useful in the context of primary care medicine, not just specialty medicine, and I'll talk about that. These are my disclosures, none of them represent a conflict for my talk today. Um, I'm going to focus mostly on stress and insomnia today, although stress is a great source of anxiety. And my area of greatest interest is in traumatic stress. So I'll talk some mostly about Trauma, stress, and insomnia, and insomnia is the primary disabling symptom of chronic stress-related disorders. The history of PTSD is very interesting. I was on the committees of the APA that developed the diagnostic criteria, but I found the first published reference to to the condition, which was virtually identical to how we diagnose it today. Came from ancient Mesopotamia, uh, where modern day Iraq is, between, between the Tigris and Euphrates rivers, and it was, it was a single case report by the physician to the to the court of Assyria to the king, and it was a case of a combat veteran, and it was beautifully described. Uh, one of the quotes from the, from the case description is, my patient sees a living person or a dead person, that's while he's sleeping or half asleep, waking up, he churns. To, to address it, but, but, but like someone who has rancid oil in his mouth, he sees, so he's unable to cry out to his bad partner. So this is a Kind of awakening out of a traumatic dream and a beautiful, excellent uh description of, of a trauma-related nightmare, which uh leads to some intrusive memories of the battlefield. And interestingly, the the physician attribute. It's the cause of this to the spirit of the enemies that his patient has vanquished on the battlefield who are now coming to get their revenge, a kind of early understanding of the role of moral injury in battle. Well, what is post-traumatic stress disorder? It's defined by 20 symptoms in 4 different categories, reliving the trauma, avoiding the reminders because they're painful, having altered views of the self and the world, such as the world is dangerous and uncontrollable and unpredictable. and persistent psychophysiological arousal of interest, the, the criteria that my committee and others in the APA develop, the 20 symptoms are quite exhaustive, but if you look at the ones in red, they are the ones that are conserved in all. Publications about PTSD starting with the one from 3700 years ago from my Assyrian war fighter, that is, nightmares, dawns in red, dissociative reactions flashing back to the trauma, feeling as though you're back there, trauma-related thoughts and feelings and images. Being triggered by reminders of the trauma, being vigilant and on guard, even in a safe situation, being jumpy and easily startled, and most important, profound sleep disturbance was difficulty falling asleep, staying asleep, and, uh, and remaining asleep. So, uh, there, uh, these are the main clinical phenotypes of PTSD conserved across all cultures and time. And I would say, of the five main symptoms we see in all patients. Painful recollections. Avoiding painful reminders, sleep disturbances. A sense of disillusionment, discouragement, and fear about the future and hyperarousal. If I asked my patients which one, if you had a magic. And you could remove any one of your, your stress-related symptoms. They almost all say, doctor, if you could restore me to healthy sleep, that would be the greatest advantage cause when I sleep well, all of my other symptoms diminish. We now know that there are 4. Separate but interrelated brain circuits that are activated in an abnormal way when we compare people with severe stress reactions and insomnia with people who've been exposed to stressors and um And uh I never developed a uh psychopathology of traumatic stress and insomnia. And the four circuits, I, if anyone would like my slides, because this is detailed, I'm not gonna go into all of the circuit details. I'd be delighted to send you all of my Lies. The First Circuit is one that is known to control emotion regulation, and people with stress disorders are highly emotionally dysregulated. They also have exaggerated focus on threat. They're overfocused on threat, and even in safe environments. They have difficulty updating their environment, meaning, If they've been in a traumatic environment, they it's very difficult. One of my Iraq warfighters that I treated, a highly accomplished person, was a special forces, had a high paying job on Wall Street after he left the military. He came to see me one Saturday morning in my office at NYU and it was a beautiful day. He drove to my office from, from uh Wall Street, where he lived. Uh, and by the time he got to my office, he was completely exhausted. I asked him why was it so difficult? He said, well, you know, Doctor Marmer, all the way up Madison Avenue, at every intersection, I had to check all four corners for explosive devices. So he is still driving on Madison Avenue on a pleasant Saturday morning, as though he's in a convoy in the war zone, failure to update the context. And then, of course, the amygdala and the wall of fear looming. A simplified version of the circuit problems in PTSD are overactivation of the amygdala and the hippocampus, which serves memory formation. And uh contextualization and an underactivity of prefrontal cortical, mostly ventral medial prefrontal cortical areas that regulate emotions. Uh, there are multiple molecular pathways that are dysregulated in PTSD. It's a complex disorder, and we have found evidence in my research and others of uh sympatho adrenergic system disturbances, disturbances of the HPA axis with abnormal cortisol levels. Uh, interestingly, lower rather than higher cortisol and chronic cases, higher in acute cases for complex reasons of receptor, uh, adjustments to stress, prolonged andennergic activation, favors exaggerated memory consolidation, and we find major abnormalities of the renin angiotensin system, leading to inflammatory processes and hypertension and stress. Disorders, very important for primary care, and then multiple disturbances of uh biogenic amines, including uh excitatory and inhibitory amines, endocannabinoids, and dopamine, which are reward circuits. So people with chronic stress, as you know, from their own patients have not only anxiety, but they have depression and lack of pleasure and interest in their life often. Uh, since this is a top for primary care, I thought you would be interested to see some of the Differences we find on routine blood testing that all of us in primary care in your practices would typically do on an annual physical exam, and when we compare patients with chronic stress and insomnia to those who've been exposed to trauma but adjusted well and are resilient, those who've developed chronic stress and insomnia have, Abnormal fasting glucose elevations, abnormal insulin and insulin tolerance, MIR, uh, elevated, uh, bad cholesterol and decreased good cholesterol. They have uh increased BMIs and weight. Uh, faster pulses, and overall, if you see the total metabolic syndrome total score, it's markedly elevated in people with stress and insomnia. So stress. Uh, in practice, David, you were saying it applies to so many people. It must be understood that stress is not just a psychological experience. Clearly, it's a profoundly psychobiological experience and affects many of the same systems that are of concern to primary care doctors in everyday practice, especially cardio metabolic, inflammatory, and immune pathways. Here's another example, look at pro-inflammatory cytokines, and those with chronic stress disorders and insomnia, compared to those who've been trauma exposed and are resilient. Marked differences in IFN. Uh, uh, TNF alpha, IFN, IL-1, IL-6, IL-10, even on a simple test, HS high sensitivity C-reactive protein, which most of us get in routine practice and primary care, marked elevations of C-reactive protein. So stress is highly pro-inflammatory. And bad for your health, bad for your cardiovascular health, bad for your metabolic health. I did a lifetime study of a representative sample of every man and every woman who served in the Vietnam War, and I followed them for 45 years. If any. If you are interested, I can give you the follow-up papers. The 45 year outcomes are noted for major increases in premature mor morbidity and mortality, mostly cardio metabolic, also increased cancer. Why increase cancer? One possible reason is that stress and insomnia causes premature aging or senescence of the immune system. Many different immune factors have been studied. My group has focused on natural killer cell senescence, and we've studied CD16, CD56 expression in natural killer cells. And if you compare The PTSD cases with the trauma exposed healthy controls, with the, you'll see a very substantial age acceleration in the senescence of natural killer cells, meaning that our primary care patients who suffer chronic stress and insomnia are much more vulnerable over time to infectious diseases and cancer, and that's exactly what we found in our lifetime study of Vietnam veterans. Uh, well, what, uh, since I started my, my medical career as a primary care doctor, and since my temper, I'm temperamentally sort of developmentally arrested at that stage, despite being a psychiatrist. I've always longed for the day when the practice of psychiatry would be laboratory-based, as was my practice of primary care. Uh, uh, a quick story, we don't have much time, but, uh, I was a medical intern at the Toronto General Hospital. I'm Canadian originally, and during my medical internship, I had a month of psychiatry, and I just come off of University of Toronto's most advanced cardio metabolic unit, and I was all ready to go with lab tests. And my, my advisor, Doctor Jacques, he said to me, I'm gonna refer you to a, go see a patient and spend, spend some time with them, and tell me if the patient has bipolar depression or unipolar. The patient's depressed, but tell me which one it is because the prognosis is very different. We know much more about that now, by the way. And the prognosis of bipolar depression is much more dangerous over time than unipolar. That's for another lecture, Steve and David, we can talk about that. But, uh, I, I spent 1520 minutes with the patient. I drew blood, I drew a blue tube and a red tube. I sent it to the lab, and I asked the lab to tell me if it was bipolar or unipolar, which kind of got me thrown off the unit. Um, but here's the remarkable thing, David. I still can't send the blood to the lab to determine if my patient has bipolar or unipolar disease, and it's been, you know, at least a century since I was a a a primary care doctor. So I'm, my group, I'm, I started a 7 university and institute consortium to develop a blood test for PTSD that we could use in primary care. Uh, in the same way that we screen for prostate cancer with PSA. And uh just uh I'll say what we, where we got, this is an earlier version of it. We have a more advanced version, and just very big. We started with 1 million features in blood, so we sequenced DNA and RNA and metabolites and proteins, started with 1 million individual features on all of our patients. We have 50 pre pre-specified. Panels using something called the wisdom of the crowd. We use advanced AI to uh reduce the number of candidates that would be associated with being a case or control. And in the end, I'm not going to go into all the details, but you can see the features that we studied carefully, uh, physiological measures, heart rate, metabolites. Uh, gene expression, mic non-coding RNAs, microRNAs, proteins, the proteins are perhaps the most interesting, uh, and Just these were Iraq and Afghanistan veterans, 100 with PTSD and 100 without without PTSD trauma exposed. 83 had complete features with all the blood features. I won't go into all the differences, we use a validation sample, but we developed a model using AI using recursive feature elimination, and we identified. A limited set of blood markers, proteins, and others, and I wanted to show you that in this panel, these are the most important features, those to the right of the vertical line were most important in the model, that is to say, if you eliminate them from the model, your ability to accurately classify PTSD in blood degrades quickly. And we ended up with 77 features, and then with further analysis, uh, uh, a total of about 22 features, uh, you can see what they are in the bond. If this is all published, anyone's interested, I will gladly share the paper with you, but it's a set of, uh, transcripts and proteins and metabolites that are And here's what's very interesting. Here is the accuracy using AEC curves, and we were with those features, we had 80% or 81% overall accuracy. That is to say, a simple panel of features from peripheral blood that any primary care doctor could draw at a at an annual physical, classified PTSD in blood with an accuracy of 80. We Now have a set of 22 proteins, which we've evolved from this. I didn't present them to you. Their accuracy is greater. And I didn't present them to you because they're, the FDA is reviewing them, and it's asking for clarifications. So we are under review to get an FDA approval for what will be the first blood test for a major psychiatric disorder. The only test we have for psychiatric. Blood tests we have now that are FDA approved are for Alzheimer's disease, which is neurological, and some for TBI that are promising. But for anxiety, depression, PTSD, bipolar disorder, schizophrenia, panic disorder, OCD, we have zero FDA approved blood tests or any other laboratory tests that could be used in high throughput screening in primary care. Um. So, we're also on the hunt for other even simpler features. Uh, I'm very interested in a human voice as a feature. We can quickly record voice. Voice is ubiquitous. We have it get it from cell phones, even from doing telemedicine visits with our patients. Without going into all the details, uh, and the background, voice was developed to communicate human emotions and attachments, especially, by the way, It was, it was adjudicated to The Varon signals of safety and danger. So we, we've been studying this, we developed a, a platform for this. Excuse me, I'll just. Working from home today. Um, uh, we developed a, a platform for this that, uh, that is, uh, using 40,000 unique biophysical elements of human speech. None of these are words, these are all the, the biophysics of speech. This platform was developed. By my colleagues at SRI who I got a call from one day and said they'd like to work on a voice test for PTSD. I asked them their background. They said, well, we've got some good engineers. We developed Siri for Apple. So I thought, I'm all in. We did it, and we did. Study, we developed the pipeline, 40,000 features. We use AI to to identify a very limited set of features, and we are able to classify who is and who is not a PTSD case simply from the human voice spectrum with nearly 90% accuracy. So we're also seeking FDA approval for this test to be used in primary care. Um, I'm gonna move quickly because this because we have limited time. If I, uh, this is all published, and I'm happy to share the publications on our blood tests and voice tests for PTSD. The only thing I want to say, David, and Steve, is I'm entirely motivated to do this, because first, And foremost, my love of primary care, secondarily, specialty care and psychiatry. I want low cost, accessible, non-invasive tests for psychiatric disorders that can be used in primary care at annual screenings. And uh my Center for Precision Psychiatry is looking for those in multiple disorders, starting with PTSD and uh moving out from there to depression and other disorders. Now, treatments of PTSD, I'm not gonna go into great detail on the treatment of PTSD in our limited time, but, uh, but we use behavior therapy, we use psychotherapies of various kinds. There are two FDA approved medications for PTSD, uh, Paxil and Zoloft. Uh, these are the, some of the, uh, FDA registration trials for Paxil. We got a very good signal. Uh, for Zoloft, I, I was one of the, um, Uh, principal investigators on the FDA registration trial for Zoloft. I was the one who testified at the FDA to get the indication of Zoloft for PTSD. It's very successful. It's well tolerated. It's very widely used. This shows also that if you treat a patient with an SSRI for stress, anxiety, and depression, you should keep going. Uh, because, uh, I don't know if you can see my cursor, but the patients get a partial response at 8 to 12 weeks, but their response continues to improve at least till 36 weeks. My advice is in primary care practice, if you're treating a patient for stress, Anxiety, depression, and insomnia, and they're having a good response to an SSRI or related anti-anxiety antidepressant medications. They, and they've been chronically symptomatic. They should be maintained on treatment at least 36 months, usually 1 year, and then tapered. Uh, venlafaxine has also been successful. Uh, I would say that Lexapro is also well tolerated and successful for stress and insomnia. So, let me, let me conclude with, uh, Very briefly, with some uh medic uh treatments for insomnia in the context of stress and anxiety. Uh, the first thing to say is we usually recommend CBTI to begin with. It's a very simple program on sleep hygiene. There are lots of good online programs the patient can use. Uh, and then we, we like very much low dose trazodone. Trazodone, as you know, is an antidepressant, not very good. Uh, but in low doses, 50 to 150 mg, it's very good for sleep and very compatible with other drugs such as the SSRIs. Nightmares are the single most disturbing symptom for, for patients with stress disorder. Uh, a lot of medications have been tried, trazodone, cyproheptadine, uh, nefazodone has a black box warning. We don't use that anymore, uh, adrenergic agents, prazosin, uh, I would say for the treatment of persistent nightmares for patients in primary care practice, the best drug is actually the old fashioned anti-hypertensive prazosin. Uh, uh, and it, it is, uh, as you know, uh, was used years ago before the current generation antidepressant drugs, uh, uh, alpha-1 and it's an alpha one adrenergic post receptor antagonist. We start at 1 mg at bedtime. We cautioned the patients that they may develop a hypotensive reaction in the first few days they take it. So we asked them to take it at bedtime, sit in a chair in bed, get used to it for a few. days and then it's well tolerated. And then we build it up gradually up to 5 to 15 mg, and it's exceptionally good uh for, for recurrent nightmares and helps with other symptoms of PTSD as well. You must caution the patient about orthostatic hypotension, and you must follow them up when you start the program and get their blood pressure on each visit, which obviously, now in primary care is done uh routinely. Uh, only increased by 1 mg every 3 to 4 days as tolerated. Uh, Mirtazapine, which is, you know, is widely used, uh, uh, antidepressant with good qualities for sleep, mirtazapine is both a neurorenergic and seronergic antidepressant mix, also has histamine one receptor actions and serotonin. And 2A and CT actions, and we like it very much. We start at 7.5 mg. We usually don't go above 15 mg, cause it's most sedative, up to 15 mg, and between 15 and 30, it becomes more activating. So horirtazapine is excellent. Uh, orexin antagonists like Suvaexin are very good. They'reexin-1 receptor antagonists. They're very good sleep. I can send you papers on that. And uh Rmeltion is very helpful for insomnia. It's a melatonin T1 and melatonin 1 and 2 receptor agonist, and it's a kind of a pharmaceutical high grade uh melatonin agent. We use 8 mg at bedtime, and uh psychotherapy for insomnia is important, CBT and other approaches. So, uh, I think at that point, uh, there's, uh, without going into all the complexities of PTSD, I think I'll stop at that point. Published October 18, 2026 Created by
CME Knowledge Hub